The reduction of faecal calprotectin during exclusive enteral nutrition is lost rapidly after food re‐introduction

Summary Background Faecal calprotectin decreases during exclusive enteral nutrition in children with active Crohn's disease. It is unknown how faecal calprotectin changes during food re‐introduction and the influence of maintenance enteral nutrition. Aims To study changes to faecal calprotectin during exclusive enteral nutrition and at food reintroduction, and explore associations with maintenance enteral nutrition. Methods Children with Crohn's disease were followed during exclusive enteral nutrition and during food‐reintroduction. Faecal calprotectin was measured before, at 33 and 54 days of exclusive enteral nutrition, and at 17, 52 and 72 days after food‐reintroduction. Maintenance enteral nutrition use was recorded with estimated weight food diaries. Data are presented with medians and Q1:Q3. Results Sixty‐six patients started exclusive enteral nutrition and 41 (62%) achieved clinical remission (weighted paediatric Crohn's disease activity index <12.5). Baseline faecal calprotectin (mg/kg) decreased after 4 and 8 weeks of exclusive enteral nutrition (Start: 1433 [Q1: 946, Q3: 1820] vs 33 days: 844 [314, 1438] vs 54 days: 453 [165, 1100]; P < .001). Within 17 days of food reintroduction, faecal calprotectin increased to 953 [Q1: 519, Q3: 1611] and by 52 days to 1094 [660, 1625] (both P < .02). Fifteen of 41 (37%) children in remission used maintenance enteral nutrition (333 kcal or 18% of energy intake). At 17 days of food reintroduction, faecal calprotectin was lower in maintenance enteral nutrition users than non‐users (651 [Q1: 271, Q3: 1781] vs 1238 [749, 2102], P = .049) and correlated inversely with maintenance enteral nutrition volume (rho: −0.573, P = .041), kcals (rho: −0.584, P = .036) and % energy intake (rho: −0.649, P = .016). Maintenance enteral nutrition use was not associated with longer periods of remission (P = .7). Faecal calprotectin at the end of exclusive enteral nutrition did not predict length of remission. Conclusions The effect of exclusive enteral nutrition on faecal calprotectin is diminished early during food reintroduction. Maintenance enteral nutrition at ~18% of energy intake is associated with a lower faecal calprotectin at the early phase of food reintroduction but is ineffective in maintaining longer term remission.


Funding information
The work of the IBD team in Glasgow is supported by the Catherine McEwen foundation. UZI is funded by NERC Independent Research Fellowship NE/ L011956/1. KG received research grants, speakers fees and had conference attendance paid by Nutricia/Nestle and DrFalk. RH has received speakers/consultancy fees or conference support from Nutricia, Dr Falk, MSD Immunology and 4D Pharma. RKR has received speaker's fees, travel support and/ or participated in medical board meetings Summary Background: Faecal calprotectin decreases during exclusive enteral nutrition in children with active Crohn's disease. It is unknown how faecal calprotectin changes during food re-introduction and the influence of maintenance enteral nutrition.

Aims:
To study changes to faecal calprotectin during exclusive enteral nutrition and at food reintroduction, and explore associations with maintenance enteral nutrition.
Methods: Children with Crohn's disease were followed during exclusive enteral nutrition and during food-reintroduction. Faecal calprotectin was measured before, at 33 and 54 days of exclusive enteral nutrition, and at 17, 52 and 72 days after food-reintroduction. Maintenance enteral nutrition use was recorded with estimated weight food diaries. Data are presented with medians and Q1:Q3.

