ORIGINAL ARTICLE

Systematic Review and Meta-analysis on the Effects of

Thiopurines on Birth Outcomes from Female and Male
Patients with Inflammatory Bowel Disease
Mona Akbari, MD, MPH,* Sveta Shah, MD,* Fernando S. Velayos, MD, MPH,† Uma Mahadevan, MD,†
and Adam S. Cheifetz, MD‡

Background: Inflammatory bowel disease (IBD) affects people during their prime reproductive years. The thiopurines (6-mercaptopurine and
azathioprine), commonly used for induction and maintenance of remission, are U.S. Food and Drug Administration (FDA) pregnancy category
D, raising concern for fetal risk. We performed a systematic review and meta-analysis to evaluate the effects of thiopurine exposure during preg-
nancy or at the time of conception on three measures of fetal risk in women and men with IBD.
Methods: A systematic search of PubMed and Web of Science using a combination of Mesh and text terms was performed to identify studies
reporting birth outcomes from IBD women and men exposed to thiopurines within 3 months of conception and/or during pregnancy. A meta-
analysis was performed using the random effects model to pool estimates and report odds ratio (OR) for three outcomes in women: low birth
weight (LBW), preterm birth, and congenital abnormalities and one in men: congenital abnormalities.
Results: In women with IBD exposed to thiopurines, the pooled ORs for LBW, preterm birth, and congenital abnormalities were 1.01 (95%
confidence interval [CI] 0.96, 1.06), 1.67 (95% CI 1.26, 2.20), and 1.45 (95% CI 0.99, 2.13), respectively. In men, the pooled OR for congenital
abnormality was 1.87 (95% CI 0.67, 5.25).
Conclusions: Thiopurine exposure in women with IBD was not associated with LBW or congenital abnormalities, but was associated with pre-
term birth. Exposure in men at the time of conception was not associated with congenital abnormalities.
(Inflamm Bowel Dis 2012;000:000–000)
Key Words: thiopurines, mercaptopurine, birth outcomes, congenital abnormalities, inflammatory bowel disease

I nflammatory bowel disease (IBD) most commonly

presents in young adults, during the childbearing years.
Thiopurines, 6-mercaptopurine (6-MP) and its prodrug, aza-
thioprine (AZA), are commonly used to treat both Crohn’s
disease (CD) and ulcerative colitis (UC). 6-MP and AZA
are effective steroid-sparing agents and have been demon-
strated to induce and maintain remission in patients with
IBD.1–9
Additional Supporting Information may be found in the online version of this
article.
Received for publication February 20, 2012; Accepted February 22, 2012.
From the *Department of Internal Medicine, Beth Israel Deaconess Medical
Center, Boston, Massachusetts, †Division of Gastroenterology, Center for Colitis
and Crohn’s Disease, University of California, San Francisco, California,
‡
Division of Gastroenterology, Center for Inflammatory Bowel Disease, Beth
Israel Deaconess Medical Center, Boston, Massachusetts.
Reprints: Adam S. Cheifetz, Center for IBD, Beth Israel Deaconess Medical
Center, 330 Brookline Ave., Rabb 425, Boston, MA 02215 (e-mail: acheifet@
bidmc.harvard.edu).
Copyright VC 2012 Crohn’s & Colitis Foundation of America, Inc.
DOI 10.1002/ibd.22948
Published online in Wiley Online Library (wileyonlinelibrary.
com).
The U.S. Food and Drug Administration (FDA) clas-
sifies 6-MP and AZA as category D medications, indicating
that that while there may be fetal risk, the benefits from
use in pregnancy may be acceptable. Patients and physi-
cians struggle with the question of whether to continue thi-
opurines during this time period. Patients report fear of
continuing medical therapy of their IBD during pregnancy,
in part due to fear of harm to the fetus.10,11 Physicians
struggle to balance potential reported risks to the fetus12–18
with the known benefits of therapy.
The risks to the fetus associated with thiopurine use
around the time of conception and during pregnancy are
not well characterized, and the literature is sparse. Studies
are often small and outcomes rare, suggesting larger studies
with sufficient sample size are needed and a systematic
review of available studies may help to identify consistent
trends. Prior meta-analysis has reported on birth outcomes
for women with IBD compared to the general population.19
To our knowledge, there are no meta-analyses to date that
compare birth outcomes for IBD patients exposed and not
exposed to thiopurines at conception and/or during preg-
nancy. Moreover, there is no pooled analysis that has

Inflamm Bowel Dis 1

Akbari et al
looked at thiopurine effects on men with IBD. The objective
of this study was to conduct a systematic review and meta-
analysis to determine whether thiopurine exposure in women
or men with IBD at the time of conception and/or during
pregnancy is associated with adverse birth outcomes.
