SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects And Methods
  5. Result
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

Background

Exclusive enteral nutrition (EEN) is an effective first line treatment for active paediatric Crohn's disease (CD).

Aim

To examine the effect of EEN on short- and long-term clinical outcome together with anthropometric measurements.

Methods

Retrospective case-note review in newly diagnosed CD (<16 years) who completed 8 weeks of EEN. Demographics, anthropometry, disease characteristics and inflammatory markers were collected at EEN initiation and at 1, 2, 6, 12 & 24 months post treatment initiation. EEN response was determined by a patient global assessment.

Results

One hundred and nine patients were included (Males 68; Median age: 11.2 years). After 8 weeks EEN, 65 were in remission, 32 improved and 12 had no improvement. By 4 weeks, mean weight/BMI z-score (s.d.) increased (P < 0.02) and between 4 and 8 weeks (P < 0.05). Baseline inflammatory markers all improved significantly by week 4 (albumin, CRP and platelets; all P < 0.01) and ESR (P < 0.00001). 63/109(58%) relapsed during follow-up. 44/63(70%) patients completed a second course of EEN with similar response rate, but lower weight gain (3.3 vs. 5.1 kg, P < 0.05). Height z-score did not change significantly over the 24 months. Introduction of azathioprine within 6 months of diagnosis did not improve height outcomes at 24 months.

Conclusions

Weight and BMI z-score improved with EEN and changes are sustained to 2 years, but height z-score did not. Seventy per cent of patients who relapsed during 2-year follow-up managed a 2nd course of EEN. The optimal therapeutic strategies for length of EEN course and to improve linear growth are awaited.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects And Methods
  5. Result
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

Crohn's disease (CD) is a chronic inflammatory condition of the gastrointestinal tract with a continuous relapsing and remitting course. It can affect any part of the gastrointestinal tract and can occur at any age, with up to 25% of cases diagnosed in childhood.[1] CD can have a significant impact on growth, pubertal development and quality of life as well as future employment potential.[2] The incidence of paediatric CD has increased recently, not only in Scotland, but throughout the world making optimal management of these patients a priority.[3, 4]

The presentation of CD is variable with only 25% of children presenting with the classic triad of weight loss, abdominal pain and diarrhoea, with 85% of children presenting with CD reporting weight loss before or at the time of their diagnosis.[5] Growth failure is a common problem in CD commonly affecting many children at diagnosis.[6, 7] Current literature suggests that between 11 and 37% of children diagnosed with CD in childhood have a deficit in their adult height compared with expected height.[8-10] At present, there are no specific treatments to date, which have been proven conclusively to impact positively on final growth outcomes.[6, 11-13]

Exclusive enteral nutrition is the primary treatment for active CD in the UK and many centres across Europe.[14-16] It is effective in inducing clinical remission in the majority of cases at a rate comparable to that of oral steroids.[14, 17, 18] In addition, it provides nutritional support and rehabilitation.[17] With regard to the latter, there is clinical experience to suggest that EEN improves anthropometry and may promote growth (at least in the short term).[17] There is definite evidence to show that weight improves after an initial course of EEN.[14] However, to date, the impact of EEN on anthropometry and measures of nutritional status and growth during medium term follow-up have not been documented in great detail. At present, there is no specific evidence to suggest what the optimal duration of nutritional therapy should be and the current practice in most units is mostly guided by local clinical experience.[19] Whether there are sustainable differences in clinical response to EEN between shorter and longer courses or if duration of EEN is associated with improved clinical outcome and disease prognosis at follow-up has not yet been explored.

The aim of this study was to explore the impact of EEN on anthropometric parameters in children with CD at baseline, 4 and 8 weeks and their association with predicting anthropometry at 6, 12 and 24 months. Our secondary aims were to evaluate the magnitude of these changes between the first two consecutive courses of the same treatment within a follow-up period of 2 years.

