- Email: [email protected]
Clinical outcomes of endoscopic submucosal dissection for early stage esophagogastric junction cancer: A systematic review and meta-analysis
Digestive and Liver Disease 47 (2015) 37–44
Contents lists available at ScienceDirect
Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld
Digestive Endoscopy
Clinical outcomes of endoscopic submucosal dissection for early stage esophagogastric junction cancer: A systematic review and meta-analysis Chan Hyuk Park a , Eun Hye Kim a , Ha Yan Kim b , Yun Ho Roh b , Yong Chan Lee a,∗ a Division of Gastroenterology, Department of Internal Medicine, Severance Hospital, Institute of Gastroenterology, Yonsei University College of Medicine, Seoul, Republic of Korea b Biostatistics Collaboration Unit, Yonsei University College of Medicine, Seoul, Republic of Korea
a r t i c l e
i n f o
Article history: Received 6 July 2014 Accepted 10 October 2014 Available online 12 November 2014 Keywords: Barrett’s adenocarcinoma Endoscopic submucosal dissection Esophagogastric junction cancer Gastric cancer Meta-analysis
a b s t r a c t Background: Although endoscopic submucosal dissection has become widely used for the management of selected cases of early gastric cancer, the effects of endoscopic submucosal dissection for esophagogastric junction cancer have not been fully evaluated. Methods: Medline, Embase, and the Cochrane Library were searched using the primary keywords “endoscopic submucosal dissection,” “ESD,” “endoscopic resection,” “esophagogastric junction,” “gastroesophageal junction,” and “Barrett.” Six short-term clinical outcomes and three long-term oncologic outcomes were extracted. Results: A total of six studies provided data on 359 early stage esophagogastric junction cancers treated with endoscopic submucosal dissection. The pooled estimate of en bloc resection and complete resection was 98.6% (95% confidence interval 95.9–99.6%) and 87.0% (95% confidence interval 79.7–92.0%), respectively. The pooled estimate of stenosis was 6.9% (95% confidence interval 3.2–14.0%). In 269 lesions with curative resection, there was no local recurrence or distant metastasis. In contrast, three local recurrences and two distant metastases occurred in 90 lesions with non-curative resection. Conclusions: Endoscopic submucosal dissection for early stage esophagogastric junction cancer is a feasible treatment option with high en bloc resection, complete resection rates and an acceptable range of complications. When curative resections are achieved, good oncologic outcomes are likely in the management of early stage esophagogastric junction cancer by endoscopic submucosal dissection. © 2014 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
1. Introduction Endoscopic submucosal dissection (ESD) has been accepted as a standard treatment for differentiated gastric adenocarcinoma without ulceration in which the depth of invasion is clinically diagnosed as T1a and the diameter is ≤2 cm [1]. In addition, Gotoda et al. proposed criteria that suggest a low-risk for lymph node metastasis in gastric cancer [2]. Based on these observations, an expanded indication for gastric ESD was suggested. The expansion of indication has allowed many patients with early gastric cancer (EGC) to undergo ESD rather than surgery. Many reports have shown excellent oncologic outcomes of ESD in patients with EGC that
∗ Corresponding author at: Division of Gastroenterology, Department of Internal Medicine, Severance Hospital, Institute of Gastroenterology, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemun-gu, Seoul 120-752, Republic of Korea. Tel.: +82 2 2228 1960; fax: +82 2 393 6884. E-mail address: [email protected] (Y.C. Lee).
fulfilled expanded indication criteria [3–7]. ESD especially has advantages in patients with proximal gastric cancer, as these patients would usually undergo total gastrectomy if surgery was to be performed. The ESD method can preserve the whole stomach and provide better quality of life compared with surgery. Two questions have been raised regarding ESD in patients with esophagogastric junction (EGJ) cancer, which is a tumour occurring at the mucosa between the lower oesophagus and cardia. The first question concerns curative resection criteria of ESD for EGJ cancer. EGJ cancer may have clinicopathological characteristics of both esophageal and gastric malignancies [8–10]. Due to differences in the curative resection criteria for ESD were different between esophageal cancer and gastric cancer, the concern is which criteria should be applied. Generally, ESD indication for esophageal cancer is stricter than for gastric cancer. While ESD indications for esophageal cancer include differentiated mucosal cancer of <20 mm without lymphovascular invasion [11], those for gastric cancer have been expanded by Gotoda et al. [2] as follows: (a) differentiated intramucosal adenocarcinoma smaller than
http://dx.doi.org/10.1016/j.dld.2014.10.011 1590-8658/© 2014 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
38
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
3 cm in diameter without lymphovascular invasion, irrespective of ulcer findings; (b) differentiated intramucosal adenocarcinoma without lymphovascular invasion and negative for ulceration, irrespective of tumour size; (c) undifferentiated intramucosal cancer smaller than 2 cm without lymphovascular invasion and ulcer findings; and (d) differentiated adenocarcinomas which are smaller than 3 cm with minimal submucosal invasion (<500 m) and without lymphovascular invasion. Oncologic outcomes of patients who underwent ESD for early stage EGJ cancer that fulfilled the ESD indication for gastric cancer but did not fulfil the ESD indication for esophageal cancer have not been fully evaluated. Although several studies have reported treatment outcomes of ESD for early stage EGJ cancer, all of these studies were retrospective, singlecentre studies that included only a relatively small number of lesions [12–17]. The second question concerns short-term clinical outcomes, including complications of ESD for EGJ cancer. Difficult location for ESD in early stage EGJ cancer may affect complete resection and curative resection, which are related to long-term oncologic outcomes. Major complications of ESD, such as postoperative bleeding and perforation, may be influenced by the difficulty of the procedure. Stenosis at the post-ESD site is also a concern after ESD for EGJ cancer. To answer the above questions and provide more reliable estimates of the efficacy of ESD for EGJ cancer, we performed a systematic review and meta-analysis. This study was written in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) recommendations [18]. 2. Methods 2.1. Search strategy We searched for all relevant studies that performed ESD for EGJ cancer published between January 1990 and March 2014 through MEDLINE, EMBASE, and the Cochrane Library. The medical terms “endoscopic submucosal dissection,” “ESD,” “endoscopic resection,” “esophagogastric junction,” “gastroesophageal junction,” and “Barrett” were used in the search (Appendix 1). We also searched the references of screened articles to identify any additional studies. All human studies published in English were considered. The latest date for updating the search was March 24, 2014. 2.2. Study selection In the first stage of the study selection, titles and abstracts of papers searched by keywords were examined to exclude irrelevant articles. Next, the full text of all selected studies was screened according to inclusion and exclusion criteria. The inclusion criteria specified (1) studies about ESD for EGJ cancer, and (2) studies reporting at least one of the appropriate clinical outcomes, including en bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, stenosis, local recurrence, distant metastasis, and cause-specific death. The exclusion criteria ruled out (1) studies including participants without cancer, such as adenoma, (2) studies including participants who underwent endoscopic mucosal resection rather than ESD (3) non-original articles, (4) animal testing or preclinical trials, (5) abstract-only publications or unpublished studies, and (6) publications in a language other than English. Two investigators (C.H.P. and E.H.K.) independently evaluated studies for eligibility and subsequently resolved any disagreements through discussion and consensus. If no agreement could be reached, a third investigator (Y.C.L.) decided eligibility. To understand the risk of bias in individual studies, a formal quality assessment of studies was performed, along with subgroup
analysis. The methodological quality of observational studies was independently assessed by two investigators (C.H.P. and E.H.K.), using the Newcastle-Ottawa scale [19,20]. Using this scale, observational studies were scored across three categories: selection (4 questions), comparability of study groups (2 questions), and ascertainment of the exposure or outcome (3 questions). All questions were assigned a score of one point, with the exception of comparability of study groups, in which a maximum of two points were awarded. In addition, a subgroup analysis was performed after excluding studies that included patients who had undergone ESD during the initial introduction period of ESD in Japan, because ESD requires a high level of endoscopic skill to achieve desirable oncologic outcomes and has an obvious learning curve [2,21,22]. 2.3. Data extraction Using a data extraction form developed in advance, two reviewers independently extracted the following information: first author, year of publication, study design, institution, country, enrollment period, number of patients and lesions, age and sex of patients, Siewert classification of lesions [23], histologic type of lesions (Barrett’s adenocarcinoma or gastric cardia cancer), tumour size, deep submucosal invasion, lymphovascular invasion, device for ESD, result of main outcomes, follow-up period, and criteria for curative resection. Main outcomes consisted of six short-term clinical outcomes, including en bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, and stenosis, and three long-term oncologic outcomes, including local recurrence, distant metastasis, and cause-specific death. 2.4. Outcomes assessed Our primary analysis focused on assessing the clinical outcomes and long-term oncologic outcomes of ESD for patients with early stage EGJ. Our a priori hypothesis posited that the procedures performed in the initial introduction period of ESD would be indicative of heterogeneity of the main outcomes. 2.5. Statistical analysis Meta-analyses were performed by calculating pooled estimates of clinical outcomes, including complications. Taking a conservative approach, we used a random effects model [24], producing wider confidence intervals than a fixed effect model would produce. In all meta-analyses, we assessed heterogeneity with Cochran’s Q test and the inconsistency index (I2 ) [25]. If heterogeneity was found, we conducted sensitivity analyses. Statistical analyses were conducted using the statistical software Comprehensive Meta Analysis (version 2.2.064; Biostat Inc., Englewood, NJ, USA). 3. Results 3.1. Study selection A flow diagram of this systematic review is shown in Fig. 1. In summary, 172 studies were identified by our literature search. After scanning titles and abstracts, we discarded 58 identical articles retrieved through multiple search engines. Additionally 97 irrelevant articles were excluded based on the titles and abstracts. Full texts of the 17 remaining articles were reviewed, and ten non-pertinent articles were excluded. In addition, one study was excluded due to cohort overlap between studies. Finally, six studies were appropriate for the meta-analysis. These studies described a total of 353 patients with 359 lesions. All studies had been published in the past 4 years, from 2010 to 2014.
