Health outcomes in offspring born to survivors of childhood cancers following assisted reproductive technologies

Worldwide, > 5,000,000 live births have been reported following assisted reproductive technologies (ART), mostly in vitro fertilization and intracytoplasmic sperm injection [1], with annual numbers rising [2]. In 2016, 2.6% of children in Germany were born following ART [3]. While the majority of children conceived by ART are born healthy, ART has been associated with a higher risk of adverse birth outcomes than spontaneous conception [4]. These include increased occurrences of adverse perinatal events [5], childhood cancer [6], congenital malformations [7] (including heart defects [8]), hypertension [9, 10], and asthma [11]. Hypotheses explaining these observations implicate pharmaceutically induced ovulation, micromanipulation during ART procedures, effects associated with increased incidence of multiple sibling births [4], and subfertility itself [12].

Infertility is a known late effect of cancer treatment in childhood [13‐15], yet national [16, 17] and international [18, 19] guidelines on fertility protection have only been developed in recent years. An increasing number of childhood cancer survivors are employing ART to fulfill the wish for biological children [20]. Survivors report being anxious about the general health of future children and the possibility that their previous cancer treatment may increase the risk of cancer in offspring [21]. Studies have shown that offspring of non-hereditary childhood cancer survivors do not have an increased risk for malformations, genetic diseases, or non-hereditary cancers [22‐24]. Yet, whether ART affects offspring health when used by childhood cancer survivors has not been specifically addressed.

Within our FeCt¹Multicenter Offspring Study, we compared perinatal outcomes and the prevalence of childhood cancer and congenital malformations, including heart defects, in children born to German survivors following ART and after spontaneous conception. To contextualize our findings, we also compared health outcomes in the survivor offspring cohort to children in the general German population.

Here, we specifically investigate the impact of ART on the health of offspring born to childhood cancer survivors using the only survey information published to date that was tailored to this question. Nearly twice the childhood cancer survivors used ART in our surveyed cohort (4.6%) compared with numbers reported by the German In Vitro Fertilization Registry (2.6%) [34]. This distribution parallels findings from Melin et al., who compared fertility treatments among female survivors of adult cancer and their siblings (5.2% vs. 2.8%) [35]. While current guidelines recommend fertility preservation before cancer treatment [36], patients in our cohort were treated (1980–1999) before its routine implementation. Only few patients cryopreserved oocytes/sperm prior to cancer treatment. In line with results from a recent European study [37], more male survivors cryopreserved, likely because sperm collection is less invasive and costly, but possibly also due to differences in counseling [38]. Our data showed that even in cases lacking pretreatment oocytes/sperm cryopreservation, successful ART pregnancies occurred in childhood cancer survivors. The increasing numbers of childhood cancer survivors who turn to ART [34] stress the importance of establishing an information base from which to counsel childhood cancer patients and survivors.

Our analyses revealed that the prevalence of multiple sibling births after ART was nearly tenfold higher than for children conceived spontaneously (28.1% vs. 3%). The higher prevalence of multiple sibling births after ART parallels the 34% reported for the general population in 2016 by the German In Vitro Fertilization Registry [34] and is in line with the 31.5% prevalence reported in 2016 in the USA [39]. Recent studies describe a mediating effect of multiple sibling births after ART on adverse obstetric outcomes [4, 5], indicating relevance for health also reflected in our findings from multivariate analyses on perinatal outcomes (p < 0.001; OR = 3.902, 95% CI = 1.804 to 8.441, Supplementary Table 1). When known confounders including multiple sibling birth were taken into account [4, 5, 40‐43], perinatal outcomes were no different in survivor offspring, whether conceived after ART or spontaneously. We detected a modest increase in the prevalence for moderate preterm birth (32 to < 37 weeks of gestation) in childhood cancer survivors compared to the KiGGS cohort, as a representative for the general German population. These findings are reassuring, since most medical consequences occur in very (28 to < 32 weeks of gestation) or extremely (< 28 weeks of gestation) preterm infants [44]. Recent studies have cited long-term treatment effects as a possible source for their detected increased prevalence of adverse pregnancy outcomes in childhood cancer survivors. Van de Loo et al. and van Dorp et al. both found that childhood cancer survivors exposed to radiation were at higher risk for preterm delivery and low birth weight [24, 45]. The majority of survivors in our collective (58.7%) were treated in the 1980s, typically receiving higher doses of radiation and alkylating agents (in multiagent chemotherapy schedules) than are currently used [46]. Against this backdrop of progressively reduced toxicity regimens, our findings appear particularly reassuring for patients treated with the less toxic protocols for childhood cancer today. Our study confirms the known increase in multiple sibling births following ART and indicates that prevalence was no higher in childhood cancer survivors than in the general German population. Adverse obstetric outcomes in survivors, as described by preceding studies, were only reflected by a small increase of moderate preterm births in our survivor cohort.

