Abstract:Purpose To characterize physiologic uptake of 18-F FDG in children undergoing PET/CT as a step to informing efforts to optimize FDG PET image quality in children. Methods This retrospective study included 193 clinically indicated 18-F FDG PET/CT examinations from 139 patients. 3D spherical regions of interest (ROIs) in the liver and in the thigh muscle (an area of uniform low-level uptake) were used to measure counts and mean standardized uptake value by body weight (SUVmean-bw). Counts, SUVs, and liver signal to noise ratio (SNR) were assessed for associations with patient-specific predictor variables using Pearson correlation and multivariable linear regression. Results Mean patient age was 11.0 ± 5.4 (SD) years, mean liver SUVmean-bw was 1.17 ± 0.60 and mean liver counts was 5387 ± 1875 Bq/mL. On univariable analysis liver SUVmean-bw and liver counts were strongly correlated with weight ( r = 0.87, p < 0.0001), age ( r = 0.75, p < 0.0001) and total injected activity ( r = 0.85, p < 0.0001). Mean thigh counts were significantly associated only with injected activity/kilogram ( r = 0.37, p < 0.0001). On multivariable analysis, body weight and age (which is collinear with body weight) were the only significant independent predictors ( p < 0.0001). Liver SNR was moderately associated with all predictors apart from injected activity per kilogram ( r = 0.09, p = 0.23). Conclusion Liver counts on 18-F FDG PET/CT have a significant positive association with age and body weight. However, liver SNR has no significant association with injected activity per kilogram suggesting that increasing dose per kilogram may not improve image quality in young children.

Patient factors affecting 18-F FDG uptake in children