P357 Importance of stool drug monitoring in anti-TNF-treated patients

Background Anti-tumour necrosis factor (TNF) therapy has dramatically changed the treatment of inflammatory bowel disease (IBD). The relationship between clinical outcomes and serum anti-TNF levels is controversial. This study aimed to perform simultaneous analyses of the serum, mucosal and faecal infliximab and adalimumab levels to determine the mucosal expression of TNF-α and assess the relationship of the anti-TNF-α levels in these biological samples with the endoscopic and clinical activities and body composition in IBD patients. Methods Consecutive IBD patients who received maintenance anti-TNF-α therapy and underwent colonoscopy were enrolled. We assessed the clinical disease activity, collected the blood, faecal and biopsy samples. The number of TNF-α positive cells in the mucosa was detected using immunofluorescent labelling. Serum, mucosal and faecal anti-TNF-α levels, anti-drug antibody level in the serum and faecal calprotectin were determined using ELISA. Each patient underwent body composition analysis. Results Fifty consecutive patients have been analysed in the study. The number TNF-α positive cells was significantly higher in the inflamed than in the un-inflamed part of the colon (p = 0.01); however, such a difference was not detected in case of tissue drug levels. Tissue drug level did not correlate with the serum drug level. Nevertheless, tissue drug levels of the samples obtained from the un-inflamed part of the colon were significantly higher in patients who were in remission (p = 0.03467). The presence of the drug in the faeces was significantly different according to disease activity (p = 0.0022). Drug levels in the biological samples did not show any association with the presence of anti-drug antibodies. Body composition parameters had no significant impact on the serum and tissue drug levels. Faecal calprotectin showed significant correlation with faecal drug concentration (p = 0.0162) and clinical and endoscopic activity (p < 0.001). Conclusion To the best our knowledge, this is the first study to measure drug concentrations in all the relevant biological samples. Our findings highlight the importance of faecal drug monitoring because active disease may be present in spite of normal serum anti-TNF levels. Simultaneous determination of faecal drug concentration and faecal calprotectin may be more accurate for evaluating therapeutic responses in the future.