Subjects And Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects And Methods
  5. Result
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

Patient selection

The study population was newly diagnosed paediatric patients with CD who were treated at the Royal Hospital for Sick Children, Glasgow, which serves as the sole tertiary paediatric referral centre for the West of Scotland. Children with significant co-morbidity affecting growth, those with no baseline measurements, and those who were not treated with EEN, as their primary treatment at disease diagnosis were excluded from analysis. Patients who discontinued EEN within 15 days were also excluded. All children were diagnosed according to established radiological, histological and endoscopic criteria.[20, 21] We have previously published data on short-term outcomes after EEN. A proportion of these patients have also been included in the current study, which reports on longer term outcomes.[17]

Nutrition

The standard feed used was Modulen IBD (Nestle, Croydon, UK), with E028 (SHS Nutricia, Liverpool, UK) reserved for patients with proven co-existing cow's milk protein allergy. Feeds were delivered orally, or a nasogastric tube was used for patients unable to manage oral feeding, for a planned course of 8 weeks. Nutritional intake was based on the child's estimated energy requirements (EAR).[22] Normal diet was then reintroduced over 5–7 days after a course of EEN was complete, initially sloppy foods were given, then normal unrestricted diet.

Data collection

Data was collected using patient's dietetic, medical and laboratory records. Anthropometric data (height, weight) were collected at 0 (diagnosis), 4 (during course of EEN) and around 8 weeks (course completion) and at 6, 12 and 24 months. Weight was measured with minimal clothing using calibrated chair scales to the nearest 100 g and height with wall mounted stadiometers according to the Frankfurt plane position. Systemic markers of disease activity [Erythrocyte Sedimentation Rate (ESR)‚ C-Reactive Protein (CRP), platelet count (PLT) and serum albumin (Alb)] were collected from patients' case records. These were assayed according to standard clinical procedures (local reference ranges used are available on request). Patient demographics, disease phenotype and medical treatments were collected. Disease phenotype was classified using the Montreal criteria.[23] Clinical response to EEN was determined by a patient global assessment. A patient global assessment was assigned after a multidisciplinary assessment of response to treatment following review of history, examination, anthropometry and laboratory values. Children were divided into three categories: remission, improvement (response) and no improvement (active disease) as described previously.[17]

Subjects with available data continued to be followed up at 6, 12 and 24 months after diagnosis. Permission to carry out this study was obtained from the local Clinical Effectiveness Office, which clarified that no formal ethical approval was needed.

Statistical analysis

Descriptive data is presented with medians and interquartile ranges or with means and standard deviations. z-scores are used to quantify how much changes in measurements (in this case anthropometry) are relative to the mean with a – applied when a value lies below the mean and + above. A z-score of 1 is equal to 1 s.d. away from the mean. Changes during the course of EEN were explored with analysis of variance of repeated measures and Bonferroni post hoc correction. Differences between first and second course of EEN were explored with Wilcoxon signed rank test. Significance levels were set at a P-value <0.05. Analysis was carried using Minitab version 15, (Coventry, UK).

Result

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects And Methods
  5. Result
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

One hundred and sixty-one new patients with CD were identified. Fifty two did not meet the inclusion criteria and were excluded: five were newly diagnosed patients who had not yet begun treatment, five children did not have a baseline treatment recorded in their case notes, two did not match the specific diagnostic criteria for CD, one child had significant co-morbidity impacting on growth potential, and two children had only baseline measurements recorded. Fifteen had been transferred to adult care so their notes were not available for follow-up review at 2 years and 22 children received treatment with steroids at some point in the primary induction course of EEN due to failure to respond, leaving 109 children available for further analysis.

There were 68 boys. The median age at diagnosis was 11.2 years (3.3–15.7). Thirty-nine percent had ileo-colonic (L3 + L4) disease at presentation and the most common behaviour was inflammatory (B1) in 94. One hundred and seven children received Modulen and two received E028.

Mean body mass index (BMI) z-score at diagnosis was −1.3 (1.5); 34% of children were thin at diagnosis (BMI ≤−2 s.d.), and 1% were obese (BMI ≥ 2 s.d.). The mean height z-score at EEN initiation was −0.6 (1.1) and 10% of children had short stature (Height ≤−2 s.d.). The mean weight z score at diagnosis was −1.1 (1.3) and 25% of children were underweight (Weight ≤−2 s.d.).

Short term follow-up (0–8 weeks, Table 1)

Children were reviewed at a mean of 30 (s.d.: 9) days after initiation of EEN and at completion of EEN 59 (s.d.: 11) days. At 4 weeks, there was a mean weight increase in 2.5 kg (2.7) compared to baseline (P < 0.001) and by 8 weeks a total increase in 4.3 kg (3.0) compared to baseline (P < 0.001). Mean weight gain between 4 and 8 weeks was 1.7 kg (2.0) (P < 0.001) compared with weight at 4 weeks. At initiation of treatment, weight z score was -1.1 (1.3), which improved to −0.4 (0.9) at 8 weeks (P < 0.001). Mean BMI z score increased from −1.3 (1.5) at baseline to −0.4 (1.0) at 4 weeks (P < 0.001) (Table 1). There was a negative correlation between weight (r = −0.65, P < 0.0001) and BMI z score (r = −0.76, P < 0.0001) at diagnosis and their change at the end of EEN, showing that those who were more underweight with a lower BMI at diagnosis showed the greatest improvement in weight gain.

