Background and aims Few studies have been conducted addressing the safety of thiopurine treatment in pregnant women with inflammatory bowel disease (IBD). The aim of this study was to evaluate the pregnancy outcome of women with IBD who have been exposed to thiopurines.
Methods 215 pregnancies in 204 women were registered and documented in the CESAME cohort between May 2004 and October 2007. Physicians documented the following information from the women: last menstrual date, delivery term, details of pregnancy outcome, prematurity, birth weight and height, congenital abnormalities, medication history during each trimester, smoking history and alcohol ingestion. Data were compared between three groups: women exposed to thiopurines (group A), women receiving a drug other than thiopurines (group B) and women not receiving any medication (group C).
Results Mean age at pregnancy was 28.3 years. 75.7% of the women had Crohn's disease and 21.8% had ulcerative colitis, with a mean disease duration of 6.8 years at inclusion. Of the 215 pregnancies, there were 138 births (142 newborns), and the mean birth weight was 3135 g. There were 86 pregnancies in group A, 84 in group B and 45 in group C. Interrupted pregnancies occurred in 36% of patients enrolled in group A, 33% of patients enrolled in group B, and 40% of patients enrolled in group C; congenital abnormalities arose in 3.6% of group A cases and 7.1% of group B cases. No significant differences were found between the three groups in overall pregnancy outcome.
Conclusions The results obtained from this cohort indicate that thiopurine use during pregnancy is not associated with increased risks, including congenital abnormalities.
- Inflammatory bowel disease
- Crohn's disease
- ulcerative colitis
- ulcerative colitis
Statistics from Altmetric.com
- Inflammatory bowel disease
- Crohn's disease
- ulcerative colitis
- ulcerative colitis
Significance of this study
What is already known about this subject?
Women with IBD often need treatment and seek information on the potential interference of drugs during pregnancy.
The safety of purine analogues has been questioned in pregnant animals receiving high doses parenterally.
Series reporting the fate of pregnancy in women receiving thiopurines for IBD until now have been limited.
What are the new findings?
In a large cohort study in France including 84 pregnancies in patients treated with purine analogues and 129 control patients, there was no increase in the risks of congenital abnormalities (2 vs 4 in the controls).
How might it impact on clinical practice in the foreseeable future?
Based on the present study and on literature in other populations, the use of thiopurines in women with IBD and wishing to become pregnant should not be discouraged. Reassuring information should be provided to pregnant women receiving this treatment.
Thiopurines are used as effective treatments for inflammatory bowel disease (IBD) and are particularly beneficial to avoid long-term use of steroids and to prevent recurrence.1 2 This class of drugs is widely used for treatment of a variety of inflammatory disorders. For example, 30.1% of the patients in the CESAME cohort, which followed 19 486 patients in France over 3 years for cancer risk assessment, were receiving thiopurines, and 14.4% had previously received thiopurines.3 Interrupting thiopurine treatment is associated with a risk of recurrence of steroid-dependent Crohn's disease, which may be as high as 70% per year in the first years of treatment.4 While any drug exposure during pregnancy raises the risk of adverse effects on the fetus, maintaining disease remission for patients with IBD is also critical for positive pregnancy outcomes.5 6 Thus establishing the safety of thiopurine treatment in pregnant patients with IBD is essential and few studies have thoroughly evaluated this issue.7–15 The studies that have been performed showed no increase in birth defects, but were limited by their statistical power. While Norgard et al reported a 9.7% additional risk of congenital abnormalities in patients receiving thiopurines in the Danish National Registry database, this increase was not statistically significant.12 In addition, studies of pregnancy outcome in patients receiving these drugs for other conditions, including renal transplant or systemic lupus erythematosus, were also reassuring.13–16 The daily dose in patients with IBD is 2–2.5 mg/kg body weight for azathioprine and 1.5–2 mg/kg for 6-mercaptopurine (6-MP); these drugs are administered orally and their bioavailability is variable but below 50%.16 Animal studies have shown a risk of congenital defects in the offspring of rodents that received 4–13 times the human therapeutic dose but not in animals receiving doses within the human therapeutic range.16–19 In particular, Platzek et al demonstrated that the ‘no observed adverse effect level’ (NOAEL) for thiopurines is 5 mg/kg.19 Thus current human treatment regimens are within the limit suggested by these animal studies.