MATERIALS AND METHODS
Search Strategy
Two independent reviewers (M.A., S.S.) completed an
online systematic search, from the institutional library, using
PubMed (National Center for Biotechnology Information) and
Web of Science (Web of Knowledge), to identify all articles
published between 1960 and 2011 and written in English on
thiopurine use and birth outcomes in patients with CD or UC.
The following combination of Mesh and text terms was used
in the search: Inflammatory Bowel Disease; Ulcerative Colitis;
Crohn’s Disease; thiopurine; azathioprine; 6 mercaptopurine;
premature birth; infant, low birth weight; congenital abnormal-
ities; and pregnancy. Studies that were not available online
were retrieved by hand searches at the institutional library and
references from the bibliographies of related articles were
crosschecked to find additional articles. All searches were
completed by September 20 2011.
Inclusion and Exclusion Criteria
Studies had to meet the following inclusion criteria: 1)
study population included patients with a diagnosis of IBD,
CD, and/or UC; 2) outcomes for thiopurine-exposed patients
were compared to a reference group unexposed to thiopurines
around the time of pregnancy; 3) exposed patients received
AZA or 6-MP at the time of conception (defined as within 3
months of conception or last menstrual period) and/or for any
duration during pregnancy; and 4) studies reporting birth out-
comes for either low birth weight (LBW), preterm birth, and/
or congenital abnormalities.
Studies that did not report outcomes for an IBD-only
population or in which it was not possible to extract outcome
data from published results were excluded. Case reports,
series, review articles, and letters were also excluded.
Data Extraction
Information retrieved independently from each study, by
two authors (M.A., S.S.), included: first author’s name, publi-
cation date, study design, covariates used for matching or mul-
tivariate analysis, method of IBD diagnosis, characteristics
and number of the thiopurine exposed and unexposed patients,
exposure assessment, and birth outcomes (LBW, preterm birth,
and congenital abnormalities).
Quality Assessment
We assessed study quality using the Newcastle Ottawa
Scale (NOS).20,21 The NOS assesses studies on three broad cat-
egories: selection of study groups, comparability of groups, and
2
Inflamm Bowel Dis
determination of exposure (case–control studies) or outcome
(cohort studies). Studies are assigned points for various ques-
tions in each category and can have a maximum of nine points.
We considered greater than seven points as high quality.19
Statistical Analysis
In conducting and reporting our analysis, we followed
the consensus guidelines by the Meta-analysis of Observatio-
nal Studies in Epidemiology group22 and the Preferred Report-
ing Items for Systematic Reviews and Meta-Analyses.23 Birth
outcomes were reported as odds ratios (ORs), which represent
the odds of adverse outcomes for births exposed to thiopurines
compared to controls. Pooled ORs were calculated using the
random effects model. The random effects model was used to
account for variations between studies and give a more con-
servative pooled estimate.24
When possible, maximally adjusted ORs for pregnancy
outcomes of individual studies were pooled. For studies that
did not report ORs, 2
2 contingency tables were constructed
based on reported outcomes and unadjusted ORs were calcu-
lated using Woolf’s method.25
The Q test was used to assess for heterogeneity and I2
statistic to quantify the percentage of heterogeneity due to
between-study variation.26 Funnel plots and the Egger’s test
were used to evaluate for publication bias.27 Given the limita-
tions of pooling too few studies, we performed sensitivity
analyses only when three or more studies were available.
Therefore, if only two studies provided results of interest, these
results were reported descriptively rather than a pooled analysis.
We sought to characterize effects of disease activity as well as
study quality. Moreover, we performed two sensitivity analyses
to test the robustness of our results. The first sensitivity analysis
removed one study at a time to see if any one study was driv-
ing the results. The second sensitivity analysis included only
larger studies with sample size (n) greater than 100 participants.
Significance levels were set at a P < 0.05. All statistical analy-
ses were performed using Stata (College Station, TX).
RESULTS
Studies Retrieved
Electronic and manual searches yielded 314 citations,
of which 285 were eliminated on review of titles, abstracts,
and studies (Fig. 1). Of the remaining 29 citations, 22 stud-
ies were excluded because they did not distinguish between
thiopurines and other immunomodulators, such as antitumor
necrosis factor a, there was incomplete information on birth
outcomes reported, such that ORs for LBW, preterm birth,
and congenital abnormalities could not be calculated,10,28–36
there was inadequate control group,37–42 IBD-specific birth
outcomes could not be determined,43,44 or they evaluated
thiopurine use during breastfeeding45 or prior to 3 months of
conception.46 Three studies had overlapping popula-
tions12,47,48; thus, the most recent study was used.47 A total
Inflamm Bowel Dis
FIGURE 1. Search and study selection process.
