Editor's Note: At the 2026 Congress of the European Society for Medical Oncology (ESMO), a biomarker analysis of the Chinese cohort from TALAPRO-2 was presented as a poster. Understanding the differences in radiographic progression-free survival (rPFS) between the Chinese cohort and the global population, and translating molecular testing results into treatment decisions, are important issues in the individualized management of metastatic castration-resistant prostate cancer (mCRPC).Oncology Frontier – UroStream invited Professor Hao Zeng from West China Hospital, Sichuan University, to discuss the homologous recombination repair (HRR) gene mutation profile identified in this analysis. Taking prior treatment exposure and baseline patient characteristics into account, he examines the prognostic and predictive value of biomarkers and explores the clinical implications and evidentiary limitations of genetic testing and patient selection.
Oncology Frontier – UroStream: The rPFS results in the Chinese cohort differed from those in the global population. What clinical questions was this biomarker analysis designed to address? Which biomarkers were evaluated, what testing methods were used, and how can these findings help us understand treatment benefit in Chinese patients?
Professor Hao Zeng: The subgroup analysis of Chinese patients in TALAPRO-2 was presented as a poster at this year’s ESMO Congress. The main reason for conducting a separate analysis of the Chinese cohort was that its efficacy results were not entirely consistent with those of the global population.
In the global TALAPRO-2 population, median rPFS was approximately 33 months in the talazoparib plus enzalutamide group, compared with more than 20 months in the control group. In the Chinese population, however, median rPFS was approximately 15 months in the combination group and less than 10 months in the control group. Chinese patients accounted for approximately 10%–15% of the overall study population (125 patients). Although this was not a particularly large proportion, the difference in outcomes between the two groups was substantial enough to warrant investigation.
The first possible explanation is prior treatment exposure. More than 20% of the Chinese patients had received a novel androgen receptor-targeted agent before entering the castration-resistant stage. When patients have already received novel hormonal therapy in earlier treatment settings, the potential for further efficacy may be reduced after progression to CRPC, regardless of whether they receive combination therapy or enzalutamide monotherapy.
The second factor is baseline characteristics. Compared with the global population, Chinese patients generally had a higher tumor burden and a greater risk of disease progression and death, both at the initial diagnosis of metastatic hormone-sensitive prostate cancer (mHSPC) and upon entering the CRPC stage. These two factors can explain a substantial proportion of the lower numerical outcomes observed in the Chinese cohort. It is important to emphasize, however, that despite the lower absolute values, the difference between the investigational and control groups remained statistically significant.
Building on these findings, we turned our attention to the molecular level: could mutation patterns among the 125 patients also partially explain why median rPFS was only around 15 months?
The analysis did identify several differences. First, the composition of HRR mutations in Chinese and Asian populations has some distinctive characteristics, including a relatively higher proportion of CDK12 mutations. Among HRR pathway alterations, BRCA1 and BRCA2 mutations are among the most frequently observed events. Previous studies have shown that the benefit of PARP inhibitors is more clearly established in patients with BRCA1/2 mutations. CDK12 mutations are different: to date, a benefit signal has been observed only in studies involving talazoparib. Therefore, the mutation profiles of Chinese and global populations differ to a considerable extent.
Second, the number of patients with BRCA mutations in this analysis was in the single digits, which is insufficient to support reliable subgroup conclusions. Further analyses will also need to distinguish between different mutation types, particularly short variants and structural variants. If short variants predominate in the Chinese population, the overall benefit from PARP inhibitors may be relatively limited.
This subgroup analysis of 125 patients provides practical reference information for the future development of PARP inhibitor monotherapy or combination strategies in the mHSPC and CRPC settings.
Oncology Frontier – UroStream: Which biomarkers in this analysis showed the most noteworthy associations with outcomes following talazoparib plus enzalutamide? Considering the performance of the control group and the subgroup sample sizes, how should we distinguish the prognostic significance of biomarkers from their ability to predict treatment benefit?
Professor Hao Zeng: The Chinese cohort provides a limited amount of data. To identify characteristic features that can help determine in advance which patients are more likely to benefit from PARP inhibitors, we still need to look to pooled analyses from global, multicenter, prospective phase III studies. Based on these pooled data, three levels of information should be considered in sequence.
The first is prior treatment history, including the number of previous lines of therapy and whether patients have received novel androgen receptor-targeted agents or chemotherapy.
