M. Emerson1, S. Markowiak1, M. Wandtke1, F. C. Brunicardi1, J. Sferra2, S. Pannell1 1University of Toledo College of Medicine and Life Sciences,Department Of Surgery,Toledo, OHIO, USA 2Promedica Health Systems,Department Of Surgery,Toledo, OHIO, USA
Introduction: Fistula formation is a dreaded complication for patients with inflammatory bowel disease (IBD), particularly Crohn’s. Various immune-modulators are associated with fistula resolution. Statin medications are becoming more appreciated for their anti-inflammatory effects. We sought to use data mining techniques on a regional dataset of 2.6 million patients to study any association between statin use and fistula formation.
Methods: A dataset of de-identified patients was queried using data-mining techniques to identify patients with Crohn’s, Ulcerative Colitis, and non-infectious inflammatory bowel disease and compare these cases according to statin use, anti-TNF-alpha use, age, gender, ethnicity, medical comorbidities, and surgical history to test for a protective effect of statins on fistula formation. X2test and independent samples t-tests were used for the exploratory analysis with p<0.05 set for statistical significance. Afterwards, we sought to temporally associate statin use with fistula formation according to time from surgery, performed a comparison analysis for fistula formation in patients with history of diverticulitis on statins, and performed a literature review to identify potential mechanisms. Conflicting diagnosis codes were searched for as a measure of data quality.
Results: We identified 202 cases of IBD associated fistula from a total population of 5,481 patients with IBD. 605 patients were identified with fistula disease in Figure 1A shows the X2test indicating that statin use is associated with a 51% risk reduction in fistula formation for the Crohn’s patient subset (p=0.029). Figure 1B compares fistula occurrence across multiple forms of IBD and diverticulitis. Statin use in diverticulitis patients showed no protective effect on fistula formation (p=0.999). Figure 1C summarizes possible overlapping mechanisms for this phenomenon. Immune modulators were associated with lower occurrence of fistula (p<0.001) though statins remained statistically significant when adjusting for concomitant anti-TNF-alpha agents. Rates of Caucasian ethnicity and gender between the groups were similar (p>0.05) though statin patients in general had higher ages (Average age statin population 65+/-9 years vs 47+/-13 years for non-statin group p=0.01) and more medical comorbidities. Time from surgery was under-powered to draw conclusions on perioperative effect (p>0.05).
Conclusion: Statin medications may represent a low-risk, low-cost method for reducing the occurrence of fistulas for patients with IBD or history of diverticulitis. However, rates of surgical exploration were low in all patients with IBD over the 11-year study period, preventing study of the perioperative effect of statin medications on fistula formation. Caucasian patients were over-represented in this regional dataset compared to national norms, though this is typical in the United States. Improved data-mining techniques, additional observation time, or prospective trials would be required to demonstrate causation.