Table 1. Changes in anthropometry during initial course of EEN
 NumberBaselineNumber

∆ 0–4 weeks

30 ± 9.3 days

Number

∆ 0–8 weeks

58.7 ± 11.3 days

Number

∆ 4–8 weeks

29.8 ± 12.8 days

  1. A z-score of 1 is equal to 1 s.d. away from the mean.

  2. a

    All P < 0.02 compared to ‘Baseline’.

  3. b

    All P < 0.02 compared to ‘4 weeks’.

  4. c

    All P < 0.05 compared to ‘∆ 0–4 weeks’.

Age11011.2 (2.4)      
Weight10931.8 (10.4)942.5 (2.7)a974.3 (3.0)b811.7 (2.0)c
Weight mean z score109−1.1 (1.3)940.5 (0.5)a970.8 (0.6)b810.2 (0.3)c
Height106140.9 (15.9)720.4 (2.2)930.08 (2.3)590.5 (1.1)
Height z mean score106−0.6 (1.1)72−0.01 (0.3)93−0.05 (0.2)59−0.06 (0.3)
BMI10615.6 (2.5)721.3 (1.4)a930.9 (1.2)b592.1 (1.4)c
BMI z mean score106−1.3 (1.5)720.9 (1.0)a931.3 (0.9)b590.5 (0.7)c
Inflammatory markers (Table 2)

All inflammatory markers improved significantly by 4 weeks of treatment with further improvement until 8 weeks seen only for ESR, although it should be noted the median ESR was already within the normal range at 4 weeks.

Table 2. Improvement in inflammatory markers seen in first 4 weeks of initial EEN
 Number of samples0 weeks4 weeks P-value cf. 0 weeks 8 weeks P-value cf. 0 weeks
ESR (median, IQR)8640.0 (33.2)15 (23)<0.000111 (17.5)<0.0001
CRP (g/dl) (median, IQR)7029 (35)12 (8)<0.017 (0.25)<0.0001
Albumin (g/dl) (median, IQR)9631 (9)36 (6)<0.000138 (5)<0.0001
Platelets (median, IQR)103515 (232)419 (188.8)<0.003387 (153.3)<0.0001
Clinical disease activity (Table S1)

At the end of EEN 65 children achieved clinical remission, 32 improved and 12 continued to have active disease (data was missing for one child). All children gained weight at the end of EEN, but this varied according to their clinical response to the treatment. In those who had active disease at the end of EEN mean weight gain 0–8 weeks of EEN was 2.2 kg (2.1) compared with those who entered remission (n = 65) who gained a mean of 5.1 kg (3.1) (P = 0.04). In those who responded but did not achieve remission the mean weight gain was 3.3 kg (2.8) at the end of EEN which was significantly lower than those children who entered remission (P = 0.03) but still a significant increase compared to baseline (P < 0.001).

Longer term follow up (6–24 months, Figure 1)

Weight and BMI z-scores improved during follow up compared to baseline with the greatest improvement seen during the first course of EEN. Height z-score showed no significant improvement from −0.6 at baseline (1.1) to −0.8 (1.0) at 24 months. Weight and BMI z-score at diagnosis were the strongest predictors of their values at any point of the follow up. There were 63 of 109 patients who experienced a clinical relapse during follow up to 24 months post diagnosis. The median time to relapse was 6.5 months (IQR: 7 months) from induction of EEN. The size of weight or BMI z-score change at the end of the first course of EEN did not predict time to subsequent clinical relapse or anthropometry and growth status at 6, 12 and 24 months in all children.

image

Figure 1. Anthropometric changes in 109 children treated with 8 weeks of EEN show significant improvement of BMI and weight z-score sustained until 2 years of follow-up. No significant change with height z-score over the 2-year period. A z-score of 1 is equal to 1 s.d. away from the mean.

Download figure to PowerPoint

Response to EEN was associated with increased height velocity up to 6 months only in patients who achieved remission compared to those who had no response (P = 0.02) but this was predominantly because those with no response dropped height z-score significantly (full data Table S2).

