SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Record linkages
  7. Use of azathioprine and mercaptopurine (exposure information)
  8. Outcome data
  9. Data on possible confounders
  10. Statistical analysis
  11. Results
  12. Discussion
  13. Acknowledgements
  14. References

Background : Immunosuppressive therapy with azathioprine and mercaptopurine is commonly used in patients with various chronic diseases. The few existing data on the reproductive safety of these drugs after paternal use before conception are inconclusive.

Aim : To examine the risk of congenital abnormalities in children fathered by men exposed to azathioprine or mercaptopurine before conception.

Methods : This was a Danish population-based cohort study, based on data from the Prescription Database, the Medical Birth Registry and the Hospital Discharge Registry of North Jutland County, Denmark. Fifty-four exposed pregnancies, in which the father filed a prescription for azathioprine or mercaptopurine (between 1 January 1991 and 31 December 2001) before conception, were included. The controls comprised 57 195 pregnancies with no paternal azathioprine or mercaptopurine use.

Results : Four children with congenital abnormalities (underlying paternal diseases: glomerulonephritis and severe skin disease) were found in 54 exposed pregnancies (7.4%), compared with 2334 (4.1%) in controls. The adjusted odds ratio for congenital abnormalities in children fathered by men treated with azathioprine or mercaptopurine was 1.8 (95% confidence interval, 0.7–5.0).

Conclusions : Our data may indicate that paternal use of azathioprine or mercaptopurine before conception is associated with an increased risk of congenital abnormalities. However, more data are needed to determine whether the association is causal.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Record linkages
  7. Use of azathioprine and mercaptopurine (exposure information)
  8. Outcome data
  9. Data on possible confounders
  10. Statistical analysis
  11. Results
  12. Discussion
  13. Acknowledgements
  14. References

Azathioprine and mercaptopurine are immunomodulatory drugs, both commonly used in the treatment of chronic diseases, such as severe cases of inflammatory bowel disease and autoimmune disease, and in the prevention of organ transplant rejection. Azathioprine, an inhibitor of purine metabolism, is rapidly converted in the liver into a number of metabolites, including the active mercaptopurine.1–3 There are no data suggesting that azathioprine and mercaptopurine have different efficacy profiles, and therefore both drugs are usually discussed together with regard to their efficacy and safety.4, 5

Recently, concerns have been raised regarding the risk of pregnancy-related complications in the offspring when the father is treated with mercaptopurine before conception.6–11 The main concern is based on data from a cohort study reporting an increased risk of congenital abnormalities in the offspring when the fathers had taken mercaptopurine before conception (two cases of congenital abnormality in 50 exposed pregnancies and no cases in 90 controls).6 Thus, the safety of mercaptopurine in the area of reproduction has been brought into question, and attention has been drawn to a potentially teratogenic effect.6, 7, 9, 10

Pregnancy outcomes depend on both maternal and paternal factors, and the hypothesis of an increased risk of birth defects associated with paternal drug exposure is not new. Thus, for decades, animal studies have indicated that the paternal use of certain drugs may induce congenital abnormalities in the offspring, but human data are very sparse.12 The mechanisms behind a paternal drug effect on pregnancy outcome may include: (i) genetic or chromosomal damage of the spermatocytes; (ii) drugs or metabolites in seminal fluid that might influence sperm maturation and/or produce a direct effect on the uterus; (iii) a systemic effect of the drug or metabolites on the female herself by absorption through the vaginal mucosa.12, 13 In animal models, mercaptopurine has been shown to cause chromosomal damage and aberrations in the spermatocytes14–17 and congenital abnormalities in the offspring.3, 10, 11, 18

Data are insufficient with regard to a possible teratogenic effect of azathioprine or mercaptopurine after paternal use. With this background, in a population-based design, we examined the risk of congenital abnormalities in children fathered by men treated with azathioprine or mercaptopurine before conception.

Study population

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Record linkages
  7. Use of azathioprine and mercaptopurine (exposure information)
  8. Outcome data
  9. Data on possible confounders
  10. Statistical analysis
  11. Results
  12. Discussion
  13. Acknowledgements
  14. References

We conducted the study in North Jutland County, Denmark (population 496 000, which is approximately 10% of the Danish population). We included data on all women who, between 1 January 1991 and 31 December 2001, had a liveborn child, and we identified all singleton pregnancies in the county during the study period (64 082 pregnancies).

