The findings were originally reported in two Danish case-control studies in which physicians reported a fourfold increased risk of squamous cell carcinoma, and a moderate increased risk of basal cell carcinoma and cutaneous malignant melanoma in patients who used hydrochlorothiazide long-term.
And, while the new study did not find an increased risk of basal cell carcinoma and cutaneous malignant melanoma among long-term users of hydrochlorothiazide, they suggest that bendroflumethiazide “may be a safer alternative for patients at increased risk of skin cancer.” The long-term use of indapamide was associated with a moderately increased risk of cutaneous malignant melanoma but did not alter the risk of either squamous cell or basal cell carcinoma
“Our results suggest that bendroflumethiazide may be a safer alternative to hydrochlorothiazide and indapamide, especially for patients at increased risk of skin cancer, but future studies are needed to rule out a causal association between bendroflumethiazide and cutaneous malignant melanoma,” wrote authors who were led by Christoph R. Meier, PhD, a professor in pharmacy with University Hospital Basel (Switzerland) and a contributor to the Boston Collaborative Drug Surveillance Program.
This study adds to existing evidence that there is a dose-dependent increased risk of squamous cell carcinoma in users of high cumulative doses of hydrochlorothiazide, compared with non–hydrochlorothiazide users.
The study, an observational cohort study, was published earlier this year. It is based on data from the U.K.-based Clinical Practice Research Datalink. It included 271,154 new users of thiazides and thiazidelike diuretics, the majority at 87.6% having been prescribed bendroflumethiazide, 5.8% indapamide, and 3.6% hydrochlorothiazide. Outcomes were compared to those observed in 275,263 users of calcium channel blockers.
“The three primary outcomes of interest were a first-time diagnosis of cutaneous malignant melanoma, basal cell carcinoma, or squamous cell carcinoma,” the authors wrote.
Incidence rates and incidence rate ratios were estimated for both short-term and long-term users of thiazidelike diuretics and calcium channel blockers, while a propensity score (PS) analysis was done in order to control for 23 baseline covariates. The mean follow-up after PS weighting was 3.9 years for indapamide users and 5.5 years for hydrochlorothiazide users. Overall, the incidence rate ratios of squamous cell carcinoma were not markedly increased for either short-term or long-term users of thiazidelike diuretics, the authors reported.
In contrast, the incidence rate ratios of squamous cell carcinoma for hydrochlorothiazide users were increased by 29% for short-term users at an IRR of 1.29 while they were increased by almost twofold for long-term hydrochlorothiazide users at an IRR of 1.95.
Long-term use of hydrochlorothiazide was again associated with a 64% increased risk of basal cell carcinoma, compared with users of a renin-angiotensin inhibitor at a weighted IRR of 1.64.
In contrast, weighted incident rate ratios for basal cell carcinoma for both short-term and long-term thiazide users were not significantly different and results were similar for patients who took hydrochlorothiazide, indapamide, or bendroflumethiazide.
Weighted overall incident rate ratios for cutaneous malignant melanoma were not significantly different for either short-term or long-term users of thiazidelike diuretics, compared with calcium channel blocker users.
However, there was a 43% increased risk of cutaneous malignant melanoma among long-term indapamide users at a weighted IRR of 1.43, compared with calcium channel blocker users, the authors reported.
“Given the biological plausibility and the severe clinical implications of cutaneous malignant melanoma, this finding should be considered carefully,” they cautioned.
Limitations to the study include the fact that the database analyzed does not have information on sun exposure, skin characteristics, or socioeconomic status which may affect the amount of sun exposure participants received.
The authors had no conflicts of interest to declare.