Predictors of postoperative stricture after esophageal endoscopic submucosal dissection for superficial squamous cell neoplasms

• Introduction
• Patients and methods
• Results
• Univariate analysis
• Logistic multivariate analysis
• Subanalysis for patients with m2 lesions in less than three quarters circumferential extension
• Discussion
• References

Background and study aims: Although endoscopic submucosal dissection (ESD) is becoming accepted as an established treatment for superficial esophageal squamous cell neoplasms, the risks for developing postoperative stricture have not been elucidated.

Patients and methods: This was a retrospective study at a single institution. From January 2002 to October 2008, 65 patients with high-grade intraepithelial neoplasms (HGINs) or m2 carcinomas treated by ESD were enrolled. Predictors of postoperative stricture were investigated by comparing results from 11 patients who developed strictures with those from 54 patients who did not .

Results: Significant differences between the two groups were observed in longitudinal diameter (45.0 ± 15.9 mm vs. 31.5 ± 13.6 mm) and circumferential diameter (37.2 ± 8.6 mm vs. 26.8 ± 9.7 mm) of the resected specimens, and the proportion of extension to the whole circumference of the lumen (< 1 / 2/ > 1 / 2/ > 3 / 4 : 2 / 4 / 5 vs. 40 / 13 / 1), histologic depth (HGIN/m2 : 2 / 9 vs. 41 / 13), and procedure time (85.6 ± 42.8 minutes vs. 53.3 ± 30.1 minutes). Multivariate analysis revealed that circumferential extension of > 3 / 4 (odds ration [OR]: 44.2; 95 % confidence interval [CI]: 4.4 – 443.6) and histologic depth to m2 (OR: 14.2; 95 %CI: 2.7 – 74.2) are reliable risk factors. Subanalysis for each category by combinations of these risk factors revealed that patients with lesions in > 3 / 4 of the circumferential area were associated with a high rate of postoperative stricture. By contrast, patients with HGIN lesions in < 3 / 4 extension have no probability of postoperative strictures. Additionally, subanalysis of patients with m2 lesions in < 3 / 4 circumferential extension revealed that circumferential diameter can be a reliable predictor for postoperative stricture.

Conclusions: Circumferential extension and histologic depth are the reliable risk factors for postoperative strictures. In combination with circumferential diameter, we can perform effective and appropriate preventive balloon dilatations after esophageal ESD.

Introduction

Endoscopic submucosal dissection (ESD) is a novel technique in the gastrointestinal tract including the esophagus [1] [2] [3] [4] [5], and progress is being made to elucidate the clinical outcomes. We previously reported high curability for large-sized high-grade intraepithelial neoplasms (HGINs) or m2 carcinomas by ESD [6]. Even for m3 or sm1 carcinomas, ESD is expected to have equivalent curability as conventional endoscopic mucosal resection (EMR) and surgical resection. Considering the higher co-morbidities after esophagectomy [7] [8] [9], and the higher incomplete resection by conventional EMR [10] [11], ESD is becoming accepted as an established treatment for superficial esophageal squamous cell neoplasms (ESCNs).

As large-sized ESCNs become candidates for ESD, postoperative stricture has arisen as a major concern with regard to long-term outcome. Once the patient has started to experience dysphagia due to postoperative stricture, repetitive periodic balloon dilatations for long periods are required [6] ([Fig. 1]).

In some institutes, patients with large-sized lesions routinely undergo preventive balloon dilatations immediately after ESD. However, the risk of developing postoperative stricture has not yet been elucidated, and many patients undergo preventive balloon dilatations without solid criteria.

In this study, we investigated the characteristics of lesions and patients who experienced dysphagia after ESD for superficial ESCNs, and attempted to predict the development of postoperative stricture in order to provide sufficient and appropriate preventive treatment.

Patients and methods

From January 2002 to October 2008, 116 consecutive superficial ESCNs in 93 patients who provided written informed consent were treated by ESD at the University of Tokyo hospital. All of the indicated lesions were preoperatively proven as or were suspicious for squamous cell carcinoma by endoscopic biopsy, and had a preoperative diagnosis of possible HGINs or intramucosal invasive carcinoma by chromoendoscopy with iodine staining [12] and narrow band imaging [13]. Among these patients, 65 with HGINs or m2 carcinomas, which had been diagnosed postoperatively, were enrolled to this study; patients with m3 or deeper carcinomas principally require additional treatments, including surgery and chemoradiation.

