Abstract
78
Introduction: 177Lu-PSMA radioligand therapy (RLT) is a novel therapeutic option that has shown encouraging results in phase 2 trials in patients with metastatic castration-resistant prostate cancer (mCRPC). PSA-response at 12 week showed strong association with survival in mCRPC patients receiving abiraterone or enzalutamide. This concept has not been explored for 177Lu-PSMA. The aim of this retrospective analysis was to evaluate whether 1) a decline in PSA by >50% at 12 wk is associated with improved overall survival (OS) and 2) a low tumor burden at baseline is correlated with higher PSA response at 12 wk and longer OS. Materials and Methods: Patients who underwent at least 2 cycles of 177Lu-PSMA-I&T RLT under a compassionate use programme and baseline 68Ga-PSMA11 PET/CT were included in the analysis. PSA-values were collected at baseline and after the 2ndcycle (12±2 wk). Baseline 68Ga-PSMA11 PET/CTs were quantified using qPSMA, an in-house developed software, as it follows: bone lesions using an SUV-threshold of 3 and soft-tissue lesions using a liver-based SUV-threshold. Tumor volumes were classified based on the median value into high and low tumor load. PSA changes between baseline and 12 wk were calculated. PSA response at 12 wk was defined as a 50% decline in PSA relative to baseline. In this landmark analysis at 12 wk, the OS was calculated from that time point. Association between PSA changes and OS was evaluated in Cox regression analyses.
Results: Sixty-one patients were analysed. Median baseline tumor burden was 475ml (IQR: 240 to 1926ml). The median PSA change at 12 wk was -34% (IQR: -71.2% to +54.7%). The overall median follow-up was 11.3 months (IQR 6.1-16.9). Twenty-five (41%) patients achieved a PSA response at 12wk. PSA response was significantly correlated with OS (HR=1.90, 95%CI: 1.07-3.36, p=0.02) with a median OS of 16.5 (95%CI: 14.4-18.6) vs. 7.5 (95%CI: 4.8-10.3) months. Tumor burden at baseline was not correlated with PSA response (response rate of 46% and 35% for patients with high and low tumor burden, respectively). In patients without a PSA response, tumor burden at baseline was a significant prognostic factor for survival with a median OS of 10.6 (95%CI: 7.0-14.2) vs. 4.8 (95%CI: 4.2-5.3) months for low vs. high tumor load, respectively. In contrast, median OS of patients with a PSA response was similar for low and high tumor burden: 17.9 (95%CI: 10.7-25.0) vs. 16.5 (95%CI: 10.1-22.9) months, respectively.
Conclusions: PSA response at 12 wk during 177Lu-PSMA RLT is significantly associated with longer OS. The likelihood for a PSA response appears independent from tumor burden at baseline. However, in patients without a PSA response, high tumor burden at baseline is a sign of poor prognosis. Prospective trials are warranted to further validate PSA response at 12 wk as a potential surrogate end-point for PSMA-targeted radionuclide therapy.