Full Citation

Title: Maternal Contraceptive Use and Central Nervous System Tumors in Offspring

Citation Type: Journal Article

Publication Year: 2022

ISBN: 4158801905317

ISSN: 0098-7484

DOI: 10.1001/JAMA.2021.21809

NSFID:

PMCID:

PMID: 34982137

Abstract: Childhood central nervous system (CNS) tumors represent serious diseases. Due to their relative intractability, CNS tumors have surpassed leukemias as the leading cause of death due to cancer in children in the US. 1 From 2014 to 2018, the incidence rate for childhood CNS tumors was approximately 55 cases per million children per year, and the most common tumor types were as-trocytoma and medulloblastoma. 2 In this issue of JAMA, Hargreave et al 3 examined the association between maternal hormonal contraception use and CNS tumors in their offspring. The authors used data from a population-based registry from Denmark and included 1185 063 children born between 1996 and 2014. During 15 335 990 person-years of follow-up (median follow-up, 12.9 years), 725 cases of childhood CNS tumors were diagnosed. Among the mothers, 136 022 had recent hormonal contraception use, 778 843 had previous use, and 270 198 had no use. Combined hormonal contraception showed no association with CNS tumors (incidence rate, 5.0 per 100 000 person-years for children born to mothers with hormonal contraception use ≤3 months before the start of and during pregnancy vs 5.3 per 100 000 person-years for children born to mothers with no use; hazard ratio, 0.95 [95% CI, 0.74-1.23]). No significant associations were observed for different regimens (ie, combined , progestin only) or route of administration (oral, non-oral). The only significant finding was among recent use of progestin-only injections (n = 885 cases), with less than 5 cases of CNS tumors among women with recent use vs no use (haz-ard ratio, 6.74 [95% CI, 2.15-21.15]). However, as the authors acknowledged, there are many reasons to be cautious about the latter finding, including the very small number of exposed children and cases. That other nonoral progestin-only contraceptives were not associated with CNS tumors, with a negative likelihood test indicating that the products in this class may not differ in terms of risk, also suggests no elevation in risk. This may have been a chance finding among the many observations made in the study. Is it biologically plausible that maternal hormonal contraceptives , particularly progestin-only products, could increase the risk of childhood CNS tumors? 4 The authors pointed out that "…hormonal exposure in utero is recognized to cause cancer in exposed offspring," referring to the example of diethylstilbestrol. Maternal diethylstilbestrol ingestion, injection , or vaginal application during fetal organogenesis in the first trimester has been associated with congenital genitouri-nary abnormalities and cervicovaginal clear cell carcinoma in offspring. 5 Diethylstilbestrol readily crosses the placenta and acts through both an estrogen-dependent pathway (ie, through estrogen receptors) and an epigenetic pathway. A second example is 17-α hydroxyprogesterone caproate, a pro-gestin administered during pregnancy to prevent miscarriage in the 1950s and 1960s and preterm birth in the 2000s. In utero exposure has been suggested to increase the risk of cancer in offspring. 6 Based on these examples, it is not entirely implausible that a similar hormone taken during pregnancy might possibly cause cancer in offspring, but additional conditions would have to be met. First, can the hormone cross the pla-centa, and at what stage of pregnancy? The natural analogue of progestins, progesterone, is critical to implantation and has receptors on trophoblast cells. As part of the luteal-placental transition between 4 to 6 weeks after ovulation (ie, 6-8 weeks' gestation), the placenta is able to synthesize pro-gesterone as well as other enzymes to metabolize steroids such as progesterone by the end of the first trimester. Proges-terone can have bidirectional transplacental passage through the mid-to late-gestation placenta in nonhuman primates, with maternal progesterone accounting for only 7% of the steroid in the fetus. 7 Second, if synthetic progestins do cross the placenta, the timing must be considered. Both the diethylstilbestrol and 17-α hydroxyprogesterone caproate examples involved repeated maternal exposure during pregnancy, while the study by Hargreave et al 3 measured recent, shorter exposure, including prior to conception. The half-life of common pro-gestins used in contraception ranges from 12 hours for levonorgestrel (used in intrauterine devices and combined hormonal contraception) to 24 hours for medroxyprogester-one (MPA; used in injectable contraception). A highly effect ive a n d r e ve r s i b l e c o nt r a c e p t ive m e t h o d , d e p o-medroxyprogesterone acetate, is given intramuscularly at a dose of 150 mg every 3 months and acts by ovulation inhibition and endometrium atrophy. After an injection, the observed return to ovulation occurs in 30% to 56% of patients within 6 months and 75% to 100% patients within 12 months. 8 Detailed serial sampling of serum MPA and estra-diol and progesterone levels (to assess drug concentrations and ovulation) shows that MPA levels begin a gradual decline 2 to 3 months after an injection, but can be detectable for 7 to 9 months. However, ovulation occurred only when serum MPA levels were no longer detectable (<20 pg/mL) at 7 to 9 months. 9 In the present study, if maternal ovulation occurred for pregnancy to have happened, mothers were unlikely to have detectable levels of MPA. Therefore, it seems unlikely

Url: https://jamanetwork.com/journals/jama/fullarticle/2787750

User Submitted?: No

Authors: Spector, Logan G.; Moertel, Christopher L.; Su, H. Irene

Periodical (Full): JAMA

Issue: 1

Volume: 327

Pages: 39-40

Countries: