POCKET-DETECTION METHOD WITH REAL-TIME INVASION ASSESSMENT AND ENDOSCOPIC RESECTION (PRIME) ALGORITHM FOR RECTAL LESIONS: A PROSPECTIVE COHORT STUDY

Background and Aims The muscle-retracting sign (MRS) has shown moderate-to-high diagnostic accuracy in predicting deep submucosal invasion in rectal cancer. This study evaluated the technical outcomes of a diagnostic and therapeutic algorithm combining MRS evaluation and endoscopic resection for rectal lesions. Patients and Methods Prospective data from consecutive rectal lesions at risk of submucosal invasion at a tertiary centre between August 2023 and August 2025 were analyzed. An algorithm incorporating deep-margin optical diagnosis using the pocket-detection method (PDM) followed by endoscopic resection was applied. Results Among the 47 rectal lesions included, PDM was successful in all cases, with MRS detection in 34.0% (16/47) of the lesions. The technical success rate of endoscopic resection was 97.9% (46/47), with endoscopic submucosal dissection performed in 30 lesions (63.8%), endoscopic intermuscular dissection in 10 (21.3%), knife-assisted full-thickness resection in 5 (10.6%), and device-assisted full-thickness resection in 1 (2.1%) patient. En bloc, R0 resection, and R0 vertical margin rates were 100% (46/46), 93.5% (43/46) and 97.8% (45/46), respectively. Adverse events occurred in 10.6% (5/47) of patients (AGREE grade I, II and IIIa). 44.7% (21/47) of the included lesions were adenocarcinomas, and MRS accuracy for deep submucosal invasion was 85.1%. Outcomes in MRS-positive and MRS-negative lesions did not differ significantly, including technical success (93.7% vs 100%), en bloc resection (100% vs 100%), and R0 resection (93.5% vs 93.3%; all p > 0.05). Rectal preservation rate was 91.5% (43/47). Conclusion The PRIME algorithm demonstrated that deep margin optical diagnosis can reliably guide rectal endoscopic resection, yielding excellent radical resection rates and safety in both MRS-positive and MRS-negative lesions.