From the Central Person Registry in Denmark, we identified all men registered as fathers to the birth cohort. Since 1 April 1968, all persons in Denmark have been registered in the Central Person Registry; a 10-digit Civil Registry number (CPR number), which incorporates the date of birth, is assigned shortly after birth. In addition, the registry includes data on residency, date of immigration or emigration and vital status and, for each individual, the CPR numbers of the parents are recorded. During the study period, a missing CPR number of the father was recorded for 6833 singleton pregnancies (10.7%), leaving 57 249 pregnancies for analysis.

Record linkages

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Record linkages
  7. Use of azathioprine and mercaptopurine (exposure information)
  8. Outcome data
  9. Data on possible confounders
  10. Statistical analysis
  11. Results
  12. Discussion
  13. Acknowledgements
  14. References

Data on prescriptions, outcome (congenital abnormalities) and confounders were obtained from the population-based registries in North Jutland County. The CPR number was used for linkage between these registries and prescription records.

In Denmark, stillborn children are not given a CPR number, and therefore we could not assess possible congenital abnormalities amongst stillborn children in the registries used.

Use of azathioprine and mercaptopurine (exposure information)

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Record linkages
  7. Use of azathioprine and mercaptopurine (exposure information)
  8. Outcome data
  9. Data on possible confounders
  10. Statistical analysis
  11. Results
  12. Discussion
  13. Acknowledgements
  14. References

The population-based Pharmaco-Epidemiological Prescription Database of North Jutland was used to identify all fathers who had filed prescriptions for azathioprine or mercaptopurine before conception of their child. The county is served by pharmacies equipped with computerized accounting systems, from which data are sent to the Danish National Health Service. The National Health Service provides tax-supported health care for all inhabitants of the country. Apart from guaranteeing free access to general practitioners and hospitals, the insurance programme refunds part of the costs associated with the purchase of most prescribed drugs. The data transferred to the prescription database from the accounting system include the patient's CPR number, the type of drug prescribed, according to the anatomical therapeutic chemical classification system, and the date of the prescription. We identified all prescriptions with the code L04A X01 (azathioprine) and L01B B02 (mercaptopurine). Both drugs are available only on prescription.

Fathers were categorized as exposed if they, at any time before conception, had filed a prescription for azathioprine or mercaptopurine. The controls comprised all fathers who had never taken azathioprine or mercaptopurine before the time of conception in the birth cohort.

The County Hospital Discharge Registry.  Data on the type of congenital abnormality were extracted from this registry. Established in 1977, data in this registry include the dates of hospital admission and discharge, surgical procedures performed and up to 20 discharge diagnoses, classified according to the Danish version of the International Classification of Diseases, 8th revision (ICD-8) until the end of 1993, and ICD-10 after this date. The codes for malformations were 740.00–759.99 in ICD-8 and Q0.00-Q99.9 in ICD-10. Diagnoses of dislocation of the hip and undescended testis were excluded because of their low validity.19

Children with congenital abnormalities, whose fathers were exposed to azathioprine or mercaptopurine, were reviewed in the hospital records.

The County Medical Birth Registry.  Pregnancy-related information was collected from the Medical Birth Registry in the county. This registry contains detailed information on all births since 1 January 1973; the midwives and doctors responsible for the deliveries record the data. For confounder control, we used information on maternal age at the time of delivery, parity, smoking status of the mother at the first antenatal care visit and gender of the child.20

Data on paternal underlying diseases.  We aimed to assess the reason for the fathers' prescriptions of azathioprine or mercaptopurine using information on discharge diagnoses in the Hospital Discharge Registry, given before azathioprine or mercaptopurine prescription. However, on the basis of discharge diagnoses, it was not always possible to obtain information on the indication for azathioprine or mercaptopurine prescription. Despite having filed prescriptions for azathioprine or mercaptopurine, the patients might not have been admitted to hospital for the disease, and therefore not recorded in the Hospital Discharge Registry. In such cases, diagnoses were given and treatments initiated by specialists engaged in private practices (often dermatologists). Therefore, the concomitant prescription pattern of other drugs in the prescription database was used to assess the underlying cause of azathioprine or mercaptopurine exposure (mostly repeat prescriptions of topical glucocorticoid compounds). No information was available on the severity of diseases.

