Endoscopic dilation (ED) in a step-up manner is the primary modality used to manage benign esophageal strictures [1, 2]. Majority of EDs are done initially at 2–4-week intervals. Simple strictures respond well by ≤ 3 ED sessions [3,4,5]. Complex benign esophageal strictures (CBES) that are long, tight, tortuous, or multifocal can be refractory (RBES) defined as inability to reach a luminal diameter of ≥ 14 mm despite 5 dilations at 2-week intervals, or requiring ≥ 1 dilation every 4 weeks to maintain a diameter of 14 mm [6]. Based on the etiology, the success rates of ED in these patients could be as low as 60–65% [7, 8]. Additional interventions such as electro-cautery incision or stents may be required.

Following breach in mucosal barrier (example, after dilation), fibrosis sets in by 2 weeks [9, 10]. Hence, at one tertiary-care center, a strategy of dilating patients with CBES initially every week was adopted, especially in those who were referred after failed dilations done at ≥ 2-week intervals. The objective of this pilot study was to evaluate the long-term clinical outcomes in a cohort of patients where this approach was used.

Materials and methods

Data was extracted from a prospectively maintained database on a cohort of consecutive patients with CBES who underwent endoscopic dilations done initially every week as per the protocol described below and in Fig. 1. The study was approved by the local Institutional Review Board and patients/health-care power of attorney gave informed consent.

Fig. 1
figure 1

This figure shows the protocol used in this study to manage patients with complex benign esophageal strictures

Full size image

Study subjects

All patients were ≥ 18 years and had biopsy proven CBES defined as being long (> 2 cm, except anastomotic strictures that can be short), tight (< 10 mm diameter), tortuous, or multifocal. Exclusion criteria were (a) malignant stricture, (b) extrinsic compression of the esophagus, (c) fistula/leaks, (d) eosinophilic esophagitis, (e) primary motility disorders, (f) cannot keep follow-up appointments, (g) other malignancies or co-morbidities limiting life expectancy, and (h) unsafe (laryngeal penetration/aspiration) oro-pharyngeal swallowing. Complex benign esophageal strictures resulting from radiation given for head-and-neck cancer underwent swallow study at baseline to document safe swallowing. This study was repeated if patient developed episodes of aspiration.

Patients were classified into two groups. Group A included patients who failed to achieve success (as defined below) despite EDs done at ≥ 2-week intervals. Group B included patients with newly diagnosed CBES in whom the protocol was used upfront. Dysphagia was graded on a previously described scale as Grade 0: can swallow normal diet, no dysphagia; Grade 1: difficulty in swallowing some solids; Grade 2: can swallows only semi-solid/soft to liquid diet; Grade 3; can swallow only liquids, and Grade 4: cannot swallow even liquids, total dysphagia [11]

Protocol (Fig. 1)

Step-up EDs were performed starting at weekly intervals (Fig. 1) till an esophageal luminal diameter of 14 mm was reached. Attempts were made to safely go up to higher diameters so as to achieve more durable response [12]. In patients with radiation or corrosive injuries, wire-guided bougie was preferred over balloon to achieve pan-esophageal dilation as these patients may have multifocal strictures. Moreover, segments between 10 and 13 mm in diameter can be missed using a standard 9.5 mm diameter endoscope while focusing only on the tight (< 9 mm) regions if a balloon is used. Focal strictures were dilated using balloon or bougie.

All dilations were performed with anesthesia assisted sedation. Fluoroscopy was used only for guidewire placement if even the 5 mm diameter endoscope would not traverse the stricture. Based on the stricture diameter as assessed by the endoscope (< 9.5 mm standard or < 5 mm thin endoscope), and/or using open biopsy forceps, dilation was performed to a value that was considered safe (resistance-to-feel) using standard techniques and recommendations. After dilation, re-look endoscopy was performed in all patients. Patients also underwent immediate contrast fluoroscopy evaluation if perforation was suspected. All strictures were periodically biopsied (every 2–3 months).

