Conditioning intensity and probability of live birth after blood or marrow transplantation, a BMTSS report

Over 20 000 blood or marrow transplants (BMTs) are performed in the United States every year.^1,2  Improved therapeutic efficacy, ready availability of a wide variety of stem cell sources, and improved transplant strategies (such as nonmyeloablative and reduced-intensity conditioning) have resulted in an increase in the use of BMT as a curative option for a variety of hematologic malignancies and other life-threatening illnesses.³  Transplant conditioning regimens often use total body irradiation (TBI) and/or alkylating agents; both are associated with germ cell injury, gonadal dysfunction, and either transient or permanent infertility.^4-8  The observed interindividual variability in the probability of live birth after BMT is likely dependent on the conditioning intensity (myeloablative condition [high intensity], or nonmyeloablative or reduced-intensity conditioning [low intensity]), age at exposure to gonadotoxic agents, sex of the BMT recipient, as well as post-BMT morbidity (chronic graft-versus-host disease [cGVHD] and other chronic health conditions).⁹  While the prevalence of live birth after BMT has been examined previously,^7,10,11  the reports are limited by either small samples or lack of contemporary cohorts and therefore an inability to capture the impact of the newer transplant strategies (eg, conditioning intensity). Finally, the difference in pregnancy outcomes (live births, abortions, and stillbirths) when compared with biologic siblings remains unknown. We addressed these gaps by utilizing the resources offered by the BMT Survivor Study (BMTSS).

BMTSS is a collaborative effort between the University of Alabama at Birmingham, City of Hope, and University of Minnesota established to examine the long-term outcomes of individuals who have survived 2 or more years after undergoing BMT between 1974 and 2014. BMTSS also examines comparable outcomes in an unaffected comparison group composed of siblings of the BMT survivors. The Institutional Review Board (IRB) at University of Alabama at Birmingham serves as the single IRB of record; the IRBs at University of Minnesota and City of Hope have approved the BMTSS protocol. Participants have provided informed consent according to the Declaration of Helsinki. The present report includes survivors who received BMT at age ≤45 years and were ≥18 years of age at study participation. Participants in the comparison cohort were ≥18 years at study participation.

Study participation consisted of completion of the BMTSS survey designed to capture demographic characteristics and health information. Sociodemographic (race/ethnicity, annual household income, availability of health insurance, level of education), chronic health conditions as diagnosed by the health care provider, and whether the survivors consulted a doctor/took medication to help become pregnant were retrieved from the BMTSS survey. Participants (BMT survivors and siblings) provided a complete reproductive history by supplying details about all pregnancies and their outcomes during their lifetime. Data collected included age at start of each pregnancy and outcome of each pregnancy (ie, whether the pregnancy resulted in a live birth, stillbirth, or abortion [spontaneous or medical]) (supplemental Table 4). Chronic health conditions diagnosed after BMT were graded using the Common Terminology Criteria for Adverse Events, Version 5.0.¹²  Survivors’ age at BMT, sex, primary diagnosis, type of transplant (autologous; allogeneic), risk of relapse at BMT (standard risk; high risk), conditioning regimens, TBI dose (200 to 800 cGy, >800 cGy) and conditioning intensity (high or low), pre-BMT exposure to pelvic/testicular radiation or alkylating agents, and history of cGVHD were retrieved from the institutional transplant databases and/or participants’ medical record.

Of the 3769 BMT survivors eligible for participation, 832 (22.1%) were lost to follow-up. Of the 2937 patients contacted, 1052 (33.8%) refused participation, and 278 (9.5%) completed an abbreviated survey that did not include reproductive history. Out of the 1607 participants who completed the full version of the questionnaire, 32 were excluded as they underwent hysterectomy before BMT, and 195 were excluded because they had never been sexually active, yielding 1375 evaluable participants (46.8% of those successfully contacted and meeting inclusion criteria) (Figure 1). When compared with nonparticipants, participants were older at BMT (median age [years]: 31 vs 23), more likely to be female (47.6% vs 39.1%, P = .004) and non-Hispanic White (72.5% vs 64.1%, P < .0001), more likely to have received conditioning with TBI (57.3% vs 46.6%, P < .0001), MAC (80.0% vs 72.8%, P < .0001), and more likely to have had a history of cGVHD (48.7% vs 30.5%, P < .0001). Participants were less likely to have received conditioning with cytarabine (2.9% vs 4.6%, P = .03), melphalan (7.8% vs 14.2%, P < .0001), fludarabine (12.3% vs 15.2%, P = .03), and cyclophosphamide (66.5% vs 73.1%, P = .0006). Participants did not differ from nonparticipants with respect to other conditioning agents or disease status at BMT (supplemental Table 1).

