SARJ Vol 21, No 4 (2015)

SouthAfrican African South

Respiratory Respiratory Journal Journal VOLUME 21

|

NUMBER 4

South African

Respiratory

Journal

OFFICIAL JOURNAL OF THE S.A. THORACIC SOCIETY

|

DECEMBER 2015

FOXAIR

everyone and it’s yours to give

Air is for

Why wait to raise expectations in your patients?

S4 FOXAIR® 50/100, 50/250 and 50/500 ACCUHALER® - 42/21.5.4/0581; 0582; 0583. Each blister contains a mixture of salmeterol xinafoate equivalent to 50 µg of salmeterol and microfine fluticasone propionate (100 µg, 250 µg or 500 µg). S4 FOXAIR® 25/50, 25/125 and 25/250 INHALER - 42/21.5.4/0244; 0245; 0246. Each single actuation provides salmeterol xinafoate equivalent to 25 µg of salmeterol and fluticasone propionate (50, 125 or 250 µg). Applicant: GlaxoSmithKline South Africa (Pty) Ltd. (Co. Reg. No. 1948/030135/07). 39 Hawkins Avenue, Epping Industria 1, Cape Town, 7460. For full prescribing information plese refer to the package insert approved by the Medicines Regulatory Authority. All adverse events should be reported by calling the Aspen Medical Hotline number or directly to GlaxoSmithKline on +27117456000.

FO/0713/933 A16773 08/13

THE SOUTH AFRICAN

RESPIRATORY JOURNAL VOLUME 21 | NUMBER 4 | DECEMBER 2015

CONTENTS EDITORIALS 89

90

Prediction values for peak expiratory flow rates – much ado about something!

U G Lalloo Asthma control in adults – unfinished business!

U G Lalloo

ORIGINAL RESEARCH

91

Factors affecting compliance and control of asthma in patients attending the Respiratory Outpatient Department, Chris Hani Baragwanath Academic Hospital S A van Blydenstein, L Nqwata, N P K Banda, P Ashmore, M L Wong Prediction of peak expiratory flow rate in a Ugandan population S Nakubulwa, K Baisley, J Levin, J Nakiyingi-Miiro, A Kamali

96

SARJ EDITOR-IN-CHIEF Prof. K Dheda DEPUTY EDITOR Prof. C Koegelenberg SECTION EDITOR Breath-taking News: Prof. E Irusen EDITORIAL BOARD Prof. G Ainslie, Prof. E Bateman, Prof. R Green, Prof. E Irusen, Prof. M Jeebhay, Prof. P Jeena, Prof. U Lalloo, Prof. A Linegar, Prof. R Masekela, Dr K Nyamande, Dr J O’Brien, Dr R Raine, Prof. G Richards, Dr R van Zyl Smit, Prof. M Wong, Prof. H Zar INTERNATIONAL EDITORIAL BOARD Prof. Adithya Cattamanchi - USA Prof. Fan Chung - UK Prof. GB Migliori - Italy Prof. Surendra Sharma - India Prof. Wing Wai Yew - China PRESIDENT SA THORACIC SOCIETY Prof. U Lalloo

HMPG

REVIEW 101 Screening for lung cancer: A review

B E Schär

CASE REPORT

108 Hypereosinophilia as a paraneoplastic phenomenon in non-small-cell lung carcinoma

L Nqwata, M L Wong, R D Mohanlal, A B Lakha 110

BREATH-TAKING NEWS

112

WHO’S WHO

113

PRODUCT NEWS

115 EVENTS

CEO AND PUBLISHER Hannah Kikaya Email: hannahk@hmpg.co.za EDITOR-IN-CHIEF Janet Seggie EXECUTIVE EDITOR Bridget Farham MANAGING EDITOR Ingrid Nye SCIENTIFIC EDITOR Simon Nye PRODUCTION MANAGER Emma Jane Couzens DTP AND DESIGN Carl Sampson HEAD OF SALES AND MARKETING Diane Smith | Tel. 012 481 2069 Email: dianes@hmpg.co.za JOURNAL ADVERTISING Charles Duke Benru de Jager Reneé van der Ryst Ladine van Heerden ONLINE SUPPORT Gertrude Fani | Tel. 021 532 1281 Email: publishing@hmpg.co.za

The Editor The South African Respiratory Journal PO Box 13725 Mowbray 7705 Telephone: 021 650 3050 Fax: 021 650 2610 Email: sarj@iafrica.com The views expressed in individual articles and advertising material are the personal views of the authors and are not necessarily shared by the editors, the advertisers or the publishers.

No articles may be reproduced without the written consent of the publishers.

FINANCE Tshepiso Mokoena HMPG BOARD OF DIRECTORS Prof. M Lukhele (Chair), Dr M R Abbas, Dr M J Grootboom, Mrs H Kikaya, Prof. E L Mazwai, Dr M Mbokota, Dr G Wolvaardt

PRINTED BY TANDYM PRINT

EDITORIAL

Prediction values for peak expiratory flow rates – much ado about something!

Prediction values for spirometric indices utilise values derived from white populations in resource-rich settings.

Traditionally, these are discounted when applied to nonwhite populations. It is encouraging to acknowledge the article by Nokubulwa et al.[1] on prediction values for peak expiratory flow rates (PEFRs) in a Ugandan population.

PEFR may be obtained from an expiratory spirogram or, as is more commonly practised, measured directly using a peak flow meter.

Measurement of PEFR is an extremely useful test for the diagnosis and management of asthma, especially in resource-limited settings where spirometry is unavailable or unreliable. In most instances the PEFR is used to monitor asthma control and not frequently used to determine target values for control, as is done for forced expiratory volume in 1 second (FEV1). It is not clear why HIV-infected persons and smokers were included in the so-called healthy population, defined in this study as individuals with the absence of respiratory symptoms. It is well known that HIV is associated with chronic obstructive pulmonary disease (COPD). Many HIV-infected persons in developing regions have had at least one episode of tuberculosis (TB), and TB is also associated with COPD.[2] It is also well known through many epidemiological studies that smokers may not have symptoms in the presence of lung disease; there is also the phenomenon of the healthy smoker effect.[3] The study would have been very helpful in understanding the influence of these factors if an additional regression equation had been calculated which excluded these groups, or which interrogated the influence of HIV and smoking in stepwise regression models. In addition, the contribution to the PEFR of the type of floor surface in the homes is unclear. The authors do acknowledge that the influence of these factors will be presented in subsequent publications. However, one cannot accept the prediction equations presented in this article in the absence of this important information.

89 SARJ VOL. 21 NO. 4 2015

There is no information on how the mini-Wright PEFR meter was calibrated. One is not informed why the standing posture was used when the PEFR is frequently measured in the sitting position in clinical practice. Of course, there is no significant difference between standing and sitting posture, but in the process of deriving prediction values, as much noise as possible must be eliminated or controlled.[4] The same applies to the time of day the measurements were made. The article would have been more instructive if the derived prediction values for PEFR were compared to currently applied prediction values.

This article serves to underline the value of PEFR in clinical practice and the derivation of regional prediction values enhances its application and relevance. PEFR measurement must be reinforced as a useful, cheap, objective functional measure of asthma control.

Like all lung function tests careful attention to technical factors is important when measuring PEFR in clinical practice.

1. Nokubulwa S, Baisley K, Levin J, Nakiyingi-Miiro J, Kamali A, Nunn A. Prediction of peak expiratory flow rate in a Ugandan population. S Afr Respir J 2015;21(4):96-99.

DOI:10.7196/SARJ.2015.v21i4.36 2. Van Zyl-Smit RN, Brunet L, Pai M, Yew WW. The convergence of the global smoking, COPD, tuberculosis, HIV, and respiratory infection epidemics. Infect Dis Clin North Am 2010;24(3):693-703. [http://dx.doi.org/10.1016/j.idc.2010.04.012] 3. Becklake MR, Lalloo UG. The ‘healthy smoker’: A phenomenon of health selection?

Respiration 1990;57(3):137-144. [http://dx.doi.org/10.1159/000195837] 4. Lalloo UG, Becklake MR, Goldsmith CM. Effect of standing versus sitting position on spirometric indices in healthy subjects. Respiration 1991;58(3-4):122-125. [http:// dx.doi.org/10.1159/000195911]

Umesh G Lalloo Director of the Enhancing Care Foundation, Durban University of Technology, Durban, South Africa

S Afr Respir J 2015;21(4):89. DOI:10.7196/SARJ.2015.v21i4.51

EDITORIAL

Asthma control in adults – unfinished business!

Asthma treatment has advanced tremendously in the last 3 decades. Medical science has provided evidencebased treatment strategies to control over 90% of persons afflicted with asthma. Evidence of the efficacy of this strategy is manifest by a significant reduction in asthma mortality globally. South Africa (SA), despite being one of the most resourced countries on the African continent, still has an unacceptably high level of mortality.[1] This is also despite the fact that, based on the advocacy of the SA Thoracic Society, every governmentrun primary healthcare clinic has inhaled corticosteroids available on prescription for asthma patients. This begs the question – why does SA have such a relatively high level of asthma mortality?

Possible explanations include: • Poor management by healthcare providers • Poor adherence to treatment plans • Studies conducted in the Western Cape may not represent the status quo in the whole country • Treatment options available in the public sector may still be suboptimal • Cofactors such as HIV and tuberculosis may contribute to the mortality and result in misclassification in the death certification.

It is reasonable to accept that all these factors conspire to explain the status of asthma care as reflected in the mortality statistics for SA. It is important to understand asthma care in SA in order to interrogate this high mortality.

The article by van Blydenstein et al. [2] in this issue provides evidence for the poor level of control of asthma in a tertiary hospital setting in Gauteng. The study reported that about 42% of the adult asthmatics were controlled according to the criteria adopted by the SA Thoracic Society asthma guideline statement. Although the rate of control is relatively high, it appears to be much better than that recorded in a study from the USA.[3] It is especially encouraging if the authors’ statement is true that mainly severe asthmatics are referred to and managed in the respiratory clinic. This is not borne out by the relatively high forced expiratory volume in 1 second (FEV1)% and FEV1/forced vital capacity (FVC) ratio across the three groups.

The high proportion of women with asthma in this cohort is interesting. It is established that women bear a disproportionate burden of asthma deaths and near-fatal asthma.[4] This observation warrants further investigation and is not likely to be simply explained by the better help-seeking behaviour of women.

There are no data on whether the diagnosis of asthma was supported by reversibility criteria and the amount of treatment received. It is important to have additional information about what particular treatments are indeed available in this academic setting respiratory clinic and whether combination inhalers, dry-powder inhalers v. pressurised metred-dose inhalers were available. It is also important for interpretation of the severity of asthma and level of control to get an idea of the quantity of controller medication used or prescribed.

In order to better interpret the findings the reader would have also appreciated a description of the staffing of the clinic in respect of pulmonologists, registrars, dedicated asthma-trained nurses and facilities.

Overall, this simple study is laudable as it raises many questions that may form the basis for future research into asthma care in SA. One is actually impressed with the relatively high level of control achieved.

But there is still ‘unfinished business’ when one aims to have the majority of asthmatics well controlled.

1. The Global Asthma Report 2014. http://www.globalasthmareport.org/burden/ burden.php (accessed 15 November 2015).

2. van Blydenstein SA, Nqwata L, Banda NPK, Ashmore P, Wong ML. Factors affecting compliance and control of asthma in patients attending the Respiratory Outpatient Department, Chris Hani Baragwanath Academic Hospital. S Afr Respir J 2015;21(4):91-95. [http://dx.doi.org/10.7196/SARJ.2015.v21i4.43] 3. Gold LS, Yeung K, Smith N, Allen-Ramey FC, Nathan RA, Sullivan SD. Asthma control, cost and race: Results from a national survey. J Asthma 2013;50(7):783-790. [http://dx.doi.org/10.3109/02770903.2013.795589] 4. Chasm RM, Pei YV, Pallin DJ, et al. Sex differences in risk of hospitalization among emergency department patients with acute asthma. Ann Allergy Asthma Immunol 2015:115(1):70-72. [http://dx.doi.org/10.1016/j.anai.2015.03.021]

Umesh G Lalloo Director of the Enhancing Care Foundation, Durban University of Technology, Durban, South Africa

S Afr Respir J 2015;21(4):90. DOI:10.7196/SARJ.2015.v21i4.50

SARJ VOL. 21 NO. 4 2015

90

ORIGINAL RESEARCH

Factors affecting compliance and control of asthma in patients attending the Respiratory Outpatient Department, Chris Hani Baragwanath Academic Hospital S A van Blydenstein,1 MB BCh, FCP(SA), MMed (Int Med), DCh (SA); L Nqwata,1 MB ChB (WSU), FCP (SA); N P K Banda,1 MBBS, MMed; P Ashmore,2 BSc (Hons), MBBS, FCP (SA), MMed (Int Med); M L Wong,1 MB BCh, DCh (SA), FCP (SA), FCCP, FRCP (Lond)

Division of Pulmonology, Chris Hani Baragwanath Academic Hospital and Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa Division of Haematology, Chris Hani Baragwanath Academic Hospital and Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa

1 2

Corresponding author: S A van Blydenstein (savanblydenstein@gmail.com)

Background. There is a sense among respiratory physicians that asthma is poorly controlled in public sector hospitals, possibly due to poor adherence and lack of knowledge regarding inhaler technique.

Objective. To describe the status of asthma control in patients attending the Respiratory Outpatient Department at Chris Hani Baragwanath Academic Hospital.

Methods. A retrospective record review was conducted on outpatient files of asthmatics known to the Respiratory Department. Data obtained included demographics, level of control and number of admissions and exacerbations.

Results. A total of 519 patient files were reviewed, 74.2% of whom were female. The mean (standard deviation) age was 47 (16.5) years.

We found 47.2% of patients were controlled, 30.4% partially controlled and 22.4% uncontrolled. Most patients (88%) had no admissions in the previous year. About 60% had not experienced exacerbations in the previous year. There were significant differences between the three groups for number of exacerbations, both per year and per lifetime, and type of steroid prescribed. For a number of significant areas, such as forced expiratory volume in one second (FEV1%) predicted and competency of inhaler technique, a large proportion of the data (>30%) had not been documented by the attending doctor in the patient files.

Conclusion. In this population of mostly middle-aged female asthmatics, less than half the patients were well-controlled despite very few admissions or exacerbations in the previous year. Documentation by clinicians of aspects indicative of asthma control was generally poor, and better documentation should be encouraged in order to improve knowledge and highlight awareness of best practice in the management of asthma.

S Afr Respir J 2015;21(4):91-95. DOI:10.7196/SARJ.2015.v21i4.43

Asthma is a chronic, inflammator y air way disease involving airway hyper-responsiveness and intermittent airflow obstruction. Although potentially under-reported, asthma affects 20% of people in sub-Saharan Africa.[1] There is a wide range in prevalence among different geographic locations, owing in part to poverty, climate, and allergen exposure.[1] There appears to be a racial discrepancy in prevalence of asthma, with Asians being less affected than Native Americans, and a gender discrepancy, with a female preponderance among adult asthmatics.[2] Patients with asthma over the age of 65 years tend to fare worse than younger patients.[3] Identification of the level of severity of asthma and determination of the factors affecting asthma control are crucial to the design and implementation of strategies to reduce the number of exacerbations and hospital admissions, thus improving outcomes. Of vital importance is the documentation of important parameters in the management of asthma, particularly in a public hospital where patients often see different doctors at each visit. These include inhaler technique, frequency of exacerbations and number of admissions to hospital.

