Objectives: Although dual-time-point scans have been widely used to improve the diagnostic efficacy of FDG PET in differentiating between malignant and benign lesions, no optimized delayed scan time-point has yet been recommended in clinical practice. Our study aimed to explore the most appropriate time for a delayed scan by comparing early and late delayed scans.
Methods: Eighty patients with suspected malignancy were given a three-phase (64 min, 110 min, 233 min after FDG injection) PET/CT scan. The maximum standardized uptake values (SUVs) in the three-phase scans were recorded as SUV1, SUV2 and SUV3, respectively, and compared among three-phase imaging. Retention indices (RIs) of each lesion in two delayed phases were calculated according to the formulae: RI1=SUV2-SUV1/SUV1 x100% and RI2=SUV3-SUV1/SUV1 x100%. RI1 and RI2 in both malignant and benign groups were assessed through correlation analysis. The diagnostic values of two delayed scans were compared through the analysis of the receiver operating characteristic curves.
Results: One hundred and nine of 148 lesions were malignant, and 39/148 lesions benign, which were verified by pathological, clinical, laboratory or radiological examination. RI1 and RI2 in malignancy were 14.8+/-13.1% and 10.8+/-20.5% respectively, and the correlation coefficient was 0.6 (P=0.0001). RI1 and RI2 in benign lesions were 11.3+/-28.2% and 9.3+/-42.4%, respectively, and the correlation coefficient was 0.6 (P=0.0001). The area under the ROC curve for RI1 was 0.627+/-0.050 (null hypothesis: true area=0.5, P=0.0130); whereas the area under the ROC curve for RI2 was 0.563+/-0.052 (null hypothesis: true area=0.5, P=0.2321), suggesting that the late delayed scan may have no diagnostic value.
Conclusion: The retention index values in the two delayed phases have good relativity. The diagnostic value of early delayed imaging is higher than that of late delayed imaging. An early delayed scan, according to our research, should be recommended in clinical practice.