Novel management of Crohn's disease by dietary manipulation of the gut microbiome : the CD-TREAT diet

Background and aims: Diet is strongly associated with the aetiology of Crohn's Disease (CD). Treatment with exclusive enteral nutrition (EEN) in CD offers a nutritional therapy paradigm that could potentially be replicated by less restrictive dietary therapies, thereby improving compliance and tolerability. The current thesis initially aimed to explore perceptions and beliefs around the use of EEN and alternative novel, solid food-based diets expressed by adult and paediatric patients with CD, previously treated with EEN and their carers. Following that, a personalised food-based diet (CD-TREAT) was designed to replicate EEN composition. Its effect on gut microbiome, gut inflammation and clinical efficacy was assessed using a combination of randomised controlled trial (RCT) in healthy volunteers, animal models, and a pilot trial in children with active CD. Methods: An anonymous questionnaire surveyed adult patients and families of paediatric patients with CD, previously treated with EEN over 1 year. This questionnaire explored participants' demographic characteristics; acceptability of a repeat EEN course to treat a future flare (EEN repeat); their opinion on how difficult EEN would be compared to an example solid food-based diet; and their intention to participate in a future clinical trial assessing the therapeutic efficacy of a solid food-based diet in CD. Twenty-five healthy adults were randomly assigned to groups that received EEN or CD-TREAT for 7 days, followed by a 14-day washout, and then the diet they had not already received. The CD-TREAT diet was individualised for each participant by research dieticians; it excludes specific dietary components (such as gluten, lactose, alcohol) and matches the EEN in others (macronutrients, vitamins, minerals and fibre) using ordinary food. information on gastrointestinal symptoms and acceptability of the diet were collected. Additionally, faecal samples were collected before and after each experimental diet and changes to gut microbiome and metabolites were assessed using 16S rRNA sequencing and gas and liquid chromatography. Healthy HLA-B7 and HLA-B27 transgenic rats with gut inflammation received EEN, CD-TREAT, or standard chow for 4 weeks. CD-TREAT is a personalised diet which was designed to replicate EEN composition. Gut microbiome composition and metabolic activity was analysed in faeces, luminal content and gut tissue. Gut inflammation was assessed by histopathology and tissue cytokine expression, assessed by levels of mRNAs encoding 5 cytokines. Children (n=5) and adults (n=3) with relapsing luminal CD (weighted paediatric CD activity index [wPCDAI]≥12.5 for children and Harvey Bradsaw index [HBI]≥5) were recruited and treated with an 8-week course of CD-TREAT. Patients were followed up at weeks two, four, and eight; wPCDAI, quality of life (QOL) scores and markers of inflammation in blood, and faecal levels of calprotectin were measured. The primary outcome was clinical response (wPCDAI fall≥17.5 for children; HBI fall≥3 for adults) or clinical remission (wPCDAI < 12.5 for children; HBI < 5 for adults) at eight weeks. Results: Forty-one families of paediatric CD patients and 10 adult CD patients were approached for the aims of the questionnaire survey with 29 families and 4 adults sending replies (71%, 40% respectively). Most of the participants were positive on completing another EEN course, however the majority would choose a solid food-based diet alternative. Both paediatric and adult patients rated EEN to be more difficult to adhere to compared to an example solid food-based diet, with the ratings of children and their parents being strongly correlated (EEN: r = 0.83, SFD: r = 0.75, p < 0.001). The majority of the respondents would agree to participate in a clinical trial assessing the efficacy of a solid food-based diet for the management of active CD. Twenty-five healthy volunteers completed the RCT. CD-TREAT was easier to follow, more filling and palatable than EEN. CD-TREAT induced similar effects to EEN on faecal short chain fatty acids, including acetate, propionate and butyrate, sulphide, pH, and bacterial load. CD-TREAT changed the microbiome and metabolome (p < 0.001) in the same direction as EEN. Changes in taxon abundance correlated between EEN and CD-TREAT (R=0.65, p < 0.001). Animal experiments replicated the human RCT effects on microbiome, and both EEN and CD-TREAT ameliorated ileitis (mean difference in histopathology scoring, EEN: -1.25, p=0.015; CD-TREAT: -1.0, p=0.044 compared with untreated controls). Expression of inteleukin-6 (IL-6) in ileum was lower (p=0.036) in B27 rats on CD-TREAT than their counterparts on standard chow. From the enrolled CD patients, high compliance rates with the CD-TREAT dietary plan were observed in the paediatric, but not in the adult group. In the group of children, 80% (4/5) clinically responded to CD-TREAT and 60% (3/5) entered clinical remission. Faecal calprotectin decreased by a mean (SEM) of 918 (278) mg/kg or 55% of baseline values. In the adult group 67% (2/3) clinically responded to CD-TREAT and 33% (1/3) entered clinical remission. Faecal calprotectin did not change from baseline. Conclusion: Four distinct multifaceted studies demonstrated that CD-TREAT is more acceptable than EEN, replicates the EEN effect on gut microbiome, ameliorates rat ileitis and improves disease activity and inflammatory markers as induction treatment in paediatric active CD. The high tolerability of CD-TREAT suggests that this personalised, ordinary food-based diet could potentially replace EEN as the nutritional therapy of choice to treat active CD.