Sergey Kashin, MD
Endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD) have become safe procedures with a low rate of short-term adverse events for a curative and radical treatment of gastrointestinal (GI) superficial neoplasias not only in Japan but in Western countries as well.1 Both procedures have a low rate of short-term adverse events as a result of technical improvements and various endoscopic options to manage intraoperative events like bleeding or perforation. In contrast to the short-term safety of ESD and EMR, the formation of a stricture is a common and severe delayed adverse event after a widely extended resection.2 Strictures after EMR and ESD rarely can be seen in the cardia, prepyloric region of the stomach, and in colon or rectum, but it is a common occurrence after an esophageal ESD. There is a clear association between the extent of the resected area and stricture occurrence, with a stricture rate approaching 90% to 100% with circumferential esophageal ESD.3 Extensive endoscopic resections for the treatment of early esophageal neoplasia can result in fibroinflammatory strictures that require repeated interventions, which significantly alter the patients’ quality of life. Owing to a relatively high frequency of this complication, a number of strategies aimed at preventing and/or treating post-ESD esophageal stricture have been introduced, including prophylactic serial dilation, intralesional steroid injection or topical steroid gel application, and prophylactic placement of fully covered, self-expandable metal stents (SEMS). Endoscopic transplantation of tissue-engineered autologous oral mucosal epithelial cell sheets and resected gastric mucosa has also been reported for prophylaxis of post-ESD esophageal stricture, but it remains experimental.4 Despite progress in the prevention of stricture formation, the effect of various measurements is still limited.
Prevention and Management of Strictures After Endoscopic Mucosal Resection/Endoscopic Submucosal Dissection for Early Esophageal Neoplasia
Post-ESD esophageal stricture is generally defined as a narrowing of the esophageal lumen seen on esophagoscopy, which is either nonpassable with a standard diagnostic (± 10 mm in diameter) endoscope or is associated with dysphagia.5 Strictures develop in 5% to 17% of patients after esophageal ESD with risk factors including the circumference and length of the resection.6–7 The possibility of esophageal stricture increases if a circumferential mucosal defect is more than three-fourths of the esophageal circumference post-ESD (ie, if the lesion size is greater than 60%). A Japanese study reported that postoperative strictures occurred in 90% of patients with lesions of more than three-fourths the circumferential extension; these rates approach 100% after complete resection of the circumference.8,9 Therefore, a second-look endoscopy is performed routinely at 2 or 3 weeks after ESD because esophageal strictures may be evident within a few weeks after the procedure during the mucosal healing process.10 The length of longitudinal mucosal defects, location of the tumor, and depth of neoplastic invasion (> pT1m2) have also been associated with the occurrence of strictures.11 A longitudinal mucosal defect more than 30 mm in size is a significant risk factor for stricture development. Lesions in the upper esophagus are reportedly complicated with post-ESD stricture more frequently than those in the lower esophagus (50% vs 11%, respectively), because the luminal diameter of the upper esophagus is smaller than that of the lower esophagus.12
The patients who have multiple risk factors or those whose mucosal defect after ESD is more than three-fourths of the circumference of the esophagus are at the highest risk of stricture development. For this reason, the Japan Esophageal Society in the 2007 edition of Guidelines for Diagnosis and Treatment of Carcinoma of the Esophagus gave an absolute indication for endoscopic resection: The lesion should not exceed two-thirds of the circumference.13
If the resected lesion is larger than 75% of the esophageal circumference, it may be necessary to perform endoscopic balloon dilations (EBDs) to prevent postprocedural stenosis. For prevention of esophageal strictures after ESD, EBD started in the early postoperative days was effective, but the burden on the patients compromises their quality of life.14 However, in recent years it has been reported that the occurrence of stenosis after ESD and the frequency of required EBD sessions can be substantially decreased by local injection or oral administration of steroids, and such prophylactic methods have become widespread in clinical practice.15 As a result, the limitation on the circumference of a lesion was deleted from the Japanese guidelines in the current version with the following comment: “Mucosal resection covering 3/4 of the entire circumference is likely to be associated with postoperative cicatricial stenosis. Therefore, sufficient explanations should be given to the patient prior to the operation and preventive measures must be taken. The need for prevention, prophylactic measures, and treatment of complications including cicatricial stenosis should be well recognized.”16 Thus, it is now becoming possible to treat any circumferential lesions with ESD and to control post-ESD stenosis.
