No abstract

No abstract

IMPORTANCE An increasing number of children worldwide are born after the use of fertility treatment, although it remains unclear whether the treatment affects the risk of childhood cancer and whether any associations observed are due to the use of specific drugs, the use of specific procedures, or the underlying infertility. OBJECTIVE To examine the association between different types of fertility treatments and cancer risk in children. DESIGN, SETTING, AND PARTICIPANTS A retrospective cohort study based on Danish population-based registry data and the Danish Infertility Cohort (individual record linkage) that included 1 085 172 children born in Denmark between January 1, 1996, and December 31, 2012, linked with parental information. There were a total of 2217 children diagnosed with cancer (follow-up occurred during 1996-2015). EXPOSURES Maternal fertility treatment during the index pregnancy, including the use of fertility drugs (clomiphene [n = 33 835], gonadotropins [n = 57 136], gonadotropin-releasing hormone analogs [n = 38 653], human chorionic gonadotropin [n = 68 181], progesterone [n = 41 628], and estrogen [n = 16 948]) and assisted reproductive technology (in vitro fertilization [n = 19 448], intracytoplasmic sperm injection [n = 13 417], and frozen embryo transfer [n = 3356]). Each exposure was examined separately and compared with children born to fertile women. MAIN OUTCOMES AND MEASURES Hazard ratios and incidence rate differences for childhood cancer. RESULTS After 12.2 million person-years of follow-up (mean, 11.3 years), the incidence rate of childhood cancer was 17.5 per 100 000 for children born to fertile women (n = 910 291) and 44.4 per 100 000 for children born after the use of frozen embryo transfer (n = 3356). Compared with children born to fertile women, the use of frozen embryo transfer was associated with an elevated risk of childhood cancer (14 cancer cases; hazard ratio, 2.43 [95% CI, 1.44 to 4.11]; incidence rate difference, 26.9 [95% CI, 2.8 to 51.0] per 100 000), mainly due to an increased risk of leukemia (5 cancer cases; incidence rate, 14.4 per 100 000; hazard ratio, 2.87 [95% CI, 1.19 to 6.93]; incidence rate difference, 10.1 [95% CI, −4.0 to 24.2] per 100 000) and sympathetic nervous system tumors (<5 cancer cases; hazard ratio, 7.82 [95% CI, 2.47 to 24.70]). There were no statistically significant associations with the use of the other types of fertility treatment examined. CONCLUSIONS AND RELEVANCE Among children born in Denmark, the use of frozen embryo transfer, compared with children born to fertile women, was associated with a small but statistically significant increased risk of childhood cancer; this association was not found for the use of other types of fertility treatment examined.

Large population-based studies are needed to examine the effect of maternal use of fertility drugs on the risk of cancer in children, while taking into account the effect of the underlying infertility. A cohort of 123,322 children born in Denmark between 1964 and 2006 to 68,255 women who had been evaluated for infertility was established. We used a case-cohort design and calculated hazard ratios (HRs) for cancer in childhood (0-19 years) and in young adulthood (20-29 years) associated with maternal use of six groups of fertility drugs (clomiphene, gonadotropins [i.e., human menopausal gonadotropins and follicle-stimulating hormone], gonadotropin-releasing hormone analogs, human chorionic gonadotropins, progesterone and other fertility drugs). We found no statistically significant association between maternal use of fertility drugs and risk for overall cancer in childhood or young adulthood. However, with regard to specific cancers in childhood, our results showed that maternal use of progesterone before childbirth markedly increased the risks of their offspring for acute lymphocytic leukemia (any use: HR, 4.95; 95% CI, 1.69-14.54; three cycles of use: HR, 9.96; 95% CI, 2.63-37.77) and for sympathetic nervous system tumors (any use: HR, 5.79; 95% CI, 1.23-27.24; three cycles of use: HR, 8.51; 95% CI, 1.72-42.19). These findings show that maternal use of progesterone may increase the risk for specific cancers in the offspring. Additional large epidemiological studies are urgently needed to confirm our finding.

The study was supported by internal funding from the Unit of Virus, Lifestyle and Genes at the Danish Cancer Society Research Center. All authors report no conflicts of interest.