Systematic review: Safety of balloon assisted enteroscopy in Crohn’s disease

Abstract

AIM

To determine the overall and comparative risk of procedure related perforation of balloon assisted enteroscopy (BAE) in Crohn’s disease (CD).

METHODS

Systematic review (PROSPERO #CRD42015016381) of studies reporting on CD patients undergoing BAE. Seventy-three studies reporting on 1812 patients undergoing 2340 BAEs were included. Primary outcome of interest was the overall and comparative risk of procedure related perforation of diagnostic BAE in CD. Secondary outcomes of interest were risk of procedure related perforation of diagnostic double balloon enteroscopy (DBE), risk of procedure related perforation of therapeutic BAE, efficacy of stricture dilation, and clinical utility of endoscopically assessing small bowel disease activity.

RESULTS

Per procedure perforation rate of diagnostic BAE in CD was 0.15% (95%CI: 0.05-0.45), which was similar to diagnostic BAE for all indications (0.11%; IRR = 1.41, 95%CI: 0.28-4.50). Per procedure perforation rate of diagnostic DBE in CD was 0.12% (95%CI: 0.03-0.44), which was similar to diagnostic DBE for all indications (0.22%; IRR = 0.54, 95%CI: 0.06-0.24). Per procedure perforation rate of therapeutic BAE in CD was 1.74% (95%CI: 0.85-3.55). Eighty-six percent of therapeutic perforations were secondary to stricture dilation. Dilation was attempted in 207 patients and 30% required surgery during median follow-up of 18 months. When diagnostic BAE assessed small bowel disease activity, changes in medical therapy resulted in endoscopic improvement in 77% of patients.

CONCLUSION

Diagnostic BAE in CD has a similar rate of perforation as diagnostic BAE for all indications and can be safely performed in assessment of mucosal healing.

Core tip: Crohn’s disease (CD) affects the small bowel in up to 60% of patients, but evaluation of small bowel disease is often difficult. Balloon assisted enteroscopy (BAE) can evaluate the small bowel but its safety and diagnostic utility is not established. This systematic review includes 73 studies reporting on 1812 patients undergoing 2340 procedures to evaluate its safety and possible utility. We found that diagnostic BAE in CD had a similar rate of perforation as diagnostic BAE for all indications, suggesting BAE is a safe method for small bowel evaluation in CD.

INTRODUCTION

Crohn’s disease (CD) is a debilitating chronic inflammatory bowel disease (IBD), which if left untreated, leads to penetrating complications (strictures, fistulae, abscesses)[1-3]. Frequent disease activity assessment with the intent of adjusting therapy has been demonstrated to significantly reduce the risk of disease related complications[4-10]. Small bowel involvement in CD is reported to occur in up to 60% of patients, with nearly 30% of patients having isolated small bowel disease[11-13]. Conventional upper and lower endoscopy, however, are limited in their ability to assess small bowel disease activity[14-18], and clinical or biochemical markers of disease activity infrequently correlate with small bowel mucosal inflammation[19]. The presence of inflammation cannot be reliably excluded by radiological imaging alone[20], and video capsule endoscopy (VCE) carries a risk for capsule retention in structuring or penetrating disease and it lacks therapeutic capability[21-24].

Balloon assisted enteroscopy (BAE) offers the advantage of providing both diagnostic (mucosal biopsies) and therapeutic (stricture dilation) potential, but its use in CD is currently limited to symptomatic patients with negative ileocolonoscopy, VCE, and/or cross sectional imaging, and the feasibility and diagnostic utility of routine BAE in clinical practice for CD is yet to be established[20,25-29]. Therefore, we performed a systematic review of the literature to quantify the safety and therapeutic utility of this endoscopic technique. We anticipate these data will help to better characterize the growing importance of BAE in CD.

MATERIALS AND METHODS

Data sources and search strategy

The following databases were searched in October 2015: MEDLINE (PubMed, January 1946 to October 3, 2015); Cochrane Central Register of Controlled Trials (Wiley, 2015); and Embase (Embase.com, January 1974 to October 3, 2015). The search included indexed terms and text words to capture the following concepts: CD and balloon enteroscopy. There were no language or study design restrictions. The search strategy was adjusted for the syntax appropriate for each database. The reference lists of included articles and review articles were examined for additional relevant studies. The full search strategy, a priori, is available at the international prospective register of systematic reviews (PROSPERO #CRD42015016381).