Mechanisms of Postendoscopic Submucosal Dissection Stricture
In general, wound healing and stricture formation after ESD in the esophagus and stomach occur by way of the following series of events described recently by Barret et al.17 The first step is an epithelial injury, leading to a loss of the barrier function of the epithelium and the exposure of the submucosal space to several kinds of insults: mechanical through food boluses, chemical through acid or bile reflux, and microbial due to the esophageal bacterial and fungal flora.18,19 The second step is immune system activation, induction of inflammatory response, and organization of granulation tissue.20,21 A cascade of inflammatory cytokines and reactive oxygen species is activated immediately after injury, and inflammation continues until the process is complete. The third step is proliferation of epithelial cells in the border of the wound, the covering of the wound with granulation, and the growth of subepithelial granulation.22 The inflammatory response is characterized by the formation of granulation tissue that includes an inflammatory cell infiltrate and palissade vessel neoformation (in the distal part of the esophagus). The fourth step is scar formation, which after extensive mucosal injuries of the esophagus results in esophageal stricture development. Resident fibroblasts and pericytes are induced to differentiate into scar-forming myofibroblasts by fibrogenic cytokines.17 Cytokines such as transforming growth factor-beta 1, tumor necrosis factor-alpha, interleukin (IL)-1, IL-6, platelet-derived growth factor, IL-17A, and IL-13 are produced by epithelial cells, endothelial cells, and myeloid leukocytes. First, myofibroblasts are highly contractile and express α-smooth muscle actin, which forms bundles of myofilaments that promote wound contraction. Second, myofibroblasts are the key drivers of scar tissue formation through synthesis of collagen Type I. Monocyte-derived macrophages contribute to the process either as classical inflammatory cells or by stimulating myofibroblast differentiation via IL-10 production.23,24 The following 3 pathological characteristics that are unique to the process of stricture formation after esophageal ESD were demonstrated: delayed mucosal healing, severe inflammation, which results in deeper ulceration and extensive fibrosis in the submucosa, and atrophy in the muscularis propria.25 These 3 factors may be induced by deep thermal injury because of the electric current effect of endoscopic therapy or the prolonged loss of the esophageal epithelium as a barrier against the external environment, including exogenous material, saliva, food, microorganisms, or refluxate.
Strategies for Prevention and Treatment of Postendoscopic Submucosal Dissection Strictures
Strategies for prevention and treatment of post-ESD strictures consider the application of different modalities for esophageal stricture prevention according to main mechanisms of esophageal stricture formation after wide resection of the circumference (Figure 26-1). Preventing esophageal stricture after mucosal injury requires reaching one or several of the following goals17: (1) suppress the initial tissue aggression factor, (2) restore the destroyed protective barrier, (3) block inflammatory pathways, and (4) oppose the action of myofibroblasts. After endoscopic resection, the only way to reach the first goal is to ensure an optimal acid suppression by high-dose proton pump inhibitor therapy during the wound-healing process. The other 3 goals can be achieved through 4 different approaches, which can be classified as follows: a protective approach, a regenerative approach, an antiproliferative approach, and a mechanical approach. Preventive technologies will be described according to these approaches and sequence of events after mucosal injury causing stricture formation (see Figure 26-1).
At the present time the most widely used technologies refer to mechanical and antiproliferative approaches; meanwhile, the new technologies for protection and regeneration of post-ESD ulcer are still under investigation. Mechanical approaches include EBD, bougienage, and various stenting procedures; antiproliferative approaches include local injections and systemic steroid therapy, preventing excessive fibrotic formation drugs such as mitomycin C (MMC), 5-fluorouracil (5-FU), and botulinum toxin-type A (BTX-A). Tissue engineering approaches (regenerative therapy) include scaffold-based therapy and cell-based therapy.