The ESD procedure was performed as previously reported in detail elsewhere [5]. The following electrosurgical knives were mainly used: the tip of an electrosurgical snare (thin type, SD-7p-1 [Olympus Co., Tokyo, Japan], from January 2002 to December 2002), Flex knife (KD-630L [Olympus], from January 2003), and Splash needle (DN-2618A [Pentax Co., Tokyo, Japan], from February 2007); a Flush knife (DK2618LN [Fujinon Toshiba ES System, Saitama, Japan]) was used occasionally. The injection agent was 1 % 1900 kD hyaluronic acid preparation (Suvenyl; Chugai Pharmaceutical Co., Tokyo, Japan) mixed with normal saline (from January 2002 to October 2003) or 10 % glycerin plus 5 % fructose and 0.9 % saline preparation (Glyceol [Chugai Pharmaceutical Co.], from November 2003 to March 2008). The ratio of solvent solution and Suvenyl was changed from a 1 : 3 ratio (form January 2002 to March 2004) to a 1 : 7 ratio (from April 2004 to March 2008), as a result of technical advances. From April 2008, hyaluronic acid preparation was mainly used for the endoscopic procedure (Mucoup; Johnson and Johnson K.K., Tokyo, Japan).

Lesions were resected with sufficient and appropriate tumor-free margins to minimize discrepancy between the size of lesion and that of resected mucosa. Longitudinal diameter was measured from the oral edge to the anal edge of resected specimens, which were stretched appropriately. Circumferential diameter was measured in the vertical axis against longitudinal axis. All of the investigations within this study were conducted after approval by the ethics committee of our institute.

Endoscopic characteristics of the lesions were classified according to the Paris endoscopic classification [14]. Histologic assessment was classified according to the revised Vienna classification of gastrointestinal epithelial neoplasia [15].

Follow-up endoscopy with iodine staining was usually performed at the 2nd month in order to check the healing of artificial ulcers and to exclude the presence of residual tumors. Only when patients experienced dysphagia was extraordinary endoscopy performed in order to check for postoperative stricture. In cases where postoperative stricture was confirmed on endoscopy, periodic balloon dilatations were started for approximately two sessions a week until the dysphagia improved [6]. Among 65 patients, 11 patients required repetitive balloon dilatations.

Eight categorical variables and four continuous variables were selected, based on the assumption that they can have an effect on the risk of postoperative stricture. In addition, all of the variables selected could be obtained during the perioperative period to predict postoperative stricture. The categorical variables were: sex, location of lesion, circumferential extension, macroscopic type, histologic depth, electrosurgical knife used, injection agent, and operator. The operators were designated as either senior (operators who had experience of more than 11 esophageal ESDs) or junior (experience of fewer than 10 esophageal ESDs). The continuous variables were: age, longitudinal diameter, circumferential diameter of resected mucosa, and procedure time for ESD.

All variables except for procedure time, were univariately compared between the 11 patients with postoperative stricture and the 54 patients without postoperative stricture. For procedure time, five patients were excluded before June 2003 because medical records about their procedure time had been lost. Eight categorical variables and four continuous variables were analyzed by x ² test and student’s t-test, respectively. To identify independent predictive variables for postoperative stricture, the variables were examined by stepwise logistic regression analysis, using JMP software (SAS Institute, North Carolina, USA).

Results

The two-step method of circumferential EMR [16] for the prevention of postoperative stricture was not performed in any patient; instead all lesions were resected in an en bloc fashion. Eleven patients underwent between one and 19 sessions of balloon dilatation over a period of 0 – 47.3 months. Among them, four patients (36.4 %) required fewer than five sessions to improve their dysphagia. By contrast, four patients (36.4 %) required more than 15 sessions. The longest interval between sessions was approximately 12 months, in a patient who experienced exacerbation of dysphagia 12 months after the previous session. For this reason, all patients were under careful observation.

Univariate analysis

There were significant differences between the two groups in five variables: longitudinal diameter (P = 0.0062) and circumferential diameter of the resected specimen (P = 0.0020), circumferential extent of the lesion (P < 0.0001), histologic depth (P = 0.0002), and procedure time (P = 0.0069) ([Table 1]).

