Is low-dose Aspirin use associated with a reduced risk of colorectal cancer ? a QResearch primary care database analysis Prof Richard Logan, Dr Yana Vinogradova, Dr Carol Coupland, Prof Julia Hippisley-Cox, Divisions of Primary Care, and Epidemiology & Public Health, University of Nottingham, UK BSG Annual Meeting, Birmingham 15th March 2011 Cohort studies of Colorectal Cancer in NSAID users Low-Dose Aspirin in the Primary Prevention of Cancer (Cook et al. JAMA 2005) Aspirin 50mg/day Placebo Rel Risk Colon 103 / 111 0.92 Rectum 30 / 25 1.20 Colorectal 133 / 136 0.97 (0.77-1.24) Average of 10 years of treatment Incidence of colorectal cancer in aspirin trials (from Cuzick et al Lancet Oncol 2009) Number of scripts in 13-96 months prior to the index date Statins COX-2 Inhibitors Traditional NSAIDs Aspirin 1.75 1.5 1.25 1 1 .75 .5 .25 0 1 2- 13-25-37-49+ 1 2- 13-25-37-49+ 1 2- 13-25-37-49+ 1 2- 13-25-37-49+ Number of scripts odds ratio compared to no prescriptions for a drug group lower/upper limits of 95 percent confidence interval © QRESEARCH 2005 Vinogradova et al. Gastroenterology 2007 Number of scripts in 13-48 months prior to the index date Statins COX-2 Inhibitors Traditional NSAIDs Aspirin 1.75 1.5 1.25 1 1 .75 .5 .25 0 1 2-12 13-24 25+ 1 2-12 13-24 25+ 1 2-12 13-24 25+ 1 2-12 13-24 25+ Number of scripts odds ratio compared to no prescriptions for a drug group lower/upper limits of 95 percent confidence interval © QRESEARCH 2005 Vinogradova et al. Gastroenterology 2007 Lancet Oct 22 2010 (n=391) 5yr 10yr 15yr 20yr 5yr 10yr 15yr 20yr Lancet Dec 7 2010 Risk of death from CRC 5yr 10yr 15yr 20yr Study population: QRESEARCH database • Currently largest primary care database in the UK • 574 general practices across the UK • > 9 million patients including those who have died or left, as well as patients still registered • > 30 million person-years of observation Data source: QRESEARCH database • • • • • Derived from GP clinical records Patient level consolidated database Anonymised data Longitudinal data for 15+ years Validated against external and internal measures Study design & setting • Nested case control study • Study period Jan 1998-July 2008 • Cases were incident colorectal • cancer patients 5 controls matched by • Age • Sex • Practice • Calendar year Exposure assessment : • Aspirin exposure analysis restricted to subjects with +10 and +15yrs of prescribing data • any use: • at least 1 script in 13-120 months and 13 to 180 months prior to the index date (date of diagnosis in the case) • Years of exposure: • up to 1 years • 1 to 2 years • 3 to 5 years • 6 to 9 years • 10 to 14 years Statistical analysis • Multiple imputations • • • • ICE procedure in STATA 5 imputed datasets Rubin’s rule’s to combine estimates Conditional logistic regression • • Odds ratios + 95% CI ( unadjusted & adjusted) 1% significance level Study Sample: 14,948 incident cases of colorectal cancer 1998/2008 9534 cases with 10 years of medical records 35,013 controls with 10 years of medical records 6643 cases with 15 years of medical records 20,652 controls with 15 years of medical records Confounding factors : • Body mass index • in kg/m2 • Smoking status • Non-smoker • Ex-smoker • Smoker • • Socio-economic status (Townsend score for post code) • quintiles Morbidities • • • • • • • CVD Diabetes High BP Osteoarthritis Colitis Crohn’s disease Rheumatoid arthritis Aspirin – median dose prescribed 10 year cohort (n = 10,073) 15 year cohort (n = 6,506) 77% 78% 76 – 150 mg 16% 15% 151 – 300 mg 5% 5% 2% 2% Tablet dose ≥ 75 mg >300 mg Aspirin – frequency of use recommended Daily 65% Twice daily 15% Alternate days 5% Cumulative exposure to aspirin in 1 to 15 yr period prior to CRC diagnosis / index date Cases (%) Controls (%) (n=6643) (n=20652) Any use 1572 (23.7) 4934 (23.9) No use 5071 (76.3) 15718 (76.1) Unadjusted odds ratio Adjusted odds ratio (95% CI) (95% CI) 1.03 (0.96 to 1.11) 1.0 reference 0.98 (0.91 to 1.07) P-value 0.709 1.0 