Of the 1239 siblings who completed the full version of the questionnaire, 200 were excluded because they had received a hysterectomy, and 46 were excluded as they had never been sexually active, yielding 1039 evaluable siblings. For the comparison with siblings, we excluded 82 survivors who had received a hysterectomy at any time during their life and 195 who had never been sexually active from the 1607 BMT survivors who had completed the full questionnaire, yielding 1330 survivors. In addition to comparing all evaluable siblings (related or unrelated) with all BMT survivors, we were able to identify 172 survivors matched 1:1 with their nearest-age and same-sex biological sibling to control for genetic or environmental factors that could affect fertility (Figure 2).

For bivariate analyses, we used χ-square tests for categorical variables and t tests for continuous variables.

Potential risk factors for not reporting a live birth after BMT were analyzed using multivariable logistic regression. The risk factors examined included age at BMT, sex, race/ethnicity, grade 3 (severe) or grade 4 (life-threatening) chronic health conditions, annual household income, education, type of BMT (autologous or allogeneic), whether the survivor/partner tried for ≥12 months to become pregnant or took medications to help become pregnant, disease status at BMT, history of pelvic/testicular radiation and/or exposure to alkylating agents prior to BMT, and history of live birth prior to BMT. Odds ratio (OR) and associated 95% CI was used to estimate the magnitude of association between the risk of not reporting a live birth and potential risk factors.

BMT survivors were placed into 4 groups based on TBI dose and conditioning intensity: no TBI/low-intensity conditioning (patients who did not receive TBI but received low-intensity conditioning), 200 to 800 cGy TBI/low-intensity conditioning (patients who received 200 to 800 cGy TBI in the context of low-intensity conditioning), >800 cGy TBI/high-intensity conditioning (patients who received >800 cGy TBI in the context of high-intensity conditioning), and no TBI/high-intensity conditioning (patients who did not receive TBI but received high-intensity conditioning).

We matched 172 survivors with their closest-age, same-sex biologic sibling, and used conditional logistic regression (adjusted for grades 3 to 4 chronic health conditions, age at study, and history of seeking fertility assistance by consulting a doctor or taking fertility-promoting medications) to determine the magnitude of risk of pregnancy outcomes in BMT survivors when compared with their biologic siblings.

We compared BMT survivors (n = 1330) with the entire sibling comparison group (n = 1039) by using logistic regression to determine the magnitude of difference between survivors and siblings in (1) failure to report live birth and (2) pregnancy outcomes. We adjusted the analysis for sex, race/ethnicity, history of seeking fertility assistance by consulting a doctor or taking fertility-promoting medications, and grades 3 to 4 chronic health conditions.

Of the 1375 survivors, 541 were autologous BMT recipients (39.3%), and 641 were female (46.6%). The median age at transplantation was 31 years (IQR, 0-45), and at study participation was 46 years (IQR, 18-73). The median length of follow-up from BMT to BMTSS survey completion was 14.3 years (IQR, 2.4-41.4). The primary indications for BMT included Hodgkin or non-Hodgkin lymphoma (HL/NHL: 32.6%), acute lymphoblastic leukemia/acute myeloid leukemia/myelodysplastic syndrome (ALL/AML/MDS: 26.7%), severe aplastic anemia (SAA: 6.8%), and other diagnoses (23.9%).

One hundred and twenty BMT survivors reported a live birth after BMT (8.7%). Provided in Table 1 are the clinical characteristics of the 120 survivors who reported post-BMT live birth and 1255 survivors who did not. The prevalence of live birth did not vary by sex (P = .5), history of pre-BMT pelvic/testicular radiation (P = .3), or pre-BMT alkylating agent (P = .4) exposure. Multivariable analysis revealed that older age at BMT (>35 years: OR, 3.9; 95% CI, 1.32-11.80; reference: age at BMT <11 years), no medical interventions to facilitate pregnancy (OR, 3.1; 95%CI, 1.97-4.74), and >800 cGy TBI/high-intensity conditioning (OR, 3.7; 95% CI, 1.9-7.0; reference: no TBI/low-intensity conditioning) were associated with not reporting a post-BMT live birth (Table 2; Figure 3).

Compared with their nearest-age and same-sex biologic siblings, BMT survivors did not differ in annual household income, educational status, availability of current health insurance, or history of trying medical interventions to facilitate pregnancy (Table 3). BMT survivors were more likely to report no live birth (56.4% vs 40.1%, P = .003) but less likely to report spontaneous abortion (9.9% vs 20.9%, P = .005). The prevalence of stillbirth or medical abortion was comparable between survivors and their biologic siblings.

After adjusting for history of trying medical interventions to facilitate pregnancy, grades 3 to 4 chronic health conditions, and age at completing the questionnaire, we found that survivors were 2 times more likely to report no live birth (OR, 2.0; 95% CI, 1.2-3.3) compared with their matched biologic siblings. In contrast, survivors were less likely to report spontaneous abortions compared with their biologic siblings (OR, 0.44; 95% CI, 0.23-0.83) (Table 4).

