Objective: To evaluate the sites and frequencies of overall and initial lymph node (LN) metastases (LNMs) of clinical T1N0 esophageal cancer.
Background: The sites and frequencies of initial LNMs and sentinel LNs (SLNs) of esophageal cancer remain unclear.
Methods: The Japan Clinical Oncology Group JCOG0502 trial was a 4-arm prospective study that compared esophagectomy with chemoradiotherapy for clinical T1N0 esophageal cancer in both randomized and patient-preference arms. The preoperative diagnostic accuracy was evaluated for patients assigned to the surgery arm. Patients who withdrew consent and who were not treated were excluded. All patients underwent esophagectomy with D2 or greater LN dissection. From the pathologic findings, sites and frequencies of LNMs and SLNs were assessed and the frequency of skip LNMs was calculated.
Results: In total, 211 patients underwent LNM and SLN analysis. Regarding N-factor accuracy, 57 (27.0%) of 211 clinical N0 cases had pathologic LNMs. The upper mediastinal and mediastinal/abdominal regions were frequent sites of LNMs in upper and lower thoracic cases, respectively. However, in middle thoracic cases, LNMs were observed in the neck, mediastinal, and abdominal regions, and pathologic SLN spread to all 3 fields. The frequency of skip LNMs was 36.7%.
Conclusions: A clinical diagnosis of T1N0 is not sufficiently accurate, and therefore, it is unacceptable to omit LN dissection or minimize the prophylactic radiation field. SLNs, which are not location restricted, should be surveyed in all 3 fields.