Association between statin use and risk for keratinocyte carcinoma in the veterans affairs topical tretinoin chemoprevention trial.

Author(s): Dore DD, Lapane KL, Trivedi AN, Mor V, Weinstock MA

Affiliation(s): i3 Drug Safety, Waltham, Massachusetts 02451, USA. david.dore@i3drugsafety.com

Publication date & source: 2009-01-06, Ann Intern Med., 150(1):9-18.

Publication type: Multicenter Study; Randomized Controlled Trial; Research Support, N.I.H., Extramural; Research Support, Non-U.S. Gov't; Research Support, U.S. Gov't, Non-P.H.S.; Research Support, U.S. Gov't, P.H.S.

BACKGROUND: Recent evidence suggests that statins may prevent cancer. OBJECTIVE: To quantify the association between statin use and the occurrence of keratinocyte carcinoma in high-risk veterans. DESIGN: Cohort study. SETTING: 6 Veterans Affairs medical centers. PARTICIPANTS: 1037 participants of the Veterans Affairs Topical Tretinoin Chemoprevention Trial, a randomized, multicenter, double-blind, vehicle-controlled trial of topical tretinoin, 0.1%, for prevention of keratinocyte carcinoma conducted from November 1998 to November 2004. MEASUREMENTS: Time to first occurrence of keratinocyte carcinoma on the face or ears. Participants using a statin at randomization, according to the Veterans Affairs Pharmacy Benefits Management database, were considered exposed. Study dermatologists conducted physical examinations at baseline and every 6 months during follow-up. The association between statin use at randomization and the outcome was evaluated by using propensity score matching (n = 608) and Cox proportional hazards regression (n = 1037). RESULTS: Among the 1037 participants, 37% used a statin at randomization (n = 397) for a median duration of at least 900 days over a median follow-up of 3.5 years. In the propensity score-matched analysis, statin use at randomization was not associated with keratinocyte carcinoma (rate ratio, 0.92 [95% CI, 0.73 to 1.16]), a finding that was consistent with the estimates derived from the Cox proportional hazards regression (rate ratio, 0.84 [CI, 0.70 to 1.02]). Limitations: The extent of residual confounding is unknown, and the confidence bounds around the measures of association were wide. These data may not be generalizable to lower-risk populations. CONCLUSION: These data show no conclusive or consistent relationship between long-term statin use and risk for keratinocyte carcinoma.