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
39
Fig. 1. Flow diagram of studies included in this meta-analysis. ESD, endoscopic submucosal dissection
The enrollment period ranged from 2000 to 2013. In addition, all were single-centre, retrospective studies conducted in Japan. Eight or nine main outcomes were reported in each study among nine possible outcomes. One study by Hoteya et al. [13], which was the only comparative study between EGJ cancer and non-EGJ cancer, had a quality score of 9, whereas other studies had a quality score of 6. The characteristics of the studies included in the meta-analysis are summarized in Table 1. 3.2. Patient and lesion characteristics and ESD procedure Patient and lesion characteristics of the included studies are shown in Table 2. The median of the mean ages across the included studies was 68.7 years (range, 63.0–72.0 years). The median of the male sex rates was 86.3% (range, 79.3–91.3%). All studies included only Siewert type II EGJ cancers with the exception of one study by Kagemoto et al. that did not provide information on Siewert classification [17]. The median of the rates of deep submucosal invasion (>500 m from the muscularis mucosa) was 22.0% (range, 8.6–24.5%). Among the included studies, an insulation-tipped knife was the preferred knife in five of the six studies. In addition, the Hook knife, Flex knife, and Sumitomo Bakelite knife were also used for ESD. The median of the mean procedure times of the included studies was 82.0 min (range, 42.5–102.7 min).
resection rates was 98.6% (95% CI 95.9–99.6%) using the random effects model (Fig. 2A). No heterogeneity was identified (P = 0.982, I2 < 0.001). All studies, including 359 lesions, reported complete resection rates and curative resection rates [12–17]. The pooled estimate of complete resection rates was 87.0% (95% CI 79.7–92.0%; Fig. 2B). In addition, a moderate heterogeneity was identified (P = 0.039, I2 = 57.226). In cases of curative resection rates, the pooled estimate was 74.6% (95% CI 69.8–78.9%; Fig. 2C). No evidence of heterogeneity (P = 0.526, I2 < 0.001) was identified. 3.4. Complications All studies, including 359 lesions, reported post-ESD bleeding rates and perforation rates. The pooled estimate of post-ESD bleeding rates was 3.4% (95% CI 1.7–6.7%; Fig. 3A). No heterogeneity was identified among the studies (P = 0.382, I2 = 5.442). The pooled estimate of perforation rates was 2.5% (95% CI 1.3–5.0%; Fig. 3B). Heterogeneity was not observed (P = 0.852, I2 < 0.001). Stenosis rate was reported in five studies, incorporating 231 lesions [12,14–17]. The pooled estimate of stenosis rates was 6.9% (95% CI 3.2–14.0%; Fig. 3C). No significant heterogeneity (P = 0.144, I2 = 41.567) was observed. All of 13 patients with stenosis were treated successfully with endoscopic balloon dilatation. 3.5. Long-term oncologic outcomes
3.3. Procedural outcomes Five studies, including 231 lesions, provided information on en bloc resection rates [12,14–17]. The pooled estimate of en bloc
All six studies, including 359 lesions, reported local recurrence rate, distant metastasis rate, and cause-specific mortality (Table 3) [12–17]. The median of the median follow-up periods across the
40
Table 1 Characteristics of studies included in the meta-analysis. Author
Year of publication
Study design
2010
Retrospective, single-centre
Hoteya et al. [13]
2013
Retrospective, single-centre
Omae et al. [14]
2013
Retrospective, single-centre
Yamada et al. [15]
2013
Retrospective, single-centre
Imai et al. [16]
2013
Retrospective, single-centre
Kagemoto et al. [17]
2014
Retrospective, single-centre
Enrollment period
Patient
Lesion
Yokohama City University Medical Center Toranomon Hospital
Japan
2000–2009
58
58
Japan
2005–2010
126
128
Cancer Institute Hospital National Cancer Center Hospital Shizuoka Cancer Center
Japan
2004–2010
44
44
Japan
2001–2007
53
53
Japan
2002–2009
49
50
Japan
2006–2013
23
26
Hiroshima University Hospital
Main outcome
Study quality
En-bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, stenosis, recurrence, mortality Complete resection, curative resection, post-ESD bleeding, perforation, recurrence, mortality En-bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, stenosis, recurrence, mortality En-bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, stenosis, recurrence, mortality En-bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, stenosis, recurrence, mortality En-bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, stenosis, recurrence
Selection
Comparability
Exposure/ outcome
***
–
***
****
**
***
***
–
***
***
–
***
***
–
***
***
–
***
Asterisk represents quality scores based on the Newcastle-Ottawa scale. ESD, endoscopic submucosal dissection.
Table 2 Patient and lesion characteristics of the included studies. Author
Age, year
Male (%)
Siewart classification
Histologic type of lesion
Lesion size (mm)
Deep submucosal invasiona (%)
LVI (%)
Drug for sedation
Lifting solution
Device for ESD
Procedure time (min)
Hirasawa et al. [12]
Mean 69.3 (range 36–85) BA: mean 63.5 (SD 12.5) GC: mean 68.8 (SD 9.3) Mean 70.0 (range 42–84) Median 68 (range 39–88) Median 72 (range 41–89) Mean 63 (SD 10)
79.3
Type II
BA: 1, GC: 57
8.6
6.9
Type II
BA: 25, GC: 103
N/A
BA: 24.0 GC: 9.7
Diazepam or midazolam Diazepam
Epinephrine + saline Glycerol + epinephrine
IT-knife
83.6
90.9
Type II
BA: 13, GC: 31
11.4
2.3
N/A
Type II
BA: 28, GC: 25
24.5
7.5
N/A
Glycerol or hyaluronic acid N/A
Mean 82 (range 22–275) BA: mean 84.9 (SD 52.3) GC: mean 107.0 (SD 64.4) N/A
84.9
Mean 20.3 (range 3–50) BA: mean 20.2 (SD 17.6) GC: mean 21.8 (SD 15.8) Mean 17 (range 5–47) Median 20 (3–47)
IT-knife
87.8
Type II
BA: 7, GC: 43
22.0
10.0
Midazolam
N/A
IT-knife
91.3
N/A
BA: 26
23.1
0.0
N/A
Glycerol + epinephrine
Hook knife, IT-knife nano, SB knife
Hoteya et al. [13]
Omae et al. [14] Yamada et al. [15] Imai et al. [16] Kagemoto et al. [17]
Median 22 (range 5–70) Mean 19.0 (SD 13.6)
Hook knife, Flex knife
IT-knife
Median 70 (range 20–480) Median 42.5 (range 10–157) Mean 93 (SD 48)
LVI, lymphovascular invasion; ESD, endoscopic submucosal dissection; BA, Barrett’s adenocarcinoma; GC, gastric cardia cancer; IT, insulation-tipped; SB, Sumitomo Bakelite; SD, standard deviation; N/A, not available a SM2 invasion (>500 m from the muscularis mucosa) or more.
