Long-term outcomes of ESD versus EMR for early-stage colorectal cancer and precancerous lesions: a meta-analysis

The management of early-stage colorectal cancer (Early CRC) and precancerous lesions (PL) has evolved towards minimally invasive endoscopic techniques. While endoscopic mucosal resection (EMR) remains standard, it often fails to achieve en bloc resection for large lesions, increasing recurrence risk. Endoscopic submucosal dissection (ESD) was developed to overcome this limitation. This meta-analysis compares the long-term efficacy and safety of ESD versus EMR for Early CRC and PL. We systematically searched PubMed, Embase, Web of Science, and the Cochrane Library for observational studies from inception to August 2025. Study selection, data extraction, and quality assessment using the Newcastle–Ottawa Scale were performed independently by two reviewers. Outcomes were pooled as the logarithm of the odds ratio (log OR) with 95% confidence intervals (CI) using RevMan 5.3 and Stata 18.0. Heterogeneity, publication bias, and robustness were assessed. Thirteen studies involving 3,250 patients were included. ESD significantly reduced recurrence risk (log OR = -1.92, 95% CI [-2.74, -1.10], P < 0.001) and achieved higher en bloc (log OR = 2.45, 95% CI [2.03, 2.88], P < 0.001) and complete resection rates (log OR = 1.79, 95% CI [0.68, 2.90], P = 0.002) compared to EMR. ESD carried a higher perforation risk (log OR = 1.42, 95% CI [0.81, 2.02], P < 0.001), while the delayed bleeding analysis was inconclusive (log OR = -0.29, 95% CI [-0.76, 0.19], P = 0.24), with sensitivity analysis showing directional reversal upon exclusion of a single study. Subgroup analyses confirmed ESD's superiority in both Early CRC and PL. Results were robust in sensitivity analysis, with minimal publication bias. ESD was associated with superior long-term efficacy for Early CRC and PL, with lower recurrence and better resection outcomes than EMR, despite a higher perforation risk. However, given the observational design of the included studies, these findings may be influenced by confounding by indication, as ESD was preferentially used for larger and more complex lesions. Furthermore, the evidence base is predominantly derived from Asian centers, and applicability to Western populations should be interpreted with caution. The evidence regarding delayed bleeding is inconclusive and requires cautious interpretation. Treatment should be individualized based on lesion characteristics and patient factors.