OA: Cancer incidence attributable to the use of oral contraceptives and hormone therapy in Alberta in 2012

open access:
Cancer incidence attributable to the use of oral contraceptives and hormone therapy in Alberta in 2012

Background:

Hormonal contraceptives and hormone replacement therapies are classified as carcinogenic to humans (group 1) by the International Agency for Research on Cancer. We sought to estimate the proportion and total number of cancers attributable to the use of oral contraceptives and hormone therapy in Alberta in 2012.

Methods:

Population attributable risks were used to estimate the proportion of attributable cases for each associated cancer site. Relative risk estimates were obtained from the most relevant and recent epidemiologic literature. Prevalences of the use of oral contraceptives and hormone therapy in Alberta were collected from Alberta's Tomorrow Project. Specific cancer incidence data were obtained from the Alberta Cancer Registry for the year 2012.

Results:

Overall, 6.3% of breast cancers (n = 135) diagnosed in Alberta in 2012 were estimated to be attributable to the use of oral contraceptives, and the exposure potentially prevented about 57.3% of endometrial cancers (n = 276) and 29.1% of ovarian cancers (n = 52). About 15.5% of breast cancers (n = 258) and 8.9% of ovarian cancers (n = 13) were estimated to be attributable to the use of hormone therapy, whereas 11.3% of endometrial cancers (n = 48) were possibly prevented by the exposure.

Interpretation:

Based on our estimates, oral contraceptive use resulted in a net protective effect among the cancer sites studied, thus reducing the cancer burden in Alberta in 2012. The use of hormone therapy was estimated to increase the cancer burden in the province, therefore the risk and benefit of hormone therapy should be carefully considered before use.

Strengths and limitations

A strength of our study was the inclusion of CIs around the population attributable risk estimates, thereby providing the precision of the estimates which was not done in the UK study. However, for estimates associated with ever oral contraceptive use and current exposure to hormone replacement therapies (Appendix 1, Supplementary Table 2), the wide confidence intervals highlight the lack of precision for those estimates. For example, although we estimated that 136 cases of breast cancer are attributable to ever using oral contraceptives, this estimate could range from 5 to 260 cases.

The restricted prevalence data available on current oral contraceptive exposure and cessation of hormone preparation use in Alberta created some limitations in this study. Increased risk of breast cancer was mostly found among current and recent users of oral contraceptives,^1,2,25 and 10 years after cessation of use, the potential risk returned to baseline.^1,2 Thus, the observed attributable risk for breast cancer related to ever using oral contraceptives was likely overestimated for women who stopped taking oral contraceptives more than 10 years ago. For endometrial and ovarian cancer, the estimate should be accurate because the protective effect of oral contraceptives on these cancers lasts for 20 years or more.¹ In addition, because of limited availability of data, conducting sensitivity analyses to verify the impact of different latency periods on population attributable risk estimates was not possible. In a Canadian national study, where the population attributable risks of hormone therapy use on breast cancer were estimated over a 12-year period, the attributable risk decreased from 11.5% in 2000 to 5.2% in 2006 because hormone therapy use decreased after the Women's Health Initiative Study was published in 2002.²⁶ The same trend likely occurred in Alberta; however, because no Alberta-specific data concerning hormone use with this level of detail were available, we were unable to conduct these sensitivity analyses.

Conclusion

Although oral contraceptives and hormone therapy are both classified as carcinogens, the net effect of population attributable risks are different when multiple cancer sites are considered. Our analyses showed that the benefit of oral contraceptive use exceeds the potential risk among the cancer sites investigated because the number of cancers possibly reduced by oral contraceptive use was more than twice the number potentially associated with the exposure. Oral contraceptive use likely reduced the cancer burden in Alberta in 2012. In contrast, hormone therapy was estimated to increase the cancer burden in the province by about 200 excess cancer cases in 2012. The risks and benefits of hormone therapies should be carefully considered before their use.