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
Hirasawa et al. [12]
Country
Institution
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
41
Fig. 2. Forest plots of the included studies analyzing the procedural outcomes including the en bloc resection rates (A), the complete resection rates (B), and the curative resection rates (C). CI, confidence interval
included studies was 35.8 months (range, 33.0–73.2 months). In all but one study [12–16], curative resection was determined according to the curative resection criteria for gastric cancer. One remaining study used the curative resection criteria for esophageal cancer for determining curative resection [17]. In 269 lesions with curative resection, no local recurrence or distant metastasis was seen. On the contrary, three local recurrences and two distant metastases occurred in 90 lesions with non-curative resection. Three cause-specific deaths were identified in a total of 359 lesions (pooled estimate 1.7% [95% CI 0.7–4.0%]; P-value for heterogeneity = 0.791; I2 < 0.001). All cause-specific deaths occurred in patients with non-curative resection. 3.6. Subgroup analysis In the subgroup analysis, after excluding two studies that included patients who had undergone ESD during the initial introduction period of ESD in Japan in 2000–2001 [12,15], four studies described a total of 242 patients with 248 lesions [13,14,16,17]. The pooled estimates of en bloc resection, complete resection, and curative resection were 98.5% (95% CI 95.0–99.6%), 90.9% (95% CI
84.6–94.7%), and 75.1% (95% CI 69.3–80.1%), respectively. No heterogeneity was identified (en bloc resection: P = 0.957, I2 < 0.001; complete resection: P = 0.253, I2 = 26.425; curative resection: P = 0.528, I2 < 0.001). The pooled estimates of bleeding rates, perforation rates, and stenosis rates were 2.9% (95% CI 1.0–7.7%), 2.4% (95% CI 1.1–5.5%), and 7.3% (95% CI 2.3–20.8%), respectively. No significant heterogeneity was identified (bleeding: P = 0.276, I2 = 22.532; perforation: P = 0.794, I2 < 0.001; stenosis: P = 0.127, I2 = 51.519). In the subgroup, distant metastasis was identified in two patients with non-curative resection. 4. Discussion This systematic review and meta-analysis showed that ESD for early stage EGJ cancer is a feasible treatment option with excellent en bloc resection rates. Generally, ESD has an advantage over endoscopic mucosal resection in terms of achievement of en bloc resection regardless of lesion size [26,27]. In this study, the en bloc resection rate of ESD for early stage EGJ cancer was almost 100%. Considering that complete resection rates of ESD for EGC are usually 90–95% [4–7], the complete resection rate of ESD for early stage EGJ cancer of 87.0% (95% CI 79.0–92.0%) in this meta-analysis is a very reasonable rate.
42
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
Fig. 3. Forest plot of the included studies analyzing the complications, including the post-ESD bleeding rates (A), the perforation rates (B), and the stenosis rates (C). ESD, endoscopic submucosal dissection; CI, confidence interval
Although the en bloc resection rate and complete resection rate of ESD for early stage EGJ cancer were acceptable, the learning curve for ESD was a concern before conducting the meta-analysis [22]. Since the first endoscopic resection using an insulation-tipped knife was introduced in 1999 [28], many improvements in the instruments used for ESD have been made. To minimize the effect of the learning curve, we conducted a subgroup analysis after excluding studies that included patients who had undergone ESD during the initial introduction period in Japan. While the pooled estimate of complete resection rates in all of the six included studies was 87.0% (95% CI 79.7–92.0%), the same estimate in the subgroup analysis was 90.9% (95% CI 84.6–94.7%). Furthermore, previously observed moderate heterogeneity for complete resection rate among the studies was removed in the subgroup analysis. We believe these results imply a learning curve effect in ESD for EGJ cancer. It is well known that the difficulty of performing ESD depends on the lesion location (especially the upper third of the stomach), and complete resection can be hindered by such location-related difficulties [29,30]. In contrast to a high en bloc resection rate and complete resection rate, the curative resection rate of ESD in early stage EGJ cancer seems to be lower than that of ESD in gastric
cancer [4,31]. The low curative resection rate of ESD for EGJ cancer in this meta-analysis, however, is caused by heterogeneity in the study population. All studies in the meta-analysis included lesions that did not fulfil both absolute and expanded indications while other previous studies that reported a curative resection rate of ESD in gastric cancer included only lesions that met ESD indication criteria. For example, the median rate of deep submucosal invasion (>500 m from the muscularis mucosa) across the studies included in the meta-analysis was 22.0% (range, 8.6–24.5%). In other words, about 20% of the lesions included in the meta-analysis could not achieve curative resection regardless of ESD outcomes. Considering the difference among study populations, it is not clear whether the 74.6% of the pooled estimate of curative resection rates in EGJ cancer is inferior to that of the curative resection rate in gastric cancer. Due to the difficult location for ESD in early stage EGJ cancer, we expected that post-ESD bleeding and perforation would increase in EGJ cancer compared with gastric cancer. Pooled estimates of post-ESD bleeding rates and perforation rates in EGJ cancer, however, were comparable to the outcomes in gastric cancer [21,32]. Judging from procedural outcomes and complications, including
3/359 2/90 3/90 0/269 0/269
This value included the follow-up period of patients with both junctional- and non-junctional gastric cancer in the study.
Total
a
0/58 1/128 0/44 2/53 0/50 0/26 0/12 1/33 0/7 0/17 1/14 0/7 0/12 0/33 0/7 3/17 0/14 0/7 0/46 0/95 0/37 0/36 0/36 0/19
Local recurrence
0/46 0/95 0/37 0/36 0/36 0/19 Gastric cancer Gastric cancer Gastric cancer Gastric cancer Gastric cancer Esophageal cancer Hirasawa et al. [12] Hoteya et al. [13] Omae et al. [14] Yamada et al. [15] Imai et al. [16] Kagemoto et al. [17]
Non-curative resection
Local recurrence
Distant metastasis Curative resection Curative resection criteria Author
Table 3 Long-term clinical outcomes of endoscopic submucosal dissection for esophagogastric junction cancer.
Distant metastasis
Cause-specific mortality
Periods of follow-up, months
Median 36.6 (range 4–94) Median 34 (range 2–96)a Mean 33 (range 6–64) Median 73.2 (range 16.8–132) Median 47 (range 22–97) Median 35 (range 2–85)
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
43
post-ESD bleeding and perforation, ESD appears to be an acceptable treatment option for EGJ cancer. However, considerable concerns remain about stenosis at postESD sites. The pooled estimate of stenosis rates was highest among the pooled estimates of complications in the meta-analysis. A previous study with a meta-analysis of ESD for superficial squamous esophageal cancer demonstrated that the pooled estimate of stenosis rates was 11.6% (95% CI 8.2–16.2%) [33]. Other studies on pyloric stenosis after ESD reported that stenosis rates were about 3–7% [34,35]. In comparison with stenosis rates after ESD for esophageal cancer and pyloric cancer, we think that the stenosis rate after ESD for EGJ cancer is not significant. Nevertheless, stenosis is obviously a major complication of ESD for EGJ cancer. One of the most interesting points in treating a patient with early stage EGJ cancer after ESD is determining which curative resection criteria should be applied. In the present meta-analysis, five of the six included studies assessed curative resection according to curative resection criteria for gastric cancer. The one remaining study used curative resection criteria for esophageal cancer as a definition of curative resection. All 269 patients with curative resection showed no local recurrence or distant metastasis regardless of curative resection criteria. Although more large-scale data are necessary to form a definite conclusion, we believe that curative resection criteria for gastric cancer can be used for assessing curative resection in EGJ cancer. In contrast to the long-term oncologic outcomes in patients with curative resection, outcomes in patients with non-curative resection were not as good. Among 90 patients with non-curative resection, 3 (3.3%) patients had local recurrence and 2 (2.2%) had distant metastasis. Three of the local recurrences were identified in only one study by Yamada et al., which had the longest follow-up duration among the included studies [15]. The three local recurrences developed at 3.6, 4.6, and 7.4 years after ESD, respectively. Considering that the other five studies had a median follow-up duration of around 3 years, they may have underestimated poor outcomes in patients with non-curative resections. Although this is the first meta-analysis of ESD outcomes in the management of early stage EGJ cancer, this study has several limitations. Most importantly, all of the included studies were performed in Japan, which is the leading country for advanced endoscopic procedures, and Japanese endoscopists have extensive expertise in therapeutic endoscopy, including ESD. Therefore, our results may not be generalizable to Western countries. Second, all studies included in the meta-analysis had a single-centre, retrospective design. Prospectively collected data in each institution, however, would minimize selection bias. Third, most of the included studies were observational studies on EGJ cancer without group comparisons. Therefore, ESD outcomes in the present meta-analysis could not be compared directly with those for other cancers, including gastric cancer. Despite these limitations, we believe that our data form the basis for understanding clinical outcomes of ESD for EGJ cancer. A high en bloc resection rate and complete resection rate were seen after ESD for EGJ cancer. Although about 7% of lesions induced stenosis after ESD, other major complications, including post-ESD bleeding and perforation, were in the reasonably acceptable range. When curative resections are achieved, good oncologic outcomes are likely in the management of early stage EGJ cancer by ESD. Conflict of interest None declared. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.dld.2014.10.011.
