Endoscopic diagnosis and treatment of esophageal adenocarcinoma: introduction of Japan Esophageal Society classification of Barrett's esophagus.

Endoscopic surveillance of Barrett's esophagus has become a foundation of the management of esophageal adenocarcinoma (EAC). Surveillance for Barrett's esophagus commonly involves periodic upper endoscopy with biopsies of suspicious areas and random four-quadrant biopsies. However, targeted biopsies using narrow-band imaging can detect more dysplastic areas and thus reduce the number of biopsies required. Several specific mucosal and vascular patterns characteristic of Barrett's esophagus have been described, but the proposed criteria are complex and diverse. Simpler classifications have recently been developed focusing on the differentiation between dysplasia and non-dysplasia. These include the Japan Esophageal Society classification, which defines regular and irregular patterns in terms of mucosal and vascular shapes. Cancer invasion depth is diagnosed by endoscopic ultrasonography (EUS); however, a meta-analysis of EUS staging of superficial EAC showed favorable pooled values for mucosal cancer staging, but unsatisfactory diagnostic results for EAC at the esophagogastric junction. Endoscopic resection has recently been suggested as a more accurate staging modality for superficial gastrointestinal cancers than EUS. Following endoscopic resection for gastrointestinal cancers, the risk of metastasis can be evaluated based on the histology of the resected specimen. European guidelines describe endoscopic resection as curative for well- or moderately differentiated mucosal cancers without lymphovascular invasion, and these criteria might be extended to lesions invading the submucosa (≤ 500 μm), i.e., to low-risk, well- or moderately differentiated tumors without lymphovascular involvement, and < 3 cm. These criteria were confirmed by a recent study in Japan.

Introduction

Esophageal adenocarcinoma (EAC) is an aggressive disease with an increasing incidence in the Western world [1-3]. Although no equivalent data are available for Eastern countries, the rate of EAC is expected to increase in Asia because of the decreasing prevalence of Helicobacter pylori infection and Westernization of the diet [4, 5]. Survival of patients with EAC correlates with disease stage, with a 5-year-survival rate of about 20% in patients with locally advanced disease [6]. The poor survival of patients with advanced EAC indicates the need for its early detection [7, 8]. Endoscopic surveillance of Barrett’s esophagus (BE) has become a foundation of the management of EAC, especially in Western countries [9-11], and this trend has accelerated in line with recent developments in advanced imaging and endoscopic resection technologies.

Surveillance and classification of lesions in patients with BE

Surveillance for EAC in patients with BE commonly involves periodic upper endoscopy, with biopsies of suspicious areas and random four-quadrant biopsies [12]. However, this biopsy protocol is time consuming, carries a risk of sampling error, and is hampered by low patient compliance [13]. New endoscopic techniques have, therefore, been developed to improve the recognition of specialized intestinal metaplasia (SIM), dysplasia, and cancer, by enhancing mucosal morphology. The most widely used such modality is narrow-band imaging (NBI) [14], and targeted biopsies sampled by this method allowed the detection of more dysplastic areas, therefore, reducing the number of biopsies required [15].

Several groups have described specific mucosal and vascular patterns characteristic for the diagnosis of lesions in BE using NBI [16-22]. These classification systems suggested that irregular mucosal pattern and vessels are predictive of dysplasia, while a ridged/villous pattern is predictive of SIM; however, despite promising initial findings, subsequent validation studies of these classification systems have reported unfavorable results [23-27]. Furthermore, the proposed criteria were complex and diverse, thus limiting their use in daily clinical practice, with the complexity associated with the concept of differentiating between SIM and non-SIM and between dysplasia and non-dysplasia within the same classification.

Simpler classifications have recently been developed focusing on differentiating between dysplasia and non-dysplasia, with the aim of improving the clinical utility of the classification [28, 29]. The new classifications classify most mucosal or vascular descriptors as “regular” for non-dysplastic and “irregular” for dysplastic BE (Table 1). These simple descriptors make the classifications easy to apply in clinical practice, with acceptable sensitivity, specificity, and inter-observer agreement for the diagnosis of dysplasia in BE (Table 1).

