Publications Repository - Helmholtz-Zentrum Dresden-Rossendorf

Quantitative assessment of radiation- and chemotherapy response with 18F-FDG (fluorodeoxyglucose) whole body PET has attracted increasing interest in recent years. In most published work the standardized uptake value (SUV) is utilized for this purpose. In the context of therapy response assessment the reliability of lesion SUVs, notably their test-retest stability, thus becomes of crucial importance. However, in a recent study substantial test-retest variability (TRV) of SUVs was demonstrated. The purpose of the present study was to investigate if the tumor-to-blood standard uptake ratio (SUR) can improve test-retest variability of tracer uptake. Methods: 73 patients with advanced non small cell ling cancer (NSCLC) from the prospective multicenter trials ACRIN 6678 (N = 34) and Merck MK-0646-008 (N = 39) were included in this study. All patients underwent two 18F-FDG PET/CT investigations at two different days (time difference: (3.6 +/- 2.1) days (range: 1-7)) prior to therapy. For each patient up to seven tumor lesions were evaluated. For each lesion maximum and peak SUV was determined. Blood SUV was determined as mean value of a three-dimensional aorta ROI that was delineated in the attenuation CT and transferred to the PET image. SUR values were computed as ratio of tumor SUV and blood SUV and were uptake time corrected to 75 min p.i.. TRV was quantified as TRV = 1.96 x RMS, where RMS is the root mean square deviation of the fractional paired differences of SUV and SUR, respectively. The combined effect of blood normalization and uptake time correction was inspected by considering the ratio RTRV = TRVSUR /TRVSUV reflecting the reduction in test-retest variability of SUR relative to SUV. RTRV was correlated with the group averaged value dCFmean of the quantity dCF = |CF - 1|, where CF is the numerical factor that converts individual ratios of paired SUV values into corresponding SUR ratios. This correlation analysis was performed by successively increasing a threshold value dCFmin and computing dCFmean and RTRV for the remaining sub-group of patients/lesions with CF >= CFmin. Results: The group-averaged test-retest variability of SUV and SUR was TRVSUV=32.1 and TRVSUR=29.0, respectively, which corresponds to a reduction of variability of SUR by a factor RTRV = 0.9 in comparison to SUV. This rather marginal improvement can be understood as of consequence of the untypically low intra-subject variability of blood SUV and uptake time and accordingly small dCF values in the investigated prospective study groups. In fact, sub-group analysis with increasing thresholds dCFmin revealed a very pronounced negative correlation (Spearman's rho = -0.99, P<0.001) between RTRV and dCFmean where RTRV ~ 0.4 in the dCFmin = 20% sub-group, corresponding to a more than twofold reduction of TRVSUR compared to TRVSUV. Conclusion: Variability of blood SUV and uptake time have been identified as causal factors contributing to the test-retest variability of lesion SUV. Therefore, test-retest variability of lesion uptake measurements can be reduced when replacing SUV by SUR as the uptake measure. The improvement becomes substantial for the level of variability of blood SUV and uptake time typically observed in the clinical context.