Oral steroid prophylaxis is effective in preventing esophageal strictures after large endoscopic resection

Introduction

In recent decades, the management of superficial esophageal lesions has evolved significantly with the addition of therapeutic endoscopy, i.e., endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD). EMR for Barrett’s esophagus with high-grade dysplasia (HGD) and intramucosal adenocarcinoma (IMC) is performed by multiband or cap mucosectomy, with a success rate of up to 87% for neoplasia [1-4]. Furthermore, after complete endoscopic resection of Barrett’s esophagus with HGD or IMC, the incidence of metachronous neoplasia appears to decrease [5-7]. Complete Barrett’s eradication can be performed by endoscopic resection or radiofrequency ablation, but its effectiveness is limited because of the high risk of iatrogenic stricture, which ranges between 17% and 88% [4,5,8,9]. Treatment of these stenosis requires multiple sessions of dilation, impairing the patient’s quality of life.

Previous studies have shown that either local [10-12] or systemic corticosteroids [13-15] can effectively reduce the need for endoscopic balloon dilation after large ESD. We report our experience from using prophylaxis with oral steroids to prevent iatrogenic stenosis after endoscopic resection of Barrett’s esophagus extending to at least half the circumference. The primary aim of our retrospective study was to determine the rate of stricture under oral steroid prophylaxis after large endoscopic resection. Our secondary aim was to determine the association of this preventive strategy with adverse effects and complications.

Patients and methods

Treatment protocol

EMR was performed by 4 operators experienced in esophageal EMR and 2 operators with poor experience in esophageal EMR under their direct supervision. Barrett’s esophagus was examined under high-definition white-light endoscopy (EG29-i10, EG-2990Zi, PENTAX MEDICAL©, Tokyo, JAPAN) and, if necessary, by chromoendoscopy with acetic acid. All resections were performed using a multiband mucosectomy technique (Duette; Cook Medical, Winston-Salem NC). If necessary, a submucosal injection of saline mixed with indigo carmine was performed. A microprocessor-controlled electrosurgical generator (ERBE VIO 300; ERBE, Tübingen, Germany) was used. Resection was performed from the palisade vessels of the gastroesophageal junction distally and was extended to remove at least a hemi-circumference of Barrett’s esophagus (Fig. 1). The Prague classification for Barrett’s esophagus [16] was used in all patients and the Paris classification [17] was used for nodular or suspicious lesions. Lesions classified 0-IIc and 0-III by the Paris classification were considered ineligible for EMR. For simplicity, we used the codes R0 for healthy vertical margins and R1 for positive vertical margins (lateral margins could not be evaluated given the large piecemeal EMR).

Figure 1 Example of large esophageal endoscopic mucosal resection

Oral prednisolone was started 1 day after EMR in a dosage of 30 mg/day. The dose was then gradually tapered in decrements of 5 mg/day every 2 weeks for 1 month, followed by decrements of 5 mg/day every week for the next 4 weeks.

Results

Between July 2013 and September 2015, 31 patients (27 men, 4 women, sex ratio 7:1; mean age 63 years) with biopsy-proven Barrett’s LGD, HGD or IMC met the inclusion criteria. Patients’ characteristics and resection details are shown in Table 1.

Table 1 Patient demographics and clinical outcome features

The median circumferential extent and maximum Barrett’s mucosa length were respectively 3 cm (range C0-C10) and 5 cm (range M2-M11). Pre-EMR histology showed 19 patients with IMC, 8 with HGD and 4 with LGD. Based on the EMR specimen, definitive histology was LGD in 13 patients, IMC in 16 and submucosal carcinoma in 2 (sm1=1, sm2=1). The resection was complete (R0) in 28 patients, requiring 1-3 sessions (median 2). Both patients with submucosal cancer had positive vertical margins (R1). They underwent surgery and definitive histology was pT1bN0M0.

Oral prednisolone in a dosage of 30 mg/day was started the day after EMR, once the patients were permitted oral intake. All patients included in this study followed the steroid protocol.

