1. NON-STEROIDAL ANTI-INFLAMMATORY DRUGS AND PANCREATIC CANCER RISK: A NESTED CASE-CONTROL STUDY Marie Bradley, Carmel Hughes, Marie Cantwell and Liam Murray Cancer Epidemiology and Prevention Research Group, & School of Pharmacy, Queen’s University Belfast

2. Overview Pancreatic cancer Chemoprevention of cancer with NSAIDs GPRD Nested case-control study examining the effect of NSAIDs on pancreatic cancer risk using GPRD

3. Pancreatic Cancer In the United Kingdom (UK) in 2006 there were around 7,600 cases of pancreatic cancer diagnosed. Rapidly fatal Presents at an advanced stage 5th leading cause of cancer mortality, in the western world Five year survival rate of 1%. As incidence and mortality increase with age, the pancreatic cancer burden will rise in the future, due to population ageing and increased life expectancy.

4. Known risk factors for pancreatic cancer 1. Smoking 4. Age 2. Genetic factors/ 5. Obesity Family history 3. Chronic pancreatitis Important to identify other aetiological factors for pancreatic cancer in order to improve prospects for primary prevention.

5. Chemoprevention of cancer using cyclooxygenase (COX) inhibitors Large body of epidemiological and experimental evidence suggesting that exposure to aspirin and NSAIDs is associated with a reduced risk of colorectal cancer through inhibition of COX-2. COX-2 expression is increased in inflammation and carcinogenesis. COX- 2 blockade has been shown in experimental studies to induce apoptosis, inhibit endothelial cell migration, neovascularisation and the invasive potential of pancreatic cancer cells. However evidence is limited.

6. Aims Primary Aim: “Investigation of the role of NSAIDs and aspirin in the aetiology/prevention of pancreatic cancer using the General Practice Research Database (GPRD )”

7. The general practice research database (GPRD) GPRD is the world’s largest computerised database of anonymised longitudinal patient records from primary care. Started up in 1987. 3.7 million (up to standard) UTS patients. 434 UTS practices across the UK. 5.5% of the UK population. Representative of UK population in terms of age, sex, and geographic distribution.

8. Methods Case-control study. Source population from (GPRD). Patients <85years old and had ≥5 years of follow-up, prior to the diagnosis date. Cases had a diagnosis of primary malignant neoplasia of the exocrine pancreas between 01/1995 and 06/2006. Up to 7 controls matched with each case on general practice site, sex and year of birth. Primary exposure of interest was exposure to NSAIDs/aspirin in the five years prior to the diagnosis date One year lag time.

9. Methods Total dose of NSAIDS per patient (the average daily dose in units of Defined Daily Dose [DDD]). Total duration of NSAID use per patient. The dose (DDDs) and duration (days) of NSAID use per patient were divided into quartiles. The effect of dose (DDDs) and duration (years) combined. Effect of different NSAID groups.

10. Analyses Conditional logistic regression analyses were used to generate odds ratios (OR) and 95% confidence intervals (95% CI) associated with NSAID use compared to non- use. All estimates of OR were adjusted for smoking status, Body Mass Index (BMI), alcohol use, and history of chronic pancreatitis, prior cancer, diabetes. Also adjusted for use of steroids, PPIs, H 2 RAs and hormone replacement therapy (HRT).

11. Results 1141 cases and 7954 controls were identified. Analysis examined: –Effect of any use –Effect of dose –Effect of duration of use –Effect of dose and duration –Effect of different groups of NSAIDs

12. 1. Dose of NSAIDs used Period of NSAID use Five years before index date, excluding the year before index date Entry into GPRD until one year before index date NSAID Dose (DDDs) Cases (n) Controls (n) Adjusted OR (95% CI)* NSAID Dose (DDDs) Cases (n) Controls (n) Adjusted OR (95% CI)* OR per 200 DDDs increase 60741780.99 (0.94-1.04) OR per 200 DDDs increase 79653890.99 (0.97-1.01) Category 1: No use: 53437761.0 Category 1: No use 34525651.0 Category2: 0-9.2 15610580.98 (0.80-1.20) Category 2: 0-31.5 19213541.01 (0.83-1.23) Category 3: 9.25-26.5 14510590.91 (0.74-1.11) Category 3: 31.55-90.5 21713571.11 (0.91-1.34) Category 4: 26.55-100.5 16010201.04 (0.85-1.26) Category 4: 90.55-298 19813331.03 (0.84-1.26) Category 5: >100.5 14610410.93 (0.76-1.14) Category 5: >298 18913450.96 (0.78-1.17) P for trend 0.620.87

