Veterinaria Italiana, Volume 49 (1), January-March 2013 by Veterinaria Italiana

ISSN 0505-401X

RIVISTA DI SANITÀ PUBBLICA VETERINARIA

VOLUME 49 (1) - GENNAIO-MARZO/JANUARY-MARCH 2013

Rivista trimestrale di Sanità Pubblica Veterinaria, edita dall’Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise “G. Caporale” A quarterly journal devoted to veterinary public health, veterinary science and medicine, published by the Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise ‘G. Caporale’ in Teramo, Italy

Volume 49 (1), 2013

Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise “G. Caporale” Campo Boario, 64100 TERAMO, Italia telefono +39 0861 3321 fax +39 0861 332251 www.izs.it

Questa rivista è nata nel 1950 con il nome di Croce Azzurra. Dal 1954 si chiamerà Veterinaria Italiana.

Direttore Editor-in-Chief Giovanni Savini

Comitato direttivo Managing Scientific Board Romano Marabelli Fernando Arnolfo

Membri onorari Honorary Members Louis Blajan - France James H. Steele - United States of America

Comitato di redazione Editorial Board Hassan Abdel Aziz Aidaros – Egypt Ayayi Justin Akakpo – Senegal Nicola T. Belev – Bulgaria Stuart C. MacDiarmid – New Zealand J. Gardner Murray – Australia Yoshihiro Ozawa – Japan Alexander N. Panin – Russia

Victor E. Saraiva – Brazil Aristarhos M. Seimenis – Greece Arnon Shimshony – Israel Samba Sidibé – Mali Gavin R. Thomson – South Africa Carlo Turilli – Italy Norman G. Willis – Canada

Comitato scientifico Scientific Advisory Board L. Garry Adams – United States of America Menachem Banai – Israel Elie K. Barbour – Lebanon A.C. David Bayvel – New Zealand Giorgio Battelli – Italy Roy G. Bengis – South Africa Ingrid E. Bergmann – Argentina Peter F. Billingsley – United States of America Silvio Borrello – Italy Canio Buonavoglia – Italy Mike Brown – United Kingdom Gideon Brücknerr – South Africa Giovanni Cattoli – Italy Bernadette Connolly – United Kingdom Julio De Freitas – Brazil Piergiuseppe Facelli – Italy Gianluca Fiore – Italy Cesidio Flammini – Italy Riccardo Forletta – Italy Bruno Garin-Bastuji – France Giorgio Giorgetti – Italy Rob Gregory – New Zealand Anwar Hassan – Malaysia

Barry J. Hill – United Kingdom Katsuyuki Kadoi – Japan Bruce Kaplan – United States of America R. Paul Kitching – Canada Corinne I. Lasmézas – France N. James MacLachlan – United States of America Salvatore Magazzù – Italy Franco Mutinelli – Italy Klaus Nielsen – Canada Lisa Oakley – New Zealand Massimo Palmarini – United Kingdom Attilio Pini – Italy Santino Prosperi – Italy Franco M. Ruggeri – Italy Domenico Rutili – Italy Paul Sutmoller – The Netherlands Peter M. Thornber – Australia Silvio Arruda Vasconcellos – Brazil Patrick Wall – Ireland Alexander I. Wandeler – Canada Kazuya Yamanouchi – Japan Cristóbal Zepeda – United States of America Stéphan Zientara – France

Segreteria di redazione Associate Editors Monica Bucciarelli, Guido Mosca, Mariarosaria Taddeo, Carlo Turilli Recensioni Book reviews Manuel Graziani Progetto grafico e web Graphic and web design Paola Di Giuseppe Amministrazione Administration Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise “G. Caporale” Campo Boario, 64100 Teramo, Italia veterinariaitaliana@izs.it Stampa Printer Giservice srl, Teramo, Italia http://www.izs.it/vet_italiana/index.html © 2013 Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise “G. Caporale” Campo Boario, 64100 Teramo, Italia

ISSN 0505-401X Formato elettronico Electronic format ISSN 1828-1427 Stampato su carta ecologica TCF Printed on 50% recycled, 100% chlorine- and acid-free environmentally friendly paper Aut. Trib. Teramo n. 299 del 16/05/1990 Sped. in Abb. Post. Art. 2 comma 20/c L. 66/96 DCB/DC Abruzzo Pescara

Volume 49 (1), 2013

Thomson Reuters Science Journal Citation Reports® database (JCR/Science Edition®) Journal impact factor 2011: 0.667 • National Library of Medicine’s MEDLINE/ PubMed system • Thomson Reuters Science Citation Index Expanded™ (SciSearch®) • CABI’s Full-Text Repository • Directory of Open Access Journals (DOAJ) • Elsevier’s SciVerse Scopus

Le pubblicazioni dell’Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise “G. Caporale” (IZSAM) sono protette dalla legge internazionale sul copyright. Gli estratti possono essere letti, scaricati, copiati, distribuiti, stampati, recuperati; è consentito inoltre il collegamento ai file pdf di Veterinaria Italiana. Informazioni per fini commerciali devono essere richieste all’IZSAM. Le traduzioni a stampa e gli adattamenti sono consentiti previa autorizzazione scritta da parte dell’IZSAM. Le opinioni espresse negli articoli pubblicati sono esclusivamente sotto la responsabilità degli autori. L’eventuale citazione di specifiche Ditte o prodotti, siano essi brevettati o meno, non implica che essi siano stati consigliati dall’IZSAM e vengano preferiti ad altri di simile natura non menzionati nei testi. Si ringrazia Fabrizio Piccari per la gentile concessione dell’immagine di copertina. Publications of the Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise ‘G. Caporale’ (IZSAM) are protected by international copyright law. Users are permitted to read, download, copy, distribute, print, search abstracts; besides they can link to Veterinaria Italiana full pdf files. Should information be required for commercial purposes, prior written permission must be sought from the IZSAM. Published translations and adaptations also require prior written approval from the IZSAM. The views expressed in signed articles are solely the responsibility of the authors. The mention of specific companies or products of manufacturers, whether or not patented, does not imply that these have been endorsed or recommended by the IZSAM in preference to others of a similar nature that are not mentioned. Acknowledgement is made to Fabrizio Piccari for the cover image.

EDITORIAL ‘If we want things to stay as they are, things will have to change’.............. 5 Nazmun Nahar, Main Uddin, Rouha Anamika Sarkar, Emily S. Gurley, M. Salah Uddin Khan, M. Jahangir Hossain, Rebeca Sultana & Stephen P. Luby Exploring pig raising in Bangladesh: implications for public health interventions................................................ 7

Indagine sulle problematiche relative all’allevamento di maiali in Bangladesh: implicazioni per possibili interventi di sanità pubblica (riassunto)...............................................7

Darem Tabbaa & Aristarco Seimenis Population displacements as a risk factor for the emergence of epidemics...................................................................19

Trasferimenti di popolazioni e fattori di rischio per le emergenze epidemiche (riassunto)......................................................................................................19

Giulia Morosetti, Marica Toson & Christian Piffer Indagine su lesioni provocate da animali in Provincia Autonoma di Bolzano nell’anno 2010 ....................................25 Lesions caused by animals in the Autonomous Province of South Tyrol in 2010: Fact-finding for prevention.................................................................................. 37 Giovanna Fusco, Giuseppe Aprea, Giorgio Galiero, Achille Guarino & Maurizio Viscardi Escherichia coli, Salmonella spp., Virus dell’Epatite A e Norovirus in Molluschi Bivalvi nel 2011-12 in Sud Italia..........................51 Escherichia coli, Salmonella spp., Hepatitis A Virus and Norovirus in bivalve molluscs in Southern Italy................................................................................. 55 Salvatora Angela Angioni, Giampiero Scortichini, Gianfranco Diletti, Fabrizia Perletta, Roberta Ceci & Nicola Ferri Studio degli effetti tossici del ritardante di fiamma PBDE-47 su vongola Chamelea gallina (Linnaeus, 1758)...........................................59 Study of the toxic effects of flame retardant PBDE-47 on the clam Chamelea gallina (Linnaeus, 1758) ........................................................... 69 Rossella Lelli, Umberto Molini, Gaetano Federico Ronchi, Emanuela Rossi, Paola Franchi, Simonetta Ulisse, Gisella Armillotta, Sara Capista, Siegfried Khaiseb, Mauro Di Ventura & Attilio Pini Vaccino inattivato e adiuvato per il controllo delle infezioni da sierotipo 9 del virus della peste equina: valutazione dell’efficacia in cavallo e cavia......................................................................79 Inactivated and adjuvanted vaccine for the control of the African horse sickness virus serotype 9 infection: evaluation of efficacy in horses and guinea-pig model................................................ 89

Volume 49 (1), 2013 Edyniesky Ferrer, Paolo Calistri, Osvaldo Fonseca, Carla Ippoliti, Pastor Alfonso, Simona Iannetti, María A. Abeledo, Octavio Fernández, María I. Percedo & Antonio Pérez Estimation of the sensitivity of the surveillance system for avian influenza in the western region of Cuba......................................99

Validità del sistema di controllo dell’influenza aviaria nella regione occidentale dell’isola di Cuba (riassunto).................................................................99

Shahrzad Azizi, Farzad Shahrani Korani & Ahmad Oryan Pneumonia in slaughtered sheep in south-western Iran: pathological characteristics and aerobic bacterial aetiology.................109

Casi di polmonite in pecore macellate nel sud ovest dell’Iran: caratteristiche patologiche ed eziologia batterica aerobica (riassunto)................................. 109

Antonietta Di Francesco, Raffaella Baldelli, Manuela Donati, Claudia Cotti, Patrizia Bassi & Mauro Delogu Evidence for Chlamydiaceae and Parachlamydiaceae in a wild boar (Sus scrofa) population in Italy...........................................119

Chlamydiaceae e Parachlamydiaceae riscontrate in una popolazione di cinghiali selvatici in Italia (riassunto)..................................................... 119

COMUNICAZIONE BREVE/SHORT COMMUNICATION Mara Battilani, Andrea Balboni, Massimo Giunti & Santino Prosperi Coinfezione da parvovirus felino e canino in un gatto.............................123 Co-infection with feline and canine parvovirus in a cat............................................. 127 LIBRI/Book reviews Salvatore Montinaro Sanità animale............................................................................131 Stefano Izzi Sicurezza della filiera della pesca.......................................................133

EDITORIAL ‘If we want things to stay as they are, things will have to change’ “Se vogliamo che tutto rimanga come è, bisogna che tutto cambi” (Il gattopardo, Giuseppe Tomasi di Lampedusa)

It is now few months that I have been appointed as new editor in chief of the Veterinaria Italiana and first of all I would like to express my sincerest gratitude to Ferdinando Arnolfo, the Director of the Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise ‘G. Caporale’, and his staff for the honour they have bestowed on me. It is for me a very new and exciting challenge. I firmly believe the journal has the potential to turn into a successful and important voice in the scientific veterinary community. As you might be aware, Veterinaria Italiana is an indexed journal that recently received the impact factor, thus acquiring international recognition alongside major scientific journals. In view of this, it is only right and proper to acknowledge the previous team for their collective work for bringing the journal to its the current position. First of all, I would like to thank the previous director, Vincenzo Caporale, who has always strongly believed in Veterinaria Italiana, putting in a great deal of effort and resources when others doubted the success of this initiative. Special thanks are also due to the previous associate editor, Gill Dilmitis, who has been one of the pillars of the journal taking care of all author needs always with extreme enthusiasm and competency. I am afraid that Mariarosaria Taddeo, the new associate editor, and I will have the daunting task of meeting their high standards and level of competence; even more so because we realise that, as in numerous other activities, it is easy to start but it is always harder to maintain the status quo. Nonetheless, if on one hand we know that it is a difficult task, on the other hand we are sure that we and the other team members will do our best to improve the current position of the journal. We have already created a new and modern format for Veterinaria Italiana so that it may compete with other international and better known scientific journals. Soon Veterinaria Italiana will be provided with an electronic platform and the DOI (Digital Object Identifier). Both, I am sure will be extremely beneficial to editors, referees, and authors alike. Many other innovations will be introduced to improve the readability and quality. All these changes, however, will not distract from our main purpose which is to provide information about new research results on veterinary and public health field, ensuring their reliability and integrity through relevant, strong and comprehensive proofs.

EDITORIAL Veterinaria Italiana will still represent the voice of authors who do not have the opportunity of disseminating their results widely, of course, being the Italian journal in Veterinary field with the highest impact factor, it is privileged to have the opportunity of publishing the results of the best Italian research teams on public and animal health. For myself, I know that a journal is a powerful educational tool. I will try to ensure independence of editorial work and good editorial practice as defined by international editorial organizations, bearing in mind that the work invested in a journal is never wasted as it stimulates and provokes research and culture of scientific communication. Giovanni Savini Editor-in-Chief Direttore

Exploring pig raising in Bangladesh: implications for public health interventions Nazmun Nahar1, Main Uddin1, Rouha Anamika Sarkar1, Emily S. Gurley1, M. Salah Uddin Khan1, M. Jahangir Hossain1, Rebeca Sultana1 & Stephen P. Luby1, 2 International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR,B), GPO Box 128, Dhaka 1000, Bangladesh nahar.nazmun@yahoo.com 2 Centers for Disease Control and Prevention (CDC), 1600 Clifton Rd, Atlanta, GA 30333, United States of America 1

Keywords Bangladesh, Discrimination, Pig-human interactions, Pig raisers, Public health, Social stigma, Zoonoses.

Summary Pigs are intermediate hosts and potential reservoirs of a number of pathogens that can infect humans. The objectives of this manuscript are to understand pig raising patterns in Bangladesh, interactions between pigs and humans, social stigma and discrimination that pig raisers experience and to explore the implications of these findings for public health interventions. The study team conducted an exploratory qualitative study by interviewing backyard pig raisers and nomadic herders (n = 34), observing daily interactions between pigs and humans (n = 18) and drawing seasonal diagrams (n = 6) with herders to understand the reasons for movement of nomadic herds. Pig raisers had regular close interaction with pigs. They often touched, caressed and fed their pigs which exposed them to pigs’ saliva and feces. Herders took their pigs close to human settlements for scavenging. Other domestic animals and poultry shared food and sleeping and scavenging places with pigs. Since pigs are taboo in Islam, a majority of Muslims rejected pig raising and stigmatized pig raisers. This study identified several potential ways for pigs to transmit infectious agents to humans in Bangladesh. Poverty and stigmatization of pig raisers make it difficult to implement health interventions to reduce the risk of such transmissions. Interventions that offer social support to reduce stigma and highlight economic benefits of disease control might interest of pig raisers in accepting interventions targeting pig borne zoonoses.

Indagine sulle problematiche relative all’allevamento di maiali in Bangladesh: implicazioni per possibili interventi di sanità pubblica Parole chiave Allevatore di suini, Bangladesh, Discriminazione sociale, Interazione maiale-uomo, Sanità pubblica, Suini, Zoonosi.

Riassunto I maiali sono ospiti intermedi e potenziali portatori di agenti patogeni infettivi per l’uomo. Questo articolo identifica le conseguenze delle metodiche impiegate nell’allevamento di maiali in Bangladesh, le interazioni tra maiale e uomo, la discriminazione sociale a cui sono sottoposti gli allevatori di maiali, analizzando le possibili implicazioni su eventuali interventi di sanità pubblica. Lo studio qualitativo è stato condotto intervistando allevatori di maiali stanziali e nomadi e osservando le interazioni giornaliere tra uomo e maiale. Lo studio riporta, inoltre, i diagrammi effettuati in collaborazione con gli allevatori per individuare le ragioni del nomadismo stagionale. Pur evidenziando regolari contatti tra esseri umani e animali, lo studio ha permesso di accertare come in Bangladesh alcune abitudini favoriscano la trasmissione di agenti infettivi dal maiale all’uomo. Allo stesso tempo, si è potuto rilevare come la povertà e i dettami della religione islamica contribuiscano ad ostacolare l’attuazione di interventi di sanità pubblica atti a prevenire il rischio di infezione. Programmi di supporto, che sottolineino agli allevatori i benefici economici del controllo delle malattie, potrebbero essere d’aiuto per rendere accettabili interventi pubblici mirati al controllo delle zoonosi suine. Veterinaria Italiana 2013, 49 (1), 7-17

Nahar et al.

Pig raising in Bangladesh and public health interventions

Introduction Pigs are the intermediate hosts and potential reservoirs of a number of viruses, bacteria, and parasites that can also infect humans (9, 13, 17, 21, 35, 44, 48, 49). These types of interspecies transmissions of diseases are defined as zoonotic diseases. The first recognized human Nipah virus outbreak, which caused 105 deaths in Malaysia, resulted from human contact with sick pigs (10, 15, 34). Pigs are an important amplifying host for Japanese encephalitis virus, which is endemic in South and Southeast Asia (43, 49). In 2009, a novel strain of influenza A, H1N1, which included genetic material from a swine influenza virus, became a global pandemic (36). Several human pathogens are present in Bangladesh, which can infect both pigs and people, though there is limited evidence of the zoonotic transmission of these pathogens (7, 18, 26, 27). However, studies from the neighboring country India suggested several disease transmissions between pigs and humans. Epidemiological investigations among pig farmers found that 18.5% of them were affected by Taenia solium taeniasis and 15% of them were affected by Neurocysticercosis caused by helminth parasite Taenia solium, which can be transmitted between humans and between humans and pigs (39, 40). Another study from India identified 3% of slaughtered pigs as having Neurocysticercosis in their brains (38). Investigation of multiple outbreaks of trichinellosis, with high mortality rates, from 2008 to 2011 in India found that all the cases consumed pork before getting the illness (42). A genomic analysis of a human group A rotavirus G6P (6) strain reported pig-to-human transmission of this virus in eastern India (31). A study examining serum samples from pigs and plasma samples from pig handlers and the general population from urban and rural areas found presence of hepatitis E virus in some pigs. A very high number of pig handlers were positive for antibodies against hepatitis E indicating the possibility of zoonotic transmission (50). A temporal relationship of Japanese encephalitis virus transmission has been identified in pigs, mosquitoes and humans that was significantly associated with Japanese encephalitis in humans in India (6). India borders Bangladeshi land from all three sides and there are many cultural similarities between the countries. Similar to India, in Bangladesh, impoverished communities raise pigs. They consume pork and live in close proximity with their domestic pigs in a poor condition, which might put them at risk of zoonoses as identified in India (4, 40). Pigs as a zoonotic reservoir have not been explored in detail in Bangladesh, partially because pig raising is commonly ignored in this predominantly Muslim country. In Bangladesh, published data are

available on the number of cattle, goats, chickens and ducks (2) but not on the number of pigs and pig raisers. After the publicity regarding swine flu in 2009, the Bangladesh Department of Fisheries and Livestock conducted a census to estimate the number of domestic pigs but until July 2012 there is no published data available. Since pigs are taboo in Islam, raising pigs is often associated with social stigma in Bangladesh (4). Social stigma and discrimination have public health implications (45). Stigma can affect peoples’ lives socially, economically physically and psychologically (45, 47). Stigma affects the quality of life through isolating a community, limiting their social and health opportunities and making them reluctant to seek help as they are concerned about discrimination and rejection (51). In this manuscript, we describe pig raising practices, interactions between pigs and humans, social stigma and discrimination that pig raisers experience and the implications of these findings for public health interventions.

Material and methods Materials and methods of this study have been described elsewhere (32). The following is a brief description of the study design and methods.

Study design This is an exploratory qualitative study. Such studies can contribute substantially to issues that have received little prior investigation (37). Exploratory qualitative studies often generate hypotheses to frame further investigations (28). In addition, qualitative investigation prioritizes the views of the population that is being studied (24) which is helpful when there is an interest in developing interventions.

Study population We included both backyard pig raisers and mobile pig herders as our informants, since these are the two main ways of raising pigs in Bangladesh. ‘Sweepers’ and indigenous communities in Bangladesh raise pigs in their backyards. We selected two ‘sweepers’ communities from two districts and a matrilineal Mandi indigenous community from central Bangladesh who raised backyard pigs. We selected mobile pig herds from the central northern part of Bangladesh.

Study sites We purposively selected three backyard pig raising communities from three districts. These were Faridpur, Chapainobabgonj and Tangail District (Figure 1). As pig herders were mobile, it was difficult for the data collection team of five qualitative researchers

Veterinaria Italiana 2013, 49 (1), 7-17

Nahar et al.

Pig raising in Bangladesh and public health interventions

Data collection The team collected data from August 2007 until September 2008 (Table I). They conducted in-depth interviews with backyard pig raisers and herders in Bengali. They recorded the interviews by audio recorders. They also performed observations and took notes on the interactions of pigs with humans and other animals. They worked with herders from each herd to make a diagram of the seasonal movements from one grazing location to another.

Data analysis The team transcribed the recorded interviews verbatim. The three authors (N.N., M.U. and R.A.S.) reviewed all the interviews and observation notes to make a coding list based on the study objectives and the emerging themes from the data. They manually coded the interviews, grouped related codes and compared the findings from all research tools. Figure 1. Field sites of the pig raising study in Bangladesh, 2007-2008. including three of the authors (N.N., M.U. and R.S.) to identify the herders. The team visited the weekly pig market in Mymensingh District and developed a rapport with a pig herder. Later, the team visited his herd and interviewed him and asked him to take them to another pig herd that he knew. He took them to the next pig herd and they interviewed the new herders and asked the herders to take the team to another herd. In this way, they selected six pig herds from six districts in Bangladesh: Mymensingh, Tangail, Sherpur, Sirajgonj, Bogra and Pabna Districts.

Rapport building A crucial element of this study was to build rapport with the pig raisers. Although pig raisers never refused to participate, initially they were often unwilling to provide any time for interviews unless the team went to them with an introduction from someone that they knew and could trust. Later, they explained that since the greater Muslim community did not accept pig raising, they were skeptical about the motivations for the study.

Sampling The team selected people as informants who directly cared for pigs. Both for backyards and herds, they continued interviewing pig raisers until receiving similar information repeatedly. In this process, they ended up including 34 informants, 17 each from backyards and herds.

Veterinaria Italiana 2013, 49 (1), 7-17

Ethical consideration The Ethical Review Committee of icddr,b reviewed and approved this study. Since it was difficult to gain trust of pig raisers for written consent, the team obtained informed consent from all the pig raisers before conducting interviews and observations.

Results Demographics Backyard pig raisers raised pigs mostly in their backyards with poultry and cattle. These pigs were semi-scavengers, which means they were partially fed by their owners and they also searched for their own food. Both males and females provided daily care to the pigs. Mobile herders moved pigs from one district to another. Their pigs were free scavengers. Herding pigs was a male occupation (Table II). Backyard pig raisers and herders used hand pumps to collect drinking water. When herders were far away from villages, they used water from a ditch, pond, river or irrigation canal to drink, cook and clean utensils. Backyard pig raisers used non-sanitary latrines. These latrines had open canals that carried Table I. Research tools used to collect data in the pig raising study in Bangladesh, 2007-2008. Research tools Indepth interview Observation Seasonal diagram

Number Backyard Herd 17 17 12 6 0 6

Duration 40-90 mins 6-8 h 20-30 mins

Nahar et al.

Pig raising in Bangladesh and public health interventions

Table II. Demographic characteristics of the pig raisers of backyards and herds, Bangladesh, 2007-2008. Demographic characteristics Sex Male Female Age Average age in years [range] Education Average years of attendance [range] Religion Hindu ‘lower caste’ Christian Occupation Cleaner (sweepers ) Agricultural work Pig raising in nomadic herds Average number of pigs [range] Faridpur sweeper colony Tangail Chapainobabgonj In the herds

Backyard Frequency (%)

Herd Frequency (%)

10 7

(59) (41)

17 0

(100) (0)

[18-60]

[14-50]

[0-12]

[0-5]

12 5

(71) (29)

17 0

(100)

11 6 0

(65) (35) (0)

0 0 17

(0) (0) (100)

3 2 52

[2-9] [1-3] [34-92]

Pig raisers from Tangail District owned small homesteads in a village. They kept pigs tied with a rope in the west corner of the homestead close to the house to avoid the smell of the pig feces coming into the house, since they believed that air does not flow from the west. They did not let their pigs roam freely.

Mobile herds 104 [40-195]

feces to a pond or river or to drain latrine water outside. Their small children defecated in the open space. Mobile herders practiced open defecation.

Settings: close proximity of pigs and humans Backyard pig raising The backyard pig raising communities differed in terms of home settings, pig keeping places and pig raising patterns which influenced pig caring and feeding practices. The Faridpur sweeper colony was a congested area in the city, mostly with one-room households built on government-owned small pieces of land where almost every square inch of land was used. On one side, there was the entrance to the community and the rest was surrounded by a pond. Because of the scarcity of land, pig raisers kept pigs together in five different places, adjacent to their houses and in pigpens along the edge of the pond. Two households kept one pig each in front of their houses tied with a rope. The team noticed that pigs often freely roamed around inside the colony compound. The pig raising community from Chapainobabgonj District consisted of three households located in the area around the local market in a peri-urban area. They were related by marriage. One household

owned a house with a pigpen in the backyard near the market. The other two households were landless and their houses were built over a pond (on a wooden platform) in the middle of the market. During the daytime, all the raisers took pigs out to the surrounding areas in their locality to feed them. The household with the pigpen kept pigs in the pigpen at night and sometimes the other two households also kept their pigs in this pigpen. The two households who did not have a pigpen reported that they kept their pigs in the nearest mango gardens (minimum one kilometer away) at night during July to February. In March, the beginning of mango season, they moved their pigs close to an abandoned swamp in the village. They also tied their pigs in front of their huts when pigs became sick.

Herders moved the pigs from one district to another based on the seasonal availability of food and water for pigs. They stayed in a place for a few days until the food was finished. During the dry season (November to April), they moved their pigs to lowlands close to the swamp or river. In that period, the water level went down and tubers and roots became available for the pigs. Sometimes, because of minimal rainfall, the water dried up and the herders moved their pigs close to the villages to get water from irrigation systems. During the rainy season (May to October), they moved the herd to highland areas, i.e. villages and gardens where water did not stand. Herders reported that caring for 30 to 90 adult pigs required three to seven people. Usually they were family members or close relatives. When each of them had only a few pigs, sometimes they kept all their pigs together and raised them in a herd. When the number of pigs increased, they employed people from their own caste to take care of the pigs. In mobile herds, pigs scavenged their own food from dawn to dusk. Overnight, herders made a tent using a polythene sheet, in an open field close to a body of water keeping a distance from village homesteads. They kept their pigs close to the tent. One herder described a usual day: “I wake up before sunrise. I take the pigs out for feeding. Others prepare food. Then we eat. For the whole day, we move with pigs, keep our eyes on them; we follow them everywhere, making sure that they don’t destroy crops. In the evening, we stop in a place to spend the night. One cooks, another brings water from the village and the last one looks after the pigs. We prepare food

Veterinaria Italiana 2013, 49 (1), 7-17

Nahar et al.

and eat. Then two of us go to sleep and one stays awake to keep an eye on the pigs. Yesterday I started by 8:30 am and went to sleep by 3 am.” The herders reported that once in every two to four months, each of them took a few days off and visited their families. However, they reported that they never took the herd to the place where their families were living because they were poor. They generally had neither enough space to keep pigs at home nor enough money to buy food for pigs. In addition, if they brought pigs to their homes, local people might protest and look down on them, since they sometimes did not disclose to their Muslim neighbors that they raised pigs.

Feeding pigs The informants from Faridpur mainly fed pigs with leftover human food that they collected from restaurants. They either cleaned the restaurant or paid a minimal fee in exchange of leftovers (US$ 0.07 for a bucket). In Tangail, they provided husks and rice that are the byproducts of homemade rice liquor. They spent about US$ 0.28 per day purchasing husks. In Chapainobabgonj, pigs scavenged in the surrounding areas. Herders led the pigs to scavenge in abandoned land, empty grain fields, orchards or villages where pigs looked for roots, leaves and seeds of many varieties of plants and grasses, worms, insects, and rotten fruits and vegetables. They also scavenged human feces, garbage dumping places and occasionally carcasses of cattle and poultry. On a few occasions, the study team observed that people asked pig raisers to bring the pig herd to clean their open toilet by eating feces or clean a place by eating rotten potatoes.

Interactions of pigs with humans, birds and animals In backyards and herds, we observed direct and indirect physical interactions of pigs with humans, other domestic animals and birds (Table III).

Exposure to pig feces Pig-keeping places both in backyards and herds were muddy with pigs’ urine and feces, which the field team never observed to be cleaned. Pigs’ bodies were smeared with this mud. Sometimes pigs splattered mud on pig raisers’ clothes while they were caring for or serving food to the pigs (Table III). In backyards, we observed dry and fresh pig feces in the surroundings of pig-raising communities. When it rained during an observation, the rainwater was trapped in the yard with pig feces and children played, glided and jumped in that water and mud.

Veterinaria Italiana 2013, 49 (1), 7-17

Pig raising in Bangladesh and public health interventions

Caressing pigs Both herders and backyard pig raisers often caressed their pigs. Backyard pig raisers believed that pigs did not return to the home unless they were ‘tamed’. The best way to tame pigs was through caressing them while they fed them. They also held or caressed the piglet to calm it so they could inspect it for lice. While caressing, sometimes pigs jerked their bodies and dirt splattered on pig raisers’ body and clothes.

Exposure to pig saliva Pig raisers offered pigs the palms of their hands to lick food from them. Both in backyards and in herds, we noticed that pigs tried to eat food for human consumption from raisers’ plates and bowls or scavenged for food from cooking pots used for human consumption, which were waiting to be cleaned.

Taking care of piglets Pig raisers reported that they took care of the newborn piglets by cleaning them after they were born. Herders kept the mother pig with the newborns adjacent to their tent. They carried young piglets on their back in a piece of cloth that they wore and slept on. They frequently brought the piglets out of that cloth by hand and let them be nursed by their mothers.

Children with pigs In backyards, children often played with piglets by touching, holding, caressing and hugging them. Children also beat, kicked and chased pigs. Sometimes children provided food to the pigs and caressed pigs. They also came in close contact with the mud of the pig-keeping places.

Sharing the same water source The team observed that pigs, cattle, poultry and humans used the same water source. Herders washed their cooking pots and utensils in the ditch in which pigs bathed and drank water. A herder drank water from a ditch after a few pigs drank water and made their bodies wet from the same ditch. Pigs went down in ponds and rivers where people bathed and ducks swam. The team also observed a dead pig floating in the river where other people were bathing.

Interaction with other animals Cattle, dogs, poultry and wild birds often scavenged food from pig-keeping places or from the food pot of the pigs. In backyards, raisers kept pigs, cows and goats closely together or under the same shed. The following note from an observation report from a backyard shows the complexity of pig, poultry and human interactions.

Nahar et al.

Pig raising in Bangladesh and public health interventions

Table III. Observed direct and indirect interactions of pigs with humans, birds and animals in backyards and herds, Bangladesh, 2007-2008. Number of pig-raising communities with observed contact Backyard Herds (%) (%) N=3 N=6

Type of contact Direct contact between pig and human Caressing pigs

(100)

Beating, kicking, pushing, chasing pigs with hands and legs

(67)

(100)

Pigs licking the palm of the pig raisers while feeding

(100)

(33)

Carrying piglets in their clothes

(0)

(33)

Touching own nose, mouth, face after touching and caressing pigs

(100)

Wiping hands on their clothes during or after touching or caressing pigs and later wiping childrens' nosesor mouths with the same cloth

(67)

(0)

Touching pig feces or feces mixed with mud by hand

(100)

Stepping on and/or feeding pigs, sitting and standing barefoot on feces mixed with mud

(100)

Playing with piglets while holding, hugging, kissing, bathing and chasing

(100)

(0)

Babies crawling on the ground with pigpen mud dropped from chickens feet

(33)

(0)

Feeding pigs sitting and standing barefoot on feces mixed with mud

(67)

(0)

Pig putting mouth close to the plates of the raisers while they were eating

(33)

(50)

Pig putting mouth on cooking pots and utensils with or without food for human consumption

(67)

(50)

Pig raisers drinking water from and/or washing cooking pots and utensils in the ditches in which pigs bathed and drank water

(0)

(33)

Pigs eating human feces from toilets

(0)

(33)

Pigs standing in the pond containing human excreta from the toilet

(33)

(0)

Chickens entering human houses and stepping on the chairs, table and bed after scavenging in pig-keeping places

(100)

(0)

Chickens scavenging in the pig-keeping places

(100)

(0)

Pigs, poultry, cattle and wild birds sharing food and water

(100)

Direct contact with children

Indirect contact with surroundings

Interaction of pigs with other animals and birds

“One of the roosters went to the pig pen and scavenged food from the pot and from the ground of the pig keeping place. The rooster walked on the pig’s feces mixed with mud and leftover food of the pigs. Then it came inside the house, scavenged food from the yard, walked on the dining table for a while and then went in front of the kitchen.”

Hand washing practices Pig raisers almost never washed their hands after touching and caressing pigs. Backyard caregivers occasionally rinsed their hands after touching the pig food. Informants from all study sites said that they used soap when they bathed and washed clothes. Herders reported less frequent use of soap than backyard informants. Herd owners provided one bar of soap (usually a local brand of a medicated soap) to each herder once a month to bathe with

and wash clothes. However, the team did not notice any soap use during their observation of herds. Both backyard raisers and herders explained that they were too poor to buy soap to wash their hands. Backyard raisers reported using soap, mud or ash to wash their hands after feeding pigs but our observational data often did not support this report. One of the backyard raisers explained that he did not wash his hands after providing food to pigs as he did not consider pigs as a dirty or filthy animal.

Pig raisers’ experience of stigmatization Disrespect and violence All the informants reported that Bengali Muslims often showed disrespect and expressed verbal and physical abuse and violence towards pigs and pig

Veterinaria Italiana 2013, 49 (1), 7-17

Nahar et al.

raisers because they considered pigs a ‘forbidden animal’. Being ‘forbidden’ meant that, for example, touching, raising, eating, seeing, or uttering the term ‘pig’ in local language ‘shuor’ (in standard Bengali shukor) were strongly offensive to Bengali Muslims. People often did not allow the pig raisers to use the main road with their pigs because of odor and thus pig raisers had to look for other travel routes. People sometimes beat the pig raisers and injured pigs if pigs spoiled anything. A backyard pig raiser shared his experience: “In the last Chaitra mash (Bengali month referring from mid March to mid April), one day, when the owner of a mango garden was collecting green mangos, our pigs spoiled 74 kg of mangos (worth US$10-12). The man became so angry that he tied me to a mango tree and beat me. My legs were bleeding. I was crying and begging, ‘Please don’t hit me anymore. I will reimburse you’. He was not listening to me. He was saying, ‘I will get the reimbursement by killing you’.” We noticed that while a herd passed by looking for food in villages, many villagers of all age groups beat pigs when they could reach one. Often this beating annoyed the herders but they rarely expressed their annoyance in front of the villagers because then the villagers might force them to leave the locality.

Pig raising in Bangladesh and public health interventions

name sometimes helped them to get a job as the authorities preferred to employ Muslims. A backyard pig raiser portrayed their situation as follows: “Pigs are prohibited to Muslims. Since we raise pigs, usually they do not come to our houses and eat anything from us. We also do not go to their houses. We live in fear because we are a minority here. Muslims are comfortable with other Hindus who do not raise pigs. They do business together.” Herders said that people prohibited them from bringing herds and pitching a tent on their land. We also observed herders persuading villagers to allow them to access their communities by explaining that pigs would clean the locality by eating garbage and weeds. The following statement is from a herder: “It’s difficult to get a place for us to stay at night. People complain when they see us making a tent near their houses. They say, ‘Our cattle will travel through the path, you can’t keep your pigs here.’ They say that we are lower caste. Even their young children beat us and rebuke us. We can’t say anything.”

“We can’t keep ourselves clean because people might think that we have money. We can’t wear anything nice even if we like to do so. If we wear a very cheap watch, people say, ‘Shala (addressing them in offensive manner) pig raiser is wearing a watch.’ They beat us and snatch everything from us. If we feel cold, we can’t wear a shawl. They take our shawl, shirt, money, everything. And we, we are human beings. We want to live like humans. We want to have quilts on the floor to sleep but we can’t because people steal them. We raise pigs and we live like pigs. What is the point of using soap and cleaning our clothes when it increases the risk to be beaten?”

Pig raisers often faced difficulties transporting pigs using public vehicles because of the negative attitude of Muslims. Besides scolding the pig raisers, people also rebuked the bus driver and conductor for letting them bring pigs on the bus. Pig raisers reported that as they paid fare for the pigs, they did not face any rejection from bus authorities. When they needed to transport pigs to a distant place, they put pigs in gunny bags to make them inconspicuous. If pigs made sounds and other passengers asked them what they had in their bags, they often answered that they had kalkhashi (black goats). They also called castrated pigs khashi which commonly refers to a castrated goat in Bangladesh. There were two reasons for using alternative words to refer to pigs; one was to avoid language that might be offensive to Muslims and the other was to conceal the fact that they were actually transporting pigs. Even if people understood, they might not be very angry because uttering or listening the word kalkhashi is not unholy like shuor. Otherwise, if they directly said ‘shuor’, people might force them to leave the vehicle. We noticed that pig raisers used several different terms to refer pigs during interviews, such as jongli (wild) and mal (goods).

Social exclusion and coping strategies

Resistance

Pig raisers often expressed their fear of being a minority and a pig raiser. They added that Muslims strictly reject them by not socializing with them. A few pig raisers from Faridpur said that when they went out to work or for other purposes, they often hid their identity as pig raisers to be accepted by others. Sometimes they used a Muslim alias publicly and in official documents because a Muslim

“Our main identity is that we all are human beings. It’s a manmade difference that I am Hindu, you are Muslim, he is Christian. But there is only one God for all human

Insecurity All the herders talked about their personal security, and their security concerns were greatest at night. They were often victims of theft, robbery and physical abuse. The following quotation reflects what almost all the herders reported to us:

Veterinaria Italiana 2013, 49 (1), 7-17

Pig raisers often made arguments against the discrimination, domination and rejection that they faced in their everyday life. The following quotation was from a backyard pig raiser who was working in a hospital as a cleaner.

Pig raising in Bangladesh and public health interventions

beings. You call him Allah; I call Sri- Krishna; he calls Jesus. In fact, we all go to the same place after death. We burn the dead body, while you bury. After burning, the ash and the residues are mixing with earth. Again, after death God will arrange the judgment in the same court for all human being. Even in our country, being the son of a cleaner, if I commit a murder, my judgment will be held in the same court as you if you (Muslim) do the same crime. If you cut your body, you will get red blood just like I would get. As we raise pigs, Muslims show hatred to us. They can tell us many things but will they give money to feed us? If our pigs do not harm them, what is their problem? But they do not consider our situation.”

Discussion Qualitative studies can significantly contribute to infectious disease control and prevention by understanding peoples’ perception and behavior in relation to disease transmission and translating local concerns into appropriate heath interventions (23, 29, 37). In this qualitative study, we identified different pig-raising patterns in Bangladesh, pig interactions with humans and other animals, social stigma that pig raisers experienced and their struggle for survival. Our findings can help to outline future strategies for health interventions, identify difficulties that can appear in this process and indicate possible solutions. Although pigs have not yet been confirmed to be an important source of zoonotic diseases in Bangladesh, studies from neighboring countries where big zoonoses have been confirmed also suggest a close interaction between pigs and people in these contexts. In India and Nepal, low income indigenous communities raise pigs in their backyards (25, 38). They have poor housing with the practice of open defecation and pigs live very close to human dwellings (38, 39, 40). These pigs are free ranging, loose to scavenge in garbage dumps and sewerage and have been identified to be infected by Neurocysticercosis and trichinellosis when pig raisers have been found to be infected by Taenia solium taeniasis (25, 38). Pig-raising communities in Bangladesh raise pigs in similar settings, practice open defecation and let pigs access the garbage. Considering these similarities, Bangladeshi pig raisers and their pigs might be exposed to similar diseases as in India and Nepal. (32). Indeed, in one backyard study site, we observed pigs were fed restaurant scraps that could also include the meat of poultry and cattle which may pass pathogens to pigs. Slaughtering and consuming sick birds and animals is common in Bangladesh (8, 46). From our study, we do not know if pigs eat the meat of pigs that died of illness, a pathway that could efficiently transmit pig pathogens; however, we found that

Nahar et al.

pigs eat dead poultry and other dead animals, so it is likely that at least occasionally they also eat dead pigs. We identified several types of contact that people had with domestic pigs, which might provide a pathway for zoonoses, especially when people came in contact with pig saliva and feces very frequently. Other domestic animals and poultry had frequent interactions with pigs that might lead to cross species transmission and could be a pathway to spread disease to humans (14, 16). Poverty has significant implications on the health of the population (20, 41). Pig raisers in Bangladesh live within extreme resource constraints that severely restrict options to modify their practices. Simple behavior change recommendations like “wash your hands with soap after touching a pig”, “keep a safe distance from pigs” or “don’t feed your pigs carcasses from animals that died” are unlikely to be implementable and so are unlikely to substantially reduce the risk of transmission of pathogens from pigs to people. A similar example comes from the low income Roma community in Romania that raises pigs to maintain their livelihood. This community has been affected by several trichinellosis outbreaks (33). Because of poverty, the Roma are known to consume the carcasses of pigs that died in uncertain circumstances and health interventions are often unsuccessful in changing their behavior. Although we do not know exactly which diseases pig raisers in Bangladesh suffer from, or the health and economic burden of these diseases, we do know that lowincome populations and their domestic animals are more likely to be affected by infectious disease and that zoonoses have an important economic impact at individual and national levels (41). Since diseases can promote poverty, controlling these diseases would be one step toward improving the economic productivity of the low-income pig raisers (20). One option would be highlighting economic benefits of controlling human and pig disease. This might be in the interest of pig-raising households as well as the government, because it is related to poverty reduction that can contribute to the national economy (12, 19). To pursue the economic benefit of disease prevention, which affects both pigs and humans, OneHealth surveillance could be initiated to detect and treat both human and animal diseases that we described elsewhere (32). Briefly, the OneHealth approach could be accomplished through the collaboration between public health and veterinary sectors to work together to combat zoonoses (22). Pig raisers in Bangladesh face substantial social stigma. Such stigma has important community health implications (5). Stigma can impede access to health care and as a result, delayed treatment, which contributes to the spread of disease. Social marginalization can cause poverty and neglect

Veterinaria Italiana 2013, 49 (1), 7-17

Nahar et al.

Pig raising in Bangladesh and public health interventions

that make the population susceptible to infectious disease. Stigmatized populations can distrust health authorities and decline cooperating during health emergencies. Finally, stigma can create mass panic among citizens by distorting public perception of risk. The study team experienced difficulties accessing the pig raisers in the beginning of our study. We also know the unwillingness of health practitioners to provide treatment created mistrust and compelled pig raisers to treat their pigs by themselves (32). Thus it is likely that stigma will present a barrier to engaging pig raisers in public health interventions. Since the majority of Muslims do not accept pigs, targeting only pig-borne diseases in a separate health intervention risks increasing stigma and marginalization. Considering their vulnerable social situation, at the very beginning it would be worthy to develop interventions that will provide social support to reduce the effect of stigma (51). In addition, highlighting pigsâ&#x20AC;&#x2122; contributions to agriculture and the environment could provide acceptance of pig raising to broader Muslim communities. When pig herds pass by they often eat weeds with roots and dig the soil of the agricultural fields, which makes it easier for farmers to cultivate their lands. Pigs also clean the environment by eating organic garbage when there are minimal efforts to use such waste as fertilizer. However, messages regarding garbage need to be carefully framed, such that they do not include human feces in garbage list.

not collect data to compare backyards and herds or explore the relationships between Hindu and Christian pig keepers. Our intention was to develop a basic understanding of both types of pig raising in Bangladesh and establish a good relationship with the pig raisers so that we could use this network for future quantitative studies.

The primary limitation in our research was that most of the field team members were Muslims. Although we invested substantial time in building rapport, these social differences could not be eliminated because of the existing Islamic hegemony in Bangladeshi society. These differences likely limited our ability to ask questions and interpret information. Our study is a qualitative exploratory study with a modest sample and the study sites were purposively selected. This does not allow us to generalize our result like a study with a large random sample but it is an effective way to develop a nuanced, in-depth understanding of the perspectives of affected individuals to understand the issues and motivations pertaining to complex human behavior (30). Findings from this study were based on pig raisersâ&#x20AC;&#x2122; reports and our observations. It would be useful to incorporate the views of the majority Muslims towards pigs, especially to understand the hostility and rejection we documented in our study. We conducted this study with two main types of pig raisers in Bangladesh, though we did

Acknowledgement

Veterinaria Italiana 2013, 49 (1), 7-17

Bangladesh is a low income country with an underresourced health care system (1). The majority of the rural population use untrained or informally-trained health practitioners to address most of their health concerns (3, 11). Quality primary healthcare from the public sector is often unavailable or inaccessible by the rural population, a situation which increases the risk of widespread transmission of new human pathogens before they are recognized (11). In addition, the high population density in Bangladesh makes people more vulnerable to the spread of infectious diseases to the wider population. The issue of pig zoonoses might appear to be the local problem of a minority population, pig raisers. However, emergence of a new disease can also appear as a global threat, as we already experienced with swine flu in 2009 (36). The next initiative would be working with pig raisers to explore how social stigma and marginalization could be minimized as a prerequisite for implementing interventions to improve their health and economy, which might also contribute to reducing the risk of zoonoses.

We are indebted to our study participants for their time, cooperation and invaluable information. Icddr,b acknowledges with gratitude the commitment of CDC to the Centreâ&#x20AC;&#x2122;s research efforts. We thank Dawlat Khan and Utpal Kumar Mondal for their assistance in data collection. We appreciate the thoughtful review of Elizabeth Oliveras. We are grateful to Nadia Ali Rimi for her careful review and editing.

Grand support This study was funded by the Centers for Disease Control and Prevention (CDC), CoAg Grant Number was: 5-U01-CI000298-03.

Conflict of interest/competing interests The authors declare no conflicts of interest.

Pig raising in Bangladesh and public health interventions

Nahar et al.

References 1. Afsana K. 2004. The tremendous cost of seeking hospital obstetric care in Bangladesh. Reprod Health Matters, 12, 171-180. 2. Alam J. 1995. Livestock resources in Bangladesh: present status and future potential. University Press Limited, Dhaka, Bangladesh, 5-15. 3. Amin R., Chowdhury S.A., Kamal G.M. & Chowdhury J. 1989. Community health services and health care utilization in rural Bangladesh. Soc Sci Med, 29, 1343-1349. 4. Asaduzzaman A. 2001. The ‘pariah’ people: an ethnography of the urban sweepers in Bangladesh. University Press Limited, Dhaka, Bangladesh, 2-7. 5. Barrett R. & Brown P.J. 2008. Stigma in the time of influenza: social and institutional responses to pandemic emergencies. J Infect Dis, 197 Suppl 1, S34-S37. 6. Borah J., Dutta P., Khan S. A. & Mahanta J. 2013. Epidemiological concordance of Japanese encephalitis virus infection among mosquito vectors, amplifying hosts and humans in India. Epidemiol Infect, 141, 74-80. 7. Brooks W.A., Alamgir A.S., Sultana R., Islam M.S., Rahman M., Fry A.M., Shu B., Lindstrom S., Nahar K., Goswami D., Haider M.S., Nahar S., Butler E., Hancock K., Donis R.O., Davis C.T., Zaman R.U., Luby S.P. & Uyeki T.M. 2009. Avian influenza virus A (H5N1), detected through routine surveillance, in child, Bangladesh. Emerg Infect Dis, 15, 1311-1313. 8. Chakraborty A., Khan S.U., Hasnat M.A., Parveen S., Islam M.S., Mikolon A., Chakraborty R.K., Ahmed B.N., Ara K., Haider N., Zaki S.R., Hoffmaster A.R., Rahman M., Luby S.P. & Hossain M.J. 2012. Anthrax outbreaks in Bangladesh, 2009-2010. Am J Trop Med Hyg, 86, 703-710. 9. Chua K.B., Goh K.J., Wong K.T., Kamarulzaman A., Tan P.S., Ksiazek T.G., Zaki S.R., Paul G., Lam S.K. & Tan C.T. 1999. Fatal encephalitis due to Nipah virus among pigfarmers in Malaysia. Lancet, 354, 1257-1259. 10. Chua K.B. 2003. Nipah virus outbreak in Malaysia. J Clin Virol, 26, 265-275. 11. Cockcroft A., Andersson N., Milne D., Hossain M.Z. & Karim E. 2007. What did the public think of health services reform in Bangladesh? Three national community-based surveys 1999-2003. Health Res Policy Syst, 5, 1. 12. Conteh L., Engels T. & Molyneux D.H. 2010. Socioeconomic aspects of neglected tropical diseases. Lancet, 375, 239-247. 13. Furst T., Keiser J. & Utzinger J. 2012. Global burden of human food-borne trematodiasis: a systematic review and meta-analysis. Lancet Infect Dis, 12, 210-221. 14. Garten R.J., Davis C.T., Russell C.A., Shu B., Lindstrom S., Balish A., Sessions W.M., Xu X., Skepner E., Deyde V., Okomo-Adhiambo M., Gubareva L., Barnes J., Smith C.B., Emery S.L., Hillman M.J., Rivailler P., Smagala J., de Graaf M., Burke D.F., Fouchier R.A., Pappas C., Alpuche-Aranda C.M., Lopez-Gatell H., Olivera H., Lopez I., Myers C.A., Faix D., Blair P.J., Yu C., Keene

K.M., Dotson P.D., Jr., Boxrud D., Sambol A.R., Abid S.H., St George K., Bannerman T., Moore A.L., Stringer D.J., Blevins P., Demmler-Harrison G.J., Ginsberg M., Kriner P., Waterman S., Smole S., Guevara H.F., Belongia E.A., Clark P.A., Beatrice S.T., Donis R., Katz J., Finelli L., Bridges C.B., Shaw M., Jernigan D.B., Uyeki T.M., Smith D.J., Klimov A.I. & Cox N.J. 2009. Antigenic and genetic characteristics of swine-origin 2009 A(H1N1) influenza viruses circulating in humans. Science, 325, 197-201. 15. Goh K.J., Tan C.T., Chew N.K., Tan P.S., Kamarulzaman A., Sarji S.A., Wong K.T., Abdullah B.J., Chua K.B. & Lam S.K. 2000. Clinical features of Nipah virus encephalitis among pig farmers in Malaysia. N Engl J Med, 342, 1229-1235. 16. Gray G.C., Trampel D.W. & Roth J.A. 2007. Pandemic influenza planning: shouldn’t swine and poultry workers be included? Vaccine, 25, 4376-4381. 17. Greger M. 2007. The human/animal interface: emergence and resurgence of zoonotic infectious diseases. Crit Rev Microbiol, 33, 243-299. 18. Hossain M.J., Gurley E.S., Montgomery S., Petersen L., Sejvar J., Fischer M., Panella A., Powers A.M., Nahar N., Uddin A.K., Rahman M.E., Ekram A.R., Luby S.P. & Breiman R.F. 2010. Hospital-based surveillance for Japanese encephalitis at four sites in Bangladesh, 2003-2005. Am J Trop Med Hyg, 82, 344-349. 19. Hotez P.J., Molyneux D.H., Fenwick A., Kumaresan J., Sachs S.E., Sachs J.D. & Savioli L. 2007. Control of neglected tropical diseases. N Engl J Med, 357, 1018-1027. 20. Hotez P.J., Fenwick A., Savioli L. & Molyneux D.H. 2009. Rescuing the bottom billion through control of neglected tropical diseases. Lancet, 373, 1570-1575. 21. Hotez P.J. & Alibek K. 2011. Central Asia’s hidden burden of neglected tropical diseases. PLoS Negl Trop Dis, 5, e1224. 22. ICDDRB. 2011. Improving Human Health through a One Health Approach in Bangladesh. Health Sci Bull, 9, 17-20. 23. Inhorn M.C. & Brown P.J. 1990. The anthropology of infectious disease. Annu Rev Anthropol, 19, 89-117. 24. Jones R. 1995. Why do qualitative research? BMJ, 311, 2. 25. Joshi D.D., Moller L.N., Maharjan M. & Kapel C.M. 2005. Serological evidence of Trichinellosis in local pigs of Nepal. Vet Parasitol, 132, 155-157. 26. Labrique A.B., Zaman K., Hossain Z., Saha P., Yunus M., Hossain A., Ticehurst J. & Nelson K.E. 2009. Population seroprevalence of hepatitis E virus antibodies in rural Bangladesh. Am J Trop Med Hyg, 81, 875-881. 27. Luby S.P., Hossain M.J., Gurley E.S., Be-Nazir A., Banu S., Khan S.U., N. H., Rota P.A., Rollin P.E., Comer J.A., E K., T.G K. & Rahman M. 2009. Recurrent Zoonotic Transmission of Nipah Virus into Human Bangladesh, 2001–2007. Emerg Infect Dis, 15, 1229-1235. 28. Malterud K. 2001. Qualitative research: standards, challenges, and guidelines. Lancet, 358, 483-488.

Veterinaria Italiana 2013, 49 (1), 7-17

Nahar et al.

29. Manderson L. 1998. Applying medical anthropology in the control of infectious disease. Trop Med Int Health, 3, 1020-1027. 30. Marshall M.N. 1996. Sampling for qualitative research. Fam Pract, 13, 522-525. 31. Mukherjee A., Dutta D., Ghosh S., Bagchi P., Chattopadhyay S., Nagashima S., Kobayashi N., Dutta P., Krishnan T., Naik T.N. & Chawla-Sarkar M. 2009. Full genomic analysis of a human group A rotavirus G9P[6] strain from Eastern India provides evidence for porcine-to-human interspecies transmission. Arch Virol, 154, 733-746. 32. Nahar N., Uddin M., Gurley E., Khan M.S.U., Hossain J., Sultana R. & Luby S.P. 2012. Pig illnesses and epidemics: a qualitative study on perceptions and practices of pig raisers in Bangladesh. Vet Ital, 48, 157-165. 33. Neghina R. 2010. Trichinellosis, a Romanian neverending story. An overview of traditions, culinary customs, and public health conditions. Foodborne Pathog Dis, 7, 999-1003. 34. Parashar U.D., Sunn L.M., Ong F., Mounts A.W., Arif M.T., Ksiazek T.G., Kamaluddin M.A., Mustafa A.N., Kaur H., Ding L.M., Othman G., Radzi H.M., Kitsutani P.T., Stockton P.C., Arokiasamy J., Gary H.E., Jr. & Anderson L.J. 2000. Case-control study of risk factors for human infection with a new zoonotic paramyxovirus, Nipah virus, during a 1998-1999 outbreak of severe encephalitis in Malaysia. J Infect Dis, 181, 1755-1759. 35. Pavio N., Meng X.J. & Renou C. 2010. Zoonotic hepatitis E: animal reservoirs and emerging risks. Vet Res, 41, 46. 36. Peiris J.S., Poon L.L. & Guan Y. 2009. Emergence of a novel swine-origin influenza A virus (S-OIV) H1N1 virus in humans. J Clin Virol, 45, 169-173. 37. Pope C. & Mays N. 1995. Reaching the parts other methods cannot reach: an introduction to qualitative methods in health and health services research. BMJ, 311, 42-45. 38. Prakash A., Kumar G.S., Rout M., Nagarajan K. & Kumar R. 2007. Neurocysticercosis in free roaming pigs - a slaughterhouse survey. Trop Anim Health Prod, 39, 391-394. 39. Prasad K.N., Prasad A., Gupta R.K., Pandey C.M. & Singh U. 2007. Prevalence and associated risk factors of Taenia solium taeniasis in a rural pig farming

Veterinaria Italiana 2013, 49 (1), 7-17

Pig raising in Bangladesh and public health interventions

community of north India. Trans R Soc Trop Med Hyg, 101, 1241-1247. 40. Prasad K.N., Verma A., Srivastava S., Gupta R.K., Pandey C.M. & Paliwal V.K. 2011. An epidemiological study of asymptomatic neurocysticercosis in a pig farming community in northern India. Trans R Soc Trop Med Hyg, 105, 531-536. 41. Seimenis A. 2012. Zoonoses and poverty - a long road to the alleviation of suffering. Vet Ital, 48, 5-13. 42. Sethi B., Butola K.S., Kumar Y. & Mishra J.P. 2012. Multiple outbreaks of trichinellosis with high mortality rate. Trop Doct, 3 April 2012, 10.1258/td.2012.110335. 43. Solomon T. 2006. Control of Japanese Encephalitis Within Our Grasp? N Engl J Med, 355, 869-871. 44. Steyer A., Poljsak-Prijatelj M., Barlic-Maganja D. & Marin J. 2008. Human, porcine and bovine rotaviruses in Slovenia: evidence of interspecies transmission and genome reassortment. J Gen Virol, 89, 1690-1698. 45. Stuber J., Meyer I. & Link B. 2008. Stigma, prejudice, discrimination and health. Soc Sci Med, 67, 351-357. 46. Sultana R., Nahar N., Rimi N.A., Azad S., Islam M.S., Gurley E.S. & Luby S.P. 2012. Backyard poultry raising in Bangladesh: a valued resource for the villagers and a setting for zoonotic transmission of avian influenza. A qualitative study. Rural Remote Health, 12, 1927. 47. Tsutsumi A., Izutsu T., Islam A.M., Maksuda A.N., Kato H. & Wakai S. 2007. The quality of life, mental health, and perceived stigma of leprosy patients in Bangladesh. Soc Sci Med, 64, 2443-2453. 48. Van Reeth K. 2007. Avian and swine influenza viruses: our current understanding of the zoonotic risk. Vet Res, 38, 243-260. 49. Vaughn D.W. & Hoke C.H., Jr. 1992. The epidemiology of Japanese encephalitis: prospects for prevention. Epidemiol Rev, 14, 197-221. 50. Vivek R. & Kang G. 2011. Hepatitis E virus infections in swine and swine handlers in Vellore, Southern India. Am J Trop Med Hyg, 84, 647-649. 51. Wang J.W., Cui Z.T., Cui H.W., Wei C.N., Harada K., Minamoto K., Ueda K., Ingle K.N., Zhang C.G. & Ueda A. 2010. Quality of life associated with perceived stigma and discrimination among the floating population in Shanghai, China: a qualitative study. Health Promot Int, 25, 394-402.

