Long-Term Outcome After Endoscopic Resection for Esophageal Squamous Cell Carcinoma Invading the Muscularis Mucosae Without Lymphovascular Invasion

Abstract Background Patients with esophageal squamous cell carcinoma (ESCC) treated with endoscopic resection (ER) are generally observed without additional treatment, if they are diagnosed with pT1a-MM (muscularis mucosae) without lymphovascular invasion (LVI), in accordance with community standards. Although previous reports indicate a metastatic recurrence rate of approximately 5.0% during follow-up, these studies have limitations, including insufficient follow-up duration and non-standardized pathological diagnoses. This study aimed to clarify the long-term outcomes of these patients and the significance of central pathological diagnosis (CPD). Methods Patients who underwent ER for ESCC between 2008 and 2015 and were diagnosed with pT1a-MM without LVI were retrospectively recruited from 47 institutions affiliated with the Gastrointestinal Endoscopy Study Group in the Japan Clinical Oncology Group. We evaluated the cumulative recurrence rate (CRR), overall survival (OS), recurrence-free survival (RFS) based on additional treatments after ER (observation, chemoradiotherapy, and surgery). CRR included both lymph node and distant recurrence as events. In the observation group, CPD was performed to assess invasion depth and LVI, and CRR was analyzed according to each CPD result. Results A total of 540 patients were enrolled. Among them, 485 (89.8%) underwent observation, while 45 (8.3%) received chemoradiotherapy and 10 (1.9%) underwent surgery as additional treatment. Over a median follow-up of 7.0 years, the 5-year CRR/OS/RFS of the observation, chemoradiotherapy, and surgery groups were 4.9%/92.7%/89.6%, 2.2%/93.3%/91.1%, and 10.0%/100%/90.0%. Among 31 patients with metastatic recurrence, 18 died from ESCC. CPD in 476 cases reclassified 75.2% as pT1a-MM without LVI, 11.8% as pT1a-LPM without LVI, and 13.0% as pT1a with LVI or pT1b-SM, with corresponding 5-year CRR rates of 4.9%, 1.8%, and 8.3%. Conclusion This study revealed that CRR in this population was not negligibly low, however, the current observation strategy was considered acceptable. Additionally, CPD facilitated the rational stratification of CRR.