44
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
References [1] Japanese gastric cancer treatment guidelines 2010 (ver. 3). Gastric Cancer 2011;14:113–23. [2] Gotoda T, Yamamoto H, Soetikno RM. Endoscopic submucosal dissection of early gastric cancer. Journal of Gastroenterology 2006;41:929–42. [3] Choi JH, Kim ER, Min BH, et al. The feasibility and safety of the endoscopic submucosal dissection of superficial gastric neoplastic lesions in patients with compensated liver cirrhosis: a retrospective study. Gut Liver 2012;6:58–63. [4] Kim DY, Hong SJ, Cho GS, et al. Long-term efficacy of endoscopic submucosal dissection compared with surgery for early gastric cancer: a retrospective cohort study. Gut Liver 2014;8:519–25. [5] Lee H, Yun WK, Min BH, et al. A feasibility study on the expanded indication for endoscopic submucosal dissection of early gastric cancer. Surgical Endoscopy 2011;25:1985–93. [6] Ahn JY, Jung HY, Choi KD, et al. Endoscopic and oncologic outcomes after endoscopic resection for early gastric cancer: 1370 cases of absolute and extended indications. Gastrointestinal Endoscopy 2011;74:485–93. [7] Park CH, Shin S, Park JC, et al. Long-term outcome of early gastric cancer after endoscopic submucosal dissection: expanded indication is comparable to absolute indication. Digestive and Liver Disease 2013;45:651–6. [8] Suh Y, Han D, Kong S, et al. Should adenocarcinoma of the esophagogastric junction be classified as esophageal cancer? A comparative analysis according to the seventh AJCC TNM classification. Annals of Surgery 2012;255: 908–15. [9] Yamashita H, Katai H. Should adenocarcinoma of the esophagogastric junction be classified as esophageal cancer? Limited to Siewert Type II, Yes. Annals of Surgery 2014, http://dx.doi.org/10.1097/SLA.0000000000000589 [in press]. [10] Pedrazzani C. Should adenocarcinoma of the esophagogastric junction be classified as gastric or esophageal cancer, or else as a distinct clinical entity? Annals of Surgery 2014, http://dx.doi.org/10.1097/SLA.0000000000000524 [in press]. [11] Pech O, May A, Rabenstein T, et al. Endoscopic resection of early oesophageal cancer. Gut 2007;56:1625–34. [12] Hirasawa K, Kokawa A, Oka H, et al. Superficial adenocarcinoma of the esophagogastric junction: long-term results of endoscopic submucosal dissection. Gastrointestinal Endoscopy 2010;72:960–6. [13] Hoteya S, Matsui A, Iizuka T, et al. Comparison of the clinicopathological characteristics and results of endoscopic submucosal dissection for esophagogastric junction and non-junctional cancers. Digestion 2013;87:29–33. [14] Omae M, Fujisaki J, Horiuchi Y, et al. Safety, efficacy, and long-term outcomes for endoscopic submucosal dissection of early esophagogastric junction cancer. Gastric Cancer 2013;16:147–54. [15] Yamada M, Oda I, Nonaka S, et al. Long-term outcome of endoscopic resection of superficial adenocarcinoma of the esophagogastric junction. Endoscopy 2013;45:992–6. [16] Imai K, Kakushima N, Tanaka M, et al. Validation of the application of the Japanese curative criteria for superficial adenocarcinoma at the esophagogastric junction treated by endoscopic submucosal dissection: a long-term analysis. Surgical Endoscopy 2013;27:2436–45. [17] Kagemoto K, Oka S, Tanaka S, et al. Clinical outcomes of endoscopic submucosal dissection for superficial Barrett’s adenocarcinoma. Gastrointestinal Endoscopy 2014;80:239–45. [18] Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care
[19]
[20]
[21]
[22]
[23]
[24] [25] [26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
interventions: explanation and elaboration. Journal of Clinical Epidemiology 2009;62:e1–34. Wells GA, Shea B, O’Connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses; 2014. Available at: http://www.ohri.ca/programs/clinical epidemiology/oxford.asp [accessed May 25]. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. European Journal of Epidemiology 2010;25:603–5. Akasaka T, Nishida T, Tsutsui S, et al. Short-term outcomes of endoscopic submucosal dissection (ESD) for early gastric neoplasm: multicenter survey by Osaka university ESD study group. Digestive Endoscopy 2011;23:73–7. Hong KH, Shin SJ, Kim JH. Learning curve for endoscopic submucosal dissection of gastric neoplasms. European Journal of Gastroenterology and Hepatology 2014;26:949–54. Siewert J. Carcinoma of the cardia: carcinoma of the gastroesophageal junction – classification, pathology and extent of resection. Diseases of the Esophagus 1996;9:173–82. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Statistics in Medicine 2002;21:1539–58. DerSimonian R, Laird N. Meta-analysis in clinical trials. Controlled Clinical Trials 1986;7:177–88. Park YM, Cho E, Kang HY, et al. The effectiveness and safety of endoscopic submucosal dissection compared with endoscopic mucosal resection for early gastric cancer: a systematic review and metaanalysis. Surgical Endoscopy 2011;25:2666–77. Lian J, Chen S, Zhang Y, et al. A meta-analysis of endoscopic submucosal dissection and EMR for early gastric cancer. Gastrointestinal Endoscopy 2012;76:763–70. Gotoda T, Kondo H, Ono H, et al. A new endoscopic mucosal resection procedure using an insulation-tipped electrosurgical knife for rectal flat lesions: report of two cases. Gastrointestinal Endoscopy 1999;50:560–3. Imagawa A, Okada H, Kawahara Y, et al. Endoscopic submucosal dissection for early gastric cancer: results and degrees of technical difficulty as well as success. Endoscopy 2006;38:987–90. Goto A, Nishikawa J, Okamoto T, et al. Outcomes of endoscopic submucosal dissection for early gastric cancer and factors associated with incomplete resection. Hepato-Gastroenterology 2013;60:46–53. Ohnita K, Isomoto H, Shikuwa S, et al. Early and long-term outcomes of endoscopic submucosal dissection for early gastric cancer in a large patient series. Experimental and Therapeutic Medicine 2014;7:594–8. Matsui N, Akahoshi K, Nakamura K, et al. Endoscopic submucosal dissection for removal of superficial gastrointestinal neoplasms: a technical review. World Journal of Gastrointestinal Endoscopy 2012;4:123–36. Kim JS, Kim BW, Shin IS. Efficacy and safety of endoscopic submucosal dissection for superficial squamous esophageal neoplasia: a meta-analysis. Digestive Diseases and Sciences 2014;59:1862–9. Iizuka H, Kakizaki S, Sohara N, et al. Stricture after endoscopic submucosal dissection for early gastric cancers and adenomas. Digestive Endoscopy 2010;22:282–8. Coda S, Oda I, Gotoda T, et al. Risk factors for cardiac and pyloric stenosis after endoscopic submucosal dissection, and efficacy of endoscopic balloon dilation treatment. Endoscopy 2009;41:421–6.