Population displacements as a risk factor for the emergence of epidemics Darem Tabbaa1 & Aristarco Seimenis2 1

DVM, PhD, Department of Public Health - Faculty of Veterinary Medicine, Al Baath University, Hama, Syria daremtabbaa@yahoo.com 2 DVM, PhD, WHO/Mediterranean Zoonoses Control Center, Athens, Greece

Keywords Animal health, Conflict, Emergence, Epidemic, Infections, Population displacement, Public health.

Summary Wars and civil conflicts have been terrible experiences since ancient times but, regretfully, they are always present even in the 21st century. Their catastrophic effects are still lived by many populations displaced from their native areas. Conflicts, particularly the civil ones, create disruption in most aspects of national structures and populations, which are forced to move to more or less safer or even distant areas, survive under downgraded conditions. They are usually housed in temporary shelters in overcrowded camps and contaminated environment. Water and food are neither safe nor sufficient. Malnutrition, lack or weak sanitary care and long-term stress lead these populations to being vulnerable to severe infections. Under such conditions there are high rates of morbidity and mortality, with elders and children being the main victims. Public health, animal health, municipalities and other inter-related sectors should work on preparedness plans well in advance in order to provide ways and means to face emergencies. Zoonotic and other communicable disease outbreaks should not be left uncontrolled, as their impact would be an additional burden for the country under unrest. Guidance should be provided on how to best articulate an emergency management plan from the early detection of outbreaks up to their control. These aspects are briefly exposed together with the imperative request for alleviation of suffering and of the multitude of hazards conflict-affected populations have to face.

Trasferimenti di popolazioni e fattori di rischio per le emergenze epidemiche Parole chiave Conflitto, Emergenza, Epidemia, Infezione, Salute animale, Salute pubblica, Trasferimento di popolazione, Zoonosi.

Riassunto Guerre e conflitti civili sono esperienze terribili da tempi remoti, purtroppo, ancora oggi di attualità. I loro effetti catastrofici sono ancora vissuti da numerose popolazioni costrette a lasciare i propri territori. I conflitti, in particolare quelli civili, sono causa del crollo di molte strutture nazionali. Le popolazioni sono costrette a muoversi verso territori, o altri paesi, non sempre sicuri, a volte lontani, e a sopravvivere in condizioni precarie. Di solito sono alloggiate in rifugi temporanei e inadeguati, in campi sovraffollati e ambienti contaminati. Acqua e alimenti non sono sufficienti e sicuri. Malnutrizione, cure sanitarie scarse o assenti, associate a stress protratti, rendono queste popolazioni vulnerabili a diverse infezioni. In queste situazioni i tassi di morbilità e mortalità sono molto alti. Anziani e bambini sono le principali vittime. Per far fronte a queste emergenze, sanità pubblica, sanità animale, autorità e altre strutture correlate devono approntare specifici piani preventivi per utilizzare vie e mezzi, per quanto possibile, semplificati e appropriati. Focolai di zoonosi e altre malattie infettive, se lasciati senza controllo, avrebbero effetti disastrosi e costituirebbero un peso supplementare per la popolazione del paese. I piani preventivi devono prevedere linee guida sulle modalità più attendibili per articolare e mettere in funzione la gestione delle emergenze epidemiche, dall’identificazione precoce di eventuali focolai sino al loro controllo. L’articolo evidenzia questi aspetti, oltre alla richiesta impositiva di limitare le sofferenze e i numerosi pericoli ambientali e sanitari alle popolazioni vittime di conflitti. Veterinaria Italiana 2013, 49 (1), 19-23

Population displacements and epidemic emergence

Tabbaa & Seimenis

Introduction The reality that we are facing in our times consists of conflicts exploding for different reasons in different countries or even regions, which lead to population displacement into more or less distant areas, in search of security. They are usually housed into temporary settlements or camps with overcrowded and rudimentary shelters, unsafe water, food and sanitation, and increased exposure to disease vectors. Such situations frequently lead to infectious disease outbreaks usually as a result of substantial and exacerbated synergic risk factors: e.g. changes due to deterioration in the environment, in human living conditions, and in vulnerability to biological pathogens and disease transmission (6, 25). In long-term emergencies, populations have high rates of morbidity and mortality due to the breakdown of sanitary services, flight of trained staff, failure of existing disease control programs, and destroyed infrastructures. Moreover, high levels of malnutrition, low or absent vaccine coverage and long-term stress make people more sensitive to infection and disease. In addition, long-term conflicts affect entire communities because of chronic lack of investment in public and animal health, education, public works for the development of the environment, etc. These changes in human conditions, in the ecosystem of pathogens, and in the environment facilitate the occurrence and transmission of infectious diseases (epidemiologic

triad). Trans-boundary movements of refugees, relief workers, animals, goods, immigrants, etc. may cause international spread of infectious disease outbreaks and epidemics (14, 15, 25) (Figure 1). This review aims at a summary description of potential infectious disease outbreaks and/or epidemics frequently manifested during and after conflicts, as well as the disruption conditions created. Prevention and control measures to be considered by animal health and public health services as well as by humanitarian organisations and professionals in addressing public health challenges in countries under unrest, are also summarised.

Conflict situations and risk factors enhancing disease emergence and transmission Environmental deterioration and communicable diseases Leishmaniasis extension was reduced in Syria at the end of 2009 as a result of the implementation of vector control programs. However, with the onset of conflicts in 2011, which has continued almost without interruption, such a control program collapsed and enabled leishmaniaisis re-emergence and the number of cases has been increasing due to the displacement of population in different safer provinces.

Ebola, Marburg, and CCHF Influenza (H5N1) Lassa fever Monkeypox Nipah and Hendra

Rift Vallery fever SARS-CoV Yellow fever Poliovirus

Plague Tularemia Malaria Trypanosomiasis

Figure 1. Distribution of emerging or re-emerging infectious disease outbreaks and countries affected by conflicts between1990-2006 (indicated in yellow in the map). Circles indicate diseases of viral origin, stars indicate diseases of bacterial origin, and triangles indicate diseases of parasitic origin. Source: Office for the Coordination of Humanitarian Affairs, World Health Organization, 2012.

Veterinaria Italiana 2013, 49 (1), 19-23

Tabbaa & Seimenis

There was a significant recrudescence of sleeping sickness (human African trypanosomiasis) in the 1990s, predominantly in conflict-affected Angola, Democratic Republic of Congo (DRC), and Southern Sudan. In particular, the DRC has had a dramatic resurgence of this disease as a direct consequence of conflicts. Control measures were interrupted in the 1990s because of conflicts, which resulted in more than 150 000 new cases from 1989 up to 1998, with 26,000 cases during this year. Since 1998, detection and treatment have been reinforced in Africa, and the number of new cases has substantially decreased among the general population. However, despite intensification of control measures, all major outbreaks in 2005 occurred in conflict-affected countries, i.e. Angola, DRC, and Southern Sudan (11, 20, 24, 25).

Inadequate surveillance, early-warning response systems and control measures Ongoing conflicts are hampering access to human and animal populations for routine monitoring, timely delivery of supplies and implementation of control measures before and during an outbreak. Moreover, vaccination campaigns are also interrupted during protracted conflicts due to long-term inadequate logistic support and/or ongoing insecurity.

Treatment ineffectiveness and development of drug resistance Pathogens’ resistance to drugs and disease complications may develop more rapidly in conflict situations because of inappropriate diagnoses or inappropriate drug regimens and outdated drugs. Treatment compliance may be poor due to purchasing of insufficient quantities of drugs, selling or saving them by patients, or interruption of treatment in sudden displacement or irregular access to healthcare facilities. In addition, private pharmacies, which can flourish in conflict situations because of no implementation of regulations, can compound this problem with drugs of unknown quality and acceptance of prescriptions from unqualified people (7).

Improving early detection and response to infectious diseases in conflict situations During emergencies, veterinary and public health officers should work together with the communities to create an environment in which public health risks are reduced as much as possible, and the safety and dignity of emergency-affected populations is enhanced. They should focus mainly on the following: • provision of clean, safe and adequate quantities of water and food;

Veterinaria Italiana 2013, 49 (1), 19-23

Population displacements and epidemic emergence

• improved sanitation and vector control; • promotion of essential activities towards preserving animal and public health as well as supporting the most acceptable level of environment achievable; • distribution of items essential for preserving health and hygiene. Early detection and response of many emerging infectious diseases require the best functional and effective veterinary and public health care systems that can be achieved. These systems involve timely investment in animal health and public health services and primary health care infrastructures, trained human resources and provision of essential drugs, general supplies, vaccines, equipment, etc. United Nations agencies, international organisations and NGO’s are providing crucial humanitarian assistance to many conflict-affected populations in coordination with relevant authorities. However, veterinary and public health services that should investigate for the possible expansion of zoonoses and other communicable diseases are becoming particularly weak and need support (1, 3, 9, 10, 13, 18). In such settings good hygiene and standard infection control precautions in animal and public health facilities must be planned beforehand in order to reduce the potential for environmental and nosocomial transmission and expansion of diseases, particularly zoonotic ones. Correct guidance must be given on the rationale for infection control and use of protective personal equipment (PPE) as well as isolation where possible.

Preparedness and response All countries should be prepared to face any emergency situation based upon suitable contingency planning. For example, it is imperative that the technical capacity of all humanitarian health partners together with the ministries of health and agriculture in charge of disease surveillance, prevention, and control should be particularly enhanced in conflict-affected countries, in order to ensure the most feasible effective implementation of infectious disease interventions. This could be achieved through availability of internationally accepted standards, IHR guidelines, and tools adapted to conflict situations. They should be supported by previous specific training of veterinary and public health planners, as well as animal and human health facility staff. Mobilisation of international experts to provide technical field support as required is important. The capacity of national staff must be increased beforehand so that it is readily available, especially in times of heightened insecurity, when some of the staff often

Population displacements and epidemic emergence

remain behind in an area under conflict and still continue to work (2, 3, 19).

to implement core capacities for detection and response to epidemics (22).

The establishment of surveillance systems in developing countries relies on close partnerships with international organisations, NGOs and community groups. Effective surveillance systems in emergencies involve selection of a small number of syndrome-based priority events, using standard surveillance forms, and simplified case definitions, health facilitiesâ&#x20AC;&#x2122; weekly reporting of major importance data, following IHR standards, immediate reporting if set alert thresholds are passed, and establishing community mechanisms for identifying disease clusters (22).

A consistent, transparent and objective policy is needed wherever and whenever possible, for military humanitarian interventions, as well as extensive civil-military liaisons and close cooperation with other humanitarian agencies.

Adequate laboratories for pathogen confirmation must be identified in advance, and support should be provided regarding training and supplying equipment and reagents. Ideally, mechanisms should be formulated beforehand for specimen transport and stockpiling of essential drugs, supplies, and outbreak investigation kits. Data should be analysed locally and regular feedback provided to veterinary and public health partners. The revised 2005 IHR provide a global legal framework to guide the response to animal and public health events of international concern. Conflict-affected countries represent one of the weakest links in global health security and should be prioritised by the international community in provision of technical and operational support

Tabbaa & Seimenis

Conclusions War conflicts, particularly civil ones, are creating serious emergencies and disasters in different countries or even regions of the world. The major impact is the vulnerability of displaced populations to different kinds of infections due to the disruption of public and animal health care systems. Poverty, misery and mourns are the inevitable consequences of the use for conflicts needs of resources which could be used otherwise. Early detection, containment, response and control of emerging infectious diseases in conflict situations are major challenges because of multiple risk factors that promote disease transmission and hinder control even more than those in many resourcepoor settings. Beyond the global public and animal health imperatives, to prevent the emergence and international spread of infectious diseases, there is also a substantial and moral imperative to alleviate suffering from the effects of a multitude of hazards on already vulnerable conflict-affected populations (2, 3).

Veterinaria Italiana 2013, 49 (1), 19-23

Tabbaa & Seimenis

Population displacements and epidemic emergence

References 1. Brown C. 2004. Emerging zoonoses and pathogens of public health significance â&#x20AC;&#x201C; an overview. Rev Sci Tech, 23 (2), 435-442.

13. Maudlin I., Eisler M.C. & Welburn S.C. 2009. Neglected and endemic zoonoses. Phil Trans R Soc B, 364, 2777-2787.

2. Chornel B.B. 2003. Control and prevention of emerging zoonoses. JVME, 30 (2), 145-147.

14. Morse S.S. 1995. Factors in the emergence of infectious diseases. Emerg Infect Dis, 1, 7-15.

3. Cutler S.J., Fooks A.R. & van der Poel W.H.M. 2010. Public health threat of new, reemerging, and neglected zoonoses in the industrialized world. Emerging Infectious Dis, 16 (1), 1-8.

15. Morse S.S. 2004. Factors and determinants of disease emergence. Rev Sci Tech, 23, 443-451.

4. Fisher-Hoch S.P. 2005. Lessons from nosocomial viral hemorrhagic fever outbreaks. Br Med Bull, 73-74, 123-137. 5. Gayer M. & Connolly M.A. 2006. Tuberculosis control in refugee and displaced populations. In Tuberculosis: a comprehensive international approach, 3rd Ed. (M.C. Raviglione, ed.). Informa Healthcare, New York, 907-918. 6. Gayer M., Legros M., Formenty P. & Connolly M.A. 2007. Conflict and emerging infectious diseases. Emerg Infect Dis, 13 (11), 1625-1631. 7. Githui W.A., Hawken M.P., Juma E.S., Godfrey-Faussett P., Swai O.B. & Kibuga D.K. 2000. Surveillance of drugresistant tuberculosis and molecular evaluation of transmission of resistant strains in refugee and nonrefugee populations in North-Eastern Kenya. Int J Tuberc Lung Dis, 4, 947-955. 8. Gruthmann J.P., Klovstad H., Boccia D., Hamid N., Pinoges L. & Nizou J.Y. 2006. A large outbreak of hepatitis E among a displaced population in Darfur, Sudan, 2004: the role of water treatment methods. Clin Infect Dis, 42, 1685-1691. 9. Jaffry K.T., Ali S., Rasool A., Raza A. & Gill Z.J. 2009. Zoonoses. Int J Agric Biol, 11, 217-220. 10. Katare M. & Kumar M. 2010. Emerging zoonoses and their determinants. Vet World, 3 (10), 481-484. 11. Lutumba P., Robays J., Miaka mia Bilenge C., Mesu V.K., Molisho D., Declercq J. 2005. Trypanosomiasis control, Democratic Republic of Congo, 1993-2003. Emerg Infect Dis, 11, 1382-1388. 12. Mantovani A., Lasagna E., Senigallesi A., Comin D., Duque C. 2006. Veterinary public health and war: a neglected chapter of human history. MZCC Information Circular, 60, 2-3.

Veterinaria Italiana 2013, 49 (1), 19-23

16. Reintjes R., Dedushaj I., Gjini A., Jorgensen T.R., Cotter B. & Lieftucht A. 2002. Tularemia outbreak investigation in Kosovo: case control and environmental studies. Emerg Infect Dis, 8, 69-73. 17. Seimenis A. 2012. Zoonoses and poverty, a long road to the alleviation of suffering. Vet Ital, 48 (1), 5-13. 18. Sharp T.W., Wightman J.M., Davis M.J., Sherman S.S. & Burkle F.M. Jr. 2001. Military assistance in complex emergencies: what have we learned since the Kurdish relief effort? Prehosp Disaster Med, 16, 197-208. 19. Tabbaa D. 2008. Control of zoonoses in emergency situations: lessons learned during recent outbreaks (gaps and weaknesses of current zoonoses control programmes). Vet Ital, 44 (4), 611-620. 20. van Nieuwenhove S., Betu-Ku-Mesu V.K., Diabakana P.M., Declercq J. & Bilenge C.M. 2001. Sleeping sickness resurgence in the DRC: the past decade. Trop Med Int Health, 6, 335-344. 21. World Health Organization (WHO) 2006. Vaccine preventable diseases 2006: monitoring system. 2006 Global summary. WHO, Geneva, WHO/IVB/2006. 22. World Health Organization (WHO) 2002. Early Warning and Response Network (EWARN), Southern Sudan. Weekly Epidemiol Rec, 77, 26-27. 23. World Health Organization (WHO) 2005. International Health Regulations, 2nd Ed. WHO, Geneva, 1-74 (www. who.int/ihr/eu). 24. World Health Organization (WHO) 2006. African trypanosomiasis (sleeping sickness). WHO, Geneva, Fact Sheet No. 259. 25. World Health Organization (WHO) 2012. Zoonoses and the Millennium development goals â&#x20AC;&#x201C; rationale and context. In Research priorities for zoonoses and marginalized infections. WHO, Geneva, WHO Technical Report Series No. 971, 119 pp.

Indagine su lesioni provocate da animali in Provincia Autonoma di Bolzano nell’anno 2010 Giulia Morosetti1, Marica Toson2, Christian Piffer1 Azienda sanitaria dell’Alto Adige - Servizio Veterinario, Comprensorio di Bolzano, via Laura Conti, 4 - 39100 Bolzano, Italia giulia.morosetti@asbz.it 2 Centro Regionale di Epidemiologia Veterinaria “Giovanni Vincenzi”, Istituto Zooprofilattico Sperimentale delle Venezie, Viale dell’Università, 10 - 35020 Legnaro (PD), Italia 1

Parole chiave Aggressività, Cane, Fattore di rischio, Gatto, Indagine epidemiologica, Morsicature, Sanità pubblica veterinaria.

Riassunto Le lesioni provocate da animali, in particolare cani, rappresentano un problema sanitario. L’opinione pubblica reagisce al problema spesso in modo sensibile. Per gestire e prevenire efficacemente questi episodi è essenziale valutare l’impatto del fenomeno sulla salute pubblica e identificare i principali fattori di rischio connessi. La presente indagine effettuata in Provincia Autonoma di Bolzano ha avuto l’obiettivo di rilevare i vari parametri epidemiologici utili alla comprensione del problema a livello locale. Nello studio sono stati riportati l’incidenza e l’impatto sui servizi sanitari delle lesioni umane riferite a varie specie animali nell’anno 2010, nonché un’analisi più dettagliata degli episodi causati da cani con il profilo degli individui aggrediti e degli animali coinvolti. Sono stati analizzati vari fattori (geografici, contestuali, stagionali e relazionali) che possono essere associati ad episodi di reazione aggressiva di cani. Sulla base dei dati raccolti sono state elaborate raccomandazioni per la prevenzione di situazioni a rischio. Veterinaria Italiana 2013, 49 (1), 25-36

Introduzione Gli incidenti dovuti a lesioni provocate da animali, soprattutto morsicature di cane, rappresentano un problema di sanità pubblica importante con costi sia in termini umani sia sanitari per le conseguenze legate ai traumi e allo stress post-traumatico, con variazioni di entità nei vari Paesi del mondo. La comprensione del fenomeno e dei fattori di rischio legati a questi episodi sono il primo passo per prevenirli e limitarne il numero. In Italia nord-orientale e nei Paesi confinanti, dal 2009, la rabbia è riemersa con un peso economico non indifferente a causa delle misure di controllo necessarie per prevenire il diffondersi dell’infezione. Lo studio della morbosità derivante da lesioni di animali, in particolare cani, e l’analisi della distribuzione spaziale degli eventi possono contribuire alla valutazione del rischio e a progettare interventi preventivi mirati. Nell’anno 2010, vista l’assenza di indagini sul fenomeno delle lesioni da animali relative alla Provincia Autonoma di Bolzano (PAB), è stato condotto uno studio riguardante l’intero territorio.

Materiali e Metodi Tipo di indagine È stato condotto uno studio d’osservazione (crosssectional) e retrospettivo dei dati di sorveglianza medico-veterinaria notificati nella PAB, nell’anno 2010, in un’area di rilevazione comprendente i 4 Comprensori della ASL dell’Alto Adige, corrispondenti al territorio della PAB.

Raccolta dati Come fonte dei dati sono stati utilizzati i questionari standardizzati di notifica compilati dai reparti di emergenza-urgenza degli ospedali e dai medici del territorio contenenti dati su età e sesso delle persone ferite, luogo e contesto dell’episodio, nonché tipo di animale coinvolto. In aggiunta sono stati raccolti e analizzati altri questionari standardizzati, riguardanti le valutazioni sui cani morsicatori, effettuati dai veterinari ufficiali del Servizio Veterinario ASL. Per la valutazione del contesto, i veterinari hanno utilizzato un metodo standardizzato di quantificazione del temperamento e dell’equilibrio psicologico del cane, nonché della capacità di gestione del

Morosetti et al.

Lesioni provocate da animali nella provincia di Bolzano

Tabella I. Frequenza relativa del tipo di binomio cane morsicatore/conduttore per tipo di valutazione del veterinario ufficiale. Tipo di valutazione 1. Cane equilibrato e ben gestito, la reazione dell'animale è comprensibile e giustificabile dal contesto 2. Cane equilibrato e abbastanza ben gestito, il proprietario può migliorare 3. Cane equilibrato, ma il proprietario non lo gestisce correttamente 4. Proprietario gestisce bene, ma il cane è pauroso/ansioso/reattivo 5. Proprietario migliorabile nella gestione, il cane è pauroso/ansioso/reattivo 6. Proprietario senza controllo, il cane è pauroso/ansioso/reattivo 7. Cane aggressivo, ma il padrone lo gestisce bene 8. Cane aggressivo, il padrone può migliorare la gestione e il mantenimento del cane 9. Cane aggressivo, proprietario senza controllo, indicato patentino Totale

Tabella II. Distribuzione per specie animale del numero di lesioni totali. Specie animale Cani Gatti Altri animali Totale

Numero di lesioni 390 52 18* 460

% di lesioni 84,8 11,3 3,9 100,0

Lim Inf 95% C.I. 81,1 8,6 2,4

Lim Sup 95% C.I. 87,9 14,6 6,3

Numero cani 127 20 1 31 40 3 0 5 4 231

% cani 55,0 8,7 0,4 13,4 17,3 1,3 0,0 2,2 1,7 100,0

Tabella III. Distribuzione per residenza in Alto Adige del numero di lesioni totali. Residenza in Alto Adige Sì No Totale

Numero di lesioni 397 63 460

% di lesioni 86,3 13,7 100,0

*Delle lesioni causate da altri animali fanno parte tre morsi di vipera e uno di pipistrello.

proprietario, con uno schema di punteggio prestabilito che ha permesso di allocare i casi in 9 categorie di possibili binomi cane-conduttore (Tabella I). I dati sulla popolazione umana residente sono stati ottenuti dall’ultima rilevazione ASTAT 2009 (2). I dati sugli accessi al pronto soccorso (P.S.) sono stati gentilmente forniti dall’osservatorio epidemiologico provinciale.

Definizione del caso Sono stati considerati tutti gli episodi notificati dai reparti di emergenza-urgenza e dai medici del territorio riferiti ad accessi per lesioni da animale, indipendentemente dalla natura della lesione (trauma, morsicatura, graffio, livido, ecc.) tra il 1 gennaio 2010 e il 31 dicembre 2010 nella PAB.

Analisi statistica È stata elaborata una rappresentazione del fenomeno delle lesioni da animale per il territorio provinciale mediante analisi descrittiva e ottenuta l’incidenza annuale delle lesioni da animale comprese quelle canine. È stata analizzata e rappresentata la distribuzione spazio-temporale degli episodi per stimare le aree e la popolazione a maggiore rischio. È stata realizzata una rappresentazione delle caratteristiche più comuni dell’individuo aggredito, del cane morsicatore e del contesto del caso.

Le incidenze calcolate sono tassi grezzi o specifici d’incidenza cumulata, calcolati come il numero di nuovi casi avvenuti nei residenti dell’Alto Adige nell’anno 2010 sul numero di abitanti all’inizio dell’anno considerato (popolazione sana a rischio). Per misurare la forza di associazione tra variabili dicotomiche è stato calcolato l’odds ratio con il relativo intervallo di confidenza esatto al 95%. Per valutare la significatività dell’associazione è stato impiegato il test del Chi-quadro. Per mettere a confronto le distribuzioni delle età degli individui feriti da cani e di quelli feriti da gatti è stato usato il test non parametrico di WilcoxonMann-Whitney per il confronto fra le mediane di due campioni indipendenti. Per determinare il grado di concordanza tra individuo aggredito e padrone nel reale contesto della morsicatura è stato calcolato il test kappa di Cohen. Per valutare l’andamento nel tempo dell’incidenza annua è stato usato il test non parametrico per trend su gruppi ordinati di Cuzick (1985). Per alcune percentuali sono stati calcolati i relativi intervalli di confidenza (intervalli di confidenza esatti per variabili binomiali) nell’ipotesi che i casi osservati costituiscano un campione rappresentativo di una popolazione più generica. Per alcune variabili quantitative sono stati riportati alcuni indici di sintesi quali: media, mediana, minimo, massimo e scarto quadratico medio o deviazione standard (SD).

Veterinaria Italiana 2013, 49 (1), 25-36

Morosetti et al.

Lesioni provocate da animali nella provincia di Bolzano

Software utilizzati EPI Info 3.5.1 (EPIINFO™, CDC, Atlanta, USA); Stata 9 (StatacorpLP 4905 Lakeway Drive, TX 77845, USA); Microsoft Word, Excel (Microsoft Office 97/2000); ESRI™ ArcMap 10.0, USA.

Risultati Le Tabelle II, III e IV riassumono la situazione riguardante il fenomeno delle lesioni da animale in provincia di Bolzano. Le lesioni da cane, preponderanti rispetto a quelle di altre specie animali, hanno interessato principalmente persone residenti.

Ricorso a cure mediche Quasi la metà degli episodi è stata presa in carico dal comprensorio sanitario di Bolzano (Figura 1). In questo comprensorio si è registrato anche il numero più elevato di cani nella PAB (n = 14.048; 41% dei cani registrati in anagrafe canina provinciale). La frequenza del ricorso a cure mediche per lesione da animale è riportata nelle Tabelle V e VI. Gli accessi al P.S. sono stati 398 (81,5%), in 62 casi (13,5%) il paziente si è rivolto ai servizi di medicina del territorio.

Per quanto riguarda l’importanza delle lesioni da animale rispetto ad altre cause, nella Tabella VII è possibile osservare come le lesioni da animale incidano in minima parte rispetto ad altre cause di accesso per infortunio ai P.S. ospedalieri dei comprensori sanitari di Merano, Bressanone e Brunico, eccetto quello di Bolzano (dati non disponibili).

Periodi di maggiore impatto Il picco degli accessi a cure mediche per lesione da animale si è osservato in estate (Figura 2).

Prognosi e gravità della lesione In 252 casi su 460 accessi (54,8%) è stata riportata la prognosi, la cui durata media è risultata di 7,9 giorni (SD = 5,2646; Min = 1 - Max = 56; Median = 7). Solo in 8 casi su 460 (1,74%) è stato necessario il ricovero ospedaliero. Di questi, 5 casi dovuti a lesioni da cane e 3 da morso di vipera. I cani sono stati gli animali maggiormente coinvolti nelle lesioni gravi, purtroppo la descrizione dettagliata della lesione è stata riportata dai medici solo in maniera sporadica e insufficiente per effettuarne l’analisi approfondita.

Nel comprensorio di Merano vi è stata la proporzione più alta degli accessi alla medicina del territorio rispetto ai P.S. ospedalieri, seguita dal comprensorio di Brunico. Quasi tutte le notifiche di lesioni da animale a Bolzano e a Bressanone sono risultati provenire da P.S.

Tabella IV. Distribuzione per specie animale del numero di lesioni causate nei residenti. Specie animale Cani Gatti Altri animali Totale (a)

Numero di lesioni 337 47 13 397

% di lesioni 84,9 11,8 3,3 100,0

BRU 14%

BX 17%

Incidenza su 100 mila residenti(a) 66,9 9,9 2,6 78

Incidenza calcolata sulla popolazione residente in Alto Adige nel 2010 = 503399 abitanti

BZ 49% ME 20%

Figura 1. Lesioni da animali: distribuzione per Comprensori sanitari su un totale di 460.

Tabella V. Distribuzione per comprensorio e tipo di accesso a cure mediche. Comprensorio sanitario Bolzano Bressanone Merano Brunico Totale

Veterinaria Italiana 2013, 49 (1), 25-36

Numero di accessi al P.S. 211 71 65 51 398

Numero di accessi al medico 15 8 25 14 62

Totale

% di accessi

226 79 90 65 460

49,1 17,2 19,6 14,1 100,0

% accessi al P.S. per comprensorio su totale accessi al P.S. 53,0 17,8 16,3 12,9 100,0

Morosetti et al.

Lesioni provocate da animali nella provincia di Bolzano

Tabella VI. Distribuzione per comprensorio dei soli accessi al Pronto Soccorso.

40 30 20 10

Tabella VII. Distribuzione per causa degli accessi al Pronto Soccorso, escluso Bolzano. Causa di accesso Lesioni da animali Incidente stradale Traumat. ambiente di lavoro Traumat. ambiente domestico Violenza altrui Autolesione/tentato suicidio Incidente sportivo Incidente sciistico Incidente scolastico Altro tipo di incidente Altri sintomi Totale

Numero 187 2.056 6.936 11.691 551 45 11.152 2.897 1.347 12.886 106.156 155.717

% 0,12% 1,30% 4,50% 7,50% 0,40% 0,00% 7,20% 1,90% 0,90% 8,30% 68,20% 100,00%

Dicembre

Novembre

Ottobre

Settembre

Mese

Agosto

Luglio

Giugno

Maggio

Aprile

0 Marzo

La percentuale è calcolata sul numero totale degli accessi a cure mediche per ogni singolo comprensorio.

Febbraio

(a)

Gennaio

Bolzano Bressanone Merano Brunico Totale

Numero di % accessi Lim Inf Lim Sup accessi al P.S. al P.S.(a) 95% C.I. 95% C.I. 211 93,4 89,3 96,2 71 89,9 81,0 95,5 65 72,2 61,8 81,1 51 78,5 66,5 87,7 398 81,5

Numero di casi

Comprensorio

Figura 2. Andamento stagionale delle lesioni da animale in Alto Adige (n=460). Tabella VIII. Distribuzione delle lesioni per sesso della vittima. Sesso Maschi Femmine Totale

Numero di lesioni 224 236 460

% di lesioni 48,7 51,3 100,0

Lim Inf 95% C.I. 44,1 46,6

Lim Sup 95% C.I. 53,4 55,9

Tabella IX. Distribuzione del numero di lesioni per sesso della vittima e specie dell’animale coinvolto. Sesso Maschi Femmine Totale

Cane 199 191 390

Altra specie 25 45 70

Totale 224 236 460

Dati sull’individuo ferito Come evidenziato nella Tabella VIII, non c’è grande differenza fra la percentuale di individui feriti di sesso maschile e femminile. Se però si tiene conto della specie dell’animale morsicatore e del sesso dell’individuo aggredito è possibile evidenziare un’associazione significativa (si veda la Tabella IX). Se si considera il fattore sesso, l’odds ratio degli individui maschi di essere feriti da un cane è risultata di 1,87 (exact = 95%, C.I. = 1,08 - 3,32, χ2 = 5.57, P = 0.018). Gli individui di sesso femminile hanno mostrato una probabilità di essere ferite da un gatto circa due volte maggiore rispetto agli individui di sesso maschile (n = 35 vs. n = 17, OR = 2,12, exact = 95%, C.I. = 1,11 - 4,17, χ2 = 6.01, P = 0,0142). Nella maggior parte dei casi si è trattato di donne adulte. I casi di morsicatura da cane più gravi hanno coinvolto individui di sesso femminile. In 5 casi, tutti dovuti a morsicature da cani (1,4%), è stato necessario il ricovero e in 4 di questi l’individuo ospedalizzato è risultato di sesso femminile. L’età di tutti gli individui femminili ricoverati è risultata su-

periore ai 60 anni. L’età media di individui di sesso maschile ospedalizzati è risultata essere di 26 anni. In 24 casi (6,2%) l’individuo ferito è risultato residente fuori provincia. L’età media degli individui feriti è di 39,7 anni, vi sono stati casi sia tra bambini molto piccoli sia tra individui molto anziani (SD = 22.5350; Min 1 - Max 88; Median = 41; Mode = 9). La Figura 3 mostra che la classe di età ferita da cani con più frequenza è quella tra 5-9 anni, con leggera prevalenza delle bambine. Relativamente alle ferite determinate da gatti, la maggior parte dei casi ha riguardato classi di età oltre i 22 anni. Nella Figura 4 è riportato un tasso di incidenza riferito al 2010 per le lesioni da cane. La classe d’età con la percentuale più alta di individui traumatizzati da animali è stata quella di 5-9 anni con un tasso di incidenza pari a 129,6 casi per 100.000 abitanti. L’età degli individui feriti da cani è risulta significativamente diversa da quella degli individui feriti da gatti (test non parametrico di Wilcoxon-MannWhitney per il confronto fra mediane, mediana = 41 vs 50,5; P = 0,0015). Raggruppando ulteriormente le

Veterinaria Italiana 2013, 49 (1), 25-36

Morosetti et al.

Lesioni provocate da animali nella provincia di Bolzano

Maschi

Numero di casi di lesione

Femmine

16 14 12 10 8 6 4 2 0

0-4

5-9

10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79

80 e oltre

Classi di età

Figura 3. Frequenza assoluta delle lesioni da cane per classi di età e sesso della persona ferita (n=390).

140,0

129,6

120,0

95,4

Incidenza

100,0 80,0

66,8 68,1

88,2 64,5

60,0 40,0

70,6

77,2

50,6 36,7

68,1 70,6

68,1 49,9

50,9 53,0

34,5

20,0 0,0

0-4

5-9

10-14 15-19 20-24 25-29 30-34 34-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79

Classe di età

80 e oltre

Figura 4. Tassi di incidenza età-specifici delle lesioni da cane (per 100 mila ab.). Casi residenti (n=337). classi di età si ha la distribuzione dei casi riportata nella Tabella X. Si può osservare come, nonostante il numero assoluto di lesioni negli individui minori sia inferiore a quello degli individui adulti, l’incidenza dei primi sia più alta.

Analisi dei casi di lesione da cane Nel 2010, nella PAB, sono stati notificati n = 390 casi di lesioni provocate da cane con un’incidenza di 66,9 per 100.000 abitanti. È stato possibile verificare l’andamento nel tempo per il solo comprensorio di Bolzano (Figura 5), in cui negli ultimi 15 anni si è osservato un trend di diminuzione dell’incidenza annua (numero

Veterinaria Italiana 2013, 49 (1), 25-36

di nuovi casi di lesione su popolazione media del relativo anno). Il test non parametrico per trend su gruppi ordinati di Cuzick (1985) ha evidenziato una diminuzione significativa nel tempo dell’incidenza annua (z = -3,49; P > |z| = 0,000). In tre comprensori (Bolzano, Bressanone, Brunico) la proporzione dei cani morsicatori rispetto al numero di cani presenti nel relativo territorio è risultata simile, nel comprensorio di Merano è risultata pari a circa la metà.

Distribuzione geografica La Figura 6 mostra come le zone urbane, in particolare Bolzano e dintorni, abbiano la frequenza mag-

Morosetti et al.

Lesioni provocate da animali nella provincia di Bolzano

96,7

80,9 84

70,4

69,5

60 40 20

2009

2008

2007

2006

0 2005

61,2 65,7 74,9 66,9

105,9

101,5

2004

55 206 75 337

101,9

2003

14,6 63,3 22,1 100

111,6 112,6

100

2002

57 247 86 390

128,8 130,9 126,6

127,4

2001

Anziani (>65) Adulti Minori (<18) Totale

120

134,3

2000

Gruppo di età

Incidenza Lesioni Proporzione Lesioni in su 100 mila da cane % residenti ab (solo residenti)

140

1999

Tabella X. Distribuzione delle lesioni da cane per gruppo di età.

160

1998

I cani presenti in anagrafe all’inizio dello studio erano 32.236 (maschi = 17.462; Femmine = 14.774). Nei casi in cui il sesso del cane era noto (n = 292, 74,8%) erano coinvolti significativamente più cani maschi rispetto alle femmine: 215 maschi (73.6%; 95% C.I. 68.2%-78.6%) 77 femmine (26.4%; 95% C.I. 21.4%‑31.8%). In merito, è stata rilevata una forte

1997

Caratteristiche del cane

In 5 casi (1,4%) è stato necessario un ricovero dopo una morsicatura da cane. Le razze coinvolte sono: 1 rottweiler, 1 pastore scozzese, 2 meticci e uno di razza sconosciuta. In 4 di questi 5 casi è risultato che fossero cani di sesso maschile. L’età media dei cani (n = 256) è di 4,9 anni (SD = 3,396; Min 0 - Max 19; Median = 4).

1996

Dalla Tabella XI si evince che nel 41,7% dei casi di morsicatura il cane è risultato di proprietà del morsicato o aveva contatti regolari con l’individuo ferito.

1995

Relazione della persona ferita con il cane

evidenza di associazione tra sesso maschile del cane e maggiore probabilità di lesione all’individuo (OR = 2,38, 95% C.I. 1,82 - 3,12; χ2 MH = 44,95, P < 0,0001).

Incidenza annua

giore di casi. Soltanto in 18 casi (4,6%) il proprietario del cane è risultato residente fuori provincia.

Anno

Figura 5. Incidenza annua di lesioni da cane (per 100 mila abitanti) nel comprensorio sanitario di Bolzano (1995 – 2009).

Figura 6. Distribuzione geografica dei casi.

Veterinaria Italiana 2013, 49 (1), 25-36

Morosetti et al.

Lesioni provocate da animali nella provincia di Bolzano

Tipologia delle razze coinvolte

secondo i raggruppamenti di razze canine stabilite dall’ENCI con l’aggiunta del gruppo di meticci (comprendente le razze non riconosciute) e dei pit bull.

In 328 casi è stata riportata la razza del cane aggressore. Sono stati raggruppati i cani aggressori

Numericamente sono stati più frequenti i casi provocati da meticci 39,9% e cani da pastore 27,7% (Figura 7). Tuttavia rapportando il numero dei casi di morsicatura al numero di soggetti del relativo gruppo di razza registrati nell’anagrafe canina provinciale, le razze con l’indice di rischio più alto sono state: bassotto e, ancora una volta, cane da pastore e pit bull (Figura 8).

Tabella XI. Relazione tra il cane morsicatore e la persona ferita. Relazione con il Frequenza cane morsicatore Proprietario 24 Contatti regolari 77 Sconosciuto 141 Totale 242

% 9,9% 31,8% 58,3% 100,0%

Lim Inf 95% C.I. 6,5% 26,0% 51,8%

Lim Sup 95% C.I. 14,4% 38,1% 64,5%

Bassotti Cani da pastore Pitbull

Gruppo razze

Pinscher-Schnauzer-Molossi Meticci e Razze non ENCI Cani di tipo primitivo Cani da ferma Cani da riporto Terrier Segugi Cani da compagnia Levrieri 0

1,8

% sul totale dei casi di lesione

Figura 7. Frequenza relativa dei casi di lesione per gruppo di razza canina.

Bassotti Cani da pastore Pitbull

Gruppo razze

Pinscher-Schnauzer-Molossi Meticci e Razze non ENCI Cani di tipo primitivo Cani da ferma Cani da riporto Terrier Segugi Cani da compagnia Levrieri 0

0,2

0,4

0,6

0,8

1,2

1,4

1,6

Indice di rischio

(% lesioni dovute al gruppo razza/% presenza razza nella popolazione canina)

Figura 8. Indice di rischio di lesione relativo al gruppo di razza canina.

Veterinaria Italiana 2013, 49 (1), 25-36

Morosetti et al.

Lesioni provocate da animali nella provincia di Bolzano

Il contesto è stato descritto in 320 di 390 casi (82,1%). Più della metà dei casi in cui un cane ha ferito un individuo (54,6%) è avvenuta in luoghi pubblici o esterni alle abitazioni, nel 39,6% dei casi nell’abitazione del proprietario e nel 5,8% in quella dell’individuo morsicato (95% C.I. = 49,2% - 59,9%; 34,4% 45,0%; 3,7% - 8,9%, rispettivamente). Nella maggioranza degli episodi il cane ha morso per scopi difensivi. Frequenti sono stati i casi in cui il cane ha leggermente ferito l’individuo con cui giocava, camminava o correva, al suo davanti. Sono stati feriti anche individui intervenuti per separare cani che litigavano o che avevano in qualche modo infastidito l’animale durante la toelettatura, la medicazione oppure che avevano toccato l’animale in punti doloranti o sensibili.

Rapporto tra cane e individuo ferito Nelle Tabelle XII e XIII sono rappresentati i contesti in cui è avvenuta la lesione dell’individuo, tenendo conto del tipo di rapporto tra morsicato e cane. Un individuo conosciuto è risultato ferito con più frequenza, in genere, in modo lieve in un contesto di Tabella XII. Contesto del morso quando la persona era ben conosciuta al cane. Contesto

Frequenza

Alimentazione Altro Difesa Disturbo Gioco Lite Movimento Totale

9 4 16 19 29 17 6 100

9,0% 4,0% 16,0% 19,0% 29,0% 17,0% 6,0% 100,0%

Lim Inf 95% C.I. 4,2% 1,1% 9,4% 11,8% 20,4% 10,2% 2,2%

Lim Sup 95% C.I. 16,4% 9,9% 24,7% 28,1% 38,9% 25,8% 12,6%

Tabella XIII. Contesto del morso quando la persona era sconosciuta al cane.

Contesto

Frequenza

Alimentazione Altro Difesa Disturbo Gioco Lite Movimento Totale

4 8 46 25 14 16 26 139

2,9% 5,8% 33,1% 18,0% 10,1% 11,5% 18,7% 100,0%

Lim Inf 95% C.I. 0,8% 2,5% 25,4% 12,0% 5,6% 6,7% 12,6%

Lim Sup 95% C.I. 7,2% 11,0% 41,6% 25,4% 16,3% 18,0% 26,2%

gioco con l’animale. In vari casi è risultato che il cane ha reagito per il disturbo subito durante una manipolazione o ha ferito accidentalmente il proprio conduttore mentre questi cercava di separarlo da un altro cane durante un litigio. Un individuo sconosciuto è risultato ferito più spesso in un contesto di difesa del padrone, del territorio o della proprietà. Frequentemente è stato riferito un attacco di un passante in movimento oppure una reazione ad un tentativo di approccio da parte di individuo estraneo al cane.

Contesto della lesione a minori I minori (<18 anni) hanno mostrato una probabilità superiore di essere soggetti a una morsicatura associata al contesto di gioco rispetto agli adulti (OR = 3,2; χ2 MH 9,5; P = 0002; 95% C.I. = 1,5 - 6,2). Per quanto riguarda le lesioni ai minori (<18), nei bambini sotto i 14 anni di età è stato riscontrato un odds ratio 3,5 volte maggiore rispetto ai ragazzi più grandi. Si è anche notato che le lesioni avvengono con più probabilità dopo un’interazione iniziata dal bambino piuttosto che dal cane (P - Fischer exact = 0,01; 95% C.I. 1,3 - 9,3).

Concordanza tra le dichiarazioni degli individui morsicati e dei conduttori di cani In 212 episodi su 320 il contesto è stato riferito sia dall’individuo morsicato che dal conduttore del cane. Nel complesso si è potuto osservare una concordanza discreta tra i contesti descritti. (Landis and Koch (19): Agreement = 62,26%; Expected Agreement = 16,74%; Kappa = 0,5468; SD = 0,0301, Z = 18,19; P =0,0000) (Figura 9). L’osservazione in dettaglio dei vari contesti riportati ha permesso di stabilire che la concordanza migliore si è avuta nei casi di lite tra cani (Kappa = 0,816) e nel caso in cui il cane giocava (Kappa = 0,568) o sta-

Alimentazione

Contesto della morsicatura

Contesto

Gioco Lite tra cani Difesa Disturbo Movimento Altro 0

0,2

0,4

0,6

0,8

Indice di concordanza k di Cohen

Figura 9. Concordanza tra il contesto dell’attacco riferito da vittima e conduttore del cane.

Veterinaria Italiana 2013, 49 (1), 25-36

Morosetti et al.

Lesioni provocate da animali nella provincia di Bolzano

va ricevendo cibo (Kappa = 0,603). Scarso accordo è stato osservato in altri contesti, ad esempio in quello in cui l’individuo ferito riferiva una determinata interazione (es. manipolazione, carezza) con il cane mentre il proprietario dell’animale dava un’altra interpretazione dell’accaduto.

Aspetti veterinari Valutazione veterinaria del cane morsicatore Il cane è rimasto sconosciuto o non verificabile dal veterinario in 105 dei 390 casi (26,9%), (Figura 10). Per 231 casi è risultata disponibile la valutazione del veterinario ufficiale con la distribuzione indicata nella Tabella I. Gli episodi attribuibili a carenze nella gestione del cane da parte del proprietario sono stati 73 (31,6%).

Vaccinazione antirabbica Lo stato vaccinale del cane è risultato noto nel 64,1% delle lesioni (n = 250). In 16 di questi casi (6,4%) i cani non sono risultati soggetti a vaccinazione valida oppure non sono risultati vaccinati.

Discussione L’indagine riferita all’anno 2010 fa parte di una serie di attività promosse dal Servizio veterinario della ASL della PAB per comprendere meglio il fenomeno delle lesioni da animale, in particolare da cane, a livello locale. L’obiettivo principale è stato quello di riuscire ad elaborare strategie di prevenzione riguardanti danni alle persone e produrre indicazioni per una gestione

Frequenza casi

35 30

27,4

21,8

16,2

15 10 5 0

Comprensori sanitari

di Bolzano (BZ), Merano (ME), Bressanone (BX), Brunico (BR)

Figura 10. Proporzione dei casi con cane sconosciuto/irreperibile sul totale dei casi del singolo comprensorio.

Veterinaria Italiana 2013, 49 (1), 25-36

più adeguata degli animali da compagnia in un contesto di possibile rischio di trasmissione di rabbia, la cui presenza negli animali selvatici è riemersa nel territorio provinciale a partire dalla fine del 2009. Per quanto riguarda la completezza dei dati raccolti è possibile che siano sfuggiti casi di lesioni meno gravi, curate a domicilio oppure dal medico di base meno abituato a notificare alle autorità veterinarie. Vi sono stati anche alcuni casi, che hanno coinvolto persone o animali non residenti, in cui il follow up dell’animale non è stato possibile. Appare anche plausibile che casi di natura lieve che coinvolgano un cane proprio, di amici o parenti, non vengano riferiti al servizio sanitario. Non tutti i casi di accesso a cure mediche per lesioni da cani vengono in seguito notificati e lesioni molto banali sono curate a domicilio (14). In alcuni studi emerge che sono notificate meno del 50% delle morsicature da cani (18, 27). È verosimile che a favore della completezza delle notifiche nella presente indagine può aver influito la ricomparsa, come già detto, della rabbia silvestre, oltre alla campagna informativa alla popolazione sulla gravità della malattia e le relative misure di vaccinazione e controllo avviate tra il 2009-2010. Evenienze che hanno favorito gli individui con lesioni da animale a ricorrere, comunque, a cure mediche anche in casi lievi.

Impatto sui servizi sanitari aziendali A parte pochi casi gravi, le lesioni provocate da animali raramente rendono necessario un ricovero ospedaliero (27). Nella PAB il tasso di ospedalizzazione è stato più basso rispetto ad altri studi. In Florida (USA) è stato rilevato un tasso di ospedalizzazione dopo accesso ai reparti di emergenza-urgenza del 4,7% (28). In questa indagine solo per l’1,74% dei pazienti che hanno fatto ricorso ai servizi sanitari a causa di lesioni da animale è stato necessario il ricovero. Le differenti percentuali rilevate dai due studi possono trovare giustificazione nella diversa possibilità di accesso alle cure nelle due realtà studiate. Considerando solo i casi riferiti a cani, la frequenza di ricovero è risultata ancora più bassa (1,3%). Nonostante il ricorso al morso appartenga al normale repertorio comportamentale del cane, non sempre le lesioni vengono provocate da un morso, infatti nella gran parte degli episodi le persone si sono rivolte ai servizi sanitari per graffi o lividi (1, 6, 10, 23, 26, 27, 31), questo aspetto è stato osservato anche nel presente studio. In vari casi il cane aveva semplicemente giocato in maniera irruenta con la persona senza mordere, provocando ematomi o escoriazioni superficiali.

Morosetti et al.

Lesioni provocate da animali nella provincia di Bolzano

Non è stato possibile analizzare nel dettaglio la tipologia o gravità delle lesioni/traumi a causa della compilazione sporadica del modulo di notifica. Tuttavia nei pochi casi gravi in cui la documentazione è risultata completa, il paziente ha comunque avuto bisogno di cure lunghe e complesse. Nei 5 casi di attacco da cane con sequele importanti la degenza media è stata di 28 giorni. Il costo delle cure mediche dovute a morsicature da cane non è da sottovalutare, infatti oltre al pericolo legato alla rabbia, questi traumi sono frequentemente fonte di infezione (9); in moltissimi casi in aggiunta alle medicazioni il paziente ha ricevuto una profilassi antirabbica e antitetanica. Sono da considerare legati a questi tipo di incidente anche i costi indiretti riguardanti le assenze da lavoro, o scuola nel caso di minori (in media 8 giorni). Non sono poi valutabili i costi derivanti da traumi psicologici causati dalla reazione dell’animale.

Rilevanza delle lesioni da animale per gli accessi ai P.S. Dai dati rilevati emerge che lo 0,12% degli accessi ai reparti di emergenza-urgenza (P.S.) degli ospedali di Merano, Bressanone e Brunico hanno riguardato lesioni da animale. L’ospedale di Bolzano ha ricevuto il maggior numero di casi di lesioni da animale (n=211) che, in percentuale sul totale degli accessi al suo P.S. ospedaliero, rappresentano lo 0,25%. Dati simili sugli accessi ai P.S. sono stati osservati a Bologna (0,21%) (22) e a Lucerna (0,36%) (20). Come osservato anche in Piemonte (14), la maggior parte dei casi sono stati notificati alle autorità veterinarie dai P.S. e solo una piccola parte degli accessi ha riguardato i medici di medicina generale (MMG) o i servizi di igiene e sanità pubblica (SISP), anche se, come già accennato, è possibile che non tutti i casi di accesso ai servizi del territorio siano stati rilevati.

Conclusioni Sulla base dei risultati ottenuti e al fine di prevenire per quanto possibile casi di lesioni a persone da parte di animali, con particolare riferimento ai cani, si ritengono opportune le seguenti raccomandazioni. La sorveglianza del fenomeno è essenziale e il flusso informativo medico-veterinario deve essere migliorato e armonizzato. Un attenzione particolare deve essere rivolta alla messa in sicurezza delle interazioni tra cani e bambini. Appare essenziale che questi ultimi vengano educati e informati a rispettare gli spazi e le esigenze del cane e degli animali da compagnia in generale. Vista la facilità con cui alcune razze canine, originariamente selezionate per particolari compiti, riescono a diventare popolari e diffuse, appare di primaria importanza la consulenza di esperti sia nella scelta di razza e tipologia canina, appropriate al contesto famigliare e allo stile di vita, sia nella socializzazione del cucciolo. Molte lesioni si verificano per incapacità gestionale del proprietario o perché il cane è mantenuto in condizioni di frustrazione, se non addirittura di malessere. È necessario studiare metodologie di informazione/educazione che mettano i proprietari in grado di gestire correttamente l’animale.

Rispetto ad altre cause di accesso per infortunio ai P.S., le lesioni da animale sono tra i motivi meno frequenti rispetto ad altre tipologie di incidente (es. sportivo, domestico). Purtroppo questa informazione di dettaglio manca per il comprensorio sanitario di Bolzano a cui appartengono quasi la metà dei casi notificati.

Appare, inoltre, importante informare su pratiche illegali e rischi dovuti all’acquisto di cuccioli da venditori con pochi scrupoli. Questi animali possono sviluppare patologie comportamentali anche gravi derivanti dall’adozione troppo precoce e dall’insufficiente socializzazione intra e interspecifica nel periodo evolutivo.

Incidenza delle lesioni

Si sottolinea l’importanza di promuovere un allevamento cinofilo qualitativo con un’adeguata scelta dei riproduttori, non solo da un punto di vista fisicomorfologico, ma anche di equilibrio caratteriale.

La popolazione nella PAB è di 503.399 abitanti (ASTAT 2010). All’inizio del 2010 nella PAB sono risultati iscritti in anagrafe 32.236 cani (Servizio veterinario, Relazione aziendale annuale 2010, non pubblicata) con un rapporto uomo-cane 15,6 : 1. L’incidenza delle lesioni per l’anno 2010 è risultata

pari a 78,9 per 100.000 abitanti (contesto urbano e rurale), leggermente superiore a quella rilevata nel contesto urbano di Bologna pari a 58,4 per 100.000 ab., con un rapporto uomo-cane di 19 : 1 (22), ma inferiore a quella osservata in Svizzera nel 1996, in zone urbane e rurali, pari a 192,5 per 100.000 ab., con un rapporto uomo-cane di 16:1 (20), simile a quella rilevata nella PAB. Anche l’incidenza delle lesioni da gatto in questa indagine è risultata inferiore a quella rilevata in Svizzera: 9,9 contro 77,7 per 100.000 abitanti (20).

Si ribadisce la necessità di una formazione adeguata degli operatori sanitari e tecnici addetti al settore e di occasioni di interscambio di esperienze con un approccio integrato e multidisciplinare per garantire flussi informativi, interventi di profilassi vaccinale

Veterinaria Italiana 2013, 49 (1), 25-36

Morosetti et al.

umana e animale e adozione di tecniche di valutazione comportamentale dei cani. Nell’organizzare la presenza del personale addetto ai servizi di pronto soccorso e ai controlli veterinari degli animali morsicatori è opportuno tenere conto della stagionalità tipicamente estiva del fenomeno. Nelle zone con presenza di rabbia silvestre, si sottolinea l’importanza di sensibilizzare la popolazione a rivolgersi immediatamente al medico per acquisire l’offerta attiva della vaccinazione antirabbica e antitetanica.

Lesioni provocate da animali nella provincia di Bolzano

Ringraziamenti Il presente studio non sarebbe stato possibile senza la fattiva collaborazione, per la rilevazione e la comunicazione dei casi avvenuti nella PAB, di medici e personale infermieristico dei reparti ospedalieri di emergenza-urgenza e dei servizi di medicina territoriale. Gli autori desiderano ringraziare veterinari ufficiali e personale tecnico delle quattro sedi dei servizi veterinari ASL. Si ringraziano anche l’Istituto provinciale di statistica (ASTAT) e l’osservatorio epidemiologico provinciale per i dati cortesemente forniti.

Bibliografia 1. Archer J. 1988. The behavioural biology of aggression. Cambridge University Press, UK, 272 p. 2. ASTAT (Istituto provinciale di statistica) 2010. Statistisches Jahrbuch 2010/ Annuario Statistico 2010. Auswertung der Bevölkerungsregister der Gemeinden/ elaborazione dei registri anagrafici comunali, 107.

15. Fedele V., Gnaccarini M., Laurenti P., Marino M. & Meia B. 2009. Monitoraggio delle morsicature nel pinerolese negli anni 1998-2008 (II parte). Argomenti, XI(1), 61-66. 16. Gershman K.A., Sacks J.J. & Wright J.C. 1994. Which dogs bite? A case-control study of risk factors. Pediatrics, 93(6), 913-917.

3. Beaver B.V. 2009. Canine Social Behaviour In Canine Behaviour: Insights and Answers, 2nd Ed, St Louis, Saunders Elsevier, 148-150.

17. Hart B.L. 1997. Selecting, raising, and caring for dogs to avoid problem aggression. J Am Vet Med Assoc, 210, 1129-1134.

4. Beaver B.V. 1993. Canine aggression. Appl Anim Behav Sci, 37(1), 81-82.

18. Kahn A., Robert E., Piette D., De Keuster T., Lamoureux J. & Leveque A. 2004. Prevalence of dog bites in children. A telephone survey. Eur J Pediatr, 163(7), 424.

5. Beaver B.V. 1983. Clinical classification of canine aggression. Appl Anim Ethol, 10(1-2), 35-43. 6. Borchelt P.L. 1983. Aggressive behavior in dogs kept as companion animals: classification and influence by sex, reproductive status, and breed. Appl Anim Behav Sci, 10, 54-61. 7. Cattarossi D. & Martuzzi F. 2007 Cani mordaci in Italia: indagine sulle razze di appartenenza e considerazioni sulla normativa vigente. Veterinaria, 21(2), 19-29. 8. Chomel B.B. & Trotignon J. 1992. Epidemiologic surveys of dog and cat bites in the Lyon area, France. Eur J Epidemiol, 8(4), 619-624. 9. Cleveland S., Kaare M., Knobel D. & Laurenson M.K. 2006. Canine vaccination: providing broader benefits for disease control. Vet Microbiol, 117, 43-50. 10. Coppinger R.P. & Coppinger L. 2001. Dogs: a startling new understanding of canine origin, behaviour and evolution. Scribner N.Y. 352 p. 11. Dehasse J. 2002. Le chien aggressif. Publibook com éditions, Paris, 350 p. 12. De Keuster T. 2009. Epidemiology of dog bites. Proceedings 34th WSAVA Congress, Sao Paulo, Brazil, July, 21-24. 13. De Keuster T., Lamoureux J. & Kahn A. 2006. Epidemiology of dog bites: a Belgian experience of canine behaviour and public health concerns. Vet J, 172, 482-487. 14. Fedele V., Gnaccarini M., Laurenti P., Marino M. & Meia B. 2008. Monitoraggio delle morsicature nel pinerolese negli anni 1998-2008. Argomenti, X(4), 53-59.