At each weekly ED, how much the stricture has regressed (or not) in diameter was assessed. If the stricture had partially regressed (“two-steps-forwards and one-step-back”), dilation was performed to higher diameters and the weekly dilations continued till a diameter of ≥ 14 mm was achieved. Following this, the frequency of dilations required to maintain this diameter was determined by increasing (or decreasing) the intervals between dilations. Steroids were injected into the stricture if the patients continued to require frequent dilations.

Additional measures

If the patient continued to require weekly dilations and the stricture kept regressing to its original diameter (“two-steps-forward and two-steps-back”), additional measures were used. For short strictures, electrocautery-incision with needle-knife was performed (Fig. 2, Video 1) [13,14,15,16,17]. Lumen apposing metal stents (LAMS) were also used for short (< 1 cm) strictures (Fig. 3) [18, 19]. For longer strictures, fully-covered self-expanding metal stent (FC-SEMS, 18/23 mm diameter) were used [20,21,22,23,24]. Stents were sutured in place only if they migrated and required re-positioning [25, 26]. All patients where the stent bridged across the GE junction were give proton pump inhibitors and advised to follow anti-reflux measures. Stents were removed after 4 weeks. Combination techniques such as electrocautery-incision followed by FC-SEMS placement in the same session was also used in patients with tight strictures (< 5 mm) to further “tear” open the incised stricture and seal micro-perforation if any (Video 1).

Fig. 2
figure 2

Electrocautery needle-knife incision of an anastomotic stricture

Full size image
Fig. 3
figure 3

Lumen apposing metal stent (LAMS) across an anastomotic stricture

Full size image

One week after additional measures (electrocautery/stent removal), weekly EDs were re-initiated. This was done to assure that the gains made by the additional measures were not lost by waiting longer.

Objectives: success and failure

Primary objective was to determine the proportion of patients who achieved success defined as achieving a luminal diameter of ≥ 14 mm and patient remaining dysphagia-free (grade ≤ 1) with none to no more than 1 dilation required every 3 months to maintain this diameter/dysphagia-free status. These patients were advised to call if dysphagia recurred. Secondary objectives were to determine what proportion of patients needed additional measures besides dilation to achieve success, did not achieve 14 mm diameter, failure rates, and adverse events (AE).

Patients who did not achieve success (14 mm diameter) as per the above criteria were classified as failures. Those who achieved success and then reverted to > 1 dilation every 3 months were also considered as failures. These patients were offered training in self-dilation. [27, 28] They were trained in real-time on their first visit watching a fluoroscopy screen while performing self-dilation with a weighted dilator (Maloney®, Fig. 4) and then instructed to dilate themselves at home at least once daily. Patients were advised to call for “rescue” ED if they developed discomfort/increasing resistance while passing the dilator.

Fig. 4
figure 4

Training in self-dilation of esophageal stricture. A Initially the patient is trained to self-dilate while watching fluoroscopy monitor in real-time as biofeedback. B Two months later at one follow-up clinic visit, patient demonstrating self-dilation (no fluoroscopy). Patient preferred to stand instead of the recommendation to sit for concerns of falling. The tape on the dilator (arrow) marks the level up to which the dilator needs to be inserted and is based on prior endoscopy assessment of the stricture location

Full size image

Patients who refused self-dilation and required EDs at the rate of more ≥ 2 every 3 months were also considered as failures. Similarly, those who preferred long-term stenting were counted as failures despite being dysphagia-free.

Follow-up data was collected each time during scheduled endoscopies. Those who responded were followed at approximately 3 months, 6 months, and then yearly (some via telephone).

Statistical analysis

Statistical analysis was done using StatDirect 3.3.5. Descriptive statistics were used to summarize demographic and clinical characteristics of the patients. Measures of central tendency (mean and median for normally distributed variables and for skewed distribution respectively) and measures of spread (standard deviation and range) was used for continuous data while frequency listing was used for categorical data summarized using proportions. Continuous variables were compared using t-test for normal data and Mann–Whitney test for skewed data while the Chi-square or Fisher’s exact test were used to compare categorical variables. P value of < 0.05 was used as a determinant of statistical significance.