Our impression is that the control of asthma in patients attending the Respiratory Outpatient Department (ROPD) at Chris Hani

91 SARJ VOL. 21 NO. 4 2015

Baragwanath Academic Hospital (CHBAH), Johannesburg, is suboptimal. However, no study has been performed at our hospital to interrogate this hypothesis, or to quantify the proportion of our patients with severe disease and/or poor control. We are of the opinion that these data will provide important information on asthma management relevant to the South African (SA) healthcare system.

Methods

We undertook a retrospective analysis of scheduled visits by patients with asthma attending the ROPD at CHBAH. Data collected included demographics, symptoms, disease severity, medication prescription and usage, control of asthma, inhaler technique competency and spirometry. This information was captured by the investigator using a data collection sheet that is routinely completed as part of our patients’ clinic records. The sources of information were the patients’ files, held at ROPD. Definitions of asthma control and indicators of asthma severity were based on those advocated by local SA Thoracic Society guidelines.[4] Acute exacerbations and severity were determined by the need for oral steroid use, days off work, antibiotic use, number of Emergency Department visits (exacerbations), intensive care unit admissions and general ward

ORIGINAL RESEARCH admissions due to asthma in the prior 12 months. Examination findings and spirometry results were also captured. Age, gender and comorbidities were noted in an attempt to identify any factors that were associated with poor control or severity of asthma. Treatment prescriptions were documented, as well as any comments regarding use of medication in the previous month. The diagnosis of asthma was based on clinical assessment. The diagnosis of gastro-oesophageal reflux was based on either reporting of symptoms, barium swallow or gastroscopy. Any patient ≥18 years known to have asthma who presented for a scheduled visit to ROPD within the defined period was included.

Patient demographics and clinical characteristics of the cohort group were summarised using descriptive statistics. All categorical data were analysed using the χ2-test, unless the frequency was ≤5, in which case Fisher’s exact test (two-tailed) was used. Student’s t-test was used for numerical variables when comparing well-controlled asthmatics with poorly controlled asthmatics.

Ethical approval was obtained from the Human Research Ethics Committee (Medical) of the University of the Witwatersrand (M150458), and from the CHBAH Medical Advisory Committee.

Results

A total of 586 patient files were reviewed. Sixty patients were younger than 18 years of age, leaving 526 eligible for analysis. Of these, a further 7 were excluded as their level of control was not determined. Of the study population, 74.2% were female. The median (standard deviation (SD)) age was 46 (16.5) years. With regard to levels of asthma control, 47.2% (245/519) were controlled, 30.4% were partially controlled (158/519), and 22.4% were uncontrolled (116/519). There was no significant difference in terms of age and median forced expiratory volume in one second (FEV1)/forced expiratory vital capacity (FVC) between the controlled, partially controlled and uncontrolled groups (Table 1). The controlled group, however, had statistically significantly higher percent predicted and median FEV1% than the partially controlled and uncontrolled groups (p<0.05).

The risk posed to asthma control by hospital admissions is shown by the fact that the number of admissions per year was lowest in the controlled group, compared with the uncontrolled (odds ratio (OR) 8.14, p<0.0001), and the partially controlled groups (OR 3.99, p<0.0001) (Table 2). Only the uncontrolled group had significantly more admissions per lifetime than the controlled group (OR 2.4, p=0.0025). There were similar differences in the acute exacerbations whether determined per week or per year between the controlled, partially controlled and uncontrolled groups of patients.

In almost half the cases, doctors did not document the patients’ inhaler technique (Table 2). Doctors assessed patients’ inhaler technique as ‘fair to good’ in only 59 partially controlled and 31 uncontrolled asthmatics v. the controlled group, which was significantly different from the uncontrolled group (Table 2).

The uncontrolled patients were also statistically more likely to be prescribed both inhaled and oral corticosteroids (CSs) than the controlled group (Table 2).

In all three groups of asthma control there was a median of one comorbidity. Fig. 1 illustrates the variety of comorbidities found in the cohort, and the numbers found within the levels of asthma control.

There were no significant differences between levels of asthma control for patients with co-existent gastro-oesophageal reflux disease

(GERD), eczema, hypertension, diabetes mellitus, HIV infection, heart disease, depression and other psychiatric disorders, and obesity.

There was a significant difference in asthma control in patients with co-existent rheumatoid arthritis, allergic rhinitis and previous tuberculosis. Rheumatoid arthritis was more frequent in those patients with controlled asthma compared with uncontrolled (p=0.018) and partially controlled asthma (p=0.032). Allergic rhinitis was also more commonly diagnosed in the controlled group compared with the partially controlled group (p=0.023). Previous tuberculosis was more common in the partially controlled group than the controlled group (p=0.020) (Table 3).

Discussion

Our data show that a large number of adult asthmatic patients attending the ROPD at CHBAH hospital remain either partially controlled or uncontrolled. Our patients were mainly middleaged women with one comorbidity. The level of asthma control is associated with %-predicted and median FEV1, number of admissions and exacerbations and steroid use.

The overwhelming preponderance (74.2%) of female asthmatics attending our clinic was somewhat surprising. A higher incidence of asthma among females in the adult population is described,[5-7] yet our figures are higher than generally reported. We are a tertiary level facility, and preferentially follow up patients whose asthma is difficult to control. The cluster analysis study of severe asthma (Moore et al.[8]) also found that women constituted the majority in all five cluster phenotypes, although the proportions ranged from 53% to 80%.

Females may have greater health-seeking behaviour than men.[9] This phenomenon remains an area for further study.

The median (SD) age was 47 (16.5) years, with no significant age difference between the groups stratified by the three levels of control.

This is likely a reflection of the catchment population, but is also in keeping with other studies.[7] Occupational asthma in adults, as is described by Burney et al.,[10] was unfortunately not examined in this study, as we did not record employment or type of occupation. We are thus unable to comment on the relative contribution that occupational asthma plays in the prevalence of asthma in our adult population.

We found that the majority of our patients were either partially controlled or uncontrolled, similar to the findings of an American study which showed that 74% of patients were partially controlled or uncontrolled (according to GINA guidelines).[11] A recent Italian study demonstrated that only 9.1% of asthmatics were controlled[7] and in an SA study, Mash et al.[12] reported that only 31.5% of asthmatics in the Western Cape were controlled.

The median FEV1, when expressed as a percentage of predicted, decreased with the level of control. This trend was echoed by the absolute FEV1 (L). However, the FEV1/FVC ratio was not significantly different between the groups stratified by level of control. A possible explanation for this unchanged ratio could be a persistent airflow limitation. In a number of patients, the data were obtained from spirometry values documented by the attending doctor, and the flow volume curves were not examined by the investigator.

Most patients (77.2%) had had no admissions in the previous year, and 60.1% had not experienced exacerbations in the previous year.

Predictably, the poorer the control, the more admissions patients were

SARJ VOL. 21 NO. 4 2015

92

ORIGINAL RESEARCH

Table 1. Patient characteristics Total group

Partially Well-controlled Well-controlled controlled v. partially (N=245) (N=158) controlled

Uncontrolled (N=116)

Well-controlled v. uncontrolled

Age (years), median (SD)

47 (16.5)

45 (17.9)

48 (15.6)

NS

46 (14.8)

NS

Gender (male/female), n

134/385

79/166

32/126

p=0.012; 23/93 OR 0.54; 95% CI 0.33 - 0.86

p=0.02; OR 0.52; 95% CI 0.31 - 0.88

FEV1 % predicted, median (SD) 81 (24.9)

85 (24.4)

79 (24.1)

0.015

70 (24.7)

0.000

FEV1 (L), median (SD)

1.8 (1.8)

2.0 (2.4)

1.7 (0.7)

0.010

1.6 (0.7)

0.001

FEV1/FVC, median (SD)

72 (13.5)

72.5 (13.0)

72 (13.0)

NS

70.5 (15.2)

NS

CI = confidence interval; OR = odds ratio; NS = non-significant.

Table 2. Factors associated with levels of asthma control Controlled (N=245), n (%)

Partially controlled (N=158), n (%)

Uncontrolled (N=116), n (%)

p-value

OR

95% CI

p-value

OR

95% CI

None

84 (34.3)

39 (24.7)

20 (17.2)

0.0031

2.5

1.39 - 4.51

0.25

1.39

0.84 - 2.29

Any

94 (38.4)

59 (37.3)

56 (48.3)

Unknown

67 (27.3)

60 (38)

40 (34.5)

None

209 (85.3)

110 (69.6)

61 (52.6)

<0.0001

8.16

4.55 - 14.93

<0.0001

3.71

2.09 - 6.59

At least one

21 (8.6)

41 (25.9)

50 (43.1)

Unknown

15 (6.1)

7 (4.4)

5 (4.3)

None

173 (70.6)

58 (36.7)

13 (11.2)

<0.0001

16.79

8.50 - 33.18

<0.0001

4.73

2.90 - 7.72

One or more

45 (18.4)

57 (36.1)

60 (51.7)

Unknown

27 (11.0)

43 (27.2)

43 (37.1)

None

218 (89.0)

93 (58.9)

30 (25.9)

<0.0001

6.78

3.23 - 14.22

<0.0001

6.93

3.46 - 13.90

One or more

12 (4.9)

40 (25.3)

62 (53.4)

Unknown

15 (6.1)

25 (10.2)

24 (20.7)

Not checked

155 (63.3)

78 (49.4)

68 (58.6)

0.008

1.77

1.18 - 2.65

Checked

90 (36.7)

80 (50.6)

48 (41.4)

Fair to good

76 (31.0)

59 (37.3)

30 (25.9)

Poor

14 (5.7)

21 (13.3)

18 (15.5)

Unknown

155 (63.3)

78 (49.4)

68 (58.6)

54 (22.0)

50 (31.6)

51 (44.0)

Patients with admissions per lifetime

Controlled v. uncontrolled

Controlled v. partially controlled

Patients with an admission prior year

Patients with AE last year

Patients with AE per week

Patient inhaler technique recorded NS

Quality of inhaler technique 0.041

2.41

1.1 - 5.27

<0.0001

2.83

1.76 - 4.55

NS

CS use Inhaled plus oral CS*

AE = acute exacerbations. *Numbers in rows may not add up to totals in the first row as there were missing data points.

93 SARJ VOL. 21 NO. 4 2015

0.035

1.67

1.06 - 2.62

ORIGINAL RESEARCH

None Other TB Cancer Heart disease COPD Rheumatoid arthritis

Uncontrolled

Psyciatric, excluding depression Depression

Partially controlled Controlled Total

Allergic rhinitis Eczema Obesity Diabetes mellitus GERD Hypertension 0

50

100

150

200

Patients, n

Fig. 1. Comorbidities in the cohort and in the three levels of asthma control. (TB = tuberculosis) likely to experience (OR 8.16 for admissions per year, and OR 2.5 for admissions in a lifetime). In keeping with a study conducted in the Western Cape,[12] about 1 in 5 patients (23%) were admitted in the prior year for asthma-related morbidity. If one examines data per lifetime, 59% (209/352) of patients had at least one admission to hospital for asthma-related morbidity. This reiterates the importance of prior hospital admission as a marker for suboptimal asthma control.

Predictably, there were significant differences in the number of exacerbations of asthma as measured by presentations to the Emergency Department, with the lowest numbers occurring in the controlled group when compared with either the partially controlled or uncontrolled groups.The statistically significantly increased likelihood for the uncontrolled group to receive both inhaled and oral CSs may be a marker for more severe asthma, or reflect a poorer inhaler technique.

Inhaler technique was documented by the attending physician in only 57.1% of patients. There was no correlation between the level of asthma control and whether the attending doctor checked inhaler technique; neither was there evidence that those patients who were poorly controlled had their inhaler technique checked more frequently

than those who were well controlled. One could argue that this may be one of the reasons why these patients were poorly controlled. The lack of documentation that the inhaler technique had been checked in these patients is a disappointing result for an aspect of management which is critical for good asthma control, and is included in recommendations to improve asthma control.[12] Attending physicians need to be fastidious about checking inhaler technique and recording it at every visit.

Although the quality of the inhaler tech­ nique was recorded in only 57% of patients, a significant difference was seen when compared with the level of asthma control.

The majority (75.7%) of patients whose inhaler technique was checked had fair to good technique, with an OR of 2.41 when compared with the number in the uncontrolled group of patients.

There was a median of one comorbidity in all three groups of control. Interestingly, despite the increased use of oral CSs in the uncontrolled group, there was no significant increase in CS-associated diseases, for example diabetes, obesity and hypertension.

As has been found internationally,[13] GERD was noted in our study population. However, our figure of 33% of asthmatics with GERD is relatively low and there was no observed

increase in its prevalence within either the uncontrolled or the partially controlled groups compared with the controlled group. This is difficult to interpret, as not all patients were subjected to gastroscopy or barium swallows in order to document the presence of GERD, neither were symptoms specifically recorded on our questionnaire.

It may be of interest to closely interrogate and examine the patients in the two poorly controlled groups for GERD, although the contribution of concomitant GERD with asthma may vary according to the asthma phenotype, and not with the level of control.[13] Allergic rhinitis was found in only 20.3% of the cohort, considerably lower than the 67% found in a Japanese study.[13] Allergic rhinitis was more commonly diagnosed in the controlled group compared with the uncontrolled group (p=0.023), in conflict with the Japanese data which described allergic rhinitis as an aggravating factor for poorer control.[13] However, allergic rhinitis may be under-diagnosed in the poorly controlled groups.

Rheumatoid arthritis was more frequent in those patients with controlled asthma compared with uncontrolled (p=0.018) and partially controlled (p=0.032), suggesting that it may be protective in the control of asthma. Possible explanations for this include the healthy user bias, or that the immunosuppressive drugs used to treat rheumatoid arthritis, particularly methotrexate, may have a beneficial effect in asthma.

Conclusion

In this population of mostly middle-aged female asthmatics, less than half the patients were well-controlled despite relatively few admissions or exacerbations in the previous year. Control of asthma has a significant impact on the number of exacerbations and admissions during the year, leading to higher healthcare costs in those patients who were not well-controlled. Documentation by clinicians of aspects indicative of asthma control and the routine checking of inhaler technique should be mandatory in order to improve knowledge and highlight awareness of best practice in the management of asthma.

