Introduction

Stricturing Crohn’s disease (CD) is accompanied by a high rate of surgical bowel resection with a risk of short bowel syndrome. Therefore, strictureplasty (SP) and endoscopic balloon dilatation (EBD) have been applied to avoid, or at least delay, the need for bowel resection. This MEDLINE review gives an overview of the epidemiology as well as the natural and therapeutical course of stricturing CD and about the safety and efficacy of SP and EBD.

Epidemiology of Crohn’s disease

Crohn’s disease is a relapsing, transmural inflammatory disease of the gastrointestinal tract with an incidence of up to 20.2 cases per 100,000 persons per year in Western societies [1, 2]. The European Collaborative Study on Inflammatory Bowel Disease (EC-IBD) reported an incidence of nine cases per 100,000 persons per year in Europe [3]. Similar numbers were observed in the US (7/100,000/year) [4]. Ott et al. [5] recently reported 6.6 newly diagnosed cases per 100,000 persons per year (95% confidence interval: 5.6-7.7) in the region of Oberpfalz, Germany (1,089,000 inhabitants). The prevalence of CD in Germany is estimated to be between 1/500 and 1/800 or approximately 100,000–150,000 patients [6, 7]. Similar numbers have been reported from North America where 400,000-600,000 patients are affected by CD [8].

According to data from the EC-IBD, the highest age-specific incidence is observed between 15 and 34 years of age [3]. This means that CD starts during the educational phase and lasts through the entire work life of most of the patients. Thus, total costs of CD are composed of direct as well as indirect costs (non-productive time, sick payments, etc.). The German Evidence-based Consensus Conference on CD has estimated an annual cost of 3 billion Euros for treating CD patients in Germany [9].

Natural course of Crohn’s disease

CD can affect the entire gastrointestinal tract from the mouth to the anus and the localization as well as the clinical behavior of CD (stricturing, penetrating as well as non-stricturing, non-penetrating type) can change over time. The anatomical localization of CD and its behavior over time were analyzed in 297 patients regularly followed-up at the University Hospital of Liège, Belgium [10]. The authors did a 10-year evaluation (n = 125) of CD localization and CD behavior according to the Vienna Classification System [11]. Affection of the colon and upper gastrointestinal tract were stable over the time while affection of the ileal region decreased and affection of the ileocolonic segment increased during the observation period. After 10 years of CD history, 43.3% suffered from CD in the terminal ileum, 30% from ileocolonic, 23.3% from colonic disease, and 3.3% from upper gastrointestinal affection. CD behavior also changed over time: at the time of diagnosis the majority of patients (73.7%) suffered from non-stricturing, non-penetrating disease. After 10 years, 37.2% suffered from penetrating and 27.1% from stricturing CD. The vast majority of Crohn strictures (98.8%) were found in the colon, ileocolonic region, and ileum after 10 years.

These observations were confirmed by a prospective study from Oslo, Norway, with 197 CD patients completing a 10-year follow-up. A total of 53% of patients had developed stricturing or penetrating CD after 10 years [12]. This means that stricturing CD is relatively uncommon at the time of diagnosis but will become more likely over the years. This phenomenon is probably due to the chronic inflammation in CD, which is accompanied by cell death, consecutive scarring [13], and contraction of the scar.

The need for intervention in stricturing Crohn’s disease

An analysis of 2,002 CD patients from Paris, France showed that 13% of them suffered from stricturing CD behavior at the time of diagnosis [14]. Of these patients, 646 were enrolled in a prospective follow-up study in 1995 with an observation period until 2000. Seventy-nine percent of the patients with stricturing behavior of CD had already undergone surgery prior to their inclusion in the follow-up study. Twenty-two percent underwent (further) surgical intervention. This observation was verified by a population-based study from Oslo, Norway [12]: The authors observed a 64% cumulative rate of surgery in patients with stricturing CD after 10 years, and stricturing CD behavior was an independent risk factor for the need of surgery. These data indicate that stricturing CD leads to the need for intervention in the majority of patients.

The impact of immunosuppressants on the need for intervention

The impact of modern immunosuppressants on the need for surgery in CD was evaluated by a French group in 2005 [15]. This work shows that there was a significant increase in the use of immunosuppressive drugs after 1982 and 1998. The absolute number of intestinal resections and the percentage of CD patients with a need for resection did not change during the same observation period. Although immunosuppressants had been used more frequently over the last 25 years, there was no significant decrease in the need for surgery. Recent evidence seems to support the hypothesis that medical treatment [16] and cessation of smoking [17] after surgery help to avoid a surgical recurrence of CD.

