Analysis of factors associated with local recurrence after endoscopic resection of gastric epithelial dysplasia: a retrospective study

Gastric adenoma or dysplasia can be defined as a precancerous lesion or an atypical change originating from the stomach epithelium. Approximately 11% of gastric dysplasias are reported to progress to cancer within 4 years. Moreover, 8–59% of gastric dysplasias are associated with gastric cancer [1].

Endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD) are widely used techniques for the treatment of gastric epithelial dysplasia. Because high-grade dysplasia (HGD) is an obvious precancerous lesion, aggressive treatment such as surgical resection or endoscopic treatment is required. However, low-grade dysplasia (LGD) is associated with a relatively low (3–9%) incidence of gastric cancer [2]. Histological examination of resected lesions following endoscopic treatment for dysplasia may lead to an upgrade of the final diagnosis. Therefore, there is a continuing debate as to whether an aggressive LGD treatment or selective treatment of lesions with risk factors for cancer is needed. Therefore, this study aimed to evaluate the long-term prognosis following endoscopic procedures for gastric epithelial dysplasia, to investigate differences in local recurrence rates according to treatment modality, and to evaluate risk factors associated with the local recurrence of dysplasia.

Previous studies comparing clinical outcomes according to endoscopic resection methods have focused mostly on early gastric cancer, and there have been few studies on gastric dysplasia alone. Therefore, this study was carried out to compare the local recurrence rates for EMR and ESD and to identify the risk factors of local recurrence. The results showed that the complete resection rate was significantly higher, and the local recurrence rate was significantly lower, in patients with gastric epithelial dysplasia treated with ESD rather than with EMR.

According to the National Cancer Information Center of Korea, as of 2016, stomach cancer had the highest incidence, at 35%, compared to that for other solid cancers [7]. In Korea, gastroduodenoscopy during a health checkup is recommended every 2 years, starting at the age of 40 years. Therefore, the diagnosis of gastric dysplasia, as well as early gastric cancer, is increasing. According to the Correa hypothesis, gastric dysplasia is a precancerous lesion that progresses from gastric atrophy and intestinal metaplasia to adenocarcinoma through hypoplasia or dysplasia [8]. However, a previous study on the natural course of gastric dysplasia showed that LGD progressed to adenocarcinoma at a relatively low rate of about 0–23%, while HGD showed a higher progression rate of about 10–81% [9]. In a recent study that followed patients with gastric dysplasia for 7 years, only 7.8% of patients with LGD cases progressed to cancer during follow-up, while 63.6% of those with HGD cases progressed to cancer [10]. In the present study, the final pathological diagnosis was not changed after endoscopic resection in 81.6% of LGD patients, but 49.2% of HGD patients were diagnosed with adenocarcinoma after endoscopic resection (Fig. 2). Thus, there is no question that HGD requires endoscopic resection or surgical treatment because of its potential for cancer progression and the coexistence of cancer cells. In contrast, LGD has a relatively low risk of malignant transformation, and spontaneously regresses in 32–59% of patients in previous studies [11‐16]. The American Society for Gastrointestinal Endoscopy and the British Society of Gastroenterology guidelines recommend endoscopic resection for gastric dysplasia of any size, if possible [17, 18]. The European guidelines also recommend grading and resecting dysplasia in patients with visible endoscopic lesions. If there is no visible endoscopic lesion, it is necessary to confirm the lesion by magnification chromoendoscopy and/or narrow-band imaging. If the lesion is confirmed, a biopsy is performed, and if the diagnosis is LGD, follow-up endoscopy should be performed within 12 months [19]. In Korea, because LGD may sometimes progress to cancer, endoscopic resection is performed, unless it is impossible because of advanced age or comorbid disease.

Gastric dysplasia is mainly treated by endoscopic resection, but argon plasma coagulation (APC) is also used. Several studies have shown that the removal of gastric dysplasia through APC curettage is a good option because of the short hospitalization period, low medical costs, and low complication rates [20, 21]. However, APC is effective for only relatively small-sized LGD lesions (2.0 cm), and the local recurrence rate is higher than that with other endoscopic methods [21]. In addition, since the tissue cannot be collected, determining the final histological diagnosis of the lesion is difficult. Therefore, ESD and EMR are considered as standard treatments for gastric dysplasia, and are used as an additional method to remove remaining lesions after endoscopic resection.

In comparison with conventional EMR, ESD requires a long treatment time and advanced operator skills, and has disadvantages associated with complications such as perforation. However, the advantages of ESD are the en bloc resection of large lesions, high complete resection rate, and low local recurrence rate [9, 22‐26].

In a 2010 study on the predictive factors for local recurrence after endoscopic resection for early gastric cancer, a larger lesion size and treatment with EMR were associated with increased an incomplete resection rate, and had a significant impact on local recurrence [27]. Data from this study showed that en bloc resection can be an important predictor of local recurrence. Histologic type, comorbidities, location of the lesion, color of the lesion, gross type, and presence of H. pylori infection were not correlated with local recurrence; only the size of the lesion and an incomplete resection were risk factors of local recurrence. In this study related to recurrence after resection of gastric dysplasia, most cases of relapse after endoscopic treatment could be treated with additional endoscopic resection (97%, 28/29) Also, this shows that even for the lesions resected with EMR, the most of them could achieve complete eradication on a subsequent endoscopy by repeating the EMR.

The present study has several limitations. First, the study was retrospective in nature and was conducted at a single center. Endoscopic resection was performed by four endoscopists, and EMR or ESD procedures were chosen according to their subjective judgment, in the absence of definitive treatment guidelines for gastric dysplasia. The individual endoscopists may also have influenced clinical outcomes; although they were trained at the same institution, they differed in experience and preferences, such as the use of particular knives and electrocoagulation modes. Thus, the en bloc resection rate may vary depending on the endoscopist. Therefore, multicenter and prospective studies are needed to confirm our results. Second, the duration of follow-up was not constant. We analyzed data from patients who underwent endoscopic resection between January 2011 and December 2015 and were followed for at least 1 year after the procedure. Therefore, our observation periods ranged between 7 (long-term) and 3 years (short-term). Third, in the present study, although we investigated H. pylori infection and eradication treatments, the effects of the presence or absence of H. pylori and eradication were not clear. Thus, further studies are needed to determine their association with local recurrence. Fourth, accurate tumor size measurement is critical for selecting proper candidates for endoscopic resections (ER) of gastric neoplasia. However, size discrepancy between endoscopic size and pathologic size often occurs during ER for gastric tumor.

One strength of the present study is its focus on the endoscopic resection of gastric dysplasia, which is a pre-cancerous lesion, and its determination of the local recurrence rate according to the endoscopic resection method. Furthermore, the discrepancy in the pathological diagnosis before and after endoscopic resection of gastric dysplasia was investigated.

The present study showed that approximately 49.2% of HGD cases were upgraded histologically to adenocarcinoma after endoscopic resection. In addition, local recurrence rates were higher in patients with lesions > 2 cm, histologically incomplete resection. Therefore, in cases of gastric dysplasia with HGD or a lesion > 2 cm, ESD is preferred to EMR due to its higher incomplete resection rate, which can help prevent local recurrence and also EMR may be possible for low-grade dysplasia that is less than 2 cm without surface changes such as redness, depression and nodularity.