The literature in PubMed, Embase, and Cochrane Library up until April 27, 2023, was systematically searched using the following term combinations: (“Olaparib or AZD2281 or KU0059436”) and (“prostatic neoplasms or prostatic carcinoma or prostate cancer”) and (“randomized controlled trial”). The search strategies for the three databases are fully provided in Supplementary Table 1. Besides, potential articles were manually searched in related studies and reviews. The eligibility of titles and abstracts in all articles was independently assessed by two reviewers (ZL and BZ).

On the basis of PICOS criteria, two reviewers independently assessed the studies by the initial literature search, subsequently incorporating the studies that met the eligibility criteria. The studies included in this meta-analysis met the following criteria:

The exclusion criteria were as follows: phase I trials, letters, comments, reviews, or meta-analyses; studies without sufficient data and proper control drugs; and repeat publications.

When disagreements arose, a third reviewer was involved and evaluated all the eligible articles until a consensus was reached.

In accordance with the Cochrane Handbook guidelines, two investigators (YQ and JY) independently assessed and extracted the proper data including: first author, publication year, NCT number, inclusion criteria, phase of study, intervention and control group, median follow-up, and survival endpoints.

The ROB of the included RCTs was assessed using version 1.0 of the Cochrane Handbook’s ROB tool, which involves five domains: random sequence generation, allocation concealment, blinding, incomplete outcome data, and selective reporting. The ROB of each included study was assessed independently by two reviewers (ZL and YS). Each domain was classified as having a high, low, or unclear ROB.

The primary outcomes were rPFS and OS. The secondary outcomes included TFST, TSST, PFS2, confirmed PSA response, ORR, and adverse events.

rPFS is defined as the time from randomization to radiological progression (assessed by the investigator per the Response Evaluation Criteria in Solid Tumors [RECIST] 1.1 for soft tissue or the Prostate Cancer Working Group-3 [PCWG3] criteria for bone) or death from any cause (17).

OS is defined as the time from randomization to death from any cause (18).

TFST and TSST are defined as the time from randomization to the earlier of the first subsequent or the second subsequent anti-cancer therapy start date following study treatment discontinuation, or death, respectively.

PFS2 is defined as objective radiological progression by RECIST 1.1 for soft tissue, symptomatic progression, a rise in PSA level, or death in the absence of overall progression (19).

Confirmed PSA response is defined as a reduction in PSA level of 50% or more on two consecutive occasions at least 4 weeks apart compared with baseline (20).

ORR is defined as the percentage of patients with a complete response or partial response in soft tissue disease and a bone scan status of non-progressive disease (21).

The adverse events from olaparib combined with abiraterone and the control assessed in this article included total adverse events, grade 3 or worse adverse events, anemia, fatigue or asthenia, nausea, and diarrhea.

Statistical analysis was performed using Stata software version 12.0. For survival outcomes, the hazard Ratios (HRs) and 95% confidence intervals (CIs) for rPFS, OS, TFST, TSST, and PFS2 were extracted from each included article. Pooled HRs were obtained using the generic inverse of variance method with a random-effect model (22). When an HR < 1, it implied that the related endpoints of olaparib combined with abiraterone were associated with a lower risk than the control group. For dichotomous variables (i.e., ORR, confirmed PSA response and adverse events), relative risk (RR) was used to assess the differences between combination therapy and control interventions. The random-effect model was used to analyze all quantitative data. The results were analyzed and presented in forest plot. I² statistic and the Cochrane Q statistic were used to assess between-study heterogeneity. A value of I² greater than 50% and a p-value < 0.10 were considered indicative of a substantial level of heterogeneity (23). A p-value < 0.05 was considered statistically significant.

The results of all records identified in the search are shown in a flow diagram (Figure 1). A total of 603 studies were identified through the systematic literature search. Two potential records were also identified from the references of the reviews by hand searching. After 108 duplicate articles and 461 irrelevant articles were excluded on the basis of title or abstract screening, a total of 36 full-text articles and conference abstracts were considered potentially eligible for this review. Subsequently, 34 studies were excluded, including four with phase I trials; 11 with letters, comments, reviews, or meta-analyses; 14 with insufficient data; three with improper control drugs; and two with repeat publications. Finally, only 2 RCTs were included for assessment in this meta-analysis (24, 25).

The characteristics of the included studies in this meta-analysis are summarized in Table 1. Trials involving a total of 938 patients, including one phase II (24) and one phase III (25) trials, were conducted in Europe, America, Asia, and Oceania. All participants had histologically confirmed mCRPC, which included 470 men receiving olaparib combined with abiraterone and 468 men receiving abiraterone combined with placebo. Olaparib was administered orally at a dose of 300 mg two times daily, and abiraterone was administered orally at a dose of 1000 mg per day. More details about the characteristics of patients at baseline are shown in Supplementary Table 2. The ROB graph (Figure 2) indicated that the ROB was low across all domains.

To the best of our knowledge, this study is the first meta-analysis pooling the data from two RCTs involving a total of 938 patients with mCRPC, and comparing the combination use of olaparib with abiraterone versus abiraterone monotherapy. A key finding of this systematic review is that the olaparib combination therapy offers a considerable rPFS benefit for patients with mCPPC compared with abiraterone monotherapy. Furthermore, patients who received olaparib experienced significant benefits in terms of PFS2, TFST, TSST, and confirmed PSA response compared with those who received abiraterone. Meanwhile, the olaparib combination therapy showed a similar effect on OS, ORR, and several adverse events compared with abiraterone monotherapy.

