Purpose/Objectives: First clinical experiences employing PSMA (prostate-specific membrane antigen) with alpha particle-emitting radionuclide actinium-225 (225Ac) exhibited astonishing responses in metastasized castration-resistant prostate cancer (mCRPC) patients. To address the toxicities associated with its stand-alone administration, a tandem approach combining [225Ac]Ac-PSMA-617 and [177Lu]Lu-PSMA-617 (actinium-lutetium) has been developed as an optimized strategy, balancing treatment efficacy while reducing adverse effects. This study explores circulating tumor DNA (ctDNA) analysis, specifically tumor fraction (TFx) estimation via ichorCNA, as a biomarker to track treatment response and resistance in mCRPC patients receiving tandem actinium-lutetium therapy. By integrating clinical data with whole genome sequencing, we aim to decode resistance mechanisms and relapse patterns hidden in ctDNA, paving the way for personalized treatment strategies. Materials/Methods: Within the framework of ethical approval (S-882/2020), this study involved collecting blood samples from mCRPC patients scheduled bimonthly [225Ac]Ac-PSMA-617/[177Lu]Lu-PSMA-617 cycles. Circulating free DNA (cfDNA) was extracted and prepared for ultra-low-pass whole genome sequencing (ULP-WGS), with a depth of 5x. The study employed ichorCNA algorithm via R (version 3.3.1) for analyzing the sequencing data, which estimates genome-wide copy number alterations (CNA) and TFx from the sparse sequencing data. Results: This study examined a cohort of 78 patients with mCRPC undergoing actinium-lutetium treatment, with a median age of 75.5 years (55–91). Pairwise analysis of TFx and PSA across treatment cycles revealed a strong correlation, with both markers showing significant changes during treatment. However, TFx, unlike prostate-specific antigen (PSA), effectively distinguished between metastatic stages (p = 0.027), highlighting its potential as a more precise biomarker for disease burden. A Cox hazards model indicated that higher pre-treatment TFx levels were associated with a 5-fold increased risk of relapse in the subsequent treatment cycle (p = 0.0259). Hierarchical clustering of logR values for CNA across the whole genome identified two distinct patient clusters: Cluster 1, with lower TFx values and a low CNV burden, and Cluster 2, with higher TFx values and an increased CNV burden (p = 8.09e-08). Survival analysis showed that patients with a lower CNV burden had a longer median survival (13.8 vs. 8.3 months, log-rank p = 0.112) suggesting CNV burden as a potential prognostic factor in mCRPC. Conclusions: This study highlights the potential of ctDNA analysis, specifically TFx estimation, as a valuable biomarker for tracking treatment response and disease progression in mCRPC patients receiving tandem actinium-lutetium therapy.

AMGHAR Mariam;  RAUSCH Tobias;  HIELSCHER Thomas;  OZGUR Hilal;  BAUDER-WÜST Ulrike;  RATHKE H;  ROSCHER Mareike;  BRUCHERTSEIFER Frank;  MORGENSTERN Alfred;  BENES Vladimir;  KRATOCHWIL Clemens;  BENESOVA-SCHÄFER Martina; 

2026-07-03

ELSEVIER SCIENCE INC

JRC143696

1879-355X (online),   

https://www.sciencedirect.com/science/article/pii/S0360301625018772?__cf_chl_f_tk=aSQXyBpmNhNCMcinOFBwTsNGCFyjXMzM2J4HUypZF8k-1783000213-1.0.1.1-O3Ml2xYrXRxEdflBdm.CnthMvMTEVJQ5Olg_n181zeI,    https://publications.jrc.ec.europa.eu/repository/handle/JRC143696,   

10.1016/j.ijrobp.2025.06.1272 (online),