Repeat FDG-PET for predicting pathological tumor response and prognosis after neoadjuvant treatment in nonsmall cell lung cancer: comparison with computed tomography.

BACKGROUND The efficacy of fluorodeoxyglucose-positron emission tomography (FDG-PET) for predicting pathological tumor response and prognosis after neoadjuvant chemoradiotherapy followed by surgery in locally advanced nonsmall cell lung cancer (NSCLC) was compared to the predictive value of tumor size as determined by computed tomography (CT). METHODS Thirty-seven consecutive NSCLC patients who received FDG-PET and CT scans both before and after neoadjuvant chemoradiotherapy were enrolled in this study. The percentage point changes in maximum standard uptake value (SUV) on PET and tumor size on CT after neoadjuvant treatment were defined as the SUV ratio and the size ratio, respectively, and were compared with pathological tumor response and prognosis after surgery. A major pathological response was defined as residual viable tumor cells corresponding to less than one-third the size of the original tumor. RESULTS Nineteen and 18 patients showed major and minor pathological responses, respectively, after neoadjuvant treatment. The optimal cutoff values for predicting a major pathological response were 0.6 for the SUV ratio and 0.79 for the size ratio. The SUV ratio predicted the pathological tumor response with higher accuracy than the size ratio did (P = 0.04). Neither the SUV ratio nor the size ratio predicted prognosis after surgery. CONCLUSION For predicting the pathological tumor response after neoadjuvant chemoradiotherapy, the SUV ratio on FDG-PET is superior to the size ratio on CT in patients with NSCLC. However, neither the SUV ratio nor the size ratio could predict prognosis.