Study selection and extraction

Studies were included for analysis if they met the following inclusion criteria: Randomized controlled trials, cohort studies, published meeting abstracts, or case series of 5 or more consecutive patients with CD, undergoing BAE for diagnostic or therapeutic purposes. Review articles and studies with fewer than 5 patients with CD were excluded. Studies with insufficient data for adverse outcomes and follow-up were excluded only after attempting to contact the primary author(s). The included population was patients of all ages with CD undergoing BAE for diagnostic and/or therapeutic purposes with clearly reported adverse outcomes. Comparative studies to assess diagnostic performance, therapeutic utility, and safety of BAE were included.

Two reviewers (Arulanandan A and Dulai PS) independently evaluated each of the articles for eligibility. Inclusion decisions for each article were made independently based on the eligibility criteria, with disagreements being resolved by a third reviewer (Kalmaz D) and consensus. A reviewer (Arulanandan A contacted the primary author(s) as required to obtain any necessary missing data from the original publications and because no language restrictions were applied, publications were translated into English as required. The reviewers followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses standards for systematic review.

Outcomes

Our primary outcome of interest was the safety of diagnostic BAE [single (SBE) or double (DBE) balloon enteroscopy) in CD, for which we calculated per procedure perforation rate and compared that to per procedure perforation rate when using diagnostic BAE for all indications. The expected rate of perforation when utilizing diagnostic BAE for all indications was derived from a Japanese database of 29068 patients[30].

Secondary outcome of interest was the safety of diagnostic DBE in CD, for which we calculated per procedure perforation rate and compared that to per procedure perforation rate when using diagnostic DBE for all indications. The expected rate of perforation when utilizing diagnostic DBE for all indications was derived from a systematic review of 9047 DBEs[31]. Other outcomes of interest were risk of procedure related perforation of therapeutic BAE, efficacy of stricture dilation, and clinical utility of endoscopically assessing small bowel disease activity.

Statistical analysis

We used the random-effects model described by DerSimonian and Laird to calculate pooled rates (and 95%CI) of perforation with BAE, stratified by indication (diagnostic vs therapeutic)[32]. Due to lack of consistent reporting in multiple studies, we did not perform a quantitative meta-analysis of factors associated with perforation, but rather discussed them qualitatively. The STATA “incidence ratio (IR)” command (version 10.0; STATA, College Station, TX, United States) was used to make comparisons for perforation rates, and the relative rates for perforation were calculated as incidence rate ratios (IRRs).

RESULTS

Patient and study characteristics

Of the 638 studies identified, 73 studies reporting on 1812 CD patients undergoing 2340 BAE procedures (DBE: n = 2027, SBE: n = 187, BAE not-specified: n = 126) were included in the final analysis[28,33-104] (Figure 1, Tables 1 and 2). The majority of studies involved international institutions (85%)[28,33,34,36-40,42-47,51-56,58,61-78,80-93,96,97,99-105], reporting retrospectively (74%)[34,35,37-43,47-51,53-57,59,60,62,63,65,67-69,71-73,75-85,87-91,93-96,99,101,102,104], on their experience with BAE. Of the 1812 patients included in the analysis, 597 (33%) were newly diagnosed with CD and 1215 (67%) had a known diagnosis of CD prior to undergoing BAE. Among the patients with known CD where the indication for the procedure was clearly documented, BAE was performed primarily for monitoring of disease activity (61%), known strictures (16%), or small bowel obstruction (4%). One thousand nine hundred and thirty-eight BAE (83%) were diagnostic and 402 BAE (17%) were therapeutic procedures.

Figure 1.

Studies identified and reasons for exclusion. VCE: Video capsule endoscopy; SE: Spiral enteroscopy; IBD: Inflammatory bowel disease; BAE: Balloon-assisted enteroscopy; CD: Crohn’s disease.

Table 1.

Study demographics n (%)

	BAE studies, n = 73	DBE, n = 60	SBE, n = 11
	Patients, n = 1812	n = 1509	n = 187
Published manuscripts	47 (64)	39 (65)	8 (72)
United States center	11 (15)	9 (15)	0 (0)
Prospective	19 (26)	14 (23)	6 (55)
Published pre-2010	18 (25)	17 (28)	1 (10)
Crohn’s patients per study
5-10	26 (36)	23 (38)	3 (27)
11-25	25 (34)	20 (33)	5 (45)
> 25	22 (30)	17 (28)	3 (27)
Age, mean ± SD	42.9 ± 15.4	42.4 ± 15.5	45.0 ± 15.8

Three studies and 116 patients were solely reported as “balloon assisted enteroscopy”; 1 study included both DBE and SBE. DBE: Double balloon enteroscopy; SBE: Single balloon enteroscopy; BAE: Balloon assisted enteroscopy.

