3371 - Predictors of PSA Response and Hematologic Changes in Patients Undergoing Lu-177-PSMA-617 Radiopharmaceutical Therapy: A Multivariate Analysis

12:45pm - 02:00pm PT

Hall F

Screen: 34

POSTER

Presenter(s)

Edmond Yaghoubian, MD - UPMC, Pittsburgh, PA

E. Yaghoubian¹, M. Jelenik², M. Becker¹, K. Bennet¹, K. Zia¹, F. E. Escorcia³, R. L. Wong¹, P. Chablani¹, L. Appleman¹, A. Saoudi¹, M. S. S. Huq⁴, A. C. Olson⁵, H. Wang⁶, H. D. Skinner⁴, and R. B. Patel⁷; ¹UPMC Hillman Cancer Center, Pittsburgh, PA, ²University of Pittsburgh, Pittsburgh, PA, ³Molecular Imaging Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, ⁴Department of Radiation Oncology, UPMC Hillman Cancer Center, Pittsburgh, PA, ⁵UPMC-Shadyside Hospital, Pittsburgh, PA, ⁶University of Pittsburgh School of Medicine, Pittsburgh, PA, ⁷University Hospitals of Cleveland - Seidman Cancer Center, Cleveland, WI

Purpose/Objective(s):

Prostate-specific antigen (PSA) response to Pluvicto (PRPT) varies, and the predictive role of baseline clinical and laboratory markers remains unclear. This study evaluates the relationship between baseline PSA, hematologic and metabolic markers, prior chemotherapy, and PSA kinetics after each injection and throughout treatment. We hypothesize that pre-treatment factors correlate with PSA response and hematologic trends, aiding personalized treatment strategies.

Materials/Methods:

A mixed-effects model with SAS PROC MIXED analyzed repeated patient measures. PSA values were log-transformed for normality. Univariate analyses identified significant predictors, which were incorporated into a multivariate model. Lab values were normalized as relative changes from baseline, and post-injection measurements were averaged.

Results:

Baseline PSA, estimated glomerular filtration rate (eGFR), aspartate aminotransferase (AST), and hemoglobin (Hgb) correlated with PSA across injections. After injection one, PSA (1.002, p < 0.0001) and eGFR (0.025, p = 0.040) were predictive. AST (0.029, p = 0.017) and Hgb (-0.392, p = 0.004) were significant after injection two. By injection four, age (-0.058, p = 0.044) and neutrophils (0.055, p = 0.043) became predictors. Prior chemotherapy positively impacted PSA response most notably after injection five (-0.165, p = 0.002). At injection six, PSA (0.775, p < 0.0001) and eGFR (0.049, p = 0.023) were the strongest predictors.

Over the full treatment, univariate analysis identified AST (0.029, p = 0.0125), Hgb (-0.366, p = 0.0062), PSA (0.957, p < 0.0001), and eGFR (0.031, p = 0.0238) as significant, but in multivariate analysis, only baseline PSA (0.865, p < 0.0001) remained predictive.

PRPT induced hematologic and metabolic changes. Absolute lymphocytes declined at all injections (-0.1628 to -0.3659, p < 0.0001). Hgb declined at injections 1, 3, 4, 5, and 6, with the largest drop at injection six (-0.0760, p < 0.0001). Platelet (-0.0653 to -0.2612, p < 0.0001) and WBC counts (-0.1013 to -0.2280, p < 0.0001) declined. AST increased at injections 2, 3, and 4 (0.1642, p = 0.0004), and alanine aminotransferase (ALT) rose at injection one (0.1254, p = 0.0389).

Conclusion:

Baseline PSA is the most significant predictor of PSA response to PRPT, with higher initial levels correlating with persistently elevated PSA throughout treatment. AST, Hgb, and eGFR also influence PSA dynamics but play secondary roles. Prior chemotherapy appears to strengthen treatment efficacy, particularly after the fifth injection. Hematologic toxicity is evident, necessitating close monitoring of blood counts. Elevated AST and ALT suggest potential hepatic stress. These findings underscore the importance of predictive modeling to refine patient selection and optimize treatment monitoring. Prospective validation is warranted.