Arch Gynecol Obstet (2009) 279:293–297 DOI 10.1007/s00404-008-0707-3
O R IG IN AL AR T I CL E
Dipyrone use during pregnancy and adverse perinatal events Tatiane da Silva Dal Pizzol · Lavínia Schüler-Faccini · Sotero Serrate Mengue · Maria Isabel Fischer
Received: 12 March 2008 / Accepted: 2 June 2008 / Published online: 21 June 2008 Springer-Verlag 2008
Abstract Objective To evaluate the risk of adverse perinatal events among newborns exposed to dipyrone during gestation. Design and Setting The present study is a secondary analysis of Brazilian study of gestational diabetes (EBDG), a cohort of women attended at healthcare units of the Brazilian national health system (SUS) located in six Brazilian state capitals, between February 1991 and June 1995. Sample A total number of 5,564 women aged 20 years and over who were between their 21st and 28th week of pregnancy were followed up. Methods A structured questionnaire was used to obtain data on the pregnant women, their pregnancies, and their use of medications. Other data and the outcomes congenital abnormalities, intrauterine death, preterm birth, or low birth
T. da Silva Dal Pizzol (&) Departamento de Produção e Controle de Medicamentos, Faculdade de Farmácia, Universidade Federal do Rio Grande do Sul, Av. Ipiranga, 2752, CEP 90640-000 Porto Alegre, RS, Brazil e-mail: tatiane.silva@ufrgs.br L. Schüler-Faccini Departamento de Genética, Universidade Federal do Rio Grande do Sul, Porto Alegre, RS, Brazil S. S. Mengue Programa de Pós-Graduação em Epidemiologia, Faculdade de Medicina, Universidade Federal do Rio Grande do Sul, Porto Alegre, RS, Brazil M. I. Fischer Centro de Informações sobre Medicamentos, Conselho Regional de Farmácia do Rio Grande do Sul e Faculdade de Farmácia, Universidade Federal do Rio Grande do Sul, Porto Alegre, RS, Brazil
weight were obtained from the medical records. To estimate the odds ratios after adjustment for the potential confounding factors, logistic regression modeling was developed. Main outcome measures Congenital abnormalities, intrauterine death, preterm birth, and low birth weight. Results Dipyrone use was reported by 555 pregnant women (11.5%). Their exposure to this medication did not present any association with the outcomes of congenital abnormalities (OR 1.11; 95% CI, 0.58–2.10), intrauterine death (OR 0.69; 95% CI, 0.33–1.43), preterm birth (OR 0.94; 95% CI, 0.73–1.20), or low birth weight (OR 0.88; 95% CI, 0.64–1.22), in the crude analysis. This absence of associations was maintained after performing logistic regression analysis. Conclusions The data suggest that the exposure to dipyrone during pregnancy does not increase the risk of congenital abnormalities and other adverse events as outcomes from pregnancy. Keywords Dipyrone · Congenital abnormalities · Intrauterine death · Preterm birth · Low birth weight
Introduction Dipyrone, which was withdrawn from the market in the United States and some other countries because of a possible association with agranulocytosis and aplastic anemia, continues to be used in some parts of Europe, South America and Asia. According to the international agranulocytosis and aplastic anemia study (IAAA), the incidence of agranulocytosis is 4.4 per million/year, with a fatality rate of 10% [1] A systematic review considered that dipyrone is a drug involved in cases of agranulocytosis with deWnite or probable
123
294
causality, with data from IAAA, that has been both criticized and defended [2]. In Brazil, dipyrone is the principal analgesic used, with 31.8% of the market, followed by paracetamol (29.7%) and acetylsalicylic acid (27.1%) [3]. Studies on the use of medication during pregnancy have revealed that dipyrone is also the analgesic that is most used during this period of women’s lives [4, 5]. The fact that dipyrone has been withdrawn from the market in many countries, including the United States and some European countries, has to a large degree contributed towards the scarcity of evidence on the safety of this medication during pregnancy. Preliminary studies have suggested that dipyrone might have a carcinogenic eVect. Sharpe et al. [6] found a positive association between antenatal exposure to dipyrone and Wilms’ tumor among low-income pregnant women (OR 10.9; 95% CI, 2.4–50.0), but not among medium (OR 0.66; 95% CI, 0.17–2.60) or high-income pregnant women (OR 2.56; 95% CI, 0.72–9.12). In another case-control study that evaluated exposure to various chemical agents including medications, in relation to increased risk of acute childhood leukemia, a positive association was found between dipyrone and leukemia (OR 5.84; 95% CI, 2.09–16.30) [7]. Three case reports have associated dipyrone use at high doses during the third trimester of pregnancy with oligohydramnios [8–10]. One case of persistent pulmonary hypertension in a newborn following dipyrone use eleven days before delivery has been reported [11]. No controlled studies investigating these possible associations were located in the literature. Information on the teratogenicity of dipyrone in women is also rare, coming only from two studies published in 2005 and 2007. In one of the studies, no association was found between dipyrone and congenital abnormalities, in a comparison with paracetamol (OR 1.55; 95% CI, 0.26–9.05) [12]. In the other study, a positive association was found between reported dipyrone use and diaphragmatic defects (OR 2.7; 95% CI, 1.0–6.8), cardiovascular abnormalities (OR 1.3; 95% CI, 1.0–1.7) and various other abnormalities (OR 1.8; 95% CI, 1.1–2.9), in comparison with population-based controls [13]. However, when compared with malformed controls, no diVerence was found in relation to dipyrone use. Considering the need for information on the safety of dipyrone during pregnancy, the objective of the present study was to investigate the risk of adverse perinatal events after exposure to dipyrone during gestation.