Establishing potential risks of drugs during pregnancy requires controls and knowledge of the risks in the general population. Clinical classifications of pregnancy outcomes include live births, elective terminations and fetal losses—that is, spontaneous abortions (loss prior to 20 weeks postconception), and fetal deaths/stillbirths (loss beyond 20 weeks postconception).20 Based on US national data, only 62% of clinically recognised pregnancies result in a live birth, 22% end in elective termination and 16% result in spontaneous abortions (1 in 7 known pregnancies) or fetal death/stillbirth (1 in 200 known pregnancies).20 21 Preterm birth and low birth weight (<2500 g) occur in 1 in 8 and in 1 in 12 live births, respectively.21 Similar data are observed in France.22
The present study aimed to assess the pregnancy and fetal outcome of a large cohort of women with IBD who had been exposed to thiopurines, exposed to other drug treatments or not treated.
Subjects and methods
The present study is a nested study, focusing on pregnancy, within the large cohort CESAME study of 19 486 patients (11 006 women) with IBD in France.3 This database was initiated in 2004 and completed by the end of 2007; patients were included and regularly followed by 680 gastroenterologists with the objective of determining if immunosuppressive treatment for IBD is associated with an increased risk of cancer. In 2005, the members of the group were invited optionally to declare cases of pregnancies and to fill in forms on the CESAME website. The protocol was approved by the Institutional Review Boards of the French National Society of Gastroenterology (SNFGE), Association François Aupetit (AFA—the largest nationwide French IBD patient association), and Groupe d'Etude Thérapeutique des Affections Inflammatoires du Tube Digestif (GETAID). Specifically written, informed patient consent was not required for this observational study.
Information was recorded on website forms by the gastroenterologists belonging to the CESAME panel. Data were collected retrospectively for the year 2005 and prospectively from 1 January 2006 until 1 January 2007 (closure of the database). Details recorded included: a detailed account of IBD diagnosis, current and past medications, smoking history and alcohol consumption, and obstetric history with previous pregnancies and previous abortions (elective or spontaneous). Pregnancy details recorded included: last menstrual date, live births with delivery term, spontaneous abortions, elective abortions (contraceptive and therapeutic), any congenital birth abnormalities, birth weight and height, and any complications during pregnancy or after birth. Pregnancies were also recorded, and prematurity was defined as a gestational age of <37 weeks, while birth weight was defined as low if under 2500 g. Hypotrophy (intrauterine growth restriction) was defined as a birth weight under the 10th percentile and hypertrophy as a birth weight above the 90th percentile of the standard French growth curves of Leroy and Lefort, based on birth weight and gestational term.23
Groups and comparison
Pregnancies were classified into three groups depending on the drug exposure (table 1). Group A: pregnancies in women exposed to thiopurines (alone or associated with another treatment—aminosalicylates, corticosteroids or antitumour necrosis factor α (TNFα). Group B: pregnancies in women receiving a drug other than thiopurines. Group C: pregnancies in women not receivieng any medication. Multiple pregnancies from a given patient might appear in different categories. Concomitant medication was also taken into account in groups A and B, leading to further division into subgroups. Group A was divided into thiopurines only, thiopurines associated with aminosalicylates, thiopurines associated with corticosteroids and thiopurines associated with anti-TNFα. Group B included subgroups for patients receiving aminosalicylates, those receiving corticosteroids, those receiving anti-TNFα and two patients who received methotrexate.
Calculations were performed using the Stata 10 (StataCorp LP, College Station, Texas, USA) software. Descriptive statistics were calculated for all variables. Quantitative (continuous) data included mean, SD and range. For qualitative (binary) data, frequencies with 95% CIs were calculated. To identify pregnancy outcome variables associated with drug exposure, we performed two-by-two comparisons between groups A, B and C. Quantitative variables were compared with the Student t test or, when this was not applicable, the Wilcoxon signed rank test. Qualitative variables were compared with the χ2 test or, when this was not applicable, the Fisher exact test. All statistical tests were two-tailed and statistical significance was considered at p<0.05 for all comparisons.
Seventy-five gastroenterologists of the CESAME group reported 215 pregnancies in 204 women (1.79% of the women followed in the CESAME cohort). The subset of CESAME investigators who participated in the pregnancy study reported all pregnancies that occurred in their personal patient file. The characteristics of pregnancies are shown in table 1. The majority (87%) of cases were reported in the prospective part of the study (2006 and 2007); 28 pregnancies were reported retrospectively for 2005 (10, 12 and 6 for groups A, B and C, respectively). There was no trend for differences in the outcome data obtained in the two parts of the study (data not shown), and the data provided pools of the retrospective and prospective periods. In group A (81 women), 86 pregnancies resulted in 55 live births. Thiopurine was the only treatment in 57 cases and was used in combination with other drugs in the remaining cases. In group B (80 women), 84 pregnancies resulted in 56 live births. In group C (43 women), 45 pregnancies resulted in 27 live births. The details of the three groups, according to their IBD diagnosis, are shown in table 2.