of seven studies met inclusion criteria; five measured thio-
purine exposure in women47,49–52 and three in men.49,53,54
Study Quality and Characteristics
A total of 3045 women and 217 men with IBD were
identified as having eligible pregnancies. A detailed list of
study quality and characteristics is shown in Table 1. Four
studies described pregnancy outcomes in a population of CD
and UC females49–52 and one in a CD-only female popula-
tion.47 Three studies were performed in a population of men
with CD and UC.49,53,54 Studies identified IBD patients
and treatment exposure through medical records,49,52–54
national registries,47,51 nationwide prescription databases,47
interviews,49,51,53 and questionnaires.50,53 Birth outcomes
were identified by medical records,52,53 international classifi-
cation of disease codes from national registries,47,51 inter-
views,49,53,54 and questionnaires.50,53 The exposure included:
AZA or 6-MP,47,52,53 AZA only,51 6-MP only,49,54 and ‘‘thi-
opurines.’’50 All but one study50 were retrospective case–
control studies by design.
Effects of Thiopurines on Birth Outcomes
Confounding was accounted for by matching52 and/or
multivariate analysis.47,51,52 Two of the five studies in
women attempted to control for disease activity using hospi-
talization as a surrogate marker.47,52 One of the three studies
in men excluded those with female partners who were
treated with immunomodulators for any indication,53 and
another study indicated that no female partners reported ill-
nesses or exposures to toxins during pregnancy.54 In estimat-
ing a pooled OR, we calculated unadjusted ORs from two
studies that reported events in cases and controls.49,50
Low Birth Weight
Five studies reported on the incidence of LBW for
pregnancies from mothers with IBD. LBW was defined as
less than 2500 g. Pooled analysis for the association
between thiopurine use and LBW was not statistically sig-
nificant (Fig. 2; OR 1.01, 95% confidence interval [CI]
0.96, 1.06; P ¼ 0.831). In the two studies that accounted
for disease activity, the association for LBW was not statis-
tically significant.47,52
Preterm Birth
Five studies reported on the incidence of preterm
birth, defined as gestational age less than 37 weeks, for
pregnancies from mothers with IBD. A statistically signifi-
cant difference was observed in pooled analysis for women
taking thiopurines (Fig. 3; OR 1.67; 95% CI 1.26, 2.20;
P < 0.001). In the two studies that accounted for disease
activity, the association between thiopurines and preterm
birth was not statistically significant in one, OR 2.28 (95%
CI 0.67, 7.73),52 but was significant in the other, OR 4.2
(95% CI 1.4, 12.5).47
Congenital Abnormalities
Pooled analysis for five studies that reported on
maternal thiopurine exposure and congenital defects was
not statistically significant (Fig. 4; OR 1.45; 95% CI 0.99,
2.13; P ¼ 0.055). The reported congenital defects for
infants born to mothers on thiopurines were: hyperplastic
heart (n ¼ 1),52 ventricular septal defect (n ¼ 2),51 atrial
septal defect (n ¼ 2),51 hydrocephalus (n ¼ 1),49 cataract
(n ¼ 2),47,50 cervical angioma (n ¼ 1),50 encephalocele
(n ¼ 1),47 sternocleidomastoid malformation (n ¼ 1),47
congenitae aliae cutis (n ¼ 1),47 and others (n ¼ 16). The
association between thiopurines and congenital defects
were not significant in the two studies that accounted for
disease activity.47,52
Three studies reported on congenital defects in
infants born to fathers on thiopurines. Pooled analysis for
congenital anomalies was not statistically significant
(Fig. 5; OR 1.87; 95% CI 0.67, 5.25; P ¼ 0.236). The con-
genital abnormalities reported for infants born to fathers
taking thiopurines were: interauricular communication with
3
Akbari et al Inflamm Bowel Dis

TABLE 1. Study Characteristics

Maternal Exposure to Thiopurines
Pregnancies Covariates for
Publication Study Patients with UC CD Exposed to Multivariate Analysis Study
First Author Year Design IBD (n) Patients (n) Patients (n) AZA/6-MP (n) or Matching Quality

Francella 2003 R 79 24 55 39 1, 2a ***

Norgard 2007 R 648b — 648 20 2-5 *******
Cleary 2009 R 2051 NS NS 324 2, 5-8 *****
Coelho 2010 P, R 204d 44 155 86 None ****
Shim 2011 R 63 NS NS 19 2, 4, 5, 7, 9-16a *****
Paternal Exposure to Thiopurines
Francella 2003 R 76 27 49 37 1, 2a ***
Rajapakse 2000 R 57 23 34 13 None ****
Teruel 2010 R 84 NS NS 46 7, 11, 17-18a ****
a
Studies do not provide adjusted ORs for outcomes: LBW, preterm, and/or congenital anomalies.
b
Total number of births.
c
Number of women with IBD exposed to thiopurines only.
d
Five patients were ‘‘IBD unclassified.’’