Arch Gynecol Obstet (2009) 279:293–297
cohort that investigated diabetes and glucose intolerance among pregnant women, in relation to prevalence, risk factors, incidence of obstetric and neonatal complications, prognostic factors, and diagnostic criteria. The Wrst phase of this study included enrollment of pregnant women, interviews, anthropometric measurements, and glucose tolerance tests. In the second phase, the pregnancies were followed up by means of reviewing the medical records. The third phase consisted of collecting information of the delivery. Women aged 20 years and over between their 21st and 28th gestational weeks, if reporting a negative history of diabetes mellitus prior to the pregnancy were enrolled. These women were attended at antenatal services of healthcare units within the Brazilian national health system (Sistema Único de Saúde, SUS) located in six Brazilian state capitals. Of the initial 5,564 consecutive women enrolled from February 1991 to June 1995, 4,831 (86.8%) there was information on the outcomes analyzed. They constitute the study sample for these analyses. The data relating to the inclusion criteria were obtained from the pregnant women’s medical records or the antenatal notes. On the occasions when it was the Wrst consultation, the data could be obtained directly from the pregnant woman. Variables analyzed
Material and methods
The interview was structured so as to obtain answers to questions on sociodemographic characteristics and also on tobacco, alcohol and drug use and reproductive history. Variables relating to the current pregnancy, the delivery and the health condition of the neonate were collected from the mother’s medical records. To Wnd out whether the pregnant women had used dipyrone, a general question was asked about the medications used during the pregnancy up to the time of the interview; this was done during the Wrst phase of the EBDG. The products mentioned that contained dipyrone were coded according to a list specially created for the EBDG. The data relating to the outcomes analyzed were collected from the mother’s medical records during phase III of the study. Congenital abnormalities were classiWed in accordance with the ninth international classiWcation of diseases. Intra-uterine death was deWned as the death of a product from the conception prior to its complete expulsion from the mother’s body. The birth of a child with less than 37 weeks of gestation was classiWed as a preterm birth. Newborns weighing less than 2,500 g were classiWed as presenting low birth weight.
Design and sample
Statistical analysis
The present study is a secondary analysis of the Brazilian gestational diabetes study (EBDG), which was a multicenter
Descriptive, bivariate and multivariate analyses were carried out with the aid of the SPSS v.13.0 software. To
123
Arch Gynecol Obstet (2009) 279:293–297
estimate the statistical signiWcance of associations, the Chi-squared test was applied and the results were presented with 95% CI. To estimate the OR after adjustment for the potential confounding factors reported in the literature, logistic regression modeling was developed. In the Wrst stage, the variables that presented statistical signiWcance in the individual analyses (deWned as P < 0.25) were selected. Confounding variables that presented statistical signiWcance in the multivariate model (P < 0.05) were kept in the Wnal model. The research center was incorporated in the model, independent of the signiWcance level, to adjust for the diVerence between pregnant women in the State of São Paulo and other pregnant women, in relation to age and schooling. The original study was approved by the ethics committees at each research center, and women consented to participate after being informed about the nature of the study. The data originating from the questionnaires, which were the subject of this secondary analysis, were treated within the principles of conWdentiality.