The results of pregnancy and fetal outcomes are shown in table 3. Pregnancy failures (spontaneous or elective abortions) were as follows: group A, 31 (36%); group B, 28 (33%); and group C, 18 (40%), with no significant difference between the groups. Results obtained in subgroups according to co-medication are shown in table 4.
The two pregnancies occurring in patients receiving methotrexate were interrupted. In addition, therapeutic abortions were performed in a woman with active ulcerative colitis who was concerned about the possible effects of the pregnancy on her digestive disease and in two cases after diagnosis of fetal illness (one lethal complex cardiac defect and one umbilical arterial thrombosis). The indication of the other elective abortions was not specified.
There were no significant differences in the risks of prematurity, birth weight and hypotrophy between the three groups (table 3). Nineteen cases of low birth weight, nine cases of hypotrophy according to the gestational age and six cases of hypertrophy were reported.
Congenital abnormalities, including one case with a complex cardiac defect leading to a therapeutic abortion, were observed in two cases of group A and four cases of group B (differences non-significant (NS), tables 3 and 4). Details of these cases are shown in table 5.
Eleven cases of complications occurring during pregnancy or after birth were reported: three in group A (5.4% of live births), five in group B (8.9% of live births) and three in group C (11.1% of live births) (difference NS, details are shown in table 6).
In this series, the outcome of pregnancy (live births, congenital anomalies, prematurity or hypotrophy) in patients with IBD did not differ between those exposed to thiopurines during pregnancy, those exposed to other drugs during pregnancy or those exposed to no drugs at all. The rates of live births and birth defects were in the range of the normal population; however, those of prematurity and hypotrophy were notably high.
Establishing the safety of drugs during pregnancy is difficult, and epidemiological studies are immensely important. Conclusions drawn from these types of studies must be interpreted with caution. The present study has several limitations that should be noted. We cannot exclude for instance that some minor side effects were not declared, particularly in the retrospective part of the study which represented 13% of the series. In addition, details regarding the activity of IBD at conception and throughout pregnancy, as well as the doses of the various treatments, were not recorded. Notably, disease activity may particularly increase the risk of prematurity.5 6 Moreover, details on congenital anomalies and pregnancy complications were provided by the patients and not officially reported by physicians. Although the number of patients in this study is high in comparison with previous cohorts, the sample size is still limited to detect low risks with satisfactory statistical power. Thus in our study, considering an α risk of 0.05 (two-sided) and a power of 85%, a sample size of 85 women in each comparison group allowed the detection of a fivefold increase (from 4% to 20%) in malformations.
The opportunity to compare the risks of pregnancy between groups of patients suffering from the same disease but receiving different treatment regimens is advantageous. Most patients in group A were exposed to thiopurines throughout the entire pregnancy as their doctors considered that cessation of treatment could have led to a relapse, deleterious for both the mother and the fetus.4 5 No observable trend emerged between the three groups, except for low birth weight and prematurity.
The data collected in this study were in the range of the general population concerning the rate of live births (64% in our study vs 65.9% of clinically recognised pregnancies in the USA20 21) and congenital anomalies (3.6% in our series vs 3.2% in France22). Previous series of patients with IBD did not show any significant increase of congenital anomalies with thiopurines,5–16 excluding one series in which only 11 women exposed to thiopurines were evaluated.10 When looking at the various malformations reported in detail by us and others, there appears to be no pattern of specific or a spectrum of birth defects.16 Cleary and Källen recently used the Swedish Medical Birth Register to identify 476 women exposed to azathioprine in early pregnancy and compared their infants with 1 163 554 controls; the risks of congenital malformations were not significantly increased (6.2% vs 5.2% among the IBD control group).15 In addition, the risk of cardiac malformations was not increased; however, the risk of specific ventricular and atrial septal defects was higher (nine cases; ie, an OR of 3.18). One such case was described by Francella et al,9 but only two other cases in the literature were identified by Polifka and Friedman in their review of 27 series.16 One of our cases, which led to a therapeutic abortion, was also a complex cardiac malformation but the mother was not taking thiopurines. In a recent international study, Goldstein et al14 compared 189 pregnant women on azathioprine who contacted one of seven teratogen information services in Israel, Canada and Europe, with a cohort of 230 pregnant women who contacted two of the services and took non-teratogenic treatments during their pregnancy. The rate of major malformations did not differ between groups: 3.5% in the azathioprine group versus 3.0% in the control group (OR 1.17; 95% CI 0.37 to 3.69; p=0.775).
