Longitudinal Tumor Mutation Burden Dynamics in Advanced Prostate Cancer Using Circulating Tumor DNA Profiling

Circulating tumor DNA (ctDNA) genomic testing is increasingly used in advanced prostate cancer (aPC) to identify actionable alterations, however the longitudinal nature of tumor mutation burden (TMB) remains poorly characterized. Current guidelines recommend somatic testing at diagnosis of metastatic hormone-sensitive or castration-resistant disease, but do not address serial testing. While prior studies suggest modest increases in TMB with advanced disease, longitudinal trends and treatment-associated changes have not been systematically examined. Here we evaluate how TMB changes over time and in relation to treatment exposures in a large real-world aPC cohort.

We retrospectively identified 714 men with aPC who underwent ≥ 1 ctDNA-based genomic profiling test with evaluable TMB between November 2018 and May 2025. For patients with multiple tests, sequential TMB values were mapped to discrete treatment episodes, used as a surrogate for lines of therapy. Univariable linear mixed-effects models with fixed effects for timeframe and treatment exposures, and random intercepts for repeated measurements, assessed associations between TMB, episode number, age, individual therapies, and treatment combinations. An exploratory multivariable mixed-effects model was also performed.

Median age at diagnosis was 63.3 years (interquartile ranges [IQR] 56.8-69.0); median prostate specific antigen 13.9 ng/mL (IQR 6.8-60.9). Median first-test TMB was 7.8 mutations per megabase (mut/Mb) (IQR 4.8-11.2). TMB increased significantly across episodes (P = .001). Age was associated with higher TMB (+ 0.20 mut/Mb per year, P < .0001). In univariable analyses, radiotherapy (RT) was associated with higher TMB (+ 1.9 mut/Mb, P = .004). In multivariable models adjusting for age and treatment episode, RT remained associated with higher TMB (+ 2.4 mut/Mb, 95% confidence intervals 1.1-3.7; P = .0004). Combination regimens, particularly androgen deprivation therapy + androgen receptor pathway inhibitors + RT, were associated with marked increases in absolute TMB.

We demonstrate that TMB is dynamic and rises over time and with select treatment exposures. These findings reinforce how TMB should be interpreted within the broader context and not necessarily viewed as a standalone threshold for initiating immunotherapy in advanced prostate cancer.