Background
Studies highlighting changes in bacterial composition in the ileoanal pouch are limited by heterogeneity in analysis techniques and sampling strategies Therefore, caution must be used when interpreting microbiota data. Similar to findings in IBD, a decrease in bacterial diversity and ‘dysbiosis’ are associated with acute and chronic inflammation in the pouch. Changes in Clostridium spp. and E. coli are associated with inflamed pouches and treatment response. This study aimed to compare the bacterial microbiota composition in patients with chronic pouchitis who responded to antibiotics vs. those who did not.

Methods
Patients with confirmed chronic pouchitis defined by a pouch disease activity score ≥ 7 were treated with antibiotics. If patients were already on antibiotics, they were offered the opportunity to stop. Follow up was at 4 weeks to check clinical status. Patients who came off antibiotics who flared were given the opportunity to restart the antibiotics to prevent deterioration. Patients were analysed as either on antibiotics if they received antibiotics 2 weeks prior to the clinic or off antibiotics if they had stopped all antibiotics 2 weeks prior to follow-up. Stool was collected from patients on follow-up and DNA was extracted from this stool. Sequencing was performed on an Illumina platform. Statistical analysis was performed using STAMP 2.1.3 software with Welch’s two-sided t-test for comparing two groups with false discovery rate correction.

Results
There were 28 patients in the cohort; 23 patients with chronic pouchitis and 5 healthy controls who had never had pouchitis. Ten patients were female. The median age of the cohort was 47 years (range 26–74 years). A total of 12 samples on antibiotics and 11 off antibiotics. There were 10 responders and 13 non-responders. There were no differences between responders and non-responders and no differences in those taking antibiotics vs. those not taking antibiotics with chronic pouchitis. Healthy controls had significantly higher proportions of clostridium genera (p = 0.01) when compared with patients who fail to respond to antibiotics. Healthy controls had significantly higher proportions of lachnospiraceae species (p = 0.03) when compared with those that responded to antibiotic treatment for chronic pouchitis.

Conclusion
In a small cohort with chronic pouchitis we were unable to detect any bacterial differences between those that responded to antibiotics vs. those that did not. Despite this a healthy pouch had significantly higher proportions of Firmicutes species and genera which has been supported in other literature.