Abstract 2161: Metformin and cancer risk: A cohort study in the UK Clinical Practice Research Datalink analyzed like a randomized trial

Abstract INTRODUCTION: The biguanide metformin may reduce cancer incidence according to recent meta-analyses of observational studies. Conversely, a meta-analysis of randomized controlled trials did not support the hypothesis that metformin lowers cancer risk. However, many of the published pharmaco-epidemiological studies suffer from severe biases and most randomized clinical trials are not designed or powered to look specifically at cancer safety outcomes. Therefore, we emulated the design and statistical analysis of a randomized trial in a large retrospective cohort study within the United Kingdom (UK) Clinical Practice Research Datalink (CPRD) to investigate associations comparing initiators of metformin to initiators of other oral first-line hypoglycemic agents among individuals newly diagnosed with diabetes in relation to risk of total cancer and by cancer site. METHODS: We retrieved data on all participants newly diagnosed with type 2 diabetes between January 1, 1987 and December 31, 2010. Eligible participants had no prior history of cancer and at least two years of follow-up in the CPRD after the diagnosis of diabetes. We also excluded the initial 12 months of follow-up after the first antidiabetic prescription. Treatment naive participants with diabetes at CPRD enrolment that first started using (initiators) metformin during follow-up were compared to initiators of other oral hypoglycemic agents. Exposure to certain antidiabetic drugs was classified only based on the initial twelve-month treatment period. We mimicked the statistical analysis of a randomized trial in an observational dataset using the intention-to-treat (ITT) principle. Cox proportional hazards models were used to estimate hazard ratios (HR) and 95% confidence intervals (95% CI) after adjustment for known or suspected risk factors for cancer. We also calculated adherence-adjusted estimates using inverse probability weighting. RESULTS: A total of 95,820 participants were included. Of these, 51,484 (54%) individuals started taking metformin monotherapy and 18,264 (19%) started sulfonylurea monotherapy. A total of 3,805 first incident cancers were identified in a maximum follow-up of 24 years. Compared to initiators of sulfonylurea monotherapy, initiators of metformin had a similar incidence of total cancer (HR, 0.96; 95% CI, 0.89-1.04). There was no evidence for an association between use of metformin and cancers of the colorectum (HR, 0.92; 95% CI, 0.76-1.13), prostate (HR, 1.02; 95% CI, 0.83-1.25), lung (HR, 0.85; 95% CI, 0.68-1.07), postmenopausal breast (HR, 1.03; 95% CI, 0.82-1.31), or any other cancer. The estimates remained similar after adjustment for non-adherence. CONCLUSIONS: We found no evidence for a decreased or increased risk of any cancer among those exposed to metformin versus sulfonylurea. These findings need to be confirmed in large RCTs. Citation Format: Konstantinos K. Tsilidis, Despoina Capothanassi, Naomi Allen, Evangelos Rizos, David Lopez, Karin van Veldhoven, Carlotta Sacerdote, Deborah Ashby, Paolo Vineis, Ioanna Tzoulaki, John Ioannidis. Metformin and cancer risk: A cohort study in the UK Clinical Practice Research Datalink analyzed like a randomized trial. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 2161. doi:10.1158/1538-7445.AM2014-2161