There were no significant differences in age, sex, location of lesion, macroscopic type, electrosurgical knife, injection agent, and operator.

Logistic multivariate analysis

A stepwise logistic regression analysis using the five possible variables derived by univariate analysis revealed two possible risk factors for postoperative strictures: lesion extending to more than three-quarters of the circumferential area and histologic depth to m2 ([Table 2]).

Consequently, all patients were allocated into one of four categories by the combination of these two risk factors, and incidence ratios of postoperative stricture for each category were calculated ([Table 3]).

These analyses revealed that patients with lesions in more than three-quarters of the circumferential area were associated with a high rate of postoperative stricture. By contrast, patients with HGIN lesions in less than three-quarters of the circumferential area have no probability of postoperative strictures. For further evaluations of predictors, we performed subanalysis for patients with m2 lesions in less than three-quarters of the circumferential extension.

Subanalysis for patients with m2 lesions in less than three quarters circumferential extension

Univariate analysis was performed for 19 patients including six patients who experienced postoperative stricture. Among six categorical variables (excluding circumferential extension and histologic depth), and four continuous variables, circumferential diameter differed significantly between the two groups ([Table 4]).

The area under the receiver operating characteristic (ROC) curve for circumferential diameter was 0.79, indicating a good prediction of postoperative stricture ([Fig. 2]).

A cut-off value of 30 mm had a sensitivity of 100 % and a specificity of 54 % for predicting postoperative stricture occurrence.

Discussion

Postoperative stricture after esophageal ESD is a major complication in long-term follow-up. Esophageal stricture can evoke severe dysphagia, and result in a decrease in quality of life or, occasionally, aspiration pneumonia. Our previous study reported that 90 % of patients with lesions in more than three-quarters of the circumferential extension experienced postoperative stricture [6]. For this reason, many endoscopists perform preventive balloon dilatations for such patients. Because there are no solid criteria for the use of preventive balloon dilatations, many endoscopists may make the decision to perform preventive balloon dilatation based only on vague impressions of endoscopic findings.

Our study revealed that circumferential extension and histologic depth of lesion can be reliable independent predictors for postoperative stricture. Additionally, circumferential diameter of the resected specimen is a useful predictor for patients with m2 lesions in less than three-quarters of the circumferential area. We can easily understand that resection of large-sized lesions results in postoperative stricture, but it is not known why a difference in histologic depth between intraepithelial neoplasms and microinvasive carcinoma affects postoperative stenosis. Further studies are needed to elucidate the mechanisms involved including a possible immunological response, and pre-existing fibrous cells and/or chemical factors in the submucosal layer; however, there is still a possibility that histologic depth is confounded by an as yet unknown important predictor.

In this study, we can propose a possible flow chart to predict postoperative stricture based on our results ([Fig. 3]).

The advantage of this flow chart is that the indication for preventive balloon dilatation is obtained based only on perioperative information, and consequently we can start preventive dilatation immediately after ESD.

Giovannini et al. reported the efficacy of the two-step method of circumferential endoscopic resection to prevent postoperative stricture [16]. This method can be one option for the treatment of large-sized lesions to prevent postoperative stricture. However, from the viewpoint of curability of malignancy, it is desirable to resect large-sized lesions in an en bloc fashion because the lateral spread of squamous cell carcinoma is not accurately assessed with reconstruction after piecemeal resection [11] [17].

A limitation of our study may be its retrospective design in a single institute with a limited number of patients. However, given that there is no current definite predictor of postoperative stenosis for esophageal ESD, this study may be quite helpful to endoscopists when deciding whether or not to perform preventive dilatations for postoperative stricture, although the predictive score should be confirmed by a prospective study in the near future.