Contents lists available at ScienceDirect
Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld
Digestive Endoscopy
Clinical outcomes of endoscopic submucosal dissection for early stage esophagogastric junction cancer: A systematic review and meta-analysis Chan Hyuk Park a , Eun Hye Kim a , Ha Yan Kim b , Yun Ho Roh b , Yong Chan Lee a,∗ a Division of Gastroenterology, Department of Internal Medicine, Severance Hospital, Institute of Gastroenterology, Yonsei University College of Medicine, Seoul, Republic of Korea b Biostatistics Collaboration Unit, Yonsei University College of Medicine, Seoul, Republic of Korea
a r t i c l e
i n f o
Article history: Received 6 July 2014 Accepted 10 October 2014 Available online 12 November 2014 Keywords: Barrett’s adenocarcinoma Endoscopic submucosal dissection Esophagogastric junction cancer Gastric cancer Meta-analysis
a b s t r a c t Background: Although endoscopic submucosal dissection has become widely used for the management of selected cases of early gastric cancer, the effects of endoscopic submucosal dissection for esophagogastric junction cancer have not been fully evaluated. Methods: Medline, Embase, and the Cochrane Library were searched using the primary keywords “endoscopic submucosal dissection,” “ESD,” “endoscopic resection,” “esophagogastric junction,” “gastroesophageal junction,” and “Barrett.” Six short-term clinical outcomes and three long-term oncologic outcomes were extracted. Results: A total of six studies provided data on 359 early stage esophagogastric junction cancers treated with endoscopic submucosal dissection. The pooled estimate of en bloc resection and complete resection was 98.6% (95% confidence interval 95.9–99.6%) and 87.0% (95% confidence interval 79.7–92.0%), respectively. The pooled estimate of stenosis was 6.9% (95% confidence interval 3.2–14.0%). In 269 lesions with curative resection, there was no local recurrence or distant metastasis. In contrast, three local recurrences and two distant metastases occurred in 90 lesions with non-curative resection. Conclusions: Endoscopic submucosal dissection for early stage esophagogastric junction cancer is a feasible treatment option with high en bloc resection, complete resection rates and an acceptable range of complications. When curative resections are achieved, good oncologic outcomes are likely in the management of early stage esophagogastric junction cancer by endoscopic submucosal dissection. © 2014 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
1. Introduction Endoscopic submucosal dissection (ESD) has been accepted as a standard treatment for differentiated gastric adenocarcinoma without ulceration in which the depth of invasion is clinically diagnosed as T1a and the diameter is ≤2 cm [1]. In addition, Gotoda et al. proposed criteria that suggest a low-risk for lymph node metastasis in gastric cancer [2]. Based on these observations, an expanded indication for gastric ESD was suggested. The expansion of indication has allowed many patients with early gastric cancer (EGC) to undergo ESD rather than surgery. Many reports have shown excellent oncologic outcomes of ESD in patients with EGC that
∗ Corresponding author at: Division of Gastroenterology, Department of Internal Medicine, Severance Hospital, Institute of Gastroenterology, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemun-gu, Seoul 120-752, Republic of Korea. Tel.: +82 2 2228 1960; fax: +82 2 393 6884. E-mail address: [email protected] (Y.C. Lee).
fulfilled expanded indication criteria [3–7]. ESD especially has advantages in patients with proximal gastric cancer, as these patients would usually undergo total gastrectomy if surgery was to be performed. The ESD method can preserve the whole stomach and provide better quality of life compared with surgery. Two questions have been raised regarding ESD in patients with esophagogastric junction (EGJ) cancer, which is a tumour occurring at the mucosa between the lower oesophagus and cardia. The first question concerns curative resection criteria of ESD for EGJ cancer. EGJ cancer may have clinicopathological characteristics of both esophageal and gastric malignancies [8–10]. Due to differences in the curative resection criteria for ESD were different between esophageal cancer and gastric cancer, the concern is which criteria should be applied. Generally, ESD indication for esophageal cancer is stricter than for gastric cancer. While ESD indications for esophageal cancer include differentiated mucosal cancer of <20 mm without lymphovascular invasion [11], those for gastric cancer have been expanded by Gotoda et al. [2] as follows: (a) differentiated intramucosal adenocarcinoma smaller than
http://dx.doi.org/10.1016/j.dld.2014.10.011 1590-8658/© 2014 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
38
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
3 cm in diameter without lymphovascular invasion, irrespective of ulcer findings; (b) differentiated intramucosal adenocarcinoma without lymphovascular invasion and negative for ulceration, irrespective of tumour size; (c) undifferentiated intramucosal cancer smaller than 2 cm without lymphovascular invasion and ulcer findings; and (d) differentiated adenocarcinomas which are smaller than 3 cm with minimal submucosal invasion (<500 m) and without lymphovascular invasion. Oncologic outcomes of patients who underwent ESD for early stage EGJ cancer that fulfilled the ESD indication for gastric cancer but did not fulfil the ESD indication for esophageal cancer have not been fully evaluated. Although several studies have reported treatment outcomes of ESD for early stage EGJ cancer, all of these studies were retrospective, singlecentre studies that included only a relatively small number of lesions [12–17]. The second question concerns short-term clinical outcomes, including complications of ESD for EGJ cancer. Difficult location for ESD in early stage EGJ cancer may affect complete resection and curative resection, which are related to long-term oncologic outcomes. Major complications of ESD, such as postoperative bleeding and perforation, may be influenced by the difficulty of the procedure. Stenosis at the post-ESD site is also a concern after ESD for EGJ cancer. To answer the above questions and provide more reliable estimates of the efficacy of ESD for EGJ cancer, we performed a systematic review and meta-analysis. This study was written in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) recommendations [18]. 2. Methods 2.1. Search strategy We searched for all relevant studies that performed ESD for EGJ cancer published between January 1990 and March 2014 through MEDLINE, EMBASE, and the Cochrane Library. The medical terms “endoscopic submucosal dissection,” “ESD,” “endoscopic resection,” “esophagogastric junction,” “gastroesophageal junction,” and “Barrett” were used in the search (Appendix 1). We also searched the references of screened articles to identify any additional studies. All human studies published in English were considered. The latest date for updating the search was March 24, 2014. 2.2. Study selection In the first stage of the study selection, titles and abstracts of papers searched by keywords were examined to exclude irrelevant articles. Next, the full text of all selected studies was screened according to inclusion and exclusion criteria. The inclusion criteria specified (1) studies about ESD for EGJ cancer, and (2) studies reporting at least one of the appropriate clinical outcomes, including en bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, stenosis, local recurrence, distant metastasis, and cause-specific death. The exclusion criteria ruled out (1) studies including participants without cancer, such as adenoma, (2) studies including participants who underwent endoscopic mucosal resection rather than ESD (3) non-original articles, (4) animal testing or preclinical trials, (5) abstract-only publications or unpublished studies, and (6) publications in a language other than English. Two investigators (C.H.P. and E.H.K.) independently evaluated studies for eligibility and subsequently resolved any disagreements through discussion and consensus. If no agreement could be reached, a third investigator (Y.C.L.) decided eligibility. To understand the risk of bias in individual studies, a formal quality assessment of studies was performed, along with subgroup
analysis. The methodological quality of observational studies was independently assessed by two investigators (C.H.P. and E.H.K.), using the Newcastle-Ottawa scale [19,20]. Using this scale, observational studies were scored across three categories: selection (4 questions), comparability of study groups (2 questions), and ascertainment of the exposure or outcome (3 questions). All questions were assigned a score of one point, with the exception of comparability of study groups, in which a maximum of two points were awarded. In addition, a subgroup analysis was performed after excluding studies that included patients who had undergone ESD during the initial introduction period of ESD in Japan, because ESD requires a high level of endoscopic skill to achieve desirable oncologic outcomes and has an obvious learning curve [2,21,22]. 2.3. Data extraction Using a data extraction form developed in advance, two reviewers independently extracted the following information: first author, year of publication, study design, institution, country, enrollment period, number of patients and lesions, age and sex of patients, Siewert classification of lesions [23], histologic type of lesions (Barrett’s adenocarcinoma or gastric cardia cancer), tumour size, deep submucosal invasion, lymphovascular invasion, device for ESD, result of main outcomes, follow-up period, and criteria for curative resection. Main outcomes consisted of six short-term clinical outcomes, including en bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, and stenosis, and three long-term oncologic outcomes, including local recurrence, distant metastasis, and cause-specific death. 2.4. Outcomes assessed Our primary analysis focused on assessing the clinical outcomes and long-term oncologic outcomes of ESD for patients with early stage EGJ. Our a priori hypothesis posited that the procedures performed in the initial introduction period of ESD would be indicative of heterogeneity of the main outcomes. 2.5. Statistical analysis Meta-analyses were performed by calculating pooled estimates of clinical outcomes, including complications. Taking a conservative approach, we used a random effects model [24], producing wider confidence intervals than a fixed effect model would produce. In all meta-analyses, we assessed heterogeneity with Cochran’s Q test and the inconsistency index (I2 ) [25]. If heterogeneity was found, we conducted sensitivity analyses. Statistical analyses were conducted using the statistical software Comprehensive Meta Analysis (version 2.2.064; Biostat Inc., Englewood, NJ, USA). 3. Results 3.1. Study selection A flow diagram of this systematic review is shown in Fig. 1. In summary, 172 studies were identified by our literature search. After scanning titles and abstracts, we discarded 58 identical articles retrieved through multiple search engines. Additionally 97 irrelevant articles were excluded based on the titles and abstracts. Full texts of the 17 remaining articles were reviewed, and ten non-pertinent articles were excluded. In addition, one study was excluded due to cohort overlap between studies. Finally, six studies were appropriate for the meta-analysis. These studies described a total of 353 patients with 359 lesions. All studies had been published in the past 4 years, from 2010 to 2014.
