From Genomic Testing to Olaparib Treatment: Real-World Utilization and Outcomes in BRCA-Mutated Metastatic Castration-Resistant Prostate Cancer Authors: Dianne Bosch, Kim J M van der Velden, Aart Beeker, Hendrik P van den Berg, André M Bergman, Maarten J van der Doelen, Alfons J M van den Eertwegh, Mathijs P Hendriks, Thomas M A Kerkhofs, Addy C M van de Luijtgaarden, Niven Mehra, Reindert J A van Moorselaar, Metin Tascilar, Walter L Vervenne, Nir I Weijl, Carin A Uyl-de Groot, Peter F A Mulders, Malou C P Kuppen, Inge M van Oort Published: 2026 ABSTRACT AbstractBackgroundIn November 2020, olaparib became the first approved poly adenosine diphosphate (ADP)-ribose polymerase inhibitor (PARPi) for metastatic castration-resistant prostate cancer (mCRPC) with BRCA1/2 mutations (BRCAm) in the Netherlands. As randomized clinical trials include fitter patients, their findings may not fully reflect real-world outcomes.ObjectiveThe aim was to evaluate genomic testing practice and subsequent use and outcomes of olaparib monotherapy in a real-world BRCAm mCRPC population.MethodsData were derived from ten hospitals in the Dutch CAPRI-3 registry, including mCRPC patients diagnosed between 2016 and 2021. Those receiving olaparib in standard-of-care treatment after its national approval (from November 2020) were analyzed and grouped as taxane-naïve (TN) or post-taxane (PT). The primary endpoint was overall survival (OS).ResultsAmong 1996 mCRPC patients, genomic analysis (somatic and/or germline) was performed in 23.4% (range 3.8–63.2% across hospitals), identifying BRCAm in 11.3% of patients. Tested patients differed significantly in age, comorbidities, and prior treatment. Among 35 eligible BRCAm patients, 27 (77.1%) received olaparib. TN patients (8/27) were significantly older and initiated olaparib at an earlier line of therapy. Median OS was 28.7 months (95% confidence interval (CI) not reached) in TN versus 10.5 months (95% CI 9.6–11.5) in PT patients (p = 0.003). Limitations include the retrospective design and small subgroups.ConclusionsGenomic testing application remained limited and uneven across centers. Most eligible patients received olaparib; TN patients seemed to benefit most.Supplementary InformationThe online version contains supplementary material available at 10.1007/s40801-026-00548-8. Key Points Uneven genomic testing in practice: In this large real-world cohort of 1996 Dutch metastatic castration-resistant prostate cancer (mCRPC) patients (2016–2021; follow-up to January 2024), genomic testing was limited and inconsistently applied across medical centers.Benefit of earlier treatment: Despite the small number of eligible BRCAm patients, most (27/35) received olaparib, with more favorable outcomes observed in the earlier initiated taxane-naïve setting.Call to action for personalized care: Lowering the threshold for genomic testing is essential to optimize personalized therapy in metastatic prostate cancer. Introduction In the last decades, several new life-prolonging drugs have become available for metastatic castration-resistant prostate cancer (mCRPC) [1]. In the last years, poly adenosine diphosphate (ADP)-ribose polymerase inhibitors (PARPi) have gained interest, as approximately 13% of mCRPC patients harbor germline and/or somatic BRCA1/2 mutations (BRCAm), associated with response to PARPi [2–5]. In 2020, olaparib was the first PARPi to be approved as monotherapy for mCRPC patients who progressed on prior treatment with an androgen receptor pathway inhibitor (ARPI) (i.e., enzalutamide or abiraterone acetate plus prednisone [AAP]). The European Medicines Agency (EMA) indication restricted use to BRCAm mCRPC (November 2020) [6], whereas the United States (US) Food and Drug Administration (FDA) indications also included other homologous recombination repair gene mutations (HRRm) (May 2020) [7]. Approval was based on the results of cohort A from the PROfound trial, demonstrating a clear benefit in both radiographic progression-free survival (rPFS) (7.4 vs. 3.6 months) and overall survival (OS) (19.1 vs. 14.7 months) in the subgroup of BRCAm compared to the physician's choice of a second ARPI [8, 9]. Nowadays, ARPI