Study limitations This study had several limitations, because it was a retrospective review, and as such,

SARJ VOL. 21 NO. 4 2015

94

ORIGINAL RESEARCH Table 3. Comorbidities in the cohort, and in levels of asthma control Total cohort (N=487), n (%)

Controlled (N=224), n (%)

Partially controlled (N=150), n (%)

Uncontrolled (N=113), n (%)

Controlled v. uncontrolled

Controlled v. partially controlled

p-value

OR

95% CI

p-value OR

95% CI

Hypertension

162 (33.3)

73 (32.6)

61 (40.7)

28 (24.8)

0.18

0.68

0.41 - 1.13

0.14

1.42

0.92 - 2.18

Gastrooesophageal reflux

162 (33.3)

78 (34.8)

53 (35.3)

31 (27.4)

0.21

0.71

0.43 - 1.16

1.00

1.02

0.66 - 1.58

Diabetes mellitus

50 (10.3)

18 (8.0)

17 (11.3)

15 (13.3)

0.18

1.75

0.85 - 3.62

0.37

1.46

0.73 - 2.94

Obesity

39 (8.0)

15 (6.7)

12 (8.0)

12 (10.6)

0.30

1.66

0.74 - 3.67

0.79

1.21

0.55 - 2.67

Allergic rhinitis

99 (20.3)

56 (25.0)

22 (14.7)

21 (18.6)

0.24

0.68

0.39 - 1.20

0.023

0.52

0.30 - 0.89

Depression

5 (1.0)

2 (0.9)

2 (1.3)

1 (0.9)

1.00

0.99

0.089 11.05

1.00

15.00

0.21 - 10.77

Rheumatoid arthritis

12 (2.5)

11 (4.9)

1 (0.7)

0 (0)

0.018

-

-

0.032

0.13

0.017 - 1.02

COPD

17 (3.5)

6 (2.7)

5 (3.3)

6 (5.3)

0.35

2.04

0.64 - 6.47

0.76

1.25

0.38 - 4.18

Heart disease

12(2.5)

2 (0.9)

8 (5.3)

2 (1.8)

0.60

2.00

0.28 14.39

0.017

6.25

1.31 -29.87

Cancer

2 (0.4)

2 (0.9)

0 (0)

0 (0)

0.55

-

-

0.52

-

-

Psychiatric disease

4 (0.8)

2 (0.9)

0 (0)

2 (1.8)

0.60

2.00

0.28 14.39

0.52

-

-

Previous tuberculosis

17 (3.5)

5 (2.2)

11 (7.3)

1 (0.9)

0.45

0.39

0.045 3.39

0.02

3.47

1.18 - 10.19

HIV

34 (7.0)

14 (6.3)

13 (8.7)

7 (6.2)

0.50

1.26

0.73 -2.18

0.92

1.01

0.60 - 1.70

Eczema

6 (1.2)

5 (2.2)

1 (0.7)

0 (0)

-

-

-

0.50

1.42

0.65 - 3.12

Other

77 (15.8)

37 (16.5)

20 (13.3)

20 (17.7)

0.92

1.09

0.60 - 1.98

0.49

0.78

0.43 - 1.40

None

98 (20.1)

43 (19.2)

29 (19.3)

26 (23.0)

-

-

-

-

-

-

COPD = chronic obstructive pulmonary disease.

accurate record-keeping was a concern, as demonstrated by the large number of data points missing, particularly in the fields of checking the inhaler technique, and lung functions results.

There was no characterisation of the phenotype of asthma as a predictor for control.

Acknowledgements. Our thanks to the staff and patients at ROPD for their willingness to help, and to Leporogo Academic Services for their assistance with compiling the study.

References 1. V Gemert F, van der Molen T, Jones R, Chavannes N. The impact of asthma and COPD in sub-Saharan Africa. Prim Care Respir J 2011;20(3):240-248.

2. Slejko JF, Ghushchyan VH, Sucher B, et al. Asthma control in the United States, 2008 - 2010: Indicators of poor asthma control. J Allergy Clin Immunol 2014;133(6):15791587. [http://dx.doi.org/10.1016/j.jaci.2013.10.028] 3. Yawn BP. Factors accounting for asthma variability: Achieving optimal symptom control for individual patients. Prim Care Respir J 2008;17(3):138-147.

4. Lalloo UAG, Wong M, Abdool-Gaffar S, et al. Guidelines for the management of chronic asthma in adolescents and adults. SA Fam Pract 2007;49(5):19-31.

95 SARJ VOL. 21 NO. 4 2015

5. Hansen S, Probst-Hensch N, Keidel D, et al. Gender differences in adult-onset asthma: Results from the Swiss SAPALDIA cohort study. Eur Respir J 2015;46(4):1011-1020. [http://dx.doi.org/10.1183/13993003.02278-2014] 6. Ehrlich RI, White N, Norman R, et al. Wheeze, asthma diagnosis and medication use: A national adult survey in a developing country. Thorax 2005;60(11):895-901. [http:// dx.doi.org/10.1136/thx.2004.030932] 7. Corrado A, Renda T, Polese G, Rossi A. Assessment of asthma control: The SERENA study.

Respir Med 2013;107(11):1659-1666. [http://dx.doi.org/10.1016/j.rmed.2013.08.019] 8. Moore WC, Meyers DA, Wenzel SE, et al. Identification of asthma phenotypes using cluster analysis in the Severe Asthma Research Program. Am J Respir Crit Care Med 2010;181(4):315-323. [http://dx.doi.org/10.1164/rccm.200906-0896OC] 9. Wang Y, Hunt K, Nazreth I, Freemantle N, Petersen I. Do men consult less than women? An analysis of routinely collected UK general practice data. BMJ Open.

2013;3(003320). [http://dx.doi.org/10.1136/bmjopen-2013-003320] 10. Burney P, Jarvis D, Perez-Padilla R. The global burden of chronic respiratory disease in adults. Int J Tuberc Lung Dis 2015;19(1):10-20. [http://dx.doi.org/10.5588/ijtld.14.0446] 11. Gold LS, Yeung K, Smith N, Allen-Ramey FC, Nathan RA, Sullivan SD. Asthma control, cost and race: Results from a national survey. J Asthma 2013;50(7):783-790. [http://dx.doi.org/10.3109/02770903.2013.795589].

12. Mash B, Rhode H, Pather M, et al. Quality of asthma care: Western Cape Province, South Africa. S Afr Med J 2009;99(12):892-896.

13. Ishizuka T, Hisada T, Kamide Y, et al. The effects of concomitant GERD, dyspepsia, and rhinosinusitis on asthma symptoms and FeNO in asthmatic patients taking controller medications. J Asthma Allergy 2014;7:131-139. [http://dx.doi.org/10.2147/JAA.S67062]

ORIGINAL RESEARCH

Prediction of peak expiratory flow rate in a Ugandan population S Nakubulwa,1 MSc; K Baisley,2 MSc; J Levin,3 PhD; J Nakiyingi-Miiro,1 PhD; A Kamali,1 PhD; A Nunn,4 MSc Medical Research Council/Uganda Research Unit on AIDS, Entebbe, Uganda London School of Hygiene and Tropical Medicine, London, UK 3 School of Public Health, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa 4 Medical Research Council Clinical Trials Unit, University College London, UK 1 2

Corresponding author: A Nunn (andrew.nunn@ucl.ac.uk)

Background. Peak expiratory flow rate (PEFR) measurement is one of the commonly used methods for assessing lung function in general practice consultations. The reference values for use by this method are mainly from Caucasian populations; data for African populations are limited. The existence of ethnic and racial differences in lung function necessitates further generation of PEFR reference values for use in African populations.

Objective. To generate equations for predicting PEFR in a Ugandan population.

Methods. The PEFR study was cross-sectional and based in rural south-western Uganda. Participants were aged 15 years or more, without respiratory symptoms and were residents of the study area. Multiple regression equations for predicting PEFR were fitted separately for males and females. The model used for PEFR prediction was: logePEFR = intercept + a(age, y) + b(logeage) + c(1/height in cm), where a, b and c are the regression coefficients.

Results. The eligible study population consisted of 774 males and 781 females. Median height was 164 cm (males) and 155 cm (females).

The majority of participants had never smoked (males 76.7%; females 98.3%). The equation which gave the best fit for males was logePEFR = 6.188 – 0.019age + 0.557logeage – 199.945/height and for females: logePEFR = 5.948 – 0.014 age + 0.317logeage – 85.147/height.

Conclusion. The curvilinear model obtained takes into consideration the changing trends of PEFR with increasing age from adolescence to old age. It provides PEFR prediction equations that can be applied in East African populations.

S Afr Respir J 2015;21(4):96-99. DOI:10.7196/SARJ.2015.v21i4.36

Pulmonary function tests play an important role in medicine by enabling the detection of airway obstruction, lung disease and its severity. [1] The methods for the testing include peak expiratory flow rate (PEFR), forced expiratory value in 1 second (FEV1) and forced vital capacity (FVC).[2] PEFR, the maximum flow of air during a forceful exhalation, is one of the most widely used tests for assessing lung function in epidemiological surveys and is commonly used in general practice consultations.[2,3] It is cheap, easy to use and does not require an electrical power supply. It is therefore convenient for use in resource-limited settings and in field studies.[3-5] Appropriate normal reference values for different populations are needed during the assessment of pulmonary function.[2] Studies have been conducted to develop prediction equations that can be used globally to obtain normal reference values; however, the data on which these equations are based are mostly from Caucasian populations, and data for African settings are limited.[1,6,7] These existing prediction equations may overestimate PEFR in black Africans by 12 - 15%.[3] Recognised ethnic and racial differences in lung function argue for the development of PEFR prediction equations for use in African populations.[8,9] Of the few studies on PEFR prediction equations that have been conducted in African populations, some have assumed a linear relationship between PEFR and age, not taking account of the increase in lung function during adolescence before a subsequent decline in later life. A study in Sudanese, South Sudanese and Tanzanian males showed a decline in PEFR from 17 to 70 years of age.[10] This finding differed from that of a study in Nigeria, where the decline in PEFR started at 30 years for males.[3] The available data would suggest that a linear fall in PEFR occurs from an age in the late 20s after maximal PEFR has been attained.[11,12] The objective of this study, therefore, was to obtain regression equations for predicting PEFR in an African population, using a curvilinear

model that considers changing trends of PEFR with age, given the fact that there are limited data on this topic in Africa.

Methods

Study design This was a cross-sectional study in which measurements of PEFR were obtained during a 1994/1995 survey of a general population cohort (GPC) in Uganda. Data were also collected on sociodemographic factors, general health and HIV status.

Study population In 1989/1990, the GPC was set up by the Medical Research Council/ Uganda Virus Research Institute (MRC/UVRI) Unit on AIDS to study the dynamics of HIV-1 infection in a rural African population.

Census, socioeconomic and medical surveys have been conducted annually until the present time. During the time of the PEFR survey, the cohort comprised all residents in 15 villages in the subcounty of Kyamulibwa, a rural area in south-western Uganda, about 40 km from Lake Victoria. The study population consisted of mainly subsistence farmers, whose staple diet was matooke (cooked bananas) with groundnuts. More details of the cohort have been published elsewhere.[13,14] In each annual medical survey, all adults aged ≥13 years are invited to participate; the survey includes determination of HIV-serostatus and other health indicators. During the 1994/1995 survey, participants had their PEFR function measured to construct a prediction model for an East African population and assess factors related to subnormal PEFR. PEFR measurements PEFR measurements were performed using mini Wright flow meters, which were calibrated at regular intervals. Participants were asked to

SARJ VOL. 21 NO. 4 2015

96

ORIGINAL RESEARCH make three attempts, and if readings were unsatisfactory, two further attempts were made. Measurements were recorded to the nearest 10 L/min. The maximum of these attempts was recorded as the observed PEFR. Measurements were made with participants in standing position.

Data management and statistical analysis Data were entered by two independent operators using dBase III+ software (Ashton-Tate, USA). Data were analysed with Stata version 11.0 (Stata Corp., USA).

Demographic characteristics of the 1994/95 survey participants were compared with those of the larger population that took part in the census, to assess whether the sample from which the PEFR data were collected was representative of the population.

Multiple regression equations were fitted separately for males and females for predicting PEFR as follows: logePEFR = intercept + a(age, y) + b(logeage) + c(1/height in cm), where a, b and c were the regression coefficients.

Table 1. Characteristics of the study population Variable

Males (N=774), Females (N=781), n (%) n (%)

Age group (years) 15 - 24

351 (45.4)

314 (40.2)

25 - 34

165 (21.3)

156 (20.0)

35 - 44

86 (11.1)

123 (15.8)

45 - 54

53 (6.9)

90 (11.5)

55 - 64

55 (7.1)

62 (7.9)

65+

64 (8.3)

36 (4.6)

Median (IQR)

26 (18 - 42)

29 (19 - 43)

164 (157 - 168)

155 (151 - 159)

Never

667 (87.4)

700 (90.7)

Sometimes

22 (2.9)

35 (4.5)

Always

74 (9.7)

37 (4.8)

Negative

736 (96.0)

736 (94.4)

Positive

31 (4.0)

44 (5.6)

Height (cm), median (IQR) Indoor cooking (N=1 535)

The model was adapted from a study in the UK by Nunn and Gregg,[12] which did not assume a linear relationship between age and PEFR but allowed for the fact that PEFR initially increases with age before reaching a maximum, after which it declines.

Observed PEFR in males and females was plotted separately against age. The regressions of PEFR on age using median height (calculated separately for males and females) were obtained from the equations generated in this study and were graphically presented.

Finally, PEFR values predicted by our model were plotted against age and compared with values from equations obtained from other studies in Africa, assuming a height of 164 cm for males and 155 cm for females (i.e. the median heights in our study).

Selection of healthy subjects Prediction models were restricted to healthy individuals, defined as those who had not had: (i) a cold, a sore throat or discharge from the ears for the previous 4 weeks; (ii) a cough lasting for 1 month or more; and (iii) wheezing in the chest. Smokers and HIV-positive individuals were included, provided they satisfied the symptom requirements mentioned above. The analysis was restricted to adults aged ≥15 years to be consistent with other studies in adults.[3] Ethical considerations The GPC has ethical approval (renewed annually) from the Uganda Virus Research Institute (UVRI) Science and Ethics Committee, ref. GC 127, and is also covered in the approval of ‘General MRC Research Programme’ by the National Council of Science and Technology (NCST) dated 15 October 1997, ref. MV 279/2 as ‘Population Dynamics of HIV-1 transmission’.

Results

Study population A total of 2 904 subjects were resident in the study area and consented to participate in the medical survey; of these, 2 505 were aged ≥15 years.

A total of 710 subjects were excluded from the PEFR prediction models on account of having respiratory symptoms at the time, 175 on account of failing to perform the PEFR test correctly and a further 65 had

HIV status (N=1 547)

Smoking (N=1 517) Never smoked

565 (76.7)

767 (98.3)

Stopped smoking

31 (4.2)

2 (0.3)

Smoke now

141 (19.1)

11 (1.4)

Currently married

330 (45.7)

359 (50.1)

Never married

341 (47.2)

215 (30.0)

Widowed

8 (1.1)

61 (8.5)

Divorced

43 (6.0)

82 (11.4)

Cement

90 (11.8)

125 (16.2)

Cow dung

6 (0.8)

4 (0.5)

Ordinary murram

664 (87.4)

641 (83.3)

Marital status (N=1 439)

Floor material (N=1 530)

97 SARJ VOL. 21 NO. 4 2015

Table 2. Predicted normal PEFR values for varying age and height by gender Height (cm) Age (years)

150

160

170

PEFR (L/minute)

Males 15

439.1

477.2

513.6

25

484.5

526.6

566.8

35

485.2

527.3

567.5

45

463.3

503.6

542.0

55

430.1

467.5

503.2

15

416.5

431.5

445.2

25

426.7

442.2

456.2

Females

35

413.8

428.7

442.4

45

390.6

404.7

417.5

55

362.8

375.9

387.8

ORIGINAL RESEARCH

The PEFR prediction equations The equation that gave the best fit of the data for males was: logePEFR = 6.188 ‒ 0.019age + 0.557logeage ‒ 199.945/height For females the equation was: logePEFR= 5.948 ‒ 0.014age + 0.317logeage ‒ 85.147/height Table 2 shows normal predicted PEFR values with varying age and height for both males and females, obtained using the PEFR prediction equations generated in this study.

