Integrated endoscopic and medical approaches for Crohn’s disease-associated strictures

1) Technical Aspects and Patient Selection

Most EBD procedures are performed using through-the-scope hydrostatic balloons, typically with diameters of 12–20 mm and dilation durations of approximately 60–120 seconds [13,17-19,26,29-31]. The average stricture length in the reported series was 2–5 cm, and single or sequential incremental inflation was used. Strictures are considered technically amenable if they are: (1) short (<5 cm); (2) non-angulated; (3) without deep ulceration or fistula; and (4) reachable by colonoscopy or balloon enteroscopy [17,19,26]. Double-balloon enteroscopy (DBE) has expanded access to small bowel strictures beyond the reach of conventional endoscopes, with Japanese and European groups demonstrating its feasibility and low complication rates [11,26,28-32,44].

2) Short-Term Efficacy Outcomes of EBD

Short-term outcomes of EBD have been consistently favorable across multiple studies. Technical success rate, defined as the ability to traverse a stricture with an endoscope after dilation, typically ranges between 85% and 95% [10-13,17,19,21-23,28,29,32,45], reflecting the reproducible feasibility of the procedure. Clinical success rate, generally defined as improvement in obstructive symptoms without the need for surgery, varies between 70% and 90% [9,11,15,17,21-25,27], depending on disease phenotype and stricture characteristics. No differences in the technical success rates were observed between de novo and anastomotic strictures [14].

In early European series, 2 studies [10,32] reported technical success rates >90% and short-term symptom improvement in approximately 80% of patients. Subsequent large-scale cohorts [17,22] confirmed comparable short-term outcomes in Asian populations, with clinical success rates of 70%–87%. Notably, a study from Korea [23] demonstrated a 93.3% short-term clinical success rate following technically successful dilation, supporting the high reproducibility of the procedure.

Meta-analytical evidence reinforces these findings, demonstrating the robust short-term efficacy of EBD for CD-associated intestinal strictures. A large-scale meta-analysis [46] pooled data from 25 studies (>1,500 patients) and reported technical and clinical success rates of 89% and 80.8%, respectively, confirming the high initial effectiveness of EBD. Another meta-analysis [47], reviewing 24 studies involving 1,163 patients, found technical and clinical success rates of 90.6% and 70.2%, respectively, and highlighted the influence of stricture length on short-term outcomes. The most recent analysis [48], which included 1,570 patients across 26 studies, reported comparable technical and clinical success rates of 87.6% and 69.7%, further confirming that EBD achieves consistent short-term symptomatic improvements across different populations and endoscopic approaches. A prospective cohort study reported that EBD resulted in short-term technical and clinical success for upper gastrointestinal stenosis associated with CD [21].

3) Long-Term Outcomes of EBD

Long-term outcomes after EBD have been variable, reflecting differences in stricture location, disease phenotype, and follow-up duration. Across representative studies, the overall surgery rate after EBD ranged approximately between 20% and 40% [9,11-17,19,21,26,29-32]. One-year surgery-free survival generally ranges between 63%–80%, decreasing to 45%–60% at 3 years, depending on disease location and number of strictures treated. A retrospective study from Japan investigating DBE-based series observed cumulative surgery-free rates of 63.1%, 59.9%, and 56.2% at 1, 2, and 3 years, respectively, with surgery ultimately required in 48.6% of patients during follow-up [19]. Notably, patients with multiple strictures had a significantly higher risk of surgery than those with a single lesion (3-year surgery-free survival of 33.3% vs. 88.9%). Similarly, a retrospective study focused on small intestinal strictures reported surgery-free rates of 79% at 2 years and 73% at 3 years [31], whereas another retrospective study including a mixed small- and large-bowel cohort, found a 3-year surgery-free rate of 76.7% [17]. In contrast, a retrospective study from Italy demonstrated a progressive decline in symptom-free survival over time (1 year 76%, 2 years 55%, 3 years 46%), underscoring the relapsing nature of fibrostenotic disease [10]. A retrospective study from Japan reported that anastomotic strictures had significantly superior surgery-free survival than de novo strictures (log-rank P<0.05) [12].