Veterinaria Italiana 2013, 49 (1), 25-36

19. Landis J.R. & Koch G.G. 1977. The measurement of observer agreement for categorical data. Biometrics, 33, 159-174. 20. Matter H.C. 1998. The epidemiology of bite and scratch injuries by vertebrate animals in Switzerland. Eur J Epidemiol, 14(5), 483-490. 21. Notari L. & Goodwin D. 2007. A survey of behavioural characteristics of pure bred dogs in Italy. Appl Anim Behav Sci, 103(1-2), 118-130. 22. Ostanello F., Gherardi A., Caprioli A., La Placa L. Passini A. & Prosperi S. 2005. Incidence of injuries caused by dogs and cats treated in emergency departments in a major Italian city. Emerg Med J, 22, 260-262. 23. Overall K.L. & Love M. 2001. Dog bites to humans: demography, epidemiology and risk. J Am Vet Med Assoc, 218(1-2), 1923-1934. 24. Palacio J., Leon M., Gacia-Belenguer S. 2005. Aspectos epidemiologicos de las mordeduras caninas. Gac Sanit, 19(1), 50-58. 25. Patrick G.R. & O’Rourke K.M. 1998. Dog and cat bites: epidemiologic analyses suggest different prevention strategies. Public Health Rep, 113(3), 252-257. 26. Sacks J.J., Sinclair L., Gilchrist J., et. al. 2000. Breeds of dogs involved in fatal attacks in the United States between 1979 and 1998. J Am Vet Med Assoc, 217, 836-840. 27. Schalamon J., Ainoedhofer H., Singer G., Petnehazy T., Mayr J., Kiss K. & Höllwarth M.E. 2006. Analysis of dog bites in children who are younger than 17 years. Pediatrics, 117(3), e374-379.

Lesioni provocate da animali nella provincia di Bolzano

28. Shelton K. 2010. Dog bites: epidemiology and prevention. Florida Department of Health, Bureau of Epidemiology, Epi Update, May 2010, 1-5. (http://www. doh.state.fl.us/disease_ctrl/epi/Epi_Updates/2010_ index.html, ultimo accesso 22 marzo 2013).

Morosetti et al.

29. Svartberg K. & Forkman B., 2002. Personality traits in the domestic dog (Canis familiaris). Appl Anim Behav Sci, 79, 133-155. 30. Thompson P.G. 1997. The public health impact of dog attacks in a major Australian city. Med J, 167, 129-132.

Veterinaria Italiana 2013, 49 (1), 25-36

Lesions caused by animals in the Autonomous Province of South Tyrol in 2010: Fact-finding for prevention Giulia Morosetti1, Marica Toson2 & Christian Piffer1 Alto Adige Public Health Service – Veterinary Service, District of Bolzano, via Laura Conti, 4 - 39100 Bolzano, Italy giulia.morosetti@asbz.it 2 Regional Centre of Veterinary Epidemiology “Giovanni Vincenzi” Istituto Zooprofilattico Sperimentale delle Venezie, Viale dell’Università, 10 - 35020 Legnaro, Padua, Italy 1

Keywords Bites, Cat, Dog, Dog aggression, Epidemiological survey, Lesions, Risk factors, Veterinary public health.

Summary Lesions caused by animals, in particular by dogs, are a health issue to which public opinion often reacts sensitively. To effectively manage and prevent these events, it is therefore essential to evaluate the public health impact of this phenomenon and to identify the main connected risk factors. The aim of the present survey in the Autonomous Province of Bolzano was to collect various epidemiological variables helpful in understanding the problem at local level. The incidence and impact on Health Services of human lesions by several animal species for the year 2010 is presented, as well as a more detailed analysis of dog bites, giving a profile of the victims and of the animals involved. Different factors (geographical, contextual, seasonal and relational) that can be associated with episodes where dogs react aggressively to humans are illustrated. On the basis of the collected data, recommendations are given to prevent risk situations. Veterinaria Italiana 2013, 49 (1), 37-50

Introduction Incidents due to lesions caused by animals, above all those due to dog bites, represent an important public health problem, with costs both in human and health care terms because of the consequences associated with post-traumatic stress and trauma of different levels of severity in the various countries of the world. Understanding of the phenomenon, and of the factors connected to these episodes, is the first step to prevent and limit these events. Since 2009, in north east Italy and the countries bordering it, rabies has re-emerged with a significant economic impact on account of the measures that have to be taken to prevent the infection from spreading. The study of morbidity deriving from lesions caused by animals, in particular dogs, and the analysis of the site distribution of the events, can contribute to the assessment of the risks connected with lesions caused by animals and to the planning of targeted interventions for prevention. Given the absence of investigations into the phenomenon of animal lesions caused by animals in the Province of Bolzano, in 2010 a study was undertaken for the whole territory of the Autonomous Province of Bolzano (APB).

Materials and Methods Type of investigation An observational (cross-sectional) retrospective study of the data for medical- veterinary surveillance was carried out in the APB in 2010 in an area including the four Health Districts of the Alto Adige Public Health Service, corresponding to the territory of the APB.

Data collection As a source of data, the standardised report questionnaires compiled by the hospital emergency departments (HEDs) and the doctors of the province were used. These contain data for age, the sex of the persons injured, the site and context of the event and the type of animal involved. In addition, other standardised questionnaires were collected and analysed concerning the assessment of the dogs responsible for the bites by the veterinary officers of the Veterinary Service of the Public Health Service. To assess the context, the veterinary officers used a standardised method for quantifying the

Morosetti et al.

Lesions caused by animals, fact-finding for prevention

Table I. Relative frequency of attacking dog/owner type according to assessment of veterinary officer. Type 1. Balanced dog. Managed correctly: the reaction of the dog is understandable and can be justified by the context 2. Balanced dog: Management could be better, the owner needs to improve his/her skills 3. Balanced dog, however the owner is not able to manage the dog correctly 4. The owner is able to manage the dog, but the animal is scared/anxious/reactive 5. The owner needs to improve his/her skills and the dog is scared/anxious/reactive 6. The owner has no control over the dog and the animal is scared/anxious/reactive 7. Aggressive dog, however the owner has good control 8. Aggressive dog, the owner needs to improve his/her skills and management 9. Aggressive dog. The owner has no control over the animal. Retraining class is suggested. Total

Table II. Distribution of total number of lesions according to animal species. Animal species Dogs Cats Other animals Total

Number of lesions 390 52 18* 460

% of lesions 84.8 11.3 3.9 100.0

Lower lim Upper lim 95% C.I. 95% C.I. 81.1 87.9 8.6 14.6 2.4 6.3

Number of assessments 127 20 1 31 40 3 0 5 4 231

% of cases 55.0 8.7 0.4 13.4 17.3 1.3 0.0 2.2 1.7 100.0

Table III. Distribution of total number of lesions according to residence in South Tyrol. Residence of injured person Number of lesions % of lesions Resident in South Tyrol 397 86.3 Non-resident 63 13.7 Total 460 100.0

*Three snake bites and one bat bite among the lesions caused by other animals.

temperament and psychological balance of the dog, as well as the skill of its owner in managing it, with a pre-established points scheme that made it possible to place the cases in nine categories of possible dogmanager combinations (see also Table I ). The data for the resident human population was obtained from the last survey, ASTAT 2009 (2). Data for visits to the Public Health Service were kindly provided by the provincial epidemiological authority.

Definition of case All the episodes reported by the HEDs and by the doctors of the province concerning visits by a person for lesions caused by animals, regardless of the nature of the lesion (traumas, bites, scratches, bruises, etc.), between the 1st of January 2010 and the 31st of December 2010 in the APB were considered to be cases.

Statistical analysis A picture of the phenomenon of lesions caused by animals was given for the province by means of a descriptive analysis, and the data for the annual incidence of lesions caused by animals, including those for dogs, was obtained. The sitetime distribution of the events was analysed and described to assess the areas and population at

greatest risk. A profile was given of the typical characteristics of the victim, of the dog involved, and the context of the case. The rates calculated are rough or specific rates of cumulative incidence, calculated as the number of new cases involving residents of the Alto Adige in 2010 out of the number of inhabitants at the beginning of the year under consideration (healthy population at risk). To measure the strength of association between dichotomous variables, the odds ratio was calculated with an exact relative confidence interval of 95%, while the Chi-square test was used to evaluate the significance of the association. To make a comparison between the age distribution of the persons injured by dogs and those injured by cats, the non-parametric Wilcoxon-MannWhitney test was used to compare the medians of two independent samples. To establish the degree of agreement between victim and master about the real context of the bite, Cohenâ&#x20AC;&#x2122;s K-test was conducted. To assess the trend for annual incidence, Cuzickâ&#x20AC;&#x2122;s non-parametric test for trends on ordered groups (1985) was used. For some percentages, the relative confidence intervals (exact confidence intervals for binomial variables) was calculated, with the hypothesis that the cases observed are a representative sample of a more generic population. Finally, for some quantitative variables, a few indices of synthesis were given, such as average, median, minimum, maximum and standard deviation (SD).

Veterinaria Italiana 2013, 49 (1), 37-50

Morosetti et al.

Lesions caused by animals, fact-finding for prevention

Software used EPI Info 3.5.1; Stata 9; Microsoft Word, Excel; ESRI ArcMap.

Results

Table VII it can be seen that animal lesions account for a minimal number of visits to the Public Health Service hospitals in the Health Districts of Merano, Bressanone and Brunico compared to other causes of visit, with the exclusion of the Bolzano Health District, for which data are not available.

Tables II to IV summarise the situation for the phenomenon of lesions caused by animals in the Province of Bolzano. Dog lesions predominate over those caused by other animal species, and mainly involve residents.

Periods of greatest impact

Visits for medical treatment

Prognosis and severity of lesion

Almost half of the cases were treated by the Bolzano Health District (Figure 1). The highest number of dogs in the APB (n= 14,048, 41% of the dogs registered in the provincial dog register) is also registered in this district.

The prognosis was reported in 252 cases out of 460 visits (54.8%), with an average duration of recovery of 7.9 days (SD 5.2646, Min 1 - Max 56, Median 7).

The frequency of visits for medical treatment for lesions caused by animals is described in Tables V and VI. There were 398 visits to the HEDs (81.5%), while in 62 cases (13.5 %) the patient went to the general practitioner surgeries of the province. In the Merano Health District there was the highest proportion of visits to general practitioner surgeries compared to HEDs, followed by Brunico, while almost all the reports of lesions caused by animals in Bolzano and Bressanone come from their HEDs.

The peak of visits for medical treatment for lesions caused by animals was found to be in the summer months (Figure 1).

In only eight out of 460 cases (1.74%) was hospital admission necessary. Of these, 5 were due to lesions caused by dogs and three for viper bites. Dogs were the animals most commonly involved in severe lesion cases; unfortunately, a detailed description of the lesion was only occasionally reported, making in-depth analysis impossible.

BRU 14%

With regard to the importance of lesions caused by animals compared to other causes of accident, in BX 17%

Table IV. Distribution of number of lesions to residents by animal species. Animal species Dogs Cats Other animals Total (a)

Number of lesions 337 47 13 397

% of lesions 84.9 11.8 3.3 100.0

Incidence/100,000 residents (a) 66.9 9.9 2.6 78

Incidence calculated on population resident in South Tyrol in 2010 = 503,399 inhabitants

BZ 49% ME 20%

Figure 1. Distribution of animal lesion cases according to Health District.

Table V. Distribution by Health District and type of medical care visit. Health District Bolzano Bressanone Merano Brunico Total

Veterinaria Italiana 2013, 49 (1), 37-50

Number of visits to Hospital Emergency Dept (HED) 211 71 65 51 398

Number of visits to General Practitioner 15 8 25 14 62

Total

% of visits

226 79 90 65 460

49.1 17.2 19.6 14.1 100.0

% visits to District HED on total number of visits to HED 53.0 17.8 16.3 12.9 100.0

Morosetti et al.

Lesions caused by animals, fact-finding for prevention

Table VI. Distribution of visits to Hospital Emergency Departments by Health District.

72.2

61.8

81.1

Brunico

78.5

66.5

87.7

Totale

398

81.5

(a)

This percentage is calculated on the total number of medical care visits of each health district.

Number

Animal lesions

187

0.12%

Road accident

2,056

1.30%

Injury at workplace

6,936

4.50%

Domestic accident

11,691

7.50%

Violence by other person

551

0.40%

Self- harm, attempted suicide

0.00%

Sports accident

11,152

7.20%

Skiing accident

2,897

1.90%

Injury at school

1,347

0.90%

Other type of accident

12,886

8.30%

Other symptoms

106,156

68.20%

Total

155,717

100,00%

Characteristics of the injured person As shown in Table VIII, there is not a great difference between the percentage of male and female persons injured. If, however, the species of the animal involved and the sex of the victim is considered, a significant relationship with the sex of the person injured is revealed (Table I). Taking the factor of sex into consideration, the odds of being injured by a dog for men was found to be 1.87 (exact 95% C.I. 1.08 - 3.32, χ2 = 5.57, P = 0.018). Women were shown to be twice as likely than men to be injured by a cat (n = 35 vs. n = 17, OR = 2.12, exact 95% C.I. 1.11 ‑ 4.17, χ2 = 6.01, P = 0.0142); in most cases the victims were adult women. The most serious cases of dog bites involved women. In five cases, all due to dog bites (1.4%), hospital admission was necessary. In 4 of these the person admitted to hospital was a woman. The age of all the women injured and admitted to hospital was above 60, while the man admitted was 26. In 24 cases (6.2%), the person injured resided outside the province.

10 0

Month

Figure 2. Seasonal pattern of animal lesions in South Tyrol (n=460).

Table VII. Distribution by reason for visit to Hospital Emergency Departments, Bolzano Health District excluded. Reason for access

December

Merano

November

95.5

October

81.0

September

89.9

August

July

Bressanone

June

96.2

May

89.3

April

93.4

March

211

January

Bolzano

February

Number of % of visits Lower lim Upper lim visits to HED to HED (a) 95% C.I. 95% C.I.

Number of cases

Health District

Table VIII. Distribution of lesions according to gender of victim. Gender Male Female Total

Number of lesions 224 236 460

% of lesions 48.7 51.3 100.0

Lower lim Upper lim 95% C.I. 95% C.I. 44.1 53.4 46.6 55.9

Table IX. Distribution of number of lesions according to gender of victim and animal species. Gender Male Female Total

Dog 199 191 390

Other species 25 45 70

Total 224 236 460

On average, the injured person was 39.7 years old, but there were cases of both very young children and very old people (SD 22.5350, Min 1-Max 88, Median 41, Mode 9). Figure 3 shows that the age group most frequently injured by dogs was children from 5 to 9 years old, with a slightly greater prevalence of females. Figure 4 shows the rate of incidence of dog lesions for 2010. The age group with the highest percentage of individuals injured by animals is that of ages 5 to 9 with an incidence rate of 129.6 cases per 100,000 inhabitants. For cats, most cases concern age groups over 22. The age of persons injured by dogs is significantly different from that of persons injured by cats (the Wilcoxon-Mann-Whitney non-parametric test for comparison between medians, median = 41 vs 50.5, P = 0.0015). Further aggregation of the age groups gives the distribution of cases shown in Table IX. It can be seen that although the total number of lesions for minors is less than for adults, incidence in the former remains the highest.

Veterinaria Italiana 2013, 49 (1), 37-50

Morosetti et al.

Lesions caused by animals, fact-finding for prevention

Male

Female

Number of lesion cases

16 14 12 10 8 6 4 2 0

0-4

5-9

10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80 and over

Age class

Figure 3. Absolute frequency of dog lesions according to age group and gender of injured person (n=390).

140,0

129,6

120,0

95,4

Incidence

100,0 80,0

66,8 68,1

88,2 64,5

60,0 40,0

70,6

77,2

50,6 36,7

68,1 70,6

68,1 49,9

50,9 53,0

34,5

20,0 0,0

0-4

5-9

10-14 15-19 20-24 25-29 30-34 34-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80 and over

Age class

Figure 4. Age specific incidence of dog lesions (per 100,000 inhbts.). Resident cases (n=337).

Analyses of cases of dog lesions In 2010, 390 cases of lesions caused by dogs were reported in the APB, with an incidence of 66.9 per 100,000 inhabitants. It was possible to establish the trend over time only for the Health District of Bolzano (Figure 5), in which a diminishing trend in annual incidence in the last 15 years can be observed (number of cases of lesions out of the mean population for the relative year). Cuzick’s nonparametric test for trend on ordered groups (1985) shows a significant decline in annual incidence over time (z = -3.49, P > |z| = 0.000). In three Health Districts (Bolzano, Bressanone and Brunico) the proportion of ‘biting’ dogs compared to

Veterinaria Italiana 2013, 49 (1), 37-50

the number of dogs in the relative district is similar, while in the Health District of Merano it is almost half.

Geographical distribution Figure 6 shows that the urban zones, in particular Bolzano and its environs, have the greatest frequency of cases. In only 18 (4.6%) cases did the dog’s owner reside outside the province.

Relationship of the injured person to the dog In Table X it can be seen that in 41.7% of dog bite cases the dog was known by the victim and was

Morosetti et al.

Lesions caused by animals, fact-finding for prevention

105.9

101.5

96.7

80.9 84

70.4

69.5

60 40 20

2009

2008

2007

2006

2005

2004

0 2003

61.2 65.7 74.9 66.9

101.9

2002

55 206 75 337

111.6 112.6

100

2001

14.6 63.3 22.1 100

128.8 130.9 126.6

127.4

2000

57 247 86 390

120

134.3

1999

Elderly (>65) Adults Minors ( <18) Total

140

1998

Age group

Incidence Number Proportion Lesions in per 100,000 of dog % residents inhs. (only lesions residents)

160

1997

Table X. Distribution of dog lesions by age group.

The average age of the dogs in the cases for which the age was reported (n = 256) was 4.9 years (SD 3.396; Min 0 - Max 19; Median 4).

1996

There were 32,236 dogs (Males = 17,462, Females = 14,774) in the register at the beginning of the study. In the cases in which the dog’s sex was known (n = 292, 74.8%), significantly more male dogs than female dogs were involved: 215 males (73.6%; 95% C.I. 68.2%-78.6%), 77 females (26.4%; 95% C.I. 21.4%-31.8%). There was strong evidence of an association between the male sex of the dog and the greater possibility of a lesion to the person (OR 2.38, 95% C.I. 1.82-3.12, χ2 MH 44.95, P<0.0001).

1995

Characteristics of the dog

In 5 cases (1.4%) admission to hospital was necessary after a dog bite. The breeds involved were: 1 Rottweiler, 1 Collie, 2 mongrels and 1 of unknown breed. In 4 of these 5 cases the sex of the dog was known and was male.

Annual incidence

owned by the person bitten or had regular contact with the injured person.

Year

Figure 5. Annual incidence of dog lesions (per 100,000 inhs.) in the Health District of Bolzano (1995 – 2009).

Figure 6. Geographic distribution of cases.

Veterinaria Italiana 2013, 49 (1), 37-50

Morosetti et al.

Lesions caused by animals, fact-finding for prevention

Type of breeds involved

to the categorisations of dog breeds established by the ENCI (Italian Kennel Club), with the addition of the group of mongrels (including unknown breeds) and that of pit bulls.

In 328 cases the breed of the dog involved was reported. The dogs that bit were grouped according

Numerically the cases caused by mongrels (39.9%) and sheepdogs (27.7%) were more frequent (FigureÂ 7). However, when the number of bite cases are compared with the number of dogs of the same breed recorded in the provincial dog register, the breeds with the highest rate of risk are found to be the dachshunds and once again the sheepdogs, followed by the pit bulls (Figure 8).

Table XI. Relationship between attacking dog and injured person. Relationship with dog Owner Regular contacts Unknown Total

Number of cases 24 77 141 242

% 9.9% 31.8% 58.3% 100.0%

Lower lim Upper lim 95% C.I. 95% C.I. 6.5% 14.4% 26.0% 38.1% 51.8% 64.5%

Dachshund Sheepdogs

Dog Breed Groups

Pitbulls Pinscher-Schnauzer-Mastiffs Mongrels, Non ENCI Breeds Primitive, Nordic Type Pointers/Gun dogs Spaniels/Retrievers Terriers Hounds Small Companion dogs Greyhounds 0

1,8

% on total number of lesion cases

Figure 7. Relative frequency of lesion cases according to dog breed group.

Dachshund Sheepdogs

Dog Breed Groups

Pitbulls

Pinscher-Schnauzer-Mastiffs Mongrels, Non ENCI Breeds Primitive, Nordic Type Pointers/Gun dogs Spaniels/Retrievers Terriers

Hounds Small Companion dogs Greyhounds 0

0,2

0,4

0,6

0,8

1,2

1,4

1,6

Risk Index

(% lesion caused by the breed group/% presence of the breed group in the total dog population)

Figure 8. Lesion risk index of the different dog breed groups.

Veterinaria Italiana 2013, 49 (1), 37-50

Morosetti et al.

Lesions caused by animals, fact-finding for prevention

Context The context was described in 320 out of 390 cases (82.1%). More than half of the cases in which a dog injured a person (54.6%) happened in public places or outside homes, in 39.6% of cases in the owner’s home and in 5.8% in that of the person bitten (95% C.I. 49.2%-59.9%; 34.4%-45.0%; 3.7%-8.9% respectively). In most of the episodes the dog bit for reasons of self-defence and there were frequent cases in which the dog slightly injured the person who was playing with it or attacked a person walking or running past the dog. There were also cases of people being injured when intervening to separate two dogs fighting each other or who in some way irritated the dog during grooming, medication or by touching it in painful or sensitive places.

Relationship between the dog and the injured person Tables XI and XII show the contexts in which the lesion to the person occurred, taking into account the type of relationship between the victim and the dog. A known person was more frequently injured, and

generally slightly injured, while playing with the dog. In various cases the dog had reacted against being disturbed while being handled or had accidently injured the person in charge of it while he/she was trying to separate it from another dog during a fight. An unknown person was more often injured in the context of the dog defending its territory, its owner, or the owner’s home or possessions. Frequently cases were reported of an attack on persons walking or running past or in reaction to an attempt by a stranger to approach the dog.

Context of lesions to minors With minors (<18) there is a greater probability of the bite being associated with the context of playing than there is with adults (OR = 3,2; χ2 MH 9.5; P = 0.002; 95% C.I. 1.5-6.2). With regard to lesions to minors (<18), in children under the age of 14 there is an odds ratio (OR) of a lesion occurring after an interaction begun by the child rather than the dog of 3.5 times higher than there is with older young people (P [Fischer exact] = 0.01; 95% C.I. 1.3-9.3).

Agreement between the declarations of the victims and the dog managers

Context Food Other Defence Disturbance Play Dog fight Movement Total

Frequency of cases 9 4 16 19 29 17 6 100

% 9.0% 4.0% 16.0% 19.0% 29.0% 17.0% 6.0% 100.0%

Lower lim Upper lim 95% C.I. 95% C.I. 4.2% 16.4% 1.1% 9.9% 9.4% 24.7% 11.8% 28.1% 20.4% 38.9% 10.2% 25.8% 2.2% 12.6%

Table XIII. Context of attack when injured person was not known to the dog. Context Food Other Defence Disturbance Play Dog fight Movement Total

Frequency of cases 4 8 46 25 14 16 26 139

% 2.9% 5.8% 33.1% 18.0% 10.1% 11.5% 18.7% 100.0%

Lower lim Upper lim 95% C.I. 95% C.I. 0.8% 7.2% 2.5% 11.0% 25.4% 41.6% 12.0% 25.4% 5.6% 16.3% 6.7% 18.0% 12.6% 26.2%

In 212 episodes out of 320 the context was reported by both the person bitten and the dog manager. Generally speaking, quite good agreement can be observed on the basis of the interpretation of Landis and Koch (19) between the context described by the two persons involved (Landis and Koch: Agreement = 62.26%, Expected Agreement = 16.74%, Kappa = 0.5468; SD = 0.0301, Z = 18.19, P = 0.0000) (Figure 9). If the various contexts reported are examined in detail, the greatest agreement between the two was in the cases of fighting between dogs

Feeding

Context of attack

Table XII. Context of attack when injured person was well-known to the dog.

Play Dog fight Defense Disturb Movement NA 0

0.2

0.4

0.6

0.8

Overall Kappa statistic k=0.5468

Figure 9. Agreement between injured person and dog owner on context of attack.

Veterinaria Italiana 2013, 49 (1), 37-50

Morosetti et al.

Lesions caused by animals, fact-finding for prevention

(Kappa = 0.816) and in the case in which the dog played (Kappa = 0.568) or was receiving food (Kappa = 0.603). Little agreement is found in other contexts, in which the injured person reported a certain type of interaction (for example, handling, stroking) with the dog, while the dog owner gave a different interpretation of what happened.

Veterinary aspects Veterinary assessment of the dog responsible for biting The dog remained unknown/unascertainable by the veterinary officer in 105 of the 390 cases (26.9%), (Figure 10). For 231 cases, the veterinary officer’s assessment was available with the distribution shown in Table XIII. 73 (31.6%) of the episodes could be attributed to poor management of the dog by its owner.

Anti-rabies vaccination The dog’s vaccination status was known in 64.1% of the lesion cases (n= 250). In 16 of these cases (6.4%) the dogs did not have a valid vaccination or had not been vaccinated.

Discussion This survey refers to the whole of 2010 and is part of a series of activities started by the Veterinary Service of the APB to better understand the phenomenon of lesions from animals, in particular from dogs, at

Frequency of cases

35 30

27,4

21,8

16,2

15 10 5 0

Health District

Bolzano (BZ), Merano (ME), Bressanone (BX), Brunico (BR)

Figure 10. Proportion of cases where attacking dog remained unknown or untraceable in each health district.

Veterinaria Italiana 2013, 49 (1), 37-50

a local level. The main aim was to prepare strategies for the prevention of damage to persons and to produce recommendations for better management of animals kept as companion pets in the context of the possible risk of transmission of rabies, whose presence among wild animals had re-emerged in the province since the end of 2009. With regard to the completeness of the data collected, it is possible that cases of less severe lesions treated at home or by the general practitioner, less used to notifying the veterinary authorities according to fixed procedures, have escaped inclusion. There were also some cases involving non-resident persons or animals in which it was not possible to follow up the animal. It also seems plausible that minor cases involving one’s own dog or one belonging to friends or relations do not get reported to the Health Service. Also according to Fedele et al. (14) not all cases of visits for medical treatment of dog lesions are reported and very slight lesions are treated at home; according to Kahn et al. (18) and Schalamon et al. (27), fewer than 50% of dog bites are reported. On the other hand, a factor contributing to the completeness of the reports in this survey may have been the reappearance of wild animal rabies in the province since 2009 and the campaign launched between 2009-2010 to inform the population about the severity of the disease and the measures for vaccination and control. These may have induced persons with animal lesions to seek medical treatment, even in minor cases.

Impact on the Health Service agencies Apart from a small number of serious cases, lesions caused by animals rarely make hospital admission necessary (27). In the APB the rate of hospitalisation was less than in other studies. Shelton (28) found a hospitalisation rate after a visit to the HED in Florida (USA) of 4.7%, while in this survey admission was only necessary for 1.74% of patients who visited the Health Service because of lesions caused by animals. Naturally the differences may also be due to the different possibilities for access to treatment in the two territories studied. Moreover, if only the cases concerning dogs are considered, frequency of admission is even rarer (1.3%). Although biting is one of the normal forms of a dog’s behaviour, lesions are not always caused by a bite. In most episodes, people visited the Health Service for scratches or bruises (1, 6, 10, 23, 26, 27, 31) and this was observed also in this study. In various cases the dog had simply played with the person in a rather unrestrained way without biting, causing haematomas or superficial grazes. It was not possible to analyse the type or severity of the lesions/traumas because of the sporadic

Morosetti et al.

Lesions caused by animals, fact-finding for prevention

compilation of the medical diagnoses on the report form. However, in the few serious cases that occurred for which documentation was complete, the patient needed long, complex treatment: in the 5 cases of dog attacks with severe consequences, average hospitalisation was 28 days. The cost of medical treatment due to dog bites should not be underestimated. In addition to the danger associated with rabies, these traumas are frequently a source of infection (9). Furthermore, in a great number of cases, as well as treatment, the patient was given a prophylaxis against rabies and tetanus. The indirect costs connected to these kinds of incident must also be considered. They concern time off work, or off school in the case of minors (on average eight days). The costs deriving from psychological trauma caused by reaction to the dog cannot be estimated.

Proportion of animal lesions in visits to hospital emergency departments From the data collected, it emerges that between 0.15 and 0.3% of visits to the HEDs of Bolzano, Merano, Bressanone and Brunico were for lesions caused by animals. Similar data for visits to HEDs was noted by Ostanello et al. (22) in the city of Bologna (0.21%) and by Matter (20) for the pediatric hospital in Lucerne (0.36%). The hospital of Bolzano receives the largest number of cases of lesions from animals, which in percentage terms make up 0.25% of the total number of visits to the HED. As also observed in Piedmont by Fedele et al. (14), most cases were reported to the veterinary authorities of the HEDs and only a small proportion of the visits concerned general practitioners or the Public Health and Hygiene Services, although, as mentioned above, it is possible that not all the visits to the services of the province were reported. In comparison with other causes of visits due to accidents to the Public Health Service, animal lesions are among the least frequent reasons when compared to other types of accident (e.g. sports accidents, domestic accidents). Unfortunately detailed information of this type is lacking for the District of Bolzano, to which almost half the reported cases belong.

Incidence of lesions The population of the APB is 503,399 inhabitants (ASTAT 2010). At the beginning of 2010, in the APB 32,236 dogs were registered in the dog registry (2010 Annual Report of the Veterinary Service of the Public

Health Service, unpublished). This corresponds to a man/dog ratio of 15.6:1. The incidence of lesions for 2010 was 78.9 per 100,000 inhabitants (urban and rural context combined), slightly higher than that found by Ostanello et al. (22) in the urban context of Bologna, which was 58.4 per 100,000 inhabitants, with a man/dog ratio of 19:1, but much lower than that observed by Matter (19) in Switzerland in 1996, in urban and rural areas, of 192.5 per 100,000 inhabitants, with a man/dog ratio of 16:1, similar to the APB. The incidence of lesions caused by cats in this survey was also much lower than that found in Switzerland by Matter (19) of 9.9 as opposed to 77.7 per 100,000 inhabitants.

Risk factors There are limitations in the design of the study on account of the lack of checks to establish the association between bites and various factors. However, for some variables a more detailed analysis was possible and some results were found to be in line with the findings of other studies.

Age and sex of the injured person With regard to the age of the person injured by a dog, the victims most frequently involved are children under the age of 10. This datum is confirmed in various studies (14, 19, 27). According to Ostanello etÂ al. (22), the age group most affected, that is, children between the ages of 0 and 9, is also the one at greatest risk from bites to the face. In the APB, in severe cases of lesions caused by dogs and in injuries caused by cats, those most frequently involved are elderly patients. Many authors agree on the fact that the victims who suffer severe injuries are most often children and women, above all elderly women (1, 6, 10, 23, 26, 30). The average age of those injured by a dog is lower than that of those injured by a cat both in this study (39 vs. 49 years of age [ANOVA, P=0.001]) and in that of Ostanello et al. (22). These authors report that the age group with the highest incidence of dog lesions was ages 20-29, while for cat lesions it was 60-69. Shelton (28) noted that injuries caused by cats were considerably more frequent in adult women and minors under the age of 20. It is possible that the source of data influences the datum for the relative age of the victim. In fact, it is likely that children are taken to the doctor in all cases while adults are more inclined to treat themselves at home in minor cases. In this survey, keeping in mind the confounding effect of age, there seems to be an association between the sex of the person and the animal

Veterinaria Italiana 2013, 49 (1), 37-50

Morosetti et al.

responsible for biting: compared to the female sex, men appear at greater risk from lesions caused by dogs, while women are at greater risk from lesions caused by cats. This observation is in line with the findings of Ostanello et al. (22), who describe clear differences among the victims of lesions caused by animals: a greater number of male patients are injured by dogs, while injuries caused by cats more frequently concern women patients. Shelton (28), on the other hand, noted a generally higher frequency of lesions among women, with an incidence per 100,000 inhabitants of visits to the Public Health Service for bites of 103.6 for women compared to 85.5 for men. These results could depend on the fact that in the APB, the cat is usually kept as a companion animal by adult women and elderly women. The control and management of dogs, on the other hand, is often delegated to male members of the family, especially in the case of dogs, which are more difficult to look after.

Seasonality As found by other authors (14, 22, 28), the frequency of lesions peaks in summer when in the APB there is more tourism and more outdoor recreational activity, and during which dogs are probably taken out more frequently and for longer periods, with a greater chance of interacting with people and animals.

Geography In the 7,400 km2 territory of the APB, the greatest concentration of cases of lesion occurs in the large population centres (the communes of Bolzano, Appiano, Renon, Merano, Bressanone, Brunico and Vipiteno). The problem is therefore primarily one of public urban medical-veterinary health care. The provincial capital Bolzano, a densely inhabited urban area with 104,029 inhabitants and 4,550 dogs, is the commune with the greatest frequency of lesions caused by dogs (19% of the total) and a man/dog ratio of 23:1. It is probable that in urban environments, where the probability of man-dog interactions is more frequent and the areas of cohabitation smaller, people may run a greater risk of being injured by animals than in a rural environment (22).

Typology of dog In the APB the stray dog phenomenon is absent and we consider the data for the typology of the dog population registered in the present-day provincial electronic dog registry system to be sufficiently complete and representative of the local dog population.

Veterinaria Italiana 2013, 49 (1), 37-50

Lesions caused by animals, fact-finding for prevention

There are indications that some breeds tend to be more dangerous on account of their genetic characteristics or because of the type of attack (11, 21, 29), in particular the sheepdog group (16): Alsatians, maremma sheepdogs and Belgian sheepdogs (7). In the APB, mongrels are the most numerous type of dog, so it is not surprising that they are the dog group by far the most frequently involved in cases of lesions to persons. However, when the frequency of attacks by the various groups of dogs was compared with the size of that group in the dog register, other groups of breeds, not the mongrels, were found to be of greater importance. These certainly included some groups already considered aggressive by public opinion, such as sheep dogs and the pinscher-schnauzer-mastiff group, but also breeds not generally considered dangerous, such as dachshunds. When only the frequency and not the severity of the lesion are considered, in this survey the latter group occupies first place as an index of risk from biting. With regard to dachshunds, it is not surprising that the high risk index of 1.8 found in this survey is similar to that found by Cattarossi & Martuzzi (7) who put the dachshund category, with a risk index of 1.47, in sixth place in their classification. In fact, as these authors note, until a few generations ago this breed, in common also with terriers, was used to hunt into their lair combative prey that the dog had to attack and possibly kill without human intervention. They are therefore dogs that, as Notari & Goodwin (21) observe, are of medium-high aggressiveness, high reactivity and so potentially more inclined to bite than other breeds. In the APB, dachshunds are more popular and numerous than terriers. Perhaps for this reason, the terrier group, although including highly reactive breeds comparable to the dachshunds, were found to be less frequently involved in the cases of lesions reported. These data support the correctness of the decision to abolish the regulations whose provisions included a list of so-called dangerous dogs. Although in some breeds selective breeding may have aimed to produce individual dogs with strong reactivity, sense of territory, possessiveness, tenacity of bite and a general tendency to react to stimuli with offensive-defensive attacks (20), the concept of the dangerousness of breeds will certainly continue to be the subject of extensive debate, as the variables to be taken into account are not only the genetic and behavioural characteristics of the dog but many other factors, including the individual characteristics of each dog-manager pair. As De Keuster et al. (12) also underlined, the tendency to bite depends not just on genetics, but also on various factors

Morosetti et al.

Lesions caused by animals, fact-finding for prevention

connected to socialisation, education, physical and psychological health, the behaviour of the victim and, according to the authors of this survey, also that of the owner. Lesions from dogs must therefore always be considered in the human socio-cultural context in which they occur.

Sex and age of the dog Other variables that must certainly be taken into consideration in assessing the risk of biting are the animal’s sex and age. Beaver (3) emphasises that nonneutered male dogs are involved considerably more frequently in problems arising from aggressiveness, comprising more than half of all cases described. This is also confirmed by Gershman et al. (16). Moreover, the dogs involved are generally young adults (4, 5). In this survey, male adult dogs of about four years of age are involved considerably more often in cases of lesions to persons than female dogs. Hence the importance of paying particular attention to the socialisation and education of this category during the period of its development, in particular during puberty (6-18 months), especially in the case of large dogs and/or dogs with high levels of arousal. It is precisely during adolescence that male dogs may challenge the hierarchy within the family-pack, adopt forms of behaviour for the defence of territory, and indulge in playful activity with strong physical interaction and competition with other males (3).

Place Another influential factor is the place in which the event occurs. Over half the events took place in public places. Various events occurred near or inside owners’ homes, but not inside the victims’ homes. It was noticed that dogs behave aggressively much more often in the homes of their own families, in their gardens or yards, or in the immediate neighbourhood (1, 6, 10, 23, 26, 27, 30). For De Keuster (12), a small number of cases (9%) occur in public parks; most of these incidents occur because the dog was not adequately controlled (51%) or wandered unattended near its own home (31%), whereas more than half of dog bites happened while it was accompanied on walks along streets and roads and 30% on the private property of the family where the dog lives. This means various cases of lesion could be avoided by taking into account the territorial and defensive behaviour of the dog, making sure it cannot cross the boundaries of its owner’s property, and preventing it entering areas people pass through or that give access to public places, or else by ensuring that the dog is adequately supervised in areas in which it is left off the lead.

Context Interpretation of the context in which the dog’s reaction occurred varies according to whether it is reported by the person who was the victim of the attack or the dog’s manager. The dog’s owner and the injured person often do not agree on how the event happened. The closest agreement is in cases of lesions due to the intervention of a person to separate two dogs fighting each other, where it is clear what happened. In many other cases that would seem to be the result of wrong or deficient interpretation of possible warning signals from the dog, descriptions often do not match: the injured person states that he/she wanted to stroke the dog while the dog’s manager believes the person disturbed the dog. In particular, the signs of aggressiveness of a frightened dog may be less visible and not interpreted correctly by the owner. This is even more likely if the person is a stranger to the dog (3). Moreover, the context of the event would seem to differ according to whether the person is familiar with or unknown to the dog. In the former case, there is a prevalence of cases of lesion during play or as a result of the dog’s reaction to being handled (e.g. grooming of its coat), while in the latter case the most frequent cases of attack regard actions taken to defend its territory or attacks by an unattended dog on persons walking or running past. As well as the differences connected with the dog’s familiarity with the injured person, the question of the latter’s age plays an important role. In various studies, and likewise in this survey, children are clearly found to be at greater risk than adults. Although for Gershman et al. (16) most bites seem to be inflicted on people who are strangers to the dog’s family, for Schalamon et al. (27) most attacks by dogs on children are by dogs known to them (73%), but not living with them, caused by the child disturbing the dog. According to De Keuster (12), 86% of bites to children occur as a result of an interaction initiated by the children. Various authors have noted that children are bitten by a dog they know either during play or when the dog is disturbed (1, 6, 10, 23, 26, 27, 30). Children and adolescents must therefore be taught to correctly interpret particular signs, postures and sounds of the animal, and to adopt safe forms of behaviour to avoid negative interactions. This should be done under the constant supervision of an adult, with the main objective of guaranteeing the wellbeing of both dog and child. In Australia, programmes in schools teaching children how to behave safely with dogs have significantly increased the adoption of protective forms of behaviour by the children involved in

Veterinaria Italiana 2013, 49 (1), 37-50

Morosetti et al.

preventive activities compared to groups not targeted by such activities (30). The decision to bring a dog into the family must also be carefully considered and must not be made just to satisfy the incessant requests of the children. In any case, the potential benefits connected with the presence of a dog on the psycho-emotional development of the child must be set against its possible costs, including the risk of lesions. Schalamon et al. (27) suggest, for example, waiting until school age is reached before admitting a dog into the family.

Assessment of the dangerousness of dogs In a territory where wild animal rabies is present, the observation of dogs is clearly of primary importance for keeping this pathology under surveillance. However, in future, veterinary assessment must increasingly take into account aspects that are not strictly clinical but are connected with the behavioural equilibrium of the dog and the way it is managed by the person in charge of it. In 2010, for the first time an assessment form for use by veterinary officers was introduced in the APB. To create a data-collection tool that was simple to use in day-to-day activities, it was necessary to strike a compromise between the need to simplify the procedure as much as possible and at the same time to objectively collect information that was useful for assessing the animal responsible for biting. Although there are still many difficulties in making a satisfactory assessment of behavior and context, data seem to indicate that most lesions occur not so much because of the dogâ&#x20AC;&#x2122;s behavioural problems but because of deficiencies in the training/management of the animal by its owner.

Conclusions On the basis of the results obtained by this survey and in order to prevent, as far as possible, cases of lesions to people by animals, in particular dogs, the following recommendations are considered necessary: Surveillance of the phenomenon is essential and the flow of information between doctor and veterinary officer must be improved and coordinated. Particular attention must be given to making interactions between children and dogs safe. It seems essential for the latter to be taught by adults, themselves well-informed, to respect the space and needs of the dog and of companion animals in general. Given how easy it is for some dog breeds, originally selected for utilitarian purposes, to succeed in becoming popular and widespread, it would seem

Veterinaria Italiana 2013, 49 (1), 37-50

Lesions caused by animals, fact-finding for prevention

to be of primary importance to consult experts when choosing the breed and type of dog. This choice should be suited to the family context and to the puppyâ&#x20AC;&#x2122;s style of life and socialisation. Many lesions occur because of the ownerâ&#x20AC;&#x2122;s poor management of the dog or because the dog is kept in conditions that cause it frustration or even make it feel unwell. It is necessary to study ways of informing/educating that enable owners to manage their dogs correctly. It would also seem important to inform the public about illegal practices and the risks of buying puppies from unscrupulous sellers. These animals can develop behavioural pathologies, often serious ones, deriving from adoption too early and from insufficient intra- and interspecific socialisation in their period of development. It is necessary to underline the importance of promoting high quality dog breeding, with a suitable choice of breeding partners, not only from a physical-morphological point of view, but also from one of balance of character. In addition, the need must be stressed for adequate training for health care workers and those working in the dog training sector, and for opportunities for the exchange of experiences acquired in this field, with an integrated multidisciplinary approach to guarantee the flow of information, suitable treatment with prophylactic human and animal vaccination in the case of lesions, and the adoption of techniques for the assessment of dog behaviour. In organising the presence of staff for accident and HEDs and for veterinary controls on animals responsible for bites, the seasonal nature of the phenomenon, typically associated with summer, should be kept in mind. In areas in which wild animal rabies is present, it is necessary to stress the importance of raising awareness among the population about going to the doctor and of ensuring the active availability of anti-rabies and anti-tetanus vaccinations.

Acknowledgements This study would not have been possible without the active collaboration in the recording and communication of the cases occurring in the APB of the doctors and nursing staff of the HEDs, of the provincial general practitioners, and of the veterinary officials and technical staff of the four centres of the veterinary services of the Public Health Service. The authors would like to thank all of the above. They would also like to thank the Provincial Institute of Statistics (ASTAT) and the provincial epidemiological observatory for the data they have kindly provided.

Lesions caused by animals, fact-finding for prevention

Morosetti et al.

References 1. Archer J. 1988. The behavioural biology of aggression. Cambridge University Press, UK, 272 p. 2. ASTAT (Istituto provinciale di statistica) 2010. Statistisches Jahrbuch 2010/ Annuario Statistico 2010. Auswertung der Bevölkerungsregister der Gemeinden/ elaborazione dei registri anagrafici comunali, 107. 3. Beaver B.V. 2009. Canine Social Behaviour In Canine Behaviour: Insights and Answers, 2nd Ed, St Louis, Saunders Elsevier, 148-150. 4. Beaver B.V. 1993. Canine aggression. Appl Anim Behav Sci, 37(1), 81-82. 5. Beaver B.V. 1983. Clinical classification of canine aggression. Appl Anim Ethol, 10(1-2), 35-43. 6. Borchelt P.L. 1983. Aggressive behavior in dogs kept as companion animals: classification and influence by sex, reproductive status, and breed. Appl Anim Behav Sci, 10, 54-61. 7. Cattarossi D. & Martuzzi F. 2007 Cani mordaci in Italia: indagine sulle razze di appartenenza e considerazioni sulla normativa vigente. Veterinaria, 21(2), 19-29. 8. Chomel B.B. & Trotignon J. 1992. Epidemiologic surveys of dog and cat bites in the Lyon area, France. Eur J Epidemiol, 8(4), 619-624. 9. Cleveland S., Kaare M., Knobel D. & Laurenson M.K. 2006. Canine vaccination: providing broader benefits for disease control. Vet Microbiol, 117, 43-50. 10. Coppinger R.P. & Coppinger L. 2001. Dogs: a startling new understanding of canine origin, behaviour and evolution. Scribner N.Y. 352 p. 11. Dehasse J. 2002. Le chien aggressif. Publibook com éditions, Paris, 350 p. 12. De Keuster T. 2009. Epidemiology of dog bites. Proceedings 34th WSAVA Congress, Sao Paulo, Brazil, July, 21-24. 13. De Keuster T., Lamoureux J. & Kahn A. 2006. Epidemiology of dog bites: a Belgian experience of canine behaviour and public health concerns. Vet J, 172, 482-487. 14. Fedele V., Gnaccarini M., Laurenti P., Marino M. & Meia B. 2008. Monitoraggio delle morsicature nel pinerolese negli anni 1998-2008. Argomenti, X(4), 53-59. 15. Fedele V., Gnaccarini M., Laurenti P., Marino M. & Meia B. 2009. Monitoraggio delle morsicature nel pinerolese negli anni 1998-2008 (II parte). Argomenti, XI(1), 61-66. 16. Gershman K.A., Sacks J.J. & Wright J.C. 1994. Which dogs bite? A case-control study of risk factors. Pediatrics, 93(6), 913-917.

17. Hart B.L. 1997. Selecting, raising, and caring for dogs to avoid problem aggression. J Am Vet Med Assoc, 210, 1129-1134. 18. Kahn A., Robert E., Piette D., De Keuster T., Lamoureux J. & Leveque A. 2004. Prevalence of dog bites in children. A telephone survey. Eur J Pediatr, 163(7), 424. 19. Landis J.R. & Koch G.G. 1977. The measurement of observer agreement for categorical data. Biometrics, 33, 159-174. 20. Matter H.C. 1998. The epidemiology of bite and scratch injuries by vertebrate animals in Switzerland. Eur J Epidemiol, 14(5), 483-490. 21. Notari L. & Goodwin D. 2007. A survey of behavioural characteristics of pure bred dogs in Italy. Appl Anim Behav Sci, 103(1-2), 118-130. 22. Ostanello F., Gherardi A., Caprioli A., La Placa L. Passini A. & Prosperi S. 2005. Incidence of injuries caused by dogs and cats treated in emergency departments in a major Italian city. Emerg Med J, 22, 260-262. 23. Overall K.L. & Love M. 2001. Dog bites to humans: demography, epidemiology and risk. J Am Vet Med Assoc, 218(1-2), 1923-1934. 24. Palacio J., Leon M., Gacia-Belenguer S. 2005. Aspectos epidemiologicos de las mordeduras caninas. Gac Sanit, 19(1), 50-58. 25. Patrick G.R. & O’Rourke K.M. 1998. Dog and cat bites: epidemiologic analyses suggest different prevention strategies. Public Health Rep, 113(3), 252-257. 26. Sacks J.J., Sinclair L., Gilchrist J., et. al. 2000. Breeds of dogs involved in fatal attacks in the United States between 1979 and 1998. J Am Vet Med Assoc, 217, 836-840. 27. Schalamon J., Ainoedhofer H., Singer G., Petnehazy T., Mayr J., Kiss K. & Höllwarth M.E. 2006. Analysis of dog bites in children who are younger than 17 years. Pediatrics, 117(3), e374-379. 28. Shelton K. 2010. Dog bites: epidemiology and prevention. Florida Department of Health, Bureau of Epidemiology, Epi Update, May 2010, 1-5. (http://www. doh.state.fl.us/disease_ctrl/epi/Epi_Updates/2010_ index.html, accessed on 22 March 2013). 29. Svartberg K. & Forkman B., 2002. Personality traits in the domestic dog (Canis familiaris). Appl Anim Behav Sci, 79, 133-155. 30. Thompson P.G. 1997. The public health impact of dog attacks in a major Australian city. Med J, 167, 129-132.

Veterinaria Italiana 2013, 49 (1), 37-50

Escherichia coli, Salmonella spp., Virus dell’Epatite A e Norovirus in Molluschi Bivalvi nel 2011-12 in Sud Italia Giovanna Fusco, Giuseppe Aprea, Giorgio Galiero, Achille Guarino, Maurizio Viscardi Istituto Zooprofilatico Sperimentale del Mezzogiorno, Dipartimento di Sanità Animale, Via Salute 2, 80055 Portici (NA), Italia aprea_giuseppe@libero.it Parole chiave Escherichia coli, Molluschi Bivalvi, Norovirus, Salmonella spp., Sud Italia, Virus dell’Epatite A.

Riassunto La legislazione europea ha fissato limiti microbiologici, chimici e biotossicologici per i molluschi ma non ha dato indicazioni sulla contaminazione da virus a trasmissione alimentare. In questo studio sono riportati i risultati sulla contaminazione da Salmonella spp., Escherichia coli (E. coli), Virus dell’Epatite A (HAV) e Norovirus (NoV) di campioni di molluschi raccolti, nel biennio 2011-2012, in Sud Italia. Tutte le matrici analizzate di Mytilus galloprovincialis e di Solen marginatus sono risultate negative all’HAV, il 6,8% positiva a Norovirus GI (NoVGI) e l’11,9% positiva a Norovirus GII (NoVGII). Inoltre esse sono risultate negative alla contaminazione da Salmonella spp. mentre il 27% è risultata positiva a E. coli. I dati ottenuti, inoltre, hanno dimostrato l’assenza di correlazione tra la contaminazione batterica dei Molluschi Bivalvi eduli testati e quella virale. Si suggerisce alle Autorità Competenti la valutazione urgente di misure normative aggiuntive al fine di garantire la sicurezza al consumatore. Veterinaria Italiana 2013, 49 (1), 51-54

Introduzione In Italia non esiste una specifica normativa sulla ricerca di virus enterici a trasmissione alimentare ad esclusione dell’art. 11 comma 5, lett. B del Reg. CE 853/2004 e del Decreto Legislativo n. 191 del 2006. Il primo lascia la possibilità di stabilire i “requisiti igienico sanitari supplementari in collaborazione con il laboratorio di riferimento comunitario per i molluschi bivalve vivi comprese le analisi virologiche e le relative norme virologiche”. Il secondo obbliga le Regioni del territorio nazionale e le Province Autonome ad applicare un piano di sorveglianza in funzione della situazione epidemiologica di ciascun territorio nei confronti di agenti responsabili di zoonosi previsti nell’allegato I parte B del regolamento, tra i quali sono annoverati Calicivirus e il virus dell’epatite A (HAV). In Italia, i casi registrati di insorgenza di gastroenterite da Virus dell’Epatite A sono ancora numerosi. Molti di questi, segnalati in regione Campania, sono stati attribuiti al consumo di molluschi bivalve. Entrambi i virus, NoV e HAV, si diffondono per via oro-fecale. Pertanto i molluschi si contaminano indirettamente con acqua inquinata da feci di individui infetti. Inoltre, NoV e HAV, per la loro capacità di dare infezioni a basse dosi (1-10/UFP) e sopravvivere a lungo nell’ambiente esterno, sono considerati tra le principali cause di gastroenterite nell’uomo (7, 11, 18). In Italia i casi segnalati di malattia di origine alimentare da Norovirus sono ancora pochi, ma in ambito inter-

nazionale si continua a registrare un numero considerevole di episodi. Nel 2006, nel sistema di allerta comunitario per alimenti e mangimi (RASFF), è pervenuta la segnalazione di 9 casi correlati ad alimenti contaminati tra cui ostriche crude (12). Nel 2008 sono stati notificati un caso in Francia, uno in Olanda e 6 in Norvegia, tutti correlati a ostriche provenienti da Spagna, Francia e Gran Bretagna. In particolare, il caso francese è stato relazionato al genogruppo I rilevato in Crassostrea gigas. Nello stesso anno sono stati notificati in Spagna anche 5 casi correlati al consumo di telline (13). Nel 2009, in Norvegia, sono stati riportati 19 casi derivanti dal consumo di Gigas oyster di origini svedesi. Trentadue persone della Repubblica Ceca hanno anche contratto l’HAV da pomodori semi-secchi importati dalla Turchia (14). Nel 2010 sono stati notificati complessivamente 13 casi da Norovirus in Danimarca, Norvegia, Irlanda, Francia e Svezia ascrivibili al consumo di lattughe, lamponi e molluschi bivalve (15). Nel 2011 sono stati segnalati 16 episodi in Danimarca in seguito al consumo di frutti di mare e lamponi (16). Il ruolo di potenziale vettore di patologie virali riconosciuto ad alcune specie di Molluschi Bivalvi è legato alla loro peculiarità di essere organismi filtratori, potendo concentrare nei loro tessuti non solo contaminanti chimici ma anche batteri e virus quando allevati o raccolti in acque contaminate da scarichi fognari. Alla luce di quanto riferito, poiché il

Fusco et al.

Contaminazione batterica e virale nei Molluschi Bivalvi nel Sud Italia

possesso dei corretti requisiti igienico sanitari per la produzione e successiva immissione sul mercato dei molluschi viene stabilito sulla base dei criteri microbiologici e biotossicologici indicati dal Reg. (CE) n. 2073/2005, si ritiene interessante valutare in questa tipologia di alimento, oltre alla presenza di Escherichia coli e Salmonella spp., anche la presenza di HAV e NoV, ritenuti a livello internazionale tra i principali agenti causa di gastroenterite nell’uomo.

CCC TGG AAG AAA G-3’, probe HAV 150 FAM-CCT GAA CCT GCA GGA ATT AA-MGB (3). Per Norovirus GI: (FW) QNIF4 5’CGC TGG ATG CGN TTC CAT-3’ (4), (REV) NV1LCR 5’-CCT TAG ACG CCA TCA TCA TTT AC3’, probe NGI FAM-TGG ACA GGA GAY CGC RAT CTTAMRA (18). Per Norovirus GII: QNIF2 (FW) 5-‘ATG TTC AGR TGG ATG AGR TTC TCW GA-3’ (8), (REV) COG2R 5’-TCG ACG CCA TCT TCA TTC ACA-3’ (6), probe NGII FAM-AGC ACG TGG GAG GGC GAT CG-TAMRA (8).

Materiali e metodi

La retrotrascrizione e l’amplificazione sono state effettuate con il seguente profilo termico: 55°C per 60 min, 95°C per 5 min e 45 cicli a 95°C per 15 sec, 60°C per 1 min e 65°C per 1 min.

Campionamento L’indagine è stata effettuata su 59 campioni: 51 cozze (Mytilus galloprovincialis), provenienti da zone classificate A e B delle province di Napoli e Caserta e 8 cannolicchi (Solen marginatus), provenienti da banchi naturali delle province di Caserta e Salerno.

Controlli microbiologici Le metodiche utilizzate, per la ricerca di Salmonella spp. e il conteggio di Escherichia coli mediante MPN, sono state rispettivamente UNI EN ISO 6579:2004 e ISO TS 16649-3:2005.

Controlli virologici Per la ricerca del Virus dell’Epatite A (genogruppi GI e GII) è stato utilizzato un protocollo fornito dal Laboratorio Nazionale di Riferimento per il controllo delle contaminazioni virali nei Molluschi Bivalvi (Dipartimento di Sanità Pubblica Veterinaria e Sicurezza Alimentare, Reparto Adempimenti Comunitari e Sanità Pubblica dell’Istituto Superiore di Sanità) (3, 6, 8, 18). Il protocollo prevede l’utilizzo di 2 ml di una soluzione di proteinasi K (0,1 mg/ml) (Sigma-Aldrich, Milano, Italia) alla quale sono stati aggiunti 2 g di epatopancreas sminuzzato. Il preparato è stato incubato prima a 37°C in agitazione per 60 min e successivamente in bagnomaria a 65°C per 15 min. I campioni sono stati centrifugati a 3.000 x g per 5 min e il surnatante è stato prelevato e conservato a -20°C. L’estrazione dell’RNA è stata eseguita con il Kit Nucleospin RNA II (Macherey - Nagel GmbH & Co., Düren, Germany) seguendo le istruzioni della ditta produttrice. L’estrazione è stata condotta su 500 µl di campione e l’RNA è stato eluito con 100 µl di buffer. La miscela di retrotrascrizione e la PCR (one step) è stata preparata utilizzando i materiali presenti nel kit Platinum qRT PCR Thermoscript one step system (Invitrogen, Karlsruhe, Germany). I primer e le sonde utilizzate per il Virus dell’Epatite A sono stati i seguenti: (FW) HAV68 5’-TCA CCG CCG TTT GCC TAG-3’, (REV) HAV240 5’-GGA GAG

Il campione è stato considerato positivo quando ha presentato un Ct ≤ 44.