Table 2.

Studies involved international institutions on their experience with balloon assisted enteroscopy

Study	Published manuscript	2010-present	United States	Prospective	Mean age	Patients	Procedures	Therapeutic	Perforations
Akarsu et al[90]	X	X			47.8	39	39	0	0
Aktas et al[95]		X	X		53	58	58	0	0
Aktas et al[92]	X	X		X	51	31	31	2	0
Arihiro et al[96]		X			56.3	32	32	9	1
Bartel et al[41]		X	X		52.6	38	38	0	0
Bartel et al[49]		X	X		54.3	7	7	7	1
Bartel et al[50]		X	X		62.7	15	15	0	0
Chen et al[84]	X	X			51	8	8	0	0
Choi et al[82]	X	X			43.5	7	7	0	0
de Ridder et al[97]	X	X		X	15	14	14	0	0
Despott et al[33]	X			X	46.4	11	13	13	1
Di Caro et al[51]	X				52	7	7	0	0
Di Nardo et al[98]	X	X		X	13	26	26	5	0
Ding et al[85]		X			39	12	22	22	1
Domagk et al[52]	X	X		X	52	11	11	0	0
Dutta et al[99]	X	X			42	14	14	0	0
Fan et al[28]	X	X		X	Not reported	77	308	0	0
Gill et al[34]	X	X			52.7	20	20	10	2
Halloran et al[35]		X	X		44.8	21	40	40	1
Hirai et al[91]	X	X			36	65	110	110	1
Huang et al[53]	X				10	7	7	0	0
Jang et al[42]	X	X			32.7	24	32	0	0
Jeon et al[54]		X			36.4	30	39	0	0
Lakatos et al[55]	X				51.6	6	6	0	0
Li et al[43]	X				49	13	13	0	0
Liu et al[56]	X				8.5	5	5	0	0
Lurix et al[57]		X	X		59	5	5	0	0
Maaser et al[58]	X	X		X	54.9	59	59	0	0
Mann et al[59]		X	X		59	9	9	0	0
Mann et al[60]		X	X		59	23	23	0	0
Manner et al[61]				X	56	20	20	0	0
Manno et al[100]	X	X		X	61	11	11	1	0
May et al[44]	X	X		X	53	9	9	0	0
Mensink et al[36]	X	X		X	53	50	50	0	0
Milewski et al[62]	X			X	45	75	75	12	0
Moreels et al[63]		X			Not reported	6	6	0	0
Morise et al[89]					13	76	76	0	0
Morishima et al[37]	X	X			36	17	35	35	0
Moschler et al[64]					35.4	193	193	0	0
Nakano et al[47]	X	X		X	64	36	36	36	0
Navaneethan et al[94]	X	X			56.8	49	59	9	1
Ohmiya et al[38]	X	X	X		41	23	23	23	0
Parker et al[65]	X				48	11	11	0	0
Pata et al[66]		X			Not reported	16	16	4	0
Peng et al[67]	X	X		X	53	15	15	0	0
Pohl et al[39]	X				Language1	19	21	21	0
Qing et al[68]	X				36	7	7	0	0
Rahman et al[48]		X			Not reported	55	55	3	1
Roushan et al[83]	X	X	X		Not reported	7	7	0	0
Russo et al[101]	X	X			47.2	6	6	0	0
Safatle et al[93]		X			57	9	9	0	0
Schulz et al[69]		X			48.6	11	11	0	0
Seiderer et al[70]	X	X			50.8	10	10	0	0
Shen et al[71]	X			X	33.9	8	8	0	0
Shi et al[72]	X	X			13	35	35	0	0
Sidhu et al[73]	X	X			61.2	39	39	0	0
Sun et al[74]		X			52	7	7	0	0
Takenaka et al[45]	X			X	52	10	10	0	0
Tsujikawa et al[102]	X	X		X	31	17	17	7	0
Uchida et al[75]	X				48.9	6	9	1	0
Urs et al[76]	X	X			12.9	5	5	0	0
Urs et al[77]		X			12.7	7	13	0	0
Watanabe et al[88]	X	X			10.5	10	20	0	0
Watanabe et al[78]		X			Not reported	59	60	0	0
Westerhoff[79]		X			Not reported	18	18	0	0
Wiarda et al[46]			X		Not reported	18	18	0	0
Xu et al[87]	X	X		X	36	21	21	0	0
Yamada et al[40]	X	X			Not reported	46	128	27	0
Yoshida et al[103]		X			37	10	10	5	0
Yu et al[86]		X		X	Not reported	36	108	0	0
Zhang et al[80]	X	X		X	31.6	5	5	0	0
Zhi et al[81]	X	X			Language1	7	7	0	0
Zhu et al[104]	X				36.3	23	23	0	0

¹Language: Unable to translate full manuscript, data was extracted from abstract which did not disclose age.

