Abstract
1521
Objectives To evaluate the effect of using liver as compared to muscle as the internal reference tissue when calculating the hypoxic subvolume in chordomas with 18F-FMISO PET/CT. The degree of hypoxia in these tumors is hypothesized to correlate with increased probability of radioresistance, local recurrence, and metastasis, and the hypoxic subvolume can potentially be treated with a higher dose of radiation.
Methods Twenty patients with chordoma of the mobile or sacrococcygeal spine were scanned with 18F-FMISO PET/CT at two time points: 1) before radiation therapy and 2) after 19.8 - 34.2 GyRBE (relative biologic effectiveness). SUVmean was measured in liver and muscle (gluteal or paraspinal) for each scan. For one patient at the first time point, the liver was not included on the PET scan and thus not available for analysis. Hypoxic subvolumes were determined using these SUVmean measurements for muscle and liver as internal reference standards.
Results The average SUVmean for muscle and liver across all scans was 1.60 and 2.83 (p < 0.0001) with similar relative standard deviations of 23% and 22%, respectively. This relative standard deviation of approximately twenty percent highlights the variability in SUV values between patients/scans and demonstrates the need for an internal reference standard. Muscle is currently used as the internal reference tissue for 18F-FMISO PET/CT imaging in chordoma patients with 1.4 x SUVmean (muscle) utilized as the arbitrarily chosen lower threshold for hypoxic tissue. In this study, 0.8 x SUVmean (liver) was found to be the equivalent lower threshold for hypoxic tissue when using liver as the reference tissue. The absolute difference between the SUV thresholds calculated using muscle and liver for each scan was 0.27 ± 0.25 (mean ± SD). Due to this difference, in 13 out of a total of 39 scans, hypoxic tissue was calculated as present based on one reference tissue and absent based on the other reference tissue. However, if a hypoxic subvolume must be greater than 1 mL to be considered clinically significant, this effect decreases to 5 out of 39 scans and further decreases to 3 out of 39 scans if a hypoxic subvolume must be greater than 10 mL to be clinically significant. Calculating an average hypoxic SUV threshold based on both muscle and liver, i.e. [1.4 x SUVmean (muscle) + 0.8 x SUVmean (liver)] / 2, has the advantage of using both data points and is an alternative to choosing one reference tissue over the other to determine the presence of a hypoxic subvolume. When comparing the average hypoxic SUV threshold to the threshold based on muscle, hypoxic tissue was calculated as present based on one threshold and absent based on the other in 5 and 0 out of 39 scans if a hypoxic subvolume must be greater than 1 and 10 mL, respectively, and when comparing the average hypoxic SUV threshold to the threshold based on liver, the number of scans was 0 and 3 out of 39, respectively.
Conclusions Muscle is currently used as the internal reference tissue for 18F-FMISO PET/CT imaging in chordoma patients. This study demonstrates that liver uptake has similar variability and can also be used as the internal reference tissue. Inherent variability in SUV values for both muscle and liver can lead to differences in the determination of the presence of hypoxic tissue when using one as the reference tissue versus the other. Consequently, using two reference tissues, namely both muscle and liver, to calculate an average hypoxic SUV threshold may be advantageous and minimize the effect of the degree of inherent SUV variability. Future research will be required to correlate chordoma hypoxic subvolumes measured using these SUV thresholds on 18F-FMISO PET/CT with histopathologic evidence of hypoxia.