Prostate cancer is the second most common cancer after lung cancer and accounts for 7% of all new cancer diagnoses in men worldwide (1, 2). The standard of care for metastatic castration-resistant prostate cancer (mCRPC) includes taxane-based chemotherapy, radiotherapy, and novel antiandrogens (NAA), such as abiraterone, enzalutamide, darolutamide, or apalutamide. Although these treatment options have shown the ability to improve overall survival (OS) (3–6), subsequent NAA therapy is known to work only in a minority of patients and the responses are short-lived (7, 8). There is an urgent need to evaluate non-NAA monotherapy approaches in light of the growing number of patients receiving their first NAA therapy before developing castration-resistant disease.

Notably, 15-30% of patients with mCRPC have mutations in homologous recombination repair (HRR) genes (including BRCA1, BRCA2, ATM, and CHEK2) (9, 10). Olaparib has been recently approved by the U.S. Food and Drug Administration (FDA) as a potent orally bioavailable poly(ADP-ribose) polymerase (PARP) inhibitor to treat BRCA-1/2-deficient patients (11). It possesses both in vitro cellular potency and in vivo efficacy, and it led to 80% tumor inhibition when administered in combination with temozolomide in mice (12). In a preliminary phase II study, patients with mCRPC who previously received NAA treatment were administered olaparib. Tumors with HRR mutations showed more positive responses to olaparib than those without mutations (13). In preclinical studies, olaparib synergizes with agents that affect the androgen receptor (AR) pathway regardless of the HRR mutation status (14). As patients develop resistance to NAA treatment, understanding the mechanism by which co-inhibition of AR and PARP proteins leads to the inability of cancer cells to repair DNA may facilitate the use of a PARP inhibitor (PARPi) in combination with NAA. However, the overall benefits of olaparib monotherapy and combination therapy still need to be evaluated. Therefore, we performed indirect comparisons and network meta-analyses to assess the efficacy and toxicity of olaparib, olaparib combined with abiraterone, and NAA to provide a basis for clinical drug selection for patients with mCRPC.

Approximately 100,000 men in the United States have non-metastatic castration-resistant prostate cancer (nmCRPC), and 86% may progress from non-metastatic CRPC to mCRPC (22). In the past decade, olaparib was the first true targeted therapy approved for prostate cancer, as a response requires the presence of DNA damage repair aberrations (23). And thanks to the results of PROpel trial, olaparib in combination of abiraterone acetate plus predisolone was approved by the FDA to treat patients with previously untreated mCRPC harboring deleterious or suspected deleterious BRCA mutations in May 2023. In studies of olaparib, olaparib plus abiraterone, and apalutamide plus abiraterone, rPFS was the primary endpoint. Preliminary data showed that all three interventions prolonged rPFS compared to abiraterone monotherapy, and all were statistically significant. The median rPFS results for all reported trials were comparable, whereas the median rPFS was longer in the intervention group than in the abiraterone group ( Table 2 ). Notably, unlike the TOPARP-A study, in which the efficacy of olaparib monotherapy in patients with mCRPC was limited to patients with HRR mutations, a trial of olaparib in combination with abiraterone (NCT01972217) showed a potential clinical benefit for patients regardless of HRR mutations (13). In addition, while direct comparisons between olaparib and NAA monotherapy (enzalutamide or abiraterone) were made in the PROfound trial, there was no direct comparison between olaparib and NAA combination therapy in the current study. The inclusion of the ACIS trial for indirect comparison with olaparib, without considering the patient’s HRR mutation status, addressed this issue to some extent. Because all three therapies demonstrated similar results in rPFS, it is important to differentiate the utility of these drugs based on their differences. Our network meta-analysis indirectly demonstrated that olaparib had better rPFS than apalutamide plus abiraterone, but there seemed to be no difference between olaparib plus abiraterone and the other two interventions. To further define the options, the rank probability model in rPFS demonstrated that the preferred drug treatment was olaparib, followed by olaparib plus abiraterone and apalutamide plus abiraterone. The subgroup analysis did not reveal a difference in the rPFS benefit between olaparib and olaparib plus abiraterone in patients with HRR mutations.

The analyses of OS and PFS2 reported in each trial are immature. The OS reported in the preliminary studies did not attain statistical significance. Possible causes are that the patient received subsequent life-prolonging treatments, which did not include the three interventions we studied. In addition, uncertainty regarding the optimal sequencing of the combination of two active therapies in mCRPC has resulted in no statistically significant data on the OS benefit, and clinical trials to date have confirmed this view (20). Although there were no significant differences in the indirect comparisons, the magnitude and direction of PFS2 data were consistent with the primary rPFS analysis.

The development of response biomarkers to rapidly identify drug-resistant diseases and guide early therapeutic conversion remains an unmet clinical need. The value of CTC as prognostic indicators for advanced prostate cancer has been well described (24, 25). These data indicate that CTC progression with low baseline CTC counts (<5) during the first 12 weeks of chemotherapy or endocrine therapy can identify patients who will not benefit from treatment. This could guide patient response assessment during the first 12 weeks of treatment, identify early disease progression, and serve as a biomarker of efficacy in clinical trials (26). In olaparib plus abiraterone trial, CTC conversion rates were similar in both arms; whereas in the olaparib alone trial the intervention arm was higher than the comparator arm (27% vs. 10%). In the olaparib plus abiraterone and apalutamide plus abiraterone trials, the PSA response rates were similar in both arms, which might be a consequence of the potency of abiraterone and its cytostatic mode of action (27). Thus, the difference in rPFS between the two groups may be due to an increased proportion of patients with stable disease and a longer duration of response in the combination arm rather than an increase in the number of patients with a complete response.

