PURPOSE: This study explored predictive factors that affected oncologic outcomes after surgical resection or follow-up without surgery in patients with submucosally invasive colorectal carcinomas after endoscopic resection. METHODS: Oncologic outcomes in terms of lymph node metastasis or tumor recurrence were assessed according to resection margin, histology, and depth of invasion. RESULTS: Eighty-seven patients with submucosally invasive colorectal carcinomas after endoscopic resection were followed prospectively. Fifty-seven (65.5 percent) patients had risk factors of deep submucosal invasion and/or unfavorable histology. Among them, 30 underwent radical resection, and 6 patients had lymph node metastases. Twenty patients with risk factors were closely followed up and 3 recurrent carcinomas were detected. Ultimately, 9 of 57 high-risk patients (15.8 percent) exhibited lymph node metastasis or tumor recurrence. Among 30 patients without risk factors, none had lymph node metastasis or recurrent carcinoma. Univariate analysis showed that tumor budding (P = 0.003) and venous invasion (P = 0.021) were factors for lymph node metastasis. In multivariate analysis, only tumor budding was an independent predictor of lymph node metastasis (P = 0.026). CONCLUSIONS: Approximately 16 percent of patients with submucosally invasive colorectal carcinoma and risk factors benefited from subsequent surgery. Tumor budding was the most significant factor for lymph node metastasis. Observation would be appropriate for patients without risk factors after endoscopic resection.
PURPOSE: This study explored predictive factors that affected oncologic outcomes after surgical resection or follow-up without surgery in patients with submucosally invasive colorectal carcinomas after endoscopic resection. METHODS: Oncologic outcomes in terms of lymph node metastasis or tumor recurrence were assessed according to resection margin, histology, and depth of invasion. RESULTS: Eighty-seven patients with submucosally invasive colorectal carcinomas after endoscopic resection were followed prospectively. Fifty-seven (65.5 percent) patients had risk factors of deep submucosal invasion and/or unfavorable histology. Among them, 30 underwent radical resection, and 6 patients had lymph node metastases. Twenty patients with risk factors were closely followed up and 3 recurrent carcinomas were detected. Ultimately, 9 of 57 high-risk patients (15.8 percent) exhibited lymph node metastasis or tumor recurrence. Among 30 patients without risk factors, none had lymph node metastasis or recurrent carcinoma. Univariate analysis showed that tumor budding (P = 0.003) and venous invasion (P = 0.021) were factors for lymph node metastasis. In multivariate analysis, only tumor budding was an independent predictor of lymph node metastasis (P = 0.026). CONCLUSIONS: Approximately 16 percent of patients with submucosally invasive colorectal carcinoma and risk factors benefited from subsequent surgery. Tumor budding was the most significant factor for lymph node metastasis. Observation would be appropriate for patients without risk factors after endoscopic resection.