Endoscopic Balloon Dilation
Many studies have addressed the option of preventive EBD in the esophagus after EMR or ESD. A median number of 4 sessions and as many as 19 procedures were required for the management of strictures after stepwise radical EMR of Barrett’s esophagus (BE) neoplasia.26 These interventions are inconvenient for patients and may cause perforation. As for EBD in cases of strictures after ESD performed for squamous cell cancers, nearly all of the clinical trials were performed in Japan, China, or South Korea. Generally, EBD using a controlled radial expansion balloon dilator is started 3 days post-ESD and performed once or twice weekly for 8 weeks. If dysphagia persists after 8 weeks, dilation on time is continued until the symptoms subside. In each session, the EBD diameter is initially set at 13 to 15 mm. Greater diameters may be required when patients remain symptomatic, and the lumen is dilated to 18 mm.12 The frequency of balloon dilation is usually proportionate to the degree of esophageal mucosal defect and the degree of stenosis. Ezoe et al27 reported the results of EBD after EMR and ESD of 41 patients with mucosal defects accounting for more than three-fourths of the esophageal lumen perimeter. Twenty-nine patients underwent EBD within 1 week at a frequency of once a week until the mucosal defects completely resolved. Twelve previous patients were chosen as the historical blank control group in which routine EBD was conducted when they developed esophageal stenosis until stenosis was corrected. As a result, preventive endoscopic dilation reduced the incidence and severity of stenosis as well as patients’ tolerance to stenosis.
Yamaguchi and colleagues28 reported that patients underwent balloon dilation a mean of 16 times and that one patient with circumferential lesions underwent dilation up to 48 times. Studies of preventive endoscopic dilation showed disappointing results in terms of the large number of required procedures, worsening quality of life for patients, and a risk of dilation-associated adverse events.
The major complications of endoscopic dilation treatment include perforation, hemorrhage, and bacteremia. The rates of perforation during EBD have been reportedly as high as 1.1% to 9.0% in the case of post-ESD strictures,7,29 which is higher (0.3%) than that in the case of strictures secondary to other causes.30 It is generally accepted that the risk of perforation can be significantly reduced if the dilation diameter is increased by ≤ 3 mm each time. The diameter and length of esophageal stenosis before dilation are key factors that affect the required dilation efficiency and frequency (Figure 26-2).
Savary-Gilliard bougienage (Cook Medical) is a cheaper mechanical approach for the dilation of strictures after EMR or ESD because bougies can be reused. Small to large bougies are selected and used to dilate the stenosis step by step to an appropriate extent. In some studies, no significant difference in efficiency was found between these 2 dilation approaches.31,32 But in case of using a through-the-scope balloon the dilation is performed under direct view, which helps to control all the steps of the procedure.
Thus, EBD and Savary-Gilliard bougienage are the most common endoscopic dilation approaches used for esophageal benign stenosis in clinical practice. But EBD strategy frequently requires more than 10 sessions, so the safety profile of preventive EBD does not seem to be satisfying. Therefore, there is a strong demand for stricture prevention using anti-inflammatory drugs or other approaches, particularly in high-risk candidates with extended resection of early esophageal neoplasia.
Temporary stent placement may also be a promising strategy for the prevention of post-ESD strictures. SEMS or biodegradable stents have been used for the treatment of benign strictures such as anastomotic stricture and cicatricial stricture secondary to esophagitis,33 although previous studies reported the usefulness of SEMS for the treatment of benign esophageal strictures with potential late-onset complications that are related to long-term stent placement.34 However, all types of stents cannot be used in patients with stenosis because of the characteristics of esophageal stenosis after early neoplasia resection.