Results Dipyrone use was reported by 555 pregnant women (11.5%). Table 1 shows the general characteristics of the 4,831 pregnant women who were followed up until their children were born. Generally, the dipyrone users presented lower age, lower schooling levels, and greater numbers of children, in relation to the non-users (P < 0.001). The dipyrone users presented more frequently a past or present smoking habit and greater alcohol consumption. The diVerence, however, was not statiscally signiWcant. Table 2 shows that no association was found between the use of dipyrone and the congenital abnormalities, both in the unadjusted analysis and after adjustment using logistic regression analysis. The congenital abnormalities in the exposure group were: clubfoot (4), syndactyly (1), polydactyly (1), choanal atresia (1), anencephaly (1), macrocephaly (1), meningomyelocele (1), abducted fourth and Wfth left foot Wngers (1).Table 2 also displays that no association was found between the use of dipyrone and intrauterine death, preterm birth, or low birth weight, both in the unadjusted analyses and after adjustment using logistic regression analysis.
Discussion The present study evaluated the association between exposure to dipyrone during pregnancy and adverse perinatal outcomes, from a sample of 4,831 users of antenatal healthcare services within SUS, in six Brazilian state capitals. No
295 Table 1 Characteristics of the pregnant women, according to reported consumption of dipyrone during pregnancy; EBDG, 1991–1995 (n = 4,831) Used dipyrone (n = 555) n
%
Did not use dipyrone (n = 4,276)a %
n
Research center Porto Alegre
79
14.2
1,009
23.6
Salvador
189
34.1
646
15.1
Fortaleza
103
18.6
923
21.6
93
16.8
380
8.9 21.5
Rio de Janeiro São Paulo
50
9.0
921
Manaus
41
7.4
397
9.3
555
100.0
4,276
100.0
509
91.7
3,703
86.6
Age in years** 20–34 35–48
46
8.3
573
13.4
555
100.0
4,276
100.0
0–4
120
21.7
892
20.9
5–8
240
43.5
1,539
36.1
9–11
165
29.9
1,433
33.6
12–16
27
4.9
403
9.4
552
100.0
4,267
100.0
None
118
24.6
1,179
30.8
1
158
32.9
1,245
32.6
2 or more
204
42.5
1,400
36.6
480
100.0
3,824
100.0
Schooling in years**
Parity**
Smoking Never smoked
311
56.0
2,554
59.7
Former smoker
137
24.7
973
22.8
Current smoker
107
19.3
749
17.5
348
62.7
2,783
65.1
Alcohol*** Not consumed Light consumption Moderate/heavy consumption
61
11.0
526
12.3
146
26.3
967
22.6
555
100.0
4,276
100.0
a
The totals for each variable diVer because of occurrences of non-informed data ** P < 0.001 *** Light consumption = daily consumption of 0.02–0.39 g of alcohol, moderate/heavy consumption = daily consumption of 0.40–85.02 g of alcohol
associations were found between this medication and congenital abnormality, intrauterine death, preterm birth, or low birth weight. The Wndings in the literature regarding the teratogenic potential of dipyrone are inconclusive. Bar-Oz et al. [12] compared 108 pregnant women who were exposed to
123
296
Arch Gynecol Obstet (2009) 279:293–297
Table 2 Crude and adjusted odds ratio (OR) and 95% conWdence interval for congenital abnormality, intrauterine death, preterm and low birth weight, EBDG 1991–1995 Used dipyrone
Did not use dipyrone
%
%
n/total
Crude OR (95% CI)
Adjusted OR (95% CI)
n/total
Congenital abnormality
2.00
11/549
1.80
75/4,150
1.11 (0.58–2.10)
1.04 (0.54–2.00)a
Intrauterine death
1.44
8/556
2.07
88/4,257
0.69 (0.33–1.43)
0.28 (0.07–1.17)b
14.93
83/556
15.80
678/4,293
0.94 (0.73–1.20)
0.77 (0.58–1.02)c
8.07
44/545
9.03
380/4,210
0.88 (0.64–1.22)
0.77 (0.51–1.17)d
Preterm Low birth weight
The diVerences in the sums are due to the lack of data about outcomes that were analyzed a Adjusted through logistic regression for research center b Adjusted through logistic regression for research center, congenital abnormality, low birth weight, and preterm birth c Adjusted through logistic regression for research center, pre-gestational body mass index, number of antenatal consultations, hypertensive disorders, and pre-delivery hemorrhage d Adjusted through logistic regression for research center, pre-gestational body mass index, smoking, number of antenatal consultations, hypertensive disorders, pre-delivery hemorrhage, congenital abnormality, and preterm birth
dipyrone during the Wrst trimester of pregnancy with 108 pregnant women exposed to paracetamol. The rate of congenital abnormalities in the dipyrone group (3%) did not diVer signiWcantly from the rate in the paracetamol group (OR 1.55; 95% CI, 0.26–9.05). However, the interpretation of those results needs to take into consideration the limitations of the study, including the absence of blinding with regard to the exposure and the controlling for potential confounding factors. From a birth record database, Bánhidy et al. [13] investigated associations between dipyrone use in the second and third months of pregnancy and diVerent types of congenital abnormality. Exactly 22,843 cases were compared with two control groups: one group of 38,151 newborns without congenital abnormalities (population-based control) and another group of 834 newborns or fetuses with Down’s syndrome (malformed controls). Positive associations were found between reported dipyrone use and diaphragmatic defects, cardiovascular abnormalities, and various other abnormalities, in comparison with