Menopausal Hormone Therapy and Risk of Melanoma: A Nationwide Register-Based Study in Finland Free

The Women's Health Initiative trials showed that menopausal hormone therapy is associated with an increased risk of breast cancer, cardiovascular diseases, and pulmonary embolism (1, 2). After the publication of those findings in the early 2000s, prescriptions of hormone therapy declined globally (3, 4). However, hormone therapy is still prescribed, mostly to prevent or alleviate severe climacteric symptoms (5), and its safety profile is still under investigation. The debate on the possible role of hormone therapy as a risk factor for cutaneous melanoma has recently intensified after the publication of three large cohort studies on the topic. A Norwegian nationwide study reported a significant association between hormone therapy with plain estrogens and the risk of cutaneous melanoma (6). A Swedish nationwide study reported an increased risk of cutaneous melanoma in both estrogen and estrogen plus progestin hormone therapy users (7). A French study reported an association between some types of hormone therapy and the risk of cutaneous melanoma, especially in women starting their treatment shortly after menopause (8).

More evidence on this association is needed to further investigate the safety of hormone therapy and the etiology of melanoma. Therefore, we conducted a nationwide cohort study in Finland, based on the linkage of population-based registries, to estimate the association between hormone therapy use and the risk of cutaneous melanoma. We examined different types, routes of administration, and doses of hormone therapy.

All Finnish women ages 50 years or older, who had used unopposed estrogen therapy or combined estrogen–progestin therapy (EPT) between 1994 and 2007 were identified from the Medical Reimbursement Register of the National Social Insurance Institution (n = 1.1 million users). The cohort was linked by the unique personal identity code to the nationwide population-based Finnish Cancer Registry, and to Population Registry, which contains information about emigration and death. Women were followed from the date on which they cumulated a predetermined amount of hormone therapy (see details in the Exposure and outcome section below), until emigration, death, or December 31, 2014, whichever occurred first. The study was performed with permission from the Finnish National Research and Development Center for Welfare and Health (THL/588/5.05.00/2013). Because of the regulations, the data are only available at individual level or tabulated into small strata if the new user has a permission of getting the data.

The estrogen component in the EPT preparations in Finland is always estradiol, which is given at daily doses of 1 or 2 mg orally, or transdermally 25–100 μg in patches, or 0.5–1.5 mg in gel. The mode of EPT regimen was defined as sequential when progestin was added for 10–14 days to estradiol at 1–3-month intervals and as continuous when progestin was taken daily with estradiol. Norethisterone acetate, present both in oral and transdermal EPT regimens, was the most common progestin component. Medroxyprogesterone acetate and dydrogesterone, were present only in oral regimens. The remaining progestins included levonorgestrel, megestrol acetate, norethisterone, progesterone, lynesterol, trimegestone, and drospirenone.

The exposure was categorized according to type of hormone therapy (estrogen therapy or EPT), and EPT was further subdivided into sequential and continuous regiments. We analyzed the data according to predetermined durations of hormone therapy use: ≥6 months or ≥60 months. To study the trend between duration of hormone therapy use and risk of cutaneous melanoma, we divided the exposure duration in two periods: 6–59 months and ≥60 months or more. To be confident about the 6–59 months category, we included only women starting their treatment in 1995 or later, for whom we accurately know their full hormone therapy use history.

Estrogen therapy and EPT were also separated into route of administration (oral and transdermal, the latter including both patches and gel) and daily doses of oral estrogen (<1.1, 1.1–1.9, and >1.9 mg) for ≥60 months of use. Oral EPT was defined as a regimen in which both estradiol and progestin were taken orally. Transdermal EPT was defined as transdermal estrogens and oral or transdermal progestin.

The outcome was incident cutaneous melanoma using the International Classification of Diseases for Oncology (ICD-O-3 topography C44, morphology 872-9, and behavior 3).

Standardized incidence ratios (SIR) were estimated by dividing the total (summed over all age groups) numbers of observed cases by the total numbers of expected cases, and exact 95% confidence intervals (95% CI) for the SIRs were estimated on the basis of the assumption that the observed number of cutaneous melanoma cases follow a Poisson distribution. The expected numbers of cutaneous melanoma cases were calculated by multiplying the number of women-years at risk in each 5-year age group and calendar period by the corresponding average cutaneous melanoma incidence in Finnish women. For the SIR of at least 6 months of use the women-years at risk were cumulated from the first purchase of the relevant product after 6 months of use, and for the SIR of at least 60 months of use from the first purchase after 60 months of use. The follow-up ended at emigration, death, or end of 2014.