We found that ART did not raise the prevalence of childhood cancer among survivor offspring. In this respect, currently available data collected for the general population are not consistent. Williams et al. reported no increase in the overall risk for cancer among 106,013 children conceived through ART in the British general population, while the risk for hepatoblastoma and rhabdomyosarcoma appeared slightly increased [47]. Likewise, Spaan et al. reported no increase in the overall cancer risk among 24,269 children conceived after ART in the general Dutch population after a median follow-up of 21 years [48]. A meta-analysis of 25 cohort and case-control studies published in 2013, however, calculated a slightly elevated overall risk for cancer in children born after ART (RR = 1.33, 95%CI = 1.08–1.63) [49]. The authors pointed out that subfertility and potential epigenetic defects in the gametes, rather than the ART procedure itself, might be the most important predisposing factor for childhood cancer underlying these data [49]. While a meta-analysis by Hoorsan et al. from 2017 arrived at a 53% higher risk for congenital malformations following ART in the general population [50], our data showed no increased risk for congenital malformations or heart defects among offspring born to childhood cancer survivors who used ART. Hoorsan et al. discuss genetic characteristics and conditions specifically occurring in infertile couples as causative, which may differ from characteristics of our study population. When comparing the prevalence of childhood cancer and congenital malformations in survivor offspring with children from the KiGGS Study, whether conceived by ART or spontaneous conception, our study confirmed findings from large studies [22‐24], that showed no higher prevalence for childhood cancer in survivor offspring than in children in the general population. Similarly, the prevalence of congenital malformations, including heart defects, was not elevated in offspring born to our survivor cohort than the reported prevalence in the general population, in line with currently available data [51‐55]. Our findings in the survivor cohort and in comparisons with the general German population support that the use of ART by childhood cancer survivors does not put offspring at additional risk for adverse health outcomes including childhood cancer or congenital malformations.

The study setting and conduct feasibility among childhood cancer survivors in Germany posed certain limitations. Recruitment was based on previous surveys identifying survivors with biological children, potentially causing a selection bias. This approach was necessary to reduce the study burden for survivors, as required by the German Society for Pediatric Oncology and Hematology. The questionnaire-based setting could produce recall bias that could reduce data accuracy. However, all survivor parents had been treated according to standardized trial protocols, for which detailed treatment information was available through the German Society for Pediatric Oncology and Hematology. Although we were able to examine a number of potential mediating factors, we had no information on maternal age, body mass index, and infections during pregnancy, which are further factors influencing perinatal events [46, 56‐58]. Missing data of the main outcomes were rare in survivor offspring: 1.79% (0.08–3.95%) but more prevalent in the KiGGS cohort: 10.78% (1.70%–13.56%). Although we have no indication of this, it cannot be ruled out with certainty that these were not completely at random. Despite the explorative character of this study, which does not allow for confirmatory conclusions, our analyses offer new insights into health issues in offspring born to childhood cancer survivors, and the high response rate reflects the strong interest shown by survivors in these issues. The small sample size of 57 survivor offspring born after ART was adequate to detect medium effects (d 0.5 or higher). Future studies are needed to further explore the effect of ART within a larger population. Our study delivers encouraging results for survivors of childhood cancer that demonstrate that the vast majority of offspring born to survivors do not experience adverse perinatal outcomes or later health problems, independently of whether conception was spontaneous or required ART.