Risultati Salmonella spp. è risultata assente in tutti i campioni così come il Virus dell’Epatite A. Sedici campioni di molluschi (27%) sono risultati positivi per Escherichia coli. Riguardo quest’ultimo dato è utile precisare che la presenza di Escherichia coli è stata rilevata in 10 campioni di cozze e in 6 campioni di cannolicchi. Nove dei campioni positivi a Escherichia coli (56%) hanno superato i limiti di legge di 230 MPN/100g. Di essi uno è risultato provenire da acque di classe A, gli altri da acque di classe B. La presenza di RNA virale appartenente al genere Norovirus genogruppo GI è stata rilevata in 4 campioni, il genogruppo GII è stato rilevato in 7 campioni. I dettagli delle positività sono riportate in Tabella I. In merito alla copresenza di Norovirus ed Escherichia coli, i risultati ottenuti hanno evidenziato: 4 campioni positivi per NoVGI, solo in un campione di cannolicchio con presenza di Escherichia coli (78 MPN/100g), 7 campioni positivi per NoVGII, solo 2 campioni di cannolicchi con presenza di Escherichia coli (230 MPN/g e 78 MPN/100g), (Tabella II). Infine sono risultati positivi alla copresenza di entrambi i genogruppi di Norovirus un campione di cozza, conforme ai requisiti di legge per assenza di Escherichia coli, e un campione di cannolicchio, positivo a Escherichia coli (78 MPN /100g). Tabella I. Presenza di NoV, HAV, Escherichia coli e Salmonella spp. in campioni di molluschi bivalve esaminati, nel biennio 2011-2012, in Sud Italia. Campioni

Mytilus 51 galloprovincialis Solen marginatus 8 Totale 59

NoV NoV Escherichia Salmonella HAV GI GII coli spp. 3

1 4

2 7

0 0

6 16

0 0

Veterinaria Italiana 2013, 49 (1), 51-54

Fusco et al.

Contaminazione batterica e virale nei Molluschi Bivalvi nel Sud Italia

Tabella II. Correlazione tra campioni positivi per NoV e numero di Escherichia coli (MPN/100g). Campioni positivi per NoV/Escherichia coli Mytilus galloprovincialis

Solen marginatus

NoV GI + + + + -

NoV GII + + + + + + +

Escherichia coli 78 MPN/100g 230 MPN/100g

Discussione Alcuni dei risultati ottenuti nel presente studio sono stati presentati in occasione del 2nd Annual World Congress of Virus and Infection (WCVI) tenutosi, nel 2011, a Pechino, Cina (1). I dati ottenuti anche se riferibili a un numero non elevato di campioni, risultano interessanti e richiedono attente considerazioni. La quasi totalità dei campioni esaminati, eccetto gli 8 campioni di cannolicchi, sono stati prelevati dai servizi veterinari da zone di classe A e B, secondo quanto stabilito dal Reg. 854/2004 CE. Nell’attuale normativa, Escherichia coli è ritenuto un utile marker biologico per la valutazione indiretta di fecalizzazione dell’acqua. Alla luce dei risultati ottenuti che indicano un’assenza di correlazione tra Escherichia coli e Norovirus, si ritiene che l’utilizzo del solo agente patogeno Escherichia coli come indicatore biologico indiretto della qualità dell’acqua e dell’alimento sia insufficiente. Si ritiene, pertanto, necessario integrare l’attuale normativa con una disposizione che preveda anche l’impiego di un marker virale adeguato mediante l’impiego di una tecnica ad elevata sensibilità e di facile impiego. Si ricorda che un indicatore per essere definito tale e adatto allo scopo deve presentare caratteristiche biologiche e biochimiche simili agli agenti patogeni. Inoltre le metodiche utilizzate per rilevarne la presenza devono essere sensibili, possibilmente rapide, validate e standardizzate. Si comprende, pertanto, come nel caso di virus, organismi incapaci di replicarsi nell’alimento e con caratteristiche biologiche e biochimiche peculiari, la presenza in un alimento possa essere rilevata solo utilizzando marker biologici a loro molto simili (2, 9, 17).

Veterinaria Italiana 2013, 49 (1), 51-54

Nel presente studio, la quasi totalità dei campioni positivi per Norovirus sono risultati negativi per Escherichia coli se si considerano i limiti di legge stabiliti per i mitili raccolti nelle zone A e B. È interessante evidenziare come Escherichia coli sia risultata presente in 6 degli 8 campioni di cannolicchi esaminati e che 2 di essi superavano i limiti di legge consentiti. Questa specie di mollusco merita una particolare attenzione poiché vivendo insabbiata in aree marine naturali e non essendo sottoposta ad alcuna attività di depurazione, presenta valori igienico-sanitari differenti rispetto a specie che si allevano in filari in aree marine controllate. È utile evidenziare che gli indicatori batterici presentano una maggiore sensibilità ai fattori ambientali ostili e ai processi di depurazione delle acque rispetto ai virus, pertanto nei loro confronti il processo di depurazione risulta molto più efficace (5, 19). Per il Virus dell’Epatite A e per i Norovirus tale effetto non sussiste.

Conclusioni Le attuali normative non prevedono limiti per la presenza di contaminanti virali nei molluschi bivalve vivi. La mancanza di correlazione tra virus a trasmissione alimentare e batteri da contaminazione fecale rende vana la possibilità di valutare indirettamente la presenza di HAV e NoV attraverso la conta batterica di Salmonella spp. ed Escherichia coli. Per tali motivi e per i risultati presentati si ritiene utile proporre alle Autorità Competenti di valutare con urgenza misure aggiuntive alle attuali norme al fine di sorvegliare e quindi garantire la sicurezza del consumatore di molluschi bivalve con particolare riguardo a quelli provenienti da zone di classe A. Tale suggerimento è dettato dalla consapevolezza che l’idoneità dei controlli sanitari deve obbligatoriamente interessare l’Autorità Sanitaria di ogni Paese per garantire al consumatore un alimento salubre. La politica internazionale sulla sicurezza alimentare si basa essenzialmente sull’analisi del rischio, individuazione del pericolo e applicazione di strategie idonee per ridurre la contaminazione dell’alimento e conseguentemente la successiva esposizione dell’uomo al contaminante. Controllare microbiologicamente le acque attraverso i bioindicatori per destinarle in seguito all’allevamento dei mitili rappresenta l’unica misura di sicuro successo. A tal fine, per l’individuazione di contaminanti virali, i Laboratori Ufficiali si potranno avvalere di metodiche biomolecolari standardizzate, di rapida esecuzione e di sicura efficacia, come ad esempio la metodica real-time PCR.

Contaminazione batterica e virale nei Molluschi Bivalvi nel Sud Italia

Fusco et al.

Bibliografia 1. Aprea G. 2011. Norovirus and Epatitis A Virus: Foodborne Pathogens Responsible of Human Viral Gastroenteritis. 2nd Annual World Congress of Virus and Infection (WCVI), July 30 - August 1, Beijing, China. 2. Caballero S., Abad F.X., Loisy F., La Guyader F.S., Cohen J., Pintò R.M. & Bosh A. 2004. Rotavirus virus-like particles as surrogates in environmental persistence and inactivation studies. Appl Environ Microbiol, 70, 3904-3909. 3. Costafreda M.I., Bosch A. & Pintò R.M. 2006. Development, evaluation, and standardization of a real-time TaqMan reverse transcription-PCR assay for quantification of hepatitis A virus in clinical and shellfish samples. Appl Environ Microbiol, 72, 3846-3855. 4. Da Silva A.K., Le Saux J.C., Parnaudeau S., Pommepuy M., Elimelech M. & Le Guyader F.S. 2007. Evaluation of removal of noroviruses during wastewater treatment, using Real-time Reverse Transcription-PCR: different behaviors of genogroups I and II. Appl Environ Microbiol, 73, 7891-7897. 5. Franco E., Toti L., Gabrieli R., Croci L., De Medici D. & Pana A. 1990. Depuration of Mytilus galloprovincialis experimentally contaminated with hepatitis A virus. Int J Food Microbiol, 11, 321-328. 6. Kageyama T., Koiјma S., Shinohara M., Uchida K., FuKushi S., Hoshino F.B., Takeda N. & Katayama K. 2003. Broadly reactive and highly sensitive assay for Norwalk- like viruses based on realtime quantitative reverse transcription-PCR. J Clin Microbiol, 41, 1548-1557. 7. Le Guyader F.S., Parnaudeau S., Schaeffer J., Bosch A., Loisy F., Pommepuy M. & Atmar R.L. 2009. Detection and quantification of noroviruses in shellfish. Appl Environ Microbiol, 75, 618-624. 8. Loisy F., Atmar R.L., Le Saux J.C., Cohen J., Caprais M.P., Pommepuy M. & Le Guyader F.S. 2005. Use of Rotavirus virus-like particles as surrogates to evaluate virus persistence in shellfish. Appl Environ Microbiol, 17, 6049-6053. 9. Loisy F., Atmar R.L., Guillon P., Le Cann P., Pommepuy M. & Le Guyader F.S. 2005. Real time RT-PCR for noroviruses screening in shellfish. J Virol Methods, 123, 1-7. 10. Ministero della Salute - Direzione Generale per la sicurezza degli alimenti e della nutrizione, Ufficio VIII

ex VI. Raccomandazione Prot. DGSAN/VII (ex VI) 3734 del 20.04.2007. 11. Pinto R.M., Costafreda M.I. & Bosh A. 2009. Risk assessment in shellfish-borne outbreaks of hepatitis A. Appl Environ Microbiol, 75, 7350-7355. 12. Rapid Alert System for Food and Feed (RASFF). 2007. Annual Report 2006, European Communities, Luxembourg, 72 pp. (http://ec.europa.eu/food/food/ rapidalert/report2006_en.pdf, ultimo accesso 20 febbraio 2013). 13. Rapid Alert System for Food and Feed (RASFF). 2009. Annual Report 2008, European Communities, Luxembourg, 56 pp. (http://ec.europa.eu/food/food/ rapidalert/report2008_en.pdf, ultimo accesso 20 febbraio 2013). 14. Rapid Alert System for Food and Feed (RASFF). 2010. Annual Report 2009, European Communities, Luxembourg, 76 pp. (http://ec.europa.eu/food/food/ rapidalert/docs/report2009_en.pdf, ultimo accesso 20 febbraio 2013). 15. Rapid Alert System for Food and Feed (RASFF). 2011. Annual Report 2010, European Communities, Luxembourg, 64 pp. (http://ec.europa.eu/food/food/ rapidalert/docs/rasff_annual_report_2010_en.pdf, ultimo accesso 20 febbraio 2013). 16. Rapid Alert System for Food and Feed (RASFF). 2012. Annual Report 2011, European Communities, Luxembourg, 52 pp. (http://ec.europa.eu/food/food/ rapidalert/docs/rasff_annual_report_2011_en.pdf, ultimo accesso 20 febbraio 2013). 17. Skraber S., Gassiloud B. & Gantzer C. 2004. Comparison of coliforms and coliphages as tools for assessment of viral contamination in river water. Appl Environ Microbiol, 70, 3644-3649. 18. Svraka S., Duizer E., Vennema H., de Bruin E., van der Veer B., Dorresteiјn B. & Koopmans M. 2007. Etiological role of viruses in outbreaks of acute gastroenteritis in the Netherlands from 1994 through 2005. J Clin Microbiol, 45, 1389-1394. 19. Ueki Y., Shojoi M., Suto A., Tanabe T., Okimura Y., Kikuchi Y., Saito N., Sano D. & Omura T. 2007. Persistence of caliciviruses in artificially contaminated oysters during depuration. Appl Environ Microbiol, 77, 5618-5701.

Veterinaria Italiana 2013, 49 (1), 51-54

Escherichia coli, Salmonella spp., Hepatitis A Virus and Norovirus in bivalve molluscs in Southern Italy Giovanna Fusco, Giuseppe Aprea, Giorgio Galiero, Achille Guarino & Maurizio Viscardi Istituto Zooprofilatico Sperimentale del Mezzogiorno, Dipartimento di SanitĂ Animale, Via Salute 2, 80055 Portici (NA), Italy aprea_giuseppe@libero.it Keywords Bivalve molluscs, Escherichia coli, Foodborne disease, Hepatitis A, Norovirus, Salmonella spp., Southern Italy.

Summary European Legislation has fixed microbiological, chemical and biotoxicological limits for shellfish but no limits for viruses. In the present study we report the results of an investigation on Salmonella spp., Escherichia coli, Hepatitis A virus (HAV) and Norovirus (NoV) contamination in 59 bivalve shellfish collected during the years 2011-2012 in Southern Italy. All the samples of Mytilus galloprovincialis and of Solen marginatus were negative for HAV whereas 6.8% of them were positive for Norovirus GI (NoVGI) and 11.9% positive for Norovirus GII (NoVGII). Samples were also negative for Salmonella spp., while 16 of them (27%) were positive for E.Â coli. No correlation was found between E. coli and NoV contamination in bivalve molluscs. Moreover, the Competent Authorities are advised to take into serious consideration additional measures for the legislation in force in order to guarantee the consumerâ&#x20AC;&#x2122;s health. Veterinaria Italiana 2013, 49 (1), 55-58

Introduction Currently there is no specific legislation enforcing the control of enteric viruses in bivalve mussels. Only art. 11.5, letter B of Reg. CE 853/2004 allows the possibility of laying down additional health standards for live bivalve molluscs in cooperation with the Community Reference Laboratory, including virus testing procedures and virological standards. Moreover, the Legislative Decree n. 191/2006 obliges the regions of the Italian territory and the Italian autonomous provinces to apply a surveillance plan according to the different epidemiological situations against the zoonotic agents listed in the Annex I, part B of the Regulation (e.g. Calicivirus and HAV). In Italy, notification of cases of gastroenteritis from HAV is still very common. Many of them are reported in the Campania region and are related to the consumption of bivalve molluscs. Since both NoV and HAV are transmitted via the human stomach and intestines, shellfish can be indirectly affected from water contaminated with faeces of infected individuals. Moreover NoV and HAV are considered two of the main causes of gastroenteritis in humans due to their ability to give infection with very low doses (1-10/UFP) and by surviving in the external environment for a long time (7, 11, 18). In Italy, confirmed cases of foodborne disease from Norovirus are still few,

but they could be underestimated, since outside Italy a considerable number of episodes has been recorded. In 2006, in the E.U. food and feed alert system (RASFF), nine outbreaks were reported from raw oysters (12) and in 2008 two outbreaks (in France and the Netherlands) were notified together with six human cases in Norway. All the cases regarded oysters coming from Spain, France and the U.K., and the French outbreak, were related to genogroup I detected in Crassostrea gigas. In the same year five cases of HAV were reported in Spain from tellina clams (13). In 2009, 19 human cases from Norovirus were reported in Norway from eating Gigas oysters coming from Sweden. 32 persons from the Czech Republic have also contracted HAV from semi-dried tomatoes imported from Turkey (14). In 2010, a total of 13 cases of Norovirus were notified in Denmark, Norway, Ireland, France and Sweden from lettuce, raspberries and bivalve molluscs (15). In 2011, 16 outbreaks of Norovirus were reported by Denmark from consumption of oysters or mussels (10 notifications) and raspberries (5 cases originating from Serbia and one from China) (16). The potential role that mussels have as vectors of bacterial and viral diseases is closely related to their water filtration activity. Consequently they are able to concentrate in their tissues not only chemical residues but also microorganisms as contaminants of the waters they are collected from.

Fusco et al.

Bacterial and virological contamination in bivalve molluscs in Southern Italy

We know that the main hygienic and sanitary parameters for correct production and commercialization of bivalve molluscs are laid down in the Reg. (CE) n. 2073/2005. Apart from E. coli and Salmonella spp. we advise taking into serious consideration the presence of enteric viral contaminants within filter-feeding lamellibranch molluscs, in particular HAV and NoV genogroups GI and GII, which are related to the most important causes of human gastroenteritis worldwide.

Materials and Methods Sampling The survey was carried out on 59 samples of Italian bivalve molluscs. Fifty one samples were represented by Mytilus galloprovincialis collected from class A and B water production areas in Naples and Caserta provinces and 8 samples were represented by Solen marginatus, collected from natural beds in Caserta and Salerno provinces.

Microbiological control Salmonella spp. and E. coli were isolated using MPN standard methods as described in UNI EN ISO 6579:2004 and ISO TS 16649-3:2005.

Virological control A PCR protocol for detecting HAV, NoVGI and NoVGII used by the Italian National Reference Laboratory (Istituto Superiore di Sanità) for control of viral contaminations in bivalve molluscs has been adopted (3, 6, 8, 18). This protocol uses 2 mL of a K proteinase solution (0.1 mg/mL) (Sigma-Aldrich, Milan, Italy) in which 2 g of cut epatopancreas is added from each sample, incubated in agitation at 37°C for 60 min and then in a water bath at 65°C for 15 min, centrifuged at 3,000 g for 5 min. The supernatant was collected and stocked at -20°C. RNA extraction was carried on with Kit Nucleospin RNA II (Macherey- Nagel GmbH & Co., Duren, Germany). The extraction started with 500 µL of sample and the RNA was eluted with 100 µL of elution buffer. The retro-transcription mix and the one-step PCR was prepared using the reagents of the Platinum qRT PCR Thermoscript one-step system (Invitrogen, Karlsruhe, Germany). Primers and probes used for the detection of hepatitis A virus are: (FW) HAV68 5’-TCA CCG CCG TTT GCC TAG-3’, (REV) HAV240 5’- GGA GAG CCC TGG AAG AAA G-3’, probe HAV 150 FAM- CCT GAA CCT GCA GGA ATT AA-MGB (3). For Norovirus GI: (FW) QNIF4 5’CGC TGG ATG CGN TTC CAT-3’ (4), (REV) NV1LCR 5’-CCT TAG ACG CCA TCA TCA TTT AC-3’, probe NGI FAM-TGG ACA GGA GAY CGC RAT CT-

TAMRA (18). For Norovirus GII: QNIF2 (FW) 5-‘ATG TTC AGR TGG ATG AGR TTC TCW GA-3’ (8), (REV) COG2R 5’- TCG ACG CCA TCT TCA TTC ACA-3’ (6), probe NGII FAM-AGC ACG TGG GAG GGC GAT CG-TAMRA (8). The retro-transcription and amplification were carried out using the following thermal scheme: 55°C for 60 min, 95°C for 5 min and 45 cycles at 95°C for 15 sec, 60°C for 1 min and 65°C for 1 min. Each sample was considered positive when Ct was ≤ 44.

Results Samples were all negative for Salmonella spp. and HAV, while 16 of them (27%) were positive for E. coli. It is useful to underline that these samples were represented by 10 Mytilus galloprovincialis and 6 Solen marginatus. Nine samples (56%) showed E. coli counts higher than 230 MPN/100g, which is the value allowed by current legislation. Among the positive samples, 1 was collected from a class A water production area while all the others were from class B waters. Norovirus GI Viral RNA was detected in 4 samples and Norovirus GII RNA was detected in 7 samples. The details of the findings are reported in Table I. As far as the co-presence of NoV and E. coli is concerned, the results are the following: among the 4 samples positive for Norovirus GI, only 1 sample of Solen marginatus was also positive for E. coli with a value of 78 MPN/100g, while in the 7 samples positive for NoVGII, E. coli was isolated in only 2 samples of Solen marginatus with values of 78 MPN/100g and 230 MPN/100g respectively (Table II). Moreover, 2 samples were positive for NoVGI and NoVGII. 1 sample was Mytilus galloprovincialis and was negative for E. coli. The other was Solen marginatus and it was also positive for E. coli with a count of 78MPN/100 g.

Discussion Some of the results of this study have been presented at the 2nd Annual World Congress of Virus and Infection (WCVI) in Beijing, China, in 2011 (1). Table I. Norovirus, hepatits A virus, E. coli, and Salmonella spp. in samples of bivalve molluscs (2011-2012). Samples Mytilus galloprovincialis Solen marginatus Total

NoV NoV Salmonella HAV E. coli GI GII spp.

8 59

1 4

2 7

0 0

6 16

0 0

Veterinaria Italiana 2013, 49 (1), 55-58

Fusco et al.

Bacterial and virological contamination in bivalve molluscs in Southern Italy

Table II. Correlation between positive samples for norovirus detection and E. coli count (MPN/100g). Positive samples for norovirus/E. coli Mytilus galloprovincialis

Solen marginatus

NoV GI + + + + -

NoV GII + + + + + + +

E. coli 78 MPN/100g 230 MPN/100g

They showed some interesting findings, even if related to a relatively small number of samples and require some general considerations. The totality of Mytilus galloprovincialis samples were collected by the Italian public veterinary service from class A and B waters according to the Reg. CE 854/2004. Since the results of our research have shown no correlation between the concentration of E. coli and NoV, we think that the detection of the pathogen E. coli is insufficient if used as the only parameter to assure the health quality of water and molluscs. That is why it is important to integrate the legislation in force with an adequate viral marker, which must be able to survive in the environment and must be easily detectable by conventional laboratory techniques. The authors are aware that when a microbial indicator is chosen it has to have biological and biochemical characteristics similar to the target that needs to be evaluated in samples. Moreover the laboratory test used for detection must be sensitive, fast, validated and standardized. Viruses are incapable of replicating in food and they have biological and biochemical features completely different from bacteria. For these reasons, foodborne viruses must be directly detected or indirectly investigated by identifying other viruses with very similar characteristics in the samples (2, 9, 17). In our study most of the samples positive for NoV detection were negative for E. coli, if considering the legal limits required for live edible lamellibranch molluscs produced in class A and B waters. This

Veterinaria Italiana 2013, 49 (1), 55-58

result clearly highlights the absence of correlation between bacterial and viral faecal contaminants. It is interesting to underline that E. coli was present in six of the eight samples of Solen marginatus tested, and two of them showed counts higher than the legislative limits. We think that this edible mollusc deserves particular attention because it lives wildly in the sand and it is not subjected to any depuration system. For this reason it can present hygienic-sanitary features that may differ from species collected from controlled waters. Moreover it is useful to highlight that bacterial indicators, unlike viruses, present a higher sensitivity to environmental stress factors and to depuration activity, which leads to much better depuration results when compared to viruses (5, 19).

Conclusions No provisions for viral contaminants in edible live lamellibranch molluscs are currently included in the legislation. Moreover, because of the absence of correlation between faecal bacterial indicators and foodborne viruses, it is not possible to evaluate NoV and HAV presence indirectly through Salmonella spp. and E. coli counts, as shown with the results of our findings. For these reasons we propose the Competent Authorities urgently consider additional measures to update the legislation in force in order to fully control and guarantee the consumer’s health from consumption of bivalve molluscs, especially those originating in class A water production areas. In fact, the efficacy of public controls must be a priority for every Sanitarian System, Italian or European, since consumer’s health is their common concern. International food safety politics is essentially based on risk analysis, hazard identification and application of strategies to reduce food contamination, and consequently consumers’ exposure. The microbiological control of water and bivalve molluscs through bioindicators represents the only instrument we have to achieve the main goal of public health, which is ‘safe food’. For this purpose Official Laboratories can be supported by the use of molecular biology techniques that are rapid and efficient screening tests. In fact, according to the European guidelines, real-time PCR is considered by the Italian Ministry of Health the most sensitive method, and it is recommended for virus detection in food (10).

Bacterial and virological contamination in bivalve molluscs in Southern Italy

Fusco et al.

References 1. Aprea G. 2011. Norovirus and Epatitis A Virus: Foodborne Pathogens Responsible of Human Viral Gastroenteritis. 2nd Annual World Congress of Virus and Infection (WCVI), July 30 - August 1, Beijing, China. 2. Caballero S., Abad F.X., Loisy F., La Guyader F.S., Cohen J., Pintò R.M. & Bosh A. 2004. Rotavirus virus-like particles as surrogates in environmental persistence and inactivation studies. Appl Environ Microbiol, 70, 3904-3909. 3. Costafreda M.I., Bosch A. & Pintò R.M. 2006. Development, evaluation, and standardization of a real-time TaqMan reverse transcription-PCR assay for quantification of hepatitis A virus in clinical and shellfish samples. Appl Environ Microbiol, 72, 3846-3855. 4. Da Silva A.K., Le Saux J.C., Parnaudeau S., Pommepuy M., Elimelech M. & Le Guyader F.S. 2007. Evaluation of removal of noroviruses during wastewater treatment, using Real-time Reverse Transcription-PCR: different behaviors of genogroups I and II. Appl Environ Microbiol, 73, 7891-7897. 5. Franco E., Toti L., Gabrieli R., Croci L., De Medici D. & Pana A. 1990. Depuration of Mytilus galloprovincialis experimentally contaminated with hepatitis A virus. Int J Food Microbiol, 11, 321-328. 6. Kageyama T., Koiјma S., Shinohara M., Uchida K., FuKushi S., Hoshino F.B., Takeda N. & Katayama K. 2003. Broadly reactive and highly sensitive assay for Norwalk- like viruses based on realtime quantitative reverse transcription-PCR. J Clin Microbiol, 41, 1548-1557. 7. Le Guyader F.S., Parnaudeau S., Schaeffer J., Bosch A., Loisy F., Pommepuy M. & Atmar R.L. 2009. Detection and quantification of noroviruses in shellfish. Appl Environ Microbiol, 75, 618-624. 8. Loisy F., Atmar R.L., Le Saux J.C., Cohen J., Caprais M.P., Pommepuy M. & Le Guyader F.S. 2005. Use of Rotavirus virus-like particles as surrogates to evaluate virus persistence in shellfish. Appl Environ Microbiol, 17, 6049-6053. 9. Loisy F., Atmar R.L., Guillon P., Le Cann P., Pommepuy M. & Le Guyader F.S. 2005. Real time RT-PCR for noroviruses screening in shellfish. J Virol Methods, 123, 1-7. 10. Ministero della Salute - Direzione Generale per la sicurezza degli alimenti e della nutrizione, Ufficio VIII

ex VI. Raccomandazione Prot. DGSAN/VII (ex VI) 3734 del 20.04.2007. 11. Pinto R.M., Costafreda M.I. & Bosh A. 2009. Risk assessment in shellfish-borne outbreaks of hepatitis A. Appl Environ Microbiol, 75, 7350-7355. 12. Rapid Alert System for Food and Feed (RASFF). 2007. Annual Report 2006, European Communities, Luxembourg, 72 pp. (http://ec.europa.eu/food/ food/rapidalert/report2006_en.pdf, accessed on 20 February 2013). 13. Rapid Alert System for Food and Feed (RASFF). 2009. Annual Report 2008, European Communities, Luxembourg, 56 pp. (http://ec.europa.eu/food/ food/rapidalert/report2008_en.pdf, accessed on 20 February 2013). 14. Rapid Alert System for Food and Feed (RASFF). 2010. Annual Report 2009, European Communities, Luxembourg, 76 pp. (http://ec.europa.eu/food/food/ rapidalert/docs/report2009_en.pdf, accessed on 20 February 2013). 15. Rapid Alert System for Food and Feed (RASFF). 2011. Annual Report 2010, European Communities, Luxembourg, 64 pp. (http://ec.europa.eu/food/food/ rapidalert/docs/rasff_annual_report_2010_en.pdf, accessed on 20 February 2013). 16. Rapid Alert System for Food and Feed (RASFF). 2012. Annual Report 2011, European Communities, Luxembourg, 52 pp. (http://ec.europa.eu/food/food/ rapidalert/docs/rasff_annual_report_2011_en.pdf, accessed on 20 February 2013). 17. Skraber S., Gassiloud B. & Gantzer C. 2004. Comparison of coliforms and coliphages as tools for assessment of viral contamination in river water. Appl Environ Microbiol, 70, 3644-3649. 18. Svraka S., Duizer E., Vennema H., de Bruin E., van der Veer B., Dorresteiјn B. & Koopmans M. 2007. Etiological role of viruses in outbreaks of acute gastroenteritis in the Netherlands from 1994 through 2005. J Clin Microbiol, 45, 1389-1394. 19. Ueki Y., Shojoi M., Suto A., Tanabe T., Okimura Y., Kikuchi Y., Saito N., Sano D. & Omura T. 2007. Persistence of caliciviruses in artificially contaminated oysters during depuration. Appl Environ Microbiol, 77, 5618-5701.

Veterinaria Italiana 2013, 49 (1), 55-58

Studio degli effetti tossici del ritardante di fiamma PBDE-47 su vongola Chamelea gallina (Linnaeus, 1758) Salvatora Angela Angioni, Giampiero Scortichini, Gianfranco Diletti, Fabrizia Perletta, Roberta Ceci, Nicola Ferri Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise “G. Caporale”, Campo Boario, 64100 Teramo, Italia a.angioni@izs.it Parole chiave Bioaccumulo, Catena trofica, Chamelea gallina, PBDE-47, Ritardante di fiamma, Venus gallina, Vongola.

Riassunto Lo studio ha avuto l’obiettivo di valutare gli effetti del 2,2’,4,4’-tetrabromodifeniletere (PBDE‑47) su vongola Chamelea gallina (Venus gallina secondo la normativa commerciale vigente). I PBDE, impiegati in diversi prodotti industriali come ritardanti di fiamma, sono annoverati tra le sostanze pericolose dalla Direttiva 2011/65/UE. Si tratta di composti bioaccumulabili, ritenuti interferenti endocrini, genotossici e neurotossici, praticamente ubiquitari, la cui concentrazione nell’ambiente, negli ultimi anni, è aumentata in maniera considerevole. Il presente studio ha avuto l’obiettivo di verificare gli effetti del PBDE-47 su Chamelea gallina: potere tossico ed eventuali effetti dannosi sulle gonadi, capacità di bioaccumulo nei tessuti e possibile ingresso nella catena trofica. La ricerca si è avvalsa di prove sperimentali a 96 h e a 14-21gg su esemplari di vongola stabulati in acqua marina filtrata. Le prove sono state precedute da un periodo di adattamento dei molluschi della durata di 5-7gg. Le vongole sono state alimentate con alghe marine (Dunaliella tertiolecta). La scelta del composto tossico PBDE-47 è stata effettuata in considerazione delle maggiori concentrazioni, tra i congeneri di PBDE, riscontrate in alcune specie acquatiche. Lo studio ha evidenziato che le concentrazioni impiegate del contaminante non hanno alterato le funzioni vitali, causato livelli significativi di mortalità e determinato alterazioni evidenti alle gonadi di Chamelea gallina. La ricerca ha evidenziato, comunque, il potere di bioaccumulo del mollusco bivalve, permettendo al PBDE-47 l’ingresso nella catena trofica. Veterinaria Italiana 2013, 49 (1), 59-68

Introduzione I ritardanti di fiamma, polibromodifenileteri (PBDE), sono composti chimici idrofobici, inclusi tra le sostanze pericolose contemplate dalla Direttiva 2011/65/UE (26). Comunemente sono usati nei settori tessile ed elettronico, nonché nella produzione di imballaggi plastici e materiale edile, in quanto caratterizzati dalla capacità di ritardare l’estendersi delle fiamme in caso di incendio. Negli ultimi anni il loro utilizzo ha subito un incremento, con il conseguente aumento della loro concentrazione nell’ambiente (1, 22). I PBDE sono additivi che vengono miscelati con polimeri di varia natura e non chimicamente legati alle plastiche o ai tessuti. Questa particolarità può provocarne un rilascio graduale nel tempo che, da una parte diminuisce le proprietà ignifughe del manufatto, dall’altra aumenta il rischio di contaminazione ambientale (22). Queste sostanze possono essere rilasciate nell’ambiente in fase di produzione, di utilizzo, di smaltimento in discarica, durante le operazioni di ince-

nerimento e tramite i reflui industriali (7, 27). Tali inquinanti ambientali sono divenuti ubiquitari (acqua, aria e terreno) (4, 6, 22, 24). La loro presenza è stata riscontrata in pesci, uccelli, molluschi bivalve, mammiferi marini, nonché in latte materno, tessuto adiposo, sangue e siero umano (13, 22, 24). I PBDE sono ritenuti persistenti, stabili, poco sensibili ai processi di degradazione chimica e biologica, e in grado di esercitare la propria azione contaminante anche a considerevole distanza dai luoghi di emissione (22, 24). Sono noti come composti bioaccumulabili (14, 22, 24), interferenti endocrini (1), neurotossici (17, 24) e genotossici con capacità di indurre aberrazioni cromosomiche (3, 7). In letteratura sono riportati effetti dannosi sulle gonadi di molluschi bivalve (mitili) (1). La vongola Chamelea gallina (Figura 1), mollusco bivalve endobentonico, che vive aggregato in banchi a elevata densità in fondali sabbiosi o sabbio-fangosi, ha un buon potere di bioaccumulo di contaminanti ambientali. L’alimentazione, basata sulla microfagia,

Effetti tossici del PBDE-47 sulla vongola Chamelea gallina

Angioni et al.

stati impiegati esemplari di Chamelea gallina (Super Regno: Eukaryota, Regno: Animalia, Philum: Mollusca, Classe: Bivalvia, Ordine: Veneroida, Famiglia: Veneridae, Genere: Chamelea, Specie: Gallina). Gli esemplari di vongola e acqua marina per gli acquari sono stati prelevati, con frequenza mensile, in un punto di campionamento (Latitudine: 42° 27’ 279 N e Longitudine: 14° 15’ 865 E) situato ad una distanza di circa 500 m dalla costa abruzzese, a sud del porto di Pescara e ad una profondità di circa 6 m (Figura 2).

Prove ecotossicologiche

Figura 1. Chamelea gallina. lavalledelcesano.it.

I molluschi prelevati sono stati trasferiti in frigo portatile, immediatamente trasportati a temperatura refrigerata in laboratorio e selezionati in base alla taglia minima commerciabile (25 mm ± 10%). Gli esemplari sono stati ripartiti in 10 acquari in vetro temperato (40x25x28 cm) dotati di dispositivi di ossigenazione e filtrazione meccanica dell’acqua. Per ciascun acquario sono stati impiegati 16 litri di acqua di mare, precedentemente filtrata mediante

mediante un sistema filtrante (15, 18), rende la vongola particolarmente sensibile alla qualità dell’acqua, al particolato in sospensione, nonché agli stress di natura chimica. In generale, lo stato di salute della fauna ittica riflette le condizioni dell’ambiente in cui essa vive. Da qui la possibilità di utilizzare gli studi ecotossicologici sui Molluschi Bivalvi per determinare la salubrità della fauna ittica (intesa come prodotto destinato all’alimentazione umana) e la qualità degli ecosistemi acquatici. Il presente lavoro ha lo scopo di valutare, mediante la simulazione di un evento inquinante in acquario, il potere tossico e gli eventuali effetti dannosi del ritardante di fiamma 2,2’,4,4’-tetrabromodifeniletere (PBDE-47) sull’apparato riproduttore della vongola. Inoltre, è stata valutata la capacità di bioaccumulo del composto tossico nei tessuti del mollusco al fine di verificare il suo possibile ingresso nella catena trofica. La scelta del PBDE-47 è stata effettuata in considerazione della sua maggiore concentrazione riscontrata in specie acquatiche (11, 19).

Figura 2. Punto di campionamento. Latitudine: 42° 27’ 279 N, Longitudine: 14° 15’ 865 E. Fonte: Google maps.

Il presente studio, relativo al progetto IZSAM 09/07 RC - Studio degli effetti tossici dei ritardanti di fiamma, polibromodifenileteri (PBDE), su vongola Chamelea gallina, è stato finanziato dal Ministero della Salute.

Materiali e metodi Piano di campionamento e prelievo Lo studio è stato effettuato nel Centro di Biologia delle Acque dell’Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise “G. Caporale”. Sono

Figura 3. Fase di adattamento di Chamelea gallina.

Veterinaria Italiana 2013, 49 (1), 59-68

Angioni et al.

Effetti tossici del PBDE-47 sulla vongola Chamelea gallina

Figura 4. Coltura di Dunaliella tertiolecta. dispositivo costituito da tre cartucce in fibre di polipropilene disposte in successione (25 μm, 10 μm, 1 μm). Gli acquari sono stati collocati in una camera termostatata con temperatura costante di 19 ± 1°C, intensità luminosa di 65 ± 30 lux, fotoperiodo di 16 ore di luce e 8 ore di buio. Le vongole sono state stabulate per un periodo di adattamento di 5-7 giorni (Figura 3). I molluschi sono stati alimentati con Dunaliella tertiolecta, alga marina unicellulare (Figura 4) coltivata in laboratorio su terreno di coltura concentrato (20 ml x 1 L di acqua di mare) Guillard’s-F/2 50X (Marine Water Enrichment Solution SIGMA Aldrich®, USA) (2). Prima della somministrazione di Dunaliella tertiolecta, le cellule algali sono state recuperate dal mezzo di coltura mediante centrifugazione (400 RCF x 10’ a 4°C) (RCF: relative centrifugal force), contate in camera di Fuchs Rosenthal con microscopio ottico (10-40X) (Figura 5) e utilizzate ad una concentrazione di 5-15 X 103cellule/ml (2). Le vongole, secondo quanto descritto in letteratura, sono state alimentate nel 1° e 4° giorno (2, 23). La sperimentazione si è avvalsa di 7 prove ecotossicologiche ognuna a diversa concentrazione di PBDE-47 (0,25 - 0,50 - 1,0 - 3,0 - 9,0 - 15,0 - 30,0 μg/L) (Figura 6). Ogni prova ha previsto una prova di controllo senza aggiunta del composto tossico. In relazione alla sola prima prova (0,25 μg/L), è stata prevista una prova di “controllo solvente” con aggiunta di solo acetone (<10 mg/L), diluente del PBDE-47. Tutte le prove hanno previsto 5 repliche effettuate con lo stesso numero di molluschi (12 vongole) e alle stesse condizioni ambientali. Quotidianamente sono stati rilevati il numero di molluschi morti, i parametri chimico-fisici dell’acqua (temperatura, salinità, pH, ossigeno disciolto) e la concentrazione di ammoniaca. Il PBDE-47 (ChemService, Inc., West Chester, PA, USA), disciolto in acetone, è stato addizionato al mezzo acquoso alle concentrazioni note.

Veterinaria Italiana 2013, 49 (1), 59-68

Figura 5. Dunaliella tertiolecta (20X) in camera di Fuchs Rosenthal. Un particolare del reticolo di conta, freccia a destra.

Figura 6. Prova con PBDE-47. La scelta del range di concentrazioni del composto tossico e dei tempi di esposizione è stata effettuata secondo i criteri riportati in letteratura e in base ai risultati preliminari ottenuti nel corso della sperimentazione (1, 10). La concentrazione di acetone in acqua è stata, per tutte le prove, inferiore a 10 mg/L, valore al di sotto della No Observed Effect Concentration (NOEC) (1). Per ogni concentrazione di PBDE-47 è stata prevista la valutazione sia degli effetti tossici a 96 ore (mortalità) sia di quelli dannosi sull’apparato riproduttivo con la protrazione delle prove a 21 giorni. Durante le prove ecotossicologiche a 96 ore, i molluschi non sono stati alimentati (2, 23), a differenza delle prove a 14-21 giorni in cui l’alimento è stato somministrato 2 volte a settimana. A conclusione di ogni prova a 14-21 gg, per ciascuna concentrazione di sostanza tossica, è stato effettuato l’esame istologico dei tessuti delle gonadi sia dei molluschi esposti sia dei controlli. È stata inoltre de-

Effetti tossici del PBDE-47 sulla vongola Chamelea gallina

Esame istologico I campioni di vongole, sgusciate e conservate in formalina al 10% (v/v), sono stati sottoposti all’esame istologico delle gonadi, al fine di evidenziare eventuali alterazioni, nel Laboratorio Nazionale di Riferimento per le malattie dei molluschi dell’Istituto Zooprofilattiaco Sperimentale delle Venezie. Per ciascuna prova sono stati esaminati un numero statisticamente rappresentativo di vongole (20-30 esemplari). L’esame è stato svolto in accordo a quanto riportato in letteratura (5).

Determinazione analitica del PBDE-47 Per la determinazione del PBDE-47 nei tessuti della vongola, è stato ottimizzato un metodo validato per la determinazione di 9 polibromodifenileteri (PBDE-28-47-66-85-99-100-153-154-183) in matrici alimentari. Il metodo ha previsto l’utilizzo della tecnica di diluizione isotopica, la separazione in gas-cromatografia ad alta risoluzione (HRGC) e la rivelazione in spettrometria di massa ad alta risoluzione (HRMS). I campioni di molluschi (5-10 grammi) sono stati miscelati con una quantità di terra di diatomee 2-3 volte maggiore rispetto al peso dell’aliquota e mantenuti in stufa a 40°C per una notte. Ad ogni campione è stata aggiunta la miscela degli “standard interni”, costituita da 7 PBDE marcati con carbonio 13, rappresentativa degli analiti in esame. Successivamente, i campioni sono stati estratti con una miscela acetone/esano (20:80, v/v) attraverso estrazione accelerata con solvente, utilizzando un sistema di estrazione ASE® 200 (Dionex, Sunnyvale, California, USA) alla pressione di 1.500 psi e una temperatura di 125°C. L’estratto organico è stato portato a secco su evaporatore rotante ed è stata determinata la percentuale lipidica per via gravimetrica. Il successivo processo di purificazione è stato effettuato in due fasi. Nella prima l’estratto solubilizzato in esano è stato sottoposto ad una partizione liquido/liquido (passaggi in H2SO4 concentrato e soluzione acquosa satura di NaCl), nella seconda fase l’estratto è stato purificato per mezzo di una colonna cromatografica multistrato impaccata manualmente con 3 grammi di gel di silice neutra, 4 grammi di gel di silice impregnato con H2SO4 al 44% (p/p) e 1 grammo di sodio solfato anidro. Gli eventuali interferenti sono stati rimossi con esano, i PBDE eluiti con una miscela diclorometano/esano (10:90, v/v). L’estratto finale è stato portato a secco mediante evaporazione in corrente di azoto e ripreso con una soluzione contenente “standard di iniezione”, rap-

presentata da 2 PBDE marcati con carbonio 13 (diversi dai precedenti). La soluzione ottenuta è stata iniettata in gascromatografia ad alta risoluzione e la rivelazione dei composti in esame è stata ottenuta per mezzo di uno spettrometro di massa operante con risoluzione (R) maggiore di 9500, in single ion monitoring. Il metodo adottato ha previsto l’impiego di un sistema HRGC-HRMS costituito da un gascromatografo capillare Trace Series 2000 (ThermoQuest CE Instruments, Milano, Italia) accoppiato con uno spettrometro di massa MAT 95 XP (Thermo Fisher Scientific, Brema, Germania). L’analisi è stata condotta su una colonna capillare DB-5 MS (60 m x 0,25 mm, 0,10 μm, J&W Scientific, California, USA). Tutte le soluzioni standard di riferimento, sono state acquistate da Wellington Laboratories Inc. (Ontario, Canada).

Risultati e Discussione I valori dei parametri chimico-fisici considerati si sono mantenuti all’interno dei seguenti range: • concentrazione dell’ammoniaca (NH3): < 0,25 mg/L; • temperatura dell’acqua: 18,5 - 20,7°C; • salinità: 34,4-39,6 g/L; • pH: 8,22 - 8,37; • ossigeno disciolto: 8,69 - 9,27 mg/L. Nel corso della fase di adattamento delle vongole in acquario, è stata registrata una mortalità inferiore al 10% (Figura 7). Nella prova di tossicità a 96 ore, per tutte le concentrazioni previste, la stessa si è costantemente mantenuta al di sotto del 10%. Gli esemplari utilizzati non hanno mai manifestato segni apparenti di sofferenza, alimentandosi ed evidenziando regolarmente le valve semichiuse

Numero vongole morte

terminata la concentrazione di PBDE-47 nei tessuti al fine di verificarne l’eventuale bioaccumulo.

Angioni et al.

0,25

0,50

0 1,00

3,00

0 9,00

15,00

30,00

Concentrazione PBDE-47 (μ/L)

Figura 7. Numero di vongole morte nelle fasi di adattamento, per ciascuna delle 7 prove di tossicità.

Veterinaria Italiana 2013, 49 (1), 59-68

Angioni et al.

Effetti tossici del PBDE-47 sulla vongola Chamelea gallina

con i sifoni ben visibili all’esterno (Figura 8). Nello specifico, l’osservazione dei molluschi ha messo in evidenza un esemplare morto alle concentrazioni di PBDE pari a 0,50 - 1,0 - 9,0 - 15,0 μg/L (mortalità = 1,67%). L’analisi della varianza univariata (ANOVA)

Numero vongole morte

Le mortalità osservate a tutte le concentrazioni non sono risultate statisticamente differenti da quelle attese (0%), infatti, tutti gli intervalli di confidenza delle stesse hanno incluso lo 0% (Tabella II). Non sono stati rilevati esemplari di molluschi morti alle altre concentrazioni utilizzate (Figura 9). Dall’esame istologico non sono risultate alterazioni ai tessuti delle gonadi al termine di ciascun prova di tossicità (14 - 21 giorni). Nelle Figure 10 e 11 sono riportati, come esempio, al valore di 15 μg/L, preparati istologici di gonadi maschili e femminili visti al microscopio ottico. Si evidenzia, come la mortalità tra i molluschi bivalve esposti al composto tossico, sia sempre stata inferiore al 10% anche durante la protrazione delle prove a 21 gg, con 3 esemplari morti alla concentrazioni di PBDE pari a 0,50 μg/L e 2 esemplari morti alle concentrazioni di PBDE pari a 1 - 9 - 15 μg/L (Figura 12) (mortalità pari a 5% e 3,33% rispettivamente). Tali valori non sono significativi, in quanto l’analisi della varianza univariata (ANOVA) non ha rilevato differenze statisticamente significative (F = 1,250; p = 0,279) tra i valori delle mortalità osservate alle differenti concentrazioni (Tabella III). Inoltre, le mortalità osservate a tutte le concentrazioni non sono risultate statisticamente differenti da quelle attese (0%), infatti, tutti gli intervalli di confidenza delle stesse hanno incluso lo 0% (Tabella IV).

Figura 8. Vongole in acquario: esemplare con i sifoni all’esterno.

non ha rilevato differenze statisticamente significative (F = 0,500; p = 0,808) tra i valori delle mortalità osservate alle differenti concentrazioni (Tabella I).

Tabella I. Risultati ottenuti utilizzando il metodo ANOVA univariata (prova a 96 h). 0

0 0,25

0 0,50

1,00

3,00

0 9,00

15,00

30,00

Concentrazione PBDE-47 (μ/L)

Figura 9. Numero di vongole morte per ogni concentrazione di PBDE-47 nelle prove di tossicità a 96 h.

Fonte

Somma Gradi Media F di Pr > F dei di dei quadrati libertà quadrati Fisher Significatività

Concentrazione

0,029

0,005

Errore

3,933

413

0,010

Totale

3,962

419

0,500

0,808

Tabella II. Stima della mortalità attesa nei molluschi bivalve impiegati nella sperimentazione (prova a 96 h). Concentrazione (PBDE-47 µg/L)

Mortalità osservata

t di Student

Significatività >t

0,25 0,5 1 3 9 15 30

0,00% 1,67% 1,67% 0,00% 1,67% 1,67% 0,00%

0,000 0,935 0,935 0,000 0,935 0,935 0,000

1,000 0,350 0,350 1,000 0,350 0,350 1,000

Veterinaria Italiana 2013, 49 (1), 59-68

Intervalli di confidenza (95%) Limite inferiore Limite superiore -3,50% 3,50% -1,84% 5,17% -1,84% 5,17% -3,50% 3,50% -1,84% 5,17% -1,84% 5,17% -3,50% 3,50%

Effetti tossici del PBDE-47 sulla vongola Chamelea gallina

Angioni et al.

Figura 10. Sezione istologica di gonade maschile, senza alterazioni e con buona produzione di gameti, di Chamelea gallina esposta a 15 μg PBDE-47/L. (A) 100X; (B) 400X. A

Figura 11. Sezione istologica di gonade femminile matura, senza alterazioni, di Chamelea gallina esposta a 15 μg PBDE-47/L. (A) 100X; (B) 400X.

Numero vongole morte

Tabella III. Risultati ottenuti utilizzando il metodo ANOVA univariata (prova a 21 gg).

Fonte 1

0 0,25

0 0,50

1,00

3,00

0 9,00

15,00

30,00

Concentrazione PBDE-47 (μ/L)

Somma Gradi Media F di Pr > F dei di dei quadrati libertà quadrati Fisher Significatività

Concentrazione

0,157

0,026

Errore

8,650

413

0,021

Totale

8,807

419

1,250

0,279

Figura 12. Numero di vongole morte per ogni concentrazione di PBDE‑47, nelle prove di tossicità a 21gg. La sperimentazione condotta alla concentrazione di 0,5 µg/L è stata interrotta a 14 giorni. Tabella IV. Stima della mortalità attesa nei molluschi bivalve impiegati nella sperimentazione (prova a 21 gg*). Concentrazione (PBDE-47 µg/L)

Mortalità osservata

t di Student

Significatività >t

0,25 0,5 1 3 9 15 30

0,00% 5,00% 3,33% 0,00% 3,33% 3,33% 0,00%

0,000 1,892 1,262 0,000 1,262 1,262 0,000

1,000 0,059 0,208 1,000 0,208 0,208 1,000

Intervalli di confidenza (95%) Limite inferiore Limite superiore -5,19% 5,19% -0,19% 10,19% -1,86% 8,53% -5,19% 5,19% -1,86% 8,53% -1,86% 8,53% -5,19% 5,19%

* La sperimentazione condotta alla concentrazione di 0,5 µg/L è stata interrotta a 14 giorni.

Veterinaria Italiana 2013, 49 (1), 59-68

Angioni et al.

Effetti tossici del PBDE-47 sulla vongola Chamelea gallina

Livelli di contaminazione dei tessuti In Tabella V si riportano i risultati relativi ai livelli di PBDE-47 riscontrati nei tessuti di vongola esposti a concentrazioni crescenti della sostanza tossica, nell’ambito delle prove a 21 giorni. I risultati relativi alle sperimentazioni con concentrazione di PBDE-47 pari a 9 e 15 g/L, non sono stati riportati a causa di problemi analitici nella fase di estrazione del campione. Il basso recupero dello standard interno, infatti, non ha permesso una determinazione affidabile dell’analita in esame. L’unicità del campione non ha consentito la ripetizione dell’analisi. Le concentrazioni del PBDE-47 nel tessuto sono state riportate su prodotto fresco, prodotto secco e base lipidica. Per le concentrazioni su prodotto secco si è fatto riferimento ad un valore di umidità medio di 80,6%, dato desunto da uno studio condotto su campioni di Chamelea gallina prelevati nel Mar Adriatico (21). Tale accorgimento si è reso necessario in quanto l’esiguo numero di esemplari, costituenti il campione, non è stato sufficiente alla determinazione contemporanea dell’umidità e del PBDE-47. La frazione lipidica è stata determinata in tutti i campioni e i valori ottenuti non hanno presentato differenze significative in funzione del periodo di campionamento. I valori sono risultati compresi tra 0,73% e 1,16% con valore medio di 0,98% di lipidi. I livelli di PBDE-47 rilevati alla fine dei test di tossicità nei gruppi di controllo sono risultati trascurabili rispetto a quelli riscontrati negli individui esposti, con valori compresi tra 96 ng/kg e 130 ng/kg. Tali valori sono risultati comparabili e, a volte, inferiori a quelli riscontrati in monitoraggi condotti su alcune specie di Molluschi Bivalvi in regioni europee e asiatiche (8, 16, 19, 20, 25). L’analisi delle vongole non esposte al contaminante ha mostrato un profilo di contaminazione (rapporti dei 9 congeneri determinati) paragonabile a quelli riscontrati in altri organismi acquatici e riportati in letteratura (11, 19).

In uno studio di bioaccumulo condotto su mitili è stato messo in evidenza come l’abbondanza dei congeneri 47 e 99 nei tessuti possa essere legata ad una migliore efficienza nei meccanismi di assunzione e di accumulo dei due congeneri rispetto agli altri PBDE (20). I livelli di PBDE-47 nel tessuto delle vongole sono risultati da 1.000 a 50.000 volte maggiori nei gruppi degli esposti rispetto a quelli dei corrispondenti controlli (Tabella V). Tali livelli hanno mostrato una crescita, seppur non lineare, all’aumentare della concentrazione della sostanza tossica in acqua. I valori di contaminazione sul prodotto fresco sono risultati compresi tra 110 x 103 ng/kg nelle vongole esposte alla concentrazione minima di 250 ng/L di PBDE in acqua, e 5.870 x 103 ng/kg per quelle esposte alla concentrazione massima di 30.000 ng/L. In diversi studi di esposizione, relativi a organismi acquatici e contaminanti organici persistenti, quali PCB e PBDE, è stata individuata una relazione lineare tra quantità di contaminante determinata nel tessuto e caratteristiche idrofobiche del composto in esame, queste ultime descritte dal logaritmo del suo coefficiente di partizione ottanolo-acqua (log kow). I dati in letteratura riportano per il PBDE-47 un valore di log kow compreso tra 6,0 e 6,8 (10, 12), valore elevato che indica come tale composto abbia spiccate caratteristiche idrofobiche e sia potenzialmente pericoloso poiché facilmente accumulabile, con meccanismi di diffusione passiva, nei tessuti adiposi della vongola. Sebbene le concentrazioni di PBDE-47 impiegate nei test di tossicità siano di gran lunga maggiori rispetto a quelle riscontrate in zone particolarmente inquinate (valori massimi riportati in letteratura di circa 0,5 ng/L), la sperimentazione ha dimostrato la facilità della vongola a bioconcentrare il contaminante in esame, con la conseguente possibilità della sua introduzione nella catena alimentare (16). La bioconcentrazione è il risultato di un’as-

Tabella V. Valori analitici della frazione lipidica e delle concentrazioni di PBDE-47 nei tessuti alla fine di ciascuna prova di tossicità a 21 gg*. PBDE-47 H2O (ng/L) 250 500 1000 3.000 9.000 15.000 30.000

Lipidi controllo (%) 1,07 0,84 1,03 1,08 1,16 0,77 0,90

PBDE-47 controllo (ng/kg) 119 130 118 118 96,0 116 120

PBDE-47 controllo (ng/kg secco) 613 670 608 608 495 598 619

PBDE-47 controllo (ng/kg grasso) 11.200 15.500 11.500 10.900 8.300 15.200 13.300

Lipidi esposti (%) 0,97 0,73 1,14 1,14 1,05 0,90 0,95

PBDE-47 esposti (ng/kg) 110 x 103 97,3 x 103 122 x 103 394 x 103 5.870 x 103

PBDE-47 esposti PBDE-47 esposti (ng/kg secco) (ng/kg grasso) 566 x 103 501 x 103 627 x 103 2.030 x 103 30.300 x 103

11.300 x 103 13.300 x 103 10.700 x 103 34.600 x 103 618.000 x 103

* La sperimentazione condotta alla concentrazione di 500 ng/L è stata interrotta a 14 giorni.

Veterinaria Italiana 2013, 49 (1), 59-68

Effetti tossici del PBDE-47 sulla vongola Chamelea gallina

sunzione diretta, attraverso l’acqua, di una sostanza chimica da parte di un organismo. Infatti il fattore di bioconcentrazione (BCF) è definito come il rapporto, allo stato stazionario (velocità di assorbimento uguale a velocità di eliminazione), tra la concentrazione del contaminante nell’organismo acquatico e la corrispondente concentrazione nel mezzo acquoso. BCF = Co/Cw dove: Co = concentrazione del contaminante (ng/kg) nell’organismo acquatico; Cw = concentrazione del contaminante (ng/L) in acqua. Si può esprimere il valore di Co su prodotto intero, base lipidica o prodotto secco, dando origine a tre fattori (BCFW, BCFL, e BCFD, rispettivamente) che numericamente possono essere molto differenti tra loro. In Europa le sostanze chimiche con BCFW maggiore di 100 sono considerate ad alto potenziale di bioaccumulo e, quindi, classificate “pericolose per l’ambiente” in quanto potrebbero danneggiare gli organismi acquatici e i loro predatori (9). Nel presente studio i dati sperimentali sui livelli di PBDE-47 determinati nel tessuto della vongola sono stati impiegati per calcolare una stima inferiore del valore di BCF. Si parla di stima inferiore poichè il protocollo sperimentale adottato nei test di tossicità, il cui obiettivo principale è stato quello di stimare la tossicità acuta e cronica del PBDE-47, non ha permesso di valutare il raggiungimento dello stato stazionario. Infatti, il protocollo sperimentale ha previsto un’unica determinazione analitica del PBDE-47 nel tessuto della vongola dopo un tempo massimo di esposizione di 21 giorni. In Tabella VI sono riportati i valori di BCF calcolati alle diverse concentrazioni di esposizione. In tutti i casi il valore di BCFW è stato maggiore di 100. Tabella VI. Valori di BCF calcolati alle diverse concentrazioni di esposizione alla sostanza tossica. Concentrazione BCFW BCFD BCFL log log log PBDE-47 H2O (L/kg) (L/kg) (L/kg) BCFW BCFD BCFL (ng/L) 250 439 2.260 45.300 2,6 3,4 4,7 500 195 1.000 26.600 2,3 3,0 4,4 1.000 122 630 10.700 2,1 2,8 4,0 3.000 131 680 11.500 2,1 2,8 4,1 30.000 196 1.010 20.600 2,3 3,0 4,3 BCFW = fattore di bioconcentrazione su prodotto intero; BCFL = fattore di bioconcentrazione su base lipidica; BCFD = fattore di bioconcentrazione su prodotto secco.

Angioni et al.

In letteratura non sono riportati studi sperimentali di bioaccumulo del PBDE-47 in Chamelea gallina, ma su altri Molluschi Bivalvi come Mytilus edulis (10, 19). I valori di BCF in questi organismi sono maggiori rispetto a quelli riscontrati nel presente studio. Una tale differenza potrebbe essere associata a: • concentrazione di ammoniaca (NH3): < 0,25 mg/L; • mancato raggiungimento dello stato stazionario dovuto al breve tempo di esposizione; • inappropriata comparazione tra organismi biologicamente differenti. A conferma delle ipotesi formulate, è stato utilizzato un modello matematico (16) che mette in relazione il log kow di una determinata sostanza con il valore di BCF in Mytilus edulis: log BCFW = 0,858 log kow - 0,808 L’equazione permette, noto il valore di log kow di una sostanza, di stimare il fattore di bioaccumulo del mitilo. L’equazione di linearità impiegata porterebbe ad un risultato di log BCF pari a 4,3, valore superiore a quello sperimentalmente stimato in questo studio per Chamelea gallina.