13. 2.Duration of NSAID use Period of NSAID use Five years before index date, excluding the year before index date Entry into GPRD until one year before index date NSAID Duration (days) Cases (n) Controls (n) Adjusted OR (95% CI)* NSAID Duration (days) Cases (n) Controls (n) Adjusted OR (95% CI)* OR per 200 days increase 6074178 0.95 (0.91-1.00) OR per 200 days increase 7965389 0.97 (0.95-0.99) Category1: No use 53437761.0 Category 1:No use 34525651.0 Category 2: 0 - 33.5 1551078 0.97 (0.80-1.19) Category 2: 0 - 40.5 2041377 1.06 (0.87-1.28) Category 3: 33.55 - 156.5 1521013 1.01 (0.83-1.24) Category 3: 40.55 - 156.5 2041320 1.09 (0.90-1.33) Category 4: 156.55 – 773 1711044 1.09 (0.90-1.33) Category 4: 156.55 - 834 2141346 1.08 (0.88-1.31) Category 5: >773 1291043 0.79 (0.64-0.99) Category 4: >834 1741346 0.85 (0.69-1.06) P for trend0.240.41

14. 3. Dose and Duration of NSAID use combined since entry into GPRD Dose of NSAID Duration of NSAID use Low dose (<1.0 DDD/day)High dose (≥1.0 DDD/day) Cases (n) Controls (n) Adjusted OR (95%CI)* Cases (n) Controls (n) Adjusted OR (95%CI)* Nonuser and use for 0-1 years 50435601.070549761.0 0-2.99 years875181.17 (0.88 1.56)523321.02 (0.73-1.42) 3-4.99 years483550.95 (0.66-1.34)181800.66 (0.39-1.11) 5 years or more484210.70 (0.49-0.99)241770.83 (0.52-1.34) P for trend0.110.20

15. 4. NSAID groups Period of NSAID use Subgroups of NSAIDs Five years before index date, excluding the year before index date Entry into GPRD until one year before index date Cases (n) Controls (n) OR (95%CI) Adjusted OR (95%CI) Cases (n) Controls (n) OR (95%C) Adjusted OR (95%CI) Aspirin and derivatives 26718131.04 (0.89- 1.21) 0.95 (0.81- 1.12) 23617471.03 (0.88- 1.22) 0.95 (0.81- 1.13) Acetic Acid derivatives 24917470.99 (0.85- 1.16) 0.96 (0.82- 1.13) 40927621.06 (0.92- 1.21) 0.99 (0.86- 1.14) Coxibs 453530.87 (0.63- 1.22) 0.80 (0.56- 1.11) 473560.91 (0.66- 1.26) 0.82 (0.58- 1.16) Fenamates 14931.06 (0.59- 1.89) 1.02 (0.56- 1.86) 432671.14 (0.81- 1.60) 1.09 (0.77- 1.55) Other NSAIDs 19861.57 (0.95- 2.60) 1.43 (0.85- 2.43) 442191.44 (1.03- 2.01) 1.41 (1.0- 1.99) Oxicams 242340.71 (0.46- 1.08) 0.67 (0.43- 1.03) 624800.89 (0.68- 1.18) 0.83(0.62- 1.10) Propionic Acid Derivatives 30720231.08 (0.94- 1.25) 1.06 (0.92- 1.23) 57437151.17 (1.03- 1.32) 1.12 (0.98- 1.28) High dose aspirin302081.0 (0.67 –1.48 0.91 (0.61- 1.35) 533141.19 (0.88- 1.61) 1.10 (0.81- 1.50)

16. No significant reductions in risk of pancreatic cancer for ever use compared with never use of a NSAID. No overall association between pancreatic cancer risk and the total dose of NSAIDs prescribed and no risk reduction was seen among subjects who used the highest doses of NSAIDs. A 20% reduction in risk was seen among subjects who had been prescribed a NSAID for approximately 2 years or longer in the 5 years before diagnosis. A 30% reduction in pancreatic cancer risk was also seen in subjects who had used lower than average doses of NSAIDs for 5 years or more since entry into the GPRD. Results

17. Ever use versus never use of COX-2 inhibitors and oxicams in the five years before the index date, was associated with modest reductions in pancreatic cancer risk (20% and 30% respectively) but statistical significance was not achieved and too few subjects were exposed to these preparations to allow a more detailed analysis of use of these drugs.

18. Discussion Study strengths: –NSAID exposure was determined by prospectively collected prescription data, avoiding self-reported exposure and its associated disadvantages –Study size –Representative sample

19. Conclusion It therefore appears that long-term use (>2 years) of NSAIDs may protect against pancreatic cancer and that duration of use may be more important than dose. As there was no further reduction in risk with use of more than one DDD per day, which represents a standard anti-inflammatory dose, it does not appear that doses which exceed those that are routinely used for the main indications of these drugs are required for prevention.