Predictors of Infliximab Failure in Ulcerative Colitis

Introduction: Infliximab is frequently used in patients with ulcerative colitis who have failed to respond adequately to conventional therapies. We evaluated predictors of primary non-response and secondary loss of response in an ulcerative colitis population. Methods: Ulcerative colitis patients from a single-center database of inflammatory bowel disease were identified. Patient demographics, disease extent and severity, concomitant therapy at initiation of infliximab, and treatment outcomes were collected using the electronic medical record. Logistic regression was used to determine predictors of primary non-response to infliximab. Cox regression was used to analyze predictors of secondary loss of response. Results: Ninety-two patients with ulcerative colitis were included in the analysis. Forty-eight percent were female. Mean disease duration was 6.5 years. Fifty-nine percent had extensive colitis and the remainder had left-sided colitis or proctitis. At initiation of infliximab, 60% were on corticosteroids, 50% were on immunomodulators, and 43% were on aminosalicylates. Of the patients 22/92 (24%) had a primary non-response to infliximab. Eighteen initial responders (25%) developed a secondary loss of response to infliximab. Mean duration of infliximab use in responders was 29.5 months. Concomitant immunomodulator use increased 8-fold the odds of a primary response to infliximab induction (p=0.0008). Subjects with extensive colonic distribution had 0.18 times the odds of having a response (p=0.0083). Secondary loss of response was unaffected by initial disease severity, disease distribution, disease duration or concomitant immunomodulator therapy. Lower body mass index (BMI) at initiation was associated with increased risk of secondary loss of response (p= 0.013). A non-significant trend was seen towards an increase in secondary loss of response in patients with steroid use at initiation (p=0.064). Colectomy rates were 24% overall. Forty-five percent of primary non-responders underwent colectomy, compared with 17% of initial responders (p=0.007). Conclusion: In this retrospective analysis, extensive disease and lack of concomitant immunomodulator therapy were associated with an increased risk of primary non-response to infliximab in patients with ulcerative colitis. Colectomy rates were greater than 2-fold higher in primary non-responders. Secondary loss of response was associated with lower BMI at initiation but was not associated with initial disease severity, disease distribution, or concomitant immunomodulator therapy.