Re-dilation is required in approximately 30%–50% of patients within 3 years [10-13,15,17,21,23,24,26,32]. In a Japanese series [31], the re-dilation-free rate decreased to 47% at 3 years, while another study from Japan reported a re-dilation rate of 46.7% during long-term observation [17]. These consistent results across cohorts demonstrate that, although EBD provides durable symptom relief and postpones surgery in many patients, repeat endoscopic intervention is often necessary because of disease recurrence or progressive fibrosis.

Overall, available studies have not demonstrated a consistent difference in long-term outcomes between de novo and anastomotic strictures treated with EBD. Several retrospective studies reported no significant differences in technical success, clinical success, or subsequent surgery rates between these 2 stricture types [11,14]. In contrast, a retrospective study from Japan suggested that anastomotic strictures may achieve superior surgery-free survival compared with de novo strictures [12]. Furthermore, a retrospective study conducted in the United States suggested divergent outcomes depending on the endpoint evaluated; for example, anastomotic strictures were reported to require repeat dilation more frequently than de novo strictures, despite comparable short-term technical success [18]. However, these findings have not been consistently reproduced across different cohorts. Such discrepancies are likely attributable to heterogeneity in study design, patient populations, treatment era, and stricture characteristics, rather than to true biological differences between de novo and anastomotic strictures.

4) Concomitant Medical Therapy and Outcomes after EBD

The impact of concomitant medical therapy on long-term outcomes after EBD remains inconclusive once anatomic and inflammatory factors are taken into account.

Regarding biologics, controversial results in long-term outcomes after EBD have been reported. Early single-center retrospective studies, mainly from Japan, suggested potential benefits of concomitant biologic therapy. Concomitant infliximab induction with EBD was associated with higher clinical efficacy and surgery-free survival compared with infliximab alone [49], and prior or ongoing thiopurine use independently predicted better post-EBD outcomes [50]. However, these findings were derived from relatively small cohorts. Another study with a relatively large number of participants also reported that biologic maintenance after EBD was associated with a lower surgery risk than conventional therapy overall, with the clearest advantage observed in ulcerated (inflammatory) strictures [26]. In contrast, larger multicenter studies and registry-based analyses have not consistently demonstrated a clinically meaningful protective effect of concomitant biologic therapy. For example, data from contemporary European and U.S. cohorts did not show improved long-term outcomes with the use of biologics or immunomodulators at the time of dilation [20,24]. In the ENEIDA registry [24], the absence of anti–tumor necrosis factor therapy at the time of dilation was independently and significantly associated with higher therapeutic success; however, this finding was interpreted as reflecting confounding by indication rather than a true detrimental effect of biologic therapy. Similarly, a Japanese single-center cohort study suggested that anti-tumor necrosis factor therapy alone did not fully mitigate the risk of post-EBD events, highlighting the greater importance of residual inflammatory activity and stricture characteristics [41].

Evidence regarding adjunctive local steroid therapy, such as intralesional steroid injection, following EBD also remains controversial. In a pediatric randomized, double-blind trial, intralesional triamcinolone injection after EBD significantly reduced both re-dilation and the need for surgery for more than 12 months [51], whereas an adult randomized trial found no benefit and suggested a trend toward earlier re-dilation with intralesional triamcinolone compared with placebo after EBD for ileocolonic anastomotic strictures [52]. Despite their randomized design, these trials differed substantially in patient populations, stricture characteristics, and clinical contexts, and therefore their results should be interpreted with caution. With respect to systemic immunomodulators, a postoperative cohort from Portugal reported that thiopurine therapy was associated with a longer time to first dilation, indicating a potential protective effect. On further multivariable modeling, thiopurine treatment remained the only medication independently associated with a delayed need for EBD; however, anastomotic strictures still required more frequent repeat dilations than de novo strictures [18].

Taken together, current evidence does not support the uniform benefit of adjunctive medical therapy following EBD. The observed variability across studies is best explained by heterogeneity in study design, patient populations, disease phenotype, and procedural factors, underscoring the need for phenotype-stratified prospective studies to determine whether concomitant medical therapy confers benefit beyond complete and safe dilation.

