P080 High faecal pH and low total microbial load associate with normalisation of faecal calprotectin in children with Crohn’s disease treated with exclusive enteral nutrition; results from iPENS, a multicentre, prospective study
K Gkikas,M Lima,S McKirdy,B Nichols,C Kerbiriou,L Gervais,G Smith,L Armstrong,T Jordan,I Chalmers,H Barlow,G Al Hourani,R Muhammed,D Wands,P Narula,M Patel,U Z Ijaz,S Milling,M Gasparetto,P Henderson,D C Wilson,R Hansen,R K Russell,K Gerasimidis
DOI: https://doi.org/10.1093/ecco-jcc/jjad212.0210
2024-01-01
Journal of Crohn's and Colitis
Abstract:Abstract Background Exclusive enteral nutrition (EEN) is a main therapy for active Crohn’s disease (CD) in children, but normalisation of faecal calprotectin (FCAL) varies among patients, even in those who enter clinical remission. To better understand disease characteristics related to EEN efficacy and its mechanism of action, we compared clinical and microbial parameters between patients whose FCAL normalised against those who did not at EEN completion. Methods Children with CD, clinically responding to EEN, were recruited from 11 UK hospitals (January 2020- May 2023, NCT04225689) and provided a single faecal sample before EEN completion. Patients were divided in two groups according to levels of FCAL at EEN completion (FCAL<250 and FCAL>250 mg/kg). Levels of faecal short chain fatty acids (SCFA), faecal sample characteristics (pH, water content (%), bristol stool score) and total microbial load (qPCR) were compared between the two groups. Anthropometric and clinical parameters (blood inflammatory markers, use of immunosuppressants, disease duration, disease location) were also compared. Machine learning using feature elimination with data imputation for missing data was performed to identify associations between clinical, anthropometry, microbial parameters and FCAL normalisation. Results At EEN completion, 84 children (female, 35%) were recruited [age, median (IQR): 13.2 (11.8, 14.9 years)] with a median (Q1, Q3) FCAL of 643 (146, 2033) mg/kg. Out of 84 patients, 35 (42%) had an FCAL<250 mg/kg. Total microbial load and SCFA were measured in a subset of patients (n=44). Patients with FCAL<250mg/kg had a higher faecal pH and lower microbial load compared to those with FCAL>250mg/kg [faecal pH; FCAL<250 mg/kg: 8.3 (8.1, 8.6) vs FCAL>250 mg/kg; 7.95 (7.6, 8.3), p=0.001; microbial load (log10 16S rRNA gene copies/g): FCAL<250mg/kg: 10.7 (10.4, 10.9) vs FCAL>250mg/kg: 11.0 (10.5, 11.2), p=0.02]. Median BMI z-score was also non-significantly (p=0.052) higher in patients with FCAL<250mg/kg. The use of immunosuppressants at EEN completion, disease duration, disease location and other faecal parameters were not different between the two groups. A multicomponent random forest model (clinical, blood inflammatory markers, anthropometry, faecal parameters) predicted normalisation of FCAL with 71% accuracy (sensitivity: 69%, specificity: 71%, p=<0.001, Figure 1). Higher faecal pH, BMI z-scores and lower total microbial load were the most influential parameters relating to FCAL<250mg/kg. Conclusion We showed that the efficacy of EEN in reducing gut inflammation might be, at least in part, mediated via reducing gut bacterial biomass and modulating luminal pH and the downstream effects this may have on inflammatory members of the microbial community.
gastroenterology & hepatology