NS: Not specified; R: Retrospective; P: Prospective.
Study quality is derived from the Newcastle Ottawa scale and takes on values of one star (lowest quality) to nine stars (highest quality).
Covariates for matching or multivariate analysis: 1, Gender of affected parent; 2, Mother’s age; 3 Disease activity; 4, Gestational age; 5, Parity; 6, Body
mass index; 7, Smoking; 8, Year of birth; 9, Mode of delivery; 10, Corticosteroids; 11, 5-ASA; 12, Hospitalization; 13, Low birth weight; 14, Preterm
birth; 15, Neonatal adverse outcomes; 16, Congenital anomalies; 17, Father’s age; 18, Sulfasalazine.

persistent ductus arteriosus (n ¼ 1),53 missing thumb (n ¼
1),54 multiple anomalies (n ¼ 1),54 and meningomyelocele
(n ¼ 1).49
Sensitivity Analysis
The first sensitivity analysis, in which one study was
removed at a time, did not suggest that any one study was
driving the pooled ORs for LBW, preterm birth, and con-
genital abnormalities (results not shown). The second sensi-
tivity analysis, which analyzed three studies with greater
than 100 female patients with IBD, showed results that
were consistent with the overall pooled results. Thiopurine
exposure was not associated with LBW (OR 1.29, 95% CI
0.93, 1.79; P ¼ 0.129) or congenital abnormalities (OR

FIGURE 2. Forest plot for maternal thiopurine use and LBW, includ-
ing all studies. The OR for LBW and maternal thiopurine use, 95%
CIs for each study, and pooled analysis from a random-effects
model are depicted on a logarithmic scale. The lateral points of
the diamond provide the confidence interval for the combined
estimate.
FIGURE 3. Forest plot for maternal thiopurine use and preterm
birth, including all studies. The OR for LBW and maternal thiopur-
ine use, 95% CIs for each study, and pooled analysis from a ran-
dom-effects model are depicted on a logarithmic scale. The lateral
points of the diamond provide the confidence interval for the com-
bined estimate.

4
Inflamm Bowel Dis
FIGURE 4. Forest plot for maternal thiopurine use and congenital
anomalies, including all studies. The OR for LBW and maternal thio-
purine use, 95% CIs for each study, and pooled analysis from a ran-
dom-effects model are depicted on a logarithmic scale. The lateral
points of the diamond provide the confidence interval for the com-
bined estimate.
1.48, 95% CI 0.85, 2.60, P ¼ 0.170), but was associated
with preterm birth (OR 1.72, 95% CI 1.04, 2.84, P ¼
0.035). Sensitivity analyses on studies accounting for dis-
ease activity as well as study quality were not performed,
as there were too few studies in each group.
Publication Bias and Heterogeneity
Funnel plots are shown for overall analyses in Sup-
porting Figures 1–3. Studies remained within the 95% CI
limits. While there was some potential asymmetry for LBW
graphically visualized the Egger’s test for asymmetry was
not significant (P ¼ 0.929). The funnel plots for preterm
birth (P ¼ 0.326) and congenital anomalies (P ¼ 0.933) did
not show evidence of asymmetry and publication bias. There
was no evidence of heterogeneity amongst studies.
DISCUSSION
This study provides a unique review of the literature,
as well as estimates across studies to assess birth outcomes
from women and men with IBD on AZA or 6-MP. Our
results suggest thiopurine exposure at the time of conception
and/or during pregnancy in women and men with IBD are
not associated with increased infant congenital anomalies.
Maternal thiopurine exposure was not associated with LBW,
but was associated with increased odds of preterm births.
Thiopurine use in women was associated with approximately
a 70% increased odds of preterm births. Our findings
remained consistent in subsequent sensitivity analyses.
In our analysis, preterm birth was the only outcome
that was significantly associated with thiopurines. Whether
this association is secondary to the use of thiopurines or
that thiopurines are a marker for more severe disease is
Effects of Thiopurines on Birth Outcomes
unclear. Disease activity has been associated with negative
fetal outcomes in a number of reports.55–59 One study dem-
onstrated that 27% of births from IBD females with disease
flares were preterm, compared to 8% of births from women
without flares, P ¼ 0.02.55 Similarly, in another study pre-
term birth occurred more frequently in IBD women hospi-
talized for disease flare (70%) compared to those not hospi-
talized (7.3%), P < 0.0001.59
It is likely that mothers treated with thiopurines lead-
ing up to and at the time of conception had more severe or
active IBD than the patients from control groups. In the
study by Shim et al,52 31.6% thiopurine-exposed births and
12.2% unexposed births were from women hospitalized for
IBD flare (P < 0.001), indicating to a degree that patients
on thiopurines had more flares. In this study, 50% of
patients on thiopurines were treated with adjunctive corti-
costeroids, but only a third of them were hospitalized for
flares. Therefore, it is possible that mild to moderate dis-
ease flares managed in an outpatient setting would not be
captured by the use of ‘‘hospitalization’’ as a surrogate for
disease activity and thus underestimate disease activity. In
our meta-analysis, only two studies accounted for disease
activity and did so using hospital admissions during preg-
nancy as a surrogate. Since only two studies accounted for
disease activity, we did not provide pooled results.