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
39
Fig. 1. Flow diagram of studies included in this meta-analysis. ESD, endoscopic submucosal dissection
The enrollment period ranged from 2000 to 2013. In addition, all were single-centre, retrospective studies conducted in Japan. Eight or nine main outcomes were reported in each study among nine possible outcomes. One study by Hoteya et al. [13], which was the only comparative study between EGJ cancer and non-EGJ cancer, had a quality score of 9, whereas other studies had a quality score of 6. The characteristics of the studies included in the meta-analysis are summarized in Table 1. 3.2. Patient and lesion characteristics and ESD procedure Patient and lesion characteristics of the included studies are shown in Table 2. The median of the mean ages across the included studies was 68.7 years (range, 63.0–72.0 years). The median of the male sex rates was 86.3% (range, 79.3–91.3%). All studies included only Siewert type II EGJ cancers with the exception of one study by Kagemoto et al. that did not provide information on Siewert classification [17]. The median of the rates of deep submucosal invasion (>500 m from the muscularis mucosa) was 22.0% (range, 8.6–24.5%). Among the included studies, an insulation-tipped knife was the preferred knife in five of the six studies. In addition, the Hook knife, Flex knife, and Sumitomo Bakelite knife were also used for ESD. The median of the mean procedure times of the included studies was 82.0 min (range, 42.5–102.7 min).
resection rates was 98.6% (95% CI 95.9–99.6%) using the random effects model (Fig. 2A). No heterogeneity was identified (P = 0.982, I2 < 0.001). All studies, including 359 lesions, reported complete resection rates and curative resection rates [12–17]. The pooled estimate of complete resection rates was 87.0% (95% CI 79.7–92.0%; Fig. 2B). In addition, a moderate heterogeneity was identified (P = 0.039, I2 = 57.226). In cases of curative resection rates, the pooled estimate was 74.6% (95% CI 69.8–78.9%; Fig. 2C). No evidence of heterogeneity (P = 0.526, I2 < 0.001) was identified. 3.4. Complications All studies, including 359 lesions, reported post-ESD bleeding rates and perforation rates. The pooled estimate of post-ESD bleeding rates was 3.4% (95% CI 1.7–6.7%; Fig. 3A). No heterogeneity was identified among the studies (P = 0.382, I2 = 5.442). The pooled estimate of perforation rates was 2.5% (95% CI 1.3–5.0%; Fig. 3B). Heterogeneity was not observed (P = 0.852, I2 < 0.001). Stenosis rate was reported in five studies, incorporating 231 lesions [12,14–17]. The pooled estimate of stenosis rates was 6.9% (95% CI 3.2–14.0%; Fig. 3C). No significant heterogeneity (P = 0.144, I2 = 41.567) was observed. All of 13 patients with stenosis were treated successfully with endoscopic balloon dilatation. 3.5. Long-term oncologic outcomes
3.3. Procedural outcomes Five studies, including 231 lesions, provided information on en bloc resection rates [12,14–17]. The pooled estimate of en bloc
All six studies, including 359 lesions, reported local recurrence rate, distant metastasis rate, and cause-specific mortality (Table 3) [12–17]. The median of the median follow-up periods across the
40
Table 1 Characteristics of studies included in the meta-analysis. Author
Year of publication
Study design
2010
Retrospective, single-centre
Hoteya et al. [13]
2013
Retrospective, single-centre
Omae et al. [14]
2013
Retrospective, single-centre
Yamada et al. [15]
2013
Retrospective, single-centre
Imai et al. [16]
2013
Retrospective, single-centre
Kagemoto et al. [17]
2014
Retrospective, single-centre
Enrollment period
Patient
Lesion
Yokohama City University Medical Center Toranomon Hospital
Japan
2000–2009
58
58
Japan
2005–2010
126
128
Cancer Institute Hospital National Cancer Center Hospital Shizuoka Cancer Center
Japan
2004–2010
44
44
Japan
2001–2007
53
53
Japan
2002–2009
49
50
Japan
2006–2013
23
26
Hiroshima University Hospital
Main outcome
Study quality
En-bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, stenosis, recurrence, mortality Complete resection, curative resection, post-ESD bleeding, perforation, recurrence, mortality En-bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, stenosis, recurrence, mortality En-bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, stenosis, recurrence, mortality En-bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, stenosis, recurrence, mortality En-bloc resection, complete resection, curative resection, post-ESD bleeding, perforation, stenosis, recurrence
Selection
Comparability
Exposure/ outcome
***
–
***
****
**
***
***
–
***
***
–
***
***
–
***
***
–
***
Asterisk represents quality scores based on the Newcastle-Ottawa scale. ESD, endoscopic submucosal dissection.
Table 2 Patient and lesion characteristics of the included studies. Author
Age, year
Male (%)
Siewart classification
Histologic type of lesion
Lesion size (mm)
Deep submucosal invasiona (%)
LVI (%)
Drug for sedation
Lifting solution
Device for ESD
Procedure time (min)
Hirasawa et al. [12]
Mean 69.3 (range 36–85) BA: mean 63.5 (SD 12.5) GC: mean 68.8 (SD 9.3) Mean 70.0 (range 42–84) Median 68 (range 39–88) Median 72 (range 41–89) Mean 63 (SD 10)
79.3
Type II
BA: 1, GC: 57
8.6
6.9
Type II
BA: 25, GC: 103
N/A
BA: 24.0 GC: 9.7
Diazepam or midazolam Diazepam
Epinephrine + saline Glycerol + epinephrine
IT-knife
83.6
90.9
Type II
BA: 13, GC: 31
11.4
2.3
N/A
Type II
BA: 28, GC: 25
24.5
7.5
N/A
Glycerol or hyaluronic acid N/A
Mean 82 (range 22–275) BA: mean 84.9 (SD 52.3) GC: mean 107.0 (SD 64.4) N/A
84.9
Mean 20.3 (range 3–50) BA: mean 20.2 (SD 17.6) GC: mean 21.8 (SD 15.8) Mean 17 (range 5–47) Median 20 (3–47)
IT-knife
87.8
Type II
BA: 7, GC: 43
22.0
10.0
Midazolam
N/A
IT-knife
91.3
N/A
BA: 26
23.1
0.0
N/A
Glycerol + epinephrine
Hook knife, IT-knife nano, SB knife
Hoteya et al. [13]
Omae et al. [14] Yamada et al. [15] Imai et al. [16] Kagemoto et al. [17]
Median 22 (range 5–70) Mean 19.0 (SD 13.6)
Hook knife, Flex knife
IT-knife
Median 70 (range 20–480) Median 42.5 (range 10–157) Mean 93 (SD 48)
LVI, lymphovascular invasion; ESD, endoscopic submucosal dissection; BA, Barrett’s adenocarcinoma; GC, gastric cardia cancer; IT, insulation-tipped; SB, Sumitomo Bakelite; SD, standard deviation; N/A, not available a SM2 invasion (>500 m from the muscularis mucosa) or more.
