Susan Spillane1, Kathleen Bennett1, Linda Sharp2, Thomas I Barron3. 1. Department of Pharmacology & Therapeutics, Trinity Centre for Health Sciences, St James's Hospital, Dublin 8, Ireland. 2. National Cancer Registry Ireland, Building 6800, Cork Airport Business Park, Kinsale Road, Cork, Ireland. 3. Department of Pharmacology & Therapeutics, Trinity Centre for Health Sciences, St James's Hospital, Dublin 8, Ireland. Electronic address: barront@tcd.ie.
Abstract
BACKGROUND: Preclinical evidence suggests a role for metformin in inhibiting tumour dissemination and metastasis. Previous studies have identified associations between metformin exposure and improved colorectal cancer survival. This study aimed to examine associations between metformin exposure and the odds of presenting with disseminated disease among colorectal cancer patients. METHODS: Colorectal cancer patients diagnosed 2001-2006 were identified from the National Cancer Registry Ireland. A linked national pharmacy claims database was used to determine exposure to anti-diabetic medications prior to diagnosis. Multivariate logistic regression was used to estimate odds ratios (OR) with 95% confidence intervals (CI) for associations between metformin use (versus non-metformin anti-diabetic drugs) and odds of presenting with disseminated disease (lymph node positive/metastatic). Analyses were stratified by anti-diabetic drug co-prescription and intensity of metformin exposure. RESULTS: The study population included 241 metformin-exposed diabetics, 129 non-metformin-exposed diabetics, and 4277 non-diabetic patients. In multivariate analysis, odds of disseminated disease were lower in metformin-exposed diabetics, compared with non-metformin-exposed diabetics, though not statistically significant (OR=0.66, 95% CI 0.39-1.12). In analyses stratified by metformin dosing intensity and anti-diabetic drug co-prescription, the odds were further from unity and approached significance in diabetics with high intensity, exclusive metformin use (OR=0.52, 95% CI 0.25-1.10). CONCLUSIONS: While overall there was no statistically significant association between metformin exposure and disseminated colorectal cancer at diagnosis, there was a suggestion that high intensity, exclusive metformin use may be associated with reduced odds of disseminated disease. The number of patients in these subgroup analyses was small, and further investigation in larger studies is warranted.
BACKGROUND: Preclinical evidence suggests a role for metformin in inhibiting tumour dissemination and metastasis. Previous studies have identified associations between metformin exposure and improved colorectal cancer survival. This study aimed to examine associations between metformin exposure and the odds of presenting with disseminated disease among colorectal cancerpatients. METHODS:Colorectal cancerpatients diagnosed 2001-2006 were identified from the National Cancer Registry Ireland. A linked national pharmacy claims database was used to determine exposure to anti-diabetic medications prior to diagnosis. Multivariate logistic regression was used to estimate odds ratios (OR) with 95% confidence intervals (CI) for associations between metformin use (versus non-metformin anti-diabetic drugs) and odds of presenting with disseminated disease (lymph node positive/metastatic). Analyses were stratified by anti-diabetic drug co-prescription and intensity of metformin exposure. RESULTS: The study population included 241 metformin-exposed diabetics, 129 non-metformin-exposed diabetics, and 4277 non-diabeticpatients. In multivariate analysis, odds of disseminated disease were lower in metformin-exposed diabetics, compared with non-metformin-exposed diabetics, though not statistically significant (OR=0.66, 95% CI 0.39-1.12). In analyses stratified by metformin dosing intensity and anti-diabetic drug co-prescription, the odds were further from unity and approached significance in diabetics with high intensity, exclusive metformin use (OR=0.52, 95% CI 0.25-1.10). CONCLUSIONS: While overall there was no statistically significant association between metformin exposure and disseminated colorectal cancer at diagnosis, there was a suggestion that high intensity, exclusive metformin use may be associated with reduced odds of disseminated disease. The number of patients in these subgroup analyses was small, and further investigation in larger studies is warranted.