Abstract
PURPOSE: We examined survival of patients who underwent esophagectomy for locally advanced esophageal cancer with foci that were suspicious for metastatic disease on initial imaging but whose disease did not progress after induction chemoradiation treatment (CRT). METHODS: The impact of pre- and posttherapy staging characteristics on survival of patients who underwent esophagectomy after CRT between 2003 and 2009 was evaluated using multivariable logistic regression. Survival of patients with and without possible metastatic disease on initial imaging was compared with the log-rank test. RESULTS: During the study period, 71 (32%) of 220 patients who underwent CRT followed by esophagectomy had possible distant metastatic disease on initial imaging. Patients with initial suspicion of metastases had a 5-year survival of 24.8%. Overall survival of patients with and without possible metastatic disease on initial imaging was not significantly different (p = 0.4), but pretreatment positron emission tomography (PET) suggesting a liver lesion (hazard ratio [HR] 3.2, p = 0.003) predicted worse survival. Additional predictors of worse survival were clinical T4 status (HR 3.1, p = 0.001), post-CRT pathologic nodal status (HR 1.6, p = 0.04), and pathologically confirmed metastatic disease at or before resection (HR 3.1, p = 0.01). None of 10 patients with pathologic metastatic disease at resection lived longer than 2.5 years. CONCLUSIONS: Patients with possible liver metastases on pretreatment PET and patients with confirmed metastatic disease at the time of surgery do not benefit from resection. However, patients with pretreatment imaging that shows possible metastatic disease in sites other than the liver still have reasonable long-term survival after resection.