| INTRODUC TI ON
Crohn's disease (CD) is a chronic inflammatory condition which can affect any part of the gastrointestinal tract from mouth to anus.
Symptoms include abdominal pain, bloody diarrhoea, weight loss, along with additional extra-intestinal complications such as anaemia, skin rashes and arthritis. 1 As Crohn's disease remains incurable, current therapy aims to induce and maintain clinical remission and, when possible, intestinal mucosal and transmural healing. 2 Exclusive enteral nutrition (EEN) is a liquid-only formula diet which is highly effective in achieving clinical remission in paediatric Crohn's disease. 3 EEN is recommended as the first line treatment for active luminal Crohn's disease in children with up to 89% of patients achieving clinical remission when placed on EEN for 8 weeks. 4,5 Reviews and meta-analyses have shown that EEN is as effective as oral corticosteroids in inducing clinical remission, in paediatric patients. It also has the added benefits of mucosal healing in a significant proportion of patients plus improvement in the overall nutritional status of patients. [6][7][8][9] Achieving complete mucosal healing after a course of EEN has been associated with reduced relapse rates and complications in the medium term. 2 While the use of enteral nutrition as induction therapy of active Crohn's disease is well established, use of partial enteral nutrition, as a maintenance therapy (maintenance enteral nutrition or MEN), is less well studied. 10 Previous research suggested that MEN may be effective in prolonging periods of clinical remission. [11][12][13][14][15][16][17][18][19][20][21][22] However, these studies have generally been of retrospective design, lacked robust assessment of MEN compliance and have not examined changes in gut markers of colonic inflammation.
We have previously described the effect that EEN has on faecal calprotectin, in children with active Crohn's disease during induction treatment with EEN, which has subsequently been replicated by others. 23,24 EEN reduces the pre-treatment concentration of faecal calprotectin by an average of 48% nevertheless, less than 20%-25% of those who have completed a successful course will have a measurement of FC below 250 mg/kg. 24 In our previous research we also have described that faecal calprotectin levels increased rapidly in patients within 4 months of food reintroduction. 23 Hence, it is of interest to study how quickly faecal calprotectin rises during the early and medium phase of food reintroduction in patients who completed a successful course of EEN, and whether MEN use can modify this effect. Furthermore, the concurrent change in cytokines paralleling changes in clinical and biochemical parameters has only been described in a few studies, contrasting with the well-documented changes in patients undergoing biologic treatment. [25][26][27][28] The aim of the current study was to determine short and medium-term changes in faecal calprotectin during food reintroduction, following induction of remission with EEN, and to explore the effectiveness of MEN to influence these changes in addition to risk of subsequent clinical relapse.

| Patients
Patients with suspected IBD who attended the Royal Hospital for Children, Glasgow and neighbouring district general hospitals, were recruited prospectively between October 2014 and May 2017.
Diagnosis was based on established radiological, histological and endoscopic guidelines. 29 Disease behaviour and anatomical location were classified using the Paris classification. 30 Patients who were subsequently diagnosed with Crohn's disease were followed up and samples were collected throughout their EEN course and during food re-introduction. Patients who had already received a diagnosis of Crohn's disease and were undergoing a repeat course of EEN were also recruited. Patients who had received antibiotics in the preceding month were excluded. Written assent or consent was taken from participants and their carers according to good clinical practice. The study was approved by the NHS West of Scotland Research Ethics Committee (14/WS/1004) and was registered in clinicaltrials.gov (NCT02341248).

| Exclusive enteral nutrition and food reintroduction
Patients were treated for 8 weeks with a polymeric, casein-based formula (Modulen IBD). Paediatric dieticians calculated the feed volume to provide the energy requirements for each child based Conclusions: The effect of exclusive enteral nutrition on faecal calprotectin is diminished early during food reintroduction. Maintenance enteral nutrition at ~18% of energy intake is associated with a lower faecal calprotectin at the early phase of food reintroduction but is ineffective in maintaining longer term remission. on the UK age and sex recommendations. 31 Children who were undernourished (BMI Z-score < −2 SD) were prescribed a 10% higher energy intake while on EEN. No other food was allowed during the EEN course, with the exception of black tea or coffee, water, or a sodium-benzoate free carbonated sugar-containing lemon and lime soft drink (7UP), clear mints and syrup flavouring. Patients were encouraged to consume EEN orally, however, if patients were unable to take their prescribed formula volumes within the first 3 days, patients were then switched to nasogastric tube feeding. While on EEN, patients received regular nurse and dietetic weekly support calls and were reviewed in the clinic 4 weeks post initiation.
If the participant did not clinically improve or deteriorated within the initial treatment period, EEN therapy was discontinued and an alternative induction treatment was commenced, usually oral corticosteroids.
Patients who successfully completed their course of EEN and entered clinical remission (defined as a decrease in weighted paediatric Crohn' wPCDAI <12.5) were encouraged to continue using their enteral formula as an oral supplement (MEN), consuming approximately 20%-25% of their total EEN volume, in addition to their free habitual diet. Patients had a rapid food reintroduction, with no specific dietary recommendations given by the members of the clinical team after the end of EEN, other than to offer MEN.

| Clinical assessment
Measurements of anthropometry were obtained and calculated as Z-scores using the UK-WHO growth charts. 32 C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), serum albumin and haemoglobin were measured at EEN initiation, end of treatment and at routine clinic appointments post-EEN. Clinical disease activity was defined using the weighted paediatric Crohn's disease activity index (wPCDAI). 33 Use of maintenance drug therapies were also recorded.

| Faecal calprotectin measurements
Fresh faecal samples to measure faecal calprotectin were collected from children at enrolment, before they began treatment with EEN.

