The role of endoscopic dilation and stents in refractory benign esophageal strictures: a retrospective analysis

This is a single-center retrospective study. One thousand forty-six patients with esophageal strictures were treated by dilation or stents between January 2009 and December 2017 in our hospital, and we excluded the patients with endoscopic incision. All medical records were retrospectively reviewed. Patients were included in the study if (1) they had been diagnosed with esophageal benign stricture according to clinical manifestation, gastroscopy, and pathology, (2) they had persistent symptoms of dysphagia with at least three sessions of endoscopic therapy, and (3) they had complete medical records and follow-up information, including gender, age at first therapy, RBES etiology, number, location, length and diameter of strictures, endoscopic therapy, the period between two consecutive endoscopic interventions, and adverse events. Patients were excluded if they (1) were younger than 16 years old, (2) had received less than 3 endoscopic interventions, (3) had been diagnosed with congenital esophageal stenosis, malignant esophageal strictures, or with esophageal fistula (Fig. 1). This study was approved by the Ethics Committee of West China Hospital of Sichuan University (IRB number: ChiCTR1800016321).

Patients were treated following local clinical practice in accordance with best practices. Gastroscopes (JIF - Q260 or JIF - Q260J, Olympus Medical Systems, Tokyo, Japan) were used according to the stricture characteristics and type of treatment. Each physician had over 10 years’ experience in endoscopy. Written consent for endoscopic dilation and stent was obtained before the operations. All patients received the standard dilation, by balloons and bougies or fully covered self-expandable metal stents (SEMSs) (MTN-SE C-membrane; 60–140 m; 20 mm; MicroTech, Nanjing, China). The dilation procedures were performed with a through-the-scope balloon (Type 5842; 3 ATM-12 mm, 4.5 ATM-13.5 mm, 8 ATM-15 mm; Boston Scientific Corp; Marlborough, America) and savary bougies system (Savary-Gilliard Dilator set, type SGD-70-1; length: 70 cm, diameters: 5 mm, 7 mm, 9 mm, 11 mm, 12.8 mm, 14 mm, 15 mm; Wilson-Cook Medical, America). Dilation was the standard treatment for all patients, with other alternatives, such as stents placed on a case-by-case basis at the discretion of the endoscopist performing the procedure, after an appropriate discussion with surgeons and a dietician, as needed. Patients with RBES undergoing esophageal dilation required multiple sessions, every 2 to 4 weeks or longer. Some patients with persistent dysphagia symptoms were unable to achieve acceptable symptom relief and needed an intensive dilation schedule, (e.g. as frequent as every 2 to 3 days). However, in our study, we considered all subsequent endoscopic interventions within 7 days as one treatment cycle. In our study stent removal was after performed 6–8 weeks (range 4–10 weeks) later to allow scar tissue remodeling.

The primary outcomes were to assess clinical effectiveness and adverse events. Endoscopic therapy success was defined as a persistent dysphagia-free period, with no need for any other interventions, lasting at least 6 months [10]. Treatment failure was defined as the need for any endoscopic intervention, including surgery [11], during the follow-up period. Adverse events included esophageal fistula, bleeding, stent dysfunction (stent overgrowth, stent migration and food impaction), endoscopy to readjust or remove stents, and intolerable chest pain. Early and late adverse events were defined as those occurring within 7 days and later than 7 days after stenting, respectively. The secondary outcome was to identify factors that affected the dysphagia-free period. A dysphagia-free period was defined as a time interval between two subsequent endoscopic interventions greater than 7 days. We considered all endoscopic interventions within 7 days as only one treatment cycle. All patients were followed via telephone contact from the third therapy session until study termination or patient death to assess symptom resolution.

Patient characteristics are shown in Table 1. In total, 75 RBES patients were included (54 male; median age, 59 years) (Fig. 1). Among these 75 patients, 39 (52%), 20 (26.7%), 10(13.3%), 3 (4%), 3 (4%) showed esophageal strictures due to surgery, ESD, caustic injury, mixed etiology, and achalasia of cardia, respectively. Achalasia of cardia was considered as benign stenosis according to the UK guidelines on oesophageal dilation [12]. The mixed etiology patients causing by caustic injury firstly, after esophageal anastomosis colon surgery, had persistent dysphasia after receiving several dilation sessions. Among the 75 cases, 9 (12%), 59 (78.7%), 4 (5.3%), 3 (4%) demonstrated esophageal strictures in the cervical esophagus, thoracic esophagus, ventral esophagus, and both cervical and thoracic esophagus, respectively. 56 (74.7%) patients had a single stricture, 14 (18.7%) had two strictures, and 5(6.6%) patients had three or more strictures. The median strictures length was 2 cm (0.5 cm, 10.0 cm), the median strictures diameter was 4 mm (1.0 mm, 8.5 mm), and the median strictures final diameter reached after each dilation session was 10 mm (5 mm, 15 mm). All 75 patients received repeated endoscopic dilations. The median number of endoscopic therapies was 5 (range 3, 21), and a mean of 1.9 (range 1, 6) SEMSs were placed in 20 (26.7%) patients. The mean time of stent placement was 29 days (range 7, 67). The median follow-up time was 31 months (range 6, 106).