For both males and females, predicted PEFR at 15 and 55 years was lower than predicted PEFR from 25 to 45 years.

Fig. 1 shows the observed values of PEFR in males and females plotted against age. The figure also shows the regression of PEFR on age for males of median height 164 cm and females of 155 cm. The maximum value of PEFR occurred at around 30 years for males and 25 years for females. For both males and females, there was a steady reduction in PEFR with increasing age after the maximum value of PEFR was attained. Of the variability in PEFR, 43% was explained by the prediction equation for males, and 28% by the equation for females.

Fig. 2 shows the predicted values of PEFR for males of height 164 cm and females of height 155 cm, compared with those obtained from three other studies in Africa.[3,9,10]

Discussion

In this study, the regression equations for predicting PEFR in a Ugandan population considering the changing trends of PEFR

800 700

PEFR (L/min)

600 500 400 300 200 20

40

Age (years)

Observed PEFR

60

80

100

Predicted PEFR (model in this study)

800 700 600 PEFR (L/min)

incomplete data on age or height. Therefore, there were 1 555 people (774 male, 781 female) in the study population, of which 665 (42.8%) were aged under 25 years and only 100 (6.4%) aged ≥65 years (Table 1). The median height was 164 cm (males) and 155 cm (females).

This was in agreement with the population distribution in the area. The majority of participants had never smoked (males 76.7%, females 98.3%). Most participants (85%) lived in houses with floor material made of ordinary murram (laterite), indicating low socioeconomic status.

When a comparison was made between medical survey participants and the GPC census participants, there was 50 - 60% participation in the medical survey for all age groups. Participation rates in the medical survey were slightly lower among males compared with females, but the differences were not significant.

500 400 300 200 20

40

Age (years)

Observed PEFR

60

80

100

Predicted PEFR (model in this study)

Fig. 1. Observed and predicted PEFR in males of median height of 164 cm (top) and females median height of 155 cm (bottom). with age were generated for both males and females. Our equations could be used to derive nomograms of normal reference ranges for East African populations.

The regression equations developed in our study are similar to those from a study in Nigerian adults, because both assumed a curvilinear prediction model accounting for increasing PEFR in adolescence followed by a decline for older age groups.[3] For both studies, maximum PEFR was reached at around 30 years for males and 25 years for females before declining in older age groups.[3] However, our study differs from an earlier study on Sudanese, South Sudanese and Tanzanian males,[10] which

failed to recognise the increase in PEFR in adolescence and early adulthood, and a study on Sudanese persons,[9] which also assumed a linear relationship for age 20 and above.

Our study found 43% of the variability in PEFR in males was explained by differences in age and height, in contrast to a study on Europeans where only 30% of the variability could be explained by age and height. However, for females, 28% of the PEFR variability was explained by age and height, which was similar to that reported among Europeans.[12] A strength of our study is the large number of individuals included in our survey, which is well in excess of what

SARJ VOL. 21 NO. 4 2015

98

ORIGINAL RESEARCH African populations. A study in a different East African population would be of interest since it would help to confirm our findings.

We will report in a different publication on the effects of smoking and other factors on PEFR.

550 500

PEFR (L/min)

450 400

Acknowledgements. This research was jointly funded by the UK Medical Research Council (MRC) and the UK Department for International Development (DFID) under the MRC/DFID Concordat agreement. We wish to acknowledge the study participants from the MRC General Population Cohort and staff of MRC/UVRI Uganda unit.

350 300 250 200 15

25

35

45 Age (years)

Current study (M) Njoku & Anah[3] (M)

55

65

75

Bashir & Musa[9] (M) Mustafa[10] (Only men in study)

550 500

PEFR (L/min)

450 400 350 300 250 200 15

25

35

45 Age (years)

55

65

75

Current study (F) Njoku & Anah[3] (F) Bashir & Mus[9] a (F)

Fig. 2. Predicted PEFR for males (top) and females (bottom) comparing different studies, based on the median height of 164 cm and 155 cm, respectively. has been obtained in other published studies. [4,7] However, there may be some limitations, such as under-representation within age groups due to absence or refusal to participate. It is also possible we may have missed individuals who were healthier, if they were more likely to be working away from home or travelling, and that individuals who were sicker were more likely to be found at home. Another possible limitation was the low representation of the older age groups when obtaining the

99 SARJ VOL. 21 NO. 4 2015

prediction equations. Our total sample size for older participants was small, with only 64 males and 36 females aged 65 years or more in the analysis; therefore, our prediction equations may not be reliable for the older age groups.

Conclusion

In conclusion, this study has provided PEFR prediction equations that could be used in obtaining normal PEFR reference values for assessing basic respiratory function in East

References

1. Quanjer PH, Stanojevic S, Cole TJ, et al. Multi ethnic reference values for spirometry for the 3 - 95-year age range: The global lung function 2012 equations.

Euro Resp J 2012;40(6):1324-1343. [http://dx.doi. org/10.1183/09031936.00080312] 2. Altalag ARJ, Wilcox P, eds. Pulmonary Function Tests in Clinical Practice. London, UK: Springer-Verlag Limited, 2009:1-3.

3. Njoku CH, Anah CO. Reference values for peak expiratory flow rate in adults of African descent. Trop Doct 2004;34(3):135-140.

4. Teklu B, Sexona T, Mills RJ. Peak expiratory flow in normal Ethiopian children and adults in Addis Ababa.

Br J Dis Chest 1987;81(2):176-181.

5. Bakki B, Hammangabdo A, Talle MA, Oluwole S, Yusuph H, Alkali MB. Peak expiratory flow in normal medical students in Maiduguri, Borno State, Nigeria.

Pan Afr Med J 2012;12:73.

6. Musafiri S, van Meerbeeck JP, Musango L, et al.

Spirometric reference values for an East African population. Respiration; international review of thoracic diseases. 2013;85(4):297-304. [http://dx.doi. org/10.1159/000337256] 7. Mengesha YA, Mekonnen Y. Spirometric lung function tests in normal non-smoking Ethiopian men and women. Thorax 1985;40(6):465-468.

8. Nku CO, Peters EJ, Eshiet Al, Bisong SA, Osim EE. Prediction formulae for lung function parameters in females of South Eastern Nigeria. Niger J Physiol Sci 2006;21(1-2):43-47.

9. Bashir AA, Musa OA. Reference spirometric values in a Sudanese cohort. East Mediterr Health J 2012;18(2):151-158.

10. Mustafa KY. Spirometric lung function tests in normal men and African ethnic origin. Am Rev Respir Dis. 1977;116(2):209-213.

11. Gregg I, Nunn AJ. Peak expiratory flow in normal subjects. BMJ 1973;3(5874):282-284.

12. Nunn AJ, Gregg I. New regression equations for predicting peak expiratory flow in adults. BMJ 1989;298(6680):1068-1070.

13. Kazooba P, Kasamba I, Baisley K, Mayanja BN, Maher D. Access to, and uptake of, antiretroviral therapy in a developing country with high HIV prevalence: A population-based cohort study in rural Uganda, 2004 - 2008. Trop Med Int Health 2012;17(8):e49-57.

14. Asiki G, Murphy G, Nakiyingi-Miiro J, et al. The general population cohort in rural south-western Uganda: A platform for communicable and non-communicable disease studies. Int J Epidemiol 2013;42(1):129-141. [http://dx.doi.org/10.1093/ije/dys234]

REVIEW

Screening for lung cancer: A review B E Schär, MB ChB, FCP SA, Cert Pulm SA Phys Pulmonologist, Mediclinic Midstream, Centurion, Gauteng Corresponding author: B E Schär (bronwyn.schar@mediclinicnetwork.com)

Lung cancer (LC) is the leading cause of cancer-related death worldwide. Its overall poor prognosis is attributable to the fact that most patients remain asymptomatic until the disease is advanced and, therefore, present with late-stage incurable disease. The rationale for LC screening is that early detection of asymptomatic disease offers the opportunity for earlier intervention, at a stage when definitive cure is still feasible, which has the potential to reduce LC-related mortality and morbidity. The findings of the National Lung Screening Trial provided the first strong evidence in support of this rationale. Since its publication, several professional organisations and societies have developed guidelines recommending the implementation of LC screening with low-dose computed tomography in asymptomatic, high-risk individuals. Although the benefits of such screening programmes may be significant, they must be carefully weighed against the potential harms to the relatively large number of healthy individuals who would undergo screening. This review examines the available evidence and current recommendations for LC screening, including benefits, potentials harms and requirements for implementation of a high-quality, safe and effective programme. In addition, the costs and availability of LC screening programmes in both the global and local settings are considered.

S Afr Respir J 2015;21(4):101-107. DOI:10.7196/SARJ.2015.v21i4.37

Lung cancer (LC) is a common malignancy among men and women; an estimated 1.8 million new cases were reported worldwide in the year 2012, 58% of which occurred in the developing world.[1] It is the leading cause of cancer-related deaths, and although its incidence is similar to other common malignancies (including breast, prostate and colorectal cancer), it causes four times as many deaths.[2,3] In 2012, LC-associated deaths accounted for 19.4% of global cancer mortality.[1] Similarly, LC mortality has been reported to account for 17% of cancer-related deaths in South Africa (SA).[4] The 5-year survival rate for LC remains low (16.8%) despite advances in therapy.[5] The stage of non-small-cell LC (NSCLC) at the time of diagnosis determines prognosis. Since stage I NSCLC can be treated, and potentially cured with surgical resection, the 5-year survival rate is significantly better than for stage IV disease (60% v. <5%).[5] Furthermore, smaller tumours within stage I NSCLC correlate with improved clinical outcomes.[6,7] Although data for small-cell LC (SCLC) are limited, they also support improved prognosis in individuals diagnosed with early-stage disease.[8] The overall poor prognosis of LC is largely attributable to the fact that 75% of patients present with late-stage, inoperable disease.[9,10] The reported operability rate in SA literature is 10 - 11%, indicating that the proportion of patients who present with curable disease is even lower than that described internationally.[4,11] The rationale for LC screening is that early detection and treatment of asymptomatic LCs has the potential to reduce LC-related mortality and morbidity by increasing the overall cure rate, allowing more limited surgical resection in order to achieve cure and reducing exposure to adjuvant therapies.[8,12] The utility of sputum cytology and chest radiography (CXR) alone, or in combination, as LC screening tools has been extensively studied and no benefit has been demonstrated for either modality.[13,14] Previously, several single-arm studies suggested that screening with low-dose computed tomography (LDCT) may be beneficial,[15] but the 2010 publication of the National Lung Screening Trial (NLST) provided the first strong supporting

101 SARJ VOL. 21 NO. 4 2015

evidence for LC screening with LDCT.[16] This led various medical professional societies, clinical networks and the US Preventative Services Task Force (USPSTF) to make recommendations in support of LC screening with LDCT in high-risk individuals.[12,17-19] Subsequently, additional studies were done that have better defined which individuals are at highest risk and are most likely to benefit from LC screening.[20-22] Several characteristics of LC, together with available evidence, suggest that implementation of LC screening may be both feasible and beneficial.[12,17,23,24] This review examines current recommendations for LC screening, including benefits, potential harm and consideration of costs and availability of screening programmes.

Screening with LDCT

Since 2000, a number of cohort and randomised controlled trials (RCTs) have been performed to evaluate the utility of LDCT compared with CXR or usual care. Table 1 provides a list of select RCTs and highlights the differences in their respective inclusion criteria, screening intervals and definitions for positive screen results.[25-31] Since the ultimate objective of screening is to identify and treat early-stage LC, the participants’ fitness for lung surgery is an important factor to consider when assessing eligibility for screening.[12] Accordingly, it is noteworthy that all the trials listed excluded those individuals with comorbidities that precluded them from curative LC surgery.

In addition, some required participants to demonstrate a specified baseline effort tolerance before undergoing screening.[27] The NLST is the largest RCT of LC screening to date. The study included 53 454 men and women at 33 tertiary centres across the USA. Individuals were considered high risk and eligible for screening if they were between the ages of 55 and 74 years, had a smoking history of at least 30 pack-years, and were either current cigarette smokers or had quit smoking within the past 15 years. Participants were randomised to LC screening with either LDCT or CXR annually for 3 years, with a median duration of follow-up of 6.5 years. Any noncalcified nodule measuring ≥4 mm in diameter identified on LDCT

REVIEW or CXR was classified as a positive screen. Of the studies performed, 24.2% of LDCTs and 6.9% of radiographs were positive, with 39% of individuals in the LDCT group and 16% of those in the radiograph group having at least one positive screen during the 3-year period. Of these abnormalities 96.4% were false-positive findings (i.e. they did not lead to a diagnosis of cancer). Although most positive findings were resolved by further imaging, 11% were followed by an invasive diagnostic procedure. The rate of procedure-related complications was low (1.4% of positive screenees in the LDCT group and 1.6% of those in the CXR group experienced a complication). The trial was stopped prematurely as an interim analysis demonstrated a statistically significant reduction of 20% in LC-specific mortality and 6.7% in allcause mortality in the LDCT arm. The number needed to screen with LDCT to avoid one LC death was 320.[16] The ongoing Nederlands-Leuven Longkanker Screenings Network (NELSON) trial, which compares LDCT screening (at 1, 2 and 2.5 years) with no screening, is the second largest of the RCTs and is powered to detect a 25% decrease in LC mortality after 10 years.[30] The data from this trial will be pooled with those from the Danish LC Screening Trial before being published.[27] Unlike many of the other RCTs, the NELSON trial uses volumetric measurements to assess screen-detected nodules. This approach appears to be more specific than measurement of diameter and has a significantly lower positive screen rate (2.6% and 1.8%, respectively, in screening rounds 1 and 2), compared with the 26.4% across all rounds of screening in the NLST.[16,31] Although not yet included in any published practice guidelines, the results from the NELSON trial suggest that the use of a volumetric strategy offers the potential advantage of decreasing the number of follow-up examinations needed for participants with a positive screen result.[31]

Recommendations for screening with LDCT

Pursuant to the findings of the NLST, several professional societies have published LC screening guidelines (Table 2). Most have defined eligibility for screening based on the NLST inclusion criteria and have advocated a screening process that closely mirrors that followed in the NLST. Based on the systematic review of the benefits and harmful effects of LDCT screening, published by Bach et al.,[18] the American College of Chest Physicians published clinical practice guidelines that recommend LC screening only for individuals who meet NLST inclusion criteria and have access to treatment in a multidisciplinary centre capable of delivering comprehensive cancer care commensurate with that provided in the NLST.[17] The American Cancer Society (ACS) guidelines support screening in a similarly defined high-risk population, with the proviso that screened individuals should be in good health.[32] The USPSTF recommendation statement, published in 2014, supported LDCT screening for adults aged 55 - 80 years who have a ≥30 pack-year cigarette smoking history, and either smoke currently or quit smoking within the past 15 years.[11] The decision to extend the age range beyond that of the NLST (ages 55 - 74 years) was based on the outcomes of a statistical comparative modelling study conducted by de Koning et al.,[22] which suggested that continuation of annual screening until the age of 80 years would be advantageous. The caveat stated by the USPSTF is that screening should be discontinued either once an individual develops a health problem that substantially limits

their life expectancy or fitness to undergo curative lung surgery; or once an individual has not smoked for ≥15 years.[12] Both the American Association of Thoracic Surgeons (AATS) and the National Comprehensive Cancer Network (NCCN) have recommended screening for individuals who do not meet NSLT inclusion criteria (younger age and/or lesser smoking exposure), but have an additional risk factor (e.g. asbestos exposure, family history, chronic obstructive pulmonary disease).[19,33] Further studies are needed to better delineate the contribution of these and other risk factors in the development of LC. Potential risk factors that may be particularly relevant in SA and warrant further investigation include: asbestos exposure, silica exposure, HIV infection, and exposure to biomass fuel emissions. Since epidemiological data indicate that HIV-infected individuals have an elevated LC risk independent of smoking, evaluation of the effect of HIV infection on LC development would be informative.[34,35] Similarly, since ~20% of SA households are exposed to smoke from the burning of solid and biomass fuels, it would be valuable to examine the potential role of biomass fuel emissions in the development of LC in the local population.[36]

Potential harmful effects of screening with LDCT

An ideal LC screening tool provides maximal benefit to the small number of people in whom LC is detected early and treated with surgical resection, without causing harm to the comparatively large number of healthy individuals who undergo screening. While no perfect screening test exists, a good understanding of the potential harmful effects of screening, and discussion of the same, are pivotal in minimising risk and ensuring appropriate and effective implementation of screening programmes.[25] False-positive screens, psychological stress related to fear of having cancer, radiation exposure, incidental findings and overdiagnosis are all potential harmful effects that must be considered when contemplating LC screening.[37] In view of the fact that LDCT scanning does not provide an immediate diagnosis for positive screens, the ability to follow up and appropriately manage pulmonary nodules should be an integral part of any LC screening programme.[24] The RCTs listed in Table 1 all had clearly defined protocols for the identification, reporting and management of pulmonary nodules. There are several society guidelines available to assist in planning nodule management algorithms. Some of these are listed in Table 3.[19,24,38-40] Detection of false-positive results that require further evaluation occurs in a significant proportion of screened individuals. For example, in the NLST 24.2% of the LDCT group and 6.9% of the CXR group had positive screens. However, 96% of these abnormalities were false-positive findings that did not lead to a diagnosis of cancer.