Concomitant medication

The cumulative use of medication until maximum follow up at 2 years was azathioprine 83, aminosalicylates 73, infliximab 13, methotrexate 33, antibiotics 39 and steroids 25. As we have previously published growth outcomes relating to specific medications in detail previously[6] and given the relative proportions of medications used in this dataset we only analysed the use of early azathioprine (within 6 months of diagnosis) and height outcomes. Early use of azathioprine was associated with significant differences in height z-score at follow-up compared to patients who did not get early azathioprine but this related most likely to changes present from diagnosis as height in both groups fell after diagnosis (Table S3).

First and second course of EEN

Paired data was collected for 44 relapsed patients who had varying responses to an initial EEN course but then went on to complete a second course of EEN during the 2 year follow up period. During the first course mean weight increase during EEN was 5.1 kg (3.1) compared with 3.3 kg (2.3) during the second course (P = 0.001). Likewise mean change in weight z-score at the end of the first course was 1.0 (0.6) compared with 0.5 (0.4) after completion of the second course (P < 0.0001), both of these were significantly improved compared to baseline data at the start of the EEN course (P < 0.05). Similarly mean BMI and BMI z-score improved during first and second course but the improvement was greater during the first course 1.6 (0.9) (P < 0.05) compared to an increase in 1.3 (1.1) after the second course (P < 0.0001).

During the first course of EEN, 29/44 children achieved remission compared to 25/44 (57%) during the second course (P = 0.51). After both courses 2/44 children continued to have active disease. Of the 29 children who achieved remission after an initial course only 1 child continued to have active disease after a second course (1 child had missing data) and 2 showed improvement but did not achieve remission. In the 2 children who had active disease after their initial course of EEN but went on to have a second course, both then achieved remission when given a second course.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects And Methods
  5. Result
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

EEN has long been established as a treatment for paediatric CD both as induction treatment for those with a new diagnosis and for those with established disease.[14] EEN has been shown in meta-analyses to have similar efficacy to corticosteroids in paediatrics with the added benefits of improving nutritional status as well as avoiding steroid related side effects.[18, 24] We have shown a significant improvement in weight and BMI after completion of a first course of EEN which was sustained to the end of the 2-year follow-up. We have demonstrated that in children who received a second course of EEN, a similar rate of clinical response and an increase in weight but this was significantly less than during the first EEN course. Furthermore, the greatest improvement in weight and inflammatory markers were seen within the first 4 weeks of treatment.

Clinical response rates after a course of EEN vary from 60% to 88.5%[14, 17] and are similar to the remission/response rate of 89% reported in this study. In addition to inducing remission, EEN had the benefit of improving anthropometric parameters after the initial course as well as at follow-up.[13, 25, 26] The amount of weight gain during an 8 week course of EEN varies[26-28] which could be explained in part by different participant characteristics and study designs. Different types of formula have been used in various studies but according to current literature there are no compositional differences between them which should affect growth parameters.[29]

In studies of growth in children with CD up to 50% have poor linear growth at diagnosis.[14] Day showed improvement in rate of growth as early as 10 weeks after initiation of EEN with an increase in 3 cm, however there was no improvement in height z-scores at follow up, similar to the results of the current study.[12, 27]

Our results, as many others have shown, are that although children do gain height with treatment including EEN and have an appropriate gain in height, they continue to maintain growth on the same centile tracking as previously. This is despite the significant improvement in weight suggesting that although the children are nutritionally rehabilitated this does not translate directly to improvement in growth. Previous work examining height outcomes have shown short term improvement in height velocity in patients treated with infliximab.[30, 31] However, when Pfefferkorn et al. considered the longer term effects on growth velocity with a 2 year follow-up of a large cohort of children from North America no sustained effect was seen for infliximab.[13] Furthermore, no significant improvement was seen in height z-score at 2 year follow-up with any form of treatment. Despite extending follow-up to 6 years there continues to be no statistically significant improvement in growth velocity suggesting EEN like all other paediatric CD therapies including infliximab has not yet been shown to change growth patterns in the long term.[11] Other studies have quantified this impact on final adult height demonstrating a mean reduction of 2.4 cm below that expected, with 20% of individuals more than 8 cm below their target range.[10] In animal models of induced colitis, growth has been reduced by 30–40% which occurred independently of nutrition. In addition, when nutritional supplements were introduced to the group with colitis, weight improved but the growth deficit was maintained compared to controls.[32]

