The prostate-specific membrane antigen (PSMA)-targeted radioligand [¹⁷⁷Lu]Lu–PSMA-617 is a new standard treatment for metastatic castration-resistant prostate cancer (mCRPC), but predictive genomic biomarkers informing its rational use are unknown. We performed detailed dissection of prostate cancer driver genes across 290 serial plasma cell-free DNA samples from 180 molecular imaging-selected patients with mCRPC from the randomized TheraP trial of [¹⁷⁷Lu]Lu–PSMA-617 (n = 97) versus cabazitaxel chemotherapy (n = 83). The primary endpoint was PSA50 biochemical response, with secondary endpoints of progression-free survival (PFS) and overall survival (OS). In this post-hoc biomarker analysis, a low pretreatment circulating tumor DNA (ctDNA) fraction predicted a superior biochemical response (100% versus 58%, P = 0.0067) and PFS (median 14.7 versus 6.0 months; hazard ratio 0.12, P = 2.5 × 10−4) on [¹⁷⁷Lu]Lu–PSMA-617 independent of predictive PSMA–positron emission tomography imaging parameters, although this benefit did not extend to OS. Deleterious PTEN alterations were associated with worse PFS and OS on cabazitaxel, whereas ATM defects were observed in select patients with favorable [¹⁷⁷Lu]Lu–PSMA-617 outcomes. Comparing pretreatment and progression ctDNA revealed population flux but no evidence that alterations in individual mCRPC genes (or FOLH1) are dominant causes of acquired [¹⁷⁷Lu]Lu–PSMA-617 or cabazitaxel resistance. Our results nominate new candidate biomarkers for [¹⁷⁷Lu]Lu–PSMA-617 selection and ultimately expand the mCRPC predictive biomarker repertoire. We anticipate our ctDNA fraction-aware analytical framework will aid future precision management strategies for [¹⁷⁷Lu]Lu–PSMA-617 and other PSMA-targeted therapeutics. ClinicalTrials.gov identifier: NCT03392428.
The Nature Medicine article “Lutetium-177–PSMA-617 or cabazitaxel in metastatic prostate cancer: circulating tumour DNA analysis of the randomized phase 2 TheraP trial” presents a comprehensive analysis of circulating tumour DNA (ctDNA) to identify biomarkers that predict treatment response in metastatic castration-resistant prostate cancer (mCRPC) by Kwan et al. It is a post-hoc study that utilised data from the TheraP trial, which compared the efficacy of the PSMA-targeted radioligand therapy [¹⁷⁷Lu]Lu–PSMA-617 with cabazitaxel chemotherapy.
The researchers analysed 290 plasma cell free DNA samples from 180 patients, focusing on prostate cancer driver genes. Looking first at ctDNA fractions (ctDNA%) one of the key findings is that patients with a low pretreatment ctDNA% exhibited a higher biochemical response rate (100% vs. 58%) and longer progression-free survival (median 14.7 vs. 6.0 months) when treated with [¹⁷⁷Lu]Lu–PSMA-617 (LuPSMA), independent of PSMA-PET imaging parameters.
Similar to previous studies in unselected first- and/or second-line mCRPC the most frequently altered genes were AR (68%), TP53 (53%) and PTEN (35%). Deleterious germline and/or somatic alterations in ≥1 DNA damage repair gene were present in 31%patientes with ctDNA ≥2%, most commonly in BRCA2 (7%), ATM (7%) and CDK12 (7%).
Focusing on patients with ctDNA ≥2%, and adjusted for ctDAN%, PTEN alterations were associated with a higher PSA50 response rate and PFS improvement on LuPSMA versus cabazitaxel (3.4 versus 1.7month, p = 0.049). This difference was driven by poor outcomes on cabazitaxel. Among the patients with DNA damage repair gene alterations the most durable responses to LuPSMA coincided with deleterious ATM and BRCA2 alterations, but several patients with CDK12 mutations benefited from cabazitaxel, demanding further investigation of selected DNA damage repair genes as biomarkers in mCRPC. On comparing baseline and progression ctDNA, no driver gene alterations were found to be enriched during either LuPSMA or cabazitaxel treatment, compatible with the hypothesis that anti-cancer therapy resistance mechanism are heterogenous.
The authors suggest that their data will aid future precision management strategies for LuPSMA and other PSMA-targeted therapeutics.
LuPSMA radioligantherapy was approved in 2022 for mCRPC patients previously treated with androgen receptor pathway inhibitor (ARPI) and a taxane-based chemotherapy. Earlier this year the Food and Drug Administration expanded the indication to include adults with PSMA-positive mCRPC who have been treated with and ARPI and are considered appropriate to delay taxane-based chemotherapy.
Studies investigating LuPSMA in even earlier disease stages, like hormone sensitive, metastatic PC and localised PC are ongoing. Obviously, this leads to a rising demand of LuPSMA worldwide that is unlikely to plateau in the foreseeable future. However, the reality of complex production logistics, high treatment costs, and the limited capacity of nuclear medicine facilities poses a significant challenge to meet this growing need. Thus, research into biomarkers to guide our treatment decisions and prioritise the treatment that will benefit the individual patients the most is essential. Kwan et al present a comprehensive ctDNA genomic framework analysing samples from TheraP, the first randomised, controlled trial comparing LuPSMA versus Cabazitaxel.
Biomarkers for patient stratification
Although longitudinal analyses did not reveal specific gene alterations driving resistance—a result consistent with the prevailing view that resistance in mCRPC is multifactorial and not typically attributable to single-gene changes—the study did identify promising candidate biomarkers for patient stratification. Whilst confirmatory data is obviously necessary and some of the findings especially concerning DNA damage repair genes are merely more than observations in a small patient group, we do gain a good impression of what to look for in future studies and how ctDNA guided treatment may look like:
These findings lay important groundwork for a more personalised treatment paradigm in mCRPC. As clinical demand for LuPSMA grows, biomarker-driven approaches may help ensure that this valuable resource is allocated where it is most likely to achieve meaningful benefit.