Early On-Treatment Circulating Tumor DNA Kinetics in Metastatic Castration-Resistant Prostate Cancer: When Molecular Response Adds Value Beyond Prostate-Specific Antigen

Early on-treatment response assessment in metastatic castration-resistant prostate cancer (mCRPC) remains challenging. Prostate-specific antigen (PSA) may not fully reflect tumor biology, and conventional imaging has limited sensitivity in bone-predominant disease. Circulating tumor DNA (ctDNA) has a short half-life (approximately 1-2 hours) and can capture dynamic changes in tumor burden more rapidly than standard assessments. Although baseline ctDNA has established prognostic value, interest has shifted toward early on-treatment ctDNA kinetics as a biomarker of treatment sensitivity or resistance. We conducted a narrative review of PubMed/MEDLINE-indexed literature through April 2026, examining the biologic rationale, clinical evidence, and limitations of early ctDNA kinetics across major mCRPC treatment classes, including androgen receptor pathway inhibitors (ARPIs), taxane chemotherapy, poly (ADP-ribose) polymerase (PARP) inhibitors, and prostate-specific membrane antigen (PSMA)-targeted radioligand therapy. The strongest evidence supports ctDNA kinetics in ARPI-treated mCRPC, where persistent ctDNA detection at 4 weeks has been associated with shorter progression-free survival and overall survival (OS), and ctDNA status has predicted outcomes independent of PSA response. In chemotherapy, on-treatment cell-free DNA (cfDNA) and tumor fraction changes have been associated with response and survival, although most studies used total cfDNA rather than tumor-specific metrics. In PARP inhibitor-treated disease, cfDNA declines at week 8 have been independently associated with OS, and serial monitoring has detected reversion mutations at progression. In PSMA-targeted radioligand therapy, early ctDNA detectability has been associated with outcomes independent of PSA. Across settings, ctDNA appears most valuable when PSA and imaging are discordant. Barriers to implementation remain, including assay heterogeneity, unstandardized metrics and time points, and confounding by clonal hematopoiesis. ctDNA has not been validated as a surrogate end point. Early on-treatment ctDNA kinetics should be considered a promising investigational biomarker. Future studies should prioritize assay standardization, predefined molecular response thresholds, and treatment-specific sampling strategies.