retreatment is considered a non-effective comparator arm and a substitute for placebo [1, 10–12]. In 2022, the PROpel trial demonstrated a significant rPFS benefit regardless of HRRm in mCRPC patients who received first-line therapy of AAP plus olaparib compared to AAP plus placebo [13]. However, planned subgroup analyses on HRRm and exploratory analysis on BRCAm clearly demonstrated a stronger efficacy in the biomarker-selected populations. Following the results of the PROpel trial, the EMA extended the indication of olaparib to combinational treatment with AAP for allcomers in whom chemotherapy is not indicated (December 2022) [6], whilst the FDA restricted the combination to cases with BRCAm only (May 2023) [14]. Despite the expanded indications for olaparib, real-world (RW) data on its outcomes remain scarce. There are a few published cohorts, exhibiting short follow-up durations and primarily focus on PFS [15–17]. The aim of this study was to assess genomic testing practice and subsequent utilization and clinical outcomes of olaparib monotherapy in a RW BRCAm mCRPC population in the Netherlands. Materials and Methods Study Design The CAPRI-3 prostate cancer registry is a retrospective cohort study conducted in the Netherlands, including patients with mCRPC diagnosed between 2016 and 2021 [18]. For this analysis, data were drawn from ten hospitals, with baseline defined as the start of olaparib treatment and follow-up continued until death or the database cut-off on 7 January 2024. Eligible patients received olaparib in standard of care (SOC) after its national approval on 1 November 2020; those treated through compassionate use or clinical trials were excluded. mCRPC was defined according to the 2024 European Association of Urology (EAU) guidelines [1] or by physician documentation. Patients with insufficient data were excluded. Written informed consent was obtained from living patients by the treating physician, except for those lost to follow-up. The study was approved by the Medical Research Ethics Committee Oost-Nederland (CMO 2020-6573) and all participating institutions. Subgroups The total cohort was retrospectively stratified based on whether patients underwent genomic analysis (germline and/or somatic) following their prostate cancer (PCa) diagnosis. Among patients treated with olaparib in SOC, two subgroups were defined based on the timing of olaparib initiation: a taxane-naïve (initiation prior to, or in the absence of, taxane chemotherapy) and post-taxane subgroup (initiation following any line of taxane chemotherapy). Genomic Analysis Genomic analyses were conducted in accordance with the local protocols and technological capabilities of each participating center, encompassing both germline and somatic (i.e., tumor) testing. The database captured the presence of genetic alterations and affected genes. The types of genetic alterations (e.g., point mutations, insertions, deletions) and sequencing approaches (specific targeted panels or whole-genome sequencing) were not available. Study Outcomes In Supplementary Table 1, all study outcomes and their definitions are provided in detail (see the electronic supplementary material). The primary endpoint was OS, measured from the time of olaparib initiation. Secondary endpoints were the occurrence of progressive disease (PD), progression-free survival (PFS), rPFS, and prostate-specific antigen (PSA) response (PSA50, defined as ≥50% decline from baseline), all calculated from the start of olaparib treatment and in accordance with the Prostate Cancer Working Group 3 (PCWG3) [19] and prostate‐specific membrane antigen (PSMA) Position Emission Tomography (PET) Progression (PPP) criteria [20]. For the overall cohort, secondary endpoints included the prevalence of genomic testing and the presence of homologous recombination repair (HRR) gene alterations. At baseline measurements and prescriptions of 90 days prior to 7 days after olaparib initiation were considered. Statistical Analysis Subgroups were compared using the chi-square or Fisher’s exact tests for categorical variables and the Mann-Whitney U or Kruskall-Wallis tests for continuous variables. Median OS (mOS), PFS, and rPFS with 95% confidence intervals (CIs) were estimated using the Kaplan-Meier method, and differences between groups were assessed by the log-rank test. A two-sided p value of < 0.05 was considered statistically significant. Analyses were performed using SPSS 27.0 (IBM Corp., Armonk, NY, USA). Figures were created using GraphPad Prism version 9.0 (GraphPad Software Inc., Boston, MA, USA), R-4.1.3 (R Software Inc., San Francisco, CA, USA), and Lucidchart (Lucid Software Inc., South Jordan, UT, USA). Results Total Population and Genomic Analysis At database cut-off, the registry included 1996 mCRPC patients from ten medical centers: 1109 (55.6%) from five teaching hospitals, 697 (35.0%) from three university medical centers (UMCs), and 190 (9.5%) from two general hospitals. Genomic analysis (somatic and/or germline) was conducted in 467 patients (23.4%), with testing rates across hospitals ranging between 3.8 and 63.2%. Of these, 275 patients (58.9%) were tested at a single referral UMC. Patients who were selected for genomic analysis were significantly younger (69 vs. 77 years, p < 0.001), had lower age-adjusted Charlson Comorbidity Index (CCI) scores (CCI < 5 in 80.7% vs. 52%, p < 0.001), and received upfront intensified systemic treatment (chemotherapy ± ARPI) for metastatic hormone-sensitive prostate cancer (mHSPC) more often (35.4% vs. 12.2%, p < 0.001) (Table 1). Genomic analysis revealed HRRm in 107 of 467 patients (22.9%), including BRCAm in 53 patients (11.3%). BRCAm consisted of BRCA1 in eight patients (1.7%), BRCA2 in 44 patients (9.4%), and both BRCA1 and BRCA2 in one patient (0.2%). Table 1Patient and disease characteristics of the total cohort at mCRPC diagnosisTotal population, N = 1996Any genomic testing, N = 467No genomic testing, N = 1529PAge (years), median (IQR)75 (69–81)69 (63–74)77 (70–83)< 0.001Age-adjusted CCI, n (%)< 0.001 0–2354 (17.7)163 (34.9)191 (12.5) 3–4818 (41.0)214 (45.8)604 (39.5) 5–6513 (25.7)74 (15.8)439 (28.7) > 6311 (15.6)16 (3.4)295 (19.3) MissingMetastases, n (%)< 0.001 Nodal disease only316 (15.8)89 (19.1)227 (14.9) Bone ± nodal disease1321 (66.2)307 (65.7)1014 (66.4) Visceral ± other sites228 (11.4)63 (13.5)165 (10.8) Unknown130 (6.5)8 (1.7)123 (8.0)Received treatments for mHSPC, n (%)< 0.001 ADT only1403 (70.3)221 (47.3)1182 (77.3) ADT + EBRT146 (7.3)55 (11.8)91 (6.0) ADT + taxane chemotherapy308 (15.4)140 (30.0)168 (11.0) ADT + ART ± taxane chemotherapy43 (2.2)25 (5.4)18 (1.2) Other45 (2.3)18 (3.9)27 (1.8) Unknown51 (2.6)8 (1.7)43 (2.8)Time from mHSPC to mCRPC (months), median (IQR)19.6 (11.4–38.7)21.4 (13.0–38.6)18.9 (11.0–38.7)0.634Bold represent significant valuesADT androgen deprivation therapy, ART androgen receptor targeted therapy (either abiraterone or enzalutamide), CCI Charlson Comorbidity Index, EBRT external beam radiotherapy, IQR interquartile range, mCRPC metastatic castration-resistant prostate cancer, mHSPC metastatic hormone-sensitive prostate cancer Patient and disease characteristics of the total cohort at mCRPC diagnosis Bold represent significant values ADT androgen deprivation therapy, ART androgen receptor targeted therapy (either abiraterone or enzalutamide), CCI Charlson Comorbidity Index, EBRT external beam radiotherapy, IQR interquartile range, mCRPC metastatic castration-resistant prostate cancer, mHSPC metastatic hormone-sensitive prostate cancer Olaparib Treatment Among 1996 mCRPC patients, 81 (4.1%) received olaparib since 2018. Three patients (3.7%) received olaparib through compassionate use prior to its national approval (1 November 2020), and 51 (63.0%) received olaparib as part of a clinical trial conducted between 2018 and 2023; these patients were excluded from further analysis (Figure 1). Trial enrollment included PROpel (n = 20), Drug Rediscovery Protocol (DRUP) (n = 16), KEYLYNK-010 (n = 13), and PROfound (n = 2). Flow diagram of the patient identification process. mCRPC metastatic castration-resistant prostate cancer, SOC standard of care At the time of national olaparib approval, 1021 patients were alive, including 35 with BRCAm mCRPC. Of these, 27 patients (77.1%) were administered olaparib in SOC and met the eligibility