Crohn’s disease (CD) is a chronic inflammatory intestinal disorder that involves transmural inflammation of the GI tract and can potentially cause stricture and fistula formation [1]. Stricture in the small intestine is one of the most common intestinal complications and sometimes requires intestinal resection. The cumulative surgery rate after diagnosis of CD is reported to be 46%-62% at 5 years and 61%-75% at 10 years [2-4]. In addition, a greater number of strictureplasties is associated with a higher the risk of reoperation [5]. To avoid the risk of short bowel syndrome, intestinal resection should be kept to a minimum [6].

Endoscopic balloon dilatation (EBD) has been shown to be an effective treatment that helps to avoid surgery [7-10], and the Japanese clinical practice guidelines for IBD recommend EBD when obstructive symptoms do not improve with drug therapy alone [11]. Lian et al. [12] reported that EBD could delay surgery for 6.45 years in a study of 176 EBD-treated and 131 surgically treated CD patients. Although it is reported that EBD failure, strictures with fistula, long stricture, and multiple stenoses are factors that contribute to intestinal resection in CD patients with stenosis in the small intestine [13-17], there are few reports about the long-term outcomes of EBD in CD patients with strictures in the small intestine.

We were given the opportunity to use prototype single-balloon enteroscopes (SBEs) with a passive bending mechanism and high force transmission insertion tube. Colonoscopes with this mechanism and insertion tube have been reported to reduce the cecal intubation time and pain during the procedure [18,19]. In a prospective study of 60 patients, the use of a prototype SBE with a passive bending mechanism and high force transmission insertion tube was found to shorten of terminal ileum intubation time compared with use of the SIF-Q260, a conventional SBE [19]. However, in that study, the SBE was inserted only 20 cm from the ileocecal valve, and the results mainly reflect improvement of insertability in the colon.

We conducted this retrospective study to test our hypothesis that the prototype SBEs with a passive bending mechanism and high force transmission insertion tube can be inserted deeper than a conventional SBE in the small intestine in patients with stricturing CD who required EBD. We also analyzed whether the type of scope used affects the long-term outcomes of EBD.

A total of 94 CD patients with a mean follow-up period of 1,886 days were included in this study. Baseline characteristics at initial EBD are shown in Table 1. The mean number of EBDs performed per patient during the follow-up period was 3.3, and the mean interval between the initial EBD and the second EBD was 508 days. Among 83 patients who underwent successful EBD, 80 (96.4%) were initially dilated to more than 15 mm. Of the remaining 3 patients, 1 was initially dilated to 12 mm and 2 were initially dilated to 10 mm.

The EBD success rate was 88.3% (83/94). The 5- and 10-year cumulative surgery-free rate was 75.7% and 72.8%, respectively (Fig. 2A). Table 2 shows the results of multivariate analysis of factors contributing to intestinal resection. Elevated CDAI, long stenosis (≥2 cm), and EBD failure were identified as factors contributing to intestinal resection. The results were also confirmed by Kaplan-Meier analysis (Fig. 2B-D). Factors such as prototype scope used for initial EBD, number of stenoses, and de novo stenosis were not associated with intestinal resection. In addition, the risk of surgery increased by 2.53 times when the length of stenosis increased by 1 cm in a successful EBD case (HR, 2.53; 95% CI, 1.580-4.069; P<0.001).

During the follow-up period, 314 EBDs were performed in 94 CD patients, and 4 patients (1.29%, 4/314) had intestinal perforation as a complication. There were no cases of bleeding requiring surgery or blood transfusion.

The insertability and the scope used were compared between the groups shown in Table 3. We compared depth of insertion, procedure time, and number of EBDs between 2 procedures in the same patient at different times (Fig. 3). Compared with conventional SBE, prototype SBE has significantly improved the depth of insertion (Fig. 3C). A representative case is shown in Fig. 4. Use of the prototype SBE improved the depth of insertion and increased the number of EBDs.