The second is molecular characteristics, particularly whether androgen receptor (AR) expression remains positive in the CRPC setting. This determines whether patients still have the potential to respond to enzalutamide as the backbone treatment.
The third is the mutation profile associated with PARP inhibitor treatment. Both TALAPRO-2 and TALAPRO-3 conducted subgroup analyses according to mutated genes. Data from talazoparib studies indicate that the population deriving benefit includes patients with BRCA1 and BRCA2 mutations, as well as those with ATM and CDK12 mutations. In studies of other PARP inhibitors, by contrast, patients with ATM or ATR mutations have either not been included, or those with CDK12 mutations have had difficulty deriving benefit. These differences are related to the intrinsic synthetic lethality activity and PARP-trapping potency of individual drugs, reflecting differences at the molecular mechanism level.
For clinicians, the approach should be to test first and interpret the results carefully. Once genetic testing results are available, we need to determine whether a BRCA mutation is germline or somatic in origin, whether it is pathogenic or non-pathogenic, and even exclude the possibility of false-positive findings caused by clonal hematopoiesis.
At this year’s ESMO Congress, Professor Azad from Australia specifically discussed these issues. Only when the results are interpreted correctly can an appropriate treatment strategy be selected. At the same time, whether PARP inhibitors are used as monotherapy or in combination, and whether they are administered in the mHSPC or CRPC setting, their safety profiles require close attention from both physicians and patients.
Oncology Frontier – UroStream: From the perspective of first-line treatment decisions for Chinese patients with mCRPC, what insights do these findings offer for genetic testing and patient selection? Which conclusions are already useful as clinical references, and which still require validation in larger populations or prospective studies?
Professor Hao Zeng: In both the CRPC and mHSPC settings, a substantial body of research data is now available on novel agents used as monotherapy and in combination regimens. Future treatment strategies will become increasingly individualized. The essential prerequisite for individualized treatment is molecular biomarker testing. PARP inhibitors are just one of many targeted treatment approaches and represent only one aspect of biomarker assessment.
In the castration-resistant prostate cancer setting, for example, PI3K/AKT pathway inhibitors have already met their primary endpoints in prospective studies.
In addition, the benefit of combining lutetium-177 PSMA (¹⁷⁷Lu-PSMA) with enzalutamide depends on assessing the uptake intensity across all metastatic lesions using PSMA PET-CT. SUV uptake intensity is positively correlated with the efficacy of ¹⁷⁷Lu therapy. We also need to consider the effects of molecular factors such as PTEN loss and TP53 mutations on treatment efficacy.
The combination of radium-223 (²²³Ra) and enzalutamide, meanwhile, is supported by data from the PEACE-3 study. All these combination strategies require appropriate patient selection before treatment.
Therefore, the value of biomarker testing is not limited to the genetic level; molecular imaging is also important. Take PTEN as an example. When predicting response to AKT inhibitors, tissue-based immunohistochemistry may be more reliable, as genetic testing may fail to detect loss of the protein at the expression level. Conversely, PTEN alterations detected at the genetic level may be more valuable for prognostic assessment. This illustrates the two key aspects I mentioned earlier.
What can already be used as a clinical reference is the integration of prior exposure to novel hormonal therapies, HRR status, and tumor burden into a framework for first-line treatment decisions. What still requires validation is the magnitude of benefit associated with relatively uncommon mutations such as CDK12 and ATM in the Chinese population, as well as the influence of the distribution of different mutation types on treatment efficacy. These questions require larger patient cohorts and prospective studies.
At present, decisions regarding PARP inhibitor treatment still rely primarily on genetic testing results from blood or tissue samples.
Summary
The TALAPRO-2 Chinese cohort analysis suggests that understanding differences in treatment benefit requires consideration of prior treatment exposure, tumor burden, and the composition of HRR mutations together.
When applying biomarkers to clinical decision-making, it is important not only to perform appropriate testing but also to accurately interpret the origin and pathogenicity of detected variants, while distinguishing their prognostic significance from their ability to predict treatment response.
Some mutation-defined subgroups in the Chinese population remain small, and the corresponding benefit signals and effects of mutation type require validation in larger cohorts and prospective studies. Based on the current evidence, integrating blood- or tissue-based testing with individual disease characteristics to guide patient selection, while paying close attention to the safety of combination therapy, is an important step toward advancing individualized treatment.
Professor Hao Zeng