| Dietary assessment
During food re-introduction and prior to sample collection, at 2 weeks and 2 months post-EEN, the energy intake of participants was estimated using 3-day estimated weight food diaries. Household measures were used to estimate portion size. Patients were asked to record compliance with MEN and the exact amount consumed on the same food diaries. These diaries were analysed into WinDiets 10 Software Suite (version 10; Robert Gordon University), to calculate energy intake.
Associations were explored between faecal calprotectin and both the volume and energy intake attributable to MEN. Energy intake was expressed in kcals, as percentage of patient's estimated average energy requirement (EAR) and as a percentage of energy intake. No MEN intake was assigned to patients who returned a food diary without reporting use of MEN.

| Rates of remission and relapse within 1-year post EEN
Patients were followed up to a maximum of 12 months, following completion of EEN. Clinical relapse was defined as deterioration of disease requiring a further induction therapy as dictated by the clinical caring team.

| Statistics
Continuous data are displayed with median and interquartile range, unless otherwise stated. Differences in rates of remission based on disease location were evaluated using a χ 2 test. Two-sample t tests were performed to compare differences between two independent groups in patient characteristics. Mann-Whitney U tests were performed to explore differences in proportional changes in faecal calprotectin during food reintroduction grouped together  and four (6%) had isolated ileitis. All patients presented with inflammatory (B1) disease behaviour (Table 1) (Table S1)
These inflammatory markers improved when patients completed EEN, with 12/40 (30%) having at least one marker outside the normal range. Among those who entered clinical remission or clinically improved, and whose blood measurements were available, 22/32 (69%) patients had all normal inflammatory makers (Table   S1).

| Faecal calprotectin concentration changes during EEN
From the 66 patients who started on EEN, a faecal sample was obtained for baseline faecal calprotectin measurement in 53 (80%).
Median faecal calprotectin at the start of treatment was 1433 mg/

| Cytokine profile changes during EEN
Prior to EEN initiation, no differences in cytokine concentration were observed between responders and non-responders. Using  (Table S3). There was no change to cytokine levels in patients who did not achieve clinical remission.  (Figure 2). The increase in faecal calprotectin during food re-introduction preceded changes to other disease activity markers (including wPCDAI, ESR and albumin) which remained similar to values at the end of EEN (Table S1).    (Figure 4).  It has also been shown that faecal calprotectin reverts to pretreatment levels within 4 months post-EEN, and that up to 55% of patients in clinical remission can have a raised faecal calprotectin. 23,38 Complementing this previous work, this more detailed study demonstrated that the rise in FC post EEN to pre-treatment levels occurs more rapidly than previously recognised. 23  The heterogeneity between studies and lack of a prospective RCT study in children remains a significant limitation in this area.

| D ISCUSS I ON
While our study is not an RCT, we have attempted to prospectively study the effect of MEN while at the same time assessing its compliance with self-reported weighted food diaries, a caveat not often captured within the existing literature. 11,18,45,46 With only half of the patients returning a food diary in which to assess MEN compliance this remains a limitation in our study but is useful to highlight many patients who are "prescribed" MEN do not actually take it, suggesting that better reinforcement from the treating team may help improve compliance.
As Our data on clinical response to treatment are similar to previous reports but while patient's weight increased too, the magnitude of this effect was not as large as published historically. 5,53 This may be explained in part by our patients not having as low z-scores suggesting improved detection and earlier diagnosis. The weight gain we observed in this study is similar to other published literature. 5,44 We have demonstrated that changes in patient's weight during the first 4 weeks of treatment with EEN predicted which patients would enter clinical remission by the end of EEN.
While this is very similar to the prediction validity using changes in faecal calprotectin during the same period, using weight change as a predictor creates a circular argument with assessing disease improvement using wPCDAI. We were unable to perform this analysis using changes in systemic markers of inflammation (CRP, ESR, inflammatory cytokines) as it is not the practice in our centre to routinely venepuncture patients at the midpoint of EEN, who are responding well clinically.
In conclusion, the current study confirms previous observa- in this study should be ascertained formally within well-designed dose-response studies.

ACK N OWLED G EM ENTS
Declaration of personal interests: None.

AUTH O R S H I P
Guarantor of the article: Richard K Russell.
Author contributions: ML carried out and co-ordinated most of the research activities and laboratory analysis, statistical analysis and produced the first draft for publication. CMC carried out patient recruitment, sample collection and part of the laboratory analysis.
UZI contributed to the study design, ethical application, and funding award, and statistical analysis, and co-supervised the researcher. LG