1) Procedure and Indications for ES

The ES procedure is illustrated in Fig. 1 [60]. First, a radial incision is made, followed by a subsequent horizontal cut. Incisional dilation is performed by repeating this procedure. The devices and settings of the electrosurgical unit are the same as those used for endoscopic submucosal dissection (ESD) of colorectal tumors [57-60]. The barrier to introducing ES is expected to decrease in facilities that perform ESD as part of their daily clinical practice. The main device differs depending on the report, and the name of the dilation technique, such as NKSt or RIC, also differs. Several studies have selected an electrosurgical endo-knife for ESD with a small insulated tip (ITknife nano; Olympus, Tokyo, Japan) [57-59], whereas others have used only a needle-shaped electrosurgical knife [53,60]. Furthermore, both types of electrosurgical knives were used in several studies [54,55,61,62]. Although there are slight differences in the devices used in NKSt and RIC, their procedures are similar. Therefore, as previously described, we combined them and referred to them as ES.

Fig. 1.

Procedure for endoscopic stricturotomy (ES) in a male in his 40s with Crohn’s disease having small intestinal stenosis due to mucosal healing of Crohn’s disease. (A) Primary stenosis in the small intestine. First radial incision (yellow arrow). (B) Just after first radial incision. (C) Subsequent horizontal cut (white arrow) after radial incision. Regarding the depth of incision, several incisions should be performed with caution not to incise too deep. (D) Completion of ES. We repeated several radial incisions and horizontal cuts and completed ES. An enteroscope could pass through the dilation site. When executing the horizontal cut, slightly angling the scope toward the muscle layer can facilitate smoother manipulation. (E) Fluoroscopy before ES. Visible stenosis (orange arrow). (F) Fluoroscopy just after the completion of ES showing adequate dilation and no leakage (orange arrow). Reproduced from Moroi R, et al. Intest Res 2025;23:302-308 [60].

Although the indications for ES differ slightly depending on the study, the major indications are generally: (1) benign stenosis; (2) stenosis length <2 cm; (3) no abscess or fistula near the stenosis; and (4) fibrotic stenosis is preferable to edematous stenosis [57-60]. Stenosis ulceration does not prevent ES because the ES technique can avoid ulceration by incising only non-ulcerated areas [59,60]. Furthermore, ES location does not matter as long as the scope can reach the stenosis. ES can be performed even in small bowel strictures using balloon-assisted enteroscopy. Two studies reported ES for small bowel stenosis using balloon-assisted enteroscopy [60,61].

2) Short-Term Outcomes of ES

Most studies demonstrate technical success rates of approximately 100%, reflecting the feasibility of this novel dilation method and provide high clinical success rates, indicating its high efficacy as a dilation method [53-62].

In the earliest single-center series, NKSt achieved 100% technical success and 54.3% symptomatic improvement despite most lesions being refractory to prior balloon dilation [53]. A subsequent comparative cohort reported 100% technical success, again highlighting good early efficacy, even in complex anastomotic strictures. More recent results from a single-institutional setting reinforce these findings. We reported 100% technical success and 100% short-term symptomatic improvement in 5 patients with CDAS [57]. In our later multicenter study, technical success remained at 96.3%, and significant early improvements in abdominal pain, bloating, defecatory difficulty, and other obstructive symptoms were observed [58].

ES via balloon-assisted enteroscopy produced similar results when applied to deep small bowel strictures. We conducted a prospective study and demonstrated 100% technical success using a long electrosurgical knife, with universal early improvement in obstructive symptom scores [60]. Similarly, another retrospective study from China reported high technical feasibility and marked symptomatic improvement following ES during balloon-assisted enteroscopy for the small bowel, further extending the applicability of the technique beyond the ileocolonic anastomosis [61]. Collectively, these data show that ES achieves 95%–100% technical success and 50%–100% short-term clinical success, typically within weeks of treatment, even in cases refractory to EBD.

3) Long-Term Outcomes of ES

Across the available cohorts, long-term surgery-free rates typically range between 70% and 90%, whereas the need for repeat endoscopic treatment remains substantial, often occurring in 20%–60% of patients, depending on stricture type, location, disease activity, and follow-up duration [53-61].

An early detailed analysis reported that after ES, 15.3% of patients ultimately required surgery over a median follow-up period of 0.9 years, while 60.6% underwent additional endoscopic therapy [53]. A retrospective analysis found comparable durability, with surgery required in 9.5%, and a re-intervention rate of 57.1% over a median follow-up period of 0.8 years [54].