P values for heterogeneity were estimated with likelihood ratio tests comparing a Poisson model with a factor for type of use with a Poisson model with only intercept, both including the logarithm of the total expected number of cases as offset.

We followed 293,570 women contributing 4,461,284 person-years (Table 1), who had used hormone therapy for at least 6 months from 1994 to 2007 (131,499 estrogen therapy and 162,071 EPT users). Mean age of hormone therapy users at baseline was 59.2 years (62.1 for estrogen therapy and 56.9 for EPT users). During the mean follow-up of 15.6 years, 1,695 women received a diagnosis of cutaneous melanoma.

The SIR for women who used any hormone therapy for at least 6 months was 1.16 (95% CI, 1.11–1.22). The SIRs for women who used estrogen therapy and EPT for at least 6 months were 1.25 (95% CI, 1.16–1.33) and 1.10 (95% CI, 1.03–1.17), respectively (Table 2; P_{heterogeneity} = 0.01). Among EPT users, sequential EPT users had an increased risk of cutaneous melanoma (SIR, 1.12; 95% CI, 1.04–1.21), while continuous EPT users had no increase (SIR, 1.01; 95% CI, 0.87–1.16).

The SIR for women who used any hormone therapy for at least 60 months was 1.28 (95% CI, 1.19–1.36). The SIRs for women who used estrogen therapy and EPT for at least 60 months (n = 141,146) were 1.37 (95% CI, 1.22–1.52) and 1.23 (95% CI, 1.13–1.34), respectively, (P_{heterogeneity} = 0.15). Among EPT users, those using sequential EPT had a significant increased risk of cutaneous melanoma (SIR, 1.26; 95% CI, 1.14–1.39), while in those using continuous EPT the excess was not statistically significant (SIR, 1.16; 95% CI, 0.98–1.35). We did not find significant differences between oral and transdermal administrations in estrogen therapy users (P_{heterogeneity} = 0.42) and EPT users (P_{heterogeneity} = 0.63). Finally, we found no significant differences in the association between hormone therapy and cutaneous melanoma risk according to different doses of oral estrogens (P_{heterogeneity} = 0.52 in estrogen therapy users and 0.65 in EPT users).

When limiting the analysis to the 152,913 women who started their use in 1995 or later, for whom we are likely to know their entire hormone therapy use history, the SIRs for women who used estrogen therapy and EPT for 6–59 months were, respectively, 1.20 (95% CI, 1.06–1.35) and 1.00 (95% CI, 0.87–1.14; P_{heterogeneity} = 0.04).

In this Finnish nationwide cohort study, we found a 16% increased risk of cutaneous melanoma associated with the use of hormone therapy for at least 6 months, and a 28% with the use of hormone therapy for at least 60 months. Our results are in line with those from three recent European cohort studies (6–8). In a Swedish nationwide cohort study published in 2017, which included 898 cases, authors reported a 19% risk increase for ever users of hormone therapy compared with never users (7). In a Norwegian study published in the same year, and including 1,476 cases, authors reported a 12% risk increase for ever users and 19% for current users of hormone therapy compared with never users (6). Authors of a French cohort study, which included 444 cases, reported a 35% risk increase for ever users and 26% for current users of hormone therapy compared with never users (8). A meta-analysis which included seven case–control studies and three small cohort studies, all published between 1983 and 2009, reported a nonsignificant association of hormone therapy with cutaneous melanoma risk (pooled relative risk, 1.16; 95% CI, 0.93–1.44; ref. 9). Finally, a 2011 article from the WHI randomized-controlled trial, including only 95 cutaneous melanoma cases, reported no association between hormone therapy and cutaneous melanoma risk (10).