Conclusioni Nelle condizioni operative ottimizzate in questo studio, Chamelea gallina ha mostrato buona capacità di adattamento in acquario, permettendo la corretta esecuzione delle prove ecotossicologiche in laboratorio. Dai risultati dei test di tossicità a 96 h e a 21 gg non è stata evidenziata una relazione tra concentrazione di PBDE-47 nell’acqua marina degli acquari e mortalità di Chamelea gallina. La massima concentrazione di esposizione al PBDE-47, infatti, non ha provocato fenomeni significativi di mortalità. I risultati dell’esame istologico non hanno evidenziato alterazioni evidenti dei tessuti dell’apparato riproduttivo nelle condizioni sperimentali adottate. Le analisi chimiche indicano come le vongole esposte al PBDE-47 presentino livelli del contaminante nel tessuto migliaia di volte superiori agli esemplari non esposti, in funzione della concentrazione nell’ecosistema acquatico, dimostrando la facilità con la quale Chamelea gallina bioconcentri il contaminante. L’ingresso di PBDE-47 nella catena trofica potrebbe costituire un rischio per la sicurezza alimentare, la qualità degli ecosistemi marini e la salubrità della risorsa vongola.

Ringraziamenti Si ringrazia il Sig. Alfonso de Benedictis per il valido supporto tecnico.

Veterinaria Italiana 2013, 49 (1), 59-68

Angioni et al.

Effetti tossici del PBDE-47 sulla vongola Chamelea gallina

Bibliografia 1. Aarab N., Lemaire-Gony S., Unruh E., Hansen P.D., Larsen B.K., Andersen O.K. & Narbonne J.F. 2006. Preliminary study of responses in mussel (Mytilus edulis) exposed to bisphenol A, diallyl phthalate and tetrabromo diphenyl ether. Aquat Toxicol, 78S, S86-S92.

12. Kuramochi H., Maeda K. & Kawamoto K. 2007. Physicochemical properties of selected polybrominated diphenyl ethers and extension of the UNIFAC model to brominated aromatic compounds. Chemosphere, 67, 1858-1865.

2. Angioni S.A., Giansante C. & Ferri N. 2010. Vongola (Chamelea Gallina): valutazione degli effetti dei solidi sospesi in acqua marina nel mollusco bivalve. Vet Ital, 46, 93-99.

13. Lema S.C., Schultz I.R., Scholz N.L., Incardona J.P., Swanson P. 2007. Neural defects and cardiac arrhythmia in fish larvae following embryonic exposure to 2,2’,4,4’-tetrabromodiphenyl ether (PBDE 47). Aquat Toxicol, 82, 296-307.

3. Baršienė J., Syvokiene J. & Bjornstad., A. 2006. Induction of micronuclei and other abnormalities in mussels exposed to bisphenol A, diallyl phathalate and tetrabromodiphenyl ether - 47. Aquat Toxicol, 78S, S105-S108. 4. De Boer J., Wester P.G., van der Horst A. & Leonards P.E.G. 2003. Polybrominated diphenyl ethers in influents, suspended particulate matter, sediments, sewage treatment plants and effluents and biota from the Netherlands. Environ Pollut, 122, 63-74. 5. Delgado M. & Pérez-Camacho A. 2007. Comparative study of gonadal development of Ruditapes philippinarum (Adams and Reeve) and Ruditapes decussatus (L.) (Mollusca: Bivalvia): Influence of temperature. Scientia Marina, 71, 471-484. 6. De Wit C. A. 2002. An overview of brominated flame retardants in the environment. Chemosphere, 46, 583-624.

14. Liu Y., Zheng G.J., Hongxia Y., Martin M., Richardson B.J, Lam M.H.V. & Lam P.K.S. 2005. Polybrominated diphenil ethers (PBDEs) in sediments and mussel tissue from Hong Kong marine waters. Mar Pollut Bull, 50, 1173-1184. 15. Lucchetti A. 2003. La vongola. Biologia, pesca e consumo delle più importanti specie commerciali. La vongola Chamelea gallina (Linneo, 1758). Il Pesce, 6, 135 (http:// www.pubblicitaitalia.com/ilpesce/2003/6/4963.html ultimo accesso il 26 gennaio 2012). 16. Mai B.-X., Luo X.-J., Yu M. & Chen S.-J. 2009. PBDEs in water and aquatic biota of the Pearl river estuary, South China. Organohalogen Compd, 71, 807-12. 17. McDonald T.A. 2002. A perspective on the potential health risk of PBDEs. Chemosphere, 46, 745-755. 18. Mengoli A. 1998. Aspetti morfo-funzionali dei mitili. Laguna, 4,13 -9.

19. Mizukawa K., Takada H., Takeuchi I., Ikemoto T., Omori K. & Tsuchiya K. 2009. Bioconcentration and biomagnification of polybrominated diphenyl ethers (PBDEs) through lower-trophic-level coastal marine food web. Mar Pollut Bull, 58, 1217-1224.

8. Fernandes A., Mortimer D.N., Gem M., Dicks P., Smith F., White S. & Rose M. 2008. Brominated contaminants (PBDD/Fs and PBDEs) in shellfish. Organohalogen Compounds, 70, 1024-27.

20. Moon H-B., Kannan K., Lee S-J. & Choi M. 2007. Polybrominated diphenyl ethers (PBDEs) in sediment and bivalves from Korean coastal waters. Chemosphere, 66, 243-251.

9. Geyer H.J., Rimkus G.G., Scheunert I., Kaune A.,• Schramm K-W., Kettrup A., Zeeman M., Muir D.C.G., Hansen L.G. & Mackay D. 2000. Bioaccumulation and occurrence of Endocrine-Disrupting Chemicals (EDCs), Persistent Organic Pollutants (POPs), and other organic compounds in fish and other organisms including humans. In: The Handbook of Environmental Chemistry, vol. 2 Part J, Bioaccumulation, (B. Beek, eds). Springer-Verlag, Berlin Heidelberg, 1-166.

21. Orban E., Di Lena G., Nevigato T., Casini I., Caproni R., Santaroni G. & Giulini G. 2006. Nutritional and commercial quality of the striped venus clam, Chamelea gallina, from the Adriatic sea. Food Chemistry, 101, 1063-1070.

10. Gustafsson K., Bjork M., Burreau S. & Gilek M. 1999. Bioaccumulation kinetics of brominated flame retardants (polybrominated diphenyl ethers) in blue mussels (Mytilus edulis). Environ Toxicol Chem, 18, 1218-1224. 11. Kuiper R.V., Vethaak A.D., Cantón R.F., Anselmo H., Dubbeldam M., van den Brandhof E-J, Leonards P.E.G., Wester P.W., van den Berg M. 2008. Toxicity of analytically cleaned pentabromodiphenylether after prolonged exposure in estuarine European flounder (Platichthys flesus), and partial life-cycle exposure in fresh water zebrafish (Danio rerio). Chemosphere, 73, 195-202.

Veterinaria Italiana 2013, 49 (1), 59-68

22. Palm A. Cousins I.T., Mackay D., Tysklind M., Metcalfe C. & Alaee M. (2002). Assessing the environmental fate of chemicals of emerging concern: a case study of polybrominated diphenyl ethers. Environ Pollut, 117, 195-213. 23. Shin P.K.S., Yau F.N., Chow S.H., Tai K.K. & Cheung S.G. 2002. Responses of the green-lipped mussel Perna viridis (L) to suspended solids. Mar Pollut Bull, 45, 157-162. 24. Siddiqi M.A., Laessig R.H. & Reed K.D. 2003. Polybrominated Diphenyl Ethers (PBDEs): New Pollutants - Old Diseases. Clinical Medical & Research, 1, 281-290. 25. Umlauf G., Christoph E.H., Huber T., Mariani G., Mueller A., Skejo H. & Wollgast J. 2009. PBDEs in water, sediments and biota of the river Danube from

Effetti tossici del PBDE-47 sulla vongola Chamelea gallina

Angioni et al.

Germany to the Black Sea. Organohalogen Compd, 71, 737-742.

pericolose nelle apparecchiature elettriche ed elettroniche (Gazzetta Ufficiale L 174 del 1.7.2011).

26. Unione Europea, 2011. Direttiva 2011/65/UE del Parlamento Europeo e del Consiglio dellâ&#x20AC;&#x2122;8 giugno 2011 sulla restrizione dellâ&#x20AC;&#x2122;uso di determinate sostanze

27. Watanabe I. & Sakai S. 2003. Environmental release and behaviour of brominated flame retardants. Environ Int, 29, 665-682.

Veterinaria Italiana 2013, 49 (1), 59-68

Study of the toxic effects of flame retardant PBDE-47 on the clam Chamelea gallina (Linnaeus, 1758) Salvatora Angela Angioni, Giampiero Scortichini, Gianfranco Diletti, Fabrizia Perletta, Roberta Ceci, Nicola Ferri Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise ‘G. Caporale’, Campo Boario, 64100 Teramo, Italy a.angioni@izs.it Keywords Bioaccumulation, Chamelea gallina, Clam, Flame retardant, Food chain, PBDE-47, Venus gallina.

Summary The purpose of the study is to evaluate the effects of 2,2’,4,4’-tetrabromodiphenylether (PBDE-47) on the Chamelea gallina clam (according to current commercial regulations: Venus gallina). PBDEs, which are used as flame retardants in various industrial products, are classed as hazardous substances by Directive 2011/65/EU. They are bioaccumulative compounds, considered to be endocrine disruptors, genotoxic, neurotoxic and practically ubiquitous, and their concentration in the environment has considerably increased in recent years. The aim of this study is to establish the effects of PBDE-47 on Chamelea gallina: toxic power and any harmful effects on the gonads, bioaccumulation capacity in the tissues, and possible entry into the food chain. The research used 96-hour and 21-day experimental tests on clams housed in filtered seawater. The tests were preceded by a period of acclimatisation of the molluscs lasting five to seven days. The clams were fed on seaweed (Dunaliella tertiolecta). The choice of the toxic compound PBDE-47 was based on the high concentration, among the congeners of PBDE, found in some aquatic species. The study demonstrated that the concentration of the contaminant used did not alter the vital functions, cause significant levels of mortality or lead to evident alteration in the gonads of Chamelea gallina. However, the research demonstrated the bioaccumulation capacity of the bivalve mollusc, allowing PBDE-47 to enter the food chain. Veterinaria Italiana 2013, 49 (1), 69-77

Introduction The flame retardants known as polybromodiphenylethers (PBDE) are hydrophobic chemical compounds included among the hazardous substances contemplated by Directive 2011/65/EU (26). They are commonly used in the textile and electronics industries, and in the manufacture of plastic packaging and construction materials, because they possess the ability to delay the spread of flames in case of fire. Their use has increased in recent years, with a consequent increase in their concentration in the environment (1, 22). PBDEs are additives that are mixed with polymers of various kinds, and are not chemically bonded to plastics or tissues. This characteristic can cause their gradual release over time, which reduces the flameproofing properties of the product, and also increases the risk of environmental contamination (22). These substances can be released into the environment at the stage of manufacture, use, disposal by tipping, during incineration operations and through industrial wastewater (7, 27).

These environmental pollutants have become ubiquitous (in water, air and soil) (4, 6, 22, 24). They have been found in fish, birds, bivalve molluscs, marine mammals, and in human milk, adipose tissue, human blood and serum (13, 22, 24). PBDEs are considered to be persistent, stable, not very sensitive to chemical and biological degradation processes, and able to perform their contaminant action even at a considerable distance from the place of emission (22, 24). They are known as bioaccumulative (14, 22, 24), endocrine disrupting (1), neurotoxic (17, 24) and genotoxic compounds with the ability to induce chromosomal aberrations (3, 7). Harmful effects on the gonads of bivalve molluscs (mussels) are reported in the literature (1). The clam Chamelea gallina (Figure 1), an endobenthic bivalve mollusc, which lives in high-density shoals on sandy or sandy/muddy sea beds, has a good power of bioaccumulation of environmental contaminants. Its diet, based on microphagy by means of a filtering system (15, 18), makes the clams particularly sensitive to water quality, particulate matter in suspension and chemical stresses. In general, the health of fish

Toxic effects of PBDE-47 on the clam Chamelea gallina

Angioni et al.

Ecotoxicology tests The molluscs sampled were transferred to a portable refrigerator, immediately conveyed to the laboratory under refrigeration, and selected on the basis of minimum marketable size (25 mm ± 10%). The specimens were divided between ten aquariums made of toughened glass (40x25x28 cm), and fitted with oxygenation and mechanical water filtration devices. 16 litres of seawater, prefiltered by a device consisting of three polypropylene fibre cartridges arranged in succession (25 μm, 10 μm and 1 μm), were used for each aquarium. The aquariums were located in a thermostat-controlled chamber with a constant temperature of 19 ± 1°C, light intensity of 65 ± 30 lux, and photoperiod of 16 hours’ light and 8 hours’ darkness.

Figure 1. Chamelea gallina. lavalledelcesano.it.

The clams were housed for an acclimatisation period of five to seven days (Figure 3). The molluscs were fed on Dunaliella tertiolecta, a unicellular seaweed (Figure 4) grown in the laboratory on Guillard’s-F/2 50X concentrated (20 ml x 1 L of seawater) culture

stocks reflects the conditions of the environment in which they live. This yields the possibility of using ecotoxicology studies on bivalve molluscs to determine the health of fish stocks (namely fish products destined for human consumption) and the quality of aquatic ecosystems. The present study was designed to evaluate, by simulating a polluting event in the aquarium, the toxic power and possible harmful effects of flame retardant 2,2’,4,4’-tetrabromodiphenylether (PBDE‑47) on the reproductive apparatus of clams. The ability to bioaccumulate the toxic compound in the tissues of the mollusc was also evaluated in order to establish whether it could enter the food chain. PBDE‑47 was selected in view of the high concentration of this substance found in aquatic species (11, 19). The present study, relating to project IZSAM 09/07 RC - Study of the toxic effects of PDBE-47 flame retardants on the Chamelea gallina clam - was funded by the italian Ministry of Health.

Figure 2. Sampling point. Latitude: 42° 27’ 279 N, Longitude: 14° 15’ 865 E. Source: Google maps.

Materials and methods Sampling plan and sample taking The study was conducted at the Water Biology Centre of the ‘G. Caporale’ Institute. Specimens of Chamelea gallina (Superkingdom: Eukaryota, Kingdom: Animalia, Phylum: Mollusca, Class: Bivalvia, Order: Veneroida, Family: Veneridae, Genus: Chamelea, Species: Gallina) were used. Clam and seawater specimens for the aquariums were taken monthly from a sampling point (Latitude: 42° 27’ 279 N and Longitude: 14° 15’ 865 E) situated approximately 500 m from the coast of Abruzzo, to the south of Pescara harbour, at a depth of approximately 6 m (Figure 2).

Figure 3. Acclimatisation phase of Chamelea gallina.

Veterinaria Italiana 2013, 49 (1), 69-77

Angioni et al.

Toxic effects of PBDE-47 on the clam Chamelea gallina

Figure 4. Coltivation of seaweed (Dunaliella tertiolecta). medium (Marine Water Enrichment Solution SIGMA Aldrich®, USA) (2). Before the administration of Dunaliella tertiolecta, the seaweed cells were recovered from the culture medium by centrifugation (400 RCF x 10 min at 4°C) (RCF: relative centrifugal force), counted in a Fuchs Rosenthal chamber with an optical microscope (10-40X) (Figure 5) and used at the concentration of 5-15 x 103cells/ml (2). As described in the literature, the clams were fed on the first and fourth days (2, 23). The study used seven ecotoxicology tests, each at a different PBDE-47 concentration (0.25, 0.50, 1.0, 3.0, 9.0, 15.0 and 30.0 μg/L) (Figure 6). Each test involved a control test without the addition of the toxic compound. For the first test only (0.25 μg/L) a “solvent control” test with the sole addition of acetone (<10 mg/l), a diluent of PBDE-47 was used. All the tests involved five replications, conducted with the same number of molluscs (12 clams) under the same environmental conditions. The number of dead molluscs, the chemico-physical parameters of the water (temperature, salinity, pH and dissolved oxygen) and the ammonia concentration were recorded every day. PBDE-47 (ChemService, Inc., West Chester, PA, USA) dissolved in acetone was added to the aqueous medium at known concentrations. The range of concentrations of the toxic compound and the exposure times were chosen according to the criteria reported in the literature and on the basis of the preliminary results obtained during the study (1, 10). The concentration of acetone in water was less than 10 mg/L in all the tests, ie. below the No Observed Effect Concentration (NOEC) (1). For each PBDE-47 concentration the toxic effects were evaluated after 96 hours (mortality), and the harmful effects on the reproductive apparatus were evaluated after 21 days. The test at the PBDE47 concentration of 0.5 μg/L was terminated early, after 14 days, due to a malfunction in the aeration

Veterinaria Italiana 2013, 49 (1), 69-77

Figure 5. Dunaliella tertiolecta (20X) ) in Fuchs Rosenthal chamber. A detail of the counting grid, arrow on the right.

Figure 6. Ecotoxicology assay with PBDE-47. system. During the 96-hour ecotoxicology tests the molluscs were not fed (2, 23), whereas in the 21-day tests, food was administered twice a week. At the end of each 21-day test, for each concentration of toxic substance, histological tests were conducted on the gonad tissues of the exposed molluscs and the controls. The PBDE-47 concentration in the tissues was also determined, to establish whether bioaccumulation took place.

Histological tests The samples of clams, shelled and preserved in 10% formalin (v/v), underwent histological tests of the gonads, in order to investigate any alteration, at the Veneto Experimental Animal Disease Prevention Institute’s National Reference Laboratory for mollusc diseases. A statistically representative number of clams (20-30 specimens) was examined for each test. The test was performed as reported in the literature (5).

Toxic effects of PBDE-47 on the clam Chamelea gallina

Angioni et al.

Analytical determination of PBDE-47

Results and Discussion

To determine the levels of PBDE-47 in the clam tissue, a validated method for the determination of 9 polybromodiphenylethers (PBDE-28-47-66-85-99100-153-154-183) in food matrices was optimised. The method involved the use of the isotope dilution technique, high-resolution gas chromatography (HRGC) separation and high-resolution mass spectrometry (HRMS) detection.

The values of the chemico-physical parameters considered remained within the following ranges:

After careful solvent evaporation, gravimetric lipid determination was performed. The sample purification was performed in two steps. Firstly, the extract dissolved in hexane was subjected to liquid-liquid partitioning (passage through concentrated H2SO4 and saturated aqueous solution of NaCl). After this, the extract was purified by a multilayer chromatography column manually packed with 3 grams of neutral silica gel, 4 grams of silica gel impregnated with 44% (w/w) H2SO4 and 1 gram of anhydrous sodium sulphate. Interfering substances were removed with hexane, and PBDEs were eluted with a mixture of dichloromethane/ hexane (10:90, v/v). The final extract was evaporated to dryness under nitrogen stream and immediately dissolved with an injection of standard solution constituting of two PBDEs labelled with carbon 13 (different from the preceding ones). The solution obtained was injected into a high-resolution gas chromatograph and the test compounds were detected by a mass spectrometer operating with a resolution (R) higher than 9,500 in single ion monitoring.

• water temperature: 18.5 – 20.7°C; • salinity: 34.4 – 39.6 g/L; • pH: 8.22 – 8.37; • dissolved oxygen: 8.69 – 9.27 mg/L. During the acclimatisation stage in the aquarium, a mortality rate of less than 10% was recorded for the clams (Figure 7). In the 96-hour toxicity test, it always remained below 10% for all the concentrations specified. The specimens used showed no apparent signs of suffering; feeding normally and showing half-closed valves with the siphons clearly visible on the exterior (Figure 8). Specifically, observation of the molluscs

Number of dead clams

Samples (5-10 grams) were mixed with a quantity of diatomaceous earth two-to-three times greater than the weight of the aliquot and kept overnight in a stove at 40°C. The mixture of ‘internal standards’, consisting of seven PBDEs labelled with carbon 13, representative of the analytes being tested, was added to each sample. The samples were then extracted with an acetone/hexane (20:80, v/v) mixture by accelerated extraction with solvent, using an ASE® 200 extraction system (Dionex, Sunnyvale, California, USA) at the pressure of 1,500 psi and the temperature of 125°C.

• ammonia concentration (NH3): < 0.25 mg/L;

0.25

0.50

0 1.00

3.00

0 9.00

15.00

30.00

PBDE-47 concentration (μ/L)

Figure 7. Number of clams that died at the acclimatisation stages for each of the 7 toxicity tests.

The instrumental analysis was conducted using an HRGC-HRMS system constituting of a Trace Series 2000 capillary gas chromatograph (ThermoQuest CE Instruments, Milan, Italy) coupled to a MAT 95 XP mass spectrometer (Thermo Fisher Scientific, Bremen, Germany). The analysis was conducted on a DB-5 MS capillary column (60 m x 0.25 mm, 0.10 μm, J&W Scientific, California, USA). All the reference standard solutions were obtained from Wellington Laboratories Inc. (Ontario, Canada). Figure 8. Clams in aquarium: specimen with the siphons on the exterior.

Veterinaria Italiana 2013, 49 (1), 69-77

Angioni et al.

Toxic effects of PBDE-47 on the clam Chamelea gallina

The histological tests did not detect any alteration in the gonad tissue at the end of each toxicity test (21 days). Figures 10 and 11 show, by way of example, at the value of 15 μg/L, histological preparations of male and female gonads seen under the optical microscope. Mortality among the bivalve molluscs exposed to the toxic compound was always below 10%, even when the tests continued for 14 and 21 days, with three dead clams at PBDE concentrations of 0.50 μg/L and two deaths at PBDE concentrations of 1, 9 and 15 μg/L (Figure 12) (mortality rates of 5% and 3.33% respectively). These values are not significant, because univariate ANOVA did not detect any statistically-significant differences (F = 1.250; p = 0.279) between the mortality values observed at the different concentrations (Table III). The mortality rates observed at all concentrations were not statistically different from expected (0%); in fact, all the confidence intervals included 0% (Table IV).

Table I. Univariate analysis of variance (Anova) method (96-h Test). Sum of Degree Mean of squares free squares Fisher’s F P value

Source Concentration Residues Total

0.029 3.933 3.962

6 413 419

0.005 0.010

0.500

Tissue contamination levels Table V shows the results of PBDE-47 levels found in the tissue of clams exposed to increasing concentrations of the toxic substance in the 21-day tests. The results of the tests at PBDE-47 concentrations of 9 and 15 μg/L are not shown, due to analytical problems at the sample extraction step; low recovery of the internal standard prevented a reliable determination of the analyte tested, and as the sample was unique, the test could not be repeated. The PBDE-47 levels in the tissue were reported for wet weight, dry weight and lipid basis. For the effects of concentration on the dry weight, reference was made to a mean moisture value of 80.6%, obtained from a study conducted on samples of Chamelea gallina obtained in the Adriatic Sea (21). This approach was used because the small number of specimens constituting the sample was insufficient for simultaneous determination of moisture and PBDE-47. The lipid fraction was determined in all samples. The values obtained did not show any significant differences according to sampling period. The lipids values were between 0.73% and 1.16%, with a mean value of 0.98%.

Number of dead clams

showed one dead clam at PBDE concentrations of 0.50, 1.0, 9.0 and 15.0 μg/L (mortality = 1.67%). Univariate analysis of variance (ANOVA) did not reveal any statistically-significant differences (F = 0.500; p = 0.808) between the mortality values observed at the different concentrations (Table I). The mortality rates observed at all concentrations were not statistically different from expected (0%); in fact, all the confidence intervals included 0% (Table II). No dead molluscs were found at any of the other concentrations used (Figure 9).

0.50

1.00

0 0.25

0 3.00

0 9.00

15.00

30.00

PBDE-47 concentration (μ/L)

0.808

Figure 9. Number of dead clams for each PBDE-47 concentration used in the 96-hour toxicity tests.

Table II. Estimation of expected mortality of the bivalve molluscs used in the study (96-h Test). Concentration (PBDE-47 µg/L)

Mortality

Student’s t test

P value

0.25 0.5 1 3 9 15 30

0.00% 1.67% 1.67% 0.00% 1.67% 1.67% 0.00%

0.000 0.935 0.935 0.000 0.935 0.935 0.000

1.000 0.350 0.350 1.000 0.350 0.350 1.000

Veterinaria Italiana 2013, 49 (1), 69-77

Confidence interval (95%) Lower limit Upper limit -3.50% 3.50% -1.84% 5.17% -1.84% 5.17% -3.50% 3.50% -1.84% 5.17% -1.84% 5.17% -3.50% 3.50%

Toxic effects of PBDE-47 on the clam Chamelea gallina

Angioni et al.

Figure 10. Histological section of the male gonad without alteration and with good production of gametes of Chamelea gallina exposed to 15 μg PBDE-47/L. (A) 100X; (B) 400X. A

Figure 11. Histological section of female mature gonad without alteration of Chamelea gallina exposed to 15 μg PBDE-47/L. (A) 100X; (B) 400X.

Number of dead clams

Table III. Univariate analysis of variance (Anova) method (21day -Test). 2

Source 1

0 0.25

0 0.50

1.00

3.00

0 9.00

15.00

30.00

PBDE-47 concentration (μ/L)

Sum of Degree Mean of squares free squares Fisher’s F P value

Concentration

0.157

0.026

Residues

8.650

413

0.021

Total

8.807

419

1.250

0.279

Figure 12. Number of dead clams for each PBDE-47 concentration in the 21-day toxicity tests. The study conducted at the concentration of 0.5 µg/L was terminated after 14 days. Table IV. Estimation of expected mortality of the bivalve molluscs used in the study (21-day test*). Concentration (PBDE-47 µg/L)

Mortality

Student’s t test

P value

0.25 0.5 1 3 9 15 30

0.00% 5.00% 3.33% 0.00% 3.33% 3.33% 0.00%

0.000 1.892 1.262 0.000 1.262 1.262 0.000

1.000 0.059 0.208 1.000 0.208 0.208 1.000

Confidence interval (95%) Lower limit Upper limit -5.19% 5.19% -0.19% 10.19% -1.86% 8.53% -5.19% 5.19% -1.86% 8.53% -1.86% 8.53% -5.19% 5.19%

* The study conducted at the concentration of 0.5 µg/L was terminated after 14 days

Veterinaria Italiana 2013, 49 (1), 69-77

Angioni et al.

Toxic effects of PBDE-47 on the clam Chamelea gallina

Table V. Analytical values of the lipid fraction and the PBDE-47 concentrations in the tissues at the end of each 21-day toxicity test*. PBDE-47 H2O (ng/L) 250 500 1,000 3,000 9,000 15,000 30,000

Control lipids (%) 1.07 0.84 1.03 1.08 1.16 0.77 0.90

Control PBDE-47 (ng/kg) 119 130 118 118 96.0 116 120

Control PBDE-47 Exposed Exposed Exposed PBDE-47 Control PBDE-47 Exposed PBDE-47 (ng/kg lipids PBDE-47 (ng/kg (ng/kg fat) (ng/kg fat) dry weight) (%) (ng/kg) dry weight) 613 11,200 0.97 110 x 103 566 x 103 11,300 x 103 3 3 670 15,500 0.73 97.3 x 10 501 x 10 13,300 x 103 608 11,500 1.14 122 x 103 627 x 103 10,700 x 103 3 3 608 10,900 1.14 394 x 10 2,030 x 10 34,600 x 103 495 8,300 1.05 598 15,200 0.90 3 3 619 13,300 0.95 5,870 x 10 30,300 x 10 618,000 x 103

* The study conducted at the concentration of 500 ng/L was terminated after 14 days.

The PBDE-47 levels found at the end of the toxicity test in the control groups were negligible compared with those detected in the exposed individuals, with values of between 96 ng/kg and 130 ng/kg. These values are comparable to, and sometimes lower than, those found in monitoring studies conducted on some species of bivalve molluscs in the European and Asian regions (8, 16, 19, 20, 25). Analysis of the clams not exposed to the contaminant showed a contamination profile (ratios of the nine congeners determined) comparable with those found in other aquatic organisms and reported in the literature (11, 19). In a bioaccumulation study conducted on mussels, it was found that the abundance of congeners 47 and 99 in the tissues may be associated with more efficient mechanisms of intake and accumulation of the two congeners compared with other PBDEs (20).

Although the PBDE-47 concentrations used in the toxicity tests were much higher than those found in particularly polluted areas (the maximum values reported in the literature are approx. 0.5 ng/L), the study demonstrated the ability of the clam to bioconcentrate the contaminant in question, with the consequent possibility of its introduction into the food chain (16). Bioconcentration is the result of direct intake of a chemical substance by an organism through water. The bioconcentration factor (BCF) is defined as the ratio, at the steady state (absorption rate equal to elimination rate), between the concentration of contaminant in the aquatic organism and the corresponding concentration in the aquarium water.

The PBDE-47 levels in the tissue of the clams were 1,000 to 50,000 times higher in the exposed groups than the corresponding controls (Table V). These levels showed a slight, though non-linear, increase as the concentration of the toxic substance in water increased. On a wet-weight basis, the contamination levels were between 110 x 103 ng/kg in the clams exposed to the minimum concentration of 250 ng/L of PBDE in water, and 5,870 x 103 ng/kg for those exposed to the maximum concentration of 30,000 ng/L.

Co = concentration of contaminant (ng/kg) in the aquatic organism;

In different exposure studies relating to aquatic organisms and persistent organic contaminants such as PCB and PBDE, a linear ratio was identified between the quantity of contaminant determined in the tissue and the hydrophobic characteristics of the compound tested, the latter being described by the logarithm of its octanol-water partition coefficient (log kow). The data in the literature give a log kow value between 6.0 and 6.8 (10, 12) for PBDE-47. A high log kow value indicates that the compound has marked hydrophobic characteristics and is potentially hazardous because it easily accumulates in the adipose tissues of clams by means of passive diffusion mechanisms.

Veterinaria Italiana 2013, 49 (1), 69-77

BCF = Co/Cw where:

Cw = concentration of contaminant (ng/L) in water. The Co value can be expressed on a wet-weight, dryweight and lipid basis, giving rise to three factors (BCFW, BCFL and BCFD) which can differ considerably in numerical terms. In Europe, chemical substances with a BCFW value greater than 100 are considered to have high potential to bioaccumulate and are classified as â&#x20AC;&#x2DC;dangerous to the environmentâ&#x20AC;&#x2122;, because they could impair the health of an aquatic organism or of predators feeding on that organism (9). In the present study, the experimental data on the PBDE47 levels determined in the clam tissues were used to calculate a lower estimate of the value of BCF. The lower estimate was calculated because the experimental protocol used in the toxicity tests did not allow evaluation of whether the steady state was reached. In fact, the experimental protocol involved a single analytical determination of PBDE-47 in the tissue of the clams after a maximum exposure time of 21 days. Table VI shows the BCF values calculated at the different exposure concentrations. In all cases the value of BCFw was greater than 100.

Toxic effects of PBDE-47 on the clam Chamelea gallina

Angioni et al.

Table VI. BCF values calculated at the different exposure concentrations. Concentration PBDE-47 H2O (ng/L) 250 500 1,000 3,000 30,000

BCFW BCFD BCFL log log log (L/kg) (L/kg) (L/kg) BCFW BCFD BCFL 439 195 122 131 196

2,260 1,000 630 680 1,010

45,300 26,600 10,700 11,500 20,600

2.6 2.3 2.1 2.1 2.3

3.4 3.0 2.8 2.8 3.0

4.7 4.4 4.0 4.1 4.3

BCFW = bioconcentration factor on wet weight basis; BCFL = bioconcentration factor on fat basis; BCFD = bioconcentration factor on dry weight basis.

No experimental studies of bioaccumulation of PBDE-47 in Chamelea gallina are reported in the literature, but such studies have been conducted on other bivalve molluscs, such as Mytilus edulis (10, 19). The BCF values in these organisms are higher than those found in the present study. This difference may be associated with: • ammonia concentration (NH3): < 0.25 mg/L; • failure to reach the steady state due to the short exposure time; • inappropriate comparison between biologically different organisms. To confirm this hypothesis, a mathematical model (16) was used which relates the log kow of a given substance to the BCF value in Mytilus edulis: log BCFW = 0.858 log kow - 0.808 The equation enables the bioaccumulation factor of the mussel to be estimated once the log kow value of a substance is known. The linearity equation used

would lead to a log BCF result of 4.3, a higher value than that experimentally estimated in this study for Chamelea gallina.

Conclusion Under the operating conditions optimised in this study, Chamelea gallina showed good adaptability in the aquarium, allowing the correct conduct of ecotoxicology tests in the laboratory. The results of the 96-hour and 21-day toxicity tests did not indicate any correlation between the concentration of PBDE-47 in the seawater in the aquariums and the mortality of Chamelea gallina. In fact, exposure to the maximum concentration of PBDE-47 did not give rise to significant mortality. The results of the histological tests did not indicate any evident alterations in the tissue of the reproductive apparatus under the experimental conditions used. The chemical analyses indicated that clams exposed to PBDE-47 presented levels of contaminant in the tissue thousands of times higher than unexposed clams, depending on the concentration in the aquatic ecosystem, thus demonstrating the ease with which Chamelea gallina bioconcentrates the contaminant. The entry of PBDE-47 into the food chain may constitute a risk to food safety, the quality of marine ecosystems and the health of clam stocks.

Acknowledgements Our thanks are extended to Mr. Alfonso de Benedictis for the expert technical support.

References 1. Aarab N., Lemaire-Gony S., Unruh E., Hansen P.D., Larsen B.K., Andersen O.K. & Narbonne J.F. 2006. Preliminary study of responses in mussel (Mytilus edulis) exposed to bisphenol A, diallyl phthalate and tetrabromo diphenyl ether. Aquat Toxicol, 78S, S86-S92.

5. Delgado M. & Pérez-Camacho A. 2007. Comparative study of gonadal development of Ruditapes philippinarum (Adams and Reeve) and Ruditapes decussatus (L.) (Mollusca: Bivalvia): Influence of temperature. Scientia Marina, 71, 471-484.

2. Angioni S.A., Giansante C. & Ferri N. 2010. Vongola (Chamelea Gallina): valutazione degli effetti dei solidi sospesi in acqua marina nel mollusco bivalve. Vet Ital, 46, 93-99.

6. De Wit C. A. 2002. An overview of brominated flame retardants in the environment. Chemosphere, 46, 583-624.

7. EFSA Panel on Contaminants in the Food Chain (CONTAM), 2011. Scientific Opinion on Polybrominated Diphenyl Ethers (PBDEs) in Food. EFSA Journal, 9(5): 2156 (www.efsa.europa.eu/efsajournal accessed on 23 January 2012). 8. Fernandes A., Mortimer D.N., Gem M., Dicks P., Smith F., White S. & Rose M. 2008. Brominated contaminants (PBDD/Fs and PBDEs) in shellfish. Organohalogen Compounds, 70, 1024-27. 9. Geyer H.J., Rimkus G.G., Scheunert I., Kaune A.,• Schramm K-W., Kettrup A., Zeeman M., Muir D.C.G.,

Veterinaria Italiana 2013, 49 (1), 69-77

Angioni et al.

Hansen L.G. & Mackay D. 2000. Bioaccumulation and occurrence of Endocrine-Disrupting Chemicals (EDCs), Persistent Organic Pollutants (POPs), and other organic compounds in fish and other organisms including humans. In: The Handbook of Environmental Chemistry, vol. 2 Part J, Bioaccumulation, (B. Beek, eds). Springer-Verlag, Berlin Heidelberg, 1-166. 10. Gustafsson K., Bjork M., Burreau S. & Gilek M. 1999. Bioaccumulation kinetics of brominated flame retardants (polybrominated diphenyl ethers) in blue mussels (Mytilus edulis). Environ Toxicol Chem, 18, 1218-1224. 11. Kuiper R.V., Vethaak A.D., Cantón R.F., Anselmo H., Dubbeldam M., van den Brandhof E-J, Leonards P.E.G., Wester P.W., van den Berg M. 2008. Toxicity of analytically cleaned pentabromodiphenylether after prolonged exposure in estuarine European flounder (Platichthys flesus), and partial life-cycle exposure in fresh water zebrafish (Danio rerio). Chemosphere, 73, 195-202. 12. Kuramochi H., Maeda K. & Kawamoto K. 2007. Physicochemical properties of selected polybrominated diphenyl ethers and extension of the UNIFAC model to brominated aromatic compounds. Chemosphere, 67, 1858-1865. 13. Lema S.C., Schultz I.R., Scholz N.L., Incardona J.P., Swanson P. 2007. Neural defects and cardiac arrhythmia in fish larvae following embryonic exposure to 2,2’,4,4’-tetrabromodiphenyl ether (PBDE 47). Aquat Toxicol, 82, 296-307. 14. Liu Y., Zheng G.J., Hongxia Y., Martin M., Richardson B.J, Lam M.H.V. & Lam P.K.S. 2005. Polybrominated diphenil ethers (PBDEs) in sediments and mussel tissue from Hong Kong marine waters. Mar Pollut Bull, 50, 1173-1184.

Toxic effects of PBDE-47 on the clam Chamelea gallina

17. McDonald T.A. 2002. A perspective on the potential health risk of PBDEs. Chemosphere, 46, 745-755. 18. Mengoli A. 1998. Aspetti morfo-funzionali dei mitili. Laguna, 4,13 -9. 19. Mizukawa K., Takada H., Takeuchi I., Ikemoto T., Omori K. & Tsuchiya K. 2009. Bioconcentration and biomagnification of polybrominated diphenyl ethers (PBDEs) through lower-trophic-level coastal marine food web. Mar Pollut Bull, 58, 1217-1224. 20. Moon H-B., Kannan K., Lee S-J. & Choi M. 2007. Polybrominated diphenyl ethers (PBDEs) in sediment and bivalves from Korean coastal waters. Chemosphere, 66, 243-251. 21. Orban E., Di Lena G., Nevigato T., Casini I., Caproni R., Santaroni G. & Giulini G. 2006. Nutritional and commercial quality of the striped venus clam, Chamelea gallina, from the Adriatic sea. Food Chemistry, 101, 1063-1070. 22. Palm A. Cousins I.T., Mackay D., Tysklind M., Metcalfe C. & Alaee M. 2002. Assessing the environmental fate of chemicals of emerging concern: a case study of polybrominated diphenyl ethers. Environ Pollut, 117, 195-213. 23. Shin P.K.S., Yau F.N., Chow S.H., Tai K.K. & Cheung S.G. 2002. Responses of the green-lipped mussel Perna viridis (L) to suspended solids. Mar Pollut Bull, 45, 157-162. 24. Siddiqi M.A., Laessig R.H. & Reed K.D. 2003. Polybrominated Diphenyl Ethers (PBDEs): New Pollutants - Old Diseases. Clinical Medical & Research, 1, 281-290. 25. Umlauf G., Christoph E.H., Huber T., Mariani G., Mueller A., Skejo H. & Wollgast J. 2009. PBDEs in water, sediments and biota of the river Danube from Germany to the Black Sea. Organohalogen Compd, 71, 737-742.

15. Lucchetti A. 2003. La vongola. Biologia, pesca e consumo delle più importanti specie commerciali. La vongola Chamelea gallina (Linneo, 1758). Il Pesce, 6, 135 (http:// www.pubblicitaitalia.com/ilpesce/2003/6/4963.html accessed on 26 January 2012).

26. Unione Europea, 2011. Direttiva 2011/65/UE del Parlamento Europeo e del Consiglio dell’8 giugno 2011 sulla restrizione dell’uso di determinate sostanze pericolose nelle apparecchiature elettriche ed elettroniche (Official Journal L 174 of 1.7.2011).

16. Mai B.-X., Luo X.-J., Yu M. & Chen S.-J. 2009. PBDEs in water and aquatic biota of the Pearl river estuary, South China. Organohalogen Compd, 71, 807-12.

27. Watanabe I. & Sakai S. 2003. Environmental release and behaviour of brominated flame retardants. Environ Int, 29, 665-682.

Veterinaria Italiana 2013, 49 (1), 69-77

Vaccino inattivato e adiuvato per il controllo delle infezioni da sierotipo 9 del virus della peste equina: valutazione dell’efficacia in cavallo e cavia Rossella Lelli1, Umberto Molini2, Gaetano Federico Ronchi2, Emanuela Rossi2, Paola Franchi2, Simonetta Ulisse2, Gisella Armillotta2, Sara Capista2, Siegfried Khaiseb3, Mauro Di Ventura2, Attilio Pini2 Istituto Zooprofilattico Sperimentale della Sicilia ‘A. Mirri’, Via Gino Marinuzzi, 3, 90129 Palermo, Italia Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise ‘G. Caporale’, Campo Boario, 64100 Teramo, Italia a.pini@izs.it 3 Central Veterinary Laboratory, 24 Goethe Street, Windhoek, Namibia 1

Parole chiave Cavia, Challenge, Namibia, Peste equina, Risposta anticorpale, Sierotipo 9 del virus della peste equina, Vaccino inattivato.

Riassunto La peste equina (PE) è una malattia virale non contagiosa dei solipedi trasmessa da insetti vettori appartenenti al genere Culicoides. La malattia è endemica in numerose regioni dell’Africa e passate esperienze hanno evidenziato come l’Italia sia un paese esposto alle malattie infettive emergenti, endemiche in Africa. Un’incursione del virus della PE unitamente alla presenza del vettore Culicoides potrebbero essere causa di una emergenza epidemica. Onderstepoort Biological Products (OBP) commercializza un vaccino vivo attenuato contenente 7 dei 9 sierotipi virali, i sierotipi 5 e 9 sono esclusi dal vaccino. L’uso di tali vaccini è controverso, pertanto, da diversi anni, vengono condotti studi su prodotti inattivati o ricombinanti. Poiché la ricerca sui vaccini PE è ostacolata dalla necessità di utilizzare il cavallo per la valutazione dell’immunogenicità, in un precedente esperimento è stata studiata la risposta sierologica ai sierotipi 5 e 9, in cavie e cavalli. Nelle due specie animali è stata osservata una risposta durevole e sovrapponibile. Nel presente studio sono state saggiate per un periodo di 12 mesi le risposte sierologiche ottenute in cavie e cavalli immunizzati con il vaccino inattivato formulato con il sierotipo 9. Le risposte sierologiche nelle due specie animali si sono confermate sovrapponibili. Al challenge dell’immunità i cavalli sono risultati protetti dall’infezione e dalla malattia. Veterinaria Italiana 2013, 49 (1), 79-88

Introduzione La peste equina (PE) è una malattia virale non contagiosa dei solipedi causata da nove sierotipi. Il virus appartiene al genere Orbivirus, famiglia Reoviridae. Approssimativamente 30 delle oltre 1.500 specie di Culicoides sono ritenute capaci di trasmettere gli Orbivirus. Il vettore più importante della PE è Culicoides imicola, una specie comune in Africa e Sud Est Asiatico (20). La mallattia è endemica in numerose regioni dell’Africa Sub Sahariana ma è stata anche registrata in Africa del Nord (1965), Spagna (1966 e 1987-1990) e Medio Oriente (1959, 1961 e 1989). La diffusione dei sierotipi 4 o 9 è stata causa di epidemie nella penisola Iberica e nel Medio Oriente. Dal 1975 la malattia ha fatto la sua comparsa anche in Africa occidentale: Nigeria, Senegal e Mauritania (11). Gli eventi dell’ultimo decennio hanno dimostrato come l’Italia sia un paese a rischio di malattie emergenti endemiche in Africa e la presenza di vettori del genere Culicoides rende un’incursione del virus della

PE un evento possibile. Vi è quindi necessità di avere prodotti immunizzanti innocui ed efficaci, rapidamente disponibili, unitamente a prodotti diagnostici sensibili e specifici. Il vaccino attenuato prodotto da OBP potrebbe essere disponibile in breve tempo. Il prodotto, comunemente utilizzato nelle regioni dove la malattia è endemica, contiene 7 dei 9 sierotipi ed è commercializzato in due formulazioni: trivalente, contenente i sierotipi 1, 3, 4, e quadrivalente, contenente i sierotipi 2, 6, 7 e 8. Le due preparazioni sono somministrate a distanza di 21 giorni l’una dall’altra. Il sierotipo 5 è stato escluso dalla formulazione nel 1990 in seguito a patogenicità residua riscontrata sul campo. Il sierotipo 9 non è incluso nel vaccino poiché era considerato epidemiologicamente non rilevante in Sud Africa. Entrambi i sierotipi hanno, tuttavia, dominato nei focolai di PE occorsi dal 2006, soprattutto nella provincia Sud Africana del Capo Occidentale (18). In Namibia, il sierotipo 9 può essere isolato nella forma polmonare, cardiaca e blanda di malattia. Febbre,

Lelli et al.

Efficacia del vaccino inattivato per il sierotipo 9 del virus della peste equina

edema della fossa sopraorbitale, difficoltà respiratorie, inappetenza e riluttanza al movimento sono sintomi comunemente osservati (Di Mattia T. e Scacchia M., 2008, comunicazione personale). Tenendo in considerazione le obiezioni che riguardano l’uso di vaccini vivi attenuati, nel corso degli anni sono state oggetto di ricerca soluzioni alternative. La ricerca si è focalizzata sull’allestimento di vaccini inattivati e adiuvati (5, 9, 12, 13) e, più recentemente, sui vaccini ricombinanti (1, 4, 7, 8, 10, 17, 19). I risultati riportati in letteratura indicano che tali formulazioni sono capaci di proteggere i cavalli dalla PE ma il challenge è stato effettuato solo a un breve intervallo di tempo dall’immunizzazione a 76-98 giorni. I promettenti risultati ottenuti con i vaccini ricombinanti devono essere ancora confermati. I limiti imposti dalla necessità di usare i cavalli per saggiare l’immunogenicità di vaccini per la PE sono stati di stimolo per valutare la possibilità di ricorrere alla cavia come modello predittivo di efficacia. In una recente pubblicazione è stato dimostrato come topi knock-out (IFNAR -/-), immunizzati con il vaccino modificato Ankara che esprime la proteina VP2 del sierotipo 4, possano essere un modello per la valutazione di efficacia del vaccino (2). In un precedente esperimento (14) varie formulazioni di vaccini monovalenti inattivati e adiuvati, preparati con i sierotipi 5 e 9, sono state saggiate per verificare la loro capacità di indurre anticorpi neutralizzanti in cavia. Sulla base dei risultati ottenuti, una delle formulazioni prodotte con il sierotipo 9 è stata inoculata in tre cavalli. Sieroconversione perdurante per oltre 10 mesi dall’immunizzazione è stata registrata in entrambe le specie. I risultati sembrano dimostrare l’utilità della cavia come modello predittivo per evidenziare le proprietà antigeniche dei vaccini inattivati. Al fine di confermare l’efficacia del vaccino e la validità della cavia come modello è stato deciso di condurre il challenge dell’immunità su cavallo. L’esperimento riportato è stato condotto con l’approvazione dei Servizi Veterinari Namibiani in un’area esente da circolazione virale. Il challenge dell’immunità è stato condotto a 109 e 365 giorni dalla vaccinazione in una stalla a prova di insetto. Nel periodo precedente al challenge sono state studiate l’innocuità del prodotto e l’andamento della risposta anticorpale. La temperature corporea, i segni clinici e la viremia sono stati rilevati nel periodo post challenge. Nelle cavie la reattività sierologica è stata saggiata fino al giorno 365.

Materiali e Metodi La produzione del vaccino sierotipo 9 è stata descritta nel precedente lavoro (14). Per una più agevole lettura ne vengono riassunti i dati. Il vaccino è stato utilizzato dopo cinque mesi dalla data di produzione.

Linee cellulari Cellule BHK21 (clone 13), fornite da European Collection of Cell Cultures (ECACC), sono state utilizzate per la produzione del vaccino. Cellule VERO (ECACC) sono state usate per la determinazione dei titoli virali, l’esecuzione delle prove di sieroneutralizzazione e l’isolamento virale.

Virus Il virus PE sierotipo 9 impiegato per la produzione del vaccino, proveniente da AHS reference antigens (Bob Swanepoel collection) è stato gentilmente fornito dal dott. Hübschle†. Dopo amplificazione, purificazione e concentrazione, la sospensione virale è stata controllata in accordo a quanto descritto dalla Farmacopea Europea (3). La tipizzazione virale è stata condotta tramite neutralizzazione di tutti i 9 sierotipi del virus PE (14). Lo stesso ceppo virale utilizzato per la produzione del vaccino è stato impiegato per l’esecuzione del challenge a 109 giorni dalla vaccinazione. Un ceppo di campo isolato da un focolaio di PE occorso, nel 2008, in Namibia è stato usato per il challenge effettuato a 365 giorni dalla vaccinazione (16). I due ceppi virali, al secondo passaggio in BHK21, liofilizzati, controllati per purezza e conservati a 5°C ± 3°C hanno mostrato, rispettivamente, un titolo infettante pari a 106,9 e 106,8 TCID50/ml. Al momento del challenge, il virus è stato ricostituito con acqua distillata, diluito in terreno MEM fino ad avere un titolo pari a 105 TCID50/ml.

Inattivante Il processo di inattivazione è stato effettuato con BEI 5mM come descritto da Ronchi et al. (14).

Adiuvanti L’adiuvante ISA27VG (SEPPIC Italia S.r.l., Milano, Italia) e precedentemente saggiato nei cavalli per la valutazione delle proprietà infiammatorie, è stato emulsionato con la sospensione virale concentrata purificata e inattivata in accordo alle istruzioni fornite dalla ditta produttrice (14). Un quantitativo pari a 0,3 mg di saponina (Sigma Aldrich, St. Louis, MO USA) per dose vaccinale è stato aggiunto prima dell’utilizzo.

Cavalli Le procedure seguite per l’esecuzione degli esperimenti di vaccinazione, il successivo challenge sui cavalli e le linee guida per la cura e il mantenimento degli animali sono state approvate dal Central Veterinary Laboratory (CVL) Animal Ethics Committee. Le fasi di vaccinazione, inoculazione, salasso e monitoraggio dei cavalli sottoposti a challenge, quando

Veterinaria Italiana 2013, 49 (1), 79-88

Lelli et al.

richiesto, sono state svolte con la supervisione di un medico veterinario. La prova di efficacia del vaccino è stata eseguita presso la Bergvlug Veterinary Research Farm del Namibian Ministry of Agriculture, Water and Forestry, nel distretto di Windhoek. L’azienda che copre una superficie di 5.000 ha, si trova a 1.932 m sopra il livello del mare, le temperature notturne variano tra -8°C in inverno e 8°C in estate, al suo interno si trovano bovini, ovini, cavalli, antilopi e predatori. Nell’azienda, utilizzata nei primi del ‘900 dalle truppe tedesche per proteggere i cavalli dalla malattia, non si registrano focolai di PE. Quindici animali di età superiore ai due anni, costituenti un gruppo omogeneo sulla base dell’esame clinico, del profilo biochimico e della conta completa degli elementi corpuscolati del sangue sono stati acquistati a Karas, a 780 Km da Windhoek, dove la vaccinazione per la PE non è praticata. I sieri degli animali sono stati saggiati prima dell’acquisto e quindici giorni dopo il loro arrivo a Bergvlug con un’ELISA competitiva (c‑ELISA) per la verifica di assenza di anticorpi alla PE. Il giorno della vaccinazione (T0) un ulteriore campione di sangue è stato prelevato e saggiato mediante sieroneutralizzazione (SN) per escludere la presenza di anticorpi nei confronti dei 9 sierotipi virali. Il vaccino è stato somministrato per via intramuscolare ai 10 cavalli sul terzo medio del lato sinistro del collo con una dose pari ad 1 ml. La temperatura corporea e la reazione infiammatoria al punto di inoculo sono state monitorate e registrate nei 15 giorni successivi alla prima e seconda vaccinazione. La temperatura è stata rilevata due volte al giorno, mattina e sera. La dose di richiamo, pari a 1 ml, è stata somministrata al livello del terzo medio del lato destro del collo a 35 giorni dalla prima somministrazione (T35). Prelievi di sangue dalla giugulare sono stati effettuati, a T15, T21, T35, T42, T49, T63, T109, T175, T220, T250, T294, T321, e T365, al fine di monitorare la cinetica della risposta anticorpale tramite SN e indice sieroneutralizzante (ISN). I cavalli sono stati tenuti al pascolo fino al momento del challenge quando sono stati trasferiti in un locale a prova di insetto provvisto di porta e finestre schermate con zanzariere a prova di Culicoides, impregnate ogni giorno con insetticida Fendona (BASF Agricultural Products Ltd, Città del Capo, Sud Africa).

Cavalli sentinella I cinque cavalli non vaccinati sono stati impiegati inizialmente come animali sentinella per escludere circolazione virale. Il giorno 109 il numero dei cavalli sentinella si è ridotto a 4 in quanto uno degli animali è stato utilizzato come controllo nel primo challenge. I soggetti sentinella sono stati tenuti al pascolo assieme ai soggetti vaccinati. I sieri prelevati con cadenza mensile da questi animali sono stati saggiati con c-ELISA.

Veterinaria Italiana 2013, 49 (1), 79-88

Efficacia del vaccino inattivato per il sierotipo 9 del virus della peste equina

Cavie L’esperimento è stato condotto in accordo alla Legge Italiana 116/92. Al fine di eseguire l’esperimento su cavie è stata inviata una comunicazione scritta al Ministero della Salute, Dipartimento per la Sanità Pubblica Veterinaria, la Nutrizione e la Sicurezza degli Alimenti; Ufficio VI, Roma, Italia. Quattro cavie femmina, di peso compreso tra 350 e 500 g, sono state inoculate con 0,5 ml di vaccino nella zona prescapolare. Una dose di richiamo è stata somministrata 35 giorni dopo. Gli animali sono stati sottoposti a prelievo di sangue tramite puntura cardiaca, effettuata previa anestesia totale, con la stessa tempistica utilizzata per i cavalli. I sieri ottenuti sono stati saggiati in ISN. L’esperimento è stato condotto negli stabulari autorizzati dell’Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise “G. Caporale” di Teramo, Italia. Tutti gli animali utilizzati sono stati mantenuti in strutture di ricovero dotate di aria condizionata (tra 20°C e 24°C), con umidità relativa pari al 50-55%, con acqua e cibo ad libitum. Gli animali sono stati controllati quotidianamente tramite esame clinico fino alla fine dell’esperimento.

Sierologia I test c-ELISA sono stati effettuati utilizzando il kit commerciale Ingezim AHSV Compac Plus (INGENASA, Madrid, Spagna). Il metodo per calcolare l’ISN ottenuto attraverso diluizioni seriali in base 10 del virus contro il siero inattivato diluito 1:10 è stato descritto in precedenza (14). Le prove di SN attraverso diluizioni al raddoppio del siero contro 100 TCID50 di virus (6) sono state eseguite in parallelo. Gli ISNs sono stati espressi come log10, il titolo SN come reciproco del fattore di diluizione neutralizzante il 50% dell’attività virale.

Test di immunogenicità L’efficacia del vaccino è stata testata al T109 e T365 dopo la prima somministrazione di vaccino. Come già affermato, lo stesso virus utilizzato per la formulazione del vaccino inattivato è stato usato nel challenge nel giorno 109 dalla vaccinazione mentre il sierotipo 9, isolato in Namibia nel 2008, è stato utilizzato nel challenge nel giorno 365 dalla vaccinazione. In entrambi i casi sono stati inoculati per via intravenosa 105 TCID50 dei rispettivi virus. Nella prima prova sono stati utilizzati 5 cavalli vaccinati. Nel secondo challenge il numero di cavalli impiegati è stato pari a 4 in quanto il quinto cavallo era ingestibile. Non è stato possibile spostare l’animale nella stalla a prova di insetto senza che si corressero rischi per gli operatori. L’uso di un solo controllo per ognuno dei due challenge è stato dettato da ragioni di benessere animale, nel rispetto della logica delle 3R: Refinement, Replacement, Reduction (15).

Lelli et al.

Efficacia del vaccino inattivato per il sierotipo 9 del virus della peste equina

L’esame clinico e la rilevazione della temperatura corporea sono stati effettuati due volte al giorno per 21 giorni dopo l’effettuazione della prova di challenge. Valori ≤ 38°C sono stati considerati fisiologici. Campioni di sangue con anticoagulante sono stati prelevati quotidianamente per 28 giorni. La viremia è stata valutata attraverso la tecnica dell’isolamento virale su cellule VERO e la RT-PCR, in accordo con quanto descritto nell’OIE Manual of Diagnostic Tests and Vaccines for Terrestrial Animals (21). Campioni negativi per effetto citopatico al primo passaggio sono stati ulteriormente saggiati effettuando tre passaggi seriali della durata di 7 giorni ciascuno. Al terzo passaggio le cellule sono state rimosse ed esaminate in immunofluorescenza per escludere la presenza di virus PE.

Cattura di Culicoides Trappole luminose per insetti prodotte da Onderstepoort Veterinary Institute sono state collocate nei pressi della zona di pascolo nella stagione secca e nella stagione delle pioggie.

to la formazione di un piccolo nodulo, del diametro massimo di 5 mm, perdurante per tre giorni. In nove animali il nodulo non è risultato né dolente né caldo. Solo l’animale n. 4 ha manifestato un nodulo caldo nei giorni 36 e 37 dalla vaccinazione.