criteria. Baseline Characteristics of Patients Treated with Olaparib Median age at baseline was 71 years (interquartile range [IQR] 64–76) and differed significantly between the taxane-naïve and post-taxane subgroup (76 vs. 68 years, p = 0.013) (Table 2). Taxane-naïve patients received olaparib in earlier treatment lines as compared to post-taxane (≤ 2nd line of treatment [LOT] in 100% vs. 52.6% of patients, p = 0.026). All other baseline characteristics held no significant differences. All patients received olaparib following progression on an ARPI, and none received concomitant therapies. Table 2Patient and disease characteristics at olaparib initiation (unless otherwise specified)Olaparib, N = 27Taxane-naïve, N = 8Post-taxane, N = 19PAge (years), median (IQR)71 (64–76)76 (70.3–82.3)68 (63–73)0.013Age-adjusted CCI, n (%)0.070 0–29 (33.3)1 (12.5)8 (42.1) 3–49 (33.3)2 (25)7 (36.8) 5–67 (25.9)3 (37.5)4 (21.1) > 62 (7.4)2 (25)0 (0)ECOG performance status, n (%)0.456 02 (7.4)0 (0)2 (10.5) 15 (18.5)2 (25)3 (15.8) 20 (0)0 (0)0 (0) Missing20 (74.1)6 (75)14 (73.7)Gleason score, n (%)0.311 < 83 (11.1)2 (25)1 (5.3) ≥ 821 (77.8)5 (62.5)16 (84.2) Missing3 (11.1)1 (12.5)2 (10.5)Metastatic disease at initial diagnosis, n (%)9 (33.3)4 (50)5 (26.3)0.657Metastases, n (%)0.364 Nodal disease only4 (14.8)2 (25)2 (10.5) Bone ± nodal disease15 (55.6)5 (62.5)10 (52.6) Visceral ± other sites8 (29.6)1 (12.5)7 (36.8)PSA (ng/mL), median (IQR)38 (8.7–175)33 (3.9–99.8)45 (8.7–255)0.511 Missing, n (%)2 (7.4)0 (0)2 (10.5)ALP (U/L), median (IQR)110.5 (73.5–278.8)94.5 (74.8–136.5)134 (72.8–328.3)0.528 Missing, n (%)3 (11.1)0 (0)3 (15.8)LDH (U/L), median (IQR)238 (200.5–282)213.5 (197.5–234)255 (207.5–305)0.140 Missing, n (%)2 (7.4)0 (0)2 (10.5)Hb (mmol/L), median (IQR)7.4 (6.3–8.2)7.8 (6.4–8.2)7.4 (6.3–8.3)0.511 Missing, n (%)2 (7.4)0 (0)2 (10.5)Genomic analysis, n (%)0.459 Somatic testing19 (70.4)6 (75)6 (75) Germline testing3 (11.1)0 (0)3 (15.8) Somatic and germline testing5 (18.5)2 (25)3 (15.8)HRR gene alterations, n (%)0.487 BRCA12 (7.4)0 (0)2 (10.5) BRCA225 (92.6)8 (100)17 (89.5)Previous ARPI0.266 Abiraterone only13 (48.1)4 (50)9 (47.4) Enzalutamide only13 (48.1)3 (37.5)10 (52.6) Abiraterone and enzalutamide1 (3.7)1 (12.5)0 (0)Previous taxane chemotherapyNR Docetaxel only15 (55.6)–15 (78.9) Docetaxel and cabazitaxel4 (14.8)–4 (21.1) No taxane8 (29.6)8 (100)–Previous platinum-based chemotherapy*0.136 Carboplatin only6 (22.2)0 (0)6 (31.6)Time from mHSPC to mCRPC (months), median (IQR)11.2 (8.6–19.4)10.0 (4.8–17.8)11.2 (8.9–21.7)0.307Time from mCRPC to olaparib initiation (months), median (IQR)15.3 (7.8–24.5)9.2 (5.7–45.7)15.4 (10.4–22.1)0.418Line of treatment**, n (%)0.026 1st or 2nd line18 (66.7)8 (100)10 (52.6) ≥ 3rd line9 (33.3)0 (0)9 (47.4)Bold represent significant valuesALP alkaline phosphatase, ARPI androgen receptor pathway inhibitor, CCI Charlson Comorbidity Index, ECOG Eastern Cooperative Oncology Group, Hb hemoglobin, HRR homologous recombination repair, IQR interquartile range, LDH lactate dehydrogenase, mCRPC metastatic castration-resistant prostate cancer, mHSPC metastatic hormone-sensitive prostate cancer, NR no result, PSA prostate-specific antigen*Cisplatin, carboplatin, or etoposide**After mCRPC diagnosis Patient and disease characteristics at olaparib initiation (unless otherwise specified) Bold represent significant values ALP alkaline phosphatase, ARPI androgen receptor pathway inhibitor, CCI Charlson Comorbidity Index, ECOG Eastern Cooperative Oncology Group, Hb hemoglobin, HRR homologous recombination repair, IQR interquartile range, LDH lactate dehydrogenase, mCRPC metastatic castration-resistant prostate cancer, mHSPC metastatic hormone-sensitive prostate cancer, NR no result, PSA prostate-specific antigen *Cisplatin, carboplatin, or etoposide **After mCRPC diagnosis Overall Survival The mOS was 13.4 months (95% CI 3.7–23.1). Taxane-naïve patients had a significantly longer mOS of 28.7 months (95% CI not reached) as compared to 10.5 months for post-taxane patients (95% CI 9.6–11.5, p = 0.003) (Table 3). Table 3Outcomes of olaparib treatmentOlaparib, N = 27Taxane-naive, N = 8Post-taxane, N = 19POS* (months), median (95% CI)13.4 (3.7–23.1)28.7 (NR)10.5 (9.6–11.5)0.003 Censored, n (%)7 (25.9)5 (62.5)2 (10.5)PFS* (months), median (95% CI)8.0 (2.9–13.2)21.1 (12.6–29.7)4.6 (2.4–6.8)< 0.001 Censored, n (%)5 (19.2)4 (50)1 (5.6)rPFS* (months), median (95% CI)12.7 (10.8–14.6)21.6 (13.0–30.2)8.7 (4.3–13.1)0.052 Censored, n (%)13 (48.1)4 (50)9 (47.4)Best PSA response (%), median (IQR)− 52.9 (− 21.4 to −86.3)−84.0 (− 54.4 to − 93.5)− 32.9 (78.6 to − 84.8)0.041 PSA50, n (%)10 (55.6)6 (75)4 (21.1)0.013 Missing, n (%)9 (33.3)2 (25)7 (36.8)Bold represent significant valuesCI confidence interval, IQR interquartile range, NR no result, OS overall survival, PFS progression-free survival, rPFS radiographic PFS, PSA prostate-specific antigen, PSA50 PSA ≥ 50% decline from baseline*Measured from start of olaparib treatment Outcomes of olaparib treatment Bold represent significant values CI confidence interval, IQR interquartile range, NR no result, OS overall survival, PFS progression-free survival, rPFS radiographic PFS, PSA prostate-specific antigen, PSA50 PSA ≥ 50% decline from baseline *Measured from start of olaparib treatment Six patients received prior platinum-based chemotherapy and showed an mOS of 7.3 months (95% CI 2.7–11.9), compared to 20.1 months (95% CI 8.7–31.5, p = 0.003) in those without prior platinum exposure. Progression-Free Survival PD was observed in 22 patients (81.5%), of which 14 cases (51.9%) were classified as radiographic ± PSA, seven cases (25.9%) as PSA only, and one case (3.7%) as clinical only. Median PFS was 8.0 months (95% CI 2.9–13.2) and median rPFS 12.7 months (95% CI 10.8–14.6) (Table 3). Taxane-naïve patients had a significantly longer median PFS of 21.1 months (95% CI 12.6–29.7) as compared to 4.6 months (95% CI 2.4–6.8, p < 0.001) for post-taxane patients. PSA Response PSA50 was found in 55.6% of patients, with a median best PSA response of −52.9% (IQR −21.4% to −86.3%) (Table 3). In the taxane-naïve subgroup compared to the post-taxane subgroup, PSA50 was observed in significantly more patients (75% vs. 21.1%, p = 0.013), and best PSA response was significantly greater (−84% [IQR −54.4% to −93.5%] vs. −32.9% [IQR 78.6% to −84.8%], p = 0.041). Treatment Course A swimmer plot (Figure 2) depicts the treatment course for each individual patient. The median duration of olaparib treatment was 8.8 months (95% CI 4.2–13.5), and four patients (14.8%) remained on olaparib treatment at the time of database cut-off. Swimmer plot showing the disease and treatment course from mCRPC diagnosis until death or end of follow-up. Each bar represents an individual patient. ARPI androgen receptor pathway inhibitor, mCRPC metastatic castration-resistant prostate cancer, PB platinum-based Discussion Genomic analysis was performed in 23.4% of 1996 mCRPC patients, with a substantial variation of 3.8% up to 63.2% across medical centers, identifying BRCAm in 11.3% of patients. Despite the very limited number of eligible BRCAm patients, 77.1% (27/35) of those harboring these mutations and alive at the time of national approval were treated with olaparib in SOC. Taxane-naïve patients were older and received olaparib at an earlier LOT, showing beneficial outcomes in both OS and PFS compared to the post-taxane setting. Our genomic testing rate of 23.4% is generally low and primarily driven by a single UMC, yet it falls between the 15.7% reported in Europe and 38.2% in the US from a large, prospective, RW survey in 2020 [21]. The disparity between Europe and the US may stem from differences in genomic testing accessibility (74.6% in Europe vs. 90.3% in the US [21]), reimbursement, and variations in guideline recommendations. In May 2019, the US National Comprehensive Cancer Network (NCCN) was the first to recommend both germline and somatic testing for all patients with metastatic PCa, irrespective of family history or disease subtype [22]. In September 2020, the general European Society of Medical Oncology (ESMO) guideline took a selective approach by merely recommending germline BRCA2 testing in these patients, and only to consider somatic testing for HRRm in the mCRPC setting [12]. In contrast, the recommendations from the ESMO Precision Medicine