Statistical analysis

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Record linkages
  7. Use of azathioprine and mercaptopurine (exposure information)
  8. Outcome data
  9. Data on possible confounders
  10. Statistical analysis
  11. Results
  12. Discussion
  13. Acknowledgements
  14. References

We constructed contingency tables for the main study variables and estimated the prevalence odds ratio (OR) with 95% confidence intervals (CI) for congenital abnormalities. We performed logistic regression analyses to estimate the relative risk (prevalence OR) of congenital abnormalities amongst fathers using azathioprine or mercaptopurine before conception compared with no paternal use of these drugs before pregnancy. Analyses were adjusted for maternal age, parity, maternal smoking and gender of the child. Insufficient data were available to allow stratification for the type of underlying paternal disease.

For model control, we used the Hosmer–Lemeshow statistic. All analyses were performed using SAS version 8.02 (SAS Institute, Cary, NC, USA).

The study was approved by the Danish Registry Board (No 2003-41-3554).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Record linkages
  7. Use of azathioprine and mercaptopurine (exposure information)
  8. Outcome data
  9. Data on possible confounders
  10. Statistical analysis
  11. Results
  12. Discussion
  13. Acknowledgements
  14. References

In 54 pregnancies, the father had filed a prescription for azathioprine or mercaptopurine at some time before conception. Of these, the most common indications for azathioprine or mercaptopurine treatment were prevention of organ rejection amongst transplant recipients (n = 19, 35.2%) and Crohn's disease or ulcerative colitis (n = 17, 31.5%) (Table 1). Azathioprine was used in 52 cases (96.3%) and mercaptopurine in two (3.7%). For the 54 exposed pregnancies, the mean and range of time from the father's last prescription of azathioprine or mercaptopurine to the time of conception were 16.8 months and 0–121 months, respectively. The control group comprised 57 195 pregnancies (Table 1). In pregnancies in which the fathers had filed a prescription for azathioprine or mercaptopurine, the mothers were older, smoked less and were more often multipara. Congenital abnormalities were found in 7.4% (n = 4/54) of the children fathered by men treated with azathioprine or mercaptopurine before conception, and in 4.1% (n = 2334/57 195) of the controls (Table 1).

Table 1.  Characteristics of pregnancies fathered by men treated with azathioprine (AZA) or mercaptopurine (MP) before conception and of control pregnancies
 Pregnancies fathered by men treated with AZA/MP before conception* (n = 54)Controls* (n = 57 195)
  • *

     Represents 47 different fathers exposed to AZA or MP and 39 576 different unexposed fathers.

Children with congenital abnormalities, n (%)4 (7.4)2334 (4.1)
Mother's age (years)
 Mean (s.d.)29.7 (5.4)28.5 (4.6)
 Range20–4514–47
Mother's smoking status, n (%)13 (24.1)15473 (27.0)
Mother's parity > 1, n (%)34 (63.0)33814 (59.1)
Father's underlying disease
 Transplant recipient, n (%)19 (35.2)
 Ulcerative colitis or Crohn's disease, n (%)17 (31.5)
 Skin disease, n (%)8 (14.8)
 Rheumatic disease and connective tissue disease, n (%)7 (13.0)
 Giant follicular lymphoma (Brill–Symmers disease), n (%)2 (3.7)
 Chronic hepatitis, n (%)1 (1.9)
 Unknown, n (%)057195 (100)

In the exposed group, all cases of congenital abnormality were reviewed in the medical records. The types of malformation included: (i) polysyndactylia; (ii) oesophagus atresia; (iii) hydronephrosis and megaloureter; and (iv) ventricular septal defect. All cases of congenital abnormality were in male children. The time periods from the father's last prescription of azathioprine or mercaptopurine to the time of conception were 9 months, 25 months, 28 months and 38 months, respectively (Table 2). For the four exposed men who fathered children with congenital abnormalities, we looked at the type of other drugs prescribed in the year before the time of conception (Table 2). Within this time window, none of these fathers filed prescriptions for potential teratogens, such as ciclosporin, tacrolimus or methotrexate. Two of the four mothers giving birth to children with congenital abnormalities did not file any prescriptions during their pregnancies. One mother filed a prescription for topical hydrocortisone during the first trimester and antifungal vagitoria (miconazole) and anti-asthmatic inhalation drugs [eformoterol (formoterol), budesonide] later in pregnancy. Another mother filed a prescription for topical nystatin during the first trimester and codeine coughing mixture and penicillin later in pregnancy.