Immediate and secondary adverse events were recorded during a median follow up of 10 months (range 1-22). No digestive perforation occurred in any case. Two patients had significant bleeding requiring endoscopic hemostasis. One patient died from a sudden and massive hemorrhage 48 h after the second EMR session.

Four patients (13%) required endoscopic balloon dilation for symptomatic esophageal stricture. The endoscopic dilation was effective in all cases, requiring 1-4 sessions (median 2). The mean dilation diameter was 16.5 mm [15 mm-18 mm]. There were no adverse events related to oral steroids. The study design is shown in Fig. 2.

Figure 2 Study design R0, healthy vertical margins; R1, positive vertical margins; EMR, endoscopic mucosal resection

Discussion

Endoscopic resection has recently been accepted as first-line therapy for early esophageal cancers, such as squamous cell carcinoma and dysplasia of Barrett’s esophagus. Secondary esophageal stricture is a common and well known complication after a large resection. It necessitates some iterative endoscopic procedures and hospitalizations, and may delay or compromise treatment if more sessions are needed. Prophylactic self-expandable metal stent insertion has been proposed, but the morbidity rate is excessive [18]. Anti-inflammatory approaches for preventing esophageal strictures after endoscopic resection are based on the concept that subsequent strictures may be suppressed by inhibiting the infiltration of inflammatory cells, the hyperplasia of granulation, and fibrosis. Local injection of triamcinolone seems to be effective, but is of limited use in large circumferential resections [10,11]. Furthermore, it may cause an ulcer at the injection site.

Several research teams have studied preventive systemic corticosteroid therapy and have reported promising early results [13-15]. However, high doses of prednisolone have been associated with adverse effects, such as severe infections, peptic ulcers, hyperglycemia, psychiatric symptoms, and osteoporosis [19].

Since the publication of the above results, we have been using a low dose of systemic corticosteroids to reduce the stricture rate post EMR. This study presents the results from clinical practice in our center. This retrospective study provides more support for the use of systemic corticosteroids in this context. Indeed, the stricture rate (13%) was much lower than the 17-88% reported in the literature [4,5,8,9]. We use relatively low doses over a limited period and it seems unlikely that this has an impact in the long term. Moreover, the stricture rate increases drastically when 75% of the circumference is involved [20] and in our study 24/31 patients underwent circumferential resection in 2 to 3 sessions (Table 1). The other patients (7/31) underwent resection in 1 session of esophageal mucosa larger than the half-circumference, but less than 75% of the esophageal lumen (Table 1).

There were 3 early (within 48 h) adverse events (3 hemorrhages and no perforation). Two hemorrhages were treated endoscopically. One patient had a sudden fatal hemorrhagic shock 48 h after the second EMR session. None of the complications were related to steroid therapy. The two patients with submucosal infiltration underwent laparoscopic transhiatal esophago-gastrectomy.

Our study is, of course, limited by being a retrospective analysis and by the relatively small number of patients. Indeed, only 31 patients were able to benefit from the steroid therapy protocol in its entirety. Some patients refused and the institution of the protocol was at the discretion of the operator. Although the treatment duration was short (8 weeks), patients with a psychiatric history, known diabetes or osteoporosis were not treated. Furthermore, we excluded from this analysis patients who underwent radiofrequency ablation after EMR, so as not to add a confounding factor.

The findings of our study are consistent with those reported in the literature, and the side effects appear to be mild, although this will need to be verified by an intention-to-treat study.

In conclusion, esophageal strictures cause dysphagia, which obliges the patient to undergo repeated balloon dilation procedures or implantation of temporary stents. These esophageal strictures decrease the patient’s quality of life and have their own potential complications (perforation, failure). Anti-inflammatory approaches are promising and have been shown to be effective in several studies. However, a multicenter randomized study will be needed to determine the best route of administration and other related factors (local or oral and posology). We believe that several surgical resections related to this complication could be avoided. With developments in the ESD technique the stricture rate may increase [21]. In the future, a tissue-engineering approach could be another line of research, along with cell therapies [22,23]. New biological agents are also promising [24].