Abstract
TS17
Introduction: 18F-DCFPyl and 68Ga-PSMA-11 are both used for staging of prostate cancer; however, the VISION Trial required 68Ga-PSMA-11 PET-CT, leaving insurers and the oncology community uncertain whether 18F-DCFPyl PET-CT is also suitable for identifying appropriate candidates for 177Lu-PSMA-617 therapy. This study compares quantitative measures of tumor burden between tracers.
Methods: 12 consecutive cases receiving both 18F and 68Ga PSMA PET prior to treatment had total tumor burden segmented using a rule based MIM post-processing workflow with an absolute SUV threshold of SUV ≥ 4. Uptake above threshold was classified as lesion or physiologic uptake by 2 Nuclear Medicine Technology Students (NMTS), with review by a board-certified radiologist. Linear regression and Bland Altman (BA) analysis were performed for differences in SUV max, total lesion SUV mean and volume, and total lesion uptake (TLU; SUV mean x volume).
Results: Scans were performed within 1.0-3.2 (median 1.8) months of one another, excepting one outlier, which was omitted due to clear progression between scans. In all but one case, 18F preceded 68Ga PET. Regression analysis demonstrated high concordance for SUV max, SUV mean, total tumor volume and TLU (r²=0.59, 0.59, 0.78 and 0.77, all p<0.01). BA showed high agreement at low tumor burden levels, although with significant heteroskedasticity, showing less agreement as tumor burden became more extreme (Figure). A small bias toward higher values on 68Ga at higher tumor burden levels was observed.
Conclusions: Our early experience shows good concordance for quantitative tumor burden measures between 18F and 68Ga PSMA PET. Concordance degraded as tumor burden became more extreme, with a trend toward higher values on 18F PET; however such differences are less likely to alter management in patients with a high overall tumor burden. Total tumor burden metrics such as volume and TLU showed better predictive power than SUV in the hottest lesion or total lesion SUV mean. It is not completely clear whether time lag between 18F and 68Ga PET introduced bias into our results; parameters drifted higher for 68Ga scans, which were performed 1-3 months later, suggesting interval progression may account for quantitative differences in patients with large tumor burden. Our group believes 18F-DCFPyl and 68Ga-PSMA-11 are equivalent for establishing suitability for 177Lu-PSMA-617 therapy.
Quantitative segmentation of PET can be time consuming; in this cohort, the cooperation of an automated threshold with oversight and proofreading by NMTS and nuclear radiologists made the workflow more practical.