Table 2.  Details of azathioprine (AZA)- and mercaptopurine (MP)-exposed fathers and the four children with congenital abnormalities (CAs)
ChildType of CAUnderlying paternal diseasePaternal use of AZA or MPMonths between last AZA prescription and conceptionPaternal use of other drug types within 1 year before conception
1PolysyndactyliaGlomerulonephritis and renal transplant recipientAZA (repeat prescriptions) 9Omeprazole, frusemide (furosemide), metoprolol, felodipine, prednisone
2Oesophagus atresiaGlomerulonephritis and renal transplant recipientAZA (repeat prescriptions)25Frusemide (furosemide), prednisone
3Hydronephrosis and megaloureterSevere skin diseaseAZA (one prescription)28Topical hydrocortisone, different types of antibiotic, analgesics
4Ventricular septal defectGlomerulonephritis and renal transplant recipientAZA (repeat prescriptions)38Frusemide (furosemide), metoprolol, prednisone

The overall adjusted risk of congenital abnormalities in children fathered by men treated with azathioprine or mercaptopurine was 1.8 (95% CI, 0.7–5.0) (Table 3).

Table 3.  Crude and adjusted odds ratio (OR), with 95% confidence interval (CI), for congenital abnormalities in children fathered by men treated with azathioprine (AZA) or mercaptopurine (MP) before conception
 AZA- or MP-exposed pregnancies, n (%)Control pregnancies, n (%)Crude OR (95% CI)OR* (95% CI)
  • Adjusted for mother's age (below 25 years, 25–29 years and 30 years or more), parity (one or more than one), maternal smoking (yes/no) and gender of the child in a logistic regression model.

  • Hosmer–Lemeshow model control: P = 0.51.

Congenital abnormalities4/54 (7.4)2334/57195 (4.1)1.9 (0.7–5.2)1.8 (0.7–5.0)

In a sub-analysis, we found that, within 3 months before conception, paternal azathioprine or mercaptopurine prescriptions were filed for 19 pregnancies, but none of these men fathered children with congenital abnormalities.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Record linkages
  7. Use of azathioprine and mercaptopurine (exposure information)
  8. Outcome data
  9. Data on possible confounders
  10. Statistical analysis
  11. Results
  12. Discussion
  13. Acknowledgements
  14. References

We found an increased risk of congenital abnormalities in children fathered by men treated with azathioprine or mercaptopurine before conception. All cases of congenital abnormalities were in male children. For children with congenital abnormalities whose fathers were treated with azathioprine or mercaptopurine, the time interval between the father's last prescription for azathioprine or mercaptopurine and the time of conception was 9–38 months. In men who filed prescriptions for azathioprine or mercaptopurine within 3 months of the time of conception, no cases of congenital abnormality were identified.

For ethical reasons, no randomized trials can be designed to evaluate the safety of azathioprine or mercaptopurine with regard to reproduction. Therefore, clinical decisions have to be based on observational studies that are often vulnerable to bias and to problems with statistical precision due to the low prevalence of adverse pregnancy outcomes, in particular congenital abnormalities. The reproductive safety associated with the paternal use of azathioprine or mercaptopurine is still poorly understood. It is therefore important to report all available data on this topic, even though the number of exposed pregnancies might be considered to be relatively small and the statistical power not optimal. Studies of teratogenesis require special attention, as congenital abnormalities cannot be regarded as a single homogeneous outcome; exposure to a teratogen does not uniformly increase the rates of all congenital abnormalities, but rather tends to increase the rates of selected congenital abnormalities.21 Therefore, under ideal circumstances, studies of specific rather than overall rates of congenital abnormalities should be performed. However, no such data are available as yet, and considerably more data are needed before the risk of specific types of congenital abnormality can be evaluated. In a cohort study of this size, we would therefore not expect to find congenital abnormalities belonging to the same organ system.

Our study was based on a complete prescription database, which prevents selection bias. One strength was that exposure measurement was based on prescriptions and not on recall. Drug exposure based on self-reported use may lead to recall bias or under-ascertainment, which is a serious threat in case–control studies of congenital abnormalities.22 Another strength was that the outcome data were obtained independently of exposure measurement.

