Children born to women who were using hormonal contraception within three months of pregnancy and during pregnancy have a higher risk of developing myeloid leukemia than those born to mothers who were not using hormonal contraception, according to a nationwide cohort study published in Lancet Oncology. Results of the study suggest that exposure to hormonal contraception in utero is a potential predictor for childhood leukemia – a disease with few established risk factors, given the known association between exposure to exogenous sex hormones and tumor development.
“The strength of the associations and the biologic plausibility of hormones causing cancer in children show that these results are strong enough to cause concern,” study author Marie Hargreave, PhD, from the Danish Cancer Society Research Center in Copenhagen, told ASH Clinical News. “Our results can have direct implications for clinical guidelines on how we use hormonal contraception up to pregnancy in the near future and will potentially protect future generations of children from potential harmful exposures before birth.”
Using data from the Danish Cancer Registry and the Danish Medical Birth Registry, investigators identified 1,185,157 children born between 1996 and 2014. The researchers then used the Danish National Prescription Registry to determine children’s exposure to hormonal contraceptives, using the following definitions:
- no use: mothers who never used contraception prior to birth (n=270,291)
- previous use (defined as prescriptions filled): mothers who used hormonal contraceptives >3 months prior to the start of pregnancy (n=778,844)
- recent use: mothers who used hormonal contraceptives ≤3 months prior to and during pregnancy (n=136,022)
During a median of 9.3 years (range = 4.6-14.2 years) from birth, 606 children were diagnosed with either lymphoid leukemia (n=465) or myeloid leukemia (n=141).
Compared with children of women who reported never using hormonal contraception, children born to women with recent use had a significantly higher risk for any leukemia (primary endpoint; hazard ratio [HR] = 1.46; 95% CI 1.09- 1.96; p=0.011). Among children who were exposed to hormonal contraception in utero, the HR rose non-significantly to 1.78 (95% CI 0.95-3.31; p=0.070).
There was no statistically significant association between previous use and lymphoid leukemia risk (HR=1.23; 95% CI 0.97-1.57; p=0.089) or between recent use and lymphoid leukemia risk (HR=1.27; 95% CI 0.90-1.80; p=0.167). (See TABLE on page 44 for additional analyses.)
Conversely, children were at a greater risk of developing myeloid leukemia if mothers reported recent use – particularly use during pregnancy:
- recent use: HR=2.17 (95% CI 1.22-3.87; p=0.008)
- use during pregnancy: HR=3.87 (95% CI 1.48- 10.15; p=0.006)
In analyses of exposure windows, the risk of developing any type of leukemia appeared to increase with more recent exposure to hormonal contraception, compared with women who never used hormonal contraception:
- use within 3 to 6 months prior to pregnancy: HR=1.38 (95% CI 1.03- 1.85; p=0.031)
- use within 6 to 12 months prior to pregnancy: HR=1.22 (95% CI 0.90- 1.65; p=0.159)
- use more than 1 year prior to pregnancy: HR=1.24 (95% CI 0.96- 1.60; p=0.108)
The unadjusted incidence of leukemia was 4.33 cases per 100,000 person-years for no maternal hormonal contraception use and 6.45 for recent use. These incidence rates translated to approximately one additional leukemia case per 47,170 exposed children, or approximately 25 leukemia cases during the study follow-up period.
“The results were almost unchanged when sex and perinatal factors were included in the analyses, when children with unknown or other leukemia were excluded, [and] when children with Down syndrome were excluded,” the authors reported. However, after adjustment for maternal smoking, the association with childhood leukemia was no longer statistically significant, suggesting that maternal smoking status may also play a role in the risk of childhood leukemia.
Previous and recent use of combined oral products (estrogen and progestin) increased the risk of any type of leukemia, compared with no use, but no specific type of hormonal contraception was associated with an increased risk for lymphoid leukemia.
Limitations of the study include the possibility that some women may not have used the prescribed contraceptives or used them at a different time, which may have led to a misclassification of exposure. The authors also acknowledged missing information, like exposure to radiation in utero and perinatal infections, that may have confounded the study’s findings.
“Further large, valid studies with registry-based information of hormonal contraception and childhood cancer – including a substantially larger study cohort and more cases – will help to either firmly establish or refute the possible link between prenatal exposure to sex hormones and the risk of childhood cancer,” Dr. Hargreave added.
She noted that her team is conducting a similar study that will include a total of four Nordic countries with registry-based data on maternal use of contraception and childhood cancer. “We are exploring the risk of other childhood cancer types with respect to maternal hormonal contraception and with respect to different types of fertility drugs, as we previously found an increased risk of leukemia and neuroblastomas in children born to women after fertility treatments involving the steroid hormone progesterone,” she said.
Corresponding authors reported financial relationships with Jazz Pharmaceuticals and Shire.
Hargreave M, Mørch LS, Andersen KK, et al. Maternal use of hormonal contraception and risk of childhood leukaemia: a nationwide, populationbased cohort study. Lancet Oncol. 2018 September 6. [Epub ahead of print]