The 3 prospective randomized controlled trials identified in our search reported on a total of 37 CD patients[44,52,61]. Two of these studies compared DBE to SBE and found DBE to have comparable[52] or superior efficacy and diagnostic yield[44], while the other suggested that fluoroscopy increases insertion depth[61]. The largest prospective cohort study included 193 patients from 62 endoscopic centers in Germany over a 2-year span and had no perforations in CD patients[64].

Safety of BAE

The rate of perforation with diagnostic BAE in CD (1.5 per 1000 procedures) was similar to that reported when utilizing diagnostic BAE for all indications (1.1 per 1000 procedures, IRR = 1.41, 95%CI: 0.28-4.50). The rate of perforation with diagnostic DBE in CD (1.2 per 1000 procedures) was similar to that reported when utilizing diagnostic DBE for all indications (2.2 per 1000 procedures, IRR = 0.54, 95%CI: 0.06-2.24). The rate of perforation with therapeutic BAE in CD was 1.74% (Tables 3 and 4).

Table 3.

Diagnostic and therapeutic procedures n (%)

	Procedures, n = 2340	Perforations, n = 10	Rate (%)
Diagnostic BAE	1938 (83)	3 (30)	0.15
Therapeutic BAE	402 (17)	7 (70)	1.70
Diagnostic DBE	1666 (71)	2 (20)	0.12

BAE: Balloon assisted enteroscopy; DBE: Double balloon enteroscopy.

Table 4.

Balloon assisted enteroscopy cases with procedure related perforation

Study	Demographics	Site and characteristics of perforation	Therapy	Procedure	Outcome after perforation
Despott et al[33]	Long standing CD (> 30 yr) with 5 prior SB resections currently on azathioprine and steroids	3 jejunal strictures (2 inflammatory, 1 fibrotic) with severe ulcerations at stricture sites	Dilated to maximum of 16.5 mm	DBE - Technically difficult due to adhesion-related angulations and fixation, and strictures were significantly ulcerated	Perforation diagnosed within 8 h, patient had laparotomy and temporary jejunostomy. Patient made full recovery and jejunostomy was reversed.
Gill et al[34]	Retained video capsule in patient with known CD	Non-obstructing jejunal stricture with mild inflammation and ulceration at the stricture site	Dilated to 15 mm	DBE - otherwise not specified	Underwent surgery, outcome otherwise not specified.
Gill et al[34]	Known CD patient had previously responded well to dilation up to 15 mm	Distal obstructing ileal stricture with mild inflammation and ulceration at the stricture site	Dilated to 15 mm	DBE - otherwise not specified	Underwent surgery, outcome otherwise not specified.
Halloran et al[35]	Known CD patient who had undergone prior surgical resection.	Scarred bowel loop adhesion site	Not specified	DBE - Perforation occurred with overtube advancement and straightening of a scarred bowel loop	Outcome not specified.
Ding et al[85]	Known CD patient	SB stricture, otherwise not specified	Dilation related perforation, otherwise not specified	DBE - Dilation related perforation, otherwise not specified	Perforation diagnosed within 12 h, patient had laparotomy and resection with ileostomy.
Bartel et al[49]	Retained video capsule in patient with known CD	SB stricture, otherwise not specified	Not specified	DBE - Otherwise not specified	Emergent surgical intervention, otherwise outcome not specified.
Rahman et al[48]	Known CD patient	Ulcer at anastomosis site	Not specified	DBE - Perforation directly related to ulcer at anastomosis	Patient made full recovery after surgical resection and primary reanastomosis.
Navaneethan et al[94]	Known CD patient	Not specified	Not specified	Not specified	Underwent surgery, outcome otherwise not specified
Arihiro et al[96]	Known CD patient	SB stricture, otherwise not specified	Dilation related perforation, otherwise not specified	SBE - Dilation related perforation, otherwise not specified	Patient improved over time without any surgical intervention.
Hirai et al[91]	Known CD patient	SB stricture, otherwise not specified	Dilation related perforation, otherwise not specified	DBE - Dilation related perforation, otherwise not specified	Patient had emergency partial ileal resection and made a full recovery.

