Prognostic factors for ESD of early gastric cancers: a systematic review and meta-analysis.

Gastric neoplasms are one of the leading types of cancer in the world and early detection is essential to improve prognosis. Endoscopy is the gold-standard diagnostic procedure and allows adequate treatment in selected cases. Endoscopic submucosal dissection (ESD) has been reported to safely address most early gastric cancers (EGCs), with high curability rates. However, data on prognostic factors related to ESDs of EGCs are conflicting. Therefore, we aimed to systematically review the available literature and to perform a meta-analysis to identify the relevant prognostic factors in this context. We performed this study according to PRISMA guidelines. Comparative studies assessing the relationship between curative resection or long-term curability rates and relevant prognostic factors were selected. Prognostic factors were demographic data, lesion features (location, morphology of the lesion, size, and depth of invasion), histological findings, (HP) infection, presence of gastric a atrophy and body mass index (BMI). Finally, we also evaluated risk factors related to metachronous gastric neoplasm. The initial search retrieved 2829 records among which 46 studies were included for systematic review and meta-analysis. The total sample comprised 28366 patients and 29282 lesions. Regarding curative resection, pooled data showed no significant influence of sex [odds ratio (OR): 1.15 (0,97, 1.36)  = 0.10 I  = 47 %] , age [OR: 1.00 (0.61, 1.64)  = 1.00 I  = 58 %], posterior vs non-posterior location [OR: 1.35 (0.81, 2.27)  = 0.25 I  = 84 %], depressed vs von-depressed macroscopic type[OR: 1.21 (0.99, 1.49)  = 0.07 I  = 0 %], non-upper vs upper location [OR: 1.41 (0.93, 2.14)  = 0.10 I  = 77 %] and BMI [OR: 0.84 (0.57; 1.26)  = 0.41 I  = 0 %]. Differentiated neoplasms presented greater chance of cure compare to undifferentiated [OR: 0.10 (0.07, 0.15)  < 0.00001 I  = 0 %]. Ulcerated lesions had lower curative rates compared to non-ulcerated [OR: 3.92 (2.81, 5.47)  < 0.00001 I  = 44 %]. Lesions smaller than 20 mm had greater chance of curative resection [OR: 3.94 (3.25, 4.78)  < 0.00001 I  = 38 %]. Bleeding during procedure had lower curative rates compared to non-bleeding [OR: 2.13 (1.56, 2.93)  < 0.0001 I  = 0 %]. Concerning long-term cure, female gender [OR 1.62 (1.33, 1.97)  < 0.00001 I  = 0 %] and the mucosal over SM1 cancers were protective factors [OR: 0.08 (0.02, 0.39)  = 0.002 I  = 86 %]. Gastric atrophy [OR: 0.60 (0.45, 0.81)  = 0.0006 I  = 42 %] and the pepsinogen I/pepsinogen II ratio [OR 2.29 (1.47, 3.57)  = 0.0002 I  = 0 %] were risk factors to metachronous gastric neoplasm. Ulcerated lesions, histology, bleeding and size > 20 mm are prognostic factors concerning curative resection. Regarding long-term cure, female gender and mucosal over SM1 cancer are predictive factors. Gastric atrophy and the pepsinogen ratio are risk factors for metachronous gastric neoplasm.