BACKGROUND:
Hydrochlorothiazide, a common antihypertensive, has photosensitive properties, potentially increasing skin cancer risk. We evaluated melanoma and nonmelanoma skin cancer among hydrochlorothiazide users with 3 different cohorts as each allows assessment of different potential cofounders/effect modifiers, including race/ethnicity.
METHODS:
We built 3 cohorts using IBM MarketScan Research Databases: commercial and encounters (>3.5 million individuals, 2010–2018), a subcohort with health risk assessment respondents (415, 330), and Medicaid (509, 767, 2011–2017). Adults (aged 18+ years) entered the respective cohort with a first-filled prescription (cohort entry) for hydrochlorothiazide (the exposure of interest) or ACE (angiotensin-converting enzyme) inhibitors, an active comparator with ≥12 months of continuous enrollment with medical/pharmacy coverage at baseline. We excluded those who used hydrochlorothiazide/ACE inhibitor (including fixed-dose combination products) 12 months before cohort entry and those with prior skin cancer, HIV, or organ transplant. We compared the risk for hydrochlorothiazide versus ACE inhibitor using multivariate proportional hazards regression.
RESULTS:
Baseline characteristics were similar, aside from more Black individuals among hydrochlorothiazide users (43.3% [95% CI, 43.0%–43.6%]) than ACE inhibitor users (28.1% [95% CI, 27.9%–28.3%]). The hazard ratio (95% CI) for nonmelanoma skin cancer related to hydrochlorothiazide (versus ACE inhibitor) was 0.96 (0.91–1.00) in the commercial cohort, 1.01 (0.77–1.32) for the health risk assessment subcohort, and 1.33 (0.77–2.29) for Medicaid. For melanoma, the respective hazard ratios were 1.07 (0.95–1.20), 0.85 (0.43–1.67), and 0.93 (0.51–1.67).
CONCLUSIONS:
Our evaluation using 3 different approaches, including adjustment for race/ethnicity, did not establish a clear difference between hydrochlorothiazide and ACE inhibitor in terms of skin cancer risk.