Main Session
Oct 01
QP 20 - GU 9: Quick Pitch: Predictive Markers in Prostate Cancer

1115 - Prognostic Value of Quantitative Imaging Metrics Obtained in Routine Clinical Practice in a Cohort of Patients Treated with PSMA-RLT

08:10am - 08:15am PT
Room 155/157

Presenter(s)

John Floberg, MD, PhD - University of Wisconsin School of Medicine and Public Health, Madison, WI

J. M. Floberg1, H. Menon2, C. Kyriakopoulos3, A. V. Serritella3, R. Hutten4, and S. Y. Cho5; 1Department of Human Oncology, University of Wisconsin-Madison, Madison, WI, 2University of WIsconsin School of Medicine and Public Health, Department of Human Oncology, Madison, WI, 3University of WIsconsin School of Medicine and Public Health, Department of Medicine, Division of Hematology/Oncology, Madison, WI, 4University of Wisconsin School of Medicine and Public Health, Department of Human Oncology, Madison, WI, 5University of Wisconsin School of Medicine and Public Health, Department of Radiology, Madison, WI

Purpose/Objective(s): The mean standardized uptake value (SUVmean), a measure of average tracer uptake across all metastases on prostate specific membrane antigen (PSMA) PET imaging, is known to be a predictive marker for response to PSMA radioligand therapy (RLT). In addition, the presence of metastases that do not demonstrate PSMA uptake is a known negative prognostic marker. These metrics are therefore important in selecting patients for PSMA-RLT. However, they are cumbersome and impractical to measure in patients with many dozens of metastases in routine clinical practice. Here we investigate imaging markers measured by a commercially available image analysis tool (TRAQInform IQ, AIQ Solutions, Madison, WI) as markers of response and disease outcomes in a cohort of patients treated with standard-of-care PSMA-RLT.

Materials/Methods: We collected disease outcomes and imaging metrics from an institutional cohort of patients treated with PSMA-RLT with IRB approval. Imaging metrics collected included SUVmean, the percentage of the total volume of disease on CT imaging that does not show PSMA uptake (%PSMA neg), and the total PSMA-avid disease volume (PSMA Vol). The relationships of these markers to PSA response (50% reduction in PSA), progression free survival (PFS), radiographic PFS, and overall survival (OS) were investigated. Correlation between imaging metrics and PSA response was assessed with logistic regression. Survival analysis was performed with Cox regression and Kaplan-Meier analysis.

Results: Data from 50 patients were gathered. There was a significant difference in SUVmean between patients with and without a PSA response (p=0.003 by Mann-Whitney U). There was no difference in %PSMA neg and PSMA Vol between responders and non-responders. Of the imaging metrics, only SUVmean was associated with PSA response, when considered either as continuous variable (OR 1.63, 95% CI 1.12-2.38) or when treating it as a dichotomous variable stratified by the median (SUVmean <8 vs = 8, OR 7.14, 95% CI 1.86-27.48). An SUVmean= 8 was also associated with improved PFS (7.7 vs 5.3 mo, HR 0.53, 95% CI 0.28-0.98). No other imaging metrics were associated with PFS, and no imaging metrics were significantly associated with rPFS or OS.

Conclusion: Of the imaging metrics obtained with a commercially available image analysis software, a higher SUVmean was associated with a better response to PSMA-RLT and improved PFS. The %PSMA neg and PSMA Vol were not associated with any disease outcomes.