Table 1

New endoscopic classifications for the diagnosis of lesions in patients with Barrett’s esophagus

	BING classification	JES classification for Barrett’s esophagus
Non-dysplasia	Mucosal pattern: regular	Mucosal pattern: regular
	Vascular pattern: regular	Vascular pattern: regular flat pattern
Dysplasia	Mucosal pattern: absent or irregular	Mucosal pattern: irregular
	Vascular pattern: irregular	Vascular pattern: irregular
Diagnostic accuracy	Sensitivity 80%	Sensitivity 87%
	Specificity 88%	Specificity 97%
Reproducibility	κ = 0.68	κ = 0.77

BING Barrett’s International NBI Group, JES Japan Esophageal Society

These new classifications include the Japan Esophageal Society classification of BE [29], in which the mucosal and vascular patterns are described as either regular or irregular (Table 2), based on detailed definitions of regular and irregular in terms of mucosal and vascular shape or arrangement (Figs. 1, 2) (Table 3), thus making the findings easy to interpret. This classification also includes a flat pattern (Fig. 3) as a regular pattern corresponding to non-dysplastic histology [30]. A validation study conducted by 10 endoscopic image reviewers using 156 still images showed promising accuracy and inter-observer agreement (Table 1).

Table 2

Japan Esophageal Society classification of Barrett’s esophagus

Pattern	Visibility	Morphologic features	Regularity
Mucosal	Visible	Pit	Regular or irregular
		Non-pit
	Invisiblea
Vascular	Visible	Net	Regular or irregular
		Non-netb
	Invisible

aIncluding a flat pattern

bIncluding normal-appearing long branching vessels and thick greenish vessels suggestive of a flat pattern

Fig. 1

Barrett’s esophageal cancer showing irregular vascular pattern (net type)

Fig. 2

Barrett’s esophageal cancer showing irregular mucosal pattern (non-pit type)

Table 3

Definition of regularity in Japan Esophageal Society classification of Barrett’s esophagus

Pattern	Regular	Irregular
Mucosal
Form/size	Similar	Various
Arrangement	Regular	Irregular
Density	Low or same as surrounding area	High
White zone	Clearly visible and/or with homogeneous width	Obscure/invisible or heterogeneous width
Vascular
Form	Similar or bending and branching gently or regularly	Various or bending and branching steeply or irregularly
Caliber change	Gradual	Abrupt
Location	Between or in mucosal patterns	Beyond of regardless of mucosal patterns
Flat pattern	Completely flat surface without a clear demarcation line. Greenish thick vessels and/or long branching vessels

Fig. 3

Flat-type mucosa: completely flat surface without a clear demarcation line and greenish thick vessels

Diagnosis of cancer invasion depth

Correct preoperative staging is crucial, given that the patient’s treatment strategy is determined largely on the basis of cancer invasion depth. Non-magnified endoscopy is the primary modality for diagnosing gastrointestinal cancer, and is also helpful for diagnosing cancer invasion depth. Correlations between endoscopic macroscopic type and invasion depth of superficial EAC have been reported [31, 32], and previous studies showed that non-magnified endoscopy could accurately diagnose invasion depth in gastrointestinal cancers [33-36]. One study found that the overall correct diagnostic assessment of early esophageal cancers was high using either non-magnified endoscopy or endoscopic ultrasonography (EUS) with a 20-MHz mini-probe, with no significant differences between the two techniques (Table 4) [37]. Although its relative simplicity means that non-magnified endoscopy may be a good modality for diagnosing EAC invasion depth, the diagnosis is subjective, and more objective criteria are, therefore, needed.

Table 4

Diagnostic performances of non-magnified endoscopy and endoscopic ultrasonography for superficial esophageal adenocarcinoma (sensitivity and specificity for mucosal cancer)

Author	Country/year/sample size	Modality	Sensitivity	Specificity	Accuracy
May A [37]	Germany/2004/93	Non-magnified endoscopy	94	56	83
		EUS	91	48	79

EUS endoscopic ultrasonography

EUS can also be used to diagnose cancer invasion depth. Conventional EUS (7.5 MHz) can differentiate between advanced T3/T4 carcinomas and T1/T2 carcinomas in more than 80% of cases; however, accurate differentiation between mucosal and submucosal (SM) invasion is difficult [38-41]. However, EUS using a mini-probe (20 MHz) enables the esophageal wall to be imaged in nine layers, thus permitting the muscularis mucosa to be seen in greater detail. Mini-probe EUS can, therefore, be used to distinguish between mucosal and SM cancers, thereby improving staging accuracy.