Material and methods

We performed a MEDLINE review of the German and English literature on SP and EBD in CD from 1980 to 2009. The search term was ‘crohn AND strictur*’ in order to increase sensitivity of the search. The initial search yielded 744 articles. Case reports, reviews, and meta-analyses were excluded. Finally, 40 papers dealing with SP and 23 papers dealing with EBD were used for the review.

Treatment options for stricturing Crohn’s disease

Surgical resection

The traditional surgical strategy for structuring disease is resection of the affected segment. In the early 1980s, several groups found that the presence or absence of microscopic disease at the resection margins of CD specimens did not influence the probability of recurrence [18, 19]. This finding was affirmed by a randomized, controlled trial in 1996 [20], where patients with CD were randomized to either ‘limited resection’ (2 cm resection margins) or ‘extended resection’ (12 cm resection margins). The probability of recurrence-free survival did not differ between the two groups. Extensive resection of the terminal ileum, which is the most frequently affected region in CD, may lead to short bowel syndrome, as the remaining intestine is not able to compensate the absorption of vitamins B12, A, D, E, or Kor bile salts. Thus, the surgical strategy in CD changed towards ‘bowel-sparing’ resections. Nevertheless, recurrence rates 15 years after first surgery are reported to be between 28% and 45% [21]. The type of anastomosis (end-to-end vs. side-to-side) also does not seem to influence the probability of recurrence, as recently shown in a randomized, controlled trial (n = 139): After a mean follow-up of 11.9 months, the symptomatic recurrence rate was 21.9% in the end-to-end anastomosis group compared to 22.7% in the side-to-side anastomosis group (p = 0.92) [22].

It was thus reasonable to search for alternatives to bowel resection in stricturing CD. The development of SP, which will be described in detail in the next chapter, provided an opportunity to widen CD strictures without bowel resection.

According to the German Clinical Practice Guideline on Diagnosis and Treatment of Crohn’s Disease [9], bowel resection in stricturing CD is only indicated in case of non-feasibility of SP or EBD. The European evidence-based consensus on the diagnosis and management of Crohn’s disease recommends conventional SP when the length of the stricture is <10 cm [23].

Strictureplasty—technical considerations

The Heineke-Mikulicz (H-M) pyloroplasty, named for Walter Hermann Heineke (1834-1901) and Johann von Mikulicz-Radecki (1850-1905), is the basic principle of traditional SP. It consists of a longitudinal enterotomy at the side of stricture followed by transverse closure with avoidance of bowel resection and anastomoses (Fig. 1). This is important as recurrent resection will potentially lead to short bowel syndrome, and anastomoses bear a higher risk of anastomotic leakage [24].

Fig. 1
figure 1

The Heineke-Mikulicz strictureplasty

Full size image

Originally strictureplasty was applied as pyloroplasty of the stomach, done to enlarge the outlet of the stomach and make it non-functional. In 1977, Katariya et al. [25] published their results of strictureplasties performed in nine patients suffering from multiple tubercular strictures in the lower gastrointestinal tract. In 1982, Lee and Papaioannou [26] published nine cases of H-M strictureplasties for CD of the small intestine. Since then, H-M strictureplasty has evolved as the most frequently performed SP for short-segment stenosis in CD. In a meta-analysis from 2007, 81% (2,499/3,259) of all strictureplasties were H-M procedures [27]. As an alternative surgical strategy, the Judd procedure is a modification of the H-M strictureplasty for fistulas arising from the strictured segment with elliptical excision of the fistula and transverse closure as in H-M strictureplasty. The Moskel-Walske-Neumayer strictureplasty is applied in cases of severe prestenotic small bowel dilatation. It consists of a Y-shaped enterotomy, advancement of the proximal intestinal wall over the stricture and final transverse closure. The Finney procedure, Jaboulay procedure, and the Michelassi side-to-side isoperistaltic strictureplasty’ involve a side-to-side enterostomy and are favored for longer stenoses of up to 25-90 cm.

Strictureplasty—evaluation of safety and efficacy

To evaluate the safety and efficacy of SP in CD, we performed a MEDLINE review on the existing German and English literature. Table 1 gives an overview of the retrieved studies. There were reports of 5,896 strictureplasties in 1,958 patients. We limited the statistical analysis to the calculation of medians and means and their comparison without hypothesis tests, as the heterogeneity of the retrieved studies does not allow any meta-analysis. To assess the safety of SP, we noted the frequency of perioperative complications. The median incidence of perioperative complications was 11% (mean 11.2%). The median incidence of major complications, i.e., anastomotic leakage, intraabdominal abscess, fistulas, sepsis, hemorrhage, ileus, or other events requiring relaparotomy was 5%; the mean major complication rate was 5.7%. To evaluate the efficacy of SP, we noted the frequency of the need for recurrent surgery after SP. We preferentially used cumulative recurrence estimates from survival analyses rather than simple frequencies for our analysis. The median surgical recurrence rate was 24% (mean: 24.0%) after a median follow-up of 46 months (mean, 57.7 months).

Table 1 Literature describing the safety and efficacy of strictureplasty for stricturing CD
Full size table

The major drawback of these studies is that very few authors provided information about site-specific recurrence and that many studies evaluated different techniques of SP.

Endoscopic balloon dilatation

After EBD had become an established treatment option in esophageal, gastric, colonic, and biliary tree strictures [67], R. A. Brower in 1986 was one of the first authors to report about the successful hydrostatic balloon dilatation of a terminal ileal stricture secondary to CD in a 24-year-old woman with refractory symptoms of bowel obstruction that refused surgery [68]. In the following two decades, several institutions reported their results of the endoscopic dilatation of CD strictures. EBD is normally restricted to CD strictures in the colon and terminal ileum. Newer techniques like double-balloon endoscopy are able to reach more orally localized regions of the small bowel but their application is still experimental.

To assess the safety and efficacy of EBD in CD, we performed a MEDLINE literature review and noted the frequency of complications as well as the rate of the need for re-dilatation and surgical recurrence, i.e., the need for surgery after initial EBD. We found 23 papers reporting about 1,003 treatment sessions in 574 patients on an intention-to-treat basis. Results are shown in Table 2.

Table 2 Literature describing the safety and efficacy of endoscopic balloon dilatation
Full size table

Median technical success was 90% (mean, 90.3%). The median rate of major complications like bowel perforation or severe bleeding was 3% (mean, 4%). According to an intention-to-treat protocol, the median surgical recurrence rate was 27.6% (mean, 28.5%). After excluding patients in whom EBD was not possible due to technical reasons (e.g., analysis per protocol), the median surgical recurrence rate was 21.4% (mean, 21.5%) after a median follow-up of 21.0 months (mean, 27.3 months).

The drawback of these studies is that different authors applied different treatment protocols. They used different balloons with different dilatation pressures and different times of insufflation. Some used intralesional steroid injection, while others did not. As in SP, very few authors reported site-specific recurrences. Some authors reported in detail the number of dilatations per patient while others only reported if they were successful or not.

Comparison of results

Table 3 lists reported results of strictureplasty and EBD. These data can only give a rough overview due to the epistemological reasons mentioned above. Meta-analysis and direct comparison of the two methods was not possible due to differing treatment protocols. One can state that there is more literature about SP and that the median/mean length of follow-up is longer for those patients. There is no difference between the incidence of major complications after SP and EBD.

Table 3 Comparison of the safety and efficacy of strictureplasty and EBD in CD
Full size table

A systematic review and meta-analysis of studies describing the safety and efficacy of SP from 2000 (n = 506 patients) onward reported a pooled surgical recurrence rate of 25.5% [92]. In a meta-analysis published in 2007 (n = 1,112 patients), the pooled recurrence rate was 23.4% (95% confidence interval: 16.9-30%) [27]. These numbers are very close to the mean/median recurrence rate in the present review. For EBD in stricturing CD, there is one systematic review published in 2007 [93]. In this review, the long-term success rate of EBD is 58% with a recurrence rate of 42% after a mean follow-up of 33 months.

The need for a trial

Both methods, SP and EBD, have now been applied for more than two decades [26, 68] in parallel and have become well-established treatment options in short-segment stricturing CD [9]. Unfortunately, there is no clinical study or even a multicenter, randomized, and controlled study that directly compares their efficacy and safety. Clinical reports about EBD exclusively derive from internal departments while results of SP are only investigated by surgeons. Internists have proven EBD to be safe and effective while surgeons did so for SP, and both seem to have good reasons for applying their method.

There is no valid information at all as to whether one treatment option is superior to the other with regard to the applicability of both. Assignment to one of the treatment options strongly depends on whether the patient is primarily seen at an internal/endoscopic or surgical department. A multicenter, randomized controlled trial would provide evidence as a basis for clinical decision making.