Temperatura corporea La temperatura corporea è rimasta nel limite dei valori fisiologici fino al giorno 35 dalla vaccinazione. In seguito alla somministrazione della seconda dose, nel giorno 36, si è avuto un rialzo termico, tra 38,2°C e 38,8°C, in 4 dei 10 cavalli (Tabella 1). Durante il periodo di osservazione non sono stati rilevati altri segni clinici. Il rialzo termico nei 4 cavalli ha coinciso con la reazione infiammatoria osservata nei giorni 36 e 37 successivi alla somministrazione della dose di richiamo.

Risposta sierologica La reattività sierologica al vaccino è stata registrata nei cavalli sino al giorno dell’effettuazione del challenge.

ISNs nei cavalli

Risultati Reattività alla vaccinazione Reazione infiammatoria al punto di inoculo Dopo la somministrazione della prima dose di vaccino è risultato visibile, al punto di inoculo, un lieve ispessimento della cute che è regredito dopo 48 ore dall’ inoculazione. Al T35 la dose di richiamo ha indot-

Gli ISNs al T15 hanno mostrato una variazione, tra 2,10 e 4,27, in 8 cavalli. Gli animali n. 6 e n. 7 hanno reagito al T21 con ISNs pari a 0,7 e 1,5, rispettivamente. Dopo la dose di richiamo, al T42, gli ISNs hanno raggiunto valori ≥ 6,8 per rimanere costanti durante l’intero periodo di osservazione con l’eccezione dei cavalli n. 6 e n. 9 che hanno mostrato ISNs pari a 2,27 e 0,83 al T109 e T365, rispettivamente. La cinetica della risposta immunitaria è riportata in Tabella 2 e in Figura 1. Gli ISN ≤ 0,5 al T0 sono stati omessi.

Cavallo n. 1 2 3 4 5 6 7 8 9 10 11 controllo 12 controllo 13 controllo 14 controllo 15 controllo

T1 a 35

T36

T37

T38

T39

T40

T41

T42

T43

T44

Temperatura fisologica ≤ 38 °C

Tabella I. Temperatura corporea nei cavalli in seguito a somministrazione della dose di richiamo (T35 ).

37,1 37,8 36,6 38,4 38,4 38,2 38,8 37,9 37,7 37,8 37 37,2 37,1 37,6 35,9

37,2 37,2 36,9 37,3 37,6 37,2 37,9 36,3 37,1 36,2 36,6 37,3 37 36,5 36

36,8 36,9 36,9 37,2 37,5 37 37,6 36,4 37,2 36,8 37 37 37,3 36,5 35,9

36,7 36,9 36,5 37,1 37,8 37,1 37,6 36,5 36,8 36,9 37,1 37 37,1 36,1 35,8

36,5 36,9 36,5 36,1 36,7 36,9 37 36,6 36,8 36,3 36,8 37 36,1 36,1 35,9

36,9 37 36,4 36,6 36,9 37,3 37,1 36,9 36,9 36,8 37,1 36,9 36,9 36,8 36,1

36,1 36,3 36,1 36,5 36,5 36,5 37 36,6 36,7 36,9 36,5 36,8 36,9 36,6 35,7

36,2 36,8 36,5 36,8 36,8 37,1 36,4 36,9 36,7 37,2 36,3 36,8 36,8 36,1 35,6

36,1 36,5 36,5 36,2 36,4 36,6 36,7 36,9 36,1 36,7 36,8 36,4 36,4 36,9 35,9

Veterinaria Italiana 2013, 49 (1), 79-88

Lelli et al.

Efficacia del vaccino inattivato per il sierotipo 9 del virus della peste equina

Tabella II. Indici sieroneutralizzanti nei cavalli e nelle cavie . T15

T21

T28

T35(a)

T42

T49

T63

T109(b)

T175

T220

T250

T294

T321

T365(c)

Cavallo n° 1

4,27

4,00

5,97

6,80

Cavallo n° 2

3,10

3,33

5,97

6,80

Cavallo n° 3

3,93

3,50

5,97

6,80

Cavallo n° 4

3,85

3,33

5,97

6,80

Cavallo n° 5

0,00

0,70

0,67

0,34

6,80

2,27

Cavallo n° 6

3,10

3,50

5,97

6,80

Cavallo n° 7

0,00

1,50

2,37

5,97

6,80

Cavallo n° 8

2,27

2,50

5,97

6,80

Cavallo n° 9

3,10

3,50

5,97

6,80

0,83

Cavallo n° 10

2,10

1,50

5,97

6,80

Cavia n° 1

1,5

3,66

Cavia n° 2

2,66

3,16

5,5

4,83

4,5

4,83

4,72

4,5

4,16

4,5

2,66

5,5

4,83

4,5

4,83

4,33

4,72

4,5

4,16

3,5

Cavia n° 3

1,66

2,16

1,66

2,16

5,5

4,83

4,5

4,83

4,72

3,5

4,16

4,5

Cavia n° 4

2,33

5,5

4,83

4,5

4,83

4,67

4,72

4,5

4,16

3,5

(a) T35 dose di richiamo, (b) challenge T109, (c) challenge T365.

8,0

6,8

7,0

5,1

5,0

5,5

5,4

2,6

2,0

2,7

4,7 4,7

4,5

6,8 6,8 5,6

4,8

6,8 6,8

6,3

4,0 3,0

6,8

richiamo

6,0

Log10

6,8

4,3

4,2

4,0

2,7 2,5

2,2 1,8

1,0 0,0

Cavalli

T15

T21

T28

T35

T42

T49

T63

T109

T175

T220

T250

T294

T321

Cavie

T365

Figura 1. Media degli indici sieroneutralizzanti nelle cavie e nei cavalli.

ISNs nelle cavie

Titoli sierologici nei cavalli

La reattività sierologica al T15 è stata simile a quella registrata nei cavalli. Al T42 gli ISNs hanno raggiunto il valore massimo di 5,5 e al T365 hanno mostrato una variazione tra 3,5 e 4,5 (Tabella 2).

La Tabella 3 mostra i risultati della SN. La reattività nei cavalli n. 6 e n. 7 è risultata ritardata al T21. Un picco sierologico pari a 320 è stato registrato tra T42 e T49 per decrescere, in seguito, lentamente. I titoli sono rimasti, comunque, a livelli di positività fino al T365 dalla vaccinazione.

La cinetica della risposta immunitaria nella cavia ha mostrato un andamento parallelo a quella dei cavalli come evidenziato in Figura 1. Gli ISN al T0 sono risultati ≤ 0,5 e conseguentemente sono stati omessi.

Veterinaria Italiana 2013, 49 (1), 79-88

Lelli et al.

Efficacia del vaccino inattivato per il sierotipo 9 del virus della peste equina

Tabella III. Titoli sieroneutralizzanti nei cavalli vaccinati.

Cavallo n° 1 Cavallo n° 2 Cavallo n° 3 Cavallo n° 4 Cavallo n° 5 Cavallo n° 6 Cavallo n° 7 Cavallo n° 8 Cavallo n° 9 Cavallo n° 10

T15

T21

T28

T35(a)

T42

T49

T63

T109(b)

T175

T220

T250

T294

T321

T365(c)

40 20 40 40 5 40 5 20 40 20

40 40 40 40 5 40 10 40 40 20

160 160 160 80 20 5 80 20 80 40

80 80 160 160 5 80 40 160 80 160

320 320 320 320 80 320 320 320 320 320

320 320 320 320 320 80 320 320 320 320

320 320 320 160 40 320 320 320 320 320

160 160 160 40 20 160 160 320 160 160

160 80 80 160 80

160 80 80 160 40

160 80 80 40 40

160 80 80 20 40

(a) T35 dose di richiamo, (b) challenge T109, (c) challenge T365.

Temperatura anti-meridana

38,9

°C

39 temperatura fisiologica

37,1

37,4

39,7

39,3

38,6

38,6 38,4 37,9

37,8

37,4

37 37,4 36,6 36,7 36,7 37,3 37,0 37,1 36,9 36,9 36,8 36,5 36,4 36 36,2 36,2

37,5

Temperatura post-meridiana

T10

T11

T12

T13

T14

38,5 37,6 37,6 37,4 37,1 37,5 37,5 37,3 36,9 37,3 37,1 36,9 36,7 36,6

T15

T16

T17

T18

T19

T20

T21

Figura 2. Challenge a T109 : temperature corporee nel cavallo di controllo. 39,0 Temperatura anti-meridana (media) 38,5 38,0

Temperatura post-meridiana (media) temperatura fisiologica

°C

37,5 37,0 36,5 36,0 35,5 35,0

T10

T11

T12

T13

T14

T15

T16

T17

T18

T19

T20

T21

Figura 3. Challenge a T109 : media delle temperature corporee nei cavalli vaccinati. Le barre di errore indicano le temperature massime e minime.

Veterinaria Italiana 2013, 49 (1), 79-88

Lelli et al.

Efficacia del vaccino inattivato per il sierotipo 9 del virus della peste equina

Prove di efficacia

al movimento. Durante i 21 giorni di osservazione non sono stati rilevati segni clinici o rialzo della temperatura corporea nei cavalli vaccinati. Le Figure 2 e 3 riportano i valori della temperatura corporea anti e post-meridiana nel cavallo di controllo e nei cavalli vaccinati. Nell’animale di controllo, l’isolamento virale ha rilevato la presenza di virus dal giorno 12 al giorno 16, i titoli infettanti sono risultati variare tra 102.8 e 104.8 TCID50/ml. La RT-PCR è risultata positiva dal giorno 11 al giorno 21 dall’inoculazione. La Figura 4 riporta i risultati.

Challenge dell’immunità T109 Cinque cavalli immunizzati e un cavallo di controllo sono stati sottoposti a challenge tramite inoculazione intravenosa di 105 TCID50 particelle di virus omologo al ceppo impiegato per la produzione del vaccino. Nel cavallo di controllo si è avuto un rialzo termico, oscillante tra i 38,6°C e 40°C, nei giorni 10, 11, 12, 13, 14 e 15 dall’inoculazione. In questi giorni sono stati osservati sintomi quali edema sopraorbitale, depressione sensoriale, rifiuto del cibo e riluttanza

Nei cavalli vaccinati l’isolamento virale e la RT-PCR, condotti sui campioni di sangue nei 28 giorni seguenti al challenge, hanno dato esito negativo.

RT- PCR positiva

4,8 4,8 4,8

Log10

3,5 3

2,8

0 T0

T9 T10 T11 T12 T13 T14 T15 T16 T17 T18 T19 T20 T21 T22 T23 T24 T25 T26 T27 T28

Figura 4. Challenge a T109 : RT-PCR e titoli viremici/ml relativi al cavallo di controllo. Nei cavalli vaccinati l’isolamento virale e la RT-PCR erano negativi e non sono riportati. 41

Temperatura anti-meridana

40,1 40

°C

39,2

38,9

Temperatura post-meridiana

39,8 39

39 38,5

temperatura fisiologica

37,6

37,8 37,4 37 37,5 37,2 36,4

38,7

37,6 36,8 36,6

37,9

37,5 37,4 37,4 37,4 37,5 37,1 37,7 37,4 37,4 37,4 37,2 37,1 37,1 37 36,8 36,7

37,1 36,7

37,9 37,4

35,9

36,1

35,7 35,6

T10

T11

T12

T13

T14

T15

T16

T17

T18

T19

T20

T21

Figura 5. Challenge a T365 : temperature corporee nel cavallo di controllo. Veterinaria Italiana 2013, 49 (1), 79-88

Lelli et al.

Efficacia del vaccino inattivato per il sierotipo 9 del virus della peste equina

39,0 Temperatura anti-meridana (media) 38,5

Temperatura post-meridiana (media) temperatura fisiologica

38,0

°C

37,5 37,0 36,5 36,0 35,5 35,0

T10

T11

T12

T13

T14

T15

T16

T17

T18

T19

T20

T21

Figura 6. Challenge a T365 : media delle temperature corporee nei cavalli vaccinati. Le barre di errore indicano le temperature massime e minime. 4 RT- PCR positiva

T10 T11 T12 T13 T14 T15 T16

Log10

2,7

2,80

T17 T18 T19 T20 T21 T23

T24 T25 T26 T27 T28

Figura 7. Challenge a T365 : RT-PCR e titoli viremici/ml relativi al cavallo di controllo. Nei cavalli vaccinati l’isolamento virale e la RT-PCR erano negativi e non sono riportati.

Challenge dell’immunità T365 Quattro cavalli immunizzati e un cavallo di controllo sono stati sottoposti a challenge dell’immunità tramite inoculazione intravenosa di 105 TCID50 particelle di virus di campo isolato, nel 2008, in Namibia. Nel cavallo di controllo un primo rialzo termico della temperatura corporea è stato registrato nel giorno 6 dall’inoculazione, un secondo rialzo, compreso tra i 38,5°C e 40,1°C, nei giorni 8, 9, 10 e 11 dall’inoculazione (Figura 5). Nel periodo febbrile sono stati

osservati segni clinici paragonabili a quelli osservati nel cavallo di controllo sottoposto a challenge al T109. Il virus è stato isolato dal sangue nei campioni prelevati dal giorno 8 al giorno 11, il titolo virale infettante variava tra 102.7 e 103 TCID50/ml. La RT-PCR è risultata positiva dal giorno 7 al giorno 18 dall’inoculazione (Figura 7). Nei cavalli immunizzati non sono stati rilevati segni clinici nei 21 giorni di osservazione. Le prove di isolamento virale e RT-PCR condotti sui campioni di san-

Veterinaria Italiana 2013, 49 (1), 79-88

Lelli et al.

gue nei 28 giorni dall’inoculazione hanno confermato l’assenza virale. I due cavalli di controllo sono sopravvissuti all’infezione.

Cavalli sentinella I cavalli sentinella sono rimasti negativi per PE durante tutti i 14 mesi di osservazione.

Cattura dei Culicoides Durante la stagione estiva delle piogge il numero di Culicoides catturati nelle vicinanze della zona di pascolo è risultato compreso tra 100 e 150 unità per notte. Il numero di insetti catturati nella stagione invernale secca è risultato ridotto a circa 10.

Conclusioni Durante i 14 mesi della sperimentazione, non sono stati osservati né segni clinici né positività sierologica nei cavalli sentinella, confermando quanto già conosciuto, nei primi del ‘900, dalle autorità tedesche che avevano in uso l’azienda Bergvlug per proteggere i cavalli dalla PE. Si può ipotizzare che l’habitat, caratterizzato dall’elevata altitudine, circa 2.000 metri s.l.m., unitamente alle basse temperature non siano favorevoli alla colonizzazione di Culicoides che pur essendo presenti sono in numero tale da non avere la capacità vettoriale. Al contrario, nelle zone a più bassa altitudine, nello stesso distretto di Windhoek, i Culicoides, durante la stagione delle piogge, sono catturati a migliaia e la malattia è endemica. Il sierotipo 9 è stato utilizzato nella sperimentazione in quanto non è presente nel vaccino polivalente prodotto da OBP nonostante, nel recente passato, il sierotipo sia stato dominante nella zona del Capo Occidentale. I due test di sieroneutralizzazione non

Veterinaria Italiana 2013, 49 (1), 79-88

Efficacia del vaccino inattivato per il sierotipo 9 del virus della peste equina

hanno evidenziato differenze significative nella capacità di rilevare anticorpi, in seguito a vaccinazione, durante i dodici mesi di osservazione. Nei cavalli immunizzati i titoli sierologici sono stati rilevati già 15 giorni dopo la vaccinazione, il plateau è stato raggiunto tra i giorni 42 e 49 dalla vaccinazione. I titoli sono rimasti a livelli significativi nell’intero periodo di osservazione. La sieroconversione registrata nei cavalli a T365, 1:80, è risultata paragonabile a quella riportata in letteratura, 1:40, a 200 giorni dalla vaccinazione (5). Il vaccino preparato con il virus proveniente dalla collezione antigeni Bob Swanepoel ha dimostrato di essere efficace quando il challenge è stato eseguito a 365 giorni con un ceppo di campo recentemente isolato in Namibia. Nei cavalli vaccinati non è stata rilevata reazione febbrile o viremia, a indicare che i cavalli sono risultati protetti dall’infezione e dalla malattia e che non sono risultate differenze significative legate a un drift antigenico tra i ceppi virali utilizzati. Al contrario, i due animali di controllo utilizzati nelle prove di challenge a T109 e a T365 hanno presentato piressia, viremia e segni clinici. La reazione febbrile ha preceduto la viremia di almeno 24 ore. Entrambi gli animali sono sopravvissuti all’infezione. Nelle cavie il vaccino ha stimolato titoli anticorpali a un livello paragonabile a quello rilevato nei cavalli. Gli indici sieroneutralizzanti sono stati rilevati a partire dal quindicesimo giorno e il plateau è stato raggiunto nel giorno 42 dalla vaccinazione. I valori dei titoli sierologici sono rimasti a livelli significativi per l’intero anno di osservazione, questo aspetto conferma che la cavia potrebbe essere un indicatore delle proprietà antigeniche del prodotto in esame. In un prossimo esperimento sarà interessante comparare il modello cavia e il modello topo IFNAR -/- per la valutazione dell’efficacia dei vaccini PE.

Efficacia del vaccino inattivato per il sierotipo 9 del virus della peste equina

Lelli et al.

Bibliografia 1. Chiam R., Sharp E., Maan S., Rao S., Mertens P., Blacklaws B., Poynter N.D., Wood J. & Olivares J.C. 2009. Induction of antibody responses to African horse sickness virus (AHSV) in ponies after vaccination with recombinant Modified Vaccinia Ankara (MVA). PLos One, 4(6), e5997. 2. Castillo-Olivares J., Calvo-Pinilla E., Casanova I., Bachanek-Bankowska K., Chiam R., Maan S., Nieto J.M., Ortego J. & Mertens P.P.C. 2011. A modified vaccinia Ankara virus (MVA) vaccine expressing African horse sickness virus (AHSV) VP2 protects against AHSV challenge in an IFNAR –/– mouse model. PLos One, 6(1), e16503. 3. Direction de la Qualité du Medicament du Conseil de l’Europe (DEQM) European Pharmacopeia 2004. European Pharmacopeia, 5th Ed. Juin 2004. Aubin, Ligugé, 01/2005:0062, 628-634. 4. Du Plessis M., Cloete M., Aitchison H. & Van Dijk A.A. 1998. Protein aggregation complicates the development of baculovirus-expressed African horsesickness virus serotype 5 VP2 subunit vaccines. Onderstepoort J Vet Res, 65, 321-324. 5. Dubourget P., Préaud J., Detraz F., Lacoste F., Fabri A., Erasmus B. & Lombard M. 1992. Development, production, and quality control of an industrial inactivated vaccine against African horse sickness virus serotype 4. In Bluetongue, African Horse Sickness and Related Orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, June 1991 Paris, France CRC Press, Boca Raton, 874-886. 6. Gard G.P. & Kirkland P.D. 1993. Bluetongue virology and serology. Australian standard diagnostic techniques for animal diseases (L.A. Corner & T.J. Bagust, eds). CISRO Information Services, Melbourne, 1-17. 7. Gillespie J.H. & Timoney J.F. 1988. African horse sickness. In Hagan and Bruner. Microbiology and infectious diseases of domestic animals, Eighth Ed. Cornell University Press, London, 712-715. 8. Guthrie A.J., Quan M., Lourens C.W., Audonnet J.C., Minke J.M., Yao J., He L., Nordgren R., Gardner I.A., MacLachlan N.J. 2009. Protective immunization of horses with a recombinant canarypox virus vectored vaccine co-expressing genes encoding the outer capsid proteins of African horse sickness virus. Vaccine, 27, 4434-4438. 9. House J.A., Lombard M., Dubourget P., House C. & Mebus C.A. 1994. Further studies on the efficacy of an inactivated African horse sickness serotype 4 vaccine. Vaccine, 12 (2), 142-144. 10. Mare S. & Paweska J.T. 2005. Preparation of recombinant African horse sickness virus VP7 antigen

via a simple method and validation of a VP7-based indirect ELISA for the detection of group-specific IgG antibodies in horse sera. J Virol Methods, 125 (1), 55-65. 11. Mellor P.S. & Hamblin C. 2004. African horse sickness. Vet Res, 35, 445-466. 12. Nashwa K.M. & Rofaiil S.K. 2007. Evaluation of bivalent inactivated oil adjuvant African horse sickness vaccine in guinea-pigs and mice. In Proc. 5th Scientific Conference, 5 November. Beni-Suef. Beni Suef Medi J, 18 (1), 57-61. 13. Nurton J.P., Guthrie A.J. & Howell P.G. 2001. Development and preliminary validation of an inactivated African horse sickness virus serotype 4 vaccine. In Proc. Annual Congress of the Equine Practitioner Group of the South African Veterinary Association, January 2001. Pretoria, South Africa. World Veterinary Association, Brussels, 33-129. 14. Ronchi G.F., Ulisse S., Rossi E., Franchi P., Armillotta G., Capista S., Peccio A., Di Ventura M. & Pini A. 2012. The guinea-pig as a model for the study of the antigenic properties of an inactivated and adjuvanted vaccine for the control of African horse sickness. Vet Ital, 48 (1), 67-76. 15. Russel W.M.S. & Burch R. 1959. The principles of human experimented technique; Methuen, London, 238 p. 16. Scacchia M., Lelli R. Peccio A., Di Mattia T., Mbulu R.S., Hager A.L., Monaco F., Savini G. & Pini A. 2009. African horse sickness: a description of outbreaks in Namibia. Vet Ital, 45 (2), 265-274. 17. Scanlen M., Paweska J.T., Vershoor J.A. & van Dijk A.A. 2002. The protective efficacy of a recombinant VP2-based African horsesickness subunit vaccine candidate is determined by adjuvant. Vaccine, 20, 1079-1088. 18. Von Teichman B.F., Dungu B. & Smit T.K. 2010. In vivo cross-protection to African horse sickness serotypes 5 and 9 after vaccination with serotypes 8 and 6. Vaccine, 28, 6505-6517. 19. Wade-Evans A.M., Pullen L., Hamblin C., O’Hara R., Burroughs J.N. & Mertens P.P.C. 1997. African horse sickness virus VP7 sub-unit vaccine protects mice against a lethal, heterologous serotype challenge. J Gen Virol, 78, 1611-1616. 20. Wilson A., Mellor P.S., Szmaragad C., Mertens P.P.C. 2009. Adaptive strategies of African horse sickness virus to facilitate vector transmission. Vet Res 40(2):16. doi: 10.1051/vetres:2008054. 21. World Organisation for Animal Health (Office International des Épizooties: OIE) 2008. Manual of diagnostic tests and vaccine for terrestrial animals (mammals, birds and bees). Sixth edition. OIE, Paris, 823-838.

Veterinaria Italiana 2013, 49 (1), 79-88

Inactivated and adjuvanted vaccine for the control of the African horse sickness virus serotype 9 infection: evaluation of efficacy in horses and guinea-pig model Rossella Lelli1, Umberto Molini2, Gaetano Federico Ronchi2, Emanuela Rossi2, Paola Franchi2, Simonetta Ulisse2, Gisella Armillotta2, Sara Capista2, Siegfried Khaiseb3, Mauro Di Ventura2 & Attilio Pini2 Istituto Zooprofilattico Sperimentale della Sicilia ‘A. Mirri’, Via Gino Marinuzzi, 3, 90129 Palermo, Italy Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise ‘G. Caporale’, Campo Boario, 64100 Teramo, Italy a.pini@izs.it 3 Central Veterinary Laboratory, 24 Goethe Street, Windhoek, Namibia 1

Keywords African horse sickness, Antibody response, Challenge, Guinea-pig, Inactivated vaccine, Namibia, Serotype 9.

Summary African horse sickness (AHS) is a non-contagious viral disease of solipeds transmitted by Culicoides. The disease is endemic in most African countries. Past experience has shown that Italy is a country exposed to emerging infectious diseases endemic to Africa; an incursion of AHS virus together with the widespread presence of Culicoides vectors could be the cause of a serious epidemic emergency. A live attenuated vaccine containing seven of the nine viral serotypes, serotype 5 and 9 are excluded, is commercially available from Onderstepoort Biological Products. However, the use of live vaccines is a matter of endless disputes, and therefore inactivated or recombinant alternative products have been investigated over the years. Since research on AHS is hampered by the use of horses to evaluate vaccine potency, in a previous experiment serological response to serotypes 5 and 9 was assayed in guinea-pigs and horses. A durable and comparable serological response was observed in the two animal species. In the present study antibody response in horses and guinea-pigs, immunised with the inactivated-adjuvanted vaccine formulated with serotype 9, was tested over a period of 12 months. When immunity was challenged, horses were protected from infection and disease. Antibody response in horses and guinea-pigs compared favourably. Veterinaria Italiana 2013, 49 (1), 89-98

Introduction African horse sickness (AHS) is a non-contagious viral disease of solipeds caused by nine virus serotypes. The virus belongs to the genus Orbivirus, family Reoviridae. Approximately 30 of the over 1,500 identified species of Culicoides are believed to be capable of Orbivirus transmission. The most important vector of AHSV in the field is Culicoides imicola, a species common throughout Africa and South East Asia (20). AHS is endemic in numerous countries in sub-Saharan Africa but disease has been recorded in North Africa (1965), Spain (1966 and 1987-1990) and the Middle East (1959, 1961 and 1989). Spreading of single serotypes caused outbreaks in the Iberian Peninsula and the Middle East: serotype 4 or 9. From 1975 the disease has also spread to West Africa: Nigeria, Senegal and Mauritania (11). Events of recent decades have revealed that Italy is a country at risk of outbreaks of emerging

infectious diseases endemic to Africa. The presence of vectors of the Culicoides genus makes the spread of AHS in Italy a likely event. There is, therefore, a need for readily available, innocuous and effective immunising products as well as specific and sensitive diagnostic tests. Live attenuated vaccines could be made available by Onderstepoort Biological Products (OBP), South Africa, in a relatively short time. The vaccine commonly used in countries where the disease is endemic contains 7 of the 9 serotypes and it is manufactured in 2 components, namely: a trivalent component containing serotypes 1, 3, and 4, and a quadrivalent one containing serotypes 2, 6, 7, and 8; they are administered 21 days apart. Serotype 5 was excluded from the formulation in 1990 due to residual pathogenicity recorded in the field. Serotype 9, considered epidemiologically irrelevant in South Africa is also not included in the vaccine. These two

Lelli et al.

Efficacy of the African horse sickness virus serotype 9 inactivated vaccine

latter viral serotypes, however, have dominated AHS outbreaks occurring in 2006, particularly in the Western Cape (18). In Namibia, serotype 9 can be isolate from the pulmonary, cardiac and mild form of disease. Pyrexia, oedema of the supraorbital fossa, breathing difficulties, inappetence and reluctance to move are constant symptoms of the disease caused by this serotype (T. Di Mattia, M. Scacchia, personal communications). Taking into consideration the objections that usually accompany the use of live attenuated viral vaccines, alternative solutions have been sought over the years. Research has been focused on inactivated-adjuvanted formulations (5, 9, 12, 13) and more recently on recombinant vaccines (1, 4, 7, 8, 10, 17, 19). Results reported in the literature indicate that the former may be capable of protecting horses from AHS. However, immunity was challenged shortly after immunisation, 76-98 days post vaccination; the promising results obtained with the latter have still to be confirmed. The limit due to the need of using horses for testing potency of AHS vaccine prompted evaluation of the possibility of resorting to animal models for testing vaccine efficacy. In a recent publication it was shown that interferon alpha receptor knock‑out mice (IFNAR -/-), immunised with Modified Vaccinia Ankara virus expressing AHS virus VP2 from serotype 4 could be a model for the purpose (2). In a previous experiment (14) various formulations of monovalent, inactivated and adjuvanted vaccines prepared with serotypes 5 and 9 were tested for their capacity to induce neutralising antibodies in guinea-pigs. On the basis of the results, one of the formulations produced with serotype 9 was administered into 3 horses. Seroconversion lasting more than 10 months after immunisation was recorded in both animal species. The study has shown that guinea-pigs could be a predictive model for investigating the antigenic properties of vaccines under investigation (14). In order to confirm efficacy of the vaccine and validity of the guinea-pig model it was then decided to carry out a horse immunity challenge. The trial hereunder reported was conducted under the auspices of the Namibian Veterinary Services in an AHS-free area on immunised horses kept to pasture until performance of the challenge test carried out on days 109 and 365 after vaccination. At the time of challenge animals were transferred to insect-proof premises. In the pre-challenge period, the product innocuity and antibody response were monitored, while body temperature, clinical signs and viraemia were monitored in the post-challenge period. In guinea-pigs, serological reactivity to the vaccination was monitored till day 365.

Materials and Methods Production of serotype 9 vaccine has been described (14). For easier reading, data are summarised here; the vaccine was used 5 months after manufacturing.

Cell lines BHK21(clone 13) cells, from the European collection of cell cultures (ECACC), were used for producing the vaccine while VERO cells, from the European collection of cell cultures (ECACC), were used for infectivity titrations, serum neutralisation tests and virus isolation.

Virus Serotype 9 from the reference antigen-Bob Swanepoel collection, kindly provided by Dr Hüebschle, was used for vaccine production. After virus amplification, purification and concentration, the viral suspension was tested according to the European Pharmacopoeia (3). Confirmation of viral type was performed by virus neutralisation (VN) against all nine AHS virus serotypes (14). The same virus strain used for vaccine production was used in the immunity challenge test carried out on day 109 after vaccination, whereas a serotype 9 virus strain, responsible for an acute form of disease, isolated from an outbreak that had occurred in Namibia in 2008 was used in the challenge test on day 365 (16). The two virus strains, at the second passage in BHK21, freeze-dried, tested for purity and stored at 5°C ± 3°C had infectious titres of 106.9 and 106.8 TCID50/ml respectively. At the time of challenge, each virus was reconstituted in distilled water, diluted in MEM medium to contain 105TCID50/ml and inoculated into the horses.

Inactivant Inactivation was carried out with BEI 5mM as described by Ronchi et al. (14).

Adjuvants The adjuvant ISA27VG, kindly provided by SEPPIC (SEPPIC Italia S.r.l.) and previously tested in horses for its inflammatory properties, was emulsified in the concentrated purified and inactivated virus suspension according to supplier recommendations (14); 0.3 mg of saponin/vaccine dose (Sigma Aldrich) were added before use.

Horses Ethical clearance was obtained from the Central Veterinary Laboratory (CVL) Animal Ethics Committee

Veterinaria Italiana 2013, 49 (1), 89-98

Lelli et al.

for the use of horses in this vaccination-challenge study. Approved guidelines for husbandry, handling and care of horses were followed. Vaccination, bleeding procedures and monitoring of challenged horses were supervised by a registered veterinarian. Vaccine efficacy was tested at the Namibian Ministry of Agriculture, Water and Forestry ‘Bergvlug’Veterinary Research Farm in the Windhoek district. The farm is at 1,932 m above sea level, night temperatures vary between – 8°C in winter and 8°C in summer. The farm covers a surface of 5,000 hectares and holds cattle, sheep, a variable number of horses, antelope and predators. There is no record of outbreaks of AHS on the farm; historically, the area was used by the German army for protecting horses from AHS. Fifteen animals aged 2 years or over, comprising a homogenous herd on the basis of clinical examination, biochemical profile, and CBC test were purchased from a farm in Karas, 780 km from Windhoek, where vaccination against AHS is not practised. Animals were tested by competitive enzyme-linked immunosorbent assay (c-ELISA) for absence of anti-AHS antibodies before purchase and 15 days after their arrival in Bergflug. On the day of immunisation, time zero (T0), a further blood sample was taken to exclude the presence of antibodies against any of the nine AHS serotypes by serum neutralizing test (SN). The vaccine was administered to 10 horses on the mid third of the left side of the neck with an intramuscular dose of 1 ml. Body temperature and inflammatory reactions at the inoculation site were monitored over the following 15 days. Temperature was checked twice a day, morning and afternoon. A 1 ml booster dose was administered in the mid third of the right side of the neck at T35. Again, body temperature and inflammatory reactions at the inoculation site were monitored for further 15 days. Jugular blood samples were taken at T15, T21, T35, T42, T49, T63, T109, T175, T220, T250, T294, T321 and T365 to monitor kinetics of antibody response by SN and by serum neutralising index (SNI). Horses were kept out to pasture until time of challenge when they were transferred to the insectproof premises provided with screened door and windows, impregnated every week with Fendona insecticide (BASF Agricultural Products Ltd, Cape Town, South Africa).

Sentinel horses Five horses initially acted as sentinels to exclude circulation of AHS virus. On day 109 their number was reduced to 4, as 1 animal was used as a control for the first potency test. They shared the same pasture with the immunised subjects. Sera obtained on a monthly basis were tested by c-ELISA.

Veterinaria Italiana 2013, 49 (1), 89-98

Efficacy of the African horse sickness virus serotype 9 inactivated vaccine

Guinea-pigs The experiment was carried out in compliance with the Italian law 116/92. To perform the experiment on guinea-pigs a written communication was sent to the Ministry of Health, Department for Veterinary Public Health, Nutrition and Food Safety; Office VI, Viale Giorgio Ribotta n. 5, zip code 00144 Rome, Italy. Four female guinea-pigs weighing between 350 g and 500 g were inoculated with 0.5 ml of vaccine in the pre-scapular region; a booster dose was administered 35 days later. Animals were bled through cardiac puncture under deep general anaesthesia at the same time intervals as for the horses and sera tested by SNI. Work was performed at the authorised plant at the Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise ‘G. Caporale’ in Teramo, Italy. All animals were maintained under standard animal housing facilities in air conditioned rooms (20°C/24°C), with a relative humidity of 50%55%, and fed with pellet food and water ad libitum. Guinea-pigs were monitored with a daily physical examination until the end of the trial.

Serology The commercial kit Ingezim AHSV Compac Plus (INGENASA, Madrid, Spain) was used to perform c-ELISA. The method to calculate the SNI by log10 serial dilutions of the virus against inactivated test serum diluted 1:10 has been described (14). SN tests performed by doubling dilutions of sera against 100 TCID50 of the virus (6) were carried out in parallel. The SNIs are expressed as log10, whereas the SN titres are expressed as their reciprocal.

Challenge tests Vaccine efficacy was tested at T109 and T365 after the primary vaccination. As already stated, the same virus that was used to generate the killed vaccine was used in the challenge on day 109 whereas serotype 9 isolated in Namibia was used for the challenge on day 365. In each case 105 TCID50 of virus were inoculated intravenously. In the challenge at T109, 5 immunised horses were used while in the challenge at T365 the number had to be reduced to 4 since the fifth animal was unmanageable and could not be moved to the insect-proof premises without risking worker safety. The use of a single control for each test was dictated by reasons of animal welfare, in respect of the 3 R logic: Refinement, Replacement, Reduction (15). Clinical examinations and body temperatures were taken twice a day for 21 days post challenge. Values ≤ 38°C were considered physiological. Blood samples were collected daily in anticoagulant for 28 days. Viraemia was tested by virus isolation on VERO cells

Lelli et al.

Efficacy of the African horse sickness virus serotype 9 inactivated vaccine

and by group-specific real-time RT-PCR according to the methods described in the OIE Manual of diagnostic tests and vaccines for terrestrial animals (21). Samples negative for cytopathic effect at first passage were further tested by performing 3 serial passages of 7 days each. On the third passage, the cells were removed and examined by immunofluorescence to exclude presence of AHS virus.

38.2°C and 38.8°C was recorded in 4 of the 10 horses (Table I). No other clinical signs were observed during the observation period. The temperature rise in the 4 horses coincided with the inflammatory reaction observed at T36 and T37 after administration of the booster.

Culicoides capture

Serological reactivity to vaccine was monitored in the horses up to challenge days.

Onderstepoort light traps were set up at night near the pasture in both the dry and wet seasons.

Results Reactivity to vaccination Inflammatory reactions at the inoculation site Slight thickening of the skin lasting no more than 48 hours could be seen at the inoculation site after administration of the first vaccine dose. The booster at T35 induced formation of a small lump having a maximum diameter of 5 mm and lasting no more than 3 days. In 9 animals the lump was neither hot nor painful; only in horse number 4 was the lump was hot on days T36 and T37.

Body temperatures Body temperatures remained within the physiological range until T35. Following administration of the booster dose, at T36, a temperature rise to between

Antibody response

SNIs in horses The SNIs at T15 varied in 8 horses between 2.10 and 4.27 while horses 6 and 7 reacted at T21, with a SNI of 0.7 and 1.5 respectively. After the booster, on T42, the SNIs reached values ≥ 6.8, they remained stable for the entire observation period, except for horses number 6 and 9, which presented a SNI of 2.27 and 0.83 at T109 and T365, respectively. The kinetics of the immune response to vaccination is shown in Table II and Figure 1. The SNIs ≤ 0.5 detected at T0 are not shown.

SNIs in guinea-pigs Serological reactivity at T15 was similar to that recorded in the horses. At T42 the SNIs reached peak values of 5.5 and at T365 they varied between 3.5 and 4.5 (Table II). The kinetics of the immune response, running parallel to that of the horses, is shown in Table II and Figure 1. SNIs at T0 were ≤0.5 and consequently are omitted.

Days Horse N° 1 2 3 4 5 6 7 8 9 10 11 control 12 control 13 control 14 control 15 control

T1 a 35

T36

T37

T38

T39

T40

T41

T42

T43

T44

Physiological temperature ≤ 38 °C

Table I. Body temperatures in horses after booster dose (T35 ).

37.1 37.8 36.6 38.4 38.4 38.2 38.8 37.9 37.7 37.8 37 37.2 37.1 37.6 35.9

37.2 37.2 36.9 37.3 37.6 37.2 37.9 36.3 37.1 36.2 36.6 37.3 37 36.5 36

36.8 36.9 36.9 37.2 37.5 37 37.6 36.4 37.2 36.8 37 37 37.3 36.5 35.9

36.7 36.9 36.5 37.1 37.8 37.1 37.6 36.5 36.8 36.9 37.1 37 37.1 36.1 35.8

36.5 36.9 36.5 36.1 36.7 36.9 37 36.6 36.8 36.3 36.8 37 36.1 36.1 35.9

36.9 37 36.4 36.6 36.9 37.3 37.1 36.9 36.9 36.8 37.1 36.9 36.9 36.8 36.1

36.1 36.3 36.1 36.5 36.5 36.5 37 36.6 36.7 36.9 36.5 36.8 36.9 36.6 35.7

36.2 36.8 36.5 36.8 36.8 37.1 36.4 36.9 36.7 37.2 36.3 36.8 36.8 36.1 35.6

36.1 36.5 36.5 36.2 36.4 36.6 36.7 36.9 36.1 36.7 36.8 36.4 36.4 36.9 35.9

Veterinaria Italiana 2013, 49 (1), 89-98

Lelli et al.

Efficacy of the African horse sickness virus serotype 9 inactivated vaccine

Table II. Serum neutralising indices in immunised horses and guinea-pigs.

Horse 1 Horse 2 Horse 3 Horse 4 Horse 5 Horse 6 Horse 7 Horse 8 Horse 9 Horse 10 Guinea-pig 1 Guinea-pig 2 Guinea-pig 3 Guinea-pig 4

T15

T21

T28

T35(a)

T42

T49

T63

T109(b)

T175

T220

T250

T294

T321

T365(c)

4.27 3.10 3.93 3.85 0.00 3.10 0.00 2.27 3.10 2.10 1.5 2 1.66 2

4.00 3.33 3.50 3.33 0.70 3.50 1.50 2.50 3.50 1.50 2 2.66 2.16 2

5.97 5.97 5.97 5.97 0.67 5.97 2.37 5.97 5.97 5.97 3.66 3.16 1.66 2.33

5.97 5.97 5.97 5.97 0.34 5.97 5.97 5.97 5.97 5.97 3.16 2.66 2.16 2

6.80 6.80 6.80 6.80 6.80 6.80 6.80 6.80 6.80 6.80 5.5 5.5 5.5 5.5

6.80 6.80 6.80 6.80 6.80 6.80 6.80 6.80 6.80 6.80 4.83 4.83 4.83 4.83

6.80 6.80 6.80 6.80 6.80 6.80 6.80 6.80 6.80 6.80 4.5 4.5 4.5 4.5

6.80 6.80 6.80 6.80 2.27 6.80 6.80 6.80 6.80 6.80 4.83 4.83 4.83 4.83

6.80 6.80 6.80 6.80 6.80 4.83 4.33 4.83 4.67

6.80 6.80 6.80 6.80 6.80 4.72 4.72 4.72 4.72

6.80 6.80 6.80 6.80 6.80 4.5 4.5 3.5 4.5

6.80 6.80 6.80 6.80 6.80 4.16 4.16 4.16 4.16

6.80 6.80 6.80 6.80 6.80 4.5 3.5 4.5 3.5

6.80 6.80 6.80 0.83 6.80 4.5 3.5 4.5 3.5

(a) T35 booster dose, (b) challenge T109, (c) challenge T365.

8.0

6.8

7.0

5.1

5.0

5.5

5.4

2.6

2.0

2.7

4.7 4.7

4.5

6.8 6.8 5.6

4.8

6.8 6.8

6.3

4.0 3.0

6.8

booster

6.0

Log10

6.8

4.3

4.2

4,0

4.0

2.7 2.5

2.2 1.8

1.0 0.0

Horses

T15

T21

T28

T35

T42

T49

T63

T109

T175

T220

T250

T294

Guinea pigs

T321

T365

Figure 1. Average serum neutralising indices : guinea-pigs compared to horses.

Serological titres in horses Table III shows the serological titres recorded. Again in horses number 6 and 7 reactivity was delayed to T21. Peak titres of 320 were recorded between T42 and T49, and then they then declined slowly, although remaining positive up to T365.

Efficacy tests Immunity challenge at T109 Five immunised horses and one control underwent

Veterinaria Italiana 2013, 49 (1), 89-98

challenge by intravenous inoculation of 105 TCID50 of virus homologous to the strain used for vaccine production. In the control, pyrexia between 38.6째C and 40째C was recorded, on days 10, 11, 12, 13, 14, and 15 post inoculation (pi), during which supraorbital oedema, sensory depression, refusal of food and reluctance to move were observed. Neither temperature rise nor any other clinical sign were observed in the immunised horses over the 21 days of observation. Figure 2 and Figure 3 show morning and afternoon temperatures in the control and vaccinated horses.

Lelli et al.

Efficacy of the African horse sickness virus serotype 9 inactivated vaccine

Table III. Serum titres in immunised horses. Horse No.

T15

T21

T28

T35(a)

T42

T49

T63

T109(b)

T175

T220

T250

T294

T321

T365(c)

1 2 3 4 6 5 7 8 9 10

40 20 40 40 5 40 5 20 40 20

40 40 40 40 5 40 10 40 40 20

160 160 160 80 20 5 80 20 80 40

80 80 160 160 5 80 40 160 80 160

320 320 320 320 80 320 320 320 320 320

320 320 320 320 320 80 320 320 320 320

320 320 320 160 40 320 320 320 320 320

160 160 160 40 20 160 160 320 160 160

160 80 80 160 80

160 80 80 160 40

160 80 80 40 40

160 80 80 20 40

(a) T35 booster dose, (b) challenge T109, (c) challenge T365.

Morning temperature

38.9

째C

39 physiological temperature

37.1

37.4

39.7

39.3

38.6

38.6 38.4 37.9

37.8

37.4

37 37.4 36.6 36.7 36.7 37.3 37.1 37.0 36.9 36.9 36.8 36.5 36.4 36 36.2 36.2

37.5

Afternoon temperature

T10

T11

T12

T13

T14

38.5 37.6 37.6 37.4 37.1 37.5 37.5 37.3 36.9 37.3 37.1 36.9 36.7 36.6

T15

T16

T17

T18

T19

T20

T21

Figure 2. Challenge T109 : body temperatures in the control horse. 39.0 Morning temperature (mean) 38.5 38.0

Afternoon temperature (mean) physiological temperature

째C

37.5 37.0 36.5 36.0 35.5 35.0

T10

T11

T12

T13

T14

T15

T16

T17

T18

T19

T20

T21

Figure 3. Challenge T109 : mean body temperatures in the vaccinated horses. The error bars indicate the maximum and minimum temperatures.

Veterinaria Italiana 2013, 49 (1), 89-98

Lelli et al.

Efficacy of the African horse sickness virus serotype 9 inactivated vaccine

The virus was successfully isolated from blood samples taken from day 12 to 16. Infective titres ranged between 102.8 and 104.8 TCID50/ml; RT-PCR was positive from day 11 to 21 pi. Figure 4 shows the test results.

In the control animal an initial temperature rise of 38.9째C was recorded on the morning of day 6 pi. A second rise, 38.5째C-40.1째C, was evident on days 8, 9, 10 and 11 pi (Figure 5), during which clinical signs were consistent with those observed in the horse challenged at T109.

Virus isolation and RT-PCR tests conducted on the blood samples for 28 days after the challenge test gave negative results.

The virus was successfully isolated from blood samples taken from days 8 to 11. Infective titres ranged between 102.7 and 103 TCID50/ml; RT-PCR was positive from days 7 to 18 pi (Figure 7). Neither temperature rise (Figure 6) nor any other clinical sign were observed in the immunised horses over the 21 days of observation. Virus isolation and RT-PCR tests conducted on the blood samples for 28 days after challenge gave negative results.

Immunity challenge at T365 Four immunised horses and one control underwent the challenge test by intravenous inoculation of 105 TCID50 of virus isolated in Namibia in the 2008.

RT- PCR positive

4.8 4.8 4.8

Log10

3.5 3

2.8

0 T1

T9 T10 T11 T12 T13 T14 T15 T16 T17 T18 T19 T20 T21 T22 T23 T24 T25 T26 T27 T28

Figure 4. Challenge T109 : polymerase chain reaction and infectivity titres/ml of blood in the control horse. In immunised horses the virus isolation and RT-PCR are negative and are not reported. 41 Morning temperature

40.1 40

39.8

째C

39.2

38.9

Afternoon temperature

39 38.5

physiological temperature

37.6

37.8 37.4 37 37.5 37.2 36.4

38.7

37.6 36.8 36.6

37.9

37.5 37.4 37.4 37.4 37.5 37.1 37.7 37.4 37.4 37.4 37.2 37.1 37.1 37 36.8 36.7

37.1 36.7

37.9 37.4

35.9

36.1

35.7 35.6

T10

T11

T12

T13

T14

T15

T16

T17

T18

T19

T20

T21

Figure 5. Challenge T365 : body temperatures in the control horse. Veterinaria Italiana 2013, 49 (1), 89-98

Lelli et al.

Efficacy of the African horse sickness virus serotype 9 inactivated vaccine

39.0 Morning temperature (mean) 38.5

Afternoon temperature (mean) physiological temperature

38.0

째C

37.5 37.0 36.5 36.0 35.5 35.0

T10

T11

T12

T13

T14

T15

T16

T17

T18

T19

T20

T21

Figure 6. Challenge T365 : mean body temperatures in the vaccinated horses. The error bars indicate the maximum and minimum temperatures. 4 RT- PCR positive

T10 T11 T12 T13 T14 T15 T16

Log10

2.7

2.80

T17 T18 T19 T20 T21 T23

T24 T25 T26 T27 T28

Figure 7. Challenge T365 : polymerase chain reaction and infectivity titres/ml in the control horse. In immunised horses the virus isolation and RT-PCR are negative and are not reported. The 2 control horses recovered without apparent sequelae.

Sentinel horses Sentinel horses remained free of antibody to AHS over the entire observation period of 14 months.

Culicoides capture During the wet season, the number of Culicoides captured near the pasture varied between 100 and

150 per night. The number dropped to about 10 during the dry winter season.

Conclusions During the 14-month trial, neither clinical signs nor serological response that might suggest AHS virus circulation was observed in the sentinel animals, confirming what was already known to the German authorities who, in the early 1900,

Veterinaria Italiana 2013, 49 (1), 89-98

Lelli et al.

were using the area to keep horses to protect them from the disease. It can be assumed that the habitat, characterized by the altitude of the farm, approximately 2,000 metres, and the low nighttime temperatures, is not favourable to Culicoides, which are in such a low number that they do not have vectorial capacity. On the contrary, in the lower areas of the Windhoek district, where disease is endemic, Culicoides, during the summer rainy season, are captured in thousand. Serotype 9 was used in the trial because it is not present in the polyvalent vaccine produced by OBP despite the fact that, in the recent past, the serotype was dominant, particularly in the Western Cape. The 2 serum neutralisation tests, conducted in the course of the present trial, aimed to asses whether there was any significant difference in detecting antibody response to vaccination over the observation period of 12 months. No significant differences were recorded. In the immunised horses, neutralising antibodies were detected by both tests from day 15 after vaccination, reached a plateau between days 42 and 49, and persisted, at significant levels, for the entire observation period. The serum conversion recorded in horses at T365 compared favourably with that reported in the literature at 200 days after

Efficacy of the African horse sickness virus serotype 9 inactivated vaccine

vaccination: mean serum titre of 80 against mean titre of 40, respectively (5). The vaccine prepared with virus from the reference antigens-Bob Swanepoel collection proved to be effective when challenge was performed at T365 with a recently isolated strain from a Namibian disease outbreak. No temperature reaction was recorded or viraemia detected, indicating that horses were protected from disease and infection and that, with reference to the strains under investigation there was not a significant antigenic drift. On the contrary, the 2 control animals used in the challenge trials at T109 and T365 reacted with pyrexia, viraemia and clinical symptoms. Pyrexia was the first sign of infection preceding viraemia of at least 24 h. Both animals survived infection. In guinea-pigs, the vaccine stimulated the development of neutralising antibodies at a level comparable to those recorded in horses. SNIs were detected on day 15, reaching a plateau at day 42, persisting at significant titres for the entire 365-day observation period, thus confirming as this model could be an indicator of the antigenic potency of the product under test. In future trials it would be of interest to compare guinea pigs and IFNAR -/- models in assessing efficacy of AHS vaccines.

References 1. Chiam R., Sharp E., Maan S., Rao S., Mertens P., Blacklaws B., Poynter N.D., Wood J. & Olivares J.C. 2009. Induction of antibody responses to African horse sickness virus (AHSV) in ponies after vaccination with recombinant Modified Vaccinia Ankara (MVA). PLos One, 4(6), e5997. 2. Castillo-Olivares J., Calvo-Pinilla E., Casanova I., Bachanek-Bankowska K., Chiam R., Maan S., Nieto J.M., Ortego J. & Mertens P.P.C. 2011. A modified vaccinia Ankara virus (MVA) vaccine expressing African horse sickness virus (AHSV) VP2 protects against AHSV challenge in an IFNAR –/– mouse model. PLos One, 6(1), e16503. 3. Direction de la Qualité du Medicament du Conseil de l’Europe (DEQM) European Pharmacopeia 2004. European Pharmacopeia, 5th Ed. Juin 2004. Aubin, Ligugé, 01/2005:0062, 628-634. 4. Du Plessis M., Cloete M., Aitchison H. & Van Dijk A.A. 1998. Protein aggregation complicates the development of baculovirus-expressed African horsesickness virus serotype 5 VP2 subunit vaccines. Onderstepoort J Vet Res, 65, 321-324. 5. Dubourget P., Préaud J., Detraz F., Lacoste F., Fabri A., Erasmus B. & Lombard M. 1992. Development, production, and quality control of an industrial inactivated vaccine against African horse sickness virus serotype 4. In Bluetongue, African Horse

Veterinaria Italiana 2013, 49 (1), 89-98

Sickness and Related Orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, June 1991 Paris, France CRC Press, Boca Raton, 874-886. 6. Gard G.P. & Kirkland P.D. 1993. Bluetongue virology and serology. Australian standard diagnostic techniques for animal diseases (L.A. Corner & T.J. Bagust, eds). CISRO Information Services, Melbourne, 1-17. 7. Gillespie J.H. & Timoney J.F. 1988. African horse sickness. In Hagan and Bruner. Microbiology and infectious diseases of domestic animals, Eighth Ed. Cornell University Press, London, 712-715. 8. Guthrie A.J., Quan M., Lourens C.W., Audonnet J.C., Minke J.M., Yao J., He L., Nordgren R., Gardner I.A., MacLachlan N.J. 2009. Protective immunization of horses with a recombinant canarypox virus vectored vaccine co-expressing genes encoding the outer capsid proteins of African horse sickness virus. Vaccine, 27, 4434-4438. 9. House J.A., Lombard M., Dubourget P., House C. & Mebus C.A. 1994. Further studies on the efficacy of an inactivated African horse sickness serotype 4 vaccine. Vaccine, 12 (2), 142-144. 10. Mare S. & Paweska J.T. 2005. Preparation of recombinant African horse sickness virus VP7 antigen via a simple method and validation of a VP7-based

Efficacy of the African horse sickness virus serotype 9 inactivated vaccine

indirect ELISA for the detection of group-specific IgG antibodies in horse sera. J Virol Methods, 125 (1), 55-65.

horse sickness: a description of outbreaks in Namibia. Vet Ital, 45 (2), 265-274.

11. Mellor P.S. & Hamblin C. 2004. African horse sickness. Vet Res, 35, 445-466.

17. Scanlen M., Paweska J.T., Vershoor J.A. & van Dijk A.A. 2002. The protective efficacy of a recombinant VP2-based African horsesickness subunit vaccine candidate is determined by adjuvant. Vaccine, 20, 1079-1088.

12. Nashwa K.M. & Rofaiil S.K. 2007. Evaluation of bivalent inactivated oil adjuvant African horse sickness vaccine in guinea-pigs and mice. In Proc. 5th Scientific Conference, 5 November. Beni-Suef. Beni Suef Medi J, 18 (1), 57-61. 13. Nurton J.P., Guthrie A.J. & Howell P.G. 2001. Development and preliminary validation of an inactivated African horse sickness virus serotype 4 vaccine. In Proc. Annual Congress of the Equine Practitioner Group of the South African Veterinary Association, January 2001. Pretoria, South Africa. World Veterinary Association, Brussels, 33-129. 14. Ronchi G.F., Ulisse S., Rossi E., Franchi P., Armillotta G., Capista S., Peccio A., Di Ventura M. & Pini A. 2012. The guinea-pig as a model for the study of the antigenic properties of an inactivated and adjuvanted vaccine for the control of African horse sickness. Vet Ital, 48 (1), 67-76. 15. Russel W.M.S. & Burch R. 1959. The principles of human experimented technique; Methuen, London, 238 p. 16. Scacchia M., Lelli R. Peccio A., Di Mattia T., Mbulu R.S., Hager A.L., Monaco F., Savini G. & Pini A. 2009. African

Lelli et al.

18. Von Teichman B.F., Dungu B. & Smit T.K. 2010. In vivo cross-protection to African horse sickness serotypes 5 and 9 after vaccination with serotypes 8 and 6. Vaccine, 28, 6505-6517. 19. Wade-Evans A.M., Pullen L., Hamblin C., Oâ&#x20AC;&#x2122;Hara R., Burroughs J.N. & Mertens P.P.C. 1997. African horse sickness virus VP7 sub-unit vaccine protects mice against a lethal, heterologous serotype challenge. J Gen Virol, 78, 1611-1616. 20. Wilson A., Mellor P.S., Szmaragad C., Mertens P.P.C. 2009. Adaptive strategies of African horse sickness virus to facilitate vector transmission. Vet Res, 40(2):16. doi: 10.1051/vetres:2008054. 21. World Organisation for Animal Health (Office International des Ă&#x2030;pizooties: OIE) 2008. Manual of diagnostic tests and vaccine for terrestrial animals (mammals, birds and bees). Sixth edition. OIE, Paris, 823-838.

Veterinaria Italiana 2013, 49 (1), 89-98

Estimation of the sensitivity of the surveillance system for avian influenza in the western region of Cuba Edyniesky Ferrer1, Paolo Calistri2, Osvaldo Fonseca1, Carla Ippoliti2, Pastor Alfonso1, Simona Iannetti2, María A. Abeledo1, Octavio Fernández1, María I. Percedo1& Antonio Pérez3 Centro Nacional de Sanidad Agropecuaria (CENSA). Carretera de Tapaste y Autopista Nacional, Apartado 10, CP 32700,San José de Las Lajas, Mayabeque, Cuba ferrer@censa.edu.cu, edyferrer8123@gmail.com 2 Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise ‘G. Caporale’, Campo Boario, 64100 Teramo, Italy 3 Instituto de Medicina Veterinaria, Calle 12 No. 355, e/ 15 y 17 Vedado, CP:10400, La Habana, Cuba 1

Keywords Avian influenza, Cuba, H5 subtype, H7 subtype, LPAI, Poultry, Surveillance system.

Summary Although avian influenza (AI) virus of H5 and H7 subtypes has the potential to mutate to a highly pathogenic form and cause very high mortalities in some poultry species, most AI infections in poultry are due to low pathogenic AI (LPAI). Hence serological surveys, coupled with passive surveillance activities, are essential to detect sub-clinical infections by LPAI viruses, H5 and H7 subtypes. However the proper planning of an active surveillance system should be based on a careful estimation of its performance. Therefore, the sensitivity of the active surveillance system for AI in the western region of Cuba was assessed by a stochastic model quantifying the probability of revealing at least one animal infected by H5 or H7 subtype. The diagnostic sensitivity of the haemagglutination inhibition assay and different levels of within-flock prevalence (5%, 12% and 30%) were considered. The sensitivity of the surveillance system was then assessed under five different samples size scenarios: testing 20, 30, 40, 50 or 60 animals in each flock. Poultry flock sites in the western region of Cuba with a size ranging from 10,000 to 335,000 birds were included in the study.

Validità del sistema di controllo dell’influenza aviaria nella regione occidentale dell’isola di Cuba Parole chiave Cuba, Influenza aviaria, LPAI, Pollame, Sistema di sorveglianza, Sottotipo H5, Sottotipo H7.

