Dental floss clip traction (DFC) has been introduced to enhance visualization and shorten dissection time during endoscopic submucosal dissection (ESD) and endoscopic full-thickness resection (EFTR) for gastric neoplasms. However, its overall benefit and modifying factors remain uncertain. This study systematically evaluated the efficacy and safety of DFC-ESD and DFC-EFTR compared with conventional techniques (C-ESD and C-EFTR) and explored moderators using meta-regression and trial sequential analysis (TSA).

Following PRISMA guidelines and the Cochrane Handbook, we searched PubMed, Scopus, Ovid, Web of Science, Cochrane Central, and CNKI to October 18, 2025. Randomized and observational studies comparing DFC-ESD or DFC-EFTR with conventional techniques were eligible. Pooled estimates were computed in R (meta 7.0-0) using random-effects models, reporting mean differences (MD) or risk ratios (RR) with 95% CIs. Prespecified meta-regression assessed lesion site, ulceration, tumor size, and study design as potential effect modifiers. Trial sequential analysis determined conclusiveness and required information size. Study quality was assessed by ROB-2 and Newcastle–Ottawa Scale, and the certainty was graded via GRADE.

We included eight studies with a total of 2,183 patients. Five studies on DFC-ESD and three on DFC-EFTR. For the DFC-ESD vs C-ESD, the overall pooled procedure time showed a nonsignificant trend favoring DFC-ESD (MD = −5.4 min, 95% CI [−13.8, 2.9]; p = 0.20; I² = 79%). Subgroup analysis revealed significant reductions for non-ulcerated lesions (MD = −5.0 min; p < 0.01) and for lesions in the upper/middle greater curvature, posterior wall, lower anterior, and lower posterior wall (all p < 0.05). Meta-regression identified the lesion site as a significant moderator (p = 0.02), whereas ulceration and tumor size were not (p > 0.10).Rates of En bloc (RR = 1.00; p = 0.66), R0 (RR = 1.01; p = 0.42), complete resection (RR = 1.00; p = 0.60), perforation (RR = 0.59; p = 0.21), and delayed bleeding (RR = 1.17; p = 0.54) were comparable between groups. TSA confirmed sufficient evidence for the reduction in procedure time among non-ulcerated and posterior-wall lesions but inconclusive power for safety outcomes. For the DFC-EFTR vs C-EFTR, the analysis demonstrated shorter operative time with DFC-EFTR (MD = −8.3 min; 95% CI [−10.4, −6.2]; p = 0.03; I² = 0%). However, Hospital stay (MD = −0.01 days; p = 0.93) and pneumoperitoneum (RR = 0.46; p = 0.26) did not differ. TSA confirmed conclusiveness for operation-time reduction but suggested further evidence is required for secondary endpoints.

DFC-ESD was associated with shorter operation times, particularly in non-ulcerated lesions and those located at the upper/middle greater curvature, upper/middle posterior wall, lower anterior wall, and lower posterior wall. Similarly, DFC-EFTR demonstrated reduced procedure times compared to C-EFTR. However, given the novelty of the DFC technique, its performance may vary across gastric sites, and certain limitations remain. Therefore, further large-scale, high-quality RCTs are warranted to confirm its efficacy, clarify its optimal indications, and ensure consistent safety outcomes.