Validation of the optimal index for monitoring early tumor changes between two scans in FDG PET/CT

Objectives Quantitative analysis is provided new ways of assessing response in specific tumor type, largely guided by the new positron emission tomography response criteria in solid tumors (PERCIST). The aim of this study was to evaluate the treatment response with PERCIST and to validate the optimal index for monitoring early tumor changes between two scans.

Methods Response to therapy expressed as a percentage change in standardized uptake value (SUV) corrected for lean body mass (SUL) peak in PERCIST. 60 patients with malignant tumor underwent 2 PET/CT scans. Percentage changes in SULpeak according to PERCIST were compared with these in lesion size (short axis) on CT. The intraclass correlation coefficients (ICC) were calculated to assess reproducibility of PERCIST. Furthermore, the usefulness of the following indices for monitoring early tumor changes between pre- and post chemotherapy was assessed by using receiver operating-characteristic (ROC): SUVmax, the metabolic volume (MV), the total lesion glycolysis (TLG) and SULpeak in PERCIST.

Results Percentage changes in SULpeak of all lesions were significantly higher than these in size. Changes in size may have underestimated response to therapy. The reproducibility of PERCIST analysis (ICC=0.95) was excellent. The areas under the ROC curves were 0.65, 0.62, 0.82 and 0.91 for SUVmax, MV, TLG and SULpeak, respectively. In monitoring the response, TLG and SULpeak were more reliable than SUVmax and MV.

Conclusions Metabolic response of PET with PERCIST may be more predictive of outcome than only morphological measures because it is uniformity of measurement and reproducibility of results. Our results showed that SULpeak in PERCIST was the most optimal index in the prediction for therapeutic effect