Recent multicenter, enteroscopy-based studies have extended these observations to more complex anatomical settings. In a multicenter Japanese cohort, we demonstrated preferable short-term results, but a substantial long-term recurrence burden, with repeat endoscopic therapy required in over 50% of patients during extended follow-up [58]. Similarly, in deep small-bowel Crohn’s strictures treated using balloon-assisted enteroscopy, we reported high early technical success, although long-term durability data indicated that approximately 20% of patients eventually required another endoscopic intervention [60].

4) Effect of Concomitant Biological Agents and Steroids on Long-Term Outcomes after ES

Evidence regarding whether concomitant medical therapy modifies long-term outcomes after ES remains limited. A retrospective study reported that background anti-tumor necrosis factor or immunomodulator use was common, but not associated with differences in repeat intervention or surgery [53]. Similar findings have been observed in comparative studies of anastomotic and distal ileal strictures [55,56]. Our study showed that local triamcinolone injections did not demonstrate reduced recurrence [60]. Collectively, current evidence does not show clear beneficial or harmful effects of biologics or immunomodulators on long-term ES outcomes. Steroid exposure (systemic or local) may identify high-risk patients or impair healing; however, its causal effects remain unproven. Given the retrospective design and small sample sizes of existing studies, prospective, adequately powered analyses are needed to clarify drug–procedure interactions.

5) Safety of ES

Major adverse events of ES include perforation during the procedure and delayed bleeding. In a published series, delayed bleeding was the most frequent adverse event, whereas perforation was rare. The delayed bleeding rate ranged from 0% to 23.8%, varying widely among studies, while perforation rates consistently remained low (0%–3%) [53-63]. Although reported postprocedural bleeding rates vary widely across studies, anastomotic strictures, particularly ileocolonic anastomoses in CD, appear to have a higher propensity for delayed hemorrhage than other stricture types [57-59]. This variability is likely influenced by differences in the definition of bleeding, procedural techniques, and heterogeneity in patient characteristics, including stricture location, inflammatory activity, and postoperative anatomy. In contrast, even in small bowel lesions treated with balloon-assisted enteroscopy, both ES showed very low bleeding rates and no perforations [60,61].

6) Comparisons between ES and Other Interventions

Several studies have compared ES with other interventions, including EBD and surgery. A retrospective analysis [54] reported significantly higher procedural success for ES compared with EBD (100.0% vs. 89.5%) and higher early symptomatic improvement (72.7% vs. 45.4%). Importantly, the need for subsequent surgery was substantially lower after ES (9.5% vs. 33.5%; P=0.03) despite similar requirements for repeat endoscopic intervention (57.1% vs. 59.8%). Adverse events differed in character rather than frequency; perforation occurred only after EBD (1.1%), whereas ES was associated with a higher rate of delayed bleeding requiring transfusion (8.8% vs. 0.0%), a complication intrinsic to incision-based therapy but manageable in experienced hands.

Another retrospective study compared ES with surgery (ileocecal resection [ICR]) and demonstrated that ES achieved 97.1% technical success with 58.3% symptom improvement, whereas long-term surgery-free survival did not differ significantly from ICR (11.3% vs. 10.2%; P=0.83). Notably, major adverse events were markedly less common in ES (10.2% vs. 31.9%), highlighting a substantial morbidity advantage [55]. Similarly, in a retrospective cohort study regarding primary distal ileal strictures, both ES and ICR were reported to achieve 100% technical success, yet ES resulted in significantly fewer complications (6.9% vs. 25.0%; P=0.05) and comparable long-term surgery rates (15.4% vs. 18.8%; P=0.79). Although symptomatic improvement was numerically higher in ICR (90% vs. 50%), the equivalence in surgery-free survival underscores the mechanical durability of ES, even when symptom resolution is incomplete [56].

Direct comparison with EBD in a randomized framework demonstrated 100% technical success with ES and significantly higher short-term clinical response (88% vs. 50%) compared with EBD. ES also reduced the need for subsequent procedures or surgery over an extended follow-up, while maintaining a low and clinically manageable adverse event rate [62].

Collectively, these comparative data suggest that ES may offer superior short-term technical and clinical effectiveness compared with EBD, particularly in anatomically complex or fibrotic strictures, and might provide long-term outcomes that are approximately similar to those of ICR while potentially reducing procedure-related morbidity. However, these observations should be interpreted with caution, as most available studies are retrospective, involve limited sample sizes, and exhibit heterogeneity in patient selection, procedural techniques, and follow-up duration.