Most positive findings were resolved by serial imaging, but 11% led to performance of an invasive procedure (1.8% biopsy, 3.8% bronchoscopy, 4% surgical procedure). Although the rate of procedure-related complications was low, it is important to remember that the NLST was conducted in specialist centres, by highly skilled multidisciplinary teams.[16] There are valid concerns as to whether these procedure and complication rates can be reproduced at a community level. In fact, review of the results of the DANTE Trial, which was performed in a community setting,

SARJ VOL. 21 NO. 4 2015

102

REVIEW psychological discomfort, and does not affect health-related quality of life or long-term anxiety levels.[42] In the NLST, 8% of LDCT scans identified clinically significant abnormalities that were unrelated to LC. Such incidental findings may include abnormalities in the lung (e.g. pulmonary fibrosis or emphysema), thyroid and heart; the most common of which is coronary calcification.[43] Currently, the effect of these incidental findings is not well defined, but the potential for their existence should be discussed as part of the decision-making process prior to initiating screening.

suggests that both the rate of invasive procedures and the rate of procedure-related complications were higher than in the NLST.[26] Concerns have been raised regarding the anxiety and psychological distress that may follow a false-positive result. Studies that have evaluated patients’ responses to the presence of pulmonary nodules demonstrate that most individuals experience at least mild distress, which may be influenced by the approach of their clinician.[41] However, a systematic review that evaluated the psychosocial effects of LC screening demonstrated that it is associated with only short-term Table 1. RCTs evaluating LC screening with LDCT[25]

Trial (lead author) NLST

DANTE

DLCST

LUSI[28]

ITALUNG[29]

MILD[30]

NELSON[31]

Location

USA

Italy

Denmark

Germany

Italy

Italy

Netherlands and Belgium

Sample size

53 454

2 811

4 104

4 052

3 206

4 099

15 822

Sex

M, F

M

M, F

M, F

M, F

M, F

M, F

Age (y)

55 - 74

60 - 74

50 - 70

50 - 69

55 - 69

>49

50 - 75

Smoking history

>30 packyears, current or former smokers who have quit within the past 15 y

>20 packyears, current or former smokers

>20 pack-years, current or former smokers who have quit at age >50 y within the past 10 y

Current or former smokers who quit within the past 10 y; exposure >15 cig/d × 25 y or >10 cig/d × 30 y

>20 packyears, current or former smokers

>20 packyears, current or former smokers who have quit within the past 10 y

Current or former smokers who quit within the past 10 y; exposure >15 cig/d × 25 y or >10 cig/d x 30 y

Screening interval

Annual

Annual

Annual

Annual

Annual

Randomised: annual or biennial

Annual, biennial or every 30 months

Screening rounds, n

3

5

5

5

5

5

4

Any diameter

Not specified

Largest diameter Largest diameter Largest diameter >5 mm >5 mm >5 mm

[16]

Criteria for screenpositive PN

[26]

>4 mm

[27]

Volume

Volume >60 mm

3

>50 mm3

DLCST = Danish LC Screening Trial; LUSI = LC Screening Intervention Trial; MILD = Multicentric Italian Lung Detection Project; PN = pulmonary nodule; cig/d = cigarettes per day; M = male; F = female.

Table 2. Professional Societies’ Guidelines for LC Screening[25] Society (Reference)

Recommendation

Grade

NCCN

Annual LDCT in high-risk individuals

B

[19]

Group 1: age 55 - 79 y with >30 pack-years, current or former smokers who quit within the past 15 y Group 2: age >50 y with >20 pack-years and 1 additional risk factor* USPSTF[12]

Annual LDCT in high-risk individuals: age 55 - 80 y with >30 pack-years, current or former smokers who B have quit within the past 15 y†

ACCP/ASCO[17]

Annual LDCT in high-risk individuals: age 55 - 74 y with >30 pack-years, current or former smokers 2B who quit within the past 15 y, but only in settings that can provide multidisciplinary care similar to that provided in the NLST

ACS[32]

Annual LDCT in high-risk individuals: age 55 - 74 y with >30 pack-years, current or former smokers who B quit within the past 15 y, who are in good health

AATS[33]

Annual LDCT in high-risk individuals: age 55 - 79 y with >30 pack-years); individuals aged 50 - 79 y with B 20 pack-years and added risk of >5% of developing LC in 5 y† LC survivors in remission >5 y

ACCP/ASCO = American College of Chest Physicians/American Society of Clinical Oncologists. * Additional risk factors include occupational exposure, COPD, idiopathic pulmonary fibrosis and personal/family history of LC. † Screening may be discontinued if life expectancy is limited or >15 y has elapsed since quitting smoking.

103 SARJ VOL. 21 NO. 4 2015

C

REVIEW the Italian Lung Study (ITALUNG), which estimated 6 - 7 mSv for a baseline LDCT and three subsequent annual LDCTs.[47] According to the current LC screening recommendations (Table 2), a high-risk patient entering a screening programme at the age of 55 years may undergo 19 annual LDCT scans, with the potential for additional imaging (in the event of a positive screen that requires follow-up), so it is likely that the cumulative radiation risk to the patient may be significantly higher than initial estimates. As always, this potential harmful effect must be tempered with the significant benefits that are to be gained from screening. Based on analysis of available data, one study has suggested that any reduction in excess of 5% in overall LC mortality would outweigh the radiation risks.[48] This is an area that will require ongoing data collection and regular audit as screening programmes are implemented.

Overdiagnosis refers to the detection of a cancer that would not otherwise have become clinically relevant during the patient’s lifetime.

The extended follow-up data for 16 years from the Mayo Lung Project, which demonstrate a persistent excess of cancers in the screened group compared with the control, suggest that overdiagnosis does occur.[44] However, it is uncertain to what extent it occurs in LC screening as reported rates vary widely (5 - 51%).[45] Although it is a potentially harmful effect of screening, the current estimates available from LDCT trials suggest that the rate of overdiagnosis is relatively low (reported rate in the NLST was 18.5%, or 1.38 cases of overdiagnosis per 320 individuals needed to screen to prevent one LC death).[16] With this in mind, the benefits of screening likely outweigh the risk of overdiagnosis for appropriately selected patients.

Radiation exposure is an inevitable consequence of screening with LDCT. The effective radiation dose from a single LDCT is between 0.61 and 1.5 mSv. When compared with other radiological studies, such as a CT pulmonary angiogram (15 mSv) and a routine CT chest (8 mSv) this dose is low.[18,46] To date, the only study that has reported exposure related to initial screening and follow-up evaluations is

Implementation of LC screening

Pursuant to the publication of the USPSTF recommendation statement and the various society guidelines (Table 2), several institutions in Europe and the USA are offering screening to

Table 3. Society guidelines for the management of pulmonary nodules[24] Recommended follow-up (months)

Nodule morphology

Size (mm)

Fleischner Society

NCCN[19]

Lung-RADS[40]

Solid

<6

6 - 12, 18 - 24*

Annual screening

Annual screening

6 - 7.9

3 - 6, 9 - 12, 24*

3, 6, annual screening

6, annual screening

8 - 10

3 - 6, 9 - 12, 24*

PET scan and/or biopsy or resect

3, annual screening

≤5

None

Annual screening

Annual screening

>5

3, 12, 24, 36

6, annual screening

Annual screening up to 20 mm

≤5

3, then yearly × 3

Annual screening

Annual screening

>5

3, then biopsy or resect

As for solid

Based on size of solid component

Pure GGN Part-solid

[38,39]

RADS = reporting and data system; PET = positron emission tomography; GGN = ground-glass nodule. * Fleischner Society recommendation for the high-risk patient.

Table 4. Recommended components of a high-quality LDCT LC screening programme[17] Organisation/Society ACCP[24]

NCCN[19]

USPSTF[12]

Multisociety* guideline[18]

ACS[32]

IASLC[50]

Recommended components

Careful participant selection

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Defined screening interval and duration

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Technical specifications and quality controls for performance of LDCT scan

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Positive screen criteria for PNs

Yes

Yes

Yes

Yes

Yes

Yes

-

Structured LDCT reporting

Yes

Yes

Yes

Yes

Yes

Yes

-

PN management algorithm with a multidisciplinary team

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Smoking cessation

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Patient and provider education

Yes

Yes

Yes

Yes

Yes

-

-

Data collection

Yes

Yes

Yes

Yes

-

Yes

Yes

AATS[33]

IASLC = International Association for the Study of LC. * ACCP, ASCO, AATS.

SARJ VOL. 21 NO. 4 2015

104

REVIEW high-risk individuals. However, there remain concerns about the generalisability of the NLST findings to scenarios outside the tertiary-care setting. When compared with the high-risk US population that is currently being offered screening, the population in the NLST differed in several respects. The participants were younger, had a higher level of education and were more likely to be former smokers. [49] In addition, of the 33 centres involved in the study the majority were tertiary-care academic facilities that were designated National Cancer Institute (NCI) cancer centres.

The level of multidisciplinary expertise available at the trial centres likely accounts for the low rate of invasive procedures undertaken to follow-up positive screen results, as well as for the low mortality rate for surgical resection (1% as compared with 4% previously reported in the US general population).[16] Guidelines and policy statements from various international professional societies have recognised that a multidisciplinary team, of a similar calibre to those that participated in the NLST, is a necessity if the outcomes of the NLST are to be replicated in clinical practice.

Furthermore, they have highlighted the point that LC screening, if not conducted properly, is potentially harmful.[16,18] It is for this reason that there has been much focus on identifying the components that are necessary in order to implement a high-quality, safe and effective screening programme. Table 4 lists the components recommended by the different professional societies to guide LC screening in the community.[12,17-19,24,32,33,50]

Risk assessment and communication of risk

Several accurate and efficient risk stratification models have been developed to assist in determining an individualâ&#x20AC;&#x2122;s risk for LC (Table 5).[20,51-54] Although these prediction models were designed with a research setting in mind, none of the RCTs listed in Table 1 made use of these models to define their entry criteria. Despite the fact that these risk-prediction models were developed primarily as research tools, they do have utility in clinical practice. They can be used as information and discussion tools for patient consultations and they may assist patients in understanding their own individual risk for LC. This is a vital component of the screening process as patient understanding and participation in making decisions regarding participation (or not) are vital to the successful implementation of any screening programme.[34]

Importance of smoking cessation

Although there are a number of environmental risk factors that are associated with LC development, there is strong evidence that cigarette smoking, which is considered to be causal in ~85 - 90% of cases, is the primary risk factor.[55] Even among non-smokers, a proportion of LC cases are thought to be attributable to second-hand tobacco smoke exposure.[56] The risk of developing LC rises with increasing cumulative tobacco smoke exposure.[12] However, some studies suggest that, for a given level of smoking exposure, women are at higher risk of developing cancer than men.[57,58] Since the risk for LC only declines many years after smoking cessation, it is also important to recognise that a significant percentage of LCs occur in former smokers.[59-61] Since cigarette smoking plays a central role in LC development, promotion of smoking cessation and reduction in population smoking rates are imperative for reducing the long-term burden of disease and cannot be replaced by LC screening. The decline in LC incidence and mortality among men in the US between 1975 and 2010, which was observed in tandem with a reduction in the smoking rate among adults in the US between 1965 and 2011, lends support to smoking cessation as an effective means of lowering LC mortality.[1,62] Available evidence suggests that LDCT screening itself does not influence smoking behaviour and that despite enrolment in LC screening trials and increased awareness, some participants continue to smoke.[63] Those individuals who had a positive screen test for LC have been shown to have a 6% lower rate of smoking compared with those who had normal screens.[64] Not only is smoking cessation and LC prevention more effective than screening in lowering LC mortality, but it is also a much more cost-effective strategy.[24,65]

Cost of LDCT screening

Following the USPSTFâ&#x20AC;&#x2122;s publication of a grade B recommendation in favour of implementation of LDCT screening for high-risk individuals in 2014,[12] and after considering the guidelines and position papers of several major medical societies (Table 2), US healthcare funders agreed to fund LC screening with LDCT, at approved LC screening centres, for individuals who meet NLST eligibility.

In light of the large number of patients who may qualify for annual screening, based on the specified criteria, it is reasonable to expect that the costs may be considerable. Therefore, one of the major

Table 5. LC Risk prediction models Â

Model Bach

[51]

Spitz

[52]

Liverpool Lung Project[53]

Hoggart[54]

Modified PLCOM2012[20]

Age (y)

50 - 75

20 - 80

20 - 80

35 - 65

55 - 74

Variables

Age

Age

Age

Age

Age

Smoking

Smoking

Smoking

Smoking

Smoking

Sex

Sex

Sex

BMI

Asbestos

Dust

Asbestos

COPD

Emphysema

Family history

CXR

Family history

Pneumonia

Family history

Prior cancer

Education

BMI = body mass index; COPD = chronic obstructive pulmonary disease.

105 SARJ VOL. 21 NO. 4 2015

REVIEW considerations for policymakers in the US has been the cost and cost-effectiveness of LDCT screening.[66] In order to address some of the questions around cost-efficacy, Black et al.[67] examined the costs of LDCT screening in the NLST, estimating mean life-years, qualityadjusted life-years (QALYs), costs per person and incremental costeffectiveness ratios (ICERs) for three categories: LDCT screening, screening with CXR and no screening. According to their analysis, the cost of CT screening per QALY gained was USD81 000. However, they also noted that the ICERs varied widely in subgroup and sensitivity analyses, suggesting that modest changes in certain study assumptions could greatly alter the estimated figure. Ultimately, they concluded that cost efficacy of LC screening outside the NLST setting will depend on how the programmes are implemented at the community level.

