Variation of liver SUV on (18)FDG-PET/CT studies in women with breast cancer.

PURPOSE: It has been proposed to use liver as reference organ when assessing tumor response with PET. We investigated whether liver FDG uptake is dependent on clinical, biological, or technical factors.

METHODS

Sixty-one women with breast carcinoma had FDG-PET/CT imaging at baseline and then after the second course of neoadjuvant chemotherapy. SUV values were measured over the primary breast tumor and in a region of interest of 30-mm diameter within the liver.

RESULTS

Liver SUVmean values for the 61 women were not different before and after the 2 courses of chemotherapy (2.27 vs 2.35; P = 0.18). No significant correlation was found between the delay postinjection and liver SUV. Liver SUVmean increased with blood glucose level (P = 0.007 on baseline PET studies, and P = 0.067 on postchemotherapy studies). Liver SUVmean increased significantly with increasing patient age, which was observed on baseline (P = < 0.001) and postchemotherapy studies (P = 0.002). Again, liver SUVmean increased with body weight (P = 0.037 on baseline studies and 0.003 on postchemotherapy studies). On multivariate analysis, 2 independent variables remained correlated to liver SUV, namely, age and weight. We found no statistically significant correlation between the studied physiological variables and breast tumor SUVmax. However, there was a trend for lower SUVmax with increasing blood glucose level or age.

CONCLUSIONS

FDG liver uptake depends on many factors such as age, weight, and blood glucose level. Some of these variables seem to have an opposite impact on tumor SUVmax. These findings challenge the usefulness of liver as reference.