P-027 YI Cumulative Probability and Outcomes of Anti-tumor Necrosis Factor Use in Patients with Ulcerative Colitis

Abstract

Background: The clinical course of ulcerative colitis (UC) may be waxing and waning -there can be periods of remission alternating with periods of acute flares requiring hospitalization, need for corticosteroid use, or need for surgery. Anti-tumor necrosis factor (anti-TNF) agents are aimed at mucosal healing and preventing long-term complications. We aim to describe the cumulative probability of anti- TNF use and outcomes of a population-based cohort of UC patients in Olmsted County. Methods: The medical records of a population-based cohort of Olmsted County, Minnesota residents diagnosed with UC from 1970 through 2010 were reviewed. Patients were followed until moving from Olmsted County, death, or June 30, 2015. We recorded data on anti-TNF use including start and stop dates, need for change to a second anti-TNF, and outcomes including adverse events and serious infections (defined as need for intravenous antibiotics, hospitalization, or death) related to anti-TNF use. Cumulative probabilities for the previous variables were calculated. Results: The cohort included 494 patients with UC, of whom 42.7% were women. Median age at diagnosis of UC was 35 years (range, 1.2-91.4). There were a total of 34 patients treated with infliximab, 3 patients treated with adalimumab, and 1 patient treated with golimumab as their first anti-TNF. The cumulative probability of any anti-TNF use after UC diagnosis was 3.2% at 5 years (95% CI, 1.6%-4.8%), 5.1% at 10 years (3.1%-7.2%), 9.0% at 20 years (6.0%-12.1%), and 10.7% at 30 years (6.8%-14.6%). Nineteen of 38 patients using anti-TNF agents stopped medication within 5 years after initiation. After the first anti-TNF medication was started, the cumulative probability of discontinuation was 21.3% at 6 months (95% CI, 7.0%-33.4%), 34.9% at 1 year (17.6%- 48.5%), 43.4% at 2 years (24.8%-57.7%), and 54.3% at 5 years (33.6%-71.7%). Seven patients started a second anti-TNF agent. Among patients discontinuing use of their first anti-TNF medication, the cumulative probability of starting a second anti-TNF was 5.3% after 30 days (95% CI, 0%-14.8%), 31.6% after 90 days (7.1%-49.6%), 31.6% after 6 months (7.1%-50.2%), and 31.6% after 1 year (7.1%-50.2%). Seven patients had either an adverse event or serious infection; the cumulative probability of any adverse event or infection after medication initiation was 0 after 30 days, 2.6% after 90 days (95% CI, 0%-7.7%), 8.4% after 6 months (0%-17.1%), 8.4% after 1 year (0%-17.9%), and 11.9% after 2 years (0.1%-23.3). Conclusions: In this population-based inception cohort, only a minority of UC patients were treated with anti-TNF agents, perhaps reflective of not only the study period, but also the overall disease severity in this unselected population without referral bias. Infliximab was the most common anti-TNF used for UC. About 50% of patients initially started on an anti-TNF had discontinued medication use within 5 years of treatment initiation. Less than 50% of these patients were started on a second anti-TNF. Adverse events/serious infections related to anti-TNF use were noted in about 18% of patients.