Overall, the mean dysphagia-free period of all 75 patients was 3.4 months (95% CI, 2.5–4.3). The dysphagia-free periods in patients treated with dilation and stents were 3.7 months (95% CI, 2.7–5), and 2.3 months (95% CI, 1.5–3), respectively. To explore the trend in the dysphagia-free period over time, we used hierarchical linear models (Table 2). The result of the fixed effect intercept and slope was significant (P < 0.001). The dysphagia-free period had a linear growth trend over time, with an increase of 12 days per endoscopic therapy (95% CI, 8.7–16.7). The last dysphagia-free period was estimated to be 72 days (95% CI, 61.8–83.6). We included patient characteristics, therapy characteristics and stricture characteristics in the second level data to estimate prognostic factors associated independently with the trend in the dysphagia-free period. There was no significant correlation between the dysphagia-free period time trend and patient age, sex, etiology, stricture location, stricture number, stricture length, therapy, or initial dysphagia-free period. However, achalasia of cardia (p < 0.001, 95% CI, 97.5–210.9), postsurgical strictures (p = 0.046, 95% CI, − 4.4-47.2), stricture length more than 2 cm (p = 0.015, 95% CI, − 48- -5.6) and initial dysphagia-free period (p < 0.001, 95% CI, 36.3–68.3) tended to positively correlated with the intercept of the dysphagia-free period time-trend. The achalasia of cardia patients and postsurgical strictures with the stricture length less than 2 cm had longer dysphagia-free periods, compared with those patients with post-ESD strictures and stricture length more than 2 cm. Random-effects results showed that the variance component of the second data layer was large, due to the large variation among patients. The variance component, however, significantly decreased in the second level after excluding some stricture characteristics, especially stricture length, number and location. As such, compared with patients and therapy characteristics, stricture characteristics explain more of the variation in the time-trend of the dysphagia-free period.

Endoscopic therapy was successful in 46 (61.3%) patients. The success rate of dilation treatment (70.9%, 39/55) was higher than that of stents (35%, 7/20). Endoscopic treatment failed in 29 (38.7%) patients. In 27 (36%) patients, the end of follow-up was coincident with the last endoscopic treatment. One (1.4%) patient was treated with surgery, and another (1.4%) patient was treated with endoscopic gastrostomy feeding tubes. Nineteen patients had adverse events after endoscopic therapy. Six (8%) patients experienced esophageal fistula, caused by ESD (1/75), surgery (1/75), caustic (2/75) and mixed etiology (2/75), respectively. Three (4%) patients experienced intolerable chest pain, and one (1.3%) patient experienced bleeding. Ten (13.3%) patients experienced stent dysfunction including stent overgrowth in 3 (4%), stent migration in 6 (8%) and food impaction in 5 (6.7%) (Table 3). Fifteen patients died during follow-up, including lung and stomach tumor metastasis, severe pulmonary infection, and other causes unrelated to any endoscopic procedure.

There was a significant correlation between endoscopic treatment success and etiology, number of dilations and endoscopic therapy at univariate analysis (Table 1). In 39 patients with postsurgical stenosis, 18 patients had postoperative chemotherapy or radiotherapy. All the patients with a history of postoperative chemoradiation in our study had negative biopsies of the stricture. There was an insignificant correlation between therapy success and postoperative chemotherapy (95% CI, 0.109–1.814; p = 0.254) and postoperative radiotherapy (95% CI, 0.131–1.907; p = 0.307). Comparing bougies with balloons dilation, it showed no difference in terms of clinical resolution (p = 0.137). There was an insignificant correlation between the diameter after each dilatation session (p = 0.756) and the length of stenosis (p = 0.923) with endoscopic treatment success at univariate analysis. We included sex, age, etiology, dilation times, and location of strictures in the multivariate analysis. The multivariate analysis (Table 4) revealed that postsurgical strictures (OR, 13.082; 95% CI, 1.708–100.173; P = 0.013) and therapy times (OR, 1.714; 95% CI, 1.267–2.319; P<0.001) were the significant predictive factors for clinical resolution. The probability of endoscopic treatment failure was high in patients with postsurgical strictures. Both univariate and multivariate analyses suggested a correlation between etiology and clinical outcome. Therefore, we divided the etiology into four subgroups. The wide confidence interval of postsurgical strictures maybe caused by too many subgroups. It still had certain clinical significance. We conclude that the number of dilation was directly correlated with the probability of endoscopic treatment failure. Meanwhile, age (OR, 0.931; 95% CI, 0.868–0.998; P = 0.043) and caustic strictures (OR, 0.02; 95% CI, 0.001–0.418; P = 0.012) were related to clinical resolution. The treatment outcome tended to be better with age. But the statistical significance was weak. It hardly to say that there was clear clinical significance. Sex and other stricture characteristics were not significant predictive factors.