Although one study still demonstrated that thiopurine expo-
sure increased the risk of preterm birth,47 the other did not
find a significant effect.52 Both studies still may be under-
estimating disease activity.
Meta-analyses can improve the precision of effect
estimate and increase the power to see a difference that is
not readily identified by small studies.60 Of the seven
FIGURE 5. Forest plot for paternal thiopurine use and congenital
anomalies. The OR for LBW and maternal thiopurine use, 95% CIs
for each study, and pooled analysis from a random-effects model
are depicted on a logarithmic scale. The lateral points of the dia-
mond provide the confidence interval for the combined estimate.
5
Akbari et al
studies included in this analysis, only three had more than
100 subjects with IBD. As such, we selected meta-analysis
as the approach to study the question of birth outcomes
and thiopurine use. We included studies in which the
exposed and comparison groups comprised of births from
only an IBD population, since IBD itself can increase the
risk of adverse birth outcomes, including LBW and preterm
birth.19,61–68 We chose to analyze only IBD cases since the
question of interest is whether to stop or continue the thio-
purine, not whether having IBD confers a higher risk of
adverse birth outcomes.
Based on two sensitivity analyses, our results are not
changed by undue influence of any one study alone or by
larger studies (n > 100). It is important to highlight the rel-
atively small number of pregnancies exposed to thiopurines
in each study. The study by Cleary and Kallen51 reported a
disproportionately larger number of pregnancies that were
exposed to thiopurines. However, our sensitivity analysis
does not suggest that the findings from the Cleary and
Kallen study significantly changed the results. Moreover, in
an effort to detect publication bias the Egger’s test did not
demonstrate significance. We also did not find significant
heterogeneity that would limit interpretation of a pooled
analysis.
6-MP and AZA are classified as category D, medica-
tions that have potential fetal risk but benefits from use in
pregnancy may be acceptable. Under consideration is a
proposal to remove the general global letter grading system
that summarizes risk vs. benefit in favor of a more trans-
parent narrative that includes a risk summary, clinical con-
siderations, and a data section. We have provided an over-
all pooled estimate of risk for the three outcomes of
interest as well as detailed information on congenital
anomalies that could prove helpful in discussing the use of
6-MP and AZA during pregnancy.
Known limitations of meta-analyses include quality
of the original studies. A number of studies relied on inter-
views, questionnaires, and/or self-report for pregnancy out-
comes49,50 and thiopurine exposure50,54 without mention of
reference to primary records. These may not be the most
robust tools for exposure and outcome assessment and have
the potential to introduce misclassification and subsequent
bias due to misclassification. If present, however, it is diffi-
cult to estimate the direction of the bias. Similarly,
response rates and missing data were infrequently reported.
Another limitation is that all studies were observational
and all but one50 were retrospective in design. However,
conducting a randomized trial to evaluate the effects of
medications for IBD during pregnancy is not ethical, and
we are left with only observational studies to detect preg-
nancy outcomes. Additionally, although we attempted to
pool only adjusted ORs, some studies did not provide
adjusted ORs for birth outcomes of interest.49,50,53,54
6
Inflamm Bowel Dis
Although multivariate analysis can be difficult when small
numbers exist in each category, confounders such as con-
current medication therapy and IBD disease activity are
important to consider when evaluating birth outcomes.
Moreover, none of the studies provided adequate informa-
tion on medication duration or dose per day to determine a
dose–response relationship between thiopurines and birth
outcomes. Studies also did not provide outcomes for thio-
purine exposure by trimester, to assess risks to fetus during
different stages of fetal development. Future, larger pro-
spective studies that adjust for confounders such as disease
activity are needed to evaluate these questions.
CONCLUSIONS
In summary, we found that thiopurine use at the time
of conception and/or during pregnancy in women with IBD
is not associated with LBW or congenital abnormalities,
but is associated with preterm birth. Men with IBD
exposed to thiopurines at the time of conception are not at
increased risk for infants with congenital abnormalities.
Larger, prospective studies that adequately identify expo-
sure and outcomes as well as appropriately adjust for
potential confounders, such as disease activity, are needed
to confirm these findings.