The prototype SBEs with a passive bending mechanism and high transmission insertion tube did not affect long-term EBD outcome, but they did contribute to improving the depth of insertion in the small intestine compared with the conventional SBE. This is the first study to show that prototype SBEs with this mechanism and insertion tube improve insertability in the strictured small intestine.

The EBD success rate in this study was 88.3%, which is consistent with the 62% to 100% success rate reported in a pooled analysis and meta-analyses [10,22,23]. Our 5- and 10-year cumulative surgery-free survival rate of 75.7% and 72.8%, respectively, was equivalent to the 3-year rate of 73% reported by Hirai et al. [16] and are slightly better than the 1- and 3-year survival rates of 63.1% and 56.2%, respectively reported by Nishida et al. [13]. At our institution, patients who undergo EBD typically repeat procedures every 1 to 2 years for surveillance and evaluation regardless of the presence or absence of stenotic symptoms. In fact, 23 of the 94 patients (24.4%) had asymptomatic stenosis. The patients reported by Nishida et al. [13] had stenotic symptoms, which may be the reason for the differences in results.

In an EBD study involving strictures in both small and large intestine, it was reported that the HR for intestinal resection increased by 8% when the length of stenosis increased by 1 cm. [10]. In our study, an increase in the length of stenosis in the small intestine by 1 cm increased intestinal resection by 2.53 times. Our results suggest that increased length of stenosis in the small intestine may require more intestinal resection than increased length of stenosis in the colon.

Multivariate analysis revealed that long stricture (≥2 cm), EBD failure, and elevated CDAI were factors contributing to intestinal resection of small intestinal stricture. Long stricture (≥2 cm) [17] and EBD failure16 have already been reported as factors contributing to intestinal resection. Although elevated CDAI has been shown to be associated with intestinal resection in CD, [24] these cohorts comprised 23% (20/86) colonic stenosis and 8% (7/86) upper GI stenosis. Therefore, the novel finding in this study is that the elevated CDAI could be related to intestinal resection even in a cohort limited to the small intestine.

Until now, there have been no reports on achieving deep insertion in the small intestine using a prototype SBE with a passive bending mechanism and high force transmission insertion tube. Although Hosoe et al. [19] reported that the rate of insertion to the terminal ileum 20 cm beyond the ileocecal valve within 10 minutes was improved from 86.2% to 96.8% by using a prototype SBE in a prospective study, the deep insertability in the small intestine has not been studied. Our findings indicate that the depth of insertion can be significantly improved by using these prototype SBEs in patients with stricturing CD.

The prototype SBEs did not contribute to the cumulative surgery-free survival rate (data not shown). Depth of insertion was increased with the prototype SBEs compared with the conventional SBE, but this difference did not directly improve the approach to stenosis located deep in the small intestine. In fact, the prototype SBEs did not increase the number of EBDs compared with the conventional SBE (2.1±1.8 vs. 1.9±1.4) (Fig. 3I). However, there appeared to be some cases where stenosis could be approached using only a prototype SBE, so further accumulation of cases is necessary in the future.

There are some limitations to this study. First, this is a non-randomized retrospective study. Therefore, the procedure time during insertion in the small intestine was not recorded. Because the procedure time includes the time of insertion in the large intestine, duration of EBD, and observation period, it was not possible for us to analyze the insertion time in the small intestine. Second, depth of insertion was measured using the method described by May et al., [21] but this method may be less accurate than the method using crystal violet reported by Takenaka et al. [25]. Third, the timing of the procedures compared to analyze the insertability was different although these procedures were performed in the same patient. Therefore, the condition of the patient may have changed. Generally, the pathology of CD is progressive, so insertion of the enteroscope becomes increasingly difficult as intestinal deformity progresses. It should be noted that the insertability improved even though most of the procedures with a prototype SBE were performed later than those with the conventional SBE.

In conclusion, the prototype SBEs with a passive bending mechanism and high force transmission insertion tube did not improve long-term EBD prognosis, but they did improve the depth of insertion compared with a conventional SBE. In addition, elevated CDAI was identified as a factor contributing to intestinal resection even in a cohort limited to CD patients with strictures in the small intestine.