All members of the working group for pediatric IBD treatment guidelines were contacted via e-mail and invited to participate in a survey in July 2012. The working group members included 32 of the 36 delegates to the Japanese Society of Pediatric IBD (23 pediatric gastroenterologists, 6 gastroenterologists, and 3 gastrointestinal surgeons). The majority of hospitals involved in the treatment of pediatric CD are members of this society. The survey participants were asked to complete and return the 2 questionnaires described below via e-mail by the end of August 2012. A second e-mail was sent after 4 weeks as a reminder to those who had not yet responded.

Two types of questionnaires were developed for this study. The 16-item consensus questionnaire assessed the attitudes of physicians toward EN. The physicians were asked about their EN standard protocols, including whether they used EN as induction and maintenance therapy in pediatric patients with CD. They were also asked whether they believed that EN affected their patients' quality of life (QOL). If they indicated that in their opinion, EN worsened the patients' QOL, they were asked to report what interventions they recommended to minimize this effect.

The 14-item efficacy questionnaire collected data on patients aged <17 years who received induction therapy between 2006 and 2011. It assessed treatment details and patient demographics. Patients were diagnosed with CD based on standard criteria.¹² To eliminate bias that could have resulted from differences in the number of patients treated in different facilities, physicians were asked to provide information regarding the last 5 patients they had treated, if they had treated >5 patients. Survey items included questions on patient profiles, number of relapses, types of treatments administered as induction therapy, formulas used for EN, EN duration, and whether remission was achieved without additional treatment (Table 1). Pediatric Crohn's Disease Activity Index (PCDAI) scores at the initiation of EN treatment, 2 weeks after treatment initiation, and at the end of treatment were also obtained. Differences in disease behavior and PCDAI scores at onset and 2 weeks after initiation of treatment between patients who did and those who did not achieve remission with EN were analyzed.

Remission was defined as an International Organization for the Study of Inflammatory Bowel Diseases score of 0 or 1, and negative CRP, based on the Japanese guidelines for CD. EEN was defined as liquid enteral nutrient intake for all peroral or pernasal nutrition, irrespective of intravenous supplemental nutrition. Partial enteral nutrition (PEN) was defined as a combination of liquid EN and the “Crohn's diet” (low fat and low fiber diet). Regarding types of formula, the commercial product names of liquid EN products available in Japan (Elental® and Elental P®, Ajinomoto Pharmaceuticals, Tokyo, Japan, for elemental formula; Twin-line NF®, Otsuka Pharmaceutical, Tokyo, Japan; Peptino®, Terumo, Tokyo, Japan, for oligomeric formula; Ensure Liquid®, Abbott Japan, Tokyo, Japan; and Racol®, Otsuka Pharmaceutical for polymeric diet) were listed in the questionnaire.

The questionnaire data included only de-identified, unlinkable, and anonymous data. The study design qualified for an exemption from the requirement for informed consent by the Gunma University Hospital Institutional Review Board.

Assessment of the normality of the data was performed using SPSS version 22.0 (IBM Corp., Armonk, NY, USA). Fisher exact test was used to investigate the differences in disease behavior between patients who did and did not achieve remission with EN. The Wilcoxon signed-rank test was used to compare PCDAI scores before and after treatment with EN.

Of the 32 physicians we contacted, 23 (19 pediatric gastroenterologists, 2 gastroenterologists, and 2 surgeons) responded to the questionnaires.

Data from 58 patients treated in 15 hospitals were examined in the efficacy questionnaire. Only 1 case was reported from an adult gastroenterology department, whereas the others were reported by pediatric gastroenterologists. Five hospitals reported treating fewer than 5 cases. Of the 58 patients, 42 (72%) were male patients. The age at diagnosis ranged from 0 to 16 years, with a median age of 12 years; the median patient age at the beginning of treatment was 13 years. A total of 44 patients (77%) had newly diagnosed CD, whereas 14 patients were treated for a CD relapse.

Of 22 physicians who responded to this questionnaire, 18 (82%) recommended treatment with EN for all newly diagnosed pediatric patients with CD, whereas 4 (18%) stated that they would consider administering EN only in certain situations, such as in patients with small intestinal involvement (3/4), growth delay (2/4), or in cases of adherence to EN by patients and parents (2/4) (Table 1). In summary, all physicians recommended EN as a treatment option. In cases of CD relapse, 13 of the 22 physicians (59.1%) recommended EN for all patients, whereas 8 (36.4%) recommended it only for those who had responded to EN during the first induction therapy. Regarding maintenance therapy, EN was recommended for all patients by 14 of the 22 physicians (63.7%); in contrast, 6 (27.2%) recommended EN depending on patient phenotypes (2 severe cases, 2 patients with small intestinal involvement, 1 patient with a growth delay, and 1 patient based on age).

Of the 22 physicians, 18 (84%) recommended EEN, as opposed to PEN, as induction therapy for all patients. Regarding the type of formula used for induction therapy, elemental formula was preferred by 18 of 21 of physicians (86%), whereas a polymeric diet was preferred by 2 physicians (10%). EEN duration was determined based on symptomatic improvement according to 16 of 22 physicians (73%), whereas a limited treatment duration was used by 5 of 22 physicians (23%). A polymeric diet was used as maintenance therapy by 10 of 23 physicians (43%). The recommended caloric intake obtained from the enteral formula during maintenance therapy was approximately 30% of the total oral intake, according to the majority of the physicians.

When asked about the perceived effects of EN on their patients' QOL, 18 of 23 physicians (78%) reported that the use of EN as induction therapy had a negative effect on their patients' QOL. Enhanced support by medical staff and family members was cited as the strategy most likely to improve the QOL of pediatric patients with CD as well as to increase their adherence to treatment with EN.

The disease phenotypes of all patients at the beginning of induction therapy are shown in Table 2. The majority of patients (45/58, 78%) had small intestinal involvement. A total of 4 of 58 patients (7%) had strictures, although none had internal fistulas. PCDAI scores prior to treatment—available for 32 of the 58 patients—ranged between 10 and 77.5 (median score, 35). All patients except 1 received treatment with EN. Treatment details are shown in Table 3. Among the 57 patients treated with EN, 40 (70%) and 17 (30%) received EEN and PEN, respectively. Regarding formula types, the elemental formula Elental® was most commonly used, whereas a polymeric diet was used in only 2 patients (4%). Among patients treated with EN (including EEN, PEN, and Crohn's diet), immunosuppressive treatments such as steroids or immunomodulators were simultaneously used in 17 patients (30%). A total of 31 of 40 patients (80%) treated with EN alone achieved remission without the requirement of any additional treatment other than 5-aminosalicylic acid, which was administered to 56 patients (97%). Among the 40 patients who received EEN, 14 (35%) simultaneously received immunosuppressive therapy. Of the remaining 26 patients, 22 (85%) achieved remission without the need for additional therapy. The mean duration of EEN administration was 15.9 days. Of the 6 patients who received EEN for >30 days, 3 (50%) did not achieve remission.

To evaluate whether PCDAI scores during the early phase (after 2 weeks) of EEN treatment could predict the efficacy of EEN, scores at the initiation of treatment and 2 weeks thereafter were analyzed for the 21 patients with available data. Among the 17 patients who responded to EEN, 9 had PCDAI scores <10 and 5 had scores of 10 to 20 at 2 weeks after the initiation of EN treatment. Only 1 of 4 patients who did not achieve remission with EEN showed a PCDAI score of <20. Thus, PCDAI scores 2 weeks after EEN initiation were significantly higher in nonresponders than in responders (P <0.05, Wilcoxon signed-rank test) (Fig. 1).