Bone Marrow Transplantation (2006) 37, 1023–1029 & 2006 Nature Publishing Group All rights reserved 0268-3369/06 $30.00 www.nature.com/bmt ORIGINAL ARTICLE Prevalence of conception and pregnancy outcomes after hematopoietic cell transplantation: report from the bone marrow transplant survivor study A Carter1, LL Robison2, L Francisco1, D Smith1, M Grant1, KS Baker3, JG Gurney4, PB McGlave3, DJ Weisdorf3, SJ Forman5 and S Bhatia1,5 1Population Sciences, City of Hope Cancer Center, Duarte, CA, USA; 2Epidemiology and Cancer Control, St Jude Children’s Research Hospital, Memphis, TN, USA; 3Blood and Marrow Transplant Program, University of Minnesota, Minneapolis, MN, USA; 4General Pediatrics, University of Michigan, Ann Arbor, MI, USA and 5Hematology and Hematopoietic Cell Transplantation, City of Hope Cancer Center, Duarte, CA, USA We conducted a retrospective study to describe the malignant disorders, with over 45 000 transplants per- magnitude of compromise in reproductive function and formed worldwide annually.1 The widespread use of HCT investigate pregnancy outcomes in 619 women and can be attributed to its improved treatment efficacy, wider partners of men treated with autologous (n ¼ 241) or variety of stem cell sources, and improved transplantation allogeneic (n ¼ 378) hematopoietic cell transplantation strategies and supportive care, and has resulted in a (HCT) between 21 and 45 years of age, and surviving 2 or growing number of long-term survivors.1,2 Most pre- more years. Median age at HCT was 33.3 years and transplant conditioning regimens for HCT include alkylat- median time since HCT 7.7 years. Mailed questionnaires ing agents and/or irradiation, both of which may cause captured pregnancies and their outcomes (live birth, germ cell injury, gonadal dysfunction and infertility.3,4 The stillbirth, miscarriage). Thirty-four patients reported 54 risk of gonadal damage is higher with increasing doses of pregnancies after HCT (26 males, 40 pregnancies; eight alkylating agents, total body irradiation (TBI), and older females, 14 pregnancies), of which 46 resulted in live age in women at the time of transplant.5 births. Factors associated with reporting no conception Although successful pregnancies after HCT have been included older age at HCT (X30 years: odds ratio reported previously,6–11 the evidence suggests that female (OR) ¼ 4.8), female sex (OR ¼ 3.0), and total body survivors may be at an increased risk of spontaneous irradiation (OR ¼ 3.3). Prevalence of conception and abortions and miscarriages, pre-term delivery and low pregnancy outcomes in HCT survivors were compared to birth weight babies.12,13 To further investigate pregnancy those of 301 nearest-age siblings. Although the risk for not outcomes and describe the magnitude of compromise in reporting a conception was significantly increased among reproductive function, as indicated by prevalence of self- HCT survivors (OR ¼ 36), survivors were not significantly reported pregnancy occurrence and pregnancy outcomes, more likely than siblings to report miscarriage or stillbirth we utilized the cohort of survivors participating in the Bone (OR ¼ 0.7). Although prevalence of conception is dimin- Marrow Transplant Survivor Study (BMTSS) and com- ished after HCT, if pregnancy does occur, outcome is pared their experience with a sibling comparison group. likely to be favorable. Patients should be counseled prior to transplant regarding strategies to preserve fertility. Bone Marrow Transplantation (2006) 37, 1023–1029. Materials and methods doi:10.1038/sj.bmt.1705364; published online 10 April 2006 Keywords: late effects; conception; pregnancy outcomes; The BMTSS, a collaborative effort between the City of hematopoietic cell transplantation Hope Cancer Center and the University of Minnesota, examines the long-term outcomes of individuals who have survived 2 or more years after undergoing HCT, as well as Introduction similar outcomes in a cohort of siblings with a similar age and gender distribution. The aim of the current study was Hematopoietic cell transplantation (HCT) is being increas- to determine the prevalence of post-HCT conception, and ingly used for the treatment of malignant and non- outcome of pregnancy in 619 adult women and partners of men previously treated with autologous or allogeneic HCT. The present report from BMTSS is restricted to individuals who met the following eligibility criteria: (i) HCT between Correspondence: Dr S Bhatia, City of Hope Cancer Center, 1500 East 1974and 1998 at the City of Hope or University of Duarte Road, Duarte, CA 91010-3000, USA. E-mail: [email protected] Minnesota; (ii) age at transplant 21–45 years; and, (iii) Received 24February 2006; revised 8 March 2006; accepted 9 March survival of at least 2 years from