Background and aims: endoscopic submucosal dissection (ESD) is the standard treatment for differentiated early gastric cancer (EGC). However, its expanded indication for undifferentiated EGC is controversial. In this study, the efficacy and safety of ESD versus surgery in patients with undifferentiated EGC were compared.
Methods: four databases were searched until February 24, 2022, for studies assessing patients with undifferentiated EGC that met an expanded indication for endoscopic resection and who were treated with ESD or surgery. Primary outcomes were all-cause mortality and any recurrence. Secondary outcomes were complete histological resection, local recurrence, metachronous recurrence, synchronous recurrence, distant metastasis, overall complication, and bleeding. All meta-analyses were performed using a random-effects models. Unadjusted (risk ratio [RR]) and adjusted (aRR and hazard ratio [aHR]) estimates with 95 % confidence interval (CI) were calculated.
Results: seven cohort studies were included (n = 2637). The use of ESD was significantly associated with higher all-cause mortality compared to surgery (RR, 2.17; 95 % CI, 1.24-3.81); adjusted all-cause mortality effects were not significant (aRR, 2.28; 95 % CI, 0.95-5.47 and aHR, 1.97; 95 % CI, 0.85-4.53). ESD was associated with a higher risk of any recurrence using unadjusted (RR, 5.24; 95 % CI, 1.49-18.46) and adjusted (aRR, 7.89; 95 % CI, 1.52-40.95 and aHR, 3.73; 95 % CI, 1.17-11.90) estimates. The risk of local recurrence, synchronous recurrence and bleeding were significantly higher for ESD versus surgery. No significant differences were found for other secondary outcomes.
Conclusions: although ESD is associated with a higher risk of any recurrence, adjusted all-cause mortality is similar during follow-up. Overall complications were similar between ESD and surgery.