ACKNOWLEDGMENTS
Authors contributions: M.A.: study design, drafting
of the article, analysis and interpretation of data; critical re-
vision of the article for important intellectual content of
the article; S.S.: analysis and interpretation of data; critical
revision of the article for important intellectual content of
the article; U.M.: interpretation of data; critical revision of
the article for important intellectual content of the article;
F.S.V.: interpretation of data; critical revision of the article
for important intellectual content of the article; A.S.C.:
study concept and design, acquisition of data, analysis and
interpretation of data; critical revision of the article for
important intellectual content of the article.
REFERENCES
1. Prefontaine E, Macdonald JK, Sutherland LR. Azathioprine or 6-mer-
captopurine for induction of remission in Crohn’s disease. Cochrane
Database Syst Rev. 2010:CD000545.
2. Ardizzone S, Maconi G, Russo A, et al. Randomised controlled trial
of azathioprine and 5-aminosalicylic acid for treatment of steroid
dependent ulcerative colitis. Gut. 2006;55:47–53.
3. Candy S, Wright J, Gerber M, et al. A controlled double blind study
of azathioprine in the management of Crohn’s disease. Gut. 1995;37:
674–678.
4. Hawthorne AB, Logan RF, Hawkey CJ, et al. Randomised controlled trial
of azathioprine withdrawal in ulcerative colitis. BMJ. 1992;305:20–22.
5. Kirk AP, Lennard-Jones JE. Controlled trial of azathioprine in chronic
ulcerative colitis. Br Med J (Clin Res Ed). 1982;284:1291–1292.
6. Lemann M, Mary JY, Colombel JF, et al. A randomized, double-blind,
controlled withdrawal trial in Crohn’s disease patients in long-term
remission on azathioprine. Gastroenterology. 2005;128:1812–1818.
Inflamm Bowel Dis
7. Paoluzi OA, Pica R, Marcheggiano A, et al. Azathioprine or metho-
trexate in the treatment of patients with steroid-dependent or steroid-
resistant ulcerative colitis: results of an open-label study on efficacy
and tolerability in inducing and maintaining remission. Aliment Phar-
macol Ther. 2002;16:1751–1759.
8. Vilien M, Dahlerup JF, Munck LK, et al. Randomized controlled aza-
thioprine withdrawal after more than two years treatment in Crohn’s
disease: increased relapse rate the following year. Aliment Pharmacol
Ther. 2004;19:1147–1152.
9. Timmer A, McDonald JW, Macdonald JK. Azathioprine and 6-
mercaptopurine for maintenance of remission in ulcerative colitis.
Cochrane Database Syst Rev. 2007:CD000478.
10. Mountifield RE, Prosser R, Bampton P, et al. Pregnancy and IBD
treatment: this challenging interplay from a patients’ perspective.
J Crohns Colitis. 2010;4:176–182.
11. Julsgaard M, Norgaard M, Hvas CL, et al. Self-reported adherence to
medical treatment prior to and during pregnancy among women with
ulcerative colitis. Inflamm Bowel Dis. 2011;17:1573–1580.
12. Norgard B, Pedersen L, Fonager K, et al. Azathioprine, mercaptopur-
ine and birth outcome: a population-based cohort study. Aliment Phar-
macol Ther. 2003;17:827–834.
13. Ostrer H, Stamberg J, Perinchief P. Two chromosome aberrations in
the child of a woman with systemic lupus erythematosus treated with
azathioprine and prednisone. Am J Med Genet. 1984;17:627–632.
14. Davison JM, Dellagrammatikas H, Parkin JM. Maternal azathioprine
therapy and depressed haemopoiesis in the babies of renal allograft
patients. Br J Obstet Gynaecol. 1985;92:233–239.
15. Cote CJ, Meuwissen HJ, Pickering RJ. Effects on the neonate of pred-
nisone and azathioprine administered to the mother during pregnancy.
J Pediatr. 1974;85:324–328.
16. DeWitte DB, Buick MK, Cyran SE, et al. Neonatal pancytopenia
and severe combined immunodeficiency associated with antenatal
administration of azathioprine and prednisone. J Pediatr. 1984;105:
625–628.
17. Pirson Y, Van Lierde M, Ghysen J, et al. Retardation of fetal growth
in patients receiving immunosuppressive therapy. N Engl J Med.
1985;313:328.
18. Ben-Neriah Z, Ackerman Z. WAGR syndrome in a baby—the result
of 6-MP treatment in a father affected by Crohn’s disease? Am J Gas-
troenterol. 2001;96:251.
19. Cornish J, Tan E, Teare J, et al. A meta-analysis on the influence of
inflammatory bowel disease on pregnancy. Gut. 2007;56:830–837.
20. The Newcastle-Ottawa Scale (NOS) for assessing the quality of non-
randomised studies in meta-analyses. Available at http://www.ohri.ca/
programs/clinical_epidemiology/oxford.asp
21. Cook DA, Levinson AJ, Garside S. Method and reporting quality in
health professions education research: a systematic review. Med Educ.
