Hormone Replacement Therapy and the Risk of Melanoma in Post-menopausal Women

B.M. Hicks; K.B. Kristensen; S.A. Pedersen; L.R. Hölmich; A. Pottegård

Disclosures

Hum Reprod. 2019;34(12):2418-2429. 

In This Article

Results

Investigation of Melanoma Risk: Case Control Study Results

We identified 14 183 cases of melanoma between 1 January 2000 and 31 December 2015. Following exclusions, 8279 cases were matched to 165 580 cancer-free controls (Figure 1). Compared with controls, cases had a lower prevalence of alcohol-related disorders and COPD, were less likely to have a low comorbidity score and had longer durations of education. Other characteristics were similar between cases and controls (Table I).

Figure 1.

Flow chart of case selection. Immunosuppressant drug use includes use of azathioprine, cyclosporine or mycophenolate mofetil.

Results for HRT use associated with secondary melanoma diagnosis are presented in Table V. Compared with non-use of HRT, the use of HRT was not associated with an increased risk of secondary melanoma for post-diagnostic new-users (fully adjusted HR, 1.56; 95% CI 0.64–3.80) or continuous HRT users (fully adjusted HR, 1.26; 95% CI 0.89–1.78). Similarly, pre-diagnostic HRT use and past use of HRT were not associated with secondary melanoma risk. Analyses by HRT type revealed null associations for oestrogen and combination therapy. New-use of progestogen post-diagnosis and continuous progestogen use were associated with increases in risk of secondary melanoma; however, these were based on a small number of events (n ≤ 6).

Overall, 31.8% of cases and 28.4% of controls filled a prescription for HRT (Table II) yielding an adjusted OR of 1.18 (95% CI 1.12–1.24). A greater proportion of cases exhibited high use of HRT (≥1000 DDDs) than controls (15.5% vs. 13.3%) which corresponded to an adjusted OR of 1.21 (95% CI 1.13–1.29). ORs remained elevated across all cumulative DDD categories with no evidence of a dose-response (P for trend = 0.59). In recent high users, associations were more marked than in distant high users (OR, 1.28; 95% CI 1.17–1.41; OR, 1.14; 95% CI 1.04–1.25, respectively). Both oral and transdermal HRT were associated with increases in melanoma risk, which was more marked with transdermal HRT use (OR, 1.37; 95% CI 1.17–1.61).

Analyses by HRT type and melanoma risk are presented in Supplementary Table SIII. Overall, positive associations were observed with all HRT types. Similar to the primary analyses, for each HRT type, there was no evidence of dose-response relationships, and associations were more pronounced with recent high use. Results were similar amongst patients with exclusive use of oestrogen and oestrogen/progestogen combination therapy (Supplementary Table SIV). Likewise, associations remained largely similar across melanoma subtypes (Supplementary Table SV).

Sub-group analyses are presented in Table SIII. Overall sub-group analyses revealed similar results. However, null associations were observed for women <50 years, for nodular and acral lentiginous melanoma and for melanoma of unspecified location and of the upper limb. Additionally analysis by stage revealed associations only with localised melanoma (OR, 1.25, 95% CI 1.15 –1.36). Tests for effect measure modification showed that clinical stage modified the association (P < 0.001), whilst there was less evidence for effect modification by localisation (P = 0.09), age (P = 0.45) or melanoma type (P = 0.37). Additional analyses investigated the risk of melanoma associated with intravaginal oestrogen use, corresponding to an OR of 1.38 (95% CI 1.13–1.68) for high use (≥1000 DDDs) (Supplementary Table SVI). There was some evidence of a dose-response relationship, and associations were more marked for recent high users. In sensitivity analyses utilising a new user design (Supplementary Table SVII), estimates were attenuated, including for high HRT use (OR, 1.13; 95% CI 0.99–1.28). Similarly, null associations were observed for recent use. Finally, analyses of increasing lag periods (Supplementary Table SVIII) revealed results similar to the primary analyses Table III.

Investigation of Melanoma Prognosis: Cohort Study Results

From 8279 melanoma cases, 6575 patients with melanoma were included after excluding 1445 patients diagnosed after 2013 and 259 patients with <1-year follow-up. Patients in the cohort were followed for a median (interquartile range) of 5.1 (2.6–8.6) years. Table IV presents baseline characteristics by HRT use. Overall, new users of HRT with melanoma were younger, more likely to have a history of oral contraceptive use and had a lower comorbidity score and longer education. Other characteristics remained largely similar between groups.

Results for HRT use associated with secondary melanoma diagnosis are presented in Table V. Compared with non-use of HRT, the use of HRT was not associated with an increased risk of secondary melanoma for post-diagnostic new-users (fully adjusted HR, 1.56; 95% CI 0.64–3.80) or continuous HRT users (fully adjusted HR, 1.26; 95% CI 0.89–1.78). Similarly, pre-diagnostic HRT use and past use of HRT were not associated with secondary melanoma risk. Analyses by HRT type revealed null associations for oestrogen and combination therapy. New-use of progestogen post-diagnosis and continuous progestogen use were associated with increases in risk of secondary melanoma; however, these were based on a small number of events (n ≤ 6).

Whilst an association was observed for all-cause mortality and post-diagnostic new users (adjusted OR, 0.31; 95% CI 0.10 –0.96), this was based on a small number of events (n < 5) (Table VI). Associations by other HRT user groups and by HRT type revealed null associations. There was no evidence of an association with HRT use categories and secondary melanoma diagnosis or all-cause mortality in analyses by route of admission (Supplementary Tables SIX and SX) and by restricting the follow-up period to a maximum of 5 years (Supplementary Tables SXI and SXII).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....