A previous meta-analyses of EUS staging of superficial esophageal cancers showed favorable pooled values for mucosal cancer staging, with a sensitivity of 0.85 [95% confidence interval (CI) 0.82–0.88], specificity of 0.87 (95% CI 0.84–0.90), positive likelihood ratio of 6.62 (95%CI 3.6–12.12), and negative likelihood ratio of 0.20 (95%CI 0.14–0.30). The equivalent values for SM cancer staging were 0.86 for sensitivity (95%CI 0.82–0.89), 0.86 for specificity (95%CI 0.83–0.89), 5.13 for positive likelihood ratio (95%CI 3.36–7.82), and 0.17 for negative likelihood ratio (95%CI 0.09–0.30) [42].

However, when the results were limited to the diagnosis of EAC, the performance of EUS was not satisfactory (Table 5) [43-46] compared with its ability to diagnose esophageal squamous cell carcinoma and gastric cancer. Meta-analyses of the diagnostic accuracy of EUS for mucosal or SM micro-invasive esophageal squamous cell carcinoma showed a sensitivity of 0.87 (95%CI 0.81–0.92), specificity 0.94 (95%CI 0.88–0.98), positive likelihood ratio 11.6 (95%CI 5.4–24.7), and negative likelihood ratio 0.15 (95%CI 0.10–0.23) [47], with equivalent results for mucosal gastric cancer of sensitivity 0.87 (95%CI 0.81–0.92), specificity 0.75 (95%CI 0.62–0.84), positive likelihood ratio 3.4 (95%CI 2.3–5.0), and negative likelihood ratio 0.17 (95%CI 0.12–0.24) [48].

Table 5

Diagnostic performance of endoscopic ultrasonography for superficial esophageal adenocarcinoma (sensitivity and specificity for mucosal cancer)

Author	Country/year	Sample size	Sensitivity	Specificity	Accuracy
Thomas T [43]	UK/2010	46	94	67	85
Fernández-Sordo JO [44]	USA/2012	109	84	50	83
Bergeron EJ [45]	USA/2014	107	72	49	64
Dhupar R [46]	USA/2015	130	59	69	64

The poor diagnostic yield was probably caused by difficulties in diagnosing EAC in the distal part of the esophagus, given that the diagnostic accuracy for EAC in the distal part of the esophagus was significantly worse than that for EAC in the mid- and proximal parts of the esophagus (Table 6) [37, 49]. This emphasizes the fact that it is particularly difficult to achieve adequate water preparation in the distal esophagus by instilling fluid through the endoscopic channel, in addition to substantial motility that prevents dilatation of the distal esophagus from being maintained for longer periods.

Table 6

Diagnostic performance of endoscopic ultrasonography for superficial esophageal adenocarcinoma (sensitivity and specificity for mucosal cancer) with regard to imaging modality and lesion location

Author	Country/year/sample size	Modality	Location	Sensitivity	Specificity	Accuracy
May A [37]	Germany/2004/93	Non-magnified endoscopy	Distal	92	43	78
			Mid to proximal	97	91	95
		EUS	Distal	89	14	69
			Mid to proximal	94	91	93
Chemaly M [49]	France/2008/91	EUS	Distal	Not described	Not described	48
			Mid to proximal	Not described	Not described	87

EUS endoscopic ultrasonography

Endoscopic resection

Endoscopic resection has recently been suggested as a staging modality for superficial gastrointestinal cancers, based on the limited accuracies of EUS and non-magnified endoscopy. Endoscopic resection, in the form of endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD), allows for removal of visible lesions and histologic assessment of the resected tissue, thus facilitating accurate diagnostic staging of the disease (Figs. 4, 5, 6, 7) [50, 51].

Fig. 4

IIa type esophagogastric junctional cancer

Fig. 5

IIa type esophagogastric junctional cancer with indigo carmine staining

Fig. 6

Histology of resected specimen showed deep muscularis mucosa invasion of cancer. SMM superficial muscularis mucosa, LPM lamina propria, DMM deep muscularis mucosa

Fig. 7

Mapping of the cancer. SMM superficial muscularis mucosa, LPM lamina propria, DMM deep muscularis mucosa, MM muscularis mucosa

The various modalities of EMR include the use of a transparent cap, two-channel endoscope, and ligation. However, these modalities are limited with respect to resection size, and large lesions must be resected in several fragments. Histological assessment of cancer invasion depth can be inaccurate if lesions are resected in small fragments, and histologic evaluation of several specimens does not allow the outer margins of the neoplastic area to be identified, and complete resection, therefore, cannot be confirmed. In addition, piecemeal resection of early neoplasia in BE is associated with a high local recurrence rate, probably because of small remnants of neoplastic tissue left in situ [52-55]. ESD provides larger specimens than EMR, thus allowing more precise histological analysis and higher en bloc and curative resection rates, and potentially reducing the incidence of recurrence. A recent meta-analysis of non-randomized studies showed that ESD of early gastrointestinal tumors was superior to EMR in terms of en bloc and curative resection rates, but was more time consuming and associated with higher rates of bleeding and perforation [56].