In supplemental Table 2, we provide a comparison of the clinical and demographic characteristics between BMT survivors and all siblings (related or unrelated) that served as a comparison cohort. The siblings were older (median age: 57.7 years vs 46 years), and more likely to be female (54.8% vs 45.0%), non-Hispanic White (86.5% vs 71.9%), better educated (>high school education: 89.2% vs 84.7%), have health insurance (97.8% vs 96%), and higher income (annual household >$100 000 [42.7% vs 30.9%]). BMT survivors were more likely not to have tried medical interventions to facilitate pregnancy (77.8% vs 66.9%, P < .0001), but were less likely to report pregnancies resulting in live birth (42.6% vs 70.5%, P < .0001), spontaneous abortion (14.2% vs 19.5%, P = .0003), and medical abortion (10.5% vs 14.0%, P = .009), compared with the sibling cohort. Survivors were more likely to have grade 3 or 4 chronic health conditions compared with the sibling cohort (37.7% vs 32.5%, P = .005). Multivariable logistic regression (supplemental Table 3) after adjusting for sex, race/ethnicity, grades 3 to 4 chronic health conditions, and history of trying medical interventions to facilitate pregnancy showed that BMT survivors were more likely to report no live birth (OR, 3.0; 95% CI, 2.5-3.6), and were less likely to report spontaneous abortion (OR, 0.7; 95% CI, 0.59-0.92) or medical abortion (OR, 0.7; 95% CI, 0.55-0.94) when compared with the sibling comparison group.

In this study, we found that BMT survivors were 2 times more likely to report no live birth when compared with their biologic same-sex siblings. Nonetheless, 120 BMT survivors reported post-BMT live births, representing 8.7% of the survivor population. Among BMT survivors, the risk for not reporting a live birth after BMT was significantly higher among those who were older at BMT, those who had received >800 cGy TBI in the context of high-intensity conditioning, and those who did not try medical interventions to facilitate pregnancy.

Importantly, there was no difference in the live birth status between patients who did not receive TBI in the setting of high-intensity conditioning and those who received low-intensity conditioning (with or without TBI). Rather, it was the full-dose TBI (in the setting of high-intensity conditioning) that was associated with the lower probability of post-BMT live birth. The association between full-dose TBI and impaired reproductive health is known. Loren and colleagues⁸  reported higher pregnancy complications in female allograft recipients compared with a non-BMT population, particularly among those receiving TBI-containing conditioning regimens. However, ours is the first report to find that the probability of live birth among those exposed to low-dose TBI is comparable to patients who received chemotherapy-based high- or low-intensity conditioning.

Similar to this study, previous studies have reported that patients who underwent BMT at a younger age were more likely to report post-BMT live birth^7,10,13,14 ; this is consistent with the observed association between increasing age and gonadal dysfunction after BMT.^15,16  Our study found that female survivors or partners of male survivors who tried medical intervention to become pregnant were more likely to report post-BMT live birth when compared with those who did not. Although data on types of medications were not available, previous studies suggested that survivors benefit from fertility-enhancing interventions.¹⁷ 

Comparison with biologic siblings suggested that although survivors were less likely to report live birth compared with their biologic siblings, they were also less likely to report spontaneous abortion. These findings suggest that although the prevalence of live birth is significantly diminished after BMT if pregnancy occurs, the outcome is likely to be favorable.

Findings in this study need to be placed in the context of the limitations. It was difficult to determine the true magnitude of infertility after BMT for a variety of reasons.¹¹  For example, we did not elicit information regarding attempts at preserving fertility (sperm or oocyte cryopreservation, in vitro fertilization),¹⁸  although we did ask the survivors about needing medications to facilitate pregnancy. Second, we did not have details regarding gonadal dysfunction. Third, we did not elicit information regarding the desire to become pregnant. These limitations notwithstanding, this is the first study to compare reproductive health in BMT survivors to their closest-age, same-sex biologic siblings, allowing us to control for biologic and sociodemographic factors. In addition, this is the first study to examine the impact of conditioning intensity on post-BMT fertility. Pre-BMT counseling to discuss assisted reproductive techniques and sexual health is highly recommended, especially among young patients. Alternative chemotherapeutic regimens with lower doses of TBI could be considered for patients who wish to have children after BMT.

This study was funded by National Cancer Institute, U01 CA213140 and Leukemia Lymphoma Society, and R01CA078938.

Contributions: S.B., S.J.F., D.J.W., S.H.A., and M.A. contributed to study conception and design; S.B., L.H., J.W., E.S., N.B., A.B., H.S.T., L.F., S.H.A., and M.A. collected and assembled the data; N.B. and S.B. analyzed and interpreted the data; N.B. and S.B. drafted the manuscript; S.B., L.H., W.L., J.W., E.S., L.F., and N.B. provided administrative, technical, or material support; S.B., M.A., and S.H.A. supervised the study; and all authors critically reviewed the manuscript for important intellectual content and approved the final manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Smita Bhatia, Institute for Cancer Outcomes and Survivorship, School of Medicine, University of Alabama at Birmingham, 1600 7^th Ave South, Lowder 500, Birmingham, AL 35233; e-mail: smitabhatia@uabmc.edu.