CD: Crohn’s disease; SB: Small bowel; DBE: Double balloon enteroscopy; SBE: Single balloon enteroscopy.

Among 1812 patients who underwent a total of 2340 procedures, there were 8 reported perforations with DBE (per procedure 0.39%, 95%CI: 0.12-0.66, per patient 0.53%, 95%CI: 0.16-0.90)[33-35,48,49,85,91,94,96], 1 reported perforation with SBE (per procedure and per patient 0.53%, 95%CI: 0.0-1.57)[96], and 1 perforation with BAE not otherwise specified. Of the 8 DBE perforations, 5 were during dilation of de novo CD strictures and the other 3 were during diagnostic evaluations of an adhesion site, an anastomosis stricture, and a site that was not specified. The 1 SBE perforation was during dilation of a de novo CD stricture site. This equated to an overall per procedure and per patient perforation rate of 0.43% (95%CI: 0.16-0.67) and 0.55% (95%CI: 0.21-0.89), respectively, for BAE in CD including both diagnostic and therapeutic procedures[94]. There was minimal heterogeneity among studies (I² = 0).

Follow-up and impact of BAE on medical and surgical management

Follow-up after BAE was reported in 407 patients with a median follow-up period of 18 mo (range 10-70)[28,33-40,85,86,88,91] (Figure 2). Stricture dilation was performed in 171 patients, resulting in 5 perforations (all at de novo CD stricture sites) and 166 technically successful dilations. Dilation was not attempted in 36 patients due to: inability to insert scope up to stricture site (n = 8), inability to maintain guide wire or through-the-scope balloon at correct position of the stricture (n = 5), stricture length or severe angulation (n = 4), severe inflammation (n = 6), the presence of intra-abdominal adhesions that prohibited advancement of the DBE (n = 2), a perforation during overtube advancement at an anastomotic site (n = 1), and/or non-obstructing or severely ulcerated strictures (n = 10). During follow-up, of the 166 patients where stricture dilation was technically successful, 72 (43%) required repeat stricture dilation and 42 (25%) patients required surgery for persistent symptoms.

Figure 2.

Outcome and impact of balloon-assisted enteroscopy in patients with follow-up.

Of the remaining 200 patients, 173 had clearly reported outcomes regarding changes in medical therapy and treatment response[28,36,86,88,91]. Based on BAE findings, 139 (80%) patients either initiated therapy with an immunomodulator (azathioprine or methotrexate, n = 83), initiated therapy with an anti-tumor necrosis factor agent (n = 52), or they were switched from one anti-tumor necrosis factor agent to another (n = 4). Seven (4%) patients required surgery, and the remaining either refused a step-up in medical therapy (n = 9, 5%) or they had no change in medical therapy needed (n = 18, 10%). Of the 139 patients where a change in medical therapy was performed, 92 (66%) achieved clinical remission at the first follow-up (as defined by the CD activity index)[28,36,86,88]. Repeat BAE was performed in 133 patients with a majority achieving endoscopic improvement in disease activity or complete healing of mucosal lesions (62%)[28,36,86,88]. A third follow-up BAE was performed in 113 patients, and rates of complete or nearly-complete mucosal healing improved from 29% (n = 33) to 43% (n = 57) with changes in therapy based on BAE results[86].

DISCUSSION

In our systematic review of 1812 CD patients undergoing 2340 BAE, the rate of perforation with diagnostic BAE in CD was similar to that seen when utilizing diagnostic BAE for other indications. The rate of perforation with diagnostic DBE in CD was also similar to that seen when utilizing diagnostic DBE for other indications. Additionally, findings from diagnostic BAE in conjunction with changes in medical therapy resulted in improved clinical and endoscopic disease activity in a majority of patients. Diagnostic BAE demonstrated a meaningful change in clinical care with a similar safety profile in CD compared to other indications.