2010;45:227–238.
22. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observatio-
nal studies in epidemiology: a proposal for reporting. Meta-analysis
Of Observational Studies in Epidemiology (MOOSE) group. JAMA.
2000;283:2008–2012.
23. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for
systematic reviews and meta-analyses: the PRISMA statement. J Clin
Epidemiol. 2009;62:1006–1012.
24. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin
Trials. 1986;7:177–188.
25. Rosner B. Fundamentals of Biostatistics. 4th ed. Belmont, CA: Wads-
worth Publishing; 1995.
26. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-
analysis. Stat Med. 2002;21:1539–1558.
27. Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis
detected by a simple, graphical test. BMJ. 1997;315:629–634.
28. Naganuma M, Kunisaki R, Yoshimura N, et al. Conception and preg-
nancy outcome in women with inflammatory bowel disease: a multi-
centre study from Japan. J Crohns Colitis. 2011;5:317–323.
29. Bengtson MB, Solberg IC, Aamodt G, et al. Relationships between
inflammatory bowel disease and perinatal factors: both maternal and
paternal disease are related to preterm birth of offspring. Inflamm
Bowel Dis. 2010;16:847–855.
Effects of Thiopurines on Birth Outcomes
30. Bortoli A, Pedersen N, Duricova D, et al. Pregnancy outcome in inflam-
matory bowel disease: prospective European case-control ECCO-Epi-
Com study, 2003–2006. Aliment Pharmacol Ther. 2010;34:724–734.
31. Molnar T, Farkas K, Nagy F, et al. Pregnancy outcome in patients
with inflammatory bowel disease according to the activity of the dis-
ease and the medical treatment: a case-control study. Scand J Gastro-
enterol. 2010;45:1302–1306.
32. Moskovitz DN, Bodian C, Chapman ML, et al. The effect on the fetus
of medications used to treat pregnant inflammatory bowel-disease
patients. Am J Gastroenterol. 2004;99:656–661.
33. Raatikainen K, Mustonen J, Pajala MO, et al. The effects of pre- and
post-pregnancy inflammatory bowel disease diagnosis on birth out-
comes. Aliment Pharmacol Ther. 2010;33:333–339.
34. Bortoli A, Saibeni S, Tatarella M, et al. Pregnancy before and after
the diagnosis of inflammatory bowel diseases: retrospective case-con-
trol study. J Gastroenterol Hepatol. 2007;22:542–549.
35. Baiocco PJ, Korelitz BI. The influence of inflammatory bowel disease
and its treatment on pregnancy and fetal outcome. J Clin Gastroen-
terol. 1984;6:211–216.
36. Mahadevan U, Sandborn WJ, Li DK, et al. Pregnancy outcomes in
women with inflammatory bowel disease: a large community-based study
from Northern California. Gastroenterology. 2007;133:1106–1112.
37. Hudson M, Flett G, Sinclair TS, et al. Fertility and pregnancy in
inflammatory bowel disease. Int J Gynaecol Obstet. 1997;58:229–237.
38. Bouhnik Y, Lemann M, Mary JY, et al. Long-term follow-up of
patients with Crohn’s disease treated with azathioprine or 6-mercapto-
purine. Lancet. 1996;347:215–219.
39. Oefferlbauer-Ernst A, Reinisch W, Miehsler W, et al. Healthy off-
spring in parents both receiving thiopurines. Gastroenterology. 2004;
126:628.
40. Alstead EM, Ritchie JK, Lennard-Jones JE, et al. Safety of azathio-
prine in pregnancy in inflammatory bowel disease. Gastroenterology.
1990;99:443–446.
41. Present DH, Meltzer SJ, Krumholz MP, et al. 6-Mercaptopurine in the
management of inflammatory bowel disease: short- and long-term tox-
icity. Ann Intern Med. 1989;111:641–649.
42. Sato A, Naganuma M, Asakura K, et al. Conception outcomes and
opinions about pregnancy for men with inflammatory bowel disease.
J Crohns Colitis. 2010;4:183–188.
43. Goldstein LH, Dolinsky G, Greenberg R, et al. Pregnancy outcome of
women exposed to azathioprine during pregnancy. Birth Defects Res
A Clin Mol Teratol. 2007;79:696–701.
44. Norgard B, Pedersen L, Jacobsen J, et al. The risk of congenital
abnormalities in children fathered by men treated with azathioprine or
mercaptopurine before conception. Aliment Pharmacol Ther. 2004;19:
679–685.
45. Angelberger S, Reinisch W, Messerschmidt A, et al. Long-term fol-
low-up of babies exposed to azathioprine in utero and via breastfeed-
ing. J Crohns Colitis. 2011;5:95–100.
