Quality of life assessment during prostate cancer radionuclide therapy: Updated FACT-RNT psychometric evaluation.

Abstract

141 Background: The Functional Assessment of Cancer Therapy–Radionuclide Therapy (FACT-RNT) is a 15-item patient-reported outcome (PRO) measure used to monitor RNT-related symptoms, toxicities, and quality-of-life impacts in patients with metastatic castrate-resistant prostate cancer (mCRPC). This study re-examined the validity and reliability of the FACT-RNT using repeated assessments in an expanded, multi-institutional cohort of patients receiving RNT. Methods: Patients starting 177 Lu-PSMA-617 for mCRPC at UCLA from December 2022 to August 2023 and at Moffitt Cancer Center from October 2023 to October 2025 completed the FACT-RNT before each cycle. Patients at Moffitt also completed previously validated measures of prostate cancer disease- and treatment-related symptoms (NFPSI-17) and physical function (PG-SGA). Reliability was assessed through tests of internal consistency, split-half reliability, inter-item correlations, and test-retest reliability across RNT cycles 1-4. Correlations between the FACT-RNT and both the NFPSI-17 and PG-SGA at cycle 2 were calculated to assess convergent validity. Results: Patients (N=81) were predominantly non-Hispanic (80%) and White (69%), and were, on average, 70 years old (SD=9, range=43-91) at RNT cycle 1. Fifty-nine (73%) completed the FACT-RNT for ≥2 cycles, and 33 (41%) completed the FACT-RNT for ≥4 cycles. Internal consistency reliability across cycles was good-to-excellent (α=0.83-0.90), average split-half reliabilities were excellent (α=0.82-0.86), and average inter-item correlations were acceptable ( r =0.24-0.30). Test-retest reliability was moderate-to-good for individual cycle scores (ICC3,1=0.68, CI=0.53-0.81, p<0.001) and excellent when averaging across cycles 1-4 (ICC3,4=0.89, CI=0.82-0.94, p<0.001). FACT-RNT total score correlations between consecutive cycles were also strong (cycles 2-3, r =0.72; cycles 3-4, r =0.91). For patients completing other PROs at cycle 2 (n=23), the FACT-RNT was positively associated with symptoms as assessed by NFPSI-17 ( r =0.89, p <0.001) and negatively associated with physical functioning assessed by PG-SGA ( r =-0.61, p <0.002), suggesting strong convergent validity. Conclusions: Results from updated analyses suggest that the FACT-RNT has good-to-excellent internal consistency and split-half reliability, acceptable inter-item correlations, and moderate-to-excellent stability across administrations. As between-cycle variability in treatment-related symptoms is expected, these findings may reflect both true change in treatment-related symptoms and quality of life, and measurement reliability. The FACT-RNT also correlated as expected with related PRO measures, supporting preliminary convergent validity. Overall, this evidence supports the validity and reliability of the FACT-RNT and its use in future research and clinical practice.