Background: Esophageal cancer continues to increase in incidence. Many patients are presenting with stage II or greater disease and proceeding to neoadjuvant chemoradiation therapy before resection. Approximately 30% of patients will achieve a complete response and might not benefit from proceeding to resection. This study will examine the ability of PET to predict patients with a complete pathologic response.
Study design: A query of our IRB-approved esophageal database revealed 81 patients who underwent a pre- and postchemoradiation PET scan and then proceeded to esophageal resection. Statistical analysis was performed to determine the ability of PET to predict a complete pathologic response.
Results: When comparing posttherapy PET with final pathology, it was determined that PET could not consistently differentiate a complete pathologic response from patients who still had persistent disease. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 61.8%, 43.8%, 70%, 35%, and 56%, respectively, for patients with a complete PET response after neoadjuvant therapy.
Conclusions: A complete PET response after neoadjuvant chemoradiation is not substantially predictive of a complete pathologic response. Patients should still be referred for resection unless distant metastases are identified.