Riassunto I sottotipi H5 e H7 del virus dell’influenza aviaria sono potenzialmente in grado di mutare in forme altamente patogene e causare un’elevata mortalità in alcune specie di volatili. Tuttavia, la maggior parte delle infezioni di influenza aviaria nel pollame sono causate da virus a bassa patogenicità (LPAI). Di conseguenza, controlli sierologici uniti a protocolli di controllo passivi sono essenziali per individuare infezioni sub-cliniche causate dai sottotipi H5 e H7 di virus LPAI. Allo stesso tempo è necessario pianificare appropriatamente il controllo attivo sulla base di una stima accurata della sua efficacia. Questo articolo riporta i risultati di test riguardanti la validità di un sistema di controllo attivo per l’influenza aviaria nella regione occidentale di Cuba. I test sono stati condotti usando un modello stocastico in grado di quantificare la probabilità di individuare almeno un animale infettato dai sottotipi H5 o H7. Sono state considerate la validità diagnostica dei test di inibizione dell’emoagglutinazione e di diverse percentuali (5%, 12% e 30%) di diffusione in uno stesso allevamento di pollame. La validità del sistema di sorveglianza è stata testata ipotizzando cinque diversi campionamenti: 20, 30, 40, 50 e 60 animali provenienti dallo stesso allevamento. Nello studio sono stati inclusi allevamenti di pollame presenti nella parte occidentale dell’isola di Cuba con un numero di animali variabile tra 10.000 e 335.000. Veterinaria Italiana 2013, 49 (1), 99-107

Surveillance system for avian influenza in Cuba

Introduction Avian influenza (AI) is a highly contagious disease, listed by the World Organisation for Animal Health (Office International des Ă&#x2030;pizooties: OIE), which has attracted much attention due to the public health implications and the effects it has on the poultry industry, given the significant economic losses suffered by countries in which AI is endemic (40). AI is caused by type A strains of the influenza virus that belong to the family Orthomyxoviridae (2, 27). The viruses that cause AI are differentiated into two groups, depending on their pathogenicity, namely: highly pathogenic avian influenza virus (HPAIV) and low pathogenic avian influenza virus (LPAIV). LPAIV is mainly responsible for respiratory illnesses and low mortalities in poultry whilst HPAIV causes systemic disease, often resulting in high mortality in turkeys and chickens (33). The antigenic differences between the two surface glycoprotein haemagglutinins (HA) and neuraminidase (NA) have enabled the identification of 16 HA subtypes (H1-H16) and 9 NA subtypes (N1N9) of AI viruses. For the purposes of the Terrestrial Animal Health Code, avian influenza is a notifiable infection of poultry caused by any avian influenza virus type A belonging to the H5 or H7 subtypes or by any AI virus with an intravenous pathogenicity index (IVPI) greater than 1.2 or causing mortality in at least 75% of cases (40). Type A influenza viruses can infect a wide range of hosts and can be pathogenic to both humans and birds (45). The antigenic characteristics of influenza virus change gradually by accumulating point mutations (antigenic drift) or suddenly by genetic re-assortment (antigenic shift) in the genes primarily encoding HA and NA. The antigenic drift leads to new antigenic variants that require a replacement of the influenza strains used in the human vaccines. The antigenic shift results in the appearance of new strains and is of great importance in the occurrence of seasonal outbreaks of human influenza. The virus strains implicated in the 20th Centuryâ&#x20AC;&#x2122;s influenza pandemics come from genetic re-assortment between avian and human viruses or transmission of the virus from animals to humans, through adaptation of purely avian strains to humans (35). Occurrences of direct bird-to-human transmission of avian influenza viruses have increasingly been reported in recent years, culminating in the outbreak of H5N1 influenza among poultry in several countries in Asia, and have caused infections in humans (11). In 1997, the first human victims of AI strain H5N1 were documented in Hong Kong (10). Although several animal species have been shown to be susceptible to influenza virus infections, three animal species besides humans appear to play a more important role in the epidemiology

100

Ferrer et al.

of influenza, namely: birds, pigs and horses (6). In particular, wild birds, ducks and geese are the reservoir of influenza virus. Transmission occurs directly or indirectly through aerosols, water, feed and other materials that have been contaminated by faeces (3). The widespread epidemic of AI in birds increases the likelihood of mutational events and genetic re-assortment. Adequate surveillance, development of vaccines, outbreak preparedness and pandemic influenza planning are important when facing an epidemic (35). Awareness of AI has increased continuously in recent years. Since 1997, the year that AI caused six deaths among 18 affected patients in Hong Kong, the focus on AI increased. After 2003, with the rapid evolution and spread of subtype H5N1, which affected poultry and wild birds in over 60 countries across 3 continents (4, 41) which resulted in the implementation of strict surveillance by animal and public health agencies both nationally and internationally. The pandemic potential of this situation remains of concern. The emergence of the disease due to HPAIV of subtype H5N1 was associated with laboratory confirmation of 602 human cases, 355 of which were fatal (42). Many control policies for the prevention and response to H5N1 outbreaks have been implemented and surveillance has increased, even in countries with no history of AI (45). Animal influenza viruses continue to threaten animal and public health, food security and livelihoods, whilst H5N1 AI remains endemic in several regions of the world (36). The increased relevance of AI to animal and human health has highlighted the lack of scientific information on several aspects of the disease, which has consequently hampered the adequate management of some of the recent crises (8). Among these aspects, is epidemiology and surveillance systems have been reinforced to provide an early warning mechanism in the event of an AI incursion and to ensure rapid response in the case of an AI outbreak in poultry. Although H5 and H7 subtypes of the AI virus have the potential to mutate to the highly pathogenic form to cause very high mortalities in some poultry species, most AI infections in poultry are caused by LPAI viruses (4), which may cause sub-clinical infections and could spread unnoticed to new premises. In such cases, the disease could spread regionally or globally, resulting in serious constraints for control. Serological surveys are essential to detect subclinical infections of LPAIV H5 and H7 subtypes and are applied to complement the passive surveillance component of programmes (1). The objectives of AI surveillance systems in disease-free regions are not only early detection of all incursions of AI but also to provide evidence of its absence in the population (12). In developing countries, resources

Veterinaria Italiana 2013, 49 (1), 99-107

Ferrer et al.

Surveillance system for avian influenza in Cuba

allocated for surveillance and early warning systems are often scarce or inadequate. In these countries, therefore, the implementation of relatively low cost serological surveillance programmes would be beneficial. However, the performance of these systems must be carefully evaluated to enhance the efficacy and efficiency of activities. In particular, when epidemiological data are lacking due to the absence of infection in the territory, risk assessment methodologies may help to model the expected surveillance results under different hypothetical scenarios, thus providing the veterinary authorities and decision-makers with the information to better target the surveillance activities and perform more comprehensive cost-benefit analyses. The Cuban AI surveillance system, in compliance with the OIE (40) and guidelines of the Food and Agriculture Organization (FAO) (26) is based on serology, viral isolation and nucleic acid detection according to a complex diagnostic algorithm. Serological testing is applied extensively as a screening method in active surveillance. The sensitivity of the active component of surveillance system, based on the haemagglutination inhibition assay (HIA), was studied in the western areas of Cuba where over 70% of the commercial poultry population is located and a greater risk of contact with migratory birds is present. The objective of this work was to assess the sensitivity of the serological surveillance system for AI in the western region of Cuba. This work was part of a study to re-plan the existing AI surveillance programme in the country, including a spatial analysis of the risk of exposure of Cuban poultry flocks to the AI virus introduction through migratory birds and also taking into account the biosecurity levels of the farm.

Materials and methods A surveillance system in a disease-free area is implemented with the principal objective of early detection of any incursion of the infection into the territory, thus enabling an assessment of the sensitivity of the system in detecting at least one infected animal. A simulation model was therefore developed to assess the probability that all infected animals tested negative by HIA on the poultry farms of western regions of Cuba under different scenarios of withinflock prevalence and number of tested animals. The model was based on the following assumptions: • aIl infection is detected only by serological investigations, no clinical sign or other evidence of virus presence are considered; • all flocks have the same probability of being infected;

Veterinaria Italiana 2013, 49 (1), 99-107

• the differences of within-flock prevalence are influenced only by the uncertainty of the prevalence estimations, not by animal biological variability; • the probability of detecting antibodies in an infected animal depends on sensitivity of the test alone; • the specificity of HIA is considered equal to one and only diagnostic sensitivity is taken into consideration. The model takes into account the number of poultry farms and sizes of flocks in the western areas of Cuba that are exposed to the possible introduction of AI viruses during the Spring and Autumn migrations of wild birds across the island. In particular, 115 poultry farms housing over 10,000 individuals each have been included in the model. Notifications to the OIE World Animal Health Information System (WAHIS) during 2007‑2009 were used to calculate the within flock prevalence levels. In particular, the following three levels were considered in the model: 5.74%, 10% and 31.25% (37-39). The diagnostic sensitivity of HIA reported by Stoyanov was used (32). Table I presents the input variables and the distributions used in the model. Five scenarios with different numbers of tested animals in each flock (20, 30, 40, 50 or 60) were studied. For each scenario, the model estimates the probability that all animals tested gave negative results under the different prevalence levels. Table I. Variables and distribution used in the model. Variables

Type of distribution

N = number of susceptible birds Cumulative in each flock n = number of animal tested by HIA in each flock Prev = within flock prevalence

Distribution Reference parameters Actual bird population in the 115 flocks of the region: (18) min = 10,000 max = 335,000 5 scenarios 20, 30, 40, 50, 60

Beta

Se = sensitivity Beta of HIA P = probability that all tested animals are negative I = number of infected animals in each flock

α=10; β=174 α=3; β=30 α=15; β=48

(39) (38) (37)

α=47; β=5

(32)

(1-Se* Prev)n

Prev*N

HIA = haemagglutination Inhibition assay

101

Ferrer et al.

Surveillance system for avian influenza in Cuba

The model was developed using @Risk (Palisade Corporation) (29) and Microsoft速 Excel 2007 and the simulation results were obtained after 1 000 iterations with Latin hypercube sampling.

Results The level of failure of AI serological surveillance is estimated through the probability that all animals tested by HIA gave negative results, considering 5.74%, 10% and 31.25% within-flock prevalence, respectively (Figures 1, 3 and 5) and five different scenarios of animal numbers tested in each flock. The impact of infection is estimated by the expected number of infected animals in each flock, considering three different within-flock prevalence rates (Figures 2, 4 and 6).

95 percentile 5 percentile

Probability

0.5 0.4 0.3 0.2 0.1 0.0

The aim of AI surveillance is to provide information on the temporal and spatial variation of circulating influenza viruses, in particular the epidemiology, ecology and evolution of AI viruses, to create an early warning system for the identification of viruses that have the potential to cause human disease. Surveillance is performed through both active and passive methods. Passive surveillance includes the collection of data in the absence of a formal procedure and is intended to actively collect the information, such as through voluntary submissions of samples to diagnostic laboratories. Active surveillance is aimed at establishing a systematic process for early detection of specific diseases in a population (19). Active surveillance is based on specific targeted investigation of at-risk populations for evidence of infection that may be based on detecting exposure to the agent (antibody detection by serology) or the presence of the agent (virus or antigen detection). The methods used must be modified according to the epidemiology of the disease (14). LPAIV or its genome can be detected in an individual bird for only few days, depending on several factors, whereas antibodies elicited by LPAIV are often present for the entire production life of the infected poultry (31). The facts that the majority of AI infections are caused by LPAIV (2), reinforces the importance of active surveillance based on antibody detection to H5 or H7 AI subtypes.

0.6 mean

Discussion

Figure 1. Probability (mean, 5th and 95th values of the simulated distribution) that all tested animals give a negative results using the haemagglutination inhibition assay (within flock prevalence = 5.7%).

0.25

The active surveillance system for AI in Cuba is mainly based on antibody detection by HIA to H5 and H7 subtypes (17). HIA is considered the serological gold standard for AI with near-perfect accuracy of sensitivity greater than 98% (9). However, in order to test the active surveillance system under

0.2 mean

0.45

95 percentile 5 percentile

0.40 0.1

Probability

0.50 0.15

0.05

0.35 0.30 0.25 0.20 0.15

>16500

15501-16500

13501-14500

14501-15500

11501-12500

12501-13500

10501-11500

8501-9500

9501-10500

7501-8500

6501-7500

5501-6500

4501-5500

3501-4500

2501-3500

1501-2500

0-500

501-1500

0.10 0.05 0.00

no. infected animals in each flock

Figure 2. Expected number of infected animals in each flock (within flock prevalence = 5.7%).

102

Figure 3. Probability (mean, 5th and 95th values of the simulated distribution) that all tested animals give a negative results by haemagglutination inhibition assay (within flock prevalence = 10%).

Veterinaria Italiana 2013, 49 (1), 99-107

Ferrer et al.

Surveillance system for avian influenza in Cuba

The European Union, together with other authorities from different geographic areas, has proposed additional surveillance studies within their AI surveillance network, with the aim of controlling H5 and H7 inďŹ&#x201A;uenza viruses. Some of these studies have been focused on wild birds, proposed as a potential early warning system, particularly in those regions where there is a large interface of human-animal contact (15). In Europe, surveillance programmes for wild birds always involves virological surveillance. Anseriformes (water fowl) and charadriiformes (shorebirds and gulls) are the main sampling targets. Active surveillance is conducted on living, clinically healthy and/or clinically diseased, injured or hunted birds. Cloacal swabs, fresh faeces and tracheal or oropharyngeal samples are collected. Passive surveillance is conducted on sick and dead wild birds. Cloacal and tracheal or oropharyngeal swabs and/or tissue (brain, heart, lung, trachea, kidney and intestines) are collected for virus isolation and molecular detection (30). In recent years several studies have been conducted in other non-European countries to improve AI surveillance. In countries with large domestic duck populations, the control of AI H5N1 infection is considered an important component of the overall control programme (21). Ducks can be asymptomatic carriers of AIV and can play an important role in the transmission of the virus. A study conducted in the Republic of Korea reported that domestic ducks showed no distinctive clinical signs except for a drop in egg production in two of three H5N1 HPAI outbreaks (24). A nation-wide active surveillance of

domestic ducks was implemented to control HPAI, including virus isolation to identify infected animals and serological testing for antibody detection. The study was conducted in laying breeder ducks and demonstrated the validity of the egg yolk antibody as alternative source to serum for AI virus antibody (23). In China, Vietnam and Indonesia, vaccination of ducks and other poultry in small commercial farms, villages and households is practised with inactivated H5N1 vaccines (25). In these countries, a surveillance system is designed to determine target levels of postvaccine antibody response and to ensure that H5N1 virus is circulating in vaccinated duck flocks. Another study, conducted in 2008, investigated an alternative strategy that involved the use of an exogenous positive marker of vaccination in domestic and wild ducks, to provide the relevant authorities with a tool

0.016

mean

0.014

95 percentile 5 percentile

0.012

Probability

more exigent conditions we assumed the lowest sensitivity value reported by Stoyanov (32) in spite of differences due to subtype used.

0.010 0.008 0.006 0.004 0.002 0.000

Figure 5. Probability (mean, 5th and 95th values of the simulated distribution) that all tested animals give a negative result in the haemagglutination inhibition assay (within flock prevalence = 31.25%).

0.16

0.25

0.14 0.12

Probability

0.2

0.15

0.1

0.1 0.08 0.06 0.04

0.05 0.02

no. infected animals in each flock

Figure 4. Expected number of infected animals in each flock (within flock prevalence = 10%).

Veterinaria Italiana 2013, 49 (1), 99-107

85001-90000 >90000

10001-15000 15001-20000 20001-25000 25001-30000 30001-35000 35001-40000 40001-45000 45001-50000 50001-55000 55001-60000 60001-65000 65001-70000 70001-75000 75001-80000 80001-85000

0-5000 5001-10000

0 0-500 501-2500 2501-4500 4501-6500 6501-8500 8501-10500 10501-12500 12501-14500 14501-16500 16501-18500 18501-20500 20501-22500 22501-24500 24501-26500 26501-28500 28501-30500 30501-32500 32501-34500 34501-36500 36501-38500 38501-40500 40501-42500 >42500

no. infected animals in each flock

Figure 6. Expected number of infected animals in each flock (within flock prevalence = 31.25%).

103

Surveillance system for avian influenza in Cuba

for post-vaccination surveillance and with more accurate data on the H5 vaccine coverage (22). Although it will be virtually impossible to prevent new outbreaks of influenza in humans and animals, global animal influenza virus surveillance can play a key role in the early recognition of new threats (6). Public health authorities and international organisations such as the FAO, OIE and World Health Organization (WHO) are tracking and monitoring AI virus circulation and are continuously engaged in the monitoring and characterisation of emerging viruses (3). Early detection and early warning, rapid confirmation of suspects, rapid and transparent notification, rapid response (including containment, management of poultry movement, zoning and compartmentalization, stamping out and vaccination) are key activities when faced with an outbreak of AI (30). For this reason, the OIE has adopted new standards for the quality of national veterinary services in which disease notification systems and information systems are improved so as to ensure early and accurate epidemiological information on a worldwide basis, in particular through its early warning system. According to the OIE Terrestrial Animal Health Code, surveillance strategies must take into account many variables such as the poultry species at risk, different biosecurity levels and production systems and the frequency of contact of domestic poultry with wild birds (40). The target population for surveillance aimed at identifying the infection should cover all susceptible poultry species within a country, zone or compartment. Surveillance should include random and targeted approaches using molecular, virological, serological and clinical methods (40). In addition, the FAO indicates five points that need to be taken into account when a surveillance programme for AI is planned, which are (14): • correctly identify the population at risk; • accurately identify the susceptible animal units (intensive flocks, markets, backyard farms, individual animals, etc.); • select the target prevalence level for the sampling scheme; • identify the size of the target population; • define the confidence level (95%, 99%). In compliance with European Union regulations, a sampling scheme for detecting infection in the poultry category (excluding for ducks, geese and turkeys) must be able to detect an infected flock with a confidence level of 95%, when the betweenflock prevalence is equal to 5% and the within-flock prevalence is approximately 30% (1).

104

Ferrer et al.

According to the surveillance programme for AI implemented in Cuba, 30 birds in each flock are tested annually by HIA. The results of the simulation model clearly demonstrate that 30 birds tested in each flock would detect the infection when the within-flock prevalence of infection is approximately 30%. In this case, the probability of having at least one positive result exceeds 99%. It has been reported with populations of 10,000-∞, that a prevalence of 30% and a confidence level of 95% is sufficient to detect infection in one out of nine animals (7). According to the guidelines of the European Union, the samples within each category of poultry (except ducks, geese and turkeys) must be designed to detect at least one infected bird with a 95% confidence if the seroprevalence within each shed of the holding is 30% (1, 43). When lower levels of within-flock prevalence are considered, the probability of failing to detect the infection with 30 tested birds raises levels to 7.5% and 19.8% with 10% and 5.7% within-flock prevalence, respectively. In these circumstances, the AI surveillance system clearly shows lower sensitivity, which can be restored by increasing the number of birds tested in each flock (5). The active surveillance program of the IA in the Republic of Argentina, aims to detect 15% prevalence with 95% confidence, with 20 samples per farm (13). However, these considerations are valid when a purely random sampling scheme is considered (2) and no samples are collected on the basis of clinical signs, which will increase the probability of AI detection. In addition, not all animals will be infected at the same time during an epidemic episode and, therefore, the choice of the level of prevalence is strictly related to the rapidity of recognition of infection targeted by the surveillance system (18). Increasing the sensitivity of the surveillance system, selecting lower prevalence levels for detection and, therefore, increasing the number of samples to be taken, also means that there will be a higher probability of more rapid recognition of any introduction of infection in the target population (34). The choice must be balanced, taking into account the epidemiological characteristics of the infection, the severity of the consequences of delayed diagnosis of infection and the resources available. In this regard, the limited resources usually available mean that a careful evaluation must be made to maximise the efficiency and the efficacy of surveillance activities (18). A risk-based approach is generally the best option but requires correct and scientifically valid risk assessments. The selection of the areas to be monitored, the time and frequency of sampling and the number and distribution of samples to be collected, are all aspects that need to be evaluated

Veterinaria Italiana 2013, 49 (1), 99-107

Ferrer et al.

scientifically and objectively (44). This paper describes an attempt to scientifically evaluate the impact of different choices in the number of samples to be taken within the current AI surveillance plan in Cuba. Further aspects should also be also taken into account to increase the sensitivity of the system, such as the existence of an effective passive surveillance and the appropriate timing of investigations during the period of time of increased risk of AI introduction, represented by the migratory birds that transit in Cuba during the spring and autumn. The correct choice of the target population is as important as the other variables. Geographic areas and populations most at risk to exposure of infection must be evaluated very carefully. Settlements of migratory birds must be defined geographically, as should their proximity to poultry farms. The

Surveillance system for avian influenza in Cuba

possible exposure of the latter, considering existing biosecurity measures in place and the consequences of such exposure, in terms of animal densities and the existence of the flocks greater economic importance (genetic centres, laying hens intensive farms, etc.), must also be carefully evaluated to better target the surveillance system.

Acknowledgments This study was conducted in the framework of a twinning project funded by the OIE between the OIE Collaborating Centre in Veterinary Training, Epidemiology, Food Safety and Animal Welfare (Istituto G. Caporale in Teramo) and the Centro Nacional de Sanidad Agropecuaria (CENSA) in Havana.

References 1. Alba A., Casal J., Napp S. & Martin P.A.J. 2010. Assessment of different surveillance systems for avian influenza in commercial poultry in Catalonia (NorthEastern Spain). Prev Vet Med, 97(2), 107-118. 2. Alexander D.J. 2007. An overview of the epidemiology of avian influenza. Vaccine, 25(30), 5637-5644. 3. Alexander D.J. & Capua I. 2008. Avian influenza in poultry. World’s Poultry Sci J, 64(04), 513. 4. Alexander D.J. & Brown I.H. 2009. History of highly pathogenic avian influenza. Rev Sci Tech Off Int Epiz, 28(1), 19-38. 5. Alfonso P., Percedo M.I., Abeledo M.A. & Noda J. 2008. Algunas pautas para establecer una estrategia de vigilancia epidemiológica de la Influenza Aviar. Rev. Salud Anim. 30(2), 69-77. 6. Buscaglia C. 2004. Influenza Aviar. InVet, 6(1), 1514-6634. 7. Cannon R.M. & Roe R.T. 1982. Livestock Disease survey: a field manual for veterinarians. Bureau of Resource Science, department of primary industry. Australian Government Publishing Service, Canberra, 35 p. 8. Capua I. & Alexander D.J. 2006. The challenge of avian influenza to the veterinary community. Avian Pathology, 35(3), 189-205. 9. Comin A., Toft N., Stegeman A., Klinkenberg D. & Marangon S. 2012. Serological diagnosis of avian influenza infections in poultry: is the hemogglutination inhibition test really “the gold standard”? Influenza and Other Respiratory Viruses, doi: 10.1111/j.17502659.2012.00391.x. 10. Charisis N. 2008. Avian influenza biosecurity: a key for animal and human protection. Vet Ital, 44(4), 657-669. 11. De Jong M.D. & Hien T.T. 2006. Avian influenza A (H5N1). Journal of Clinical Virology, 35(1), 2-13. 12. European Food Safety Authority (EFSA). 2008. Scientific Opinion of the Panel on Animal Health and Welfare on a request from The European Commission on Animal

Veterinaria Italiana 2013, 49 (1), 99-107

health and welfare aspects of avian influenza and the risk of its introduction into the EU poultry holdings. The EFSA Journal, 715, 1-161. 13. Espinosa C. & Borgna P. 2010. Actividades de Vigilancia para la Influenza Aviar en la República Argentina (1998-2009). Riesgo de Introducción de la Influenza Aviar en la República de Argentina: Análisis preliminar Ministerio de Agricultura, Ganadería y Pesca. SenasaBuenos Aires: IICA: 51. 14. Food and Agriculture Organization (FAO). 2004. Guiding principles for Highly Pathogenic Avian Influenza surveillance and diagnostic networks in Asia. FAO expert meeting on surveillance and diagnosis of avian influenza in Asia, Bangkok, 21-23 July 2004. 15. Fouchier R. 2003. Animal influenza virus surveillance. Vaccine, 21(16), 1754-1757. 16. Gall D., Nielsen K., Nicola A. & Renteria T. 2007. A proficiency testing method for detecting antibodies against Brucella abortus in quantitative and qualitative serological tests. Rev Sci Tech Off Int Epiz, 26(3), 819-828. 17. Instituto de Medicina Veterinaria (IMV). 2006. Programa de prevención y control influenza aviar altamente patógena. Ministerio de la Agricultura. Impresiones MINAG: Arroyo No.27 esq. a Diaria, Habana Vieja, 2-46 pp. 18. Instituto de Medicina Veterinaria (IMV). 2009. Informe de Balance Anual. Ministerio de la Agricultura. La Habana, Cuba, 82 pp. 19. Iglesias I., Perez A.M., De La Torre A., Munoz M.J., Martinez M. & Sanchez-Vizcaino J.M. 2010. Identifying areas for infectious animal disease surveillance in the absence of population data: highly pathogenic avian influenza in wild bird populations of Europe. Prev Vet Med, 96(1-2), 1-8. 20. Jackwood M.W. & Stallknecht D.E. 2007. Molecular epidemiologic studies on North American H9 avian

105

Surveillance system for avian influenza in Cuba

influenza virus isolates from waterfowl and shorebirds. Avian Dis, 51(1 Suppl), 448-50.

infectious disease surveillance. J Vet Diagn Invest, 15(6), 501-514.

21. James C.M., Foong Y.Y., Mansfield J.P., Vind A.R., Fenwick S.G. & Ellis T.M. 2008. Evaluation of a positive marker of avian influenza vaccination in ducks for use in H5N1 surveillance. Vaccine, 26(42), 5345-5351.

35. Trampuz A., Prabhu R.M., Smith T.F. & Baddour L.M. 2004. Avian influenza: a new pandemic threat? Mayo Clin Proc, 79(4), 523-30.

22. James C.M., Foong Y.Y., Mansfield J.P., Vind A.R., Fenwick S.G. & Ellis T.M. 2008. Evaluation of a positive marker of avian influenza vaccination in ducks for use in H5N1 surveillance. Vaccine, 26(42), 5345-5351. 23. Jeong O.M., Kim M.C., Kang H.M., Ha G.W., Oh J.S., Yoo J.E. 2010. Validation of egg yolk antibody based C-ELISA for avian influenza surveillance in breeder duck. Vet Microbiol, 144(3-4), 287-292. 24. Kim H.R., Park C.K., Lee Y.J., Woo G.H., Lee K.K., Oem J.K. 2009. An outbreak of highly pathogenic H5N1 avian influenza in Korea, 2008. Vet Microbiol, 141, 362-366. 25. Lee C.W. & Suarez D.L. 2005. Avian Influenza virus: Prospects for prevention and control by vaccination. Anim Health Res Rev, 6(1), 1-15. 26. León E.A., Duffy S.J., Stevenson M.A., Lockhart C. & Späth E.J. 2009. Sistema AVE de información geográfica para la asistencia en la vigilancia epidemiológica de la influenza aviar basado en el riesgo. Organización de las Naciones Unidas Para La Agricultura y la Alimentación. Roma, Italia. Manual. 7: 71. 27. Linzitto O.R., Espinoza C., Rodríguez C.A. 2005. Reseña sobre vigilancia y prevención de la influenza aviar y el rol zoonótico ABCL, 39(004), 485-492. 28. Mansley L.M., Donaldson A.I., Thrusfield M.V. & Honhold N. 2011. Destructive tension: mathematics versus experience – the progress and control of the 2001 foot and mouth disease epidemic in Great Britain. Rev Sci Tech Off Int Epiz, 30(2), 483-498. 29. Palisade Corporation. 2002. @RISK In Risk analysis and simulation, add-in for Microsoft Excel., Palisade Corporation, Version 4.5: 31 Decker Road. Newfield, NY USA 14867. 30. Pavade G., Weber-Vintzel L., Hamilton K., Dehove A. & Zepeda C. 2011. OFFLU review of avian influenza surveillance and epidemiological projects in some European, African, and Asian countries. OIE-FAO, Network of expertise on animal influenza (http:// www.offlu.net/fileadmin/home/en/offluprojects/ pdf/OFFLU_surveillance_study.pdf accessed on 4 July 2011). 31. Spickler A.R., Trampel D.W.,Roth J.A. 2008. The onset of virus shedding and clinical signs in chickens infected with high-pathogenicity and lowpathogenicity avian influenza viruses. Avian Pathol, 37, 555-577. 32. Stoyanov Z.I. 2006. Comparison of haemagglutination inhibition (HI), immunodiffusion (ID) and ELISA tests for detecting anti avian influenza antibodies in chicken after inoculations with inactivated antigen. Rev Med Vet, 157(6), 336-340. 33. Swayne D.E. & King D.J. 2003. Avian influenza and Newcastle disease. J Am Vet Med Ass, 222(11), 1534-1540. 34. Thurmond M.C. 2003. Conceptual foundations for

106

Ferrer et al.

36. Van Kerkhove M.D., Mumford E., Mounts A.W., Bresee J., Ly S., Bridges C.B., Otte J. & Braga E.M. 2011. Highly Pathogenic Avian Influenza (H5N1): Pathways of Exposure at the Animal Human Interface, a Systematic Review. PLoS One. 6(1): 682-685. 37. World Organisation for Animal Health (Office International des Épizooties: OIE). 2007. World Animal Health Information Database (WAHID Interface). Summary of Immediate notifications. Low pathogenic avian influenza (poultry). Paris, France (http://www.oie.int/wahis/public. php?page=single_report&pop=1&reportid=5481, accessed on 13 May 2010). 38. World Organisation for Animal Health (Office International des Épizooties: OIE). 2008. World Animal Health Information Database (WAHID Interface). Summary of Immediate notifications. Low pathogenic avian influenza (poultry), Paris, France. (http://www.oie.int/wahis/public. php?page=single_report&pop=1&reportid=7116, accessed on 13 May 2010). 39. World Organisation for Animal Health (Office International des Épizooties: OIE). 2009. World Animal Health Information Database (WAHID Interface). Summary of Immediate notifications. Low pathogenic avian influenza (poultry), Paris, France. (http://web.oie.int/wahis/public. php?page=single_report&pop=1&reportid=8760, accessed on 13 May 2010). 40. World Organisation for Animal Health (Office International des Épizooties: OIE). 2011. Terrestrial Animal Health Code. Volume II, Section 10.4. Avian influenza. 20th Edition, Paris, France. (http://www. oie.int/fileadmin/Home/esp/Health_standards/ tahc/2010/es_chapitre_1.10.4.pdf., accessed on 16 February 2012). 41. World Organisation for Animal Health (Office International des Épizooties: OIE). 2012. Update on highly pathogenic avian influenza in animals (type H5 and H7). Outbreaks of highly pathogenic avian influenza (subtype H5N1) in poultry. From the end of 2003 to 24 April 2012. Paris, France. (http://www. oie.int/fileadmin/Home/esp/Animal_Health_in_ the_World/docs/pdf/Graf_avian_influenza/grafico_ IAAP_02_03_2012.pdf., accessed on 2 June 2012). 42. World Health Organization (WHO). 2012. Cumulative number of confirmed human cases for avian influenza A(H5N1) reported to WHO, 2003-2012.WHO/GIP, data in HQ. (http://www.who.int/influenza/human_ animal_interface/H5N1_cumulative_table_archives/ en/, accessed on 24 April 2012). 43. Welby S., Van Den Berg T., Marche S., Houdart P., Hooyberghs J. & Mintienl K. 2010. Redesigning the Serological Surveillance Program for Notifiable Avian Influenza in Belgian Professional Poultry Holdings. Avian Dis, 54(S1), 597-605.

Veterinaria Italiana 2013, 49 (1), 99-107

Ferrer et al.

44. Willeberg P., Paisley L.G., Lind P. 2011. Epidemiological models to support animal disease surveillance activities. Rev Sci Tech Off Int Epiz, 30(2), 603-614.

Veterinaria Italiana 2013, 49 (1), 99-107

Surveillance system for avian influenza in Cuba

45. Yee K.S., Carpenter T.E., Cardona C.J. 2009. Epidemiology of H5N1 avian influenza. Comp Immunol Microbiol Infect Dis, 32(4), 325-340.

107

Pneumonia in slaughtered sheep in south-western Iran: pathological characteristics and aerobic bacterial aetiology Shahrzad Azizi1, Farzad Shahrani Korani2 & Ahmad Oryan3 1

Department of Pathobiology, Faculty of Veterinary Medicine, Shahid Bahonar University of Kerman, Kerman, Iran Azizi.shahrzad@gmail.com 2 Graduated of Veterinary Medicine, Islamic Azad University, Shahrekord Branch, Shahrekord, Iran 3 Department of Veterinary Pathology, School of Veterinary Medicine, Shiraz University, Shiraz, Iran

Keywords Actinomyces pyogenes, Corynebacterium pseudotuberculosis, Histopathology, Iran, Klebsiella pneumoniae, Pasteurella multocida, Pneumonia, Sheep, Staphylococcus aureus.

Summary In this study, the lungs of 1,000 sheep carcasses were subjected to gross examination and those suspected to be infected with pneumonia were studied at histopathological level as well as examined for presence of bacteria. Pneumonia was detected in 42 (4.2%) carcasses. Based on histopathological lesions, 45.24% were affected with suppurative bronchopneumonia, 20.93% with interstitial pneumonia, 11.9% bronchointerstitial pneumonia, 7.14% with fibrinous bronchopneumonia and 2.38% with embolic pneumonia. In addition, 11.9% of the lungs showed lung abscesses and 2.33% were affected with pleuritis without involving pulmonary parenchyma. Bacteriological examination revealed presence of ovine pathogens, such as Pasteurella multocida (24.53%), Staphylococcus aureus (20.75%), Klebsiella pneumoniae (15.09%), Corynebacterium pseudotuberculosis (7.55%) and Actinomyces pyogenes (1.89%). The most common form of pneumonia was suppurative bronchopneumonia with moderate amounts of fibrin deposits on the pleural surface and inside the bronchioles and alveoli.

Casi di polmonite in pecore macellate nel sud ovest dell’Iran: caratteristiche patologiche ed eziologia batterica aerobica Parole chiave Actinomyces pyogenes, Corynebacterium pseudotuberculosis, Iran, Istopatologia, Klebsiella pneumoniae, Pasteurella multocida, Polmonite, Pecore, Staphylococcus aureus.

Riassunto Lo studio riporta i risultati dell’esame macroscopico di 1.000 carcasse di pecore e di esami istopatologici e di isolamento di batteri contaminanti effettuati in casi di sospetta polmonite. Casi di polmonite sono stati riscontrati in 42 carcasse (4,2%). L’esame istopatologico delle lesioni ha evidenziato broncopolmonite suppurativa (45,24%), polmonite interstiziale (20,93%), polmonite broncointerstiziale (11,9%), broncopolmonite fibrinosa (7,14%) e polmonite embolica (2,38%). Inoltre, sono stati riscontrati ascessi nell’11,9% e pleurite senza interessamento del parenchima polmonare nel 2,33% dei polmoni esaminati. L’esame batteriologico ha evidenziato la presenza di alcuni patogeni ovini: Pasteurella multocida (24,53%), Staphylococcus aureus (20,75%), Klebsiella pneumoniae (15,09%), Corynebacterium pseudotuberculosis (7,55%) e Actinomyces pyogenes (1,89%). I casi di broncopolmonite suppurativa con moderate quantità di deposito di fibrina sulla superficie della pleura e all’interno di bronchioli ed alveoli rappresentano il reperto più comune. Veterinaria Italiana 2013, 49 (1), 109-118

109

Azizi et al.

Pneumonia in slaughtered sheep in south-western Iran

Introduction Respiratory diseases are common in all species of domestic animals. Pathogenesis is multifactorial, and the diseases appear due to the interaction of infectious micro-organisms (bacteria, mycoplasma, viruses and fungi), host defence, environmental factors (22) and stress (33, 38). Respiratory diseases in all major sheep-producing countries result in lamb mortality, reduced growth rate, carcass condemnation and consequent substantial economic impact on animal husbandry because of the need to activate chemotherapeutic and vaccination programmes (15, 21). Age, geographic location, nutrition and climate are determining factors on the type of microorganism causing pneumonia. In addition, rearing systems, stress factors, climatic changes, unhygienic conditions, sudden changes in feed and a low level of herd health status are stated as predisposing factors to bacteria and viruses. A wide variety of bacteria are found in the upper respiratory tract. Airways and lung parenchyma remove the infectious agents that are deposited, so that the deeper structures are less often attacked. Most of the infectious agents are normal inhabitants of the nasopharynx. These bacteria, after growing in the nose and throat, extend downwards and produce multiple bacterial infections. Pneumonia is a common disease of sheep. Lamb losses are generally connected with bacterial infections leading to pneumonia, diarrhoea, and subsequent sepsis, which is a potential complication of pneumonia (5). Sheep represent a great resource in Iran. The highlands of the Chaharmahal-Bakhtiary Province (south-west Iran) are an important part of the national economy for sheep production. Because of the important economic impact on the sheep industry, bacterial pneumonia has been extensively studied experimentally and in the field (10). This study was designed to isolate the aerobic bacteria from different types of ovine pneumonic lungs, to describe the histopathological findings, and to evaluate the frequency of different types of pneumonia in sheep. Therefore, the aim of this work is the acquisition of basic knowledge of the pathogenic bacteria associated with ovine respiratory diseases in Iran.

Materials and methods Area Chaharmahal-Bakhtiary province is located in south-west Iran. This province is 16,533 km2 in size, withinthe central part of the Zagros mountains. The latitude in the southern zone is between 31° 4’

110

and 42° 4’ north. Longitude is 49°39’ west to 51° 21’ east. Due to the natural situation of this province, it includes various climates. The origin of rainfall in the province is mostly Mediterranean and Sudanese atmospheric flows, which enter this region from the west and south and affect the region for a period of eight months (October to May). The average rainfall of the province is about 560 mm per year.

Sample collection In this study, lungs from 1,000 native sheep slaughtered at Shahrekord slaughterhouse underwent gross examination for the presence of pneumonic lesions. No details of sex, breed or husbandry conditions of the sheep were available, and all the animals were submitted for routine slaughter. The lungs with macroscopic pneumonic lesions were obtained (organs showing parasitic lesions were excluded). Following gross inspection, the samples of apparently affected lungs were taken for pathologic and microbiologic investigations. In addition, 10 lungs appearing healthy at gross inspection were examined for isolation of bacteria as a control group.

Pathological investigation Tissue samples 1 cm3 in thickness were fixed in 10% neutral-buffered formalin for histopathological examination. The samples were then dehydrated in graded ethanol and embedded in paraffin. Sections 5 µm in thickness were stained with haematoxylin and eosin and examined by an ordinary light microscope.

Bacterial isolation Samples of the affected areas were aseptically collected and placed in sterile plates kept in an icebox and were submitted to the Bacteriology Department. The outer surface of the pneumonic lungs were first seared with a heated spatula before the cut inner surface of the lungs were cultured on blood agar by contact with addition of 5% sheep blood and McConkey agar. The plates were then aerobically incubated at 37º C for 24-48 h. Subcultures were made and pure cultures of each strain were obtained. Identification of the isolated bacteria was performed according to the standard procedures and included morphology of the colonies on blood agar plates, presence and type of haemolysis, Gram staining, cytochrome oxidase, catalase, indole production, urease production, sulfhydric acid production (TSI), oxidation/ fermentation, motility and growth ability under aerobic conditions (31). Biochemical characters of isolates were determined with commercial test kits.

Veterinaria Italiana 2013, 49 (1), 109-118

Azizi et al.

Pneumonia in slaughtered sheep in south-western Iran

Results Pathological findings of pulmonary lesions The lungs of 42 (4.2%) sheep were affected at gross level by various types of pneumonia and pulmonary abscesses. No bacterial isolation and lesions were observed in ten apparently normal lungs. Pneumonias were classified into 5 subgroups according to their macroscopic appearances (texture, exudation and distribution of lesions) and microscopic findings (26). The morphologic diagnosis and bacteriological results in the 42 cases of ovine pulmonary lesions examined as well as normal lungs at gross examination are given in Tables I and II.

exudate leaked from small airways. There were numerous small, greenish-yellow purulent foci scattered throughout the affected cranioventral lobes. In the chronic phase, abscesses of different sizes, with fibrous capsules surrounding them, were observed in the affected lobes.

Suppurative bronchopneumonia

Histopathologically, neutrophil-rich exudates were present in the alveolar spaces and lumens of the bronchioles and bronchi, and in some occasions an admixture of various amounts of cell debris, mucus, fibrin, neutrophils and macrophages were observed in these areas. The inflammatory process confined to the individual lobules and normal alveoli were seen adjacent to the alveoli filled with neutrophilic exudates. In severe cases, purulent exudates completely filled the entire lumen of alveoli and bronchioles. Due to complete or partial obstruction of the airways some of the lobules showed atelectasis and/or emphysema. The bronchiolar epithelium showed necrosis and mixed populations of neutrophils and lymphocytes infiltrated the lamina propria and formed typical peribronchiolitis. Multiple areas of necrosis and occasionally sequestration foci were consistently observed within the pneumonic portions.

The most commonly encountered pneumonia was suppurative bronchopneumonia (45.24%, n = 19/42). The gross appearance showed irregular consolidation with lobular pattern. The cranial, middle and accessory lobes were the main affected areas. Consolidation involved more right lung than left. The pulmonary parenchyma was firm in texture. Depending on the age and nature of the process, the consolidated lungs varied from dark red in acute to grey-pink and grey in chronic form (Figure 1). In the acute phase, the cut surface of the consolidated lobules was moist and purulent

In the chronic phase, thickening of alveolar walls, mainly by lymphocytes and macrophage infiltration, was evident. Cellular infiltration in the alveolar lumen, and occasionally walls, consisted of macrophages, lymphocytes and neutrophils. Goblet cells of the bronchiolar epithelium showed hyperplasia and changed the exudates to mucopurulent. Varying degrees of bronchiolar lymphatic tissue hyperplasia (BALT) was another common finding. In some cases, extensive peribronchiloar lymphoid accumulation narrowed the bronchiolar lumina. Occasionally,

Macroscopic and histopathological characteristics of the diagnosed pneumonia

Table I. The prevalence of pneumonia types, pulmonary lesions and associated bacteria in affected lungs compared with normal lungs. Type of pneumonia Purulent bronchopneumonia (n = 19/42) (45.24%) Fibrinous bronchopneumonia (n = 3/42) (7.14%) Interstitial pneumonia (n = 9/42) (21.43%) Embolic pneumonia (n = 1/42) (2.38%) Bronchointerstitial pneumonia (n = 5/42) (11.9%) Pulmonary abscesses (n = 5/42) (11.9%) Total bacteria Grossly normal lungs

Veterinaria Italiana 2013, 49 (1), 109-118

Actinomyces pyogenes

Total bacteria isolated

3 (5.66%)

–

18 (33.96%)

2 (3.77%)

–

4 (7.55%)

–

1 (1.89%)

–

2 (3.77%)

–

1 (1.89%)

–

1 (1.89%)

2 (3.77%)

1 (1.89%)

–

4 (7.55%)

3 (5.66%)

1 (1.89%)

2 (3.77%)

1 (1.89%)

8 (15.09%)

13 (24.53%) –

8 (15.09%) –

11 (20.75%) –

4 (7.55%) –

1 (1.89%) –

37 (69.81%) –

Pasteurella multocida

Klebsiella pneumoniae

Staphylococcus Corynebacterium aureus pseudotuberculosis

6 (11.32%)

5 (9.43%)

4 (7.55%)

2 (3.77%)

–

1 (1.89%)

111

Azizi et al.

Pneumonia in slaughtered sheep in south-western Iran

Table II. Details of morphological diagnosis and bacteriological results in 42 cases of ovine pneumonia and pulmonary lesions as well as apparently normal lungs. Number of cases (n = 42)

Morphological diagnosis

Purulent bronchopneumonia

Fibrinous bronchopneumoni

Interstitial pneumonia

1 5

Embolic pneumonia Bronchointerstitial pneumonia

Pulmonary abscesses

Normal lungs

severe chronic suppurative pleuropneumonia resulted in abscess formation. As shown in Tables I and II, the organism most commonly isolated from the suppurative bronchopneumonia was Pasteurella multocida. Other bacteria isolated included Klebsiella pneumoniae, Corynebacterium pseudotuberculosis and Staphylococcus aureus.

Fibrinous bronchopneumonia Fibrinous bronchopneumonia was detected in 7.14% (n = 3/42) of the affected animals (Table I). Macroscopic lesions were similar to suppurative bronchopneumonia except that fibrin was dominant and the lesions were lobar in nature. The distribution of the lesions was almost anteroventral. The apical and cardiac lobes were the most affected parts, but in severe cases infection was more extensive, involving substantial portions of the diaphragmatic lobe. The diseased portions of lung became remarkably consolidated, dark red in colour and firm in consistency due to pulmonary congestion, oedema and fibrin accumulation. A thin layer of fibrin usually covered the pleural surface of the affected lobules. Microscopically,

112

Bacteria isolated Klebsiella pneumoniae (n = 2) K. pneumoniae + Staphylococcus aureus (n = 1) K. pneumoniae + Corynebacterium pseudotuberculosis (n = 2) Pasteurella multocida (n = 5) Staphylococcus aureus (n = 3) P. multocida + C. pseudotuberculosis (n =1) No bacteria isolated (n = 5) P. multocida + S. aureus (n = 2) No bacteria isolated (n = 1) S. aureus (n = 1) P. multocida (n = 1) No bacteria isolated (n = 7) S. aureus (n=1) K. pneumoniae (n = 1) K. pneumoniae + S. aureus (n = 1) P. multocida (n = 1) No bacteria isolated (n = 2) K. pneumoniae (n = 1) P. multocida + S. aureus (n = 2) P. multocida + C. pseudotuberculosis (n = 1) Actinomyces pyogenes (n = 1) No bacteria isolated

the affected lungs were dominated by diffuse capillary congestion. Presence of multifocal areas of necrosis and variable amounts of fibrinous exudate in the lumen of the alveoli and bronchioles were the most conspicuous and predominant features. The interlobular septa and pleura were thickened by fibrin, neutrophils and oedema. Some necrotic areas were surrounded by a rim of elongated cells, often referred to as â&#x20AC;&#x2DC;oat cellsâ&#x20AC;&#x2122;, which were severely degenerated neutrophils mixed with alveolar macrophages. Fibrinous pleurisy with or without adhesion and extensive and widespread vascular thrombosis was evident in these animals. The thrombi were usually limited to small blood vessels, capillaries and lymphatics of the pneumonic lungs (Figure 2). In the bacterial examination, P. multocida and S. aureus were isolated from these cases (Table II).

Interstitial pneumonia Nine animals (n = 9/42) (21.43%) showed interstitial pneumonia. The gross lesions were distributed throughout the lungs, often with more severe involvement of dorsocaudal regions. The affected

Veterinaria Italiana 2013, 49 (1), 109-118

Azizi et al.

Pneumonia in slaughtered sheep in south-western Iran

Figure 2. Histopathological characteristics of fibrinous bronchopneumonia. Deposits of fibrinopurulent exudate in alveoli (thin arrows) and interalveolar septa (thick arrow), presence of thrombi in capillary (arrow head). Pasteurella multocida and Staphylococcus aureus were isolated from this type of pneumonia, (H&E staining), (Scale bar: 250 µm). Figure 1. Macroscopic appearance of bronchopneumonia. Dark red consolidation of cranioventral lobes (arrows). Pasteurella multocida, Klebsiella pneumoniae, Corynebacterium pseudotuberculosis and Staphylococcus aureus were isolated from this type of pneumonia. lungs were enlarged and were diffuse red to pale in appearance. Rib impressions were seen on the costal surfaces of the diaphragmatic lobes. The affected lung failed to collapse if pressed (Figure 3). No evidence of exudates could be detected in cut surfaces or air passages. The texture of the lungs was rubbery. The histopathological features of the affected lungs showed a marked increase in mononuclear cells and occasionally mild fibrosis in the interalveolar septa and presence of varying numbers of macrophages within the alveolar lumina. In some cases, hyperplasia of pneumocyte type II was seen. Microscopically, there was no obvious exudate in the alveolar spaces and airways. Interstitial pneumonia was associated with P. multocida and S. auerus in two lung samples. No bacteria were isolated from the other seven affected cases (Tables I and II).

Bronchointerstitial pneumonia Bronchointerstitial pneumonia was detected in five animals (11.9%, n = 5/42). The affected lungs were diffuse red, wet, and failed to collapse. In contrast to interstitial pneumonia, the anterior lobes of the affected lungs showed red consolidation. Histopathologic findings revealed mixed characteristic features of suppurative bronchopneumonia and interstitial pneumonia.

Veterinaria Italiana 2013, 49 (1), 109-118

Figure 3. Macroscopic appearance of interstitial pneumonia. Affected lung is pale, enlarged, uncollapsed and rib impressions are observed on diaphragmatic surface. Pasteurella multocida and Staphylococcus aureus were isolated from this type of pneumonia. While the alveolar walls were thickened mainly by lymphocytes and macrophages, neutrophils and fibrin were accumulated in the alveolar lumens and the air passages. In some cases, multinucleated giant cells were seen in the alveoli. Pneumocyte type II showed mild hyperplasia. Alveolar collapse was also seen. There were spheroidal to ovoid concentrically

113

Azizi et al.

Pneumonia in slaughtered sheep in south-western Iran

Figure 4. Lamellar corpora amylacea in the alveolar space of a lung affected by bronchointerstitial pneumonia (arrow) and thickening of alveolar space (arrow head), (H&E staining), (Scale bar:100 µm). lamellar corpora amylacea in some of the alveoli (Figure 4). BALT hyperplasia and perivascular cuffing was another microscopic feature. Pasteurella multocida, Klebsiella pneumoniae and S. aureus were isolated from three cases of bronchointerstitial pneumonia. No bacteria were isolated from the other two animals (Table II).

Figure 5. Macroscopic appearance of embolic pneumonia caused by Staphylococcus aureus. White, raised foci of 4-6 mm in size are observed throughout pulmonary lobes (arrow heads).

Embolic pneumonia Embolic pneumonia was detected in 2.38% (n = 1/42) of the affected sheep and gross characterization was by multifocal nodules of the same size distributed randomly throughout the pulmonary lobes. The gross lesions were white foci, small in size (4-6 mm), surrounded by a discrete, red, hemorrhagic halo (Figure 5). Microscopically, theses nodules showed multifocal neutrophilic aggregations that were randomly scattered throughout the pulmonary lobes. The bacteria isolated from embolic pneumonia was S. aureus.

Pulmonary abscesses In this study, the prevalence of pulmonary abscess (n = 5/42, 11.9%) was considered without concurrent bronchopneumonia. The abscesses ranged from 2 to 10 cm in diameter and occurred as single or sometimes multiple instances in one or more lobes. Some of them were very large and involved an entire pulmonary lobe. Lung abscesses containing viscous white-yellow odourless pus were found in the affected lung and mediastinal lymph nodes (Figure 6). Chronic abscesses were often surrounded by reactive fibrous walls. Bacteriological examinations from the centre of each abscess revealed growth of S. aureus, P. multocida, K. pneumoniae, C. pseudotuberculosis and Actinomyces pyogenes (Table II).

114

Figure 6. A large abscess containing viscous white-yellow odourless pus in the mediastinal lymph node. Staphylococcus aureus, Pasteurella multocida, Klebsiella pneumoniae, Corynebacterium pseudotuberculosis and Actinomyces pyogenes were isolated from pulmonary and mediastinal lymph node abscesses.

Microbiology findings A total of 37 bacterial isolates were collected from the 43 ovine pulmonary lesions cultured. Of these isolates, 56.75% (n = 21) were Gram-positive and 43.24% (n = 16) were Gram-negative. The rate of pure and mixed culture was 39.53% and 23.26%, respectively. A summary of the bacteriological results is presented in Tables I and II.

Veterinaria Italiana 2013, 49 (1), 109-118

Azizi et al.

Discussion Respiratory diseases are common in various species of domestic animals particularly the herbivores. Because of its major economic impact on the sheep industry, through consequences such as death, retarded growth and reduced weight-gains in recovered animals, slaughterhouse wastage, drugs and labour costs, bacterial pneumonia has been extensively studied experimentally and in the field. Daniel et al. (9) evaluated the prevalence and onset of lung lesions and their impact on growth of lambs and showed that severe lung lesions could lead to greatly decreased growth performance of the animals. The present study was designed to refine and correlate the histopathological pattern of ovine pulmonary lesions with their bacterial aetiologies. Of 1,000 lungs examined, 42 (4.2%) cases showed gross lesions of various types of pneumonia and pulmonary abscesses. Pneumonias were classified into 5 subgroups according to their macroscopic and microscopic appearances, including suppurative bronchopneumonia (45.24%), fibrinous bronchopneumonia (7.14%), interstitial (21.43%), bronchointerstitial (11.9%) and embolic pneumonia (2.38%). Also, prevalence of pulmonary abscesses was obtained (11.9%). In bacteriological examinations, Pasteurella multocida (24.53%), Staphylococcus aureus (20.75%), Klebsiella pneumoniae (15.09%), Corynebacterium pseudotuberculosis (7.55%) and Actinomyces pyogenes (1.89%) were detected as important aerobic bacterial agents and the causative factors of various pulmonary lesions. Because of financial limitations, it was not possible isolate anaerobic bacteria. Pasteurella multocida and S. aureus were isolated from two of the seven lungs affected with interstitial pneumonia. It was possible these bacteria were secondary invaders and/ or environmental contaminants because viruses are normally the etiologic agents of this type of pneumonia. In addition, lack of substantial exudate, thickening of alveolar septa and presence of multinucleated syncytial cells in alveoli affected with bronchointerstitial pneumonia suggested viral agents as the primary aetiology of interstitial pneumonia, with further contamination of the infected lungs by bacteria. Concentric lamellar bodies known as corpora amylacea were present in the alveolar spaces in this form of pneumonia. Lin etÂ al. (24) reported corpora amylacea in 36% of lambs affected by chronic nonprogressive pneumonia. They stated corpora amylacea may result from bronchiolar stenosis and stagnation of exudates. Brochointerstitial pneumonia had all characteristics of interstitial pneumonia, with neutrophils and cell debris additionally accumulating in some alveoli and

Veterinaria Italiana 2013, 49 (1), 109-118

Pneumonia in slaughtered sheep in south-western Iran

bronchioles. Ettorre et al. (13) studied pneumonia in lambs in Italy and reported lamellar concentrations of protein material known as corpora amylaceas in 1.9% of affected cases. Numerous studies have been conducted on the pathology and bacteriology of ovine pulmonary lesions in various geographic situations. These have shown that the pattern of pneumonia and their bacterial agents were variable from region to region, depending on managing system, meteorological situations, sanitation, and age. Oruc (28) studied the correlation between the bacteriologic agents of lamb pneumonia with their corresponding pathological pattern and showed a prevalence rate of 35.41% pneumonia among 740 diseased or dead lambs. The author described the lung lesions histopathologically as acute-catarrhal (17.56%), muco-purulent (14.50%), purulent-necrotic (9.54%), fibrinous (26.72%), fibrino-necrotic (5.73%), interstitial (18.32%), and verminous pneumonia (7.63%). The bacteriological study described in (28), Mannheimia haemolytica (Pasteurella haemolytica) (56.14%), E. coli (24.56%), and P. multocida (10.52%) were detected as the main causative agents of various pneumonas in lambs. The findings reported in (28) are contrary to that of the present study, in which M. haemolytica and E. coli were not isolated from any of the pneumonic lungs. Haziroglu et al. (17) studied the pathology and microbiology of pneumonic lungs and reported a prevalence of 3.6% among 13,588 lambs. Macroscopic lesions were identified as atypical pneumonia in the cranial lobes. In the histology they reported proliferative pneumonia in most cases and stated that it was normally accompanied by exudative characteristics. Mannheimia haemolytica was isolated from 51.6%, Mycoplasma ovipneumoniae from 43.0% of pneumonic lungs and 26.2% of the cases showed mixed infection by both organisms. They described a close relationship between M. haemolytica and exudative inflammation (p<0.05). Abubakr et al. (1) studied pneumonia in 35 experimentally-infected sheep and 28 goats by intratracheal route with bacteria isolated from naturally-diseased sheep and goats in Sudan. According to their findings, bronchopneumonia was associated with S. aureus and Streptococcus spp, fibrinous pneumonia was associated with Proteus mirabilis, Pseudomonas aeruginosa or a combination of these two bacteria with Mycoplasma arginini, while chronic interstitial pneumonia was associated with a combination of S. aureus, Streptococcus spp. and M. arginini. Ikede (18) studied the respiratory lesions in 60 dead sheep in a livestock station at Ibadan in Nigeria over a 15-month period and showed that purulent pneumonia (48%), fibrinous pneumonia (12%), interstitial pneumonia (5%), giant cell pneumonia (3%), verminous pneumonia

115

Pneumonia in slaughtered sheep in south-western Iran

(3%) and mycotic pneumonia (2%) were the main types of pneumonia in that farm. The organisms isolated from the pulmonary lesions of the above study included M. haemolytica, E. coli, Klebsiella spp., Staphylococcus spp., Mycoplasma arginini and Aspergillus fumigatus. Similar to the above findings, the purulent bronchopneumonia was the most common form of the pneumonia. Pasteurella multocida was the most common isolated microorganism from infected sheep, with a rate of 35.14% in the present study. Pasteurella spp. are normal bacterial flora of the upper respiratory tract and opportunistic pathogens that are normally responsible for respiratory infection in small ruminants. These bacteria cause septicemia in humans. Guillet et al. (16) reported P. multocida and meningitis in two-month-old twin infants after household exposure to a slaughtered sheep. This report emphasized that sheep could be a source of human contamination with Pasteurella species. Domestic celebrations involving religious sacrifices of these animals without appropriate hygiene may therefore be associated with infections of humans that are in close vicinity of the infected sheep. It has been postulated that the father of these twins who sacrificed the infected sheep was colonised with P. multocida and infected his two infants. Contact with pets or other possible sources of Pasteurella infection should be considered by using strict hygiene rules to minimize colonization of infection by contact. This organism is one of the main etiological agents of ovine pneumonic outbreaks, but compared to M. haemolytica (P. haemolytica) its significant contribution in ovine pneumonia has been largely ignored. No M. haemolytica was isolated from the affected lungs of the present study. Odugbo et al. (27) isolated P. multocida from pneumonic lungs of sheep and stated that Pasteurellae not belonging to M. haemolytica could not be neglected when considerations concerning ovine pasteurellosis are made. They explained that when the pneumonic pasteurellosis due to M. haemolytica became controlled in the field, the prevalence of P. multocida pneumonia increased significantly. Jaramillo-Argano et al. (20) isolated Mannheimia spp. and P. multocida strains from bovine pneumonic lungs in a slaughterhouse in Mexico and showed P. multocida (60.3%) and Mannheimia spp. (39.7%) as the main causative agents of bovine pneumonia in that area. However, in other studies, M. haemolytica was reported as the most frequent isolate of bovine pneumonic lungs (19, 37). In our study, K. pneumoniae (15.09%) was isolated from the pulmonary abscesses, chronic bronchopneumonias and relatively fewer cases of bronchointerstitial pneumonia, whereas in most of the previous studies, Klebsiella has rarely been isolated from the pneumonic lungs of different

116

Azizi et al.

species of animals. Gameel et al. (14) identified K. pneumoniae subspecies ozaenae from small nodules on the chest wall and in the lungs of sheep slaughtered in Al-Ahsa abattoir, Saudi Arabia. Ludford and Stevens (25) detected K. pneumoniae from a case of canine pneumonia. Boguta et al. (8) isolated this bacterial flora from the nasal cavity of 80 foals with upper respiratory tract infection, as well as 20 healthy foals. They isolated K. pneumoniae and E. coli from the lung abscesses and purulent bronchopneumonia. In addition, no bacterial species with recognised pathogenicity was isolated from the healthy animals. In agreement with the current study, K. pneumoniae has been found to be a common cause of lung abscesses in human medicine. Most of the lungs abscesses occur in patients with a predisposition to aspiration and systemic or local immune-compromised status, such as chronic lung disease, malignancies, urinary tract infections, septicaemia, soft tissue infections or diabetes mellitus (6, 7, 30, 36). Al-Tarazi (4) studied bacterial aetiologies together with histopathological changes of pneumonia in 284 lungs of slaughtered camels in the northern parts of Jordan and reported the prevalence rate of 10.2% pneumonia in camels. Based on his findings, the pathological lesions of the pneumonic lungs comprised chronic proliferative bronchopneumonia (20.69%), chronic pleuropneumonia (6.9%), interstitial pneumonia (58.6%) and lung abscesses (10.34%). M. haemolytica and Pseudomonas aeroginosa were the most frequent isolated organisms from cases of chronic proliferative bronchopneumonia and chronic pleuropneumonia. E. coli and Klebsiella spp. were isolated from interstitial pneumonia. S. aureus, A. pyogenes and haemolytic streptococci were isolated from the lung abscesses. In their study, the most frequentlyisolated organisms were E. coli (26.66%), Klebsiella spp. (14.66%), Pseudomonas aeruginosa (12%), S. aureus (10.66%), and M. haemolytica (6.66%). In our study, S. aureus (29.73%) was another potentially pathogenic bacterium that was isolated mainly from bronchopneumonia and pulmonary abscesses. This bacterium is the main inhabitant of the upper respiratory mucosa, playing a pathogenic role in immune-compromised hosts. It also has zoonotic implications, with opportunities for reciprocal transmission between man and domestic animals when natural barriers are compromised (3). In the present investigation, C. pseudotuberculosis was isolated from the suppurative bronchopneumonia and pulmonary abscesses, and A. pyogenes was isolated from one case of pulmonary abscess. A. pyogenes is considered to be a common inhabitant of the upper respiratory and genital tracts of domestic animals (32). It is an

Veterinaria Italiana 2013, 49 (1), 109-118

Azizi et al.

important opportunistic pathogen, responsible for suppurative infections in a variety of domestic animals. It has recently been reclassified as the genus Actinomyces on the basis of rRNA sequence (2, 12, 23, 35). C. pseudotuberculosis is a mycolic acid containing, facultative intracellular actinomycete associated with the development of abscesses in a variety of mammalian hosts (11, 29, 34). This study revealed a correlation between the type of pneumonia and the isolated bacteria in sheep. The results of the present study showed that P. multocida was the most common bacteria, suggesting its importance in respiratory problems of sheep in this area. P. multocida was more likely to cause bronchopneumonia with moderate amounts of fibrin. Therefore, its role in small ruminants should receive more attention. No bacterium was isolated from those cases that did not show apparent gross lesions of pneumonia nor in some samples of pneumonic lungs. This could have been due to antibiotic therapy before slaughter, the etiological

Pneumonia in slaughtered sheep in south-western Iran

agents disappearing and gross lesions recovereing or possibly organisms other than bacteria were involved in inducing the pneumonia. There might have been some other predisposing factors such as parasites, viruses and Mycoplasma spp. Further studies are required to determine the significance of other pathogens such as Verminous pneumonia, and pneumonias due to Mycoplasma organisms, viruses and other aetiologies.

