Résumé
Study question : What is the long-term impact on pregnancy outcomes of a childhood adolescent, and young adult (AYA) exposure to chemotherapy or chemo-radiotherapy?Summary answer : Whereas preterm delivery and pregnancy loss were significantly increased in chemotherapy and chemo-radiotherapy groups compared to control group, live birth rate was not affected.What is known already : The impact of radiotherapy on the uterus has been widely described. Lately, studies have underlined the potential negative effect of chemotherapy on the womb. Indeed, three recent publications have reported that adult females treated with hematopoietic stem cell transplantation (HSCT) in childhood had a reduced uterus volume, both after conditioning by total body irradiation (TBI) and after alkylating agents-based regimen. The pregnancy outcome will depend in part on the uterine volume and its vascularization. Therefore, a history of chemotherapy or chemo-radiotherapy may be associated to an increased risk of obstetrical complications.Study design, size, duration : A systematic review and meta-analysis were performed on comparative studies. Searches were conducted on Medline, Embase and the Cochrane Library from 1990 to April 2023, using the main following search terms: cancer survivors, bone marrow transplantation, chemotherapy, radiotherapy, pregnancy outcome, pregnancy complications, preterm delivery, and pregnancy loss. All studies reporting live birth rate (LBR) in adult females who have received chemotherapy and/or chemo-radiotherapy during childhood and AYA < 25 years were included.Participants/materials, setting, methods : Two independent reviewers carried out the study selection, the bias assessment using the ROBINS-I tool and data extraction. The primary outcome was the LBR. The main secondary outcomes were preterm delivery and pregnancy loss rates. For each outcome, a random-effect frequentist network meta-analysis (NMA) was carried out to estimate risk ratios with their 95% confidence intervals. Sensitivity analyses were conducted by excluding studies at high risk of bias and leave-one out method.Main results and the role of chance : After reviewing 2,328 abstracts, 25 studies were retained for review. Eight studies were included in the meta-analysis on LBR: 3,997 females became pregnant leading to 2,878 live births. The NMA shows a moderate heterogeneity/inconsistency (τ²= 0.012, I²=58% [11%; 80%]) and did not show any significant impact of treatments on LBR. Sensitivity analyses lead to the same conclusions. Six studies reported a total of 125 preterm deliveries out of 1,976 started pregnancies and were included in the NMA (τ²= 0.07, I²=19% [0%; 62%]). Chemo-radiotherapy was significantly associated with higher risk of preterm delivery compared to the control group (3.44 [1.94; 6.09]) and chemotherapy group (1.93 [1.19; 3.15]). The chemotherapy group was associated with a statistically higher risk of preterm delivery than the control group (1.78 [1.01; 3.13]). Eight studies including 606 pregnancy loss out of 3,653 started pregnancies were entered in the NMA (τ²= 0, I²=0% [0%; 65%]). Chemo-radiotherapy was associated with significantly higher risk of pregnancy loss compared to the control group (1.80 [1.34; 2.43]) and chemotherapy group (1.68 [1.26; 2.25]). No statistical difference was shown in the risk of pregnancy loss between the chemotherapy and the control groups (1.07 [0.87; 1.32]).Limitations, reasons for caution : Study selection was focused on live birth rate; therefore, the research was probably not exhaustive for the secondary outcomes. Adjustment on parity or history of hormonal replacement therapy (HRT) was not performed because of a lack of data. Subgroup analysis on the type of chemotherapy was not possible.Wider implications of the findings : This meta-analysis reports a negative impact of chemotherapy or chemo-radiotherapy on obstetric outcomes. Further studies including clear information on parity and use of HRT are needed. Moreover, a description on the type of chemotherapy could help to refine these conclusions.