Anti-TNF nonresponse in ulcerative colitis: correcting for mucosal drug exposure reveals distinct cytokine profiles.

Introduction: It remains unclear why up to 30% of ulcerative colitis (UC) patients do not respond to tumor necrosis factor inhibitors (TNFi). Validated biomarkers for nonresponse (N)R) are lacking. Most studies investigating underlying mechanisms do not differentiate between pharmacokinetic and inflammatory mechanisms. We therefore aimed to develop a framework to correct for mucosal drug exposure (MDE) and applied this to mucosal cytokine profiles previously linked to (N)R.

Methods: In a prospective international cohort, we studied patients with active moderate-severe UC starting TNFi treatment. Patients underwent endoscopy before (baseline) and after induction treatment (follow-up). NR was defined as the absence of Mayo endoscopic subscore improvement by central read or need for colectomy. The ratio of mucosal concentrations of TNFi/TNF was used to define high or low MDE. Mucosal concentrations of interleukin-6 (IL-6), Oncostatin M (OSM), interleukin-10 (IL-10), and interleukin-12/23p40 (IL-12/IL-23p40) were measured.

Results: Fifty-four UC patients were included (43 infliximab, 11 adalimumab) of whom 39 (72%) were endoscopic responders (after a median treatment of 62 days [48-96]). NR with high MDE had high IL-6 at both time points. R with low MDE exhibited low mucosal IL-10 at baseline. At follow-up, high OSM was associated with NR (irrespective of MDE) and high IL-12/IL-23p40 with R.

Conclusions: We incorporated MDE in mucosal cytokine research to avoid bias due to the insufficient presence of anti-TNF. When applied to mucosal cytokines previously linked to (N)R, IL-6 appears to drive inflammation in TNFi-resistant UC patients, while OSM seems to parallel inflammation and does not cause refractoriness.