46. Zlatanic J, Korelitz BI, Rajapakse R, et al. Complications of preg-
nancy and child development after cessation of treatment with 6-mer-
captopurine for inflammatory bowel disease. J Clin Gastroenterol.
2003;36:303–309.
47. Norgard B, Pedersen L, Christensen LA, et al. Therapeutic drug use
in women with Crohn’s disease and birth outcomes: a Danish nation-
wide cohort study. Am J Gastroenterol. 2007;102:1406–1413.
48. Langagergaard V, Pedersen L, Gislum M, et al. Birth outcome in women
treated with azathioprine or mercaptopurine during pregnancy: A Danish
nationwide cohort study. Aliment Pharmacol Ther. 2007;25:73–81.
49. Francella A, Dyan A, Bodian C, et al. The safety of 6-mercaptopurine
for childbearing patients with inflammatory bowel disease: a retro-
spective cohort study. Gastroenterology. 2003;124:9–17.
50. Coelho J, Beaugerie L, Colombel JF, et al. Pregnancy outcome in
patients with inflammatory bowel disease treated with thiopurines:
cohort from the CESAME study. Gut. 2010;60:198–203.
51. Cleary BJ, Kallen B. Early pregnancy azathioprine use and pregnancy
outcomes. Birth Defects Res A Clin Mol Teratol. 2009;85:647–654.
52. Shim L, Eslick GD, Simring AA, et al. The effects of azathioprine on
birth outcomes in women with inflammatory bowel disease (IBD).
J Crohns Colitis. 2011;5:234–238.
7
Akbari et al
53. Teruel C, Lopez-San Roman A, Bermejo F, et al. Outcomes of preg-
nancies fathered by inflammatory bowel disease patients exposed to
thiopurines. Am J Gastroenterol. 2010;105:2003–2008.
54. Rajapakse RO, Korelitz BI, Zlatanic J, et al. Outcome of pregnancies
when fathers are treated with 6-mercaptopurine for inflammatory
bowel disease. Am J Gastroenterol. 2000;95:684–688.
55. Bush MC, Patel S, Lapinski RH, et al. Perinatal outcomes in
inflammatory bowel disease. J Matern Fetal Neonatal Med. 2004;
15:237–241.
56. Fedorkow DM, Persaud D, Nimrod CA. Inflammatory bowel disease:
a controlled study of late pregnancy outcome. Am J Obstet Gynecol.
1989;160:998–1001.
57. Morales M, Berney T, Jenny A, et al. Crohn’s disease as a risk factor for
the outcome of pregnancy. Hepatogastroenterology. 2000;47:1595–1598.
58. Norgard B, Hundborg HH, Jacobsen BA, et al. Disease activity in
pregnant women with Crohn’s disease and birth outcomes: a regional
Danish cohort study. Am J Gastroenterol. 2007;102:1947–1954.
59. Reddy D, Murphy SJ, Kane SV, et al. Relapses of inflammatory bowel
disease during pregnancy: in-hospital management and birth outcomes.
Am J Gastroenterol. 2008;103:1203–1209.
60. Koretz RL. Methods of meta-analysis: an analysis. Curr Opin Clin
Nutr Metab Care. 2002;5:467–474.
Inflamm Bowel Dis
61. Kornfeld D, Cnattingius S, Ekbom A. Pregnancy outcomes in women
with inflammatory bowel disease—a population-based cohort study.
Am J Obstet Gynecol. 1997;177:942–946.
62. Fonager K, Sorensen HT, Olsen J, et al. Pregnancy outcome for
women with Crohn’s disease: a follow-up study based on linkage
between national registries. Am J Gastroenterol. 1998;93:2426–2430.
63. Elbaz G, Fich A, Levy A, et al. Inflammatory bowel disease and pre-
term delivery. Int J Gynaecol Obstet. 2005;90:193–197.
64. Baird DD, Narendranathan M, Sandler RS. Increased risk of preterm
birth for women with inflammatory bowel disease. Gastroenterology.
1990;99:987–994.
65. Dominitz JA, Young JC, Boyko EJ. Outcomes of infants born to
mothers with inflammatory bowel disease: a population-based cohort
study. Am J Gastroenterol. 2002;97:641–648.
66. Stephansson O, Larsson H, Pedersen L, et al. Crohn’s disease is a risk
factor for preterm birth. Clin Gastroenterol Hepatol. 2010;8:509–515.
67. Stephansson O, Larsson H, Pedersen L, et al. Congenital abnormalities
and other birth outcomes in children born to women with ulcerative coli-
tis in Denmark and Sweden. Inflamm Bowel Dis. 2011;17:795–801.
68. Malgarinos G, Gikas A, Delicha E, et al. Pregnancy and inflammatory
bowel disease: a prospective case-control study. Rev Med Chir Soc
Med Nat Iasi. 2007;111:613–619.