HCT, irrespective of 2006; published online 10 April 2006 disease status. The Human Subjects Committees at the Pregnancy after hematopoietic cell transplantation A Carter et al 1024 participating institutions approved the BMTSS protocol. presence of chronic graft-vs-host disease (cGVHD), im- Informed consent was provided according to the Declara- munosuppressive agents used for GVH prophylaxis (yes vs tion of Helsinki. Data collection included completion of a no) and report of pre-transplant pregnancy ending in live 255-item mailed questionnaire designed to capture a wide birth (yes vs no). Patients considered being at standard risk range of demographic characteristics and health informa- for relapse included those in first or second complete tion. Specific to this report, participants provided a remission after acute leukemia or lymphoma, those with complete pregnancy history by supplying details about all severe aplastic anemia, and first chronic phase of CML. All pregnancies and outcomes both before and subsequent to other patients were placed into the high-risk category. HCT. Data collected included age at beginning of each pregnancy, weeks the pregnancy lasted, date of delivery and Pregnancy outcomes outcome of each pregnancy (i.e. whether the pregnancy Risk factors were analyzed for the occurrence of a live birth resulted in a livebirth, stillbirth, spontaneous miscarriage or or the occurrence of an adverse pregnancy outcome medical abortion). Only post-HCT pregnancies were used (stillbirth or miscarriage). for the purpose of the current analysis. The two primary outcomes of interest were inability to conceive after Comparison with siblings. Potential risk factors for transplant, and pregnancy outcomes (i.e. livebirths, still- adverse pregnancy outcomes (miscarriage or stillbirth) births and spontaneous miscarriages). Pregnancies in included age at study participation (younger than 45 years gestation were not included in outcome analyses. We vs 45 years or older), sex, race, annual household income, compared the information obtained from the HCT availability of health insurance and highest level of survivors to that obtained from 301 closest-age siblings education. w2 tests were performed to determine whether who completed an identical BMTSS questionnaire. Ana- the rate of pre-term delivery (p37 weeks gestation) differed lyses were restricted to pregnancies occurring between the with respect to the sibling controls.16 ages of 21 and 45 for both survivors and the sibling In order to account for multiple pregnancies in or by one comparison group. survivor or one sibling (within person correlation), we used generalized estimating equations with a binominal distribu- Statistical analysis tion and a logit link to estimate the relative odds of Standard parametric and non-parametric tests were used to pregnancy outcomes among survivors compared to sib- lings. Our models also included a variance component for report descriptive analyses. To test for associations between 14,17–19 ordinal and categorical end points, we applied w2 tests for within family correlation. independence. The primary outcomes of interest were inability to conceive (yes vs no) and pregnancy outcomes (live births, miscarriages and spontaneous abortions). Data Results were analyzed with SAS version 9.1 (SAS Institute, Cary, NC, USA).14 Of the 978 patients eligible for participation in this study, 897 were successfully contacted with 619 (69%) agreeing to participate (Table 1). The median age at study participation Conception was 42.6 years (range 23.3–60.4 years), median age at Comparison with siblings. Potential risk factors for not transplant 33.3 years (range 21.0–45.0) and median length reporting a conception were analyzed using unconditional of follow-up 7.7 years (2.0–24.4 years). Twenty-nine logistic regression and included age at study participation, percent of participants had received HCT for CML, sex, race, annual household income (o$20 000 vs $20 000– 22.9% for AML, and 29.9% for HD or NHL. $60 000 vs 4$60 000), availability of health insurance, and Comparison of study participants with the potentially highest level of education (high school or less vs high school eligible non-participants showed that participants were degree with or without some college or training after high more likely to be female (47 vs 39%, P ¼ 0.02), and have school, vs college graduate). Odds ratio15 was used to received a conditioning regimen containing TBI (81 vs estimate magnitude of association between the risk of not 73%, P ¼ 0.003). Participants did not differ statistically reporting a conception and the potential risk factors. from non-participants in other conditioning agents re- ceived, race/ethnicity, primary diagnosis and length of Within survivor comparison.