Self-Expandable Metallic Stents
The major advantage of these stents in the treatment of benign esophageal stenosis is they can provide a sustained dilation effect to the stenosed segment and can be removed when the stenosis is relieved or when complications occur. Several recent studies have reported the use of temporary SEMS to treat esophageal benign stenosis, which shows that stent implantation is effective to some extent in the treatment of stenosis and can alleviate the symptoms in some patients. Nevertheless, some studies have found that the long-term effect of temporary SEMS after implantation was not as satisfactory as expected and that the incidence of complications such as granulation tissue hyperplasia, chest pain, and stent displacement was relatively high.35 As for the treatment of stenosis after ESD, Holt and colleagues36 assessed in 12 patients fully covered metal stents placed 10 days after a circumferential EMR for Barrett’s neoplasia. In 4 patients, stents had to be removed for pain or dysphagia, and 57.5% of the patients ultimately developed strictures. Bhat et al,37 in the same clinical setting, reported a 28% stricture rate after esophageal stent removal or migration. It seems that the more classical design of the stents and shorter duration of the stent placement in the last study may be the main differences between these 2 works. In a study that included 23 patients with high stricture risk after esophageal endoscopic resections, 19 of which were circumferential, treated by prophylactic stents, the authors reported a 52% early stent migration rate and a 39.2% rate of esophageal stricture.38 Wen and colleagues39 found that covered esophageal stent placement for the prevention of esophageal strictures after ESD is effective and safe. In their random control test, the fully covered esophageal stent was placed immediately after ESD at the ulcer site then removed 8 weeks later. They concluded that the proportion of patients who developed a stricture was significantly lower in the stent group than in the control group. Moreover, the number of bougie dilation procedures was significantly lower in the stent group than in the control group.
Biodegradable stents have been applied for the prevention of post-EMR/ESD strictures because of the various drawbacks of metallic and plastic stents. Tanaka et al40 were the first to use polylactide biodegradable stents in 2 patients with esophageal stenosis and obtained promising results. Similarly, in the study by Saito and colleagues,41 the stents were used in 2 patients who developed stenosis after ESD for early esophageal cancer. The mucosal defects accounted for more than three-fourths of the esophageal perimeter in both patients. Polylactide biodegradable stents were implanted after balloon dilation when the stenosis occurred, and no adverse effects or recurrence was observed in the 6 months after implantation. However, the usefulness of these studies is limited because of the high frequency of biodegradable stent migration, small number of patients, and short-term follow-up periods. Thus, further studies are needed to assess the efficacy of biodegradable stents in prevention of esophageal stenosis after ESD.
Extracellular matrix (ECM) stents are a promising combination of mechanical and tissue engineering approaches for the treatment of strictures after EMR or ESD in the esophagus. Badylak et al42 reported that an ECM biologic scaffold composed of porcine-derived small intestinal submucosa, together with a temporary metallic stent, is safe and efficacious for the prevention of esophageal strictures after endoscopic resection in dog models. Another dog model was established in 2009 by Nieponice and colleagues43 in which ECM stents were used to prevent esophageal stenosis after circumferential EMR. In that study, ECM stents were implanted through endoscopy in 5 dogs after EMR using another 5 dogs as controls. As a result, none of the dogs in the treatment group presented with esophageal stenosis and no significant cicatrices or inflammation were observed in the pathological specimens. In contrast, esophageal stenosis occurred in all 5 dogs in the control group and epithelialization and incomplete inflammation were observed at the EMR site. Biological scaffolds have been safely used by the same group of authors to treat 5 patients with BE and high-grade dysplasia undergoing long length (more than 8 cm) circumferential endoscopic resections.44 The esophageal strictures after endoscopic resection were improved by only a few sessions (0 to 9) of endoscopic dilation even though the temporary stent support prevented strictures. Surprisingly, the small perforation site healed in 18 days by covering the perforation with a biological scaffold and stent. The scaffold provides an ECM, supports strictures, and promotes cell migration. However, it may be insufficient to cover an extensive mucosal defect after esophageal ESD.45
To control inflammation and myofibroblastic proliferation, anti-inflammatory drugs, antimitotic drugs, and so-called antifibrotic drugs have been tested.