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
Hirasawa et al. [12]
Country
Institution
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
41
Fig. 2. Forest plots of the included studies analyzing the procedural outcomes including the en bloc resection rates (A), the complete resection rates (B), and the curative resection rates (C). CI, confidence interval
included studies was 35.8 months (range, 33.0–73.2 months). In all but one study [12–16], curative resection was determined according to the curative resection criteria for gastric cancer. One remaining study used the curative resection criteria for esophageal cancer for determining curative resection [17]. In 269 lesions with curative resection, no local recurrence or distant metastasis was seen. On the contrary, three local recurrences and two distant metastases occurred in 90 lesions with non-curative resection. Three cause-specific deaths were identified in a total of 359 lesions (pooled estimate 1.7% [95% CI 0.7–4.0%]; P-value for heterogeneity = 0.791; I2 < 0.001). All cause-specific deaths occurred in patients with non-curative resection. 3.6. Subgroup analysis In the subgroup analysis, after excluding two studies that included patients who had undergone ESD during the initial introduction period of ESD in Japan in 2000–2001 [12,15], four studies described a total of 242 patients with 248 lesions [13,14,16,17]. The pooled estimates of en bloc resection, complete resection, and curative resection were 98.5% (95% CI 95.0–99.6%), 90.9% (95% CI
84.6–94.7%), and 75.1% (95% CI 69.3–80.1%), respectively. No heterogeneity was identified (en bloc resection: P = 0.957, I2 < 0.001; complete resection: P = 0.253, I2 = 26.425; curative resection: P = 0.528, I2 < 0.001). The pooled estimates of bleeding rates, perforation rates, and stenosis rates were 2.9% (95% CI 1.0–7.7%), 2.4% (95% CI 1.1–5.5%), and 7.3% (95% CI 2.3–20.8%), respectively. No significant heterogeneity was identified (bleeding: P = 0.276, I2 = 22.532; perforation: P = 0.794, I2 < 0.001; stenosis: P = 0.127, I2 = 51.519). In the subgroup, distant metastasis was identified in two patients with non-curative resection. 4. Discussion This systematic review and meta-analysis showed that ESD for early stage EGJ cancer is a feasible treatment option with excellent en bloc resection rates. Generally, ESD has an advantage over endoscopic mucosal resection in terms of achievement of en bloc resection regardless of lesion size [26,27]. In this study, the en bloc resection rate of ESD for early stage EGJ cancer was almost 100%. Considering that complete resection rates of ESD for EGC are usually 90–95% [4–7], the complete resection rate of ESD for early stage EGJ cancer of 87.0% (95% CI 79.0–92.0%) in this meta-analysis is a very reasonable rate.
42
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
Fig. 3. Forest plot of the included studies analyzing the complications, including the post-ESD bleeding rates (A), the perforation rates (B), and the stenosis rates (C). ESD, endoscopic submucosal dissection; CI, confidence interval
Although the en bloc resection rate and complete resection rate of ESD for early stage EGJ cancer were acceptable, the learning curve for ESD was a concern before conducting the meta-analysis [22]. Since the first endoscopic resection using an insulation-tipped knife was introduced in 1999 [28], many improvements in the instruments used for ESD have been made. To minimize the effect of the learning curve, we conducted a subgroup analysis after excluding studies that included patients who had undergone ESD during the initial introduction period in Japan. While the pooled estimate of complete resection rates in all of the six included studies was 87.0% (95% CI 79.7–92.0%), the same estimate in the subgroup analysis was 90.9% (95% CI 84.6–94.7%). Furthermore, previously observed moderate heterogeneity for complete resection rate among the studies was removed in the subgroup analysis. We believe these results imply a learning curve effect in ESD for EGJ cancer. It is well known that the difficulty of performing ESD depends on the lesion location (especially the upper third of the stomach), and complete resection can be hindered by such location-related difficulties [29,30]. In contrast to a high en bloc resection rate and complete resection rate, the curative resection rate of ESD in early stage EGJ cancer seems to be lower than that of ESD in gastric
cancer [4,31]. The low curative resection rate of ESD for EGJ cancer in this meta-analysis, however, is caused by heterogeneity in the study population. All studies in the meta-analysis included lesions that did not fulfil both absolute and expanded indications while other previous studies that reported a curative resection rate of ESD in gastric cancer included only lesions that met ESD indication criteria. For example, the median rate of deep submucosal invasion (>500 m from the muscularis mucosa) across the studies included in the meta-analysis was 22.0% (range, 8.6–24.5%). In other words, about 20% of the lesions included in the meta-analysis could not achieve curative resection regardless of ESD outcomes. Considering the difference among study populations, it is not clear whether the 74.6% of the pooled estimate of curative resection rates in EGJ cancer is inferior to that of the curative resection rate in gastric cancer. Due to the difficult location for ESD in early stage EGJ cancer, we expected that post-ESD bleeding and perforation would increase in EGJ cancer compared with gastric cancer. Pooled estimates of post-ESD bleeding rates and perforation rates in EGJ cancer, however, were comparable to the outcomes in gastric cancer [21,32]. Judging from procedural outcomes and complications, including
3/359 2/90 3/90 0/269 0/269
This value included the follow-up period of patients with both junctional- and non-junctional gastric cancer in the study.
Total
a
0/58 1/128 0/44 2/53 0/50 0/26 0/12 1/33 0/7 0/17 1/14 0/7 0/12 0/33 0/7 3/17 0/14 0/7 0/46 0/95 0/37 0/36 0/36 0/19
Local recurrence
0/46 0/95 0/37 0/36 0/36 0/19 Gastric cancer Gastric cancer Gastric cancer Gastric cancer Gastric cancer Esophageal cancer Hirasawa et al. [12] Hoteya et al. [13] Omae et al. [14] Yamada et al. [15] Imai et al. [16] Kagemoto et al. [17]
Non-curative resection
Local recurrence
Distant metastasis Curative resection Curative resection criteria Author
Table 3 Long-term clinical outcomes of endoscopic submucosal dissection for esophagogastric junction cancer.
Distant metastasis
Cause-specific mortality
Periods of follow-up, months
Median 36.6 (range 4–94) Median 34 (range 2–96)a Mean 33 (range 6–64) Median 73.2 (range 16.8–132) Median 47 (range 22–97) Median 35 (range 2–85)
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
43
post-ESD bleeding and perforation, ESD appears to be an acceptable treatment option for EGJ cancer. However, considerable concerns remain about stenosis at postESD sites. The pooled estimate of stenosis rates was highest among the pooled estimates of complications in the meta-analysis. A previous study with a meta-analysis of ESD for superficial squamous esophageal cancer demonstrated that the pooled estimate of stenosis rates was 11.6% (95% CI 8.2–16.2%) [33]. Other studies on pyloric stenosis after ESD reported that stenosis rates were about 3–7% [34,35]. In comparison with stenosis rates after ESD for esophageal cancer and pyloric cancer, we think that the stenosis rate after ESD for EGJ cancer is not significant. Nevertheless, stenosis is obviously a major complication of ESD for EGJ cancer. One of the most interesting points in treating a patient with early stage EGJ cancer after ESD is determining which curative resection criteria should be applied. In the present meta-analysis, five of the six included studies assessed curative resection according to curative resection criteria for gastric cancer. The one remaining study used curative resection criteria for esophageal cancer as a definition of curative resection. All 269 patients with curative resection showed no local recurrence or distant metastasis regardless of curative resection criteria. Although more large-scale data are necessary to form a definite conclusion, we believe that curative resection criteria for gastric cancer can be used for assessing curative resection in EGJ cancer. In contrast to the long-term oncologic outcomes in patients with curative resection, outcomes in patients with non-curative resection were not as good. Among 90 patients with non-curative resection, 3 (3.3%) patients had local recurrence and 2 (2.2%) had distant metastasis. Three of the local recurrences were identified in only one study by Yamada et al., which had the longest follow-up duration among the included studies [15]. The three local recurrences developed at 3.6, 4.6, and 7.4 years after ESD, respectively. Considering that the other five studies had a median follow-up duration of around 3 years, they may have underestimated poor outcomes in patients with non-curative resections. Although this is the first meta-analysis of ESD outcomes in the management of early stage EGJ cancer, this study has several limitations. Most importantly, all of the included studies were performed in Japan, which is the leading country for advanced endoscopic procedures, and Japanese endoscopists have extensive expertise in therapeutic endoscopy, including ESD. Therefore, our results may not be generalizable to Western countries. Second, all studies included in the meta-analysis had a single-centre, retrospective design. Prospectively collected data in each institution, however, would minimize selection bias. Third, most of the included studies were observational studies on EGJ cancer without group comparisons. Therefore, ESD outcomes in the present meta-analysis could not be compared directly with those for other cancers, including gastric cancer. Despite these limitations, we believe that our data form the basis for understanding clinical outcomes of ESD for EGJ cancer. A high en bloc resection rate and complete resection rate were seen after ESD for EGJ cancer. Although about 7% of lesions induced stenosis after ESD, other major complications, including post-ESD bleeding and perforation, were in the reasonably acceptable range. When curative resections are achieved, good oncologic outcomes are likely in the management of early stage EGJ cancer by ESD. Conflict of interest None declared. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.dld.2014.10.011.