Results

Study subjects

A cohort of 57 consecutive patients with CBES were enrolled and underwent a total of 488 dilations (not counting dilations done daily by those performing self-dilations).

Group A: Cohort of 21 patients (mean age 65 ± 13 SD years; male 9, female 12) with CBES were enrolled. At baseline, 10 (48%) patients had esophageal lumen of < 5 mm and dysphagia grade of 3–4. In 11 patients (48%), the baseline diameter was between 5 and 10 mm, and dysphagia grade was between 2 and 3. These patients had failed a median of 6 (range of 4–12) prior dilations done at a mean interval of 17 ± 9 SD days between each dilation. Six of these patients who failed prior EDs also had steroid injected into the stricture. Weekly ED as per the protocol was initiated and patients were followed for a median of 4 years (range of 1–9 years). One patient was lost to follow-up for 5 months and then re-established care after recurrent dysphagia.

Group B: This group included 36 patients (mean age 61 ± 13 SD years; male 21, female 15) with newly diagnosed CBES [Baseline luminal diameter < 5 mm in 19 (53%) with dysphagia grade 3–4; luminal diameter between 5 and 10 mm in 17 (47%) with dysphagia grade 2–3]. Majority of these referrals were from within the institute referred by oncology, surgery, gastroenterology, and internists. Following interventions as per protocol, these patients were also followed up for a median of 4 years (range of 1–9 years).

The etiology of the strictures between the two groups were comparable (Table 1). One patient (Group A) had lichen planus induced stricture and another was NG-tube related following post-ERCP pancreatitis. Two patients in Group B had esophageal leaks managed with FC-SEMS. After the stent removal, the leaks resolved but they developed CBES (< 5 mm).

Table 1 Etiology of strictures in the two groups
Full size table

Clinical outcomes

Success: In 52 of the 57 patients (92%), a luminal diameter of ≥ 14 mm was achieved. On follow-up, the stricture recurred in 10 that required more than one ED every 3 months. Hence the overall success rate of 74%. In Group A, 57% (12 of 21) of patients who failed dilations done at the traditional 2-week intervals achieved long-term success with weekly EDs (mean 8 ± 4.7 dilation/patient). The success rate was even higher in those where this protocol was applied upfront (Group B: 83.3%; P 0.033; mean 6.5 ± 5.5 dilations/patient). Of the 12 patients who achieved success in Group A, 7 patients (58%, ITT 33%) achieved long-term success by just dilating them initially once every week. Remaining 5 patients (42%) did not respond to weekly dilations and required temporary FC-SEMS. Following stent removal, they started responding to weekly dilations and eventually achieved long-term success (Table 2). In Group B, 25 of 30 patients (83.3%, ITT 69%, mean 6.5 ± 5.5 dilations/patient) achieved long-term success by weekly dilations. In 5 (16.7%) patients additional interventions (FC-SEMS: 4, electro-cautery-incision plus SEMS: 1) followed by re-initiating weekly EDs were needed to achieve long-term success (P = 0.057 compared to Group A).

Table 2 Comparison of success rates and interventions to achieve success in the two groups
Full size table

Failures: Etiology of the strictures in those who failed to achieve success are shown in Table 3. Majority of these were post-radiation (40%) and anastomotic stricture (33%). In 5 patients, a diameter of 14 mm could not be achieved and in others, the strictures recurred requiring > 1 dilation every month after achieving a 14 mm diameter. Table 4 shows the outcomes of those who failed the protocol. Four patients initially achieved success but within 12-month returned with recurrence of dysphagia requiring more than 1 dilation every 12 weeks. Although 3 of them again achieved success, their need for dilations kept varying from one dilation every 2 to 4 months. Only 2 accepted to train in self-dilation (in one with radiation stricture, the husband dilated the patient every day). One patient with anastomotic stricture initially had a LAMS placed with good benefit. Following stent removal, the stricture recurred within 4 weeks. He developed a pseudo-diverticulum proximal to the stricture that required partial trans-esophageal diverticulotomy. As the patient continued to require frequent dilations and was not a candidate for self-dilations (diverticulum) he opted for long-term LAMS.