We found evidence suggesting that women using estrogen therapy had a higher risk elevation of cutaneous melanoma than those using EPT. In agreement with our results, in the Swedish cohort authors found a 26% risk increase in unopposed estrogen ever users and a 13% increased risk in estrogen plus progestin ever users compared with never users (7). In the Norwegian cohort, the authors reported a 45% increased risk of cutaneous melanoma associated with the current use of estrogens formulations and no increase with the use of estrogens plus progestins, compared with never use (6). In the French cohort, authors reported a cutaneous melanoma risk increase of 49% in estrogen therapy ever users and 36% in EPT ever users compared with never users (8). The fact that all four cohorts found a higher risk for estrogen therapy users than EPT users might have a biological explanation. Several in vitro experiments suggested that estrogens might increase the proliferation of melanocytes and melanoma cells, while progesterone might act as an antiproliferative and proapoptotic agent, counteracting the stimulatory effects of estrogens (11–16). As hypothesized in a previous article (6), estrogens and progestins might in fact affect the risk of cutaneous melanoma in opposite ways. Moreover, in our study only sequential EPT was associated with a significant increased risk of cutaneous melanoma, while continuous EPT was not, in agreement with the results from the Norwegian cohort (6). One could hypothesize that the risk difference between the two schedules could be explained by the smaller doses and number of days per month of progestins in the sequential compared with continuous EPT formulations. Another hypothesis is that the cyclic stimulation is a risk factor per se.

We did not find significant differences between the routes of administration of hormone therapy. Similar results were also found in the French and the Norwegian cohorts. Concerning the doses of estrogens, we did not find an increasing cutaneous melanoma risk with increasing doses. This is in contrast with the results of the Norwegian cohort, where authors reported that increasing doses of estrogens were associated with an increasing cutaneous melanoma risk, while increasing doses of progestins were associated with a decreasing cutaneous melanoma risk.

This is a large nationwide cohort study, which adds important evidence on the potential harmful effect of hormone therapy on the risk of cutaneous melanoma, especially of estrogen therapy. The use of the prescription registry ensures detailed information on the exposure, including hormone therapy types, routes of administration, and dose of estrogens, avoiding recall bias. The completeness and accuracy of the cancer registry guaranteed precise information on the outcome (17). The length of the follow-up is an additional strength of the study. The biggest limitation of the study is the lack of information on important potential confounders and effect modifiers, such as sun exposure and use of solarium, which might interact with hormone therapy exposure (8, 18). However, it was shown that in Finland socioeconomic status, which is usually associated with occupational sun exposure, recreational sun exposure, and tanning (19), is homogeneous in hormone therapy users and nonusers (20). This suggests that ultraviolet radiation exposure might be similar in hormone therapy users and nonusers, and that the lack of information on ultraviolet radiation exposure might not cause an important bias in our study. The heterogeneity analyses between subgroups (e.g., estrogen therapy vs. EPT) might be confounded by differences in the age distribution between the subgroups. However, for the main finding (use of at least 6 months of estrogen therapy vs. EPT), we calculated the SIR_ET/SIR_EPT ratio in each of the 5-year age stratum and we found no important differences among the strata, indicating that there is no important confounding (21). Finally, our estimates could be affected by another potential confounding factor: whether hormone therapy users are more concerned about their health than nonusers, and have more skin examinations and removals of precancerous lesions, the estimates might be too low.

In conclusion, menopausal hormone therapy was associated with an increased risk of cutaneous melanoma, especially estrogen therapy. EPT use of less than 5 years was not associated with an increased risk of cutaneous melanoma. The major limitation of our study was the lack of information on ultraviolet radiation exposure. More evidence on the effect of different types of hormone therapy and doses of hormones on the risk of cutaneous melanoma is needed, as well as on the possible interaction between hormone therapy use, ultraviolet radiation exposure, and cutaneous melanoma risk.

No potential conflicts of interest were disclosed.

Where authors are identified as personnel of the International Agency for Research on Cancer/World Health Organization, the authors alone are responsible for the views expressed in this article, and they do not necessarily represent the decisions, policy, or views of the International Agency for Research on Cancer/World Health Organization.

Conception and design: E. Botteri, N.C. Støer, E. Weiderpass, E. Pukkala, H. Lyytinen

Development of methodology: E. Botteri, E. Weiderpass, E. Pukkala, O. Ylikorkala

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): E. Weiderpass, E. Pukkala, O. Ylikorkala

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): E. Botteri, N.C. Støer, E. Weiderpass, E. Pukkala, O. Ylikorkala

Writing, review, and/or revision of the manuscript: E. Botteri, N.C. Støer, E. Weiderpass, E. Pukkala, O. Ylikorkala, H. Lyytinen

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): E. Weiderpass

Study supervision: E. Botteri, E. Weiderpass, E. Pukkala

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.