44
C.H. Park et al. / Digestive and Liver Disease 47 (2015) 37–44
References [1] Japanese gastric cancer treatment guidelines 2010 (ver. 3). Gastric Cancer 2011;14:113–23. [2] Gotoda T, Yamamoto H, Soetikno RM. Endoscopic submucosal dissection of early gastric cancer. Journal of Gastroenterology 2006;41:929–42. [3] Choi JH, Kim ER, Min BH, et al. The feasibility and safety of the endoscopic submucosal dissection of superficial gastric neoplastic lesions in patients with compensated liver cirrhosis: a retrospective study. Gut Liver 2012;6:58–63. [4] Kim DY, Hong SJ, Cho GS, et al. Long-term efficacy of endoscopic submucosal dissection compared with surgery for early gastric cancer: a retrospective cohort study. Gut Liver 2014;8:519–25. [5] Lee H, Yun WK, Min BH, et al. A feasibility study on the expanded indication for endoscopic submucosal dissection of early gastric cancer. Surgical Endoscopy 2011;25:1985–93. [6] Ahn JY, Jung HY, Choi KD, et al. Endoscopic and oncologic outcomes after endoscopic resection for early gastric cancer: 1370 cases of absolute and extended indications. Gastrointestinal Endoscopy 2011;74:485–93. [7] Park CH, Shin S, Park JC, et al. Long-term outcome of early gastric cancer after endoscopic submucosal dissection: expanded indication is comparable to absolute indication. Digestive and Liver Disease 2013;45:651–6. [8] Suh Y, Han D, Kong S, et al. Should adenocarcinoma of the esophagogastric junction be classified as esophageal cancer? A comparative analysis according to the seventh AJCC TNM classification. Annals of Surgery 2012;255: 908–15. [9] Yamashita H, Katai H. Should adenocarcinoma of the esophagogastric junction be classified as esophageal cancer? Limited to Siewert Type II, Yes. Annals of Surgery 2014, http://dx.doi.org/10.1097/SLA.0000000000000589 [in press]. [10] Pedrazzani C. Should adenocarcinoma of the esophagogastric junction be classified as gastric or esophageal cancer, or else as a distinct clinical entity? Annals of Surgery 2014, http://dx.doi.org/10.1097/SLA.0000000000000524 [in press]. [11] Pech O, May A, Rabenstein T, et al. Endoscopic resection of early oesophageal cancer. Gut 2007;56:1625–34. [12] Hirasawa K, Kokawa A, Oka H, et al. Superficial adenocarcinoma of the esophagogastric junction: long-term results of endoscopic submucosal dissection. Gastrointestinal Endoscopy 2010;72:960–6. [13] Hoteya S, Matsui A, Iizuka T, et al. Comparison of the clinicopathological characteristics and results of endoscopic submucosal dissection for esophagogastric junction and non-junctional cancers. Digestion 2013;87:29–33. [14] Omae M, Fujisaki J, Horiuchi Y, et al. Safety, efficacy, and long-term outcomes for endoscopic submucosal dissection of early esophagogastric junction cancer. Gastric Cancer 2013;16:147–54. [15] Yamada M, Oda I, Nonaka S, et al. Long-term outcome of endoscopic resection of superficial adenocarcinoma of the esophagogastric junction. Endoscopy 2013;45:992–6. [16] Imai K, Kakushima N, Tanaka M, et al. Validation of the application of the Japanese curative criteria for superficial adenocarcinoma at the esophagogastric junction treated by endoscopic submucosal dissection: a long-term analysis. Surgical Endoscopy 2013;27:2436–45. [17] Kagemoto K, Oka S, Tanaka S, et al. Clinical outcomes of endoscopic submucosal dissection for superficial Barrett’s adenocarcinoma. Gastrointestinal Endoscopy 2014;80:239–45. [18] Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care
[19]
[20]
[21]
[22]
[23]
[24] [25] [26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
interventions: explanation and elaboration. Journal of Clinical Epidemiology 2009;62:e1–34. Wells GA, Shea B, O’Connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses; 2014. Available at: http://www.ohri.ca/programs/clinical epidemiology/oxford.asp [accessed May 25]. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. European Journal of Epidemiology 2010;25:603–5. Akasaka T, Nishida T, Tsutsui S, et al. Short-term outcomes of endoscopic submucosal dissection (ESD) for early gastric neoplasm: multicenter survey by Osaka university ESD study group. Digestive Endoscopy 2011;23:73–7. Hong KH, Shin SJ, Kim JH. Learning curve for endoscopic submucosal dissection of gastric neoplasms. European Journal of Gastroenterology and Hepatology 2014;26:949–54. Siewert J. Carcinoma of the cardia: carcinoma of the gastroesophageal junction – classification, pathology and extent of resection. Diseases of the Esophagus 1996;9:173–82. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Statistics in Medicine 2002;21:1539–58. DerSimonian R, Laird N. Meta-analysis in clinical trials. Controlled Clinical Trials 1986;7:177–88. Park YM, Cho E, Kang HY, et al. The effectiveness and safety of endoscopic submucosal dissection compared with endoscopic mucosal resection for early gastric cancer: a systematic review and metaanalysis. Surgical Endoscopy 2011;25:2666–77. Lian J, Chen S, Zhang Y, et al. A meta-analysis of endoscopic submucosal dissection and EMR for early gastric cancer. Gastrointestinal Endoscopy 2012;76:763–70. Gotoda T, Kondo H, Ono H, et al. A new endoscopic mucosal resection procedure using an insulation-tipped electrosurgical knife for rectal flat lesions: report of two cases. Gastrointestinal Endoscopy 1999;50:560–3. Imagawa A, Okada H, Kawahara Y, et al. Endoscopic submucosal dissection for early gastric cancer: results and degrees of technical difficulty as well as success. Endoscopy 2006;38:987–90. Goto A, Nishikawa J, Okamoto T, et al. Outcomes of endoscopic submucosal dissection for early gastric cancer and factors associated with incomplete resection. Hepato-Gastroenterology 2013;60:46–53. Ohnita K, Isomoto H, Shikuwa S, et al. Early and long-term outcomes of endoscopic submucosal dissection for early gastric cancer in a large patient series. Experimental and Therapeutic Medicine 2014;7:594–8. Matsui N, Akahoshi K, Nakamura K, et al. Endoscopic submucosal dissection for removal of superficial gastrointestinal neoplasms: a technical review. World Journal of Gastrointestinal Endoscopy 2012;4:123–36. Kim JS, Kim BW, Shin IS. Efficacy and safety of endoscopic submucosal dissection for superficial squamous esophageal neoplasia: a meta-analysis. Digestive Diseases and Sciences 2014;59:1862–9. Iizuka H, Kakizaki S, Sohara N, et al. Stricture after endoscopic submucosal dissection for early gastric cancers and adenomas. Digestive Endoscopy 2010;22:282–8. Coda S, Oda I, Gotoda T, et al. Risk factors for cardiac and pyloric stenosis after endoscopic submucosal dissection, and efficacy of endoscopic balloon dilation treatment. Endoscopy 2009;41:421–6.