Table 3 Etiology of the strictures in those who failed to achieve success
Full size table
Table 4 Outcomes of patients who failed the protocol
Full size table

Mortality

Eleven patients deceased of whom 5 were in the failure group and 6 had already achieved success. Except for one patient as described below (1; failure group), all other causes for mortality were unrelated to EDs:

Mortality in failure group: (1) Radiation stricture being dilated every 4–6 weeks. Patient did not want SEMS or self-dilation. At 28 months perforation occurred during a dilation that resulted in a trachea-esophageal fistula. Covered SEMS was placed. Family opted for hospice care as the patient had severe dementia, (2) 12-month follow-up developed metastatic head-and-neck cancer, (3) 42-month follow-up, respiratory failure from smoking and worsening emphysema. (4) 24-month follow-up, recurrent lung cancer, (5) 18-month follow-up, recurrent laryngeal cancer.

Mortality in success (dysphagia-free) group: (1) 36-month follow-up, recurrent adenocarcinoma at anastomotic site with brain metastasis, (2) 20-month follow-up, myocardial infarction, (3) 78-month follow-up, metastatic colon cancer, (4) 18-month follow-up, metastatic head-and-neck cancer, (5) 36-month follow-up, cirrhosis with hepatic decompensation, and (6) 42-month follow-up, progressive multiple sclerosis with urinary sepsis, hospice care.

Other adverse events

Chest pain: following FC-SEMS in one patient with post-RFA/EMR stricture, Stent had to be removed within a week. Another patient required pain medications for 5 days following FC-SEMS.

Continuous oozing of blood following dilation in one patient with post-radiation stricture who also had significant neovascularization of the esophageal mucosa. Although this patient did not require hospitalization or blood transfusion, the continuous oozing required tamponade with a FC- SEMS that was removed after one week. No further significant bleeding was observed with subsequent dilations.

Esophago-pharyngeal reflux with nocturnal cough in one patient with gastric pull-up and anastomotic stricture following LAMS placement. Following life-style changes and raising the head-end of the bed, symptoms resolved. Options like injection botulinum toxin into the pyloric sphincter of G-POEM were not required.

Unrelated to the dilations, one patient with surgery for recurrence of cancer developed a retropharyngeal abscess that resolved with drainage and antibiotics.

Discussion

Endoscopic dilation is the primary approach in managing patients with benign esophageal strictures [1, 2]. The goal is to achieve a luminal diameter at ≥ 14 mm [6], although more durable response has been reported with larger diameters [12]. Simple strictures such as webs, rings, or peptic respond well to ≤ 3 ED sessions [3,4,5]. CBES such as those from radiation, corrosives, ESD, pills, or surgery can be refractory despite EDs done at 2-week intervals as defined by Kochman et al. [6] These patients are then characterized as “failure” to dilations and referred to tertiary-care centers for additional measures such as electrocautery-incisions or FC-SEMS that are not without risks and generally not performed in the community. In this study, instead of considering these patients as “failure” to dilations, the authors investigated the clinical outcomes of a protocol where EDs were initially done every week instead of the traditional every 2-weeks. Compared to success being defined as achieving a luminal diameter of ≥ 14 mm, this study also incorporated maintaining this diameter over long-term follow-up in the definition of success.

Using this approach, over 55% of patients who failed prior dilations done at ≥ 2-week intervals achieved long-term success and many of these patients just responded to weekly dilations without the need of stents or electro-cautery-incision. Over a median of 4-year follow-up, these patients continue to remain dysphagia-free with minimal interventions. When this protocol was applied upfront in those with newly diagnosed CBES, the success rate was even higher. The exact reason for this difference is not known. Based on the principles of wound healing as described below, it could be those who were referred after several prior dilation sessions done at ≥ 2-week intervals progressively developed worsening fibrosis as compared to the group where the protocol was applied soon after the diagnosis. Higher success rates were also reported in another study where dilations were performed every week [4]. However, they included some patients with simple strictures and the follow-up was less than 2 years.