Several studies have reported on the use of ESD for EAC and esophagogastric junction cancer [57-66]. In general, ESD is associated with favorable outcomes with acceptable en bloc resection and complication rates. However, the curative resection rate, defined as en bloc resection with cancer-free margins and minimal risk of metastasis, limited to EAC at the esophagogastric junction, was significantly lower than those for cardia and non-cardia gastric cancers (Table 7) [57, 66]. One cause of incomplete resection of esophagogastric junction EAC was positive lateral margins caused by sub-epithelial progression of the tumor proximally, which were hard to recognize before treatment, while the low accuracy of diagnosing cancer invasion depth before treatment and high lymphovascular involvement, confirmed in resected specimens, were also contributory factors.

Table 7

Outcomes of endoscopic submucosal dissection for esophagogastric junctional cancer with regard to location

Author	Location	Complete resectiona	Curative resectionb
Osumi H [66]	Esophagus	100% (55/55)	62% (34/55)
	Cardia	100% (87/87)	82% (71/87)
Hoteya S [57]	Esophagus	64% (16/25)	48% (12/25)
	Cardia	96% (99/103)	81% (83/103)

aComplete resection: en bloc resection with cancer-free margins

bCurative resection: complete resection with low risk of metastasis

Risk of metastasis

The risk of metastasis after endoscopic resection for gastrointestinal cancers is evaluated based on histologic findings of the resected specimen. Studies of esophagectomy specimens have indicated a low risk of 0.0–1.3% for mucosal EAC [67-69], thus providing the rationale for endoscopic treatment of mucosal EAC with curative intent.

The frequency of metastasis in EAC is known to increase with increasing depth of tumor invasion into the SM [70-72]. SM1 cancer, i.e., cancer invading the shallow part of the SM, remains the most controversial, with some studies reporting a relevant incidence of lymph node metastasis even in SM1 cancers [73-75]. However, when the rate of metastasis is stratified by pathologic findings, SM1 cancers without risk factors such as lymphovascular involvement and a poorly differentiated component have very low rates [76-78]. Some studies [79, 80] have accordingly suggested that a subgroup of SM cancers could be adequately treated by endoscopic resection.

European guidelines [80] indicate that endoscopic resection appears to be curative for well- or moderately differentiated mucosal cancers without lymphatic or vascular invasion, and that these criteria might be extended to lesions with invasion into the SM (≤ 500 μm), namely to low-risk tumors (well or moderately differentiated, without lymphovascular involvement, < 3 cm) (Table 8). A recent study in Japan [81] validated these criteria, showing no metastases (0/186 lesions) in patients with mucosal cancer without lymphovascular involvement and a poorly differentiated component, or in patients with SM cancer (≤ 500 μm) without lymphovascular involvement, a poorly differentiated component, and ≤ 30 mm (0/32 lesions).

Table 8

Assessment of metastasis risk based on histology of endoscopically resected specimen

European guideline [80]	Curative for	Curative criteria might be extended to
	Mucosal cancerWell or moderately differentiatedLymphovascular involvement(−)	Submucosal cancer (≤ 500 μm)Well or moderately differentiatedLymphovascular involvement(−)Tumor size < 3 cm
Report from Japan Ishihara R [81]	Very low risk (no metastasis in 186 cancers)	Low risk (no metastasis in 32 cancers)
	Mucosal cancerPoorly differentiated component (−)Lymphovascular involvement (−)	Submucosal cancer (≤ 500 μm)Poorly differentiated component (−)Lymphovascular involvement (−)Tumor size ≤ 3 cm

Future perspectives

Recent advances in endoscopic technologies have provided various tools for the management of gastrointestinal cancers. Previous studies showed the utility of such tools for the early detection and accurate staging of cancers. However, most of these studies were retrospective and limited by small sample sizes. Prospective, multicenter studies are, therefore, needed to provide more reliable evidence and facilitate the use of these tools in clinical practice.