The primary objective of this review was to evaluate the rPFS of olaparib in combination with abiraterone for patients with mCRPC. The data indicated that this combination offers a substantial rPFS benefit for patients with mCRPC, regardless of HRR mutation status (HR 0.66; 95% CI, 0.55–0.79; p < 0.01). Consistent with previous studies (10), the findings of the present study validated the synergistic interaction between PPAR inhibitor (olaparib) and androgen receptor antagonist (abiraterone), regardless of HRR mutation status. Previous studies (26–28) evaluating castration and androgen deprivation in prostate cancer revealed a potential connection between the androgen pathway and the DNA damage response. Inhibition of androgen signaling seems to decrease the expression of HRR genes and impair the ability of cells to repair DNA double-strand breaks, leading to increased sensitivity to RARP inhibition. This evidence could potentially elucidate the rationale behind the enhanced effectiveness of olaparib combination therapy.

The evaluation of the efficacy of anti-tumor drugs in clinical trials often relies on assessing the clinical outcome of OS (18). OS is considered one of the most important measures as it provides unambiguous and unbiased results. Positive outcomes in OS can serve as confirmatory evidence that a particular drug has potential to prolong life. However, in the present study, the combination of olaparib and abiraterone failed to show a statistically prolonged OS in patients with mCRPC. This finding may be attributed to factors such as the limited sample size and baseline prognostic factors (age, ECOG status and PSA concentration). A retrospective analysis of 15 studies demonstrated a robust correlation between PFS2 and OS, suggesting that PFS2 can serve as a reliable measure of long-term clinical benefit in cases where OS assessment is not feasible (29). By contrast, the present study showed a significant benefit in PFS2 (HR 0.72, 95% CI 0.56–0.93, p < 0.05). One plausible explanation for the disparity observed between rPFS and OS could be the influence of crossover and post-progression therapies (30). In clinical practice, the need for immediate subsequent treatment following disease progression may not always be necessary, because the process of clinical decision-making should involve a comprehensive evaluation of the patient’s physical health, tumor status, and clinical symptoms. Consequently, the time to subsequent therapy may be more clinically meaningful than the time to progression alone for patients. The findings of the present study suggested that the combination of olaparib and abiraterone has potential to prolong TFST (HR 0.75, 95% CI 0.63–0.89, p < 0.01) and TSST (HR 0.73, 95% CI 0.58–0.93, p < 0.05).

In this study, the ORR in the olaparib combination and control groups were 53.1% and 44.9%, respectively, without statistical significance (RR = 1.18, 95% CI 0.97–1.44, p > 0.05). The subgroup analysis revealed similar results for complete and partial responses (p > 0.05). However, a higher proportion of patients in the combination therapy group achieved a confirmed PSA response rate of at least 50% (349 [74.3%] of 470 patients) compared with the monotherapy group (304 [65.0%] of 468, RR 1.14, 95% CI 1.05–1.24, p < 0.01). A phase I clinical trial reported that abiraterone achieved a decline in PSA concentration in 58% of patients with mCRPC (31). In the present study, olaparib and abiraterone demonstrated a good activity in reducing PSA levels.

In this study, the four most commonly reported adverse events were anemia, fatigue or asthenia, nausea, and diarrhea. Olaparib combination therapy resulted in higher incidences of high-grade anemia and any-grade adverse events (fatigue or asthenia, nausea, and diarrhea; p<0.05) than abiraterone monotherapy, whereas similar effects were observed in the other adverse events (p>0.05). In total, 43.7% of patients in the olaparib combination group experienced any-grade anemia (16.0% had grade 3 or higher), compared with 14.1% of patients in the control group (2.8% had grade 3 or higher). A recent meta-analysis of 29 trials found that anemia is the most frequently reported hematological toxicity associated with PARPi (such as olaparib, rucaparib, veliparib, niraparib and talazoparib) when compared with alternative treatment (32). However, the underlying cause of anemia resulting from PARPi remains poorly understood. Preclinical evidence suggested that PARP-2 plays a crucial role in erythroid differentiation, and its deletion can lead to extravascular hemolytic anemia. This finding revealed that PARPi may influence hematopoiesis, potentially explaining the observed hematological adverse events in these studies (33, 34). Recent evidence revealed that olaparib induced temporary macrocytic anemia. However, the contribution of potential deficiencies in the vitamin B12 or folic acid pathway to the development of this anemia remains uncertain (35). A notable detail that additional toxicities were frequently observed, unrelated to hematological effects. For patients with underlying disorders, conducting a comprehensive evaluation of the risks and benefits, and closely monitoring safety during the initial stages of treatment are crucial. Considering these factors, healthcare professionals should be well-informed about the potential adverse events, including fatigue, nausea, and diarrhea, that may occur in patients receiving PARPi. They should also utilize laboratory tests, physical examinations, and clinical judgment before and during the treatment period to ensure thorough monitoring.

This meta-analysis possesses several notable strengths, including a comprehensive search across relevant databases and the inclusion of recent high-quality, well-designed, randomized, double-blinded trials. To our knowledge, this review is the first to systematically explore the efficacy and safety of olaparib and abiraterone combination therapy in patients with mCRPC. However, this systematic review had some unavoidable limitations. Firstly, the investigation of the effect of olaparib in combination with abiraterone on patients with different HRR mutation statuses is limited by the current lack of available phase II/III trials. Second, the inclusion of OS as a secondary endpoint in this study does not provide direct evidence of clinical benefit. Lastly, the number of studies (<10) included in the analysis is limited, and funnel plotting or Egger’s test was not conducted to assess the potential publication bias.