Adverse effects and quality-of-life aspects of treatment options are important factors in the decision-making process, as cancer presentation and long-term survival rates were similar in patients treated with the three interventions. The four most common AEs in patients treated with olaparib were anemia, nausea, fatigue/asthenia, and decreased appetite, consistent with the results of a meta-analysis of the risk of AEs in studies on olaparib in other tumor types (28). The most common AE reported in the phase II clinical trials of olaparib plus abiraterone was nausea. The most common ≥3 AE, however, was anemia, which was also consistent with the results reported by PROpel. As we were unable to obtain a detailed number of each adverse event in PROpel in the trial of olaparib plus abiraterone, the analysis of the AEs results in our study was based primarily on the results of a phase II clinical trial. Thus, insufficient patient numbers may account for the difference in the choice of olaparib, with or without abiraterone, for AEs and ≥3 AEs. A further study of PROfound found an increased risk of four common adverse events not significantly associated with prior taxane therapy or BRCA alterations (29). The incidence of these four AEs peaks in the first two months of treatment and can be controlled by dose adjustment and supportive therapy without stopping treatment, which suggests that the management of AEs should be accompanied by a focus on the clinical benefit of patients.

CAPTURE trial reported that BRCA patients, regardless of somatic/germline origin, exhibited considerably lower rPFS, PFS2, and OS, as well as significantly worse PFS2 and OS than non-BRCA patients in the HRR subgroup. These poor results were not attributable to the BRCA 1/2 subgroup receiving fewer active mCRPC therapies. And these situations could be avoided by adding PARPi (30). PARPi other than Olapali are also worth attention. The TRITON3 trial compared the efficacy of rucaparib monotherapy to the doctor’s choice of chemotherapy, abiraterone, or enzalutamide in chemotherapy-naive patients with BRCA or ATM mutations (31). In the experimental group, rPFS considerably improved (median 10.2 vs. 6.4 months; HR, 0.61; 95% CI 0.47-0.80; P = 0.0003). In the subgroup analysis, rucaparib maintained its rPFS advantage in the BRCA-mutated cohort, but there was no difference between the two treatment arms in patients with ATM mutations (median rPFS 8.1 vs. 6.8 months; HR, 0.97; 95% CI 0.59–1.52; P = 0.84). In another phase III trial TALAPRO-2, patients were randomized to receive enzalutamide plus talazoparib or enzalutamide plus placebo; all patients benefited from combination therapy regardless of HRR status, but the benefit was more pronounced in patients with HRR mutations (rPFS 27.9 VS. 16.4; HR, 0.46; 95% CI 0.3–0.7; P < 0.001) (32). Another FDA-approved PARPi combination therapy was based on MAGNITUDE; this trial, which included the largest BRCA1/2 cohort to date in first-line mCRPC, demonstrated that niraparib plus abiraterone significantly improved rPFS in BRCA1/2 patients compared with abiraterone plus placebo (19.5 vs. 10.9 months; HR, 0.55; P= 0.0007) (33). Consistent with PARPi monotherapy, the combination regimen was most beneficial for BRCA mutations patients, followed by HRR mutations patients, and most unfavorable for HRR proficient mCRPC patients. These results further support the importance of screening germline and somatic BRCA1/2 alterations for a more precise patients care (30).The assumption that PARPi binding to NAA is beneficial regardless of BRCA mutation status is being challenged based on recent final overall survival data from the PROpel trial, as well as interim analyses of the TALAPRO-2 and MAGNITUDE trials. The discrepancy between PROpel results and the actual FDA drug approval also highlights that the need to explore whether patients with non-HRR mutations benefit from the combination of PARPi with NAA (34). In addition, given the inconsistent inclusion and design criteria of these PARPi trials, which do not allow for direct comparison of results, the selection and sequencing of these drugs will be increasingly challenging in the clinic.

The findings of this review should be interpreted within the context of its limitations. First, our analysis relied on indirect comparisons, which have inherent limitations and can be misleading if there are significant differences in baseline study populations between studies. Second, we chose HR and OR; different utility measures provide explanations for drug efficacy that must be viewed with caution and validated further. Additionally, the included studies did not report the type of HRR-mutated genes, and the difference between PARP inhibitors combined with or without NAA in patients with HRR mutations needs to be further explored. As the analyses of AEs in the olaparib plus abiraterone trials were insufficient, we only studied overall AEs and ≥3 AEs and did not perform further data analysis on the four common AEs caused by PARP inhibitors. Despite these limitations, we carefully selected evidence, and the screened studies were high-quality randomized controlled trials with similar patient-selection criteria. We also demonstrated the utility of PARP inhibitors in prolonging the survival of patients with high-risk prostate cancer. We also provided a basis for clinicians to target specific endpoints.

We found that all three interventions including olaparib, olaparib plus abiraterone and apalutamide plus abiraterone significantly improved rPFS. In the absence of found differences in efficacy between interventions, olaparib might be the optimal choice. Apalutamide plus abiraterone might be taken into consideration as a substitute for olaparib and olaparib plus abiraterone when AEs limited their use.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/ Supplementary Material .

XC and YP designed the study and analyzed the data. QW and CR revised the images. ML, XH, and LX performed the literature search and collected data for the manuscript. XL revised the manuscript. All authors contributed to the article and approved the submitted version.