Conclusions Pasteurella multocida was the most common bacterial agent isolated from the pneumonic lungs in this area. This coincides with some studies but differs from others that considered M. haemolytica the most pathogenic bacteria. K. Pneumoniae, reported as a rare aetiology of pneumonia in small ruminants, was also highly prevalent in this area. The prevalence of K. pneumoniae as a sheep pathogen in Iran has not been previously reported.

References 1. Abubakr M.I., Elfaki M.E., Abdalla S.A. & Kamal S.M. 1981. Pathological studies on sheep and goats pneumonia in the Sudan. II. Experimental infection. Bull Anim Health Prod Afr, 29 (1), 85-94. 2. Addo P.B. & Dennis S.M. 1977. Corynebacteria associated with diseases of cattle, sheep and goats in northern Nigeria. Br Vet J, 133, 334-339. 3. Ajuwape T.P. & Aregbesola E.A. 2002. The bacterial flora of the upper respiratory tract of normal rabbits. Isr Vet Med Assoc, 57, 1-5. 4. Al-Tarazi Y.H. 2001. Bacteriological and pathological study on pneumonia in the one-humped camel (Camelus dromedarius) in Jordan. Rev Elev Med Vet Pays Trop, 54 (2), 93-97. 5. Ayelet G., Yigezu L., Gelaye E., Tariku S. & Asmare K. 2004. Epidemiologic and serologic investigation of multifactorial respiratory disease of sheep in the central highland of Ethiopia. Intern J Appl Res Vet Med, 2 (4), 274-278. 6. Bartlett J.G. 1987. Anaerobic bacterial infections of the lung. Chest, 91, 901-909. 7. Bartlett J.G. 1993. Anaerobic bacterial infections of the lung and pleural space. Clin Infect Dis, 16, S248-S255. 8. Boguta L., Gradzki Z., Borges E., Maurin F., Kodjo A. & Winiarczyk S. 2002. Bacterial flora in foals with upper respiratory tract infections in Poland. J Vet Med B, 49, 294-297. 9. Daniel J.A., Held J.E., Brake D.G., Wulf D.M. & Epperson W. 2006. Evaluation of the prevalence and onset of lung lesions and their impact on growth of lambs. Am J Vet Res, 67 (5), 890-894. 10. Daoust P.Y. 1989. Morphological study of bacterial

Veterinaria Italiana 2013, 49 (1), 109-118

pneumonia of feedlot cattle: determination of age of lesions. Can Vet J, 30 (2), 155-160. 11. Dorella F.A., Pacheco L.G., Oliveira S.C., Miyoshi A. & Azevedo V. 2006. Corynebacterium pseudotuberculosis: microbiology, biochemical properties, pathogenesis and molecular studies of virulence. Vet Res, 37 (2), 201-218. 12. Ertas H.B., Kili A., Ozbey G. & Muz A. 2005. Isolation of Arcanobacterium (Actinomyces) pyogenes from abscessed cattle kidney and identification by PCR. Turk J Vet Anim Sci, 29, 455-459. 13. Ettorre C., Sacchini F., Scacchia M. & Della Salda L. 2007. Pneumonia of lambs in the Abruzzo region of Italy: anatomopathological and histopathological studies and localization of Mycoplasma ovipneumoniae. Vet Ital, 43 (1), 149-155. 14. Gameel A.A., el-Sanousi S.M., al-Nawawi F. & al-Shazly M.O. 1991. Association of Klebsiella organisms with pulmonary lesions in sheep. Rev Elev Med Vet Pays Trop, 44 (2), 161-164. 15. Goodwin K.A., Jackson R., Brown C., Davies P.R., Morris R.S. & Perkins N.R. 2004. Pneumonic lesions in lambs in New Zealand: patterns of prevalence and effects on production. N Z Vet J, 52 (4), 175-159. 16. Guillet C.H., Join-Lambert O. & Carbonnelle E., Ferroni A. & VachĂŠe A. 2007. Pasteurella multocida sepsis and meningitis in 2-month-old twin infants after household exposure to a slaughtered sheep. Clin Infect Dis, 45 (6), e80-81 17. Haziroglu R., Diker K.S., Gulbahar M.Y., Akan M. & Guvenc T. 1994. Studies of the pathology and microbiology of pneumonic lungs of lambs. Dtsch Tierarztl Wochenschr, 101 (11), 441-443.

117

Pneumonia in slaughtered sheep in south-western Iran

18. Ikede B.O. 1978. The pattern of respiratory lesion in goats and sheep in Nigeria, Part II: Lesions in sheep. Bull Anim Prod Afr, 16, 172-185. 19. Jaramillo M.L., Aguilar R.F. & Trigo T.F.J. 1987. Serotipificación de Pasteurella haemolytica y determinación de los tiposcapsulares de Pasteurella multocida, aisladas de pulmonesneumónicos de becerros en México. Veterinaria México, 18, 185-188. 20. Jaramillo-Aragno C.J., Hernandez-Castro R., Campuzano-Ocampo V., Suarez-Guemes F., DelgadoGonzalez R. & Trigo-Tavera F. 2007. Characterisation of Mannheimia spp. and P. multocida strains isolated from bovine pneumonic lungs in two slaughterhouses in Mexico. J Anim Vet Adv, 6 (12), 1398-1404. 21. Jones G.E., Field A.C. & Gilmour J.S., Rae A.G., Nettlrton P.F. & McLauchlan M. 1982. Effect of experimental chronic pneumonia on body weight, feed intake and carcase composition of lambs. Vet Rec, 110, 168-173. 22. Lacasta D., Ferrer L.M., Ramos J.J., Gonzalez J.M. & De Las Herasc M. 2008. Influence of climatic factors on the development of pneumonia in lambs. Small Rumin Res, 80, 28-32. 23. Lechtenberg K.F., Nagaraja T.G., Leipold H.W. & Chengappa M.M. 1988. Bacteriologic and histologic studies of hepatic abscesses in cattle. Am J Vet Res, 49, 58-62. 24. Lin X., Alley M.R., Manktelow B.W. & Slack P. 1989. Pulmonary corpora amylacea in sheep. J Comp Pathol, 10 (3), 267-274. 25. LudfordG.C. & Stevens M.S. 1958. The isolation of Klebsiella pneumoniae from a case of canine pneumonia. Aust Vet J, 34 (8), 253-255.

118

Azizi et al.

29. Peel M.M., Palmer G.G., Stacpoole A.M. & Kerr T.G. 1997. Human lymphadenitis due to Corynebacterium pseudotuberculosis: report of ten cases from Australia and review. Clin Infect Dis, 24, 185-191. 30. Podschun R. & Ullmann U. 1998. Klebsiella spp. as nosocomial pathogens: epidemiology, taxonomy, typing methods and pathogenicity factors. Clin Microbiol Rev, 11, 589-603. 31. Quinn P.J., Carter M.E., Markey B.K. & Carter G.R. 1998. Clinical veterinary microbiology, 2nd, Ed., Mosby, London, 137-143, 254-258. 32. Ramos C.P., Foster G. & Collins M.D. 1997. Phylogenetic analysis of the genus Actinomyces based on 16S rRNA gene sequences: description of Arcanobacterium bernardiae comb. nov., and Arcanobacterium pyogenes comb. nov. Int J Syst Bacteriol, 47, 46-53. 33. Roy J.H.B. 1990. Respiratory infections in the calf, management of health (J.H.B. Roy, Ed.). Butterworths, London, 132-153. 34. Songer J.G., Beckenbach K., Marshall M.M., Olson G.B. & Kelley L. 1988. Biochemical and genetic characterization of Corynebacterium pseudotuberculosis. Am J Vet Res, 49, 223-226. 35. Timoney J.F., Gillespie J.H., Scott F.W. & Barlough J.E. 1988. The genus Pasteurella In Hagan and Bruner’s microbiology and infectious diseases of domestic animals, 8th Ed. (W.A. Hagan & J.F. Timoney, eds). Comstock Publishing Associates, Cornell University Press, Ithaca, New York, 104-116.

26. McGavin M.D.& Zachary J.F. 2007. Pathologic basis of veterinary disease, 4th Ed. Mosby, St Louis, 507-517.

36. Wang J.L., Chen K.Y., Fang C.T., Hsueh P.R., Yang P.C. & Chang S.C. 2005. Changing bacteriology of adult community-acquired lung abscess in Taiwan: Klebsiella pneumoniae versus anaerobes. Clin Infect Dis, 40, 915-922.

27. Odugbo M.O., Odama L.E., Umoh J.U. & Lamorde A.G. 2006. Pasteurella multocida pneumonic infection in sheep: prevalence, clinical and pathological studies. Small Rumin Res, 66, 273-277.

37. Welsh R.D., Dye L.B., Payton M.E. & Confer A.W. 2004. Isolation and antimicrobial susceptibilities of bacterial pathogens from bovine pneumonia. J Vet Diagn Invest, 16, 426-431.

28. Oruç E. 2006. The pathologic and bacteriologic comparison of pneumonia in lambs. Turk J Vet Anim Sci, 30, 593-599.

38. Wikse S.E. & Baker J.C. 1996. The Bronchopneumonias. In Large animal internal medicine, 2nd Ed. (B.P. Smith, ed.). Mosby, St Louis, 632-650.

Veterinaria Italiana 2013, 49 (1), 109-118

Evidence for Chlamydiaceae and Parachlamydiaceae in a wild boar (Sus scrofa) population in Italy Antonietta Di Francesco1, Raffaella Baldelli1, Manuela Donati2, Claudia Cotti1, Patrizia Bassi1 & Mauro Delogu1 Department of Veterinary Medical Sciences, University of Bologna, Via Tolara di Sopra 50, 40064 Ozzano Emilia (BO), Italy antoniet.difrancesco@unibo.it 2 Section of Microbiology DESOS, University of Bologna, Policlinico S. Orsola, Via Massarenti 9, 40138 Bologna, Italy 1

Keywords Chlamydiaceae, Italy, Nested PCR, Parachlamydiaceae, Wild boar.

Summary Conjunctival swabs from 44 free-living wild boars culled during a demographic control programme applied in a Regional Park located in the Northern Italy were examined by 16S rRNA encoding gene nested PCR. In total, 22 (50%) wild boars were PCR positive. Sequencing of the amplicons identified Chlamydia suis and Chlamydia pecorum in 12 and 5 samples, respectively. For one sample found PCR positive, the nucleotide sequence could not be determined. Four conjunctival samples showed ≥ 92% sequence similarities to 16S rRNA sequences from Chlamydia-like organisms, as did large intestine, uterus, and vaginal swabs from the same four animals. Amoeba DNA was found in one Chlamydia-like organism positive conjunctival swab. To our knowledge, this is the first detection of members of the Parachlamydiaceae family in wild boars, confirming a large animal host range for Chlamydialike organisms.

Chlamydiaceae e Parachlamydiaceae riscontrate in una popolazione di cinghiali selvatici in Italia Parole chiave Chlamydiaceae, Cinghiali selvatici, Italia, Nested PCR, Parachlamydiaceae.

Riassunto Il presente studio riporta i dati riguardanti tamponi congiuntivali di 44 cinghiali selvatici. I tamponi sono stati prelevati nell’ambito di un programma di controllo demografico in un parco regionale dell’Italia del Nord. Essi sono stati esaminati tramite una nested PCR riguardante il gene che codifica il 16S rRNA. I risultati della PCR sono stati positivi per 22 (50%) cinghiali selvatici. Il sequenziamento dei segmenti amplificati ha evidenziato la presenza di Chlamydia suis in 12 campioni e di Chlamydia pecorum in 5 campioni. Non è stato possibile determinare la sequenza nucleotidica di uno dei campioni PCR-positivi. In 4 tamponi congiuntivali, ed anche in tamponi prelevati da intestino, utero e vagina degli stessi animali, sono state riscontrate analogie con sequenze di 16S rRNA di paraclamidie. In un tampone congiuntivale contenente paraclamidie è stato evidenziato DNA di ameba. Questo studio riporta per la prima volta la presenza di organismi della famiglia Parachlamydiaceae in cinghiali selvatici a conferma della diffusione di questi organismi in numerose specie animali. Veterinaria Italiana 2013, 49 (1), 119-122

119

Chlamydiaceae and Parachlamydiaceae in wild boars

Introduction Parachlamydiaceae are gram-negative obligate intracellular bacteria, showing 80-90% sequence homology of rRNA genes with Chlamydiaceae (9). Parachlamydiaceae naturally infect free-living amoebae (5), but these Chlamydia-like organisms are also able to enter and multiply within human macrophages (6), pneumocytes and lung fibroblasts (3). Among Parachlamydia, Neochlamydia and Protochlamydia genera belonging to the Parachlamydiaceae family, members of the genus Parachlamydia, and especially Parachlamydia acanthamoebae, have been investigated for a potential pathogenic role in humans and animals. In humans, P. acanthamoebae is considered an emerging agent of lower respiratory tract infections, which may cause bronchitis, bronchiolitis, community-acquired pneumonia and aspiration pneumonia. In addition, P. acanthamoebae has been linked to atherosclerosis, uveitis, urogenital infection and miscarriage. Water and free-living amoebae have mostly been suggested to be the source of the human infections (reviewed by 7). Regarding the occurrence in animals, there is evidence that Parachlamydia might represent a new abortigenic agent in cattle and small ruminants (1, 16), suggesting a potential zoonotic risk from ruminant abortion material. DNA of Chlamydia-like organisms has also been detected in cervical swabs and genital tracts of sows, in semen of boars, and in conjunctival samples of koalas, cats, guinea pigs, pigs, and sheep (reviewed by 2). The DNA detection of Chlamydia-like organisms in symptomatic and asymptomatic animals indicates exposure to these bacteria but their pathogenicity remains unclear. Detailed studies relating to the occurrence of Chlamydia-like organisms in wildlife populations are lacking, nevertheless a recent report (14) described Parachlamydia spp. detection in conjunctival swabs, faeces and internal organs of wild ruminants. Wild boar (Sus scrofa) has been suggested to represent a wildlife reservoir for the same Chlamydiaceae species detected in domestic pigs, including Chlamydia suis, Chlamydia psittaci, Chlamydia abortus and Chlamydia pecorum (8, 10, 17), but no data are currently available about the occurrence of Chlamydia-like organisms in wild boar. In the present study, we report the molecular detection of Chlamydiaceae and Parachlamydiaceae in a wild boar population in Italy.

Materials and methods Conjunctival swabs from 44 free-living wild boars (21 females and 23 males), 24 of which were aged under 10 months, were collected

120

Di Francesco et al.

from February to September 2011 from animals culled during a demographic control program applied in a Regional Park (Gessi Bolognesi e Calanchi dell’Abbadessa, 48.15 km2) located in the Emilia‑Romagna region of Northern Italy. The samples were examined for Chlamydiaceae by 16S rRNA encoding gene nested PCR. DNA was extracted using a commercial DNA Blood and Tissue Kit (QIAGEN). One extraction control constituting only kit reagents was also tested. The primary PCR amplifying a 369 bp region was performed using the following forward and reverse oligonucleotide primers: 16SIGF (5’-CGGCGTGGATGAGGCAT-3’) (9) - CL1 (5’-GCGTCGCTTCGTCAGACTT-3’). Cycling conditions were as follows: 5 min of denaturation at 95° C and 30 cycles each consisting of denaturation at 94° C for 1 min, annealing at 56° C for 1 min and extension at 72° C for 1 min. A final elongation step of 5 min at 72° C completed the reaction. In the secondary PCR amplification, a 243 bp fragment, primers CL2 (5’-TTAGTGGCGGAAGGGTTAG-3’) 16SIGR (5’-TCAGTCCCAGTGTTGGC-3’) (9), were used: 1 µl of product from the first PCR step was added to a final volume of 50 µl. PCR conditions were as described above and 20 cycles were carried out. The extraction control and a distilled water negative control were included in both PCR. The amplified products were visualized after electrophoresis in 2% agarose gel by ethidium bromide staining under UV light. The secondary PCR products were purified by using a QIAquick PCR purification kit (QIAGEN) and both strands were sequenced (Bio-Fab Research, Italy). The nucleotide sequences were compared with those available in GenBank by using the BLAST server from the National Center for Biotechnology Information (http://blast.ncbi.nlm.nih.gov/Blast. cgi.). Preliminary tests were performed on the Italian C. pecorum PV5268 isolate obtained from a bovine cervical swab and the Italian C. suis MS04 isolate obtained from a pig conjunctival swab, both characterised by molecular analysis, and C. abortus S26/3 and C. psittaci 6BC reference strains, to assess if the nested PCR and following sequencing were able to detect and differentiate these Chlamydia species.

Results In total, 22/44 (50%) wild boars, 11 females and 11 males, were positive by 16S rRNA nested PCR. The chlamydial prevalence detected in wild boars aged under 10 months was higher than that found in animals aged over 10 months (62.5% vs 35%). Twelve out of the 44 (27%) animals tested were positive for C. suis (≥ 99% sequence similarity to GenBank entry AY661797.1) and 5/44 (11%) for C. pecorum (≥ 98% sequence similarity to GenBank entry HQ457465.1).

Veterinaria Italiana 2013, 49 (1), 119-122

Di Francesco et al.

Chlamydiaceae and Parachlamydiaceae in wild boars

For one animal found PCR positive, the nucleotide sequence could not be determined because of a mixed signal, probably caused by multiple infections.

44/173 (25%) sera from wild boars hunter-killed in three Italian regions; the higher antibody prevalence was found in an area with high outdoor domestic pig breeding, suggesting an epidemiologic role of domestic pig infection for wildlife infection. This suggestion was confirmed by the present data, since the wild boars sampled in this study were from areas where free-ranging husbandry of domestic pigs was present. The finding of a Chlamydia species usually infecting ruminants, C. pecorum, is not surprising, considering previous reports (12, 13) and suggests an extended host range of individual Chlamydia species.

Four out of the 44 (9%) conjunctival samples showed varying sequence similarities (92-98%) to 16S rRNA entries from Chlamydia-like organisms reported in the GenBank database. In order to evaluate the systemic diffusion of these Chlamydia‑related bacteria, tissue samples of lung (L), pulmonary limphonodes (LN), small intestine (SI), large intestine (LI), liver (LV), uterus (U) and vaginal swab (VS) collected from the four wild boars, were subjected to the DNA extraction and nested PCR, as described above. The results of the nested PCR and species identification by nucleotide sequencing are presented in Table I.

Interestingly, 4 wild boars showed PCR positivity for Chlamydia-like organisms. We observed a low degree of systemic infection, according to Regenscheit et al. (2012). The PCR positivity of wild boar genital tract and conjunctival swabs is consistent with previous reports in other animal species, as well as the PCR positivity of the large intestine and the recent detection of Chlamydia-like organisms in faecal samples of wild ruminants (14). All 4 wild boars positive for Chlamydia‑like organisms showed PCR positivity to one or one more Chlamydia species, supporting the suggestion that Chlamydia-like organisms could be a part of a multifactorial disease (11).

As Parachlamydiaceae occur as endosymbionts in protozoa, specifically Acanthamoeba spp., DNA extracted from these PCR-positive samples and water samples taken from stagnant puddles were also screened with an Acanthamoeba specific 18S rDNA gene PCR for the presence of Acanthamoeba species, according to Schroeder et al. (18). As a positive control, Acanthamoeba castellanii ATCC 50739 was used. Only one conjunctival swab reacted positive, showing 99% sequence similarity to Acanthamoeba sp. CRIB-25 DNA (GenBank entry EU273827.1).

Amoeba DNA was found in only 1 Chlamydia‑like organism-positive sample. The persistence of Chlamydia-related bacteria in the absence of an amoebal host has been presumed from other authors (4, 15), as well as suggested by the ability of these bacteria to multiply in different mammalian cells.

Discussion Chlamydial DNA was detected in 22/44 (50%) animals. This chlamydial prevalence is consistent with data by Hotzel et al. (10), confirming the wild boar as a Chlamydia wildlife reservoir. Sequencing of the 16S rRNA PCR products identified C. suis and C. pecorum in 12 and 5 samples, respectively. The prevalence of C. suis was in line with the results of a previous seroepidemiologic study performed in Italy (8), showing a specific antibody reactivity to C. suis in

Conclusions To our knowledge, this is the first detection of members of the Parachlamydiaceae family in wild boars, confirming a large animal host range for Chlamydia-like organisms. Although the pathogenicity of these Chlamydia-like bacteria

Table I. Chlamydiaceae and Chlamydia-like organisms detected by 16S rRNA gene nested PCR and sequencing in wild boar samples. Case number

Samples (% homology) LI LV

2741 f

−

C. suis (99%)

C. psittaci (99%)

CRIB38 (93%)

CRIB38 (95%)

CRIB383 (96%)

2851 m

C. psittaci4 99%

C. psittaci4 (99%)

P. acanthamoebae5 (93%)

P. acanthamoebae5 (92%)

2994 m

C. suis1 (99%)

C. suis1 (97%)

C. suis1 (99%)

−

USC96 (98%)

34514 m

−

C. pecorum7 (99%)

C. pecorum7 (94%)

CRIB383 (94%)

C. pecorum7 (99%)

CRIB383 (98%)

VS 3

CS 3

SL = lung; LN = pulmonary lymphonode; SI = small intestine; LI = large intestine; LV = liver; U = uterus; VS = vaginal swab; CS = conjunctival swab; na = not available. f = female; m = male. GenBank accession numbers of sequences: 1AY661797.1, 2CP002807.1, 3EU683886.1, 4AB001809.1, 5JN051144.1, 6FJ160741.1, 7HQ457465.1

Veterinaria Italiana 2013, 49 (1), 119-122

121

Chlamydiaceae and Parachlamydiaceae in wild boars

is still unclear, growing evidence suggests that Parachlamydia spp. may play a role in respiratory tract infections and ocular diseases in humans. The molecular detection of organisms belonging to the Parachlamydiaceae family in wild boar could be an additional component of the zoonotic potential

Di Francesco et al.

of this animal species known to be receptive to zoonotic agents such as C. psittaci and C. abortus. In view of potential exposure of hunters and other persons handling carcasses and raw game meat, the occurrence of Chlamydia-related bacteria in wild boars should be further investigated.

References 1. Borel N., Thoma R., Spaeni P., Weilenmann R., Teankum K., Brugnera E., Zimmermann D.R., Vaughan L. & Pospischil A. 2006. Chlamydia-related abortions in cattle from Graubunden, Switzerland. Vet Pathol, 43, 702-708. 2. Borel N., Pospischil A. & Greub G. 2010. Parachlamydia acanthamoebae and its zoonotic risk. Clinical Microbiology Newsletter, 32, 185-191. 3. Casson N., Medico N., Bille J. & Greub G. 2006. Parachlamydia acanthamoebae enters and multiplies within pneumocytes and lung fibroblasts. Microbes Infect, 8, 1294-1300. 4. Corsaro D., Feroldi V., Saucedo G., Ribas F., Loret J.F. & Greub G. 2009. Novel Chlamydiales strains isolated from a water treatment plant. Environ Microbiol, 1, 188-200. 5. Greub G. & Raoult D. 2002. Crescent bodies of Parachlamydia acanthamoebae and its life cycle within Acanthamoeba polyphaga: an electron micrograph study. Appl Environ Microbiol, 68, 3076-3084. 6. Greub G., Mege J.L. & Raoult D. 2003. Parachlamydia acanthamoebae enters and multiplies within human macrophages and induces their apoptosis. Infect Immun, 71, 5979-5985. 7. Greub G. 2009. Parachlamydia acanthamoebae, an emerging agent of pneumonia. Clin Microbiol Infect, 15, 18-28. 8. Di Francesco A., Donati M., Morandi F., Renzi M., Masia M.A., Ostanello F., Salvatore D., Cevenini R. & Baldelli R. 2011. Seroepidemiologic survey for Chlamydia suis in wild boar (Sus scrofa) populations in Italy. J Wildl Dis, 47, 709-712. 9. Everett K.D., Bush R.M. & Andersen A. 1999. Emended description of the order Chlamydiales, proposal of Parachlamydiaceae fam. nov. and Simkaniaceae fam. nov., each containing one monotypic genus, revised taxonomy of the family Chlamydiaceae, including a new genus and five new species, and standards for the identification of organisms. Int J Syst Bacteriol, 49, 415-440. 10. Hotzel H., Berndt A., Melzer F. & Sachse K. 2004. Occurrence of Chlamydiaceae spp. in a wild boar

122

(Sus scrofa L.) population in Thuringia (Germany). Vet Microbiol, 103, 121-126. 11. Lutz-Wohlgroth L., Becker A., Brugnera E., Huat Z.L., Zimmermann D., Grimm F., Haessig M., Greub G., Kaps S., Spiess B., Pospischil A. & Vaughan L. 2006. Chlamydiales in Guinea-pigs and Their Zoonotic Potential. J Vet Med A Physiol Pathol Clin Med, 53, 185-193. 12. Pantchev A., Sting R., Bauerfeind R., Tyczka J. & Sachse K. 2010. Detection of all Chlamydophila and Chlamydia spp. of veterinary interest using species-specific realtime PCR assays. Comp Immunol Microbiol Infect Dis, 33, 473-484. 13. Polkinghorne A., Borel N., Becker A., Lu Z.H., Zimmermann D.R., Brugnera E., Pospischil A. & Vaughan L. 2009. Molecular evidence for chlamydial infections in the eyes of sheep. Vet Microbiol, 16, 142-146. 14. Regenscheit N., Holzwarth N., Greub G., Aeby S., Pospischil A. & Borel N. 2012. Deer as a potential wildlife reservoir for Parachlamydia species. Vet J, 193, 589-92. 15. Richter M., Matheis F., Gönczi E., Aeby S., Spiess B. & Greub G. 2010. Parachlamydia acanthamoebae in domestic cats with and without corneal disease. Vet Ophthalmol, 13, 235-237. 16. Ruhl S., Goy G., Casson N., Thoma R., Pospischil A., Greub G. & Borel N. 2008. Parachlamydia acanthamoebae infection and abortion in small ruminants. Emerging Infect Dis, 14, 1966-1968. 17. Salinas J., Caro M.R., Vicente J., Cuello F., ReyesGarcia A.R., Buendia A.J., Rodolakis A. & Gortázar C. 2009. High prevalence of antibodies against Chlamydiaceae and Chlamydophila abortus in wild ungulates using two “in house” blocking-ELISA tests. Vet Microbiol, 135, 46-53. 18. Schroeder J.M., Booton G.C., Hay J., Niszl I.A., Seal D.V., Markus M.B., Fuerst P.A. & Byers T.J. 2001. Use of subgenic 18S ribosomal DNA PCR and sequencing for genus and genotype identification of acanthamoebae from humans with keratitis and from sewage sludge. J Clin Microbiol, 39, 1903-1911.

Veterinaria Italiana 2013, 49 (1), 119-122

COMUNICAZIONE BREVE Coinfezione da parvovirus felino e canino in un gatto Mara Battilani, Andrea Balboni, Massimo Giunti, Santino Prosperi Dipartimento di Medicina Veterinaria, Alma Mater Studiorum - Università di Bologna, Via Tolara di Sopra, 50 – 40064 Ozzano dell’Emilia (BO), Italia mara.battilani@unibo.it Parole chiave Gatto, Infezione, Italia, Parvovirus canino, Virus della panleucopenia felina.

Riassunto In questo studio si riporta un caso di coinfezione da parvovirus canino (CPV) di tipo 2a e da virus della panleucopenia felina (FPV) in un gatto di 3 mesi, con la presenza di una variante di parvovirus avente caratteristiche genetiche intermedie tra CPV e FPV. Il riscontro di una variante virale caratterizzata da epitopi specifici sia per il FPV che per il CPV avvalora l’importanza del meccanismo evolutivo per mutazioni multistep nella produzione di nuove varianti e nell’emergenza di nuovi virus. Questo tipo di adattamento progressivo è già stato riscontrato durante l’emergenza del CPV. Sulla base dei risultati ottenuti, è possibile ipotizzare che il CPV abbia presumibilmente iniziato un nuovo processo di riadattamento nell’ospite felino, confermando l’importanza del salto d’ospite nell’emergenza di nuovi virus. Veterinaria Italiana 2013, 49 (1), 123-125

Recentemente sono state condotte numerose ricerche nel campo delle malattie infettive emergenti, focalizzate in particolare sulle infezioni multiple e sono stati proposti diversi modelli per approfondire i meccanismi alla base delle coinfezioni. Una conoscenza approfondita delle interazioni che si stabiliscono tra i diversi patogeni nel corso delle infezioni multiple è indispensabile ai fini di prevederne la virulenza, la persistenza nonché la diffusione delle malattie infettive emergenti (1, 15). Infezioni multiple sono state descritte anche per i parvovirus autonomi, dimostrando che i parvovirus hanno un notevole grado di variabilità e che si possono rilevare contemporaneamente in un animale infetto più varianti virali (2, 3, 4, 5, 14). Tra i carnivori, i gatti sono recettivi sia alle nuove varianti del parvovirus canino (CPV-2a, 2b e 2c) che al parvovirus della panleucopenia felina (FPV). Coinfezioni con più ceppi di parvovirus potenzialmente facilitano gli eventi ricombinanti e l’elevata eterogeneità genetica (2, 4, 6, 13). In questo studio viene segnalata una coinfezione da CPV-2a e FPV in un gatto, che ha portato alla comparsa di una variante virale avente caratteristiche intermedie tra CPV e FPV, confermando che, in natura, possono instaurarsi infezioni multiple con diverse specie di parvovirus. Un gattino maschio, di 3 mesi d’età, meticcio (“Prillium”), proveniente da una colonia felina è stato ricoverato all’Ospedale Didattico Veterinario dell’Università di Bologna per diarrea, letargia ed anoressia. All’esame clinico il gatto si presentava depresso, in decubito laterale, ipotermico e con moderata disidratazione. Un prelievo di sangue

venoso è stato eseguito per un esame emocromocitometrico e un profilo biochimico di base completo e le feci sono state analizzate per cercare eventuali parassiti intestinali e il parvovirus. Il profilo ematobiochimico mostrava panleucopenia e panipoprotidemia. La sintomatologia clinica è peggiorata nei giorni successivi, con la comparsa di una diarrea acquosa emorragica, vomito, grave depressione ed il gatto è deceduto nonostante il trattamento intensivo a cui è stato sottoposto. Sulla base della sintomatologia clinica è stato avanzato il sospetto di parvovirosi, successivamente confermato dal test immunocromatografico (SNAP canine parvovirus antigen test, IDEEX Laboratories, Inc., Westbrook, Maine, USA) eseguito sulle feci raccolte durante la fase acuta della malattia. Dai campioni fecali è stato estratto il DNA virale utilizzando un kit di estrazione a colonnine, il nucleospin tissue mini kit (Macherey-Nagel, Düren, Germany) seguendo le istruzioni rilasciate dal produttore. I geni completi VP2 ed NS sono stati amplificati utilizzando un set di primers specifici per il parvovirus canino e felino (2). I prodotti di amplificazione sono stati clonati nel vettore PCR 4/TOPO utilizzando il TOPO cloning kit (Invitrogen, Carlsbad, California), e 10 cloni ricombinanti per ogni gene sono stati purificati e sequenziati. Le sequenze nucleotidiche complete dei geni VP2 ed NS sono state comparate ed allineate utilizzando l’interfaccia web di CLUSTAL W. Numerose analisi statistiche relative alla diversità nucleotidica e variabilità di sequenza sono state condotte sul set dei dati di sequenza utilizzando il software DNASP

123

Battilani et al.

Parvovirus felino e canino in un gatto

Tabella I. Parametri indicativi della variabilità di sequenza. Cloni

SynDif

NSynDif

Cloni mutati (%)

Totale mutazioni/ basi sequenziate

Frequenza di mutazione

VP2 (n = 10) NS (n = 10)

34 6

0,00913 (SE 0,00266) 0,00075 (SE 0,00026)

19 3

15 3

60 (6/10) 30 (3/10)

34/17.450 6/17.450

2 × 10–3 3,4 × 10–4

S = siti polimorfici; π = diversità nucleotidica; n = numero sequenze; SynDif = numero totale delle mutazioni sinonime; NSynDif = numero totale delle mutazioni non sinonime

versione 5.10.00 (9). Per valutare la diversità genetica della popolazione virale si sono utilizzati come indicatori la frequenza di mutazione (numero totale di mutazioni/numero totale di basi sequenziate) e la percentuale dei cloni mutati. Poichè si è visto che i fenomeni di ricombinazione intervengono nell’evoluzione dei parvovirus (11), il set di sequenze è stato analizzato per individuare eventi ricombinanti utilizzando il metodo GARD (Genetic Algorithms for Recombination Detection) implementato nell’interfaccia web di Datamonkey (8). L’analisi delle sequenze nucleotidiche dei 10 cloni del gene VP2 ha evidenziato che il clone 6 e il clone 20 sono completamente identici fra loro, come sono uguali fra loro i cloni 1, 3, 4 e 7; gli altri cloni mostrano una similarità di sequenza che varia dal 98,3 al 99,9%. L’analisi delle sequenze nucleotidiche dei 10 cloni del gene NS ha evidenziato che i cloni 1 e 10 sono identici fra loro, come lo sono i cloni 2, 4, 5, 7, 8, 9; gli altri cloni presentano una similarità di sequenza che varia dal 99,7 al 99,9%. La frequenza di mutazione rilevata nel campione era nell’ordine di 2 × 10–3, un valore analogo ad un RNA virus, e più elevato rispetto al tasso di sostituzione per anno stimato per i parvovirus dei carnivori che è nell’ordine dei 104 - 105 nt. (12). Questo risultato supporta l’importanza della coinfezione con più specie di parvovirus come potenziale fonte di complessità e diversità genetica (Tabella 1). L’analisi dei residui critici della proteina VP2 che influenzano lo spettro d’ospite felino e canino del CPV e FPV, ha evidenziato che i cloni 6, 15 e 20 erano CPV2a, mentre i cloni 1, 3, 4, 7, 9, 12, 13 erano FPV; questo dato conferma la presenza di un’infezione mista con la presenza di 2 specie di parvovirus nello stesso paziente. Nel clone 15 vi è la contemporanea presenza dei residui caratterizzanti il CPV-2a e il FPV, per cui il clone 15 lo si può considerare un virus intermedio tra il CPV e il FPV. Poiché l’analisi con Datamonkey ha escluso che si siano verificati eventi ricombinanti,

124

si è ipotizzato che tali mutazioni siano il risultato di un adattamento del virus nella nuova specie ospite. In particolare la presenza del clone 15, che contiene gli epitopi specifici del FPV e CPV, sottolinea l’importanza del meccanismo di mutazioni multistep nel generare nuove varianti e nell’emergenza di nuovi virus. Questo tipo di adattamento a multistep è già stato documentato durante l’emergenza del CPV (7) e sulla base dei nostri risultati, si può ipotizzare che il CPV abbia iniziato un nuovo processo di ri-adattamento nell’ospite felino, confermando l’importanza del meccanismo di host switching nell’emergenza di nuovi virus. Ad oggi, non vi sono studi disponibili relativamente al decorso clinico e alla prognosi dei casi che presentano una coinfezione CPV-FPV. Analogamente a quello che avviene nel corso delle infezioni multiple nell’uomo (10), il decorso clinico della parvovirosi potrebbe essere alterato e il CPV potrebbe accelerare il decorso e la gravita dell’infezione poiché il CPV rappresenta un nuovo patogeno per il gatto (9). In alternativa la coinfezione potrebbe avere implicazioni dirette sulla persistenza virale, l’interazione ospite-virus e la patogenesi (15). Non è ancora stata chiarita quale sia la patogenicità delle varianti di CPV per il gatto, in quanto i risultati sono ad oggi controversi. Il CPV è stato di frequente isolato dalle feci di gatti clinicamente sani, suggerendo che il CPV possa determinare nel gatto forme subcliniche o miti di malattia. Inoltre il CPV è stato isolato dalle cellule mononucleate del sangue (PBMCs) dei gatti, suggerendo che il CPV possa persistere nei gatti infetti nonostante la presenza di anticorpi neutralizzanti (11). Concludendo, le infezioni multiple potrebbero aumentare le possibilità che si stabiliscano infezioni persistenti nell’ospite felino, avvalorando il ruolo epidemiologico del gatto come serbatoio e fonte di nuove varianti di parvovirus.

Veterinaria Italiana 2013, 49 (1), 123-125

Battilani et al.

Parvovirus felino e canino in un gatto

Bibliografia 1. Alizon S. & van Baalen M. 2008. Multiple infections, immune dynamics, and evolution of virulence. Am Nat, 172, E150-E168. 2. Battilani M., Gallina L., Vaccari F. & Morganti L. 2007. Coinfection with multiple variants of canine parvovirus type 2 (CPV-2). Vet Res Commun, 31 (1), 209-212. 3. Battilani M., Scagliarini A., Ciulli S., Morganti L. & Prosperi S. 2006. High genetic diversity of the VP2 of a canine parvovirus strain detected in a domestic cat. Virol, 352 (1), 22-26. 4. Gottschalck E., Alexandersen S., Cohn A., Poulsen L.A., Bloom M.E. & Aasted B. 1991. Nucleotide sequence analysis of Aleutian mink disease parvovirus shows that multiple virus types are present in infected mink. J Virol, 65, 4378-4386. 5. Hoelzer K., Shalckelton L., Holmes E.C. & Parrish C.R. 2008. Within-host genetic diversity of endemic and emerging parvoviruses of dogs and cats, J Virol, 82 (22), 11096-11105. 6. Hueffer K., Parker J.S.L., Weichert W.S., Geisel R.E., Sgro J.-Y. & Parrish C.R. 2003. The natural host range shift and subsequent evolution of canine parvovirus resulted from virus-specific binding to the canine transferring receptor. J Virol, 77 (3), 1718-1726. 7. Kosakovsky Pond S.L., Posada D., Gravenor M.B., Woelk C.H. & Frost S.D. 2006Automated phylogenetic detection of recombination using a genetic algorithm. Mol Biol Evol, 23, 1891-1901. 8. Librado P. & Rozas J.A. 2009. DnaSP v5: a software for comprehensive analysis of DNA polymorphism data, Bioinformatics, 25, 1451-1452.

Veterinaria Italiana 2013, 49 (1), 123-125

9. Lin L., Verslype C., Van Pelt J., Van Ranst M. & Fevery J. 2006. Viral interaction and clinical implications of coinfection of hepatitis C virus with other hepatitis viruses. Eur J Gastroen Hepat, 18, 1311-1319. 10. Miyazawa T., Ikeda Y., Nakamura K., Naito R., Mochizuki M., Tohya Y., Vu D., Mikami T. & Takahashi E. 1999. Isolation of feline parvovirus from peripheral blood mononuclear cells of cats in northern Vietnam. Microbiol Immunol, 43 (6), 609-612. 11. Shackelton L.A., Hoelzer K., Parrish C.R. & Holmes E.C. 2007. Comparative analysis reveals frequent recombination in the parvoviruses. J Gen Virol, 88, 3294-3301. 12. Shackelton L.A., Parrish C.R., Truyen U. & Holmes E.C. 2005. High rate of evolution associated with the emergence of carnivore parvovirus. PNAS, 102 (2), 379-384. 13. Srivastava G., Wong K.Y., Chiang A.K.S., Lam K.Y. & Tao Q. 2000. Coinfection of multiple strains of EpsteinBarr virus in immunocompetent normal individuals: reassessment of the viral carrier state. Blood, 95 (7), 2443-2445. 14. Vieira M.J., Silva E., Desario C., Decaro N., Carvalheira J., Buonavoglia C. & Thompson G. 2008. Natural coinfection with parvovirus variants in dog. Emerg Infect Dis, 14 (4), 678-679. 15. Zhang P., Sadland G.J., Feng A., Xu D. & Minchella D.J. 2007. Evolutionary implications for interactions between multiple strains of host and parasite. J Theor Biol, 248, 225-240.

125

SHORT COMMUNICATION Co-infection with feline and canine parvovirus in a cat Mara Battilani, Andrea Balboni, Massimo Giunti & Santino Prosperi Department of Veterinary Medical Sciences, Alma Mater Studiorum - University of Bologna, Via Tolara di Sopra, 50 – 40064 Ozzano dell’Emilia (BO), Italy mara.battilani@unibo.it Keywords Canine parvovirus, Cat, Feline panleukopenia virus, Italy, Multiple infections, Virus.

Summary In this study we reported a case of co-infection with canine parvovirus (CPV) type 2a and feline panleukopenia virus (FPV) in a 3-month-old male kitten, with the presence of a parvovirus variant which is a true intermediate between CPV and FPV. The report of a viral variant which contained FPV- and CPV-specific epitopes stresses the importance of the mechanism of multistep mutation in the production of new variants and in the emergence of new viruses. This type of multistep adaptation has already been documented during the emergence of CPV and on the basis of our results, it was hypothesized that CPV had presumably started a new process of readaptation in the feline host, confirming the importance of viral host switching as a mechanism for the emergence of new viruses. Veterinaria Italiana 2013, 49 (1), 127-129

Research in the field of the emerging infectious diseases has been focused recently on multiple infections and various models have been proposed to investigate this topic. A more extensive knowledge of the effective interactions of pathogens during multiple infections is essential for predicting virulence, persistence and spread of the emerging infectious diseases (1, 15). Multiple infections have been described for autonomous parvovirus in several studies showing that parvoviruses have a considerable degree of variability and that several viral variants can be detected simultaneously in an infected animal (2, 3, 4, 5, 14). Among carnivores, cats are susceptible to both new variants of canine parvovirus (CPV-2a, 2b and 2c) and feline panleukopenia virus (FPV), and coinfection with multiple parvovirus strains may occur, potentially facilitating recombination and high genetic heterogeneity (2, 4, 6, 13). In this study we reported a case of co-infection with CPV type 2a and FPV in a cat, with the presence of a parvovirus variant which is a true intermediate between CPV and FPV, confirming that multiple infections with several parvovirus species may occur. A 3-month-old male kitten, mixed breed (‘Prillium’), was referred from a feline colony to the Veterinary Teaching Hospital of the University of Bologna for watery diarrhea, lethargy and anorexia. At physical examination the cat was depressed, in lateral recumbency, hypothermic and with moderate dehydration. Venous blood was collected for a minimum database and faecal samples were

screened for intestinal parasites and parvovirus. The blood profile showed panleukopenia and panhypoprotidemia. Clinical signs worsened in the next two days with the onset of watery hemorrhagic diarrhoea, vomit and severe depression and the cat died despite intensive care treatment. On the basis of clinical signs, a presumption of parvovirus infection was tentatively made, which was confirmed by an immunochromatographic test (SNAP canine parvovirus antigen test, IDEEX Laboratories, Inc., Westbrook, Maine, USA) performed on the faecal samples collected during the acute phase of the disease. Viral DNA was extracted from specimens using a nucleospin tissue mini kit (Macherey-Nagel, Düren) according to the manufacturer’s instructions. The entire VP2 and NS genes were amplified with a set of primers designed for both feline and canine parvovirus as described previously (2). The amplification products were cloned into the PCR 4/TOPO vector using the TOPO cloning kit (Invitrogen, Carlsbad, California), and ten recombinant clones for each gene were purified and sequenced. The complete VP2 and NS nucleotide sequences obtained were compared and aligned using the CLUSTAL W web interface. A variety of statistical analyses regarding nucleotide diversity and sequence variability were conducted on the sequence data set using the versatile program DNASP version 5.10.00 (9). Mutation frequency (total number of changes/total number of bases

127

Battilani et al.

Feline and canine parvovirus in a cat

Table I. Summaries of sample sequence variability. Clones

SynDif

NSynDif

Mutated clones (%)

Total mutations/ bases sequenced

Mutation frequency

VP2 (n = 10) NS (n = 10)

34 6

0.00913 (SE 0.00266) 0.00075 (SE 0.00026)

19 3

15 3

60 (6/10) 30 (3/10)

34/17,450 6/17,450

2 × 10–3 3.4 × 10–4

S = polymorphic sites; π = nucleotide diversity; n = ample size; SynDif = total number of synonymous differences; NSynDif = total number of non-synonymous differences

sequenced) and the percentage of mutated clones were used as indicators of genetic diversity of the viral population. Since genetic recombination has been assessed as a factor in parvovirus evolution (13), sequence datasets were screened for recombination using Genetic Algorithms for recombination detection (GARD) implemented in the Datamonkey web interface (8). Sequence analysis of 10 clones of the VP2 gene were completely identical for clone 6 and clone 20 as well as for clones 1, 3, 4 and 7 at the nucleotide level; the other clones showed a sequence similarity which varied from 98.3 to 99.9%. Sequence analysis of 10 clones of the NS gene were completely identical for clone 1 and clone 10 as well as for clones 2, 4, 5, 7, 8, 9 at the nucleotide level; the other clones showed a sequence similarity which varied from 99.7 to 99.9%. The mutation frequency detected in the sample was of the order of 2 × 10–3, a value analogous to the RNA virus, and higher when compared to the annual substitution rate of the order of 104 to 105 nt. estimates of carnivore parvovirus (12). This result supported the importance of co-infection with multiple species of parvovirus as a potential source of genetic complexity and diversity (Table 1). Analysis of the critical residues of the VP2 protein that affect the canine and feline host ranges of CPV and FPV, showed that clones 6, 15 and 20 were CPV-2a; in contrast, clones 1, 3, 4, 7, 9, 12, 13 were FPV, with evidence of a mixed infection involving two species of parvovirus in the same patient. Due to the concomitant presence of residues typical of both CPV-2a and FPV sequences, clone 15 could be considered a true intermediate between CPV and FPV. Since datamonkey analysis excluded the fact that recombination events took place, it has been hypothised that these coding changes were the

128

result of virus adaptation in the new host species. The report of clone 15, which contained FPV- and CPV-specific epitopes, stresses the importance of the mechanism of multistep mutation in the production of new variants and in the emergence of new viruses. This type of multistep adaptation has already been documented during the emergence of CPV (7) and on the basis of our results, it was hypothesized that CPV had presumably started a new process of readaptation in the feline host, confirming the importance of viral host switching as a mechanism for the emergence of new viruses. To date, there are no studies available on the clinical course and prognosis of cases presenting CPV-FPV coinfection. Analagous to what happens in the presence of multiple human infections (10), the clinical course of parvovirus infection could be altered and CPV could accelerate the progression of panleukopenia infection, since CPV represents a novel pathogen for cats. Alternatively, CPV-FPV co-infection could have a direct implication for virus persistence, host-viral interaction and pathogenesis (15). The pathogenicity of the CPV variants for cats has been investigated, but the results are controversial. CPV has frequently been isolated from the feces of clinically healthy cats, suggesting that CPV causes subclinical or very mild disease in this species. Furthermore, CPV was isolated from the peripheral blood mononuclear cells (PBMCs) of cats, suggesting that CPV could persistently infect cats irrespective of the presence of neutralising antibodies (11). Finally, multiple infections could increase the chance of establishing persistent infection in the feline host, increasing the epidemiological role of the cat as a reservoir and as a source of new variants of parvoviruses.

Veterinaria Italiana 2013, 49 (1), 127-129

Battilani et al.

Feline and canine parvovirus in a cat

References 1. Alizon S. & van Baalen M. 2008. Multiple infections, immune dynamics, and evolution of virulence. Am Nat, 172, E150-E168. 2. Battilani M., Gallina L., Vaccari F. & Morganti L. 2007. Coinfection with multiple variants of canine parvovirus type 2 (CPV-2). Vet Res Commun, 31 (1), 209-212. 3. Battilani M., Scagliarini A., Ciulli S., Morganti L. & Prosperi S. 2006. High genetic diversity of the VP2 of a canine parvovirus strain detected in a domestic cat. Virol, 352 (1), 22-26. 4. Gottschalck E., Alexandersen S., Cohn A., Poulsen L.A., Bloom M.E. & Aasted B. 1991. Nucleotide sequence analysis of Aleutian mink disease parvovirus shows that multiple virus types are present in infected mink. J Virol, 65, 4378-4386. 5. Hoelzer K., Shalckelton L., Holmes E.C. & Parrish C.R. 2008. Within-host genetic diversity of endemic and emerging parvoviruses of dogs and cats, J Virol, 82 (22), 11096-11105. 6. Hueffer K., Parker J.S.L., Weichert W.S., Geisel R.E., Sgro J.-Y. & Parrish C.R. 2003. The natural host range shift and subsequent evolution of canine parvovirus resulted from virus-specific binding to the canine transferring receptor. J Virol, 77 (3), 1718-1726. 7. Kosakovsky Pond S.L., Posada D., Gravenor M.B., Woelk C.H. & Frost S.D. 2006Automated phylogenetic detection of recombination using a genetic algorithm. Mol Biol Evol, 23, 1891-1901. 8. Librado P. & Rozas J.A. 2009. DnaSP v5: a software for comprehensive analysis of DNA polymorphism data, Bioinformatics, 25, 1451-1452.

Veterinaria Italiana 2013, 49 (1), 127-129

129

a cura di Manuel Graziani

LIBRI/Book reviews

Salvatore Montinaro

Sanità animale (Poletto, pp. 280, € 45,00) www.polettoeditore.it

Un manuale strutturato per fornire un quadro d’insieme della materia e chiarire il metodo logico con il quale trovare i riferimenti scientifici e normativi necessari per affrontare il lavoro di tutti i giorni: così viene presentato Sanità animale, un volume che rispetto ai pochi testi analoghi sul mercato, si distingue per l’utilizzo parco di stralci normativi, comunque sistematicamente citati, in modo da renderne agevole la consultazione. L’attenzione maggiore è spostata verso l’analisi scientifica e normativa delle problematiche affrontate, delle peculiarità delle singole malattie e dei principi d’intervento sul campo. Il medico veterinario Salvatore Montinaro, forte dell’esperienza maturata in diverse amministrazioni (Regione, ASL, IZS) riesce nell’intento di descrivere in maniera semplice, ma completa e dettagliata, il funzionamento della sanità animale e della lotta alle malattie diffusive animali. Come esplicita nell’introduzione del capitolo Norme Veterinarie e SSN: “Non si cercherà in questa sede di ripercorrere la storia della veterinaria pubblica, dal codice di Hammurabi in poi: ci si limiterà piuttosto a uno sguardo retrospettivo nella storia recente, finalizzato a identificare in modo sistematico, partendo dal generale per arrivare al particolare, i principali punti di riferimento che caratterizzano il quadro normativo della sanità pubblica veterinaria”. Anche per questo il manuale è indirizzato soprattutto a coloro che già lavorano, o che intendono lavorare, nei servizi veterinari di sanità animale delle Aziende Sanitarie Locali, ponendosi come un testo “professionale” e un sussidio tecnico da avere a portata di mano. Salvatore Montinaro, classe 1967, svolge la sua attività professionale come dirigente del servizio veterinario di sanità animale della ASL di Nuoro, inoltre, fa parte del gruppo di esperti comunitari del TAIEX ed è consulente tecnico della FAO.

131

LIBRI/Book reviews Stefano Izzi

Sicurezza della filiera della pesca (Aracne, pp. 244, € 15,00) www.aracneeditrice.it

Questo libro, firmato dall’analista ed esperto di sicurezza e intelligence Stefano Izzi, è la diretta emanazione di un progetto finalizzato al miglioramento delle attività di prevenzione e tutela alimentare. Il progetto di studio e ricerca in questione, previsto dai programmi comunitari per la formazione sui temi della sicurezza ed i piani di comunicazione e formazione, finanziato dal Ministero delle Politiche Agricole Alimentari e Forestali, esamina gli elementi della filiera della pesca in Italia con una contestualizzazione a livello internazionale. Evidenzia le vulnerabilità del “sistema filiera” e propone soluzioni per la “riduzione del rischio” sulla sicurezza alimentare. Il fine è predisporre il necessario materiale per informare e istruire tutti gli attori della filiera e consentire, sia ai giovani che agli esperti del settore, di avere piena dimestichezza del proprio ruolo all’interno della filiera e delle sue metodologie operative. Il volume è frutto di un’analisi mirata, da un lato, all’individuazione delle vulnerabilità nel generale contesto di sicurezza nel settore della pesca e, dall’altro, all’individuazione dei rischi ai quali sono esposte le figure che operano nella filiera, dei metodi operativi e delle precauzioni disposte dalle norme a tutela della persona e dell’integrità alimentare. Sicurezza della filiera della pesca è inserito nella collana diretta da Paolo Polidori “produzioni animali e sicurezza alimentare”, focalizzata su temi di ricerca nell’ambito della nutrizione e alimentazione animale, zootecnia, ispezione degli alimenti di origine animale, clinica medica e parassitologia veterinaria.

133