Background: Endoscopic resection is widely used as an effective treatment for superficial esophageal cancer. However, the risk of lymph node metastasis increases in cases of muscularis mucosae or deeper invasion, for which additional treatment such as radiotherapy or surgery is required. Accordingly, the current study investigated the efficacy and toxicity of salvage radiotherapy after non-curative endoscopic resection as an organ preservation strategy. Methods: We retrospectively reviewed 37 esophageal cancer patients who received salvage radiotherapy after non-curative endoscopic resection. The pathological invasion depths were the muscularis mucosae, submucosal layer, and muscularis propria in 14, 22, and one patient, respectively. All patients received external beam radiotherapy. Among them, eight received intraluminal brachytherapy following external beam radiotherapy. Elective nodal irradiation was administered to all patients. Twenty-five patients received concurrent platinum and fluorouracil-based chemotherapy. Results: The median follow-up time was 74 months (range: 3–212). The 5-year progression-free survival and overall survival rates were 64 and 78%, respectively. No local or regional lymph node recurrence was observed. The causes of death included esophageal cancer in one patient, metachronous esophageal cancer in one patient, other malignancies in eight patients, and other causes in six patients. Late cardiac toxicities ≥ grade 3 were observed in six patients, one of whom died of arrhythmia. Conclusions: Salvage radiotherapy after non-curative esophageal endoscopic resection is an effective treatment as an organ preservation strategy. Although muscularis mucosae and submucosal cancer have a high risk of lymph node metastasis, our results suggest that elective nodal irradiation contributes to reduced regional node metastases.

Background/Aim: Surgical resection and chemoradiotherapy (CRT) can be performed as additional treatments for superficial esophageal cancer after endoscopic resection, but the selection criteria vary depending on the institution. We retrospectively evaluated the outcomes of patients with endoscopically resected superficial esophageal cancer treated with surgical resection and CRT at our institution. Patients and Methods: The outcomes of 67 cases of additional treatment after endoscopic resection of superficial esophageal cancer, excluding adenocarcinoma, performed at our hospital from January 2000 to June 2017 were compared (30 cases in the surgery group and 37 cases in the CRT group). Results: In the surgery group, eight patients had lymph node metastasis and two had recurrence in the supraclavicular fossa lymph nodes after surgery, therefore reoperation was performed. There were no deaths from esophageal cancer, and the 5-year survival rate was 92.6%. One patient in the CRT group had a recurrence in the cervical paraoesophageal lymph node, which was resected, but no death from esophageal cancer was observed, and the 5-year survival rate was 81.0%. The 5-year survival rate was significantly better in the surgery group than in the CRT group (p=0.039). The greater number of elderly patients in the CRT group was considered to be the reason for the worse prognosis, Conclusion: Although the prognosis of esophagectomy or chemoradiotherapy (CRT) is very favorable, CRT is considered to be the preferred additional treatment after endoscopic resection of superficial esophageal cancer without lymph node metastasis from the viewpoint of organ preservation.

Background: This study aimed to propose a machine learning model to predict the local response of resectable locally advanced esophageal squamous cell carcinoma (LA-ESCC) treated by neoadjuvant chemoradiotherapy (NCRT) using pretreatment 18-fluorodeoxyglucose positron emission tomography (FDG PET) images. Methods: The local responses of 98 patients were categorized into two groups (complete response and noncomplete response). We performed a radiomics analysis using five segmentations created on FDG PET images, resulting in 4250 features per patient. To construct a machine learning model, we used the least absolute shrinkage and selection operator (LASSO) regression to extract radiomics features optimal for the prediction. Then, a prediction model was constructed by using a neural network classifier. The training model was evaluated with 5-fold cross-validation. Results: By the LASSO analysis of the training data, 22 radiomics features were extracted. In the testing data, the average accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve score of the five prediction models were 89.6%, 92.7%, 89.5%, and 0.95, respectively. Conclusions: The proposed machine learning model using radiomics showed promising predictive accuracy of the local response of LA-ESCC treated by NCRT.