Competing interests: None

References

• 1 Ohkuwa M, Hosokawa K, Boku N. et al . New endoscopic treatment for intramucosal gastric tumors using an insulated-tip diathermic knife.  Endoscopy. 2001;  33 221-226
  Thieme ConnectPubMedSearch in Google Scholar
• 2 Yamamoto H, Kawata H, Sunada K. et al . Success rate of curative endoscopic mucosal resection with circumferential mucosal incision assisted by submucosal injection of sodium hyaluronate.  Gastrointest Endosc. 2002;  56 507-512
  CrossrefPubMedSearch in Google Scholar
• 3 Yahagi N, Fujishiro M, Kakushima N. et al . Endoscopic submucosal dissection for early gastric cancer using the tip of an electrosurgical snare (thin type).  Dig Endosc. 2004;  16 34-38
  CrossrefPubMedSearch in Google Scholar
• 4 Oyama T, Tomori A, Hotta K. et al . Endoscopic submucosal dissection of early esophageal cancer.  Clin Gastroenterol Hepatol. 2005;  3 67-70
  CrossrefPubMedSearch in Google Scholar
• 5 Fujishiro M, Yahagi N, Kakushima N. et al . Endoscopic submucosal dissection of esophageal squamous cell neoplasms.  Clin Gastroenterol Hepatol. 2006;  4 688-694
  CrossrefPubMedSearch in Google Scholar
• 6 Ono S, Kodashima S, Fujishiro M. et al . Clinical outcomes of endoscopic submucosal dissection for esophageal squamous cell carcinoma (in Japanese).  Stomach Intestine. 2009;  44 (in press)
  PubMedSearch in Google Scholar
• 7 Kato H, Tachimori Y, Watanabe H. et al . Superficial esophageal carcinoma. Surgical treatment and the results.  Cancer. 1990;  66 2319-2323
  CrossrefPubMedSearch in Google Scholar
• 8 Kato H, Tachimori Y, Mizobuchi S. et al . Cervical, mediastinal, and abdominal lymph node dissection (three-field dissection) for superficial carcinoma of the thoracic esophagus.  Cancer. 1993;  72 2879-2882
  CrossrefPubMedSearch in Google Scholar
• 9 Roth J A, Putnam Jr. J B. Surgery for cancer of the esophagus.  Semin Oncol. 1994;  21 453-461
  PubMedSearch in Google Scholar
• 10 Katada C, Muto M, Manabe T. et al . Local recurrence of squamous-cell carcinoma of the esophagus after EMR.  Gastrointest Endosc. 2005;  61 219-225
  CrossrefPubMedSearch in Google Scholar
• 11 Shimura T, Sasaki M, Kataoka H. et al . Advantages of endoscopic submucosal dissection over conventional endoscopic mucosal resection.  J Gastroenterol Hepatol. 2007;  22 821-826
  CrossrefPubMedSearch in Google Scholar
• 12 Mandard A M, Tourneux J, Gignoux M. et al . In situ carcinoma of the esophagus. Macroscopic study with particular reference to the Lugol test.  Endoscopy. 1980;  12 51-57
  Thieme ConnectPubMedSearch in Google Scholar
• 13 Yoshida T, Inoue H, Usui S. et al . Narrow-band imaging system with magnifying endoscopy for superficial esophageal lesions.  Gastrointest Endosc. 2004;  59 288-295
  CrossrefPubMedSearch in Google Scholar
• 14 The Paris endoscopic classification of superficial neoplastic lesions: esophagus, stomach, and colon: November 30 to December 1, 2002.  Gastrointest Endosc. 2003;  58 3-43
  CrossrefPubMedSearch in Google Scholar
• 15 Schlemper R J, Riddell R H, Kato Y. et al . The Vienna classification of gastrointestinal epithelial neoplasia.  Gut. 2000;  47 251-255
  CrossrefPubMedSearch in Google Scholar
• 16 Giovannini M, Bories E, Pesenti C. et al . Circumferential endoscopic mucosal resection in Barrett"s esophagus with high-grade intraepithelial neoplasia or mucosal cancer. Preliminary results in 21 patients.  Endoscopy. 2004;  36 782-787
  Thieme ConnectPubMedSearch in Google Scholar
• 17 Lambert R. Treatment of esophagogastric tumors.  Endoscopy. 2003;  35 118-126
  Thieme ConnectPubMedSearch in Google Scholar

