Statin Initiation and Mortality After Colorectal Cancer Treatment.

Abstract

Observational studies have suggested improved survival among patients with colorectal cancer (CRC) who start statin therapy after diagnosis, but such studies are vulnerable to time-related biases and informative censoring. To estimate the association between statin initiation after CRC treatment and CRC-specific and all-cause mortality using target trial emulation. This nationwide cohort study used Korean National Health Insurance Service claims data from January 1, 2005, through December 31, 2015. Eligible adults had newly diagnosed CRC and no statin prescriptions in the prior 6 months. Time zero was the date of first CRC treatment. A clone-censor weight approach with stabilized inverse probability of censoring weights emulated initiation within a 6-month grace period. Data were analyzed at 3-month intervals during 60 months of follow-up from August 22, 2023, to November 22, 2025. Initiation of any statin within 6 months after first CRC treatment vs no initiation during the same grace period. The primary outcome was CRC-specific mortality; the secondary outcome was all-cause mortality. Weighted pooled logistic regression was used to estimate hazard ratios (HRs), 60-month marginal survival probabilities, and risk differences. Sensitivity analyses deliberately introduced immortal time bias and prevalent user bias. The same design was applied in a cohort with ischemic heart disease as a positive control. Among 88 516 patients (mean [SD] age, 62.7 [11.7] years; 56 424 [63.7%] male), 2071 initiated statins and 86 445 did not. Statin initiation was associated with modestly lower CRC-specific mortality compared with noninitiation (HR, 0.95 [95% CI, 0.92-0.99]); all-cause mortality was also lower but did not reach statistical significance (HR, 0.96 [95% CI, 0.93-1.00]). The 60-month absolute differences in CRC-specific and all-cause mortality were small (1.1 percentage points [pp] [95% CI, -0.1 to 2.3 pp] and 2.4 pp [95% CI, 1.0-3.8 pp], respectively). Conventional analyses yielded more protective estimates. In the positive-control cohort with ischemic heart disease, statin initiation was associated with lower all-cause but not cardiovascular disease-specific mortality. In this nationwide cohort study of adults treated for CRC, the survival benefit associated with statin initiation was modest and substantially attenuated compared with conventional observational estimates, suggesting that earlier protective associations may have reflected time-related biases and informative censoring.