Some patients in both groups did not respond to weekly dilations. Hence additional intervention such as FC-SEMS or electro-cautery-incision were needed. There was a trend toward less need for additional interventions in those where the protocol was applied upfront (P = 0.057). Interestingly, when weekly dilation protocol was re-started in a week after additional interventions, a significant proportion of patients in both groups started responding to ED when initially they were not.

The rationale behind the approach of initially starting with weekly dilations and using this approach again immediately after FC-SEMS/electro-cautery was to not lose out on the gains made by waiting too long. Although the exact duration of “waiting too long” has not been defined, we opted for the weekly intervals based on studies looking at wound healing. After the mucosal barrier is broken (such as after dilation), fibrosis with luminal distortion (stricture) sets in by 2 weeks while re-epithelization can take over 5 weeks [9, 10]. Dilating the stricture in a step-up manner every week may be one option to address this as was done in this study. Although delaying fibrosis with steroid injection into the stricture has been tried (including several patients in this study), results have been variable [29,30,31,32]. Since re-epithelization to restore the mucosal barrier can take over 5 weeks [10], another option could have been to rapidly restore the epithelial barrier as has been tried using principles of regenerative medicine such as injecting autologous pluripotent cells into the site of the mucosal break, or using autologous cell-sheets to immediately cover the breached epithelium [33,34,35,36,37,38,39,40,41,42]. Besides animal studies, these techniques have also shown promising results in human beings [38, 42]. As these techniques are expertise driven, expensive, and have regulatory limitations, the authors adopted the frequent dilation approach.

Since patients with CBES can have a variable course over time, studies looking at the response to interventions with short term follow-ups of less than 1–2 years may not represent the long-term outcomes. One of the strengths of this study was the median follow-up of 4 years. There are however several limitations such as there was no control arm. However, as several patients had failed prior dilations done at ≥ 2-weekly intervals, they acted as their own controls. Over 55% of these patients started responding to weekly dilations when previously they had failed to do so. Secondly, no cost analysis was done on frequent EDs. Although the initial cost of weekly dilations was much higher, 74% of patients were dysphagia-free with a mean of 8 ± 4.7 dilations per patient in the group who had failed previous dilations done at ≥ 2-week intervals. This was 6.5 ± 5.5 dilations per patient in the group where this protocol was applied upfront. Hence it is anticipated that over a long-term period (median follow-up of 4 years in this study), this led to significant cost savings, not to also forget a significant improvement of quality of life. Accessibility to GI endoscopy centers for weekly dilations may be an issue based on the area/country of practice. Experience from a very busy tertiary-care center like ours showed that, once patients were accepted for weekly dilations, appointment for the next dilation in a week was already successfully factored into the schedule and the intervals between dilations accordingly adjusted if the patient starts responding.

Whether the frequent dilation approach exposed the patients to higher risks of adverse events was not shown to be the case in this study. There was one dilation related perforation and one bleeding episodes among the 488 EDs performed. Although all EDs were performed by advanced endoscopy faculty, most of the gastroenterologists are trained in ED (core training) and can easily adopt the weekly ED protocol in the community setting hence obviating the need for measure such as electrocautery-incision or FC-SEMS. Lastly, since the number of patients in each group was small, sub-group analysis on the response rates based on etiology was not possible.

In summary, the protocol of initially dilating every week and then increasing the interval between dilation sessions based on response was shown to be safe and effective in the management of patients with CBES who failed dilations done at the traditional 2-week or more intervals. If these results are validated in large prospective randomized studies, it will have great implications on how we manage these patients since weekly EDs can safely be performed in the community as against electro-cautery incision or placing FC-SEMS. Better outcomes were observed in those where this approach was used sooner than later. Those who initially did not respond to weekly dilation and hence needed additional interventions such as FC-SEMS/electro-cautery, re-initiating weekly dilations led to further success. Prospective randomized studies on a larger cohort of patients will be required to do subgroup analysis on the influence of the etiology of the stricture on outcomes.