What’s Good for theGoose Should Be Good forthe Gander—6-MP Use in FathersWith Inflammatory Bowel DiseaseImmunomodulators are now a standard therapy for the treat-ment of inflammatory bowel disease (IBD), with well doc-umented efficacy in both ulcerative colitis and Crohn’sdisease (1, 2). However, data regarding their safety duringpregnancy is limited. The majority of the human safety datafor immunomodulators is derived from historical reviews oftransplant recipients (3–5) and patients with other autoim-mune diseases (6, 7). The results from these retrospectivestudies have been favorable, with no significantly increasedrisks for either spontaneous abortions or congenital anom-alies.

The data for 6-MP use during pregnancy in IBD has beenlimited to two small studies (8, 9). In the retrospective seriesby Alsteadet al., no adverse outcomes were seen in 16pregnancies in 14 women. In a second abstract, Francellaetal. reported 167 conceptions with some maternal history of6-MP use. There were no spontaneous abortions or congen-ital anomalies seen in eight women who continued 6-MPthroughout the pregnancy. In the 64 women who were on6-MP at conception, there was no significant increased rateof spontaneous abortions or congenital anomalies comparedto the control group of patients without exposure to 6-MP.

The available data for 6-MP use in men is even moresparse. Although the effects of sulphasalazine on spermmotility and density are well known (10), to date no one hasreported on adverse events secondary to fathers using im-munomodulatory therapy in IBD. Again we draw our infor-mation from encouraging reports from the transplant popu-lation. Pennet al. describe 23 children fathered by 19 men.There were 21 normal live births, one spontaneous abortion,and one child with myelomeningocele (11). The authorsconcluded that “the risk [of adverse outcomes] was notexorbitant, particularly for paternity.” In Golby’s reportfrom a multicenter registry of renal transplant patients, 38 of40 births were normal, with only two first trimester abor-tions, an incidence lower than that in the general population(12).

In this month’s issue of theAmerican Journal of Gastro-enterology, Rajapakseet al. (13) report birth defects andadverse outcomes in children fathered by men on 6-MP.They identified 50 pregnancies in which men had exposureto 6-MP. Thirteen pregnancies were fathered by men within3 months of 6-MP exposure (group 1A), and 37 with aprevious exposure before the 3 months before conception(group 1B). A group of children fathered by men withoutpreconception exposure to 6-MP was used as a controlgroup (group 2). There were four adverse events in group1A—two spontaneous abortions and two congenital anom-alies. There was one spontaneous abortion in group 1B, andtwo in the control group, without any congenital anomalies

seen in either group. There was no statistical differencebetween groups 1 and 2 for the incidence of complications(p 5 0.097). With subgroup analysis, however,p values of,0.05 were seen when group 1A was compared to group 1Bor group 2. Based upon this analysis, the conclusion was thatan increased incidence of pregnancy complications occurswhen fertilization is within 3 months of 6-MP use by thefather.

These findings must be put into proper prospective. In thegeneral population, the baseline incidence of congenitalanomalies is 3% (14) and 10–12% for spontaneous abor-tions (15). The rate of spontaneous abortion in the controlgroup was much lower (3%), and was used for the compar-ison statistic rather than the accepted 10% for the generalpopulation. This would greatly inflate any difference seen inthe incidence rate for the 6-MP group. Again, when com-paring groups 1 and 2, there was no identifiable difference.

In the subgroup analysis, an impressivep value of 0.002was seen when comparing group 1A and group 2. However,the 95% confidence interval on the estimated odds of 19.6 is3.1–122, with an N of only 13 in group 1A. A comparisonof 1A versus1B produced an odds of 16 and a 95% CI of1.6–161. Although these intervals technically fit the criteriafor statistical significance, they are far too wide to providea meaningful assessment of the true effect. To find a 5-foldincreased risk of congenital anomalies with 6-MP use, with80% power and a baseline incidence of 3%, this study wouldrequire 55 pregnancies in group 1A and 220 in group 2. Forspontaneous abortions, to find a true 15% rate of spontane-ous abortions with 6-MP use compared to a baseline 10%,a study would require 531 group 1A pregnancies and 1062group 2 pregnancies. If there really is an increased risk ofpregnancy complications with 6-MP use in IBD, we willhave to do a better job at quantifying this risk, and puttingit into perspective, given such a high prevalence rate ofspontaneous abortion and low rate for congenital anomalies.

If we are to change our clinical practice based upon asingle study, it must hold up to the rigors of study design forharm (16). In terms of validity, we do not know that groups1A and 1B are comparable. These are the groups uponwhich the conclusion is based. Were the men that fatheredthe group 1A children sicker than those in group 1B? Per-haps the group 1A fathers all have Crohn’s disease, and theyare being compared to a combination of diseases in group1B. Although the temporal relationship is correct, and ahigher risk is seen in those pregnancies conceived within 3months of 6-MP use, a true dose-response cannot be as-sumed, as the confidence intervals for each group overlap.The strength of their risks is impressive; but the precision ofthe estimate, as evidenced by the wide confidence intervals,is unacceptable.

At this point, this study has not added overwhelmingevidence to suggest any significant increased rate of adverseoutcomes (including spontaneous abortions) with 6-MP use.It is far too premature to recommend discontinuing thesedrugs in men wishing for their partners to conceive. As the

581AJG – March, 2000 Editorials

authors concede, it will take larger studies, performed in aprospective manner, to see what 6-MP use, specific diagno-sis, and disease activity do to pregnancy complication rates.In the meantime, we would do better to counsel our malepatients on the evils of cigarette smoking on spermatogen-esis (17–20).