reference Up to 365days 464 (7.0) 1369 (6.6) 1.07 (0.96 to 1.20) 1.04 (0.92 to 1.17) 0.522 1 to 3 years 453 (6.8) 1304 (6.3) 1.10 (0.98 to 1.24) 1.05 (0.93 to 1.19) 0.418 4 to 6 years 373 (5.6) 1253 (6.1) 0.97 (0.86 to 1.10) 0.91 (0.79 to 1.04) 0.160 7 to 9 years 196 (3.0) 655 (3.2) 0.99 (0.84 to 1.17) 0.90 (0.75 to 1.09) 0.279 10 to 14 years 86 (1.3) 353 (1.7) 0.83 (0.65 to 1.06) 0.73 (0.56 to 0.94) 0.017 Risk of Colorectal cancer in patients using aspirin in 1-15 yr period priorcancer to CRC diagnosis Risk of colorectal in patients using aspirin/ index date in 2 to 15 years prior to the index date Adjusted odds ratios and 95% confidence intervals Exposure (N of cases) OR (95%CI) Up to 1 year (417) 1.12 (0.99 to 1.26) 1 to 3 years (380) 0.97 (0.85 to 1.11) 4 to 6 years (378) 0.96 (0.84 to 1.10) 7 to 9 years (233) 0.87 (0.73 to 1.02) 10 to 14 years (164) 0.81 (0.66 to 0.98) .6 .8 1 1.2 Odds ratios and 95%CI are adjusted for deprivation, smoking, BMI, comorbidities, use of medication Reference group: No use of aspirin in 2 to 15 years prior the index date © QRESEARCH 2008 version 20 1.4 Cumulative exposure to aspirin in 1 to 10 yr period prior to CRC diagnosis / index date Any use Cases (%) Controls (%) (n=9534) (n=35013) 2147 (22.5) 7926 (22.6) 7387 (77.5) 27087 (77.4) Unadjusted odds ratio Adjusted odds ratio (95% CI) (95% CI) 1.03 (0.97 to 1.09) 0.98 (0.91 to 1.04) P-value 0.472 Years of use No use 1.0 reference 1.0 reference Up to 365 days 614 (6.4) 2281 (6.5) 1.01 (0.92 to 1.11) 0.98 (0.88 to 1.08) 0.635 2 to 4 years 842 (8.8) 2944 (8.4) 1.08 (1.00 to 1.18) 1.03 (0.94 to 1.13) 0.524 5 to 6 years 329 (3.5) 1194 (3.4) 1.04 (0.91 to 1.18) 0.96 (0.83 to 1.10) 0.522 7 to 10 years 362 (3.8) 1507 (4.3) 0.93 (0.82 to 1.05) 0.85 (0.74 to 0.97) 0.017 Exposure to aspirin in 1 to 10 yr period prior to CRC diagnosis / index date No use CRC cases Controls 7387 (77.5) 27087 (77.4) unadj OR 1.0 reference adj OR 1.0 reference P value Up to 1 year 547 (5.7) 1946 (5.6) 1.06 (0.96- 1.17) 1.02 (0.92- 1.13) 0.712 1 to 3 yrs 742 (7.8) 2621 (7.5) 1.07 (0.98- 1.17) 1.02 (0.93- 1.12) 0.703 4 to 6 yrs 359 (3.8) 1254 (3.6) 1.08 (0.95- 1.22) 1.00 (0.88 -1.14) 0.972 7 to 10 yrs 499 (5.2) 2105 (6.0) 0.91 (0.82- 1.01) 0.82 (0.73 -0.93) 0.001 Risk of Colorectal cancer in patients using aspirin Risk of colorectal in patients using aspirin/ index date in 1-10 yr period prior cancer to CRC diagnosis in 2 to 10 years prior to the index date Adjusted odds ratios and 95% confidence intervals Exposure (N of cases) OR (95%CI) Up to 1 year (547) 1.02 (0.92 to 1.13) 1 to 4 years (742) 1.02 (0.93 to 1.12) 5 to 6 years (359) 1.00 (0.88 to 1.14) 7 to 9 years (499) 0.82 (0.73 to 0.93) .6 .8 1 1.2 Odds ratios and 95%CI are adjusted for deprivation, smoking, BMI, comorbidities, use of medication Reference group: No use of aspirin in 2 to 10 years prior the index date © QRESEARCH 2008 version 20 1.4 Conclusions • Patients taking low dose aspirin have a reduced risk of Colorectal cancer • An 18% reduction in risk is evident after more than 7yrs of aspirin use • Effect not consistent with being COX-2 mediated Methodological strengths • Large sample size and representative population • Data electronically collected – unlikely misclassification bias • Data collected before the diagnosis – no recall bias • Excluded prescriptions 12 months prior to cancer diagnosis Baseline characteristics (15 years of data): Cases n=9,534 Controls n=35,013 Males (number, percent) 5,447 (57.1) 19,980 (57.1) Age in years (median, IQR) 71 (63 to 78) 72 (64 to 78) Months of records (median, IQR) 243 (173 to 417) 248 (176 to 1422) Body mass index (median, IQR) 26.2 (23.7 to 29.2) 26.1 (23.7 to 29.0) Smokers (number, percent) 1,325 (13.9) 4,941 (14.1) Comorbidity in CRC cases and controls: Proportion of patients with morbidities 34.1 33.8 35 cases controls 30 25 20 18.2 17.9 15.3 15.9 15 9.5 10 8.1 5 1.3 1.5 1.1 0.7 0.2 0.2 Col Crohns 0 CVD © QRESEARCH 2008 version 20 DM HBP RA OA