Effects of early medication treatment and metformin use for cancer prevention in diabetes patients: a nationwide sample cohort study in Korea using extended landmark-time analysis

Abstract

OBJECTIVES: This study investigated the effectiveness of early medication treatment and metformin use for cancer prevention in type 2 diabetes patients. METHODS: Population-based cohort data were used from the Korean National Health Insurance Service-National Sample Cohort database (KNHIS-NSC) for 2002–2013. Patient-specific medication prescription status was defined by the landmark time (LMT; a fixed time after cohort entry), considering both pre- and post-LMT prescriptions to control methodological biases in observational research. The LMT was set to 2 years. Logistic regression analysis with multivariable adjustment was conducted to analyze cancer incidence by patient-specific medication prescription status. RESULTS: Only 33.4% of the subjects were prescribed medication early (before the LMT) with compliance. Cancer incidence in individuals with early prescription and compliance was 25% lower (odds ratio [OR],0.75; 95% confidence interval [CI], 0.67 to 0.84) than in those without. As early-prescribed medications, metformin monotherapy and metformin combination therapy were associated with 34% (OR, 0.66; 95% CI, 0.51 to 0.83) and 25% (OR, 0.75; 95% CI, 0.64 to 0.88) lower cancer risk than nonuse, respectively. Patients who were prescribed late (post-LMT) but did not comply with the prescription had a 24% (OR, 1.24; 95% CI, 0.97 to 1.58) higher cancer incidence than non-users. Among patients who started monotherapy early without changes throughout the entire follow-up period, those who started on metformin had a 37% (OR, 0.63; 95% CI, 0.41 to 0.99) lower risk of cancer than non-metformin users. CONCLUSIONS: Doctors must prescribe antidiabetic medication early, and patient compliance is required, regardless of the prescription time, to prevent cancer. Metformin monotherapy or combination therapy is recommended as an early prescription.