On the other hand, no information was available on compliance. However, patient non-compliance is not likely to have a major impact, because the drugs are typically used for long-term treatment in these usually co-operative patients with severe chronic diseases; any potential misclassification of exposure would only tend to under-estimate our risk estimates. With regard to possible confounders, we were able to take into consideration the influence of the mother's age, parity, maternal smoking and gender of the child. None of these factors, however, changed our risk estimates. We had no opportunity to consider the possible influence of the type of the underlying paternal disease or disease activity, because the amount and type of data did not allow us to perform stratified analyses according to these factors. Furthermore, we were unable to evaluate a possible dose effect. From our database, we could extract information on the total amount of azathioprine or mercaptopurine prescribed (in milligrams), but could not relate this amount to the number of milligrams prescribed per day or to the length of treatment. From an evaluation of the type of maternal drug exposure during pregnancy for exposed cases of congenital abnormality, we have no reason to believe that maternal drug use confounded the association.

Our data support the evidence from the only other study with a control group that has specifically estimated the risk of congenital abnormalities after paternal mercaptopurine exposure.6 In that study, Rajapakse et al. found an increased risk of congenital abnormalities in children fathered by men treated with mercaptopurine for inflammatory bowel diseases, compared with fathers who had never been treated with mercaptopurine (two cases of congenital abnormality in 50 exposed pregnancies vs. no cases of congenital abnormality in 90 control pregnancies).6 A study by Francella et al., that included a control group, concluded that mercaptopurine appeared to be safe before or at the time of conception with regard to pregnancy outcomes in general (including congenital abnormalities).5 In that study, the rate of having a successful pregnancy after exposure to mercaptopurine (maternal or paternal) was estimated to be 0.85 (95% CI, 0.47–1.55) relative to controls, adjusted for the gender of the affected parent and age of the mother.5 However, it was not the aim of the study to estimate the separate risk of congenital abnormalities after paternal mercaptopurine exposure, and therefore it was not comparable with the method used in our study.

From the observation of the time period between the last prescription of azathioprine or mercaptopurine and the time of conception, our data may indicate a possible long-term teratogenic potential. However, a single prescription of azathioprine or mercaptopurine may include a large number of tablets providing treatment for long periods. Thus, from information on the last prescription for azathioprine or mercaptopurine, we cannot be sure for how many months the fathers were actually exposed before conception. The types of underlying disease in the azathioprine-exposed fathers who had children with congenital abnormalities may, in fact, indicate a permanent use of azathioprine. From these arguments, our finding of a possible long-term effect of azathioprine or mercaptopurine should be interpreted with caution. According to the findings of Rajapakse et al., it has been suggested that male patients should be advised to discontinue treatment with azathioprine or mercaptopurine 3 months before conception.6 This 3-month cut-off point seems relevant in light of the fact that spermatogenesis takes approximately 70–90 days in humans. Nevertheless, even if the last process of maturation of the sperm cells lasts for 3 months, it may be possible that azathioprine or mercaptopurine actually induces damage to the early primitive stages of the sperm cells. In animal models, mercaptopurine has been shown to induce mutations in very early meiotic or differentiating gonial stages.14, 17

Three of the four congenital abnormalities were found in children fathered by men with renal transplants, and the fourth father had severe skin disease. However, from these observations, no conclusion can be drawn with regard to the possible impact of the type of underlying paternal disease.