M. FujishiroMD 

Department of Gastroenterology
Graduate School of Medicine
University of Tokyo

7-3-1, Hongo
Bunkyo
Tokyo
Japan

Fax: +81-3-58008806

Email: mtfujish-kkr@umin.ac.jp

References

• 1 Ohkuwa M, Hosokawa K, Boku N. et al . New endoscopic treatment for intramucosal gastric tumors using an insulated-tip diathermic knife.  Endoscopy. 2001;  33 221-226
  Thieme ConnectPubMedSearch in Google Scholar
• 2 Yamamoto H, Kawata H, Sunada K. et al . Success rate of curative endoscopic mucosal resection with circumferential mucosal incision assisted by submucosal injection of sodium hyaluronate.  Gastrointest Endosc. 2002;  56 507-512
  CrossrefPubMedSearch in Google Scholar
• 3 Yahagi N, Fujishiro M, Kakushima N. et al . Endoscopic submucosal dissection for early gastric cancer using the tip of an electrosurgical snare (thin type).  Dig Endosc. 2004;  16 34-38
  CrossrefPubMedSearch in Google Scholar
• 4 Oyama T, Tomori A, Hotta K. et al . Endoscopic submucosal dissection of early esophageal cancer.  Clin Gastroenterol Hepatol. 2005;  3 67-70
  CrossrefPubMedSearch in Google Scholar
• 5 Fujishiro M, Yahagi N, Kakushima N. et al . Endoscopic submucosal dissection of esophageal squamous cell neoplasms.  Clin Gastroenterol Hepatol. 2006;  4 688-694
  CrossrefPubMedSearch in Google Scholar
• 6 Ono S, Kodashima S, Fujishiro M. et al . Clinical outcomes of endoscopic submucosal dissection for esophageal squamous cell carcinoma (in Japanese).  Stomach Intestine. 2009;  44 (in press)
  PubMedSearch in Google Scholar
• 7 Kato H, Tachimori Y, Watanabe H. et al . Superficial esophageal carcinoma. Surgical treatment and the results.  Cancer. 1990;  66 2319-2323
  CrossrefPubMedSearch in Google Scholar
• 8 Kato H, Tachimori Y, Mizobuchi S. et al . Cervical, mediastinal, and abdominal lymph node dissection (three-field dissection) for superficial carcinoma of the thoracic esophagus.  Cancer. 1993;  72 2879-2882
  CrossrefPubMedSearch in Google Scholar
• 9 Roth J A, Putnam Jr. J B. Surgery for cancer of the esophagus.  Semin Oncol. 1994;  21 453-461
  PubMedSearch in Google Scholar
• 10 Katada C, Muto M, Manabe T. et al . Local recurrence of squamous-cell carcinoma of the esophagus after EMR.  Gastrointest Endosc. 2005;  61 219-225
  CrossrefPubMedSearch in Google Scholar
• 11 Shimura T, Sasaki M, Kataoka H. et al . Advantages of endoscopic submucosal dissection over conventional endoscopic mucosal resection.  J Gastroenterol Hepatol. 2007;  22 821-826
  CrossrefPubMedSearch in Google Scholar
• 12 Mandard A M, Tourneux J, Gignoux M. et al . In situ carcinoma of the esophagus. Macroscopic study with particular reference to the Lugol test.  Endoscopy. 1980;  12 51-57
  Thieme ConnectPubMedSearch in Google Scholar
• 13 Yoshida T, Inoue H, Usui S. et al . Narrow-band imaging system with magnifying endoscopy for superficial esophageal lesions.  Gastrointest Endosc. 2004;  59 288-295
  CrossrefPubMedSearch in Google Scholar
• 14 The Paris endoscopic classification of superficial neoplastic lesions: esophagus, stomach, and colon: November 30 to December 1, 2002.  Gastrointest Endosc. 2003;  58 3-43
  CrossrefPubMedSearch in Google Scholar
• 15 Schlemper R J, Riddell R H, Kato Y. et al . The Vienna classification of gastrointestinal epithelial neoplasia.  Gut. 2000;  47 251-255
  CrossrefPubMedSearch in Google Scholar
• 16 Giovannini M, Bories E, Pesenti C. et al . Circumferential endoscopic mucosal resection in Barrett"s esophagus with high-grade intraepithelial neoplasia or mucosal cancer. Preliminary results in 21 patients.  Endoscopy. 2004;  36 782-787
  Thieme ConnectPubMedSearch in Google Scholar
• 17 Lambert R. Treatment of esophagogastric tumors.  Endoscopy. 2003;  35 118-126
  Thieme ConnectPubMedSearch in Google Scholar

M. FujishiroMD 

Department of Gastroenterology
Graduate School of Medicine
University of Tokyo

7-3-1, Hongo
Bunkyo
Tokyo
Japan

Fax: +81-3-58008806

Email: mtfujish-kkr@umin.ac.jp