Abstract
1625
Objectives Various methods for quantification of FDG PET studies exist, ranging from kinetic analysis to standardized uptake values (SUV). In response monitoring studies changes in plasma glucose level, fractional blood volume and arterial input function affect SUV. In contrast, they are accounted for in full kinetic analysis. The purpose of this study was to investigate the need for validating SUV for response monitoring purposes.
Methods Baseline and 2 response dynamic FDG scans were performed in 6 subjects with advanced gastrointestinal malignancies within a phase I clinical trial. Quantification of tumour uptake included full kinetic methods, using both non-linear regression (NLR) and Patlak analyses, and simplified measures such as the simplified kinetic method (SKM) and SUVBW or SUVBSA with and without plasma glucose correction.
Results 5 out of 6 subjects showed a decrease in SUVBW (-38% to +14%). Responses observed with Patlak and NLR correlated well (R2=0.99, slope=1.04±0.03). Responses seen with SUVBW and SUVBSA were significantly different (p<0.05) from that with Patlak with regression slopes of 0.52±0.08 and 0.51±0.08. Similar results were seen for SUV corrected for plasma glucose. Responses found with SKM were 41±13% lower than those with Patlak (i.e. slope=0.59, p<0.01). Further analysis revealed that these differences in response were primarily due to changes in arterial input functions.
Conclusions SUV responses may differ substantially from those seen with full kinetic analysis due to changes in delivery of FDG to the tumor following therapy. Underestimation of treatment response using SUV may result in underestimation of antitumor activity of a new drug.
- © 2009 by Society of Nuclear Medicine