Local data regarding the costs involved in LC screening are lacking for both the public and private sector. Currently, many private radiology practices do not have a specific fee structure in place for LDCT screening. However, those that do, report a cost of ZAR2 600 - 3 800 per screening test, inclusive of consultation with a physician.

As this fee is not covered by most SA medical aid funds at present, this cost is borne by the patient (direct correspondence, private hospitals in Johannesburg, October 2015). Further research and data collection in this area are required before a dialogue between the relevant roleplayers can be held. Unfortunately, given the existing strains on an already over-burdened public healthcare system, and resistance on the part of private healthcare funders to pay for screening in the private sector, LC screening will most likely only be accessible to a small minority of SA citizens for the foreseeable future.

Conclusion

LC is a deadly malignancy. Until recently, there was no proven method for early detection, but with the advent of LDCT screening, there is the potential to save lives in high-risk individuals by detecting and treating early-stage disease. However, the false-positive rate for LC screening is high and the potential benefits may be diminished if screening results in increased morbidity and mortality related to increased frequency of invasive procedures. Refinement of eligibility criteria and the use of structured, standardised reporting for CT interpretation may mitigate risk by reducing the number of false-positives identified on screening. A well-defined protocol for management of screen-detected pulmonary nodules is a vital component of any LC screening programme and may aid in minimising unnecessary surgery.

Successful implementation of LC screening relies on a comprehensive, multidisciplinary programme with an emphasis on smoking cessation and LC prevention. Balanced presentation of potential benefits, harms and costs of screening, such that individuals are able to make informed decisions about their healthcare choices, is of the utmost importance and risk-prediction models may be of use in facilitating these discussions.

In SA we face unique challenges in implementing LC screening.

Identification of risk factors that are particularly relevant to our local population (e.g. HIV infection, silica exposure and biomass emissions exposure) is an area where future research is required. Limited resources in the state sector and restricted funding by medical aids in the private sector both pose significant barriers to the equitable implementation of LC screening in our communities. Collation of

comprehensive local LC data and engaging the relevant role-players in discussion will be important initial steps towards planning a LC screening programme in the local setting.

References

1. Ferlay J, Soerjomataram I, Ervik M, et al. GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No.11. Lyon, France: International Agency for Research on Cancer, 2013. http://globocan.iarc.fr. (accessed 2 September 2015).

2. Global Burden of Disease Cancer Collaboration, Fitzmaurice C, Dicker D, et al. The Global Burden of Cancer 2013. JAMA Oncol 2015;1(4):505-527. [http://dx.doi.org/ 10.1001/jamaoncol.2015.0735] 3. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin 2012;62(1):10-29. [http://dx.doi.org/10.3322/caac.20138] 4. Wilcox PA, Oâ&#x20AC;&#x2122;Brien JA, Abratt RP. Lung cancer at Groote Schuur Hospital: A local perspective. S Afr Med J 1990;78(12):716-720.

5. Howlader N, Noone AM, Krapcho M, et al. (eds). SEER Cancer Statistics Review, 1975 - 2011 [based on the November 2013 SEER data submission, posted to the SEER website April 2014]. Bethesda, USA: National Cancer Institute, 2014.

6. Mountain CF. Revisions in the International System for Staging Lung Cancer. Chest 1997;111(6):1710-1717.

7. Flieder DB, Port JL, Korst RJ, et al. Tumor size is a determinant of stage distribution in T1 non-small cell lung cancer. Chest 2005;128(4):2304-2308. [http://dx.doi. org/10.1378/chest.128.4.2304] 8. Deffebach ME, Humphrey L. Screening for lung cancer. http://uptodate.com (accessed 5 August 2015).

9. Siegal R, Ward E, Brawley O, Jemal A. Cancer statistics, 2011: The impact of eliminating socio-economic and racial disparities on premature cancer deaths. CA Cancer J Clin 2011;61(4):212-236. [http://dx.doi.org/10.3322/caac.20121] 10. Jett JR. Current treatment of unresectable lung cancer. Mayo Clinic Proc 1993;68(6):603-611.

11. Mukansi M, Smith C, Feldman C. A study of lung cancer in Johannesburg, South Africa. South Afr J Infect Dis 2013;29(1):43-47. [http://dx.doi.org/10.1080/231200 53.2014.11441566] 12. Moyer VA. Screening for lung cancer: US Preventative Services Task Force recommendation statement. Ann Intern Med 2014;160(5):330-338. [http://dx.doi. org/10.7326/M13-2771] 13. Oken MM, Hocking WG, Kvale PA, et al. Screening by chest radiograph and lung cancer mortality: The Prostate, Lung, Colorectal, and Ovarian (PLCO) randomised trial. JAMA 2011;306(17):1865-1873. [http://dx.doi.org/10.1001/jama.2011.1591] 14. Fontana RS, Sanderson DR, Taylor WF, et al. Early lung cancer detection: Results of the initial (prevalence) radiologic and cytologic screening in the Mayo Clinic study.

Am Rev Resp Dis 1984;130(4):561-565.

15. Henschke CI, McCauley DI, Yankelevitz DF, et al. Early Lung Cancer Action Project: Overall design and findings from baseline screening. Lancet 1999;354(9173):99-105. [http://dx.doi.org/10.1016/S0140-6736(99)06093-6] 16. Aberle DR, Adams AM, Berg CD, et al. Reduced lung cancer mortality with low-dose computed tomographic screening. N Engl J Med 2011;365(5):395-409. [http://dx.doi. org/10.1056/NEJMoa1102873] 17. Detterbeck FC, Mazzone PJ, Naidich DP, Bach PB. Screening for lung cancer.

Diagnosis and management of lung cancer, 3rd ed. American College of Chest Physicians evidence-based clinical practice guidelines. Chest 2013;143(5 Suppl):e78Se92S. [http://dx.doi.org/10.1378/chest.12-2350] 18. Bach PB, Mirkin JN, Oliver TK, et al. Benefits and harms of CT screening for lung cancer: A systematic review. JAMA 2012;307(22):2418-2429. [http://dx.doi. org/10.1001/jama.2012.5521] 19. National Comprehensive Cancer Network. Lung Cancer Screening. http://www.nccn. org/professionals/physician_gls/f_guidelines.asp (accessed 12 August 2015).

20. Tammemägi MC, Katki HA, Hocking WG, et al. Selection criteria for lungcancer screening. N Engl J Med 2013;368(8):728-736. [http://dx.doi.org/10.1056/ NEJMoa1211776] 21. Kovalchik SA, Tammemagi MC, Berg CD, et al. Targeting of low-dose CT screening according to the risk of lung-cancer death. N Engl J Med 2014;369(3):245-254. [http:// dx.doi.org/10.1056/NEJMoa1301851] 22. de Koning HJ, Meza R, Plevritis SK, et al. Benefits and harms of computed tomography lung cancer screening strategies: A comparative modeling study for the US Preventative Task Force. Ann Intern Med 2014;160(5):311-320. [http://dx.doi. org/10.7326/M13-2316] 23. Peto R, Darby S, Deo H, Silcocks P, Whiteley E, Doll R. Smoking, smoking cessation, and lung cancer in the UK since 1950: Combination of national statistics with two case-control studies. BMJ 2000;321(7527):323-329. [http://dx.doi.org/10.1136/ bmj.321.7257.323]

SARJ VOL. 21 NO. 4 2015

106

REVIEW 24. Mazzone P, Powell CA, Arenberg D, et al. Components necessary for high-quality lung cancer screening. Chest 2015;147(2):295-303. [http://dx.doi.org/10.1378/ chest.14-2500] 25. Kandora NM, Silvestri GA, Tanner NT. Screening and early detection efforts in lung cancer. Cancer 2015;121(9):1347-1356. [http://dx.doi.org/10.1002/cncr.29222] 26. Infante M, Cavuto S, Lutman FR, et al. Long-term follow-up results of the DANTE trial, a randomised study of lung cancer screening with spiral computed tomography.

Am J Respir Crit Care Med 2015;191(10):1166-1175. [http://dx.doi.org/10.1164/ rccm.201408-1475OC] 27. Saghir Z, Dirksen A, Ashraf H, et al. CT screening for lung cancer brings forward early disease. The randomised Danish Lung Cancer Screening Trial: Status after 5 annual screening rounds with low-dose CT. Thorax 2012;67(4):296-301. [http:// dx.doi.org/10.1136/thoraxjnl-2011-200736] 28. Becker N, Motsch E, Gross ML, et al. Randomized study on early detection of lung cancer with MSCT in Germany: Study design and results of the first screening round.

J Cancer Res Clin Oncol 2012;138(9):1475-1486. [http://dx.doi.org/10.1007/s00432012-1228-9] 29. Lopes Pegna A, Picozzi G, Falaschi F, et al. ITALUNG Study Research Group. Fouryear results of low-dose CT screening and nodule management in the ITALUNG trial. J Thorac Oncol 2013;8(7):866-875. [http://dx.doi.org/10.1097/JTO.0b013e31828f68d6] 30. Pastorino U, Rossi M, Rosato V, et al. Annual or biennial CT screening versus observation in heavy smokers: 5-year results of the MILD trial. Eur J Cancer Prev 2012;21(3):308-315. [http://dx.doi.org/10.1097/CEJ.0b013e328351e1b6] 31. Horeweg N, van der Aalst CM, Vliegenthart R, et al. Volumetric computed tomography screening for lung cancer: 3 rounds of the NELSON trial. Eur Resp J 2013;42(6):16591677. [http://dx.doi.org/10.1183/09031936.00197712] 32. Wender R, Fontham ET, Barrera E Jr, et al. American Cancer Society lung cancer screening guidelines. CA Cancer J Clin 2013;63(2):107-117. [http://dx.doi. org/10.3322/caac.21172] 33. Jaklitsch MT, Jacobson FL, Austin JHM, et al. The American Association for Thoracic Surgery guidelines for lung cancer screening using low-dose computed tomography scans for cancer survivors and other high risk groups. J Thorax Cardiovasc Surg 2012;144(1):33-38. [http://dx.doi.org/10.1016/j.jtcvs.2012.05.060] 34. Mani H, Halgentz Jr M, Aboulafia DM. Lung cancer in HIV infection. Clin Lung Cancer 2012;13(1):6-13. [http://dx.doi.org/10.1016/j.cllc.2011.05.005] 35. Winstone TA, Man SF, Hull M, et al. Epidemic of lung cancer in patients with HIV. Chest 2013;143(2):305-314. [http://dx.doi.org/10.1378/chest.12-1699] 36. Norman R, Barnes B, Mathee AM, Bradshaw D, and the South African Comparative Risk Assessment Collaborating Group. Estimating the burden of disease attributable to indoor air pollution from household use of solid fuels in South Africa in 2000.

S Afr Med J 2007;97(8):764-771.

37. Humphrey L, Deffebach M, Pappas M, et al. Screening for lung cancer with low-dose computed tomography: A systematic review to update the US Preventative Services Task Force recommendation. Ann Intern Med 2013;159(6):411-420. [http://dx.doi. org/10.7326/0003-4819-159-6-201309170-00690] 38. MacMahon H, Austin JHM, Gamsu G, et al. Fleischner Society. Guidelines for management of small pulmonary nodules detected on CT scans: A statement from the Fleischner Society. Radiology 2005;237(2):395-400. [http://dx.doi.org/10.1148/ radiol.2372041887] 39. Naidich DP, Bankier AA, MacMahon H, et al. Recommendations for the management of subsolid pulmonary nodules detected at CT: A statement from the Fleischner Society. Radiology 2013;266(1):304-317. [http://dx.doi.org/10.1148/ radiol.12120628] 40. American College of Radiology. Lung CT Screening Reporting and Data System (Lung-RADS). http://www.acr.org/~/media/ACR/Documents/PDF/QualitySafety/ Resources/LungRADS/AssessmentCategories.pdf (accessed 9 September 2015).

41. Slatore CG, Press N, Au DH, Curtis JR, Wiener RS, Ganzini L. What the heck is a “nodule”? A qualitative study of veterans with pulmonary nodules. Ann Am Thorac Soc 2013;10(4):330-335. [http://dx.doi.org/10.1513/AnnalsATS.201304-080OC] 42. Slatore CG, Sullivan DR, Pappas M, Humphrey LL. Patient-centred outcomes among lung cancer screening recipients with computed tomography: A systematic review. J Thorac Oncol 2014;9(7):927-934. [http://dx.doi.org/10.1097/JTO.0000000000000210] 43. Kucharczyk MJ, Menezes RJ, McGregor A, Paul NS, Roberts HC. Assessing the impact of incidental findings in a lung cancer screening study by low-dose computed tomography. Can Assoc Radiol J 2011;62(2):141-145. [http://dx.doi.org/10.1016/j. carj.2010.02.008]

107 SARJ VOL. 21 NO. 4 2015

44. Marcus PM, Bergstralh EJ, Zweig MH, Harris A, Offord KP, Fontana RS. Extended lung cancer incidence follow-up in the Mayo Lung Project and overdiagnosis. J Natl Cancer Inst 2006;98(11):748-756. [http://dx.doi.org/10.1093/jnci/djj207] 45. Yankelevitz DF, Kostis WJ, Henschke CI, et al. Overdiagnosis in chest radiographic screening for lung carcinoma: Frequency. Cancer 2003;97(5):1271-1275. [http:// dx.doi.org/10.1002/cncr.11185] 46. Smith-Bindman R, Lipson J, Marcus R, et al. Radiation dose associated with common computed tomography examinations and the associated lifetime attributable risk of cancer. Arch Intern Med 2009;169(22):2078-2086. [http://dx.doi.org/10.1001/ archinternmed.2009.427] 47. Mascalchi M, Mazzoni LN, Falchini M, et al. Dose exposure in the ITALUNG trial of lung cancer screening with low-dose CT. Br J Radiol 2012;85(1016):1134-1139. [http://dx.doi.org/10.1259/bjr/20711289] 48. Brenner DJ. Radiation risks potentially associated with low-dose CT screening of adult smokers for lung cancer. Radiology 2004;231(2):440-445. [http://dx.doi.org/10.1148/ radiol.2312030880] 49. National Lung Screening Trial Research Team, Aberle DR, Adams AM, Berg CD, et al.

Baseline characteristics of participants in the randomised national lung screening trial.

J Natl Cancer Inst 2010;102(23):1771-1779. [http://dx.doi.org/10.1093/jnci/djq434] 50. Field JK, Smith RA, Aberle DR, et al. IASLC CT Screening Workshop 2011 Participants. International Association for the Study of Lung Cancer Computed Tomography Screening Workshop 2011 report. J Thorac Oncol 2012;7(1):10-19. [http://dx.doi.org/10.1097/JTO.0b013e31823c58ab] 51. Bach PB, Kattan MW, Thornquist MD, et al. Variations in lung cancer risk among smokers.

J Natl Cancer Inst 2003;95(6):470-478. [http://dx.doi.org/10.1093/jnci/95.6.470] 52. Spitz MR, Hong WK, Amos CI, et al. A risk model for prediction of lung cancer. J Natl Cancer Inst 2007;99(9):715-726. [http://dx.doi.org/10.1093/jnci/djk153] 53. Cassidy A, Myles JP, van Tongeren M, et al. The LLP risk model: An individual risk prediction model for lung cancer. Br J Cancer 2008;98(2):270-276. [http://dx.doi. org/10.1038/sj.bjc.6604158] 54. Hoggart C, Brennan P, Tjonneland A, et al. A risk model for lung cancer incidence.