1) Evidence of RCTs of SEMS for CDAS

Two randomized controlled trials (RCTs) provided head-to-head comparisons between SEMS and EBD. The ProtDilat study, an open-label, multicenter RCT, randomized patients with symptomatic Crohn’s strictures to undergo repeated EBD or removable SEMS placement. Global therapeutic success at 1 year, typically defined as absence of surgery or need for additional endoscopic intervention, was significantly higher in the EBD arm (80.5%) than in the SEMS arm (51.3%), whereas technical success and serious adverse events, including perforation (2.5% in each group), were comparable. Cost analyses within the same program further favored EBD over SEMS as the initial endoscopic strategy [65]. In contrast, a smaller single-center RCT from Sweden reported 100% technical success in both arms, but adverse events occurred in 57% of SEMS-treated patients versus none after EBD, prompting early study termination [66]. Despite this, clinical success, defined as freedom from repeat endoscopic or surgical intervention, was numerically higher with SEMS (86%, 6/7) than with balloon dilation (20%, 1/5) [66]. Because the 2 existing randomized trials provided conflicting results, further large-scale studies are needed to clarify the optimal role of SEMS in CD strictures. The ongoing ENDOCIR trial is expected to address these gaps, and its findings will be crucial for guiding future clinical practice [65].

2) Evidence of Non-RCTs on SEMS for CDAS

Non-randomized evidence suggests that short-term, removable metallic self-expandable stents represent the most promising stent-based approach for CDAS. Most published studies have evaluated fully covered metallic SEMS designed to allow temporary placement and subsequent removal, aiming to provide sustained luminal patency while minimizing tissue ingrowth and stent-related complications. In a multicenter cohort, SEMS placement followed by scheduled removal after 7 days achieved high short-term clinical success (93.5%) and acceptable long-term success (58.7%), with a remarkably low migration rate (6.5%) and no perforations, highlighting its favorable safety profile in real-world practice [67]. Comparable outcomes were reported in a single-center U.K. case series, in which 81% of patients experienced symptomatic improvement, and none required stricture-related surgery during up to 50 months of follow-up [68]. Traditional fully covered or uncovered SEMS maintained for prolonged periods have similarly demonstrated important limitations, including difficult stent extraction, impaction, and an increased need for rescue surgery, despite modest clinical success in selected cases [69-71].

3) Biodegradable Stents for CDAS

In contrast to metallic SEMS, biodegradable stents are composed of polymer-based materials and are designed to gradually degrade over time, thereby eliminating the need for endoscopic removal. A limited number of small case series have reported the use of biodegradable stents for CDAS, including 2 representative studies [72,73]. These reports suggested that biodegradable stents may provide temporary luminal patency without the requirement for stent retrieval; however, their clinical efficacy was inconsistent, and stent-related adverse events such as early collapse, migration, and recurrent obstruction were frequently observed. Given their distinct material properties, radial force profiles, and degradation behavior, biodegradable stents should be considered a conceptually separate modality from metallic SEMS, and their role in the management of CDAS remains investigational.

4) Stent Insertion and Stricture Length

Most EBD series limit treatment to strictures <5 cm because of the increased perforation risk and reduced long-term efficacy, whereas ES is generally reserved for shorter focal (1–2 cm) fibrotic lesions. In contrast, several stent studies included strictures between 5–8 cm in size, particularly in partially or fully covered SEMS cohorts. For example, a retrospective case series reported successful SEMS placement in strictures up to 8 cm in length [71]. This prospective pilot study included strictures of up to 50 mm on cross-sectional imaging [70]. A biodegradable stent series has also been used to treat strictures beyond the range typically managed with EBD, including several anastomotic lesions between 4–6 cm [72,73]. Similarly, a U.K. single-center experience with short-term removable SEMS involved strictures of ≤ 6 cm [68]. These findings suggest that stent therapy may expand endoscopic treatment options for patients with longer fibrotic or anastomotic strictures that fall outside the conventional size limits for EBD or ES. Although safety and efficacy vary among stent types, these data suggest that stenting could serve as a minimally invasive alternative to surgery, specifically for intermediate-length strictures, a subgroup in which endoscopic options have been traditionally limited.