Our data also confirm previous studies showing higher risks of prematurity (18%) and low birth weight (13.8%) in patients with IBD when compared with the general population (12.5% and 8%, respectively, in the USA).21 24 25 In the recent study by Goldstein et al,14 the mean birth weight was lower in the thiopurines group (2995 g vs 3252 g; p=0.001) and the risk of prematurity was increased (21.4% vs 5.2%; p<0.001), as was the risk of low birth weight (23% vs 6.0%; p<0.001). However, the disease activity which led to the prescription of thiopurines is most probably the cause of these differences. No other studies showed that thiopurine usage increases the risk of pregnancy outcomes in comparison with women with the same disease but not taking thiopurines. In our series, nine babies were hypotrophic according to their gestational term.23 As a patient may often receive combinatorial treatments, we analysed the outcome of pregnancies in subgroups and particularly focused upon those receiving anti-TNFα treatments. Overall, only one congenital abnormality was reported in the thiopurine/anti-TNFα group, but a very low rate of live birth was observed. The patients in our study were pregnant in the years 2005–2007 and it seems likely that the limited knowledge on the safety of anti-TNFα agents at that time contributed to the decision of the patients to interrupt pregnancy. Our data should therefore not be interpreted as suggesting deleterious effects of these drugs on pregnancy outcome. Both the European Crohn's and Colitis Organization and US Food and Drug Administration categorise anti-TNFα agents as safe during pregnancy, at least when used during the first 6 months; however, future studies need to be initiated in order to strengthen this suggestion.26
We observed four congenital abnormalities in the patients from group B. This did not differ from the results of the other groups. All the mothers had received 5-aminosalicylic acid (5-ASA); however, this drug does not increase the risk according to the literature, as confirmed by a recent meta-analysis.27
Infantile immunosuppression has been reported when the mother had received thiopurines during pregnancy.28 In our series, and in that of Francella et al, there was no statistical difference in the risk of infections between patients receiving or not receiving thiopurines,9 and no indication of a frequent risk.
Halting thiopurine treatment before pregnancy is an option in patients with a low risk of IBD recurrence on a short-term basis. Zlatanic et al assessed the complications of pregnancy and child development when mothers had stopped 6-MP before conception (72 pregnancies in 29 patients) compared with a historical group of 75 women with IBD who had not received 6-MP and had 140 pregnancies. There was greater fetal loss in the 6-MP cessation group (29.2% vs 14.3%) and two chromosomal aberrations (trisomy 22 and mosaicism).29
Based on the present study and others in the literature, the use of thiopurines in women with IBD who wish to become pregnant should not be discouraged. As measuring low risks necessitates a high statistical power, further epidemiological studies and meta-analyses, especially on cardiac septal defects, should be performed.
Clinicians who participated in the study and reported pregnancies (members of the CESAME pregnancy study group): J F Colombel, L Beaugerie, P Marteau, X Hébuterne, E Lerebours, P Baumer, J Cosnes, M Lemann, A Bourreille, J P Gendre, P Seksik, A Blain, E H Metman, A Nisard, G Cadiot, M Veyrac, B Coffin, C Boehm, M Brun, A Cortot, R Fournier, J L Gerbal, M Kaassis, B Le Gall, B Mesnard, C Michiels, S Nahon, X Roblin, L Roget, S Schneider, A M Weiss, C Wittersheim, V Abitbol, P Afchain, D Barbereau, J Bellanger, P Benfredj, A Blanquart, M Blazquez, F Boiffin, P Bord, Y Bouhnik, A Charachon, D Chaslin-Ferbus, E Cuillerier, B David, Y Emery, O Empinet, P Estable, I Etienney, O Fourdan, D Goldfain, H Grimpel, P Houcke, F Jean, R Jian, G Hecquet, I Joly-Le-Floch, B Jost, M Lagneau, P Lamy, P Lebourgeois, O Leroy, N Mallier-Gehrke, P Mamou, D Martin, J F Mockly, X Moreau, B Pornin, P Rey, P Rogier, J Sanchez, J Seroka, J Silvie, D Sondag, R Tenenbaum.
Funding Supported by grants from the Programme Hospitalier de Recherche Clinique National (PHRC) (AOM05157), Association François Aupetit (AFA), Délégation Inter-régionale de la Recherche clinique Ile de France-Assistance Publique Hôpitaux de Paris (AP-HP), Ligue contre le Cancer and Fonds de Recherche de la Société Nationale Française de Gastroentérologie (SNFGE).
Competing interests None.
Ethics approval This study was conducted with the approval of the Institutional Review Boards of the French National Society of Gastroenterology (SNFGE), Association François Aupetit (AFA—the largest nationwide French IBD patient association) and Groupe d'Etude Thérapeutique des Affections Inflammatoires du Tube Digestif (GETAID).
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.