Cancer Prev Res (Phila) 2012;5(6):834-846. [http://dx.doi.org/10.1158/1940-6207.

CAPR-11-0237] 55. Alberg AJ, Samet JM. Epidemiology of lung cancer. Chest 2003;123(1 Suppl):21S-49S. 56. Fontham ET, Correa P, Reynolds P, et al. Environmental tobacco smoke and lung cancer in non-smoking women. A multicentre study. JAMA 1994;271(22):1752-1759.

57. McDuffie HH, Klaassen DJ, Dosman JA. Men, women and primary lung cancer – a Saskatchewan personal interview study. J Clin Epidemiol 1991;44(6):537-544.

58. Osann KE, Anton-Culver H, Kurosaki T, Taylor T. Sex differences in lung-cancer risk associated with cigarette smoking. Int J Cancer 1993;54(1):44-48.

59. Burns DM. Primary prevention, smoking, and smoking cessation: Implications for future trends in lung cancer prevention. Cancer 2000;89(11 Suppl):2506-2509.

60. Halpern MT, Gillespie BW, Warner KE. Patterns of absolute risk of lung cancer mortality in former smokers. J Natl Cancer Inst 1993;85(6):457-464.

61. Tong L, Spitz MR, Fueger JJ, Amos CA. Lung carcinoma in former smokers. Cancer 1996;78(5):1004-1010.

62. Office on Smoking and Health, Centers for Disease Control and Prevention. Trends in current cigarette smoking among high school students and adults, United States, 1965 - 2011. http://www.cdc.gov/tobacco/data_statistics/tables/trends/cig_smoking/ (accessed 3 September 2015).

63. Slatore CG, Baumann C, Pappas M, Humphrey LL. Smoking behaviors among patients receiving computed tomography for lung cancer screening. Systematic review in support of the US Preventative Services Task Force. Ann Am Thorac Soc 2014;11(4):619-627. [http://dx.doi.org/10.1513/AnnalsATS.201312-460OC] 64. Tammemägi MC, Berg CD, Riley TL, Cunningham CR, Taylor KL. Impact of lung cancer screening results on smoking cessation. J Natl Cancer Inst 2014;106(6):dju084. [http://dx.doi.org/10.1093/jnci/dju084] 65. Villanti AC, Jiang Y, Abrams DB, Pyenson BS. A cost-utility analysis of lung cancer screening and the additional benefits of incorporating smoking cessation interventions. PLoS One 2013;8(8):e71379. [http://dx.doi.org/10.1371/journal. pone.0071379] 66. Sox HC. Better evidence about screening for lung cancer. N Engl J Med 2011;365(5):455-457. [http://dx.doi.org/10.1056/NEJMe1103776] 67. Black WC, Gareen IF, Samir SS, et al. Cost-effectiveness of CT screening in the National Lung Screening Trial. N Engl J Med 2014;371(19):1793-1802. [http://dx.doi. org/10.1056/NEJMoa1312547]

CASE REPORT

Hypereosinophilia as a paraneoplastic phenomenon in non-smallcell lung carcinoma L Nqwata,¹ MB ChB, FCP (SA); M L Wong,¹ MB BCh, DCh (SA), FCP (SA), FCCP, FRCP (Lond); R D Mohanlal,2 MB ChB, DMH, FCPath (SA), MMed; A B Lakha,3 MB ChB, FCP (SA), Cert Clin Haem (SA) Division of Pulmonology, Chris Hani Baragwanath Academic Hospital; and the Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa 2 Department of Anatomical Pathology, School of Pathology, Faculty of Health Sciences, University of the Witwatersrand; and National Health Laboratory Services, Chris Hani Baragwanath Academic Hospital, Johannesburg, South Africa ³ Division of Clinical Haematology, Chris Hani Baragwanath Academic Hospital; and the Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa 1

Corresponding author: L Nqwata (drlamla@yahoo.com)

Hypereosinophilia is a rare paraneoplastic finding in malignant disease, particularly lung cancer. When it occurs, it is usually indicative of metastatic disease. We describe a 52-year-old male patient with paraneoplastic hypereosinophilia associated with primary adenocarcinoma of the right lower lobe and extensive metastatic disease.

S Afr Respir J 2015;21(4):108-109. DOI:10.7196/SARJ.2015.v21i4.38

Hypereosinophilia is defined as peripheral blood eosinophilia ≥1.5 × 109/L on two different occasions at least one month apart, with or without tissue eosinophilia, but without evidence of eosinophilinduced organ damage or dysfunction.[1] Rarely, it may be encountered as a paraneoplastic phenomenon.

Case report

We report a 52-year-old male patient who presented with a 2-month history of cough and significant weight loss. His past

Fig. 1. Chest radiograph showing large mass in the right lower lobe (white arrow) and metastatic pulmonary nodule in the left midzone (red arrow).

medical history was unremarkable, except for chronic obstructive pulmonary disease (COPD) and long-standing, well-controlled epilepsy. Physical examination revealed a wasted patient with slow cognition not associated with any focal neurological deficits. Clinical examination revealed decreased breath sounds with dullness to percussion over the right lower lobe, and hepatomegaly extending 8 cm below the costal margin. Blood investigations showed marked elevation of his white cell count (peak level 114.18 × 109/L) with an absolute eosinophilia ranging from 29.81 to 82.33 × 109/L during the course of his admission. The bone marrow trephine confirmed marked eosinophilia, with no malignant infiltrate or evidence of clonal eosinophilic proliferation (absence of the FIP1L1/PDGFRA fusion gene on fluorescence in situ hybridisation (FISH) analysis).

The chest radiograph showed a well-circumscribed lesion in the

Fig. 2. CT scan demonstrating lobulated right lower lobe mass (red arrow).

SARJ VOL. 21 NO. 4 2015

108

CASE REPORT right lower lobe with a pulmonary nodule in the left upper lobe (Fig. 1). Computed tomography demonstrated a right lower lobe mass measuring 9.1 × 6.0 × 6.7 cm (Fig. 2), bilateral pulmonary nodules, multiple liver lesions and a left adrenal mass. A core biopsy of the lung mass revealed nests of tumour cells with abundant eosinophilic cytoplasm and extensive tumour necrosis.

No keratinisation or well-formed glands were noted. A panel of immunohistochemical stains was positive for CK7, MOC31 and thyroid transcription factor (TTF-1) in the malignant cells. Other stains performed, including p63, calretinin and WT1, excluded mesothelioma and squamous cell carcinoma. A final diagnosis of poorly differentiated adenocarcinoma of the lung was made.

High-dose corticosteroids and hydroxyurea were administered to the patient, with only modest reduction in the eosinophilia.

He was referred to the palliative care team and died 47 days after admission.

Discussion

Paraneoplastic hypereosinophilia is rare. It has been described with many solid tumours including thyroid, breast, genitourinary, gastrointestinal, hepatocellular and in both non-small- and small-cell lung carcinoma.[2] Several studies suggest that local production of cytokines (GMCSF, IL-3, and IL-5) by tumour cells is the most likely mechanism, particularly IL-5, which is considered to be the most important eosinophilopoetin.[3-5] These cytokines are involved in eosinophil development and maturation, with predominantly IL-5 mobilising eosinophils from bone marrow into the blood.

109 SARJ VOL. 21 NO. 4 2015

Management of eosinophilia is best achieved by tumour reduction (surgery and/or chemotherapy). However, case reports suggest that this is not always feasible, as these patients usually have extensive metastases and poor outcomes.[3-5]

Conclusion

Our case report illustrates paraneoplastic hypereosinophilia, a rare manifestation in primary lung cancer. As in other case reports,[3-5] this is associated with a poor prognosis.

Learning points: • Hypereosinophilia may occur as a paraneoplastic phenomenon in primary lung cancer.

• Paraneoplastic hypereosinophilia portends a poor prognosis.

References 1. Valent P, Klion AD, Horny HP, et al. Contemporary proposal on criteria and classification of eosinophilic disorders and related syndromes. J Allergy Clin Immunol 2012;130(3):607-612. [http://dx.doi.org/10.1016/j.jaci.2012.02.019] 2. Samoszuk M. Eosinophils and human cancer. Histol Histopathol 1997;12(3):807-812.

3. Venkatesan R, Salam A, Alawin I, Willis M. Non-small cell lung cancer and elevated eosinophil count: A case report and literature review. Cancer Treatment Communications 2015;4:55-58. [http://dx.doi.org/10.1016/j.ctrc.2015.05.002] 4. Pandit R, Scholnik A, Wulfekuhler L, et al. Non-small cell lung cancer associated with excessive eosinophilia and secretion of interleukin-5 as a paraneoplastic syndrome.

Am J Hematol 2007;82(3):234-237. [http://dx.doi.org/10.1002/ajh.20789] 5. Verstraeten AS, de Weerdt A, van den Eynden G, et al. Excessive eosinophilia as paraneoplastic syndrome in a patient with non-small-cell lung carcinoma: A case report and review of the literature. Acta Clin Belg 2011;66(4):293-297. [http://dx.doi. org/10.1179/ACB.66.4.2062571]

BREATH-TAKING NEWS

The challenge of the asthma-COPD overlap syndrome (ACOS) In the past, asthma airway inflammation was believed to be predominantly characterised by eosinophilic inflammation and type 2 helper T (Th2) lymphocytes, unlike chronic obstructive pulmonary disease (COPD), where inflammation is characterised predominantly by neutrophilic inflammation and CD8 lymphocytes. This difference in the pattern of inflammation makes the clinical extremes of asthma and COPD easily distinguishable, with differences in clinical picture and age of the patients. However, over the years, it has been observed that in older patients the presentation of asthma and COPD may converge clinically and the conditions may mimic each other. Airway remodelling that develops over time in some asthma patients leads to irreversible airway obstruction resembling COPD. In contrast, reversible airway obstruction can occur in patients with COPD, with the result that these patients may resemble those with asthma. The condition in which a person has clinical features of both asthma and COPD is called the asthma-COPD overlap syndrome (ACOS).

The prevalence of ACOS is estimated to be 15 - 45% in people with obstructive airway disease and increases with age.

In a recent review, Postma and Rabe[1] discuss studies that have shown the heterogeneity of pathophysiology of asthma and COPD. Therefore it may be difficult to distinguish asthma from COPD in patients who have pathophysiological and clinical features of both.

The authors emphasised that there is still a paucity of data on how to diagnose and treat ACOS and answers to the primary review questions (i.e. how would one make appropriate ACOS diagnosis and what is the appropriate treatment for ACOS?) are not evidencebased. Nevertheless, the review highlighted important points on pathophysiology, clinical features, diagnosis and treatment of obstructive airway disease assuming that asthma and COPD are two extreme ends of the disease spectrum with ACOS in between.

Pathophysiology of asthma, COPD and ACOS and clinical implications

Studies have shown substantial heterogeneity in progressive airway obstruction, bronchial hyper-responsiveness, reversibility of airway obstruction, atopy, airway inflammation and exhaled nitric oxide (FeNO) among patients with asthma or COPD. From early adulthood, forced expiratory volume in 1 second (FEV1) normally declines by ~25 - 50 ml annually. The decrease is greater in obstructive airway disease. Although the decrease is more in COPD than asthma, there is no convincing evidence that the rate of FEV1 decline can be used to distinguish between asthma and COPD. The global trend of increasing life expectancy shifts the median age of the population with asthma upward. This increases the probability of overlap with COPD as defined by FEV1; hence, the estimated prevalence of ACOS is highly age-dependent.

Bronchial hyper-responsiveness, which was initially thought to be a hallmark of asthma and a marker of eosinophilic inflammation, is also present in some COPD patients. It is driven by multiple factors (allergen and non-allergen) and may not be as responsive to steroids as it is in asthma.

Reversibility of airway obstruction after inhalation of a bronchodilator drug can diminish or even disappear with longstanding asthma and treatment with a bronchodilator drug or inhaled steroids. Therefore, lack of full reversibility does not rule out an asthma diagnosis.

Reversibility of airway obstruction is frequently present in COPD as well; reversibility was observed in up to 50% of patients with COPD. Atopy, which is a risk factor for asthma, has also been found to be a risk factor for COPD in two studies (ECLIPSE and EUROSCOP). COPD patients with atopy are somewhat younger, more likely male, with higher body mass index and more likely to benefit from steroid treatment.

There is evidence from bronchial biopsy/sputum studies/exhaled breath studies that there is substantial heterogeneity in mucosal inflammation. Asthma patients with severe or late-onset disease or chronic infections or who smoke, may also exhibit neutrophilic inflammation and CD8 cells in the airways, both of which were once believed to be hallmarks of COPD. Therefore, the absence of eosinophilia and the lack of a response to inhaled glucocorticoids in a patient does not rule out asthma. Similarly, some COPD patients may have eosinophilic inflammation which is not usually steroid responsive.

Exhaled FeNO levels are lower in smokers than in non-smokers, which makes measurements of FeNO levels less useful for differentiating asthma from COPD.

Diagnosis of ACOS

So far there is no agreed specific ACOS diagnosis definition to guide clinicians. However, depending on patient age, smoking history, atopy history, bronchial hyper-responsiveness, airway obstruction reversibility and symptoms, patients with obstructive airway disease can either have ‘easy’ asthma and ‘easy’ COPD or overlap syndrome (ACOS). ACOS patients can evolve either from asthma or from COPD. The Global Initiative for Asthma (GINA) and Global Initiative for Chronic Obstructive Lung Disease (GOLD) propose that if three or more features of either asthma or COPD are present, then that diagnosis is suggested; if there are similar numbers of features of asthma and COPD, the diagnosis of ACOS should be considered.

• ‘Easy asthma’ patients: characteristically young (about 21 years), with a history of atopy, non-smokers, recurrent wheezing/ dyspnoea, reversible airway obstruction and bronchial hyperresponsiveness.

• ‘Easy COPD’ patients: characteristically older (about 65 years), with no atopy history, heavy smokers (about 95 pack-years), chronic dyspnoea, no wheezing/reversible airway obstruction/ bronchial hyper-responsiveness.

• ACOS stemming from asthma: middle-aged (about 45 years) with atopy history, non-smoker, chronic dyspnoea with flares and wheezing, bronchial hyper-responsiveness, but no airway obstruction reversibility.

• ACOS stemming from COPD: middle-aged (about 45 years) with atopy, smoker (20 pack-years), chronic dyspnoea with flares, wheezing, reversible airway obstruction and with or without bronchial hyper-responsiveness.

SARJ VOL. 21 NO. 4 2015

110

BREATH-TAKING NEWS

Clinical relevance of ACOS and treatment for ACOS

Despite GINA and GOLD recognising ACOS as a clinical reality, there is still need for more evidence on clinical phenotypes and underlying mechanisms that will help in devising a specific definition of ACOS and its treatment.

There is the possible danger of overdiagnosing ACOS, thereby overtreating COPD patients with inhaled glucocorticoids, if different definitions of ACOS are used.

Inconsistent definitions used in treatment studies make it almost impossible to determine the most effective therapy for an individual ACOS patient.

For ‘easy’ asthma and ‘easy COPD’, clear step-wise treatment approaches have been provided by GINA and GOLD, respectively. These approaches also include treatment for exacerbation reduction and comorbidities.