the population-based controls, but not with the malformed controls. When the analysis was restricted to dipyrone use according to the antenatal care logbook for the pregnant women, there was also no association. According to these authors, these diVerences could result from memory bias, since the mothers of children with congenital abnormalities may make more eVorts in searching for a possible explanation for such events, including any use of medications during pregnancy. In this sample of Brazilian pregnant women, there was no diVerence between the groups exposed and not exposed to dipyrone with regard to preterm birth or low birth weight, either in the raw analyses or in the logistic regression models that were applied. This Wnding is in agreement with Bánhidy et al. [13]. No analysis on dipyrone in relation to intrauterine death was found in the literature.
123
Given that the EBDG was planned to meet other research objectives, the present analysis presents certain limitations. It is likely that the use of dipyrone in the sample studied was underestimated, since many women can have obtained this substance that is sold over the counter in Brazil, making it easily obtainable and apparently harmless. Medical records were not reviewed to verify the prescription of dipyrone during the antenatal period. Finally, the question that was made to pregnant women allowed the reported use of any medicine containing dipyrone up to, and at most, the 28th gestational week, the limit for enrollment in the study. No information was obtained on the quantities of dipyrone used or for how long it was used. Also, there was no information on exactly when dipyrone was used. Consequently, it was impossible to analyze the exposure according to gestational period, which would be a particularly important condition for investigating the teratogenic eVect of the medication. However, because this is a medication sold over the counter and used occasionally, we believe that it would be diYcult to obtain this information reliably, considering that it depends on the pregnant woman’s memory. Such use may have occurred at diVerent times during the pregnancy, and on more than one occasion. With regard to the outcome measurements, it must be emphasized that the congenital abnormalities were diagnosed during the Wrst days of life, by the doctor attending the child, without conWrmation by experts. Some abnormalities could thus be overlooked in this initial assessment, and this might also occur in relation to those that tend to be manifested later on. Considering the possible occurrence of both under and over diagnosis, the impact of potential bias in the results is uncertain. The sample investigated included only pregnant women attended by the Brazilian public healthcare system. Although
Arch Gynecol Obstet (2009) 279:293–297
this does not make results invalid, factors such as sociodemographic characteristics and cultural behavior may limit any generalization of the Wndings from the present study in relation to the pregnant women who use the complementary healthcare system. Despite these limitations, the present study presents strong points that must be highlighted. This was the Wrst prospective study to analyze the teratogenic eVect of dipyrone in a Latin American country where its outpatient use is widespread and unrestricted. Around 120 products containing dipyrone as a single drug or in associations of drugs are commercialized in Brazil, and more than 80% of the sales are made without a medical prescription [3]. In the Ministry of Health’s Essential Medications List (RENAME), the pharmaceutical forms of oral solution and injectable solution of 500 mg/ml are standardized: the Wrst of these for unrestricted use and the second for use in hospitals [14]. This study also, for the Wrst time, analyzed the eVect of dipyrone on outcomes with multifactorial etiology, such as a low birth weight and prematurity, controlling for covariables by means of logistic regression, in contrast to previous study carried out in Israel and Italy [12]. It must be emphasized that the present study was unable to evaluate any possible carcinogenic eVect crossing the placenta. Two case-control studies have shown a possible association with childhood tumors: Wilm’s tumor and leukemia [6, 7]. From in vitro and animal studies, however, dipyrone demonstrated weak mutagenic or carcinogenic potential, and only at very high doses [15, 16]. These results from humans may therefore have been due to the mothers’ memory bias or to some other factor relating to dipyrone use that was not investigated. In summary, the results from the present study suggest that dipyrone use during pregnancy does not present any teratogenic risk and is not associated with intrauterine death, preterm birth, or low birth weight. Dipyrone is still a stigmatized medication in various settings because of reports of its relation to agranulocytosis and aplastic anemia. Despite this, its use is well-established in Brazil and other countries, including among pregnant women. This reinforces the need of future investigations to determine the teratogenic eVects of this drug, and to identifying other adverse perinatal events. ConXict of interest statement None of the authors had any conXicts of interest.
