Pharmacokinetic modeling of glucose uptake and retention on 18F-FDG-PET imaging of irradiated bone

Objectives Some recent approaches to radiotherapy are employing fewer treatments but higher doses of radiation per treatment. Radiation therapy can result in decreased bone metabolism and density with subsequent fracture. This study examined pharmacokinetics of the FDG uptake in PET imaging to quantify changes in glucose metabolism caused by spinal radiosurgery in a mini-pig model. Specifically, we tested the hypothesis that FDG uptake is inversely related to radiation dose in porcine bone.

Methods Approximately 1 year-old female Yucatan mini-pigs were enrolled in a 3-arm design (n=4-5/arm) to receive spinal sterotactic radiosurgery (SRS) uniform doses of 16 Gy ±5% , 20 Gy ±5% or 24 Gy ±5% to C5-C7. Each animal underwent three PET/CT scans on a Siemens 64-slice Biograph at pre-SRS, 4 weeks post-SRS, and 3 months post-SRS. Due to neurotoxicity, the high dose group was terminated at the 3 month time point. Medium and low dose groups received two additional scans at 6 and 12 months post-SRS. PET-CT images were generated using the manufacturer’s installed software package. The PET uptake data was generated using whole bone ROIs for C4-T1 and pharmacokinetic modeling performed using an aortic image derived input function and a software package available on the Siemen’s Inveon Research Workplace 4.0 suite.

Results Curiously, Chi square analysis of compartment modeling revealed the best fit using a reversible model. Ki and VD values decreased with time for all vertebral bodies, but were greatest in C5-7 irrespective of the model. There was a statistically significant decline in Ki and VD that increased in magnitude over time. The decline was more profound in the medium dose group, but surprisingly not significant in the high dose group.

Conclusions In our study the FDG uptake in irradiated bone was best fit by reversible models. Irradiated bone shows a dose dependent decline in FDG uptake that worsens over a 1 year time period.