On the basis of data from epidemiological studies, concerns have been raised about the reproductive safety of azathioprine and mercaptopurine. Our data suggest an increased risk of congenital abnormalities in children fathered by men using azathioprine or mercaptopurine before conception. In addition, our data may indicate a more long-term teratogenic potential of azathioprine and mercaptopurine than previously assumed. Therefore, our data may complicate the question about the recommended period of discontinuation of azathioprine or mercaptopurine treatment before conception. Thus, until more data become available on the reproductive safety, we believe that it is important to be cautious about the use of azathioprine and mercaptopurine in men in the fertile age band.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study population
  6. Record linkages
  7. Use of azathioprine and mercaptopurine (exposure information)
  8. Outcome data
  9. Data on possible confounders
  10. Statistical analysis
  11. Results
  12. Discussion
  13. Acknowledgements
  14. References
  • 1
    Lamers CB, Griffioen G, Van Hogezand RA, Veenendaal RA. Azathioprine: an update on clinical efficacy and safety in inflammatory bowel disease. Scand J Gastroenterol Suppl 1999; 230: 1115.
  • 2
    Saarikoski S, Seppala M. Immunosuppression during pregnancy: transmission of azathioprine and its metabolites from the mother to the fetus. Am J Obstet Gynecol 1973; 115: 11006.
  • 3
    Armenti VT, Moritz MJ, Davison JM. Drug safety issues in pregnancy following transplantation and immunosuppression: effects and outcomes. Drug Saf 1998; 19: 21932.
  • 4
    Meuwissen SG, Ewe K, Gassull MA, et al. IOIBD questionnaire on the clinical use of azathioprine, mercaptopurine, cyclosporin A and methotrexate in the treatment of inflammatory bowel diseases. Eur J Gastroenterol Hepatol 2000; 12: 138.
  • 5
    Francella A, Dyan A, Bodian C, Rubin P, Chapman M, Present DH. The safety of mercaptopurine for childbearing patients with inflammatory bowel disease: a retrospective cohort study. Gastroenterology 2003; 124: 917.
  • 6
    Rajapakse RO, Korelitz BI, Zlatanic J, Baiocco PJ, Gleim GW. Outcome of pregnancies when fathers are treated with mercaptopurine for inflammatory bowel disease. Am J Gastroenterol 2000; 95: 6848.
    Direct Link:
  • 7
    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 Gastroenterol 2001; 96: 251.
  • 8
    Kane SV. What's good for the goose should be good for the gander — 6-MP use in fathers with inflammatory bowel disease. Am J Gastroenterol 2000; 95: 5812.
  • 9
    Korelitz BI, Rajapakse RO, Zlatanic J, Baiocco PJ, Gleim GW. 6MP for fathers with IBD at the time of conception: the value of a large clinical experience over unrelated controls. Am J Gastroenterol 2000; 95: 23867.
    Direct Link:
  • 10
    Katz JA, Pore G. Inflammatory bowel disease and pregnancy. Inflamm Bowel Dis 2001; 7: 14657.
  • 11
    Connell W, Miller A. Treating inflammatory bowel disease during pregnancy: risks and safety of drug therapy. Drug Saf 1999; 21: 31123.
  • 12
    Joffe JM. Influence of drug exposure of the father on perinatal outcome. Clin Perinatol 1979; 6: 2136.
  • 13
    Joffe JM, Soyka LF. Paternal drug exposure: effects on reproduction and progeny. Semin Perinatol 1982; 6: 11624.
  • 14
    Mosesso P, Palitti F. The genetic toxicology of mercaptopurine. Mutat Res 1993; 296: 27994.
  • 15
    Oakberg EF, Crosthwait CD, Raymer GD. Spermatogenic stage sensitivity to mercaptopurine in the mouse. Mutat Res 1982; 94: 16578.
  • 16
    Russell LB, Hunsicker PR. Study of the base analog mercaptopurine in the mouse specific-locus test. Mutat Res 1987; 176: 4752.
  • 17
    Generoso WM, Preston RJ, Brewen JG. 6-Mercaptopurine, an inducer of cytogenetic and dominant-lethal effects in premeiotic and early meiotic germ cells of male mice. Mutat Res 1975; 28: 43747.
  • 18
    Friedman JM, Polifka JE. The Effects of Drugs on the Fetus and Nursing Infant. Baltimore: Johns Hopkins University Press, 1996.
  • 19
    Larsen H, Nielsen GL, Bendsen J, Flint C, Olsen J, Sorensen HT. Predictive value and completeness of the registration of congenital abnormalities in three Danish population-based registries. Scand J Public Health 2003; 31: 126.
  • 20
    Knudsen LB, Olsen J. The Danish Medical Birth Registry. Dan Med Bull 1998; 45: 3203.
  • 21
    Mitchell AA. Special considerations in studies of drug-induced birth defects. In: StromBL, ed. Pharmacoepidemiology, 3rd edn. Chichester: John Wiley & Sons, 2002: 74963.
  • 22
    West SL, Savitz DA, Koch G, Strom BL, Guess HA, Hartzema A. Recall accuracy for prescription medications: self-report compared with database information. Am J Epidemiol 1995; 142: 110312.