Anti-Inflammatory Drugs (Corticosteroids)
Most studies have focused on steroids, either locally delivered to the esophageal wall or administered by the systemic route. The aim of local or systemic administration of steroids is to reduce injury-induced inflammation and to decrease the formation of granulation tissue. Continued treatment with steroids may inhibit collagen synthesis and cross-linking. A meta-analysis of steroid administration for the prevention of post-ESD esophageal strictures published recently by Wang and Ma46 included 12 studies with 513 patients: 5 randomized controlled trials, 5 cohort studies, and 2 case-control studies. The results showed that treatment with steroids may reduce the risk of strictures on average by 60% and decreased the need for EBD. Local injection of steroids seems to have an effect similar to that of systemic administration in terms of stricture development, and it is superior in terms of number of EBD sessions. Most publications are small retrospective case series with heterogeneous steroids injection protocols. Two larger controlled studies have been published in this field. Hashimoto et al5 demonstrated in a prospective study involving 21 patients and 20 controls with high-risk post-ESD strictures that a 3-triamcinolone injection protocol at days 3, 7, and 10 after ESD induced a statistically significant decrease in the stricture rate (19% vs 75%, P < .001). Similar results and decrease in the stricture rate after a single triamcinolone injection were showed by Hanaoka et al15 in 30 patients (10% vs 66%, P < .0001) with a 7% complication rate.5 On the other hand, Funakawa and colleagues47 in a retrospective analysis of 45 patients with high-risk post-ESD strictures, 35 of which had been treated by repeated injections of triamcinolone, found steroid injection to be safe but ineffective. In particular, 10 of the 12 (83%) patients with circumferential ESD developed strictures despite steroid injections, and even in the resections involving 75% to 100% of the circumference, the rate of stricture was not significantly different between the groups, with 34.8% (8/23) and 40% (4/10) stricture rates in the triamcinolone and control groups, respectively. Takahashi et al48 in a randomized controlled study compared triamcinolone injections after ESD with a control group of patients without any treatment after an ESD procedure. The stricture rate for the triamcinolone group was 62.5% (10/16), which is still high but is lower than that in the control group (87.5%, 14/16). Nagami and colleagues49 evaluated the prophylactic efficacy of a single locoregional triamcinolone injection for stricture formation after ESD for superficial esophageal cancer using propensity score matching analysis. A total of 150 patients were enrolled in this study and 37 patients with or without steroid injections were matched after propensity score matching analysis. The incidence of stricture formation with triamcinolone injections (80 mg) immediately after the ESD procedure decreased from 45.9% (in controls without steroids) to 18.9% (P = .016). The mean number of balloon dilations also decreased from 28 ± 4.6 in the control group without steroids to 0.6 ± 1.5 (P > .01).
In a case report Lee et al50 showed complete success in preventing stricture with prophylactic triamcinolone injection therapy in a patient who had a near whole-circumferential esophageal mucosal defect after ESD, and thus a higher potential for esophageal strictures. The results of another retrospective study by Hashimoto’s group51 showed that triamcinolone injection protocol without oral steroids or in combination with prednisolone 0.5 mg/kg for 6 weeks had better results in prevention of post-ESD strictures compared with the control group. The stricture rate was 19% for the first group, 33% for the second, and 75% for the controls (P < .001). Mori et al52 conducted a randomized head-to-head comparison between steroid injection therapy plus EBD vs steroid gel application plus EBD. No significant difference was observed in the stricture rate between these 2 groups. The authors reported that the requirement for technical expertise and the total procedure time for the gel application study arm (6.87 minutes) was lower (although not statistically significant) than that of the injection group (total procedure time 7.33 minutes). In addition, gel application provides an alternative method of stricture prevention for individuals on oral anticoagulation or antiplatelet medications, as it obviates the need for needle injection, thus lowering potential bleeding risk. In regard to local steroid injection-related adverse events, the safety and efficacy of this therapy have yet to be fully evaluated. Tsujii and colleagues53 investigated the outcomes after esophageal ESD for large lesions and complications associated with EBD performed after local dexamethasone or triamcinolone injections (4 to 8 mg/mL solution) into the submucosal tissue of the ulcer bed for prevention of post-ESD strictures. Some patients who had recurrent strictures received subsequent steroid injections concomitant with EBD. The authors noted a high risk of post-ESD complications. Perforations during EBD only occurred in 6 of 28 patients who had received local steroid therapy. All 6 perforation cases necessitated prolonged hospitalization, and 2 patients required surgical treatment. The authors speculated that local steroid injection reduces the strength and elasticity of the esophageal wall resulting in vulnerability to dilating pressure, and gentle injection into the residual submucosa without deeper insertion of the needle is essential in this procedure. Also, EBD should be performed with great caution, especially after local steroid therapy, for example, in a stepwise approach starting with a small balloon diameter to reduce risk of perforation. Despite some limitations triamcinolone injections can prevent esophageal strictures even for large mucosal defects (Figure 26-3).
