Objective To explore the value of ⁶⁸Ga-prostate specific membrane antigen (PSMA) PET/CT in the efficacy evaluation and prognosis prediction of patients with metastatic castration-resistant prostate cancer (mCRPC) after ¹⁷⁷Lu-Evans blue (EB)-PSMA radioligand therapy (RLT).

Methods A retrospective analysis was conducted on imaging and survival data from 38 mCRPC patients (age 68 (51, 75) years) who received ¹⁷⁷Lu-EB-PSMA RLT at Peking Union Medical College Hospital from January 2019 to December 2021, with follow-up data until December 31, 2022. The correlation of ⁶⁸Ga-PSMA PET/CT-related parameters with prostate-specific antigen (PSA) response, PSA-progression-free survival (PFS), and overall survival (OS) was explored, along with their prognostic predictive value. Pearson correlation analysis was performed to assess the correlation between variables. The prognostic factors for optimal PSA response and endpoint events, including PSA-PFS and OS, were identified using Logistic regression analysis and Cox proportional hazards regression model. Variables with values of P<0.2 in the univariate analysis were included in the multivariate analysis to determine independent predictive factors. The Kaplan-Meier curve analysis was used to analyze the time data of endpoint events. Among the obtained independent predictive factors, maximally selected rank statistics was used to obtain the optimal cutoff value for the ⁶⁸Ga-PSMA PET/CT-related parameters and stratified survival estimates were performed (Log-rank test).

Results Among the 38 mCRPC patients, 14 and 16 received 2 and 3 cycles of ¹⁷⁷Lu-EB-PSMA RLT, respectively. A PSA level decline of ≥50% from baseline was observed in 57.9% (22/38) of the patients. The Kaplan-Meier curve analysis revealed median PSA-PFS of 4.8 months (95%CI: 2.7–6.7) and median OS of 11.9 months (95%CI: 8.0–17.3). The Pearson correlation analysis showed a moderate correlation between baseline total lesion PSMA (TLP) and baseline PSA in all the patients (r=0.566, P=0.001), and a strong correlation between percentage change in TLP before and after treatment (ΔTLP) and percentage change in PSA before and after treatment (ΔPSA) (r=0.722, P<0.001). Univariate Logistic regression analysis was performed to identify the mean standardized uptake value (SUV_mean) of whole-body PSMA (OR=2.005, 95%CI: 1.124–3.713; P=0.007) and baseline TLP (OR=1.100, 95%CI: 1.004–1.118; P=0.012) as predictive factors of optimal PSA response. The multivariate Logistic regression analysis confirmed whole-body PSMA SUV_mean as an independent predictive factor of optimal PSA response (OR=1.910, 95%CI: 1.009–3.799; P=0.030). The Cox proportional hazards regression model results indicated that baseline PSMA-positive tumor volume (PSMA-VOL) (HR=1.810, 95%CI: 1.010–3.602; P=0.009) and baseline alkaline phosphatase (HR=1.695, 95%CI: 1.005–2.450; P=0.031) were independent predictive factors for PSA-PFS. Meanwhile, baseline PSA (HR=1.953, 95%CI: 1.082–3.450; P=0.002) and baseline PSMA-VOL (HR=4.400, 95%CI: 1.009–8.919; P=0.002) were independent predictive factors for OS. The Kaplan-Meier curve analysis demonstrated significantly prolonged PSA-PFS in patients with PSMA-VOL of ≤388 ml (6.2 months vs. 2.9 months, χ²=6.87, P=0.009) and significantly prolonged OS in those with PSMA-VOL of ≤423 ml (15.3 months vs. 7.5 months, χ²=10.05, P=0.002).

Conclusion The relevant parameters of ⁶⁸Ga-PSMA PET/CT exhibit important reference value in the efficacy evaluation and prognosis prediction of patients with mCRPC after ¹⁷⁷Lu-EB-PSMA RLT.