OBJECTIVETo compare the clinical efficacy of endoscopic submucosal dissection (ESD) and surgical resection for early gastric cancer (EGC). Method The retrospective case-control study was conducted. Clinical and pathological data of 246 patients who were diagnosed as early gastric cancer and received ESD or surgery for EGC at Peking University First Hospital from 2010 to 2015 were collected, including 171 males and 75 females with average age of (63.6±10.8) years. Eighty-one patients received ESD according to the following indications: no peritoneal lymph node metastasis in preoperative CT; differentiated mucosal cancer without ulcer findings, irrespective of tumor size; differentiated mucosal cancer with diameter ≤30 mm and ulcer; differentiated minimal submucosal invasive cancer (≤500 μm from the muscularis mucosa) with diameter ≤30 mm and without ulcer; undifferentiated mucosal cancer with diameter ≤20 mm and without ulcer; high grade intraepithelial neoplasia with diameter >20 mm; mucous lesion recurrence after EMR without chance of EMR again. One-hundred and sixty-five cases received surgery according to the findings of peritoneal lymph node metastasis in preoperative CT or the growth of carcinoma beyond the expanding criteria of ESD (surgery group). En block resection rate (removing all the lesion at once) and curative resection rate [standard: negative horizontal and vertical surgical margins; negative vessel carcinoma embolus; pT1a and pT1b (SM1); lesion diameter <3 cm with differentiated type, pT1a with ulcer or pT1b(SM1); lesion diameter <2 cm with undifferentiated type, pT1a without ulcer] were compared between two groups. According to pathological results, including tumor location, macrographic type, size, pathological type, differentiated type, invasive depth, surgical margin, vessel carcinoma embolus and lymph node metastasis, all the patients meeting the ESD curative resection criteria in both groups were subgrouped in order to compare the baseline information, surgical conditions, postoperative complications, recovery, follow-up and survival. The end of follow up was December 2016.

RESULTSThe en block resection rate was 93.8%(76/81), while curative resection rate was 91.4% (74/81) in ESD group. The en block resection rate and curative resection rate was both 100% in surgery group. According to the pathological results, 170 cases were confirmed to be in accordance with the curative resection standard, including 74 cases in ESD group and 96 cases in surgery group. Subgroup analysis showed that ESD group had older cases (t=2.939, P=0.004) and more cases with lesion in upper 1/3 of stomach (χ=8.992, P=0.011), while no significant differences in tumor size, invasion depth, degree of differentiation (t=1.875, 2.393, 3.074, all P>0.05) were observed. Compared to surgery group, ESD group had significantly shorter operative time [(76.4±46.3) minutes vs. (271.9±92.6) minutes, t=17.950, P=0.000], shorter fasting period [(3.2±1.4) days vs. (8.8±5.4) days, t=9.801, P=0.000], shorter hospital stay [(9.0±5.8) days vs. (22.1±9.1) days, t=11.471, P=0.000], less costs [(2.6±2.2) ten thousand yuan vs (7.4±3.0) ten thousand yuan, t=12.235, P=0.000] and lower morbidity of early-stage postoperative complication [1.4%(1/74) vs. 20.8%(20/96), χ=14.502, P=0.013]. One-hundred and sixty-two of 170 patients (95.3%) were followed up for median time of 28 months (range, 11 to 84 months). The recurrence rate was 2.7% (2/74) in ESD group and 4.2% (4/96) in surgery group respectively without significant difference(χ=1.787, P=0.409). Five-year overall survival rate was 97.5% and 96.5% respectively without significant difference as well (χ=0.115, P=0.735).

CONCLUSIONSESD is an effective and safe treatment of early gastric cancer. It can be used as the first protocol for well-differentiated mucosal or SM1 EGC without ulcer as well as undifferentiated mucosal EGC with diameter less than 2 cm.