297
References 1. The international agranulocytosis aplastic anemia study (1986) Risks of agranulocytosis and aplastic anemia. A Wrst report of their relation to drug use with special reference to analgesics. JAMA 256:1749–1757. doi:10.1001/jama.256.13.1749 2. Andersohn F, Konzen C, Garbe E (2007) Systematic review: agranulocytosis induced by nonchemotherapy drugs. Ann Intern Med 146:657–665 3. International panel on the evaluation of safety of dipyrone use (on line) (2007). Available on 20 April 2007 at http://www.anvisa.gov.br/divulga/informes/relatoriodipirona2.pdf 4. Mengue SS, Shenkel EP, Duncan BB, Schmidt MI (2001) Drug use by pregnant women in six Brazilian cities. Rev Saude Publica 35:415–420. doi:10.1590/S0034-89102001000500002 5. Fonseca MR, Fonseca E, Bergsten-Mendes G (2002) Prevalence of drug use during pregnancy: a pharmacoepidemiology approach. Rev Saude Publica 36:205–212 6. Sharpe CR, Franco EL (1996) Use of dipyrone during pregnancy and risk of Wilm’s tumor. Epidemiology 7:533–535. doi:10.1097/ 00001648-199609000-00014 7. Alexander FE, Patheal SL, Biondi A, Brandalise S, Cabrera ME, Chan Li C et al (2001) Transplacental chemical exposure and risk of infant leukemia with MLL gene fusion. Cancer Res 61:2542–2546 8. Catalán JL, Santonja J, Martinez L, Jiménez NV (1994) Oligoamnios asociado al uso de dipirona. Med Clin (Barc) 105:541–543 9. Sánchez de la Nieta MD, Rivera F, de la Torre M, Alcázar R, Caparrós G, Paz Alcaide M et al (2003) Acute renal failure and oligohydramnios induced by magnesium dipyrone (metamizol) in a pregnant women. Nephrol Dial Transplant 18:1679–1680. doi:10.1093/ndt/gfg240 10. Weintraub A, Mankuta D (2006) Dipyrone-induced oligohydramnios and ductus arteriosus restriction. Isr Med Assoc J 8:722–723 11. Martí Solé JJ, Pasarisas Sala M (1996) Possible association between the administration of metamizol to mothers and persistent pulmonary hypertension in neonates. An Esp Pediatr 44:387–388 12. Bar-Oz B, Clementi M, di Giantonio E (2005) Metamizol (dipyrone, optalgin) in pregnancy, is it safe? A prospective comparative study. Eur J Obstet Gynecol Reprod Biol 119:176–179. doi:10.1016/j.ejogrb.2004.07.004 13. Bánhidy F, Ács N, Puhó E, Czeizel AE (2007) A populationbased case-control teratologic study of oral dipyrone treatment during pregnancy. Drug Saf 30:59–70. doi:10.2165/00002018200730010-00006 14. Brazil. Ministry of Health. Secretary of ScienceTechnology Strategic Products. Department of Pharmaceutical Assistance Strategic Products (2007) National list of essential medicines— RENAME, 4th edn. Ministry of Health Press, Brasília, p 285 15. Izumi K, Sano N, Otsuka H, Kinouchi T, Ohnishi Y (1991) Tumor promoting potential in male F344 rats and mutagenicity in Salmonella typhimurium of dipyrone. Carcinogenesis 12:1221–1225. doi:10.1093/carcin/12.7.1221 16. Giri AK, Mukhopadhyay A (1998) Mutagenicity assay in Salmonella and in vivo sister chromatid exchange in bone marrow cells of mice for four pyrazolone derivatives. Mutat Res 420:15–25
123