Purpose: Radioligand therapy with [177Lu]PSMA-617 (PSMA-RLT) is a promising therapeutic option for metastatic castration-resistant prostate cancer (mCPRP). This study assessed the prognostic value of early PSA measurements during PSMA-RLT. Methods: 27 patients with mCRPC scheduled for PSMA-RLT were prospectively enrolled for a serial short-interval PSA-assessment. Change in PSA (∆%PSA) during two treatment cycles was correlated with biochemical response (BR) and change in tumor volume on PET (TV) after 16 weeks (w16), as well as overall survival (OS). PCWG3 criteria and the recently recommended threshold of ∆%PSA ≤ −30% were assessed for their predictive value. Results: ∆%PSA first correlated with BR, TV and OS after 4 weeks (c1w4). At c1w4, ∆%PSA ≤ −30% was associated with the biochemical response at w16 (p = 0.003) and a longer median OS (p = 0.025), whereas the PCWG3-derived threshold of ∆%PSA ≤ −50% showed no such correlation. In contrast, ∆%PSA ≥ 25% at c1w4 was associated with biochemical progression at w16 (p = 0.003) and a shorter median OS (p < 0.001). Conclusion: PSA changes as early as four weeks after PSMA-RLT allow a significant prediction of later biochemical and PET-based imaging response, as well as OS. At this early time point, a more lenient threshold for a PSA decrease of at least 30% appears better-suited for the prediction of a positive biochemical response and longer OS. In contrast, the PCWG3-derived threshold for PSA increase (+25%) reliably anticipates biochemical progression and shorter OS.

The radiation dose to the kidneys should be monitored in prostate cancer patients treated with radioligand therapy (RLT) targeting the prostate-specific membrane antigen (PSMA). We analyzed whether pretherapeutic kidney function is predictive of subsequent kidney dose and to what extend the cumulative kidney dose after multiple therapy cycles at the end of treatment can be predicted from a dosimetry based on the first cycle. Methods: Data of 59 patients treated with at least 2 cycles of 177 Lu-PSMA-617 (PSMA-RLT) were analyzed. Treatment (median: 6 GBq/cycle) was performed at 6-8 week intervals, accompanied by voxel-based 3D-dosimetry (measured kidney dose) with SPECT/CT on each of days 0-3 and once during days 6-9. Pretherapeutic kidney function (eGFR, MAG3-clearance) was correlated to the kidney doses.Cumulative kidney doses at the end of treatment were compared to a dose estimation based on the population-based mean kidney dose, individual first cycle kidney dose and mean kidney doses of cycles 1, 3 and 5 per administered activity. Results: A total of 176 PSMA-RLT cycles were performed with a median of 3 cycles per patient. The average kidney dose per administered activity of all 176 cycles was 0.67 ± 0.24 Gy/GBq (range 0.21 -1.60). MAG3-clearance and eGFR were no reliable predictors of subsequent absorbed kidney dose and showed only small effect sizes (R 2 = 0.080 and 0.014, p = 0.039 and 0.375). All simplified estimations of cumulative kidney dose correlated significantly (p < 0.001) with measured kidney doses: Estimations based on the individual first-cycle dose were more accurate than the use of the population-based average kidney dose (R 2 = 0.853 vs. R 2 = 0.560). Dose estimation was best when the doses of cycles 3 and 5 were included as well (R 2 = 0.960). Conclusion: Pretherapeutic renal function was not predictive for subsequent kidney dose during therapy. Extrapolation of individual data from dosimetry of the first cycle was highly predictive for the cumulative kidney dose at the end of treatment. This is further improved by the integration of dose information from every other cycle. In any case, because of a high interindividual variance, an individual dosimetry is advisable.

Background The recent phase III VISION-trial confirms the treatment efficacy of radioligand therapy with [177Lu]PSMA-617 (PSMA-RLT) in metastatic castration-resistant prostate cancer (mCRPC). In PSMA-RLT, the relatively low absorbed bone marrow dose allows for multiple therapy cycles with relatively low risk of haematological adverse events (hAE). However, as disease progression itself may be a cause of bone marrow impairment, the aim of this study was to assess potential relations between impairment of haematological status and response to PSMA-RLT. Methods In this retrospective analysis, haematological parameters (HP) of 64 patients with mCRPC were systematically acquired over two cycles (12–16 weeks) of PSMA-RLT from baseline to restaging. Changes in HP were analysed qualitatively (CTCAE 5.0) and quantitatively. The HP changes from baseline were compared to quantitative and qualitative biochemical and imaging response, using PCWG3 and PROMISE criteria. Results All grade 3/4 hAE observed were associated with disseminated or diffuse bone involvement as well as biochemical non-response at restaging. Quantitatively, at baseline, HP inversely correlated with biochemical and volumetric (on PET) tumour burden as well as bone involvement pattern (p ≤ 0.043). Among patients with disseminated or diffuse bone involvement, percentage changes in HP (%HP) at restaging inversely correlated with serological and imaging tumour burden (p ≤ 0.017). Biochemical non-responders showed a significant decrease in %HP (p ≤ 0.001) while HP in biochemical responders remained stable (p ≥ 0.079). Conclusion During early cycles of PSMA-RLT, qualitative and quantitative bone marrow impairment appears to be closely associated with osseous tumour burden as only patients with advanced bone involvement and non-response to therapy exhibited high-grade haematological adverse events, showing a significant decline of haematological parameters. This implies that in patients with advanced mCRPC, non-response to PSMA-RLT may be a major cause of bone marrow impairment during early treatment cycles. German Clinical Trial Register DRKS00013665. Registered 28 December 2017, retrospectively registered (www.drks.de/drks_web/navigate.do?navigationId=trial.HTML&TRIAL_ID=DRKS00013665)