In a small study by Day et al. improvement was shown in inflammatory markers (ESR, CRP, albumin and platelets) after 8 weeks EEN in 12 newly diagnosed CD patients who achieved remission.[14] However, others have shown improvement in ESR and CRP over a much shorter space of time (1 week).[33, 34] These improvements in inflammatory markers have paralleled changes in PCDAI, both seen as early as 4 weeks. In the Borelli et al. study, no further significant change in either PCDAI or inflammatory markers was seen between 4 and 8 weeks.[34] This data along with the larger dataset in the current study could, therefore, suggest that as the most significant improvement in weight and inflammatory markers is seen in the first 4 weeks of EEN the course could be discontinued or changed to supplemental rather than exclusive feeding at this time. However, consistent with other studies, we have shown that weight continues to improve up to 8 weeks of treatment.[27] So does the additional degree of nutritional benefit outweigh the adherence issues to longer courses? These issues are difficult to answer with our or indeed any retrospective study therefore should be answered by a prospective randomised trial comparing different lengths of treatment course with appropriate long term follow-up. Of note, all the studies that have demonstrated mucosal healing from EEN have been after 8 weeks or more of EEN.[34, 35] In addition, faecal inflammatory markers have been demonstrated to continue to fall upto 8 weeks of EEN suggesting until further evidence evolves that 8 weeks is the optimal course length based on data from current studies.[36]

There is a paucity of published data describing clinical outcomes from second courses of EEN in children and whether the benefits already discussed in terms of improvements in anthropometric and inflammatory markers are reproducible in subsequent EEN courses. A recent small study (n = 27) suggested a higher remission rate following a course of 8 weeks of EEN in children with newly diagnosed CD compared to those with long standing CD (80% compared to 58%).[27] Our significantly larger cohort showed a comparable remission rate between first and second courses. In addition it demonstrates that even children who failed to respond to an initial course achieved remission with a second course. This may seem counterintuitive to give children a treatment on which they have already been seen to ‘fail’ however, recent evidence suggests that the long term inflammatory response is strongly influenced by the cytokine milieu which changes with time.[37] Therefore, even if a child fails with a treatment on the first occasion, the immune response could potentially be different allowing the treatment to work on a separate occasion. Furthermore, there is benefit in terms of improvement in weight and BMI (although not as significant as the initial course). Knight et al. described a median time to relapse of 54 weeks after a course of EEN in 44 paediatric patients. After remission was achieved, patients were asked to continue supplemental nutrition but only 50% were able to do so. Of the patients who relapsed and who chose a second course of nutrition all achieved a clinical response.[38] Based on these and our results it would be reasonable to consider using a subsequent course of EEN in both children who have and have not entered remission after an initial course of EEN for management of a disease relapse.

In conclusion, our study examined the effects of EEN used for induction of remission in paediatric CD and its effect on anthropometric measurements with sustained improvements in weight and BMI, though no improvement in height SDS. The greatest improvement in both weight and inflammatory markers were seen after 4 weeks of EEN, especially in those patients who achieved remission. A prospective study is now needed to compare children given 4 and 8 week courses of EEN to determine if comparable results can be achieved by shorter periods of exclusive feeding. In addition, further therapeutic strategies to maximise height potential in children with Crohn's disease need to be explored.

Authorship

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects And Methods
  5. Result
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

Guarantor of the article: R. K. Russell.

Author contributions: FLC, KG and RKR prepared the manuscript with comments and corrections by all the authors. AP, DM, TC, EB, VG, PM and AB helped collect the patient data. KG carried out the statistical analysis. All authors have read and approved the final draft.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects And Methods
  5. Result
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

Declaration of personal interests: Dr Richard K Russell has received speaker's fees, travel support, or participated in medical board meetings with MSD Immunology, Abbott, Dr Falk, Nestle and Ferring Pharmaceuticals. Dr Paraic McGrogan has received speaker fees, travel support and participated in medical board meetings with Nestle.

Declaration of funding interests: The Yorkhill IBD team is generously supported by the Catherine McEwan Foundation and the Yorkhill IBD fund. RKR is supported by a NHS research Scotland fellowship.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects And Methods
  5. Result
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects And Methods
  5. Result
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
apt12230-sup-0001-TableS1-S3.docxWord document14K

Table S1. Children who achieved remission gained more weight than those with active disease.

Table S2. Clinical response to EEN and height and height velocity at 6, 12 and 24 months.

Table S3. The use of azathioprine in the first 6 months after diagnosis and follow up growth parameters.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.