• ‘Easy asthma’: Inhaled glucocorticoids in combination with bronchodilator drugs (short- and long-acting beta-agonists (LABAs) are the main pillars. Leukotriene-receptor antagonists are an alternative choice in milder disease cases. For severe allergic asthma with appropriate IgE levels, anti-IgE treatment is an approved option.

• ‘Easy COPD’: The main emphasis is on smoking cessation and the use of LABAs and long-acting muscarinic antagonists (LAMAs). The role of inhaled glucocorticoids is still debatable; however, it is accepted for patients with more severe disease and those with frequent exacerbations.

• ACOS stemming from asthma: There are no firm treatment

guidelines. However, due to lack of randomised clinical trials, treatment with inhaled glucocorticoids should be continued in patients with long-standing asthma, even if a component of irreversible airway obstruction develops; leukotriene modifiers may be of value in those with atopy. Combination therapy with a LAMA and a LABA is a reasonable approach for patients with more severe asthma or COPD or with overlapping conditions.

• ACOS stemming from COPD: Because of the current evidence that some COPD patients may have reversibility, eosinophilia and bronchial hyper-responsiveness, such patients may benefit from inhaled glucocorticoids.

Conclusion

With no clinical trials on ACOS, the authors believe it is still premature to recommend the designation of ACOS as a disease entity. There is need for further research to obtain a standardised definition of ACOS and its treatment; unique biomarkers will settle the debate.

Ndaziona P K Banda Division of Pulmonology, Chris Hani Baragwanath Academic Hospital and Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa

References

1. Postma DS, Rabe KF. The asthma-COPD overlap syndrome. N Engl J Med 2015;373(13):1241-1249. [http://dx.doi.org/10.1056/NEJMra1411863]

S Afr Respir J 2015;21(4):110-111. DOI:10.7196/SARJ.2015.v21i4.49

High-flow oxygen therapy in acute hypoxaemic respiratory failure High-flow oxygen therapy allows for the delivery of heated and humidified oxygen via nasal cannula at high flow rates, thereby generating low levels of positive pressure in the upper airways. Additionally, this technique allows for the titration of the fraction of inspired oxygen as well.[1-3] High-flow oxygen therapy has been shown to result in better comfort and oxygenation in patients with acute respiratory failure in previous studies. However, evidence for high-flow oxygen therapy on intubation rates and mortality is lacking.

Frat et al.[4] in a recent trial compared intubation rates within 28 days in patients who were assigned to either high-flow oxygen therapy, standard facemask oxygen therapy or non-invasive positive pressure ventilation.

Secondary outcomes were the number of ventilator-free days at day 28 and all-cause mortality at 90 days. In this multicentre, open-label trial 310 patients across 23 intensive care units in Belgium and France were randomly assigned to each group. All patients included in the study had acute hypoxaemic respiratory failure without hypercapnia, and a partial pressure to fraction of respired oxygen ratio of <300 mm Hg.

Results

The intubation rate was non-significantly different in the three groups: 38% in the high-flow oxygen group, 47% in the standard facemask oxygen group, and 50% in the non-invasive ventilation group (p=0.18).

The number of ventilator-free days at day 28 was analysed as a secondary outcome, and was significantly lower in the high-flow oxygen group (24 (8) days compared with 22 (10) in the facemask group and 19 (12) in the non-invasive ventilation group); for all analyses p=0.02. High-flow

111 SARJ VOL. 21 NO. 4 2015

oxygen therapy was also found to significantly lower 90-day mortality as a secondary outcome compared with standard oxygen therapy and non-invasive ventilation. The hazard ratio for death at 90 days was 2.01 (95% confidence interval (CI) 1.01 - 3.99) when comparing standard oxygen with high-flow oxygen therapy (p=0.046) and 2.5 (95% CI 1.31 - 4.78) with non-invasive ventilation compared with high-flow oxygen (p=0.006). Given the findings of this study, high-flow oxygen therapy appears to be a reasonable alternative to facemask oxygen and non-invasive ventilation in patients with acute hypoxaemic respiratory failure and may confer a survival benefit.

Laila Suleman Department of Pulmonology and Critical Care, Charlotte Maxeke Johannesburg Academic Hospital and University of the Witwatersrand, Johannesburg, South Africa

References

1. Chanques G, Riboulet F, Molinari N, et al. Comparison of three high flow oxygen therapy devices: A clinical physiological cross-over study. Minerva Anestesiol 2013;79(12):1344-1355.

2. Corley A, Caruana LR, Barnett AG, Tronstad O, Fraser JF. Oxygen delivery through high-flow nasal cannulae increase end-expiratory lung volume and reduce respiratory rate in post-cardiac surgical patients. Br J Anaesth 2011;107(6):998-1004. [http:// dx.doi.org/10.1093/bja/aer265] 3. Parke RL, Eccleston ML, McGuinness SP. The effects of flow on airway pressure during nasal high-flow oxygen therapy. Respir Care 2011;56(8):1151-1155. [http://dx.doi. org/10.4187/respcare.01106] 4. Frat JP, Thille AW, Mercat A, et al. High-flow oxygen through nasal cannula in acute hypoxaemic respiratory failure. N Engl J Med 2015;372(23):2185-2196. [http://dx.doi. org/10.1056/NEJMoa1503326]

S Afr Respir J 2015;21(4):111. DOI:10.7196/SARJ.2015.v21i4.48

WHO’S WHO

Prof. Risenga Frank Chauke MB ChB, MMed (Thoracic-Chir) (Medunsa), Ass FC (Cardio) SA, MBA (Gibs-UP)

Prof. Chauke is the Professor and Head of the Cardiothoracic Surgery Department at the joint newly established Sefako Makgatho Health Sciences University (SMU), formerly known as the University of Limpopo (Medunsa Campus) and Dr George Mukhari Academic Hospital complex. His mandate as a joint appointee in the complex includes service delivery, teaching and research.

Prof. Chauke completed his higher education at Bankuna High School in 1988. He went on to do an MB ChB at Medunsa, graduating in 1994, after which he embarked on a specialist course in Cardiothoracic Surgery (MMed Thoracic Chir), which he completed in 2002. He was invited to be an associate in the College of Cardiothoracic Surgeons of the Colleges of Medicine of South Africa (SA), and went on to do a course in learning facilitation through the Assessment College of SA in 2012. He then enrolled for a Masters in Business Administration through the Gordon Institute of Business Science at the University of Pretoria, which he completed in 2014. Prof. Chauke has been awarded certificates from a number of attendance courses, which include, among others, a certificate in video-assisted thoracoscopic surgery and a certificate in thoracic endovascular aneurysm repair.

He has worked at hospitals such as Letaba Hospital, Pretoria Academic Hospital and Ga-Rankuwa Hospital, now called Dr George Mukhari Academic Hospital. He is also running his limited private practice at Louis Pasteur Private Hospital.

Prof. Chauke is also the chairperson of the Bankuna Alumni association, which aims at motivating and giving career guidance to the high-school students at Bankuna High School and Nkowankowa circuit.

He sits on various committees and boards, including: • School Board of Medicine, SMU • Faculty Board of Health Science, SMU • Medical Advisory Council, SMU and Dr George Mukhari Academic Hospital • Executive Committee of SAMA Trade Union • Executive Committee of the Society of Cardiothoracic Surgeons of SA • Councillor of the College of Cardiothoracic Surgeons of SA, a division of the Colleges of Medicine of SA • Vice-Chair of SAMA Gauteng North Branch.

He previously served as: • Medunsa Student Representative Council Ad-hoc President • Executive Committee of Senior Doctors Association of SA • Chairperson of the Board of Calvary Servanthood Community Church.

Prof. Chauke serves in various business structures, including: • African Haze Trading 20cc, trading as Comprehensive Cure • Tsholofelo Entsha (Pty) Ltd. He also serves on local organising committees for national and international conferences.

Prof. Goolam Mahomed MB BCh, FCP (SA), FCCP

Prof. Goolam Mahomed is Associate Professor and Head of the Department of Intensive Care, Dr George Mukhari Academic hospital and Sefako Makgatho Health Sciences University (SMU) (formerly Medunsa Campus, University of Limpopo), South Africa (SA).

His current main research interests are: • ventilator-associated pneumonia • pu l m on ar y hy p e r t e ns i on an d pulmonary thrombo-embolism • the effects of HIV on the lung and outcomes in intensive care

• the role of vitamin-D in intensive care and pulmonary disease • drug-resistant tuberculosis (TB).

He is Chairman of the Islamic Medical Association of SA, Tshwane Branch, and Chairman of the Jacaranda Branch of the Critical Care Society of Southern Africa.

He was awarded the Medical Research Council of SA University Flagship project grant for the trial on treatment of multidrug-resistant tuberculosis (TB) – a new approach. This is a 3-year major grant worth R16.5 million over 3 years. He will be using this grant to do a multicentre study in collaboration with Prof. Keertan Dheda at the University of Cape Town Lung Institute.

SARJ VOL. 21 NO. 4 2015

112

PRODUCT NEWS

113 SARJ VOL. 21 NO. 4 2015

PRODUCT NEWS Rivaroxaban reduces length of hospital stay in patients with symptomatic venous thromboembolism The phase III EINSTEIN deep vein thrombosis (DVT) and EINSTEIN pulmonary embolism (PE) trials demonstrated the potential of oral rivaroxaban (Xarelto, Bayer) – 15 mg twice daily for 21 days, followed by 20 mg once daily – for the treatment of venous thromboembolism (VTE), a term that embraces DVT and PE. A subsequent study by van Bellen et al.,[1] published in Current Medical Research and Opinion in 2014, was undertaken to assess the length of initial hospitalisation in patients presenting with either symptomatic DVT or PE using hospitalisation records from these trials.

The authors found that overall 52% of EINSTEIN DVT patients and 90% of EINSTEIN PE patients were admitted to hospital. The proportion of hospitalised DVT patients with a length of stay 5 days or fewer, receiving rivaroxaban, was 54% compared with 31% for those receiving enoxaparin/vitamin K antagonist (VKA), the current standard of care for the treatment of patients with symptomatic DVT and PE. For patients with PE, the corresponding values were 45% and 33%. Stays of 6 - 10 days were observed in 29% of rivaroxaban-treated patients compared with 45% for enoxaparin/VKA-treated patients for DVT. For patients with PE, these values were 39% and 46% in the rivaroxaban and enoxaparin/ VKA groups, respectively. Overall, length of stay was significantly shorter in the rivaroxaban group, compared with the enoxaparin/VKA group across all analyses performed (p<0.0001).

VTE is associated with significant morbidity and mortality and therefore carries a considerable healthcare burden. Rivaroxaban is as effective as enoxaparin/VKA for the treatment of acute symptomatic DVT or PE, with the additional benefit of significantly reducing the period of hospitalisation in patients being treated for an initial DVT or PE. ‘Coupled with improved patient treatment satisfaction and no requirement for routine monitoring or dose adjustment, this presents strong advantages for treating patients with VTE with rivaroxaban,’ the authors wrote.

They concluded that a single-drug regimen with rivaroxaban may reduce the burden on healthcare systems and patients by providing effective and well-tolerated treatment. ‘The convenience of a single-drug approach with oral rivaroxaban has the potential to allow discharge based on a patient’s clinical condition and to facilitate the transition from in-hospital to outpatient care. […] However, assessment of patient risk is still warranted to identify candidates who can safely receive outpatient treatment, and patient monitoring is essential to ensure adherence to the specified dosing regimen.’

Reference

1. van Bellen B, Bamber L, Correa de Carvalho F, et al. Reduction in the length of stay with rivaroxaban as a single-drug regimen for the treatment of deep vein thrombosis and pulmonary embolism. Curr Med Res Opin 2014; 30(5):829-837. [http://dx.doi.org/10.1185/03007995.2013.879439]

For full prescribing information, refer to the package insert approved by the Medicines Regulatory Authority (MCC).

PHARMACOLOGICAL CLASSIFICATION: A.8.2 Anticoagulants.

S4 XARELTO® 10. Reg. No.: 42/8.2/1046. Each film-coated tablet contains rivaroxaban 10 mg.

INDICATION: Prevention of VTE in patients undergoing major orthopaedic surgery of the lower limbs.

S4 XARELTO® 15 and XARELTO® 20. Reg. No.: 46/8.2/0111 and 46/8.2/0112. Each film-coated tablet contains rivaroxaban 15 mg or 20 mg, respectively.

INDICATIONS: Prevention of stroke and systemic embolism in patients with non-valvular atrial fibrillation; Treatment of DVT and for the prevention of recurrent DVT and PE; Treatment of PE and for the prevention of recurrent PE and DVT. Bayer (Pty) Ltd, Co. Reg. No.: 1968/011192/07, 27 Wrench Road, Isando, 1609. Tel: 011 921 5044 Fax: 011 921 5041.

L.ZA.GM.06.2014.1007

© Bayer HealthCare Pharmaceuticals June 2014

SARJ VOL. 21 NO. 4 2015

114

EVENTS UNIVERSITY OF CAPE TOWN D E PA R T M E N T O F M E D I C I N E

GENERAL PHYSICIANS CONFERENCE 2016 THURSDAY 18 – SUNDAY 21 FEBRUARY 2016 CAPE TOWN INTERNATIONAL CONVENTION CENTRE

SAVE THE DATE! MORE INFORMATION

AN INTERACTIVE APPROACH TO COMMON MEDICAL DISORDERS AND EMERGENCIES The conference will bring together a panel of expert speakers from Academic Institutions in Southern Africa.

INTERACTIVE SESSIONS,QUIZZES AND UPDATES The topics will be relevant to specialist physicians and trainees, in the private and public sectors, and colleagues beyond our borders.

‘MEET THE EXPERTS’ WORKSHOPS 18 February, Groote Schuur Hospital Topics will include the fields of: Allergology and Clinical Immunology, Cardiology, Dermatology, Diabetes and Endocrinology, Gastroenterology, Geriatrics, Haematology, Hepatology Infectious Diseases and HIV Medicine, Lipidology, Medical Ethics, Nephrology, Neurology, Pharmacotherapy, Pulmonology, and Rheumatology.

>>

W W W. P H Y S I C I A N S 2 0 1 6 . C O . Z A

+27 (0)21 406 6733

The South African Respiratory Journal PO Box 13725 Mowbray 7705 Any correspondence to the Editor should be sent to the same address or via email to: sarj@iafrica.com The website of The South African Thoracic Society can be found at www.pulmonology.co.za

115 SARJ VOL. 21 NO. 4 2015

FOXAIR

everyone and it’s yours to give

Air is for

Why wait to

prescribe?

S4 FOXAIR® 50/250 and 50/500 ACCUHALER® - 42/21.5.4/0582; 0583. Each blister contains a mixture of salmeterol xinafoate equivalent to 50 µg of salmeterol and microfine fluticasone propionate (250 µg or 500 µg). Applicant: GlaxoSmithKline South Africa (Pty) Ltd. (Co. Reg. No. 1948/030135/07). 39 Hawkins Avenue, Epping Industria 1, Cape Town, 7460. For full prescribing information plese refer to the package insert approved by the Medicines Regulatory Authority. All adverse events should be reported by calling the Aspen Medical Hotline number or directly to GlaxoSmithKline on +27117456000. FO/0713/933 A16772 08/13

The South African Respiratory Journal acknowledges with thanks the invaluable sponsorship of the following companies: Aspen GSK Division Bayer Healthcare