3249 - Predicting Response to <sup>177</sup>Lu-PSMA-617 Using Pre-Treatment Clinical Biomarkers

12:45pm - 02:00pm PT

Hall F

Screen: 33

POSTER

Presenter(s)

Jackson Howell, MD - University of Utah Huntsman Cancer Institute, Salt Lake City, UT

J. Howell¹, J. Wilkes², N. M. Maughan², K. Morton³, E. Hu⁴, D. Gill⁵, S. Samuelson⁶, S. B. Johnson^1,7, and D. Boothe^2,6; ¹Dept. of Radiation Oncology, Huntsman Cancer Institute, University of Utah, Salt Lake City, UT, ²Dept. of Radiation Oncology, Intermountain Health, Murray, UT, ³Dept. of Radiology, Intermountain Health, Murray, UT, ⁴Intermountain Health & Summit Physician Specialists, Murray, UT, ⁵Dept. of Medical Oncology, Intermountain Health, Murray, UT, ⁶Utah Cancer Specialists, Murray, UT, ⁷Dept. of Radiology, University of Utah, Salt Lake City, UT

Purpose/Objective(s):

Lutetium-177 prostate-specific membrane antigen 617 (¹⁷⁷Lu-PSMA-617) is a beta-emitting radioligand therapy used in the treatment of metastatic castrate-resistant prostate cancer (mCRPC). Post-treatment factors such as early prostate-specific antigen (PSA) reductions predict response; however, predictive models using pre-treatment biomarkers are lacking and present a significant challenge to patient selection. We hypothesize that local progression after prior radiation (RT) (indicative of radioresistance) and pre-treatment PSMA PET/CT characteristics can be used to predict ¹⁷⁷Lu-PSMA-617 response.

Materials/Methods:

Pre-treatment clinical and imaging characteristics (demographics, mean disease standard uptake value [SUV] on PSMA PET/CT, max SUV, lesional heterogeneity), pre-cycle labs, prior treatments (RT, surgery, systemic therapy), and treatment responses were retrospectively collected for a cohort of 59 mCRPC patients who had previously received RT and were treated with ¹⁷⁷Lu-PSMA-617 at two community centers. PSA response was dichotomized between those who did and did not achieve a =50% reduction in PSA after cycle 2 (PSA_50-2) of ¹⁷⁷Lu-PSMA-617, and this variable’s correlations with pre-treatment characteristics were analyzed using univariate (UVA) and multivariate (MVA) logistic regression analysis. Progression-free survival ([PFS], radiographic progression or persistent PSA elevation) was analyzed using log-rank and Kaplan-Meier methods.

Results:

Median follow-up was 10.9 months (mo), and median PFS was 6.5 mo (95%CI [4.1 – 10.3]). Twenty patients (33.9%) were determined to have had local failure after RT, 38 (64.4%) had a mean disease SUV < 10, and 30 (50.8%) ultimately achieved PSA_50-2 on ¹⁷⁷Lu-PSMA-617. When evaluating pre-treatment clinical factors, both local failure after RT (UVA: HR 3.7, P=.025; MVA: HR 5.2, P=.024) and mean disease SUV < 10 (UVA: HR 9.6, P=.002; MVA: HR 12.5, P=.002) were independent, negative predictors of PSA_50-2. Among patients with both local failure after RT and a mean disease SUV < 10, only 14.3% achieved PSA_50-2, compared to 50.0% and 91.7% among those who had one or neither risk factor, respectively. On logistic regression analysis, patients with one or both risk factors were at heightened risk of not achieving PSA_50-2 (0 factors: reference; 1 factor: HR 11.0, P=.030; 2 factors: HR 66.0, P=.001), and the three groups demonstrated stratification of PFS outcomes (0 factors: median PFS 14.1 mo, reference; 1 factor: median PFS 5.1 mo, P=.002; 2 factors: median PFS 3.9 mo, P<.001).

Conclusion:

Local failure after RT and mean disease SUV < 10 on PSMA PET/CT are clinically relevant pre-treatment biomarkers that can be used combinatorially to predict response to ¹⁷⁷Lu-PSMA-617. Our model could help optimize patient selection and enhance prognostication. Validation in a larger, prospective cohort is warranted.