Sunanda V. Kane, M.D., M.S.P.H.Section of Gastroenterology

University of ChicagoChicago, Illinois

REFERENCES

1. Kornbluth A, Sachar DB. Ulcerative colitis practice guidelinesin adults. American College of Gastroenterology, PracticeParameters Committee. Am J Gastroenterol 1997;92:204–11.

2. Hanauer SB, Meyers S. Management of Crohn’s disease inadults [see comments]. Am J Gastroenterol 1997;92:559–66.

3. Brown JH, Maxwell AP, McGeown MG. Outcome of preg-nancy following renal transplantation. Ir J Med Sci 1991;160:255–6.

4. Gaughan WJ, Moritz MJ, Radomski JS, et al. National Trans-plantation Pregnancy Registry: Report on outcomes in cyclo-sporine-treated female kidney transplant recipients with aninterval from transplant to pregnancy of greater than fiveyears. Am J Kidney Dis 1996;28:266–9.

5. Scantlebury V, Gordon R, Tzakis A, et al. Childbearing afterliver transplantation. Transplantation 1990;49:317–21.

6. Rubaltelli FF. Serum conjugated bilirubin in newborns ofmothers treated with immunosuppressive drugs. Acta PaediatrScand 1972;61:606–8.

7. Williamson RA, Karp LE. Azathioprine teratogenicity: reviewof the literature and case report. Obstet Gynecol 1981;58:247–50.

8. Alstead EM, Ritchie JK, Lennard-Jones JE, et al. Safety ofazathioprine in pregnancy in inflammatory bowel disease.Gastroenterology 1990;99:443–6.

9. Francella A, Dayan A, Rubin P. 6-Mercaptopurine is a safetherapy for child bearing patients with inflammatory boweldisease: A case controlled study. Gastroenterology 1996;110:909.

10. Moody GA, Probert C, Jayanthi V, et al. The effects of chronicill health and treatment with sulphasalazine on fertilityamongst men and women with inflammatory bowel disease inLeicestershire. Int J Colorectal Dis 1997;12:220–4.

11. Penn I, Makowski E, Droegemueller W, et al. Parenthood inrenal homograft recipients. JAMA 1971;216:1755–7.

12. Golby M. Fertility after renal transplantation. Transplantation1970;10:201–7.

13. Rajapakse RO, Korelitz BI, Zlatanic J, et al. Outcome ofpregnancies when fathers are treated with 6-mercaptopurinefor inflammatory bowel disease. Am J Gastroenterol 2000;95:684–8.

14. Congenital malformations surveillance report, Jan. 1982–Dec.1985. Atlanta: US Department of Health and Human Services,1988.

15. Wilcox A, Weinberg CR, O’Connor JF, et al. Incidence ofearly loss of pregnancy. N Engl J Med 1988;319:189–94.

16. Levine M, Walter S, Lee H, et al. Users’ guide to the medicalliterature IV: How to use an article about harm. JAMA 1994;271:1615–9.

17. Vine MF, Margolin BH, Morrison HI, et al. Cigarette smoking

and sperm density: A meta-analysis. Fertil Steril 1994;61:35–43.

18. Potts RJ, Newbury CJ, Smith G, et al. Sperm chromatindamage associated with male smoking. Mutat Res 1999;423:103–11.

19. Rubes J, Lowe X, Moore D II, et al. Smoking cigarettes isassociated with increased sperm disomy in teenage men. FertilSteril 1998;70:715–23.

20. Chia S, Ong CN, Tsakok FM. Effects of cigarette smoking onhuman semen quality. Arch Androl 1994;33:163–8.

Reprint requests and correspondence:Sunanda V. Kane, M.D.,M.S.P.H., University of Chicago, Section of Gastroenterology,Mail Code 4076, 5841 South Maryland, Chicago, IL 60637.

Received Oct. 6, 1999; accepted Dec. 8, 1999.

Doing Battle With HCVHepatitis C virus (HCV) infection has emerged as a majorhealth care problem in the US veteran population. Thegrowing awareness of its importance is reflected in therecent mandate by the Department of Veterans Affairs toperform HCV testing on all veterans at risk for the disease(1) and to manage patients according to the Center forDisease Control and Prevention guidelines (2). Interferon/ribavirin combination therapy is available to VA patients,and Congress has earmarked $230 million for HCV diag-nosis and treatment for the year 2000 VA budget. Gastro-enterologists practicing in the VA system have noticed adrastic increase in HCV-related morbidity over the pastyears. However, the full impact of the disease is unknown.This is due to the lack of epidemiological data that wouldallow one to estimate the number of affected patients and tocalculate the HCV-related expenditures. The study by Ram-sey Cheung in this issue of theAmerican Journal of Gas-troenterology(3) provides a first systematic and detailedreport on HCV antibody seroprevalence, risk factors, geno-types, and age distribution of infected patients. The datawere obtained in a retrospective analysis of the Palo AltoVA Healthcare system. Over a period of 6 yr, (1992–1998),approximately one-fifth of the approximately 40,000 veter-ans living in this predominantly suburban area were testedfor HCV antibodies by a second generation ELISA assay.The test was positive in 2985 subjects, corresponding to apoint prevalence of 35%. The overwhelming majority ofthese patients were found to be viremic, as confirmed byPCR analysis. As pointed out by the authors, the high rate ofseropositivity is likely to be affected by selection and refer-ral bias, as patients were tested based on the presence ofHCV risk factors or abnormal liver tests. To obtain a morerepresentative seroprevalence rate, the author measuredHCV antibodies in 126 consecutive patients involved inblood-borne exposure accidents and obtained a prevalencerate of 18%. This number may still be an overestimate, as itexcludes a majority of veterans who do not use the VA fortheir healthcare. Nevertheless, the reported rate is remark-

582 Editorials AJG – Vol. 95, No. 3, 2000