Accuracy of 18F-FDG PET/CT in predicting residual disease after neoadjuvant chemoradiotherapy for esophageal cancer

Abstract

Our purpose was to prospectively investigate optimal evaluation of qualitative and quantitative 18F-FDG PET/CT in response evaluations 12–14 wk after neoadjuvant chemoradiotherapy (nCRT) in esophageal cancer patients. Methods: This was a side study of the prospective diagnostic pre-SANO trial. 18F-FDG PET/CT scans at baseline and at 12–14 wk after nCRT were qualitatively assessed for the presence of tumor. Maximum SUVs normalized for lean body mass (SULmax) were measured in all scans. The primary endpoint was the proportion of false-negative patients with tumor regression grade (TRG) 3–4 (.10% vital residual tumor) in qualitative and quantitative analyses. Receiver-operating-characteristic curve analysis for TRG1 versus TRG3–4 using SULmax, SULmax tumor-to-esophagus ratio, and D%SULmax was performed to define optimal cutoffs. Secondary endpoints were sensitivity, specificity, negative predictive value, and positive predictive value for TRG1 versus TRG2–4. Results: In total, 129 of 219 patients were analyzed. Qualitative 18F-FDG PET/CT was unable to detect TRG3–4 in 15% of patients. Sensitivity, specificity, negative predictive value, and positive predictive value in qualitative analysis for detecting TRG1 versus TRG2–4 was 80%, 37%, 42%, and 77%, respectively. In 18 of 190 patients (10%) with follow-up scans after nCRT, 18F-FDG PET/CT identified new interval metastases. Quantitative parameters did not detect TRG3–4 tumor in 27%–61% of patients. The optimal cutoff for detecting TRG1 versus TRG2–4 was a post-nCRT SULmax of 2.93 (area under receiver-operating-characteristic curve, 0.70). Conclusion: Qualitative and quantitative analyses of 18F-FDG PET/CT are unable to accurately detect TRG3–4 and to discriminate substantial residual disease from benign inflammation-induced 18F-FDG uptake after nCRT. However, 18F-FDG PET/CT is useful for the detection of interval metastases and might become useful in an active surveillance strategy with serial 18F-FDG PET/CT scanning.