Comparison of discovery rates and prognostic utility of [68Ga]Ga-PSMA-11 PET/CT and circulating tumor DNA in prostate cancer-a cross-sectional study

Kilian Kluge,Holger Einspieler,David Haberl,Clemens Spielvogel,Dominik Amereller,Gerda Egger,Gero Kramer,Bernhard Grubmüller,Shahrokh Shariat,Marcus Hacker,Lukas Kenner,Alexander Haug
DOI: https://doi.org/10.1007/s00259-024-06698-7
Abstract:Background: Circulating-tumor DNA (ctDNA) and prostate-specific membrane antigen (PSMA) ligand positron-emission tomography (PET) enable minimal-invasive prostate cancer (PCa) detection and survival prognostication. The present study aims to compare their tumor discovery abilities and prognostic values. Methods: One hundred thirty men with confirmed PCa (70.5 ± 8.0 years) who underwent [68Ga]Ga-PSMA-11 PET/CT (184.8 ± 19.7 MBq) imaging and plasma sample collection (March 2019-August 2021) were included. Plasma-extracted cell-free DNA was subjected to whole-genome-based ctDNA analysis. PSMA-positive tumor lesions were delineated and their quantitative parameters extracted. ctDNA and PSMA PET/CT discovery rates were compared, and the prognostic value for overall survival (OS) was evaluated. Results: PSMA PET discovery rates according to castration status and PSA ranges did differ significantly (P = 0.013, P < 0.001), while ctDNA discovery rates did not (P = 0.311, P = 0.123). ctDNA discovery rates differed between localized and metastatic disease (P = 0.013). Correlations between ctDNA concentrations and PSMA-positive tumor volume (PSMA-TV) were significant in all (r = 0.42, P < 0.001) and castration-resistant (r = 0.65, P < 0.001), however not in hormone-sensitive patients (r = 0.15, P = 0.249). PSMA-TV and ctDNA levels were associated with survival outcomes in the Logrank (P < 0.0001, P < 0.0001) and multivariate Cox regression analysis (P = 0.0023, P < 0.0001). Conclusion: These findings suggest that PSMA PET imaging outperforms ctDNA analysis in detecting prostate cancer across the whole spectrum of disease, while both modalities are independently highly prognostic for survival outcomes.
What problem does this paper attempt to address?