Therapeutic BAE also exhibited significant clinical utility as approximately 70% of patients in our study avoided surgery after stricture dilation, albeit with higher rates of perforation. This increased risk of perforation with stricture dilation in CD[106] must be weighed against the therapeutic benefits achieved. Efforts need to focus on risk stratification of small bowel strictures to determine which patients may be more suitable for small bowel resection as compared to endoscopic dilation. Prior studies have demonstrated that both length and location are useful prognostic factors, and short strictures located in the large bowel or at the site of prior anastomosis are likely to be most amendable to endoscopic dilation as compared to complex or lengthy de-novo strictures[107-109]. Within our study at least three of the six perforations during stricture dilation were at ulcerated sites, and at least two of the perforations were in patients with previous stricture dilations. As the alternative to symptomatic strictures is surgical resection or strictureplasty, our results suggest that stricture dilation via BAE can be done in accessible, short-length symptomatic strictures, and it has better performance in anastomotic strictures, but it may need to be avoided in presence of significant CD-related inflammation.

The relative excess risk of endoscopy associated perforations among IBD patients as compared to non-IBD patients has previously been demonstrated, with disease severity and steroid use (a surrogate for disease activity) being two of the strongest predictors for a procedure related perforation[30,110-112]. Within our systematic review, the total rate of perforation with BAE in CD (4.27 per 1000 procedures) when including both diagnostic and therapeutic procedures was nearly 4 times that reported with diagnostic BAE for all indications (1.1 per 1000 procedures), and the significant majority of this risk was seen in therapeutic procedures.

Our results demonstrate that diagnostic BAE is a safe tool in monitoring small bowel disease activity and may have a role in guiding medical treatment to achieve clinical remission and mucosal healing. The rate of perforation with diagnostic BAE in CD (1.55 per 1000) was also similar to that reported with lower endoscopy in IBD patients (1.89 per 1000)[112]. Other modalities have been demonstrated to have a reasonable diagnostic accuracy for assessing small bowel disease activity in CD[22,113-124], but they have several technical and practical limitations that prevent their routine use in clinical practice[48,113,114,121,125-127]. In our pooled analysis, findings from BAE in conjunction with changes in medical therapy resulted in improved clinical and endoscopic disease activity in a substantial majority of patients. Although further prospective studies are needed to understand the positioning of BAE in disease activity assessment and treating to a target of mucosal healing in CD, these data suggest that BAE may be a useful tool for assessing small bowel disease activity and treatment response in CD.

Our study has highlighted several key findings but it also has several limitations. There is an inherent selection bias in patients who undergo BAE that undoubtedly affects our results. The majority of analyzed studies were retrospective with variable objectives, different inclusion criteria, and limited follow-up data. Additionally, the lack of prospective studies comparing BAE to cross-sectional imaging in evaluation of small bowel disease makes evaluating the utility of diagnostic BAE difficult. Lastly, BAE has a high rate of incomplete enteroscopy[64], which likely was not portrayed in our results, as completion rates were not consistently documented. Strengths of our study included the extensive search performed and the large number of patients and procedures analyzed.

Diagnostic BAE in CD has a similar perforation rate as diagnostic BAE in other indications and may help safely guide medical therapy via assessment of mucosal healing. Therapeutic BAE may help avoid surgery in patients with symptomatic strictures, but the rate of perforation with therapeutic procedures is higher and thus providers must exercise caution when utilizing this endoscopic technique. Further efforts should focus on risk stratification of patients to ensure optimal safety and diagnostic yield and further studies are needed to identify patient and stricture characteristics at highest risk for procedure related complications.

COMMENTS

Background

Crohn’s disease (CD) often affects the small bowel, but conventional upper and lower endoscopy are limited in their evaluation of small bowel disease activity or ability to perform interventions. Balloon assisted enteroscopy (BAE) allows for direct visualization and sampling of the small bowel, however its safety and role in CD remain to be established.

Research frontiers

This systematic review reveals that diagnostic BAE in CD has similar perforation rates as diagnostic BAE for all indications, suggesting that BAE is a safe way to evaluate small bowel CD. Further randomized controlled trials are warranted to confirm these findings as well as to identify anatomical characteristics that are at highest risk for procedure related complications.

Innovations and breakthroughs

The research demonstrates that BAE has a similar perforation rate in diagnostic evaluation of small bowel CD vs other indications. This systematic review is the most up-to-date overview of this subject matter.

Applications

Diagnostic BAE in CD has a similar perforation rate as in other indications, thus it can be safely performed in diagnostic evaluation of small bowel CD.

Peer-review

The present manuscript is well written. The efficacy, safety and long-term prognosis of balloon dilation using BAE should be separately shown between patients with small bowel strictures and patients with anastomotic strictures. This should be analyzed in the manuscript.