Oral steroid treatment does not require an endoscopic procedure, therefore the burden on patients is not heavy. Two studies conducted by Isomoto and Yamaguchi’s group28 showed the efficacy of oral steroids in preventing post-ESD esophageal strictures. In a small prospective study that included 7 patients (4 patients received oral prednisolone and 3 patients underwent EBD after ESD), strictures developed in 2 (50%) in the steroids group and in all 3 patients without oral prednisolone (100%). Similar results were shown in a larger prospective study.28 Out of 19 participants who received oral prednisolone, post-ESD stricture developed in 5.3% (1 case) as compared to 31.8% (7 out of 21 cases) in the group that received EBD without steroid therapy. Recently Zhou and colleagues54 suggested an optimal administration program of oral prednisone therapy and demonstrated that it is safe and effective for the prevention of esophageal stricture in patients after complete or semicircular ESD for esophageal squamous cell carcinoma. Thirteen of 23 patients included in the study received oral prednisolone at a dose of 30 mg/day on the third day post-ESD, and then tapered gradually (30, 25, 20, 15, 10, and 5 mg for 14 days). Postprocedural esophageal stricture was significantly lower in the prednisolone group (23.1%) compared to the control group (80%) (P < .05). A significantly higher number of EBD sessions were performed (P < .05) in the controls (13.5) than in the steroids group (0.69). There were no adverse events related to oral prednisolone or the procedure itself, and no treatment-related mortality was observed during the 12-month follow-up.54 But this was a nonrandomized controlled study performed in a single endoscopic center.
Oral steroids have to be administered with caution because if the cumulative dose of prednisolone is greater than 1000 mg over 8 weeks, there is a potential risk of prednisolone-related adverse events,55 including immune suppression, infection, optical damage, psychiatric disturbance, diabetes, peptic ulceration, and osteoporosis, which could represent severe clinical problems.56–58 According to the guidelines of the European League Against Rheumatism, the risk of bone fractures appears at doses of up to 7.5 mg per day during 3 months.59 Kataoka et al60 reported promising results of a short-period, low-dose oral prednisolone treatment for the prevention of stricture after ESD for early-stage esophageal cancers involving up to 50 % of the circumference. With an average cumulative dose of 420 mg of prednisolone within 3 weeks, the incidence of stenosis was 17.6 %, whereas it was 68 % without prevention. In this study a low cumulative dose of 420 mg for 3 weeks was highly beneficial in preventing the induction of osteoporosis. The success with steroid therapy in preventing post-ESD strictures in many cases is suggestive of its efficacy. The question that remains unanswered is why the results are not uniform (Table 26-1). Possibilities include variations in technique and individual patient characteristics.
Other Antiscarring Agents
Local administration of controlled-release antiscarring agents can prevent further esophageal post-ESD stricture development. According to recent publications MMC, 5-FU, and BTX-A, which are used as antiscarring agents for the treatment of hypertrophic skin scars, may prevent postoperative esophageal strictures in animal models and human.
Antimitotic (Antineoplastic) Drugs
MMC, which is an anthracycline chemotherapeutic agent derived from some Streptomyces species, can simultaneously inhibit fibroblast proliferation and reduce scar formation when topically applied to a surgical lesion or wound. MMC has been used for treating strictures in a variety of anatomical locations and in the prevention and elimination of scars in fields including ophthalmology, plastic surgery, otolaryngology, urology, orthopedics, and the upper GI tract.63,64 In a retrospective study of 5 patients who developed refractory esophageal stenosis after ESD and needed repeated balloon dilation, Machida and colleagues65 found no recurrence or drug adverse effects in the 4.8 months after the injection of MMC at the site of EBD.