Co-exposure of microscopic colitis-associated drugs does not affect paracellular permeability in vitro

Abstract

Background: Increasing evidence has associated microscopic colitis (MC) with exposure to frequently prescribed drugs (eg, NSAIDs and PPIs). The associations may be affected by duration and recency of use or by concomitant exposure. One of the proposed underlying mechanisms is a direct effect of the associated drugs on the colonic barrier. However, supportive data are lacking. The aims of this study were to assess the association between MC and single or concomitant use of NSAIDs and PPIs and to study their effects on colonic barrier function in a cell culture model. Methods: First, a case-control study was conducted. Cases of MC were identified in the national pathology registry and linked to the Outpatient Pharmacy Database of the Dutch PHARMO Institute. Each case was matched to 5 non-MC controls by age, gender, postal code, and follow-up period. The case's index date determined that of the control. Dispensings within the first 60 days before index date were excluded. Exposure was classified as current (61-90 days), recent (91-150 days), or past use (> 150 days) according to the time since last prescription. In current users, duration of continuous use and average daily dose were assessed. Conditional logistic regression was applied to quantify the strength of the associations. Second, CaCo-2 monolayers were basolaterally exposed to 0, 10, 25, or 100 μM of a frequently used PPI (omeprazole), NSAID (diclofenac), or both, for 2 hours. The transepithelial electrical resistance (TEER) and paracellular permeability (using FITC-D4) were assessed. Results: In total, 1 118 cases and 5 590 controls were identified. Current (OR 2.1, 95% CI 1.5-2.9) and recent (OR 2.2, 95% CI 1.6-2.9) use of NSAIDs and current (OR 2.3, 95% CI 1.8-3.0) and recent (OR 3.1, 95% CI 2.4-4.0) use of PPIs was associated with an increased risk of MC. Long-term use (> 24 months) of these drugs attenuated this risk. The highest risks were observed with current (OR 2.6, 95% CI 1.7-4.1) and recent (OR 3.4, 95% CI 2.2-5.2) concomitant use of NSAIDs and PPIs. In the in vitro study, a dosedependent drop in TEER was observed for NSAID, but not for PPI treated cells. Co-exposure to NSAID and PPI did not result in an additive effect. None of the conditions resulted in significant changes in FITC-D4 permeation. Conclusions: Observational data confirm an association between MC and current and recent PPI or NSAID exposure, especially in case of concomitant use. Although these drugs influenced TEER in a cell culture model, we did not observe an effect on paracellular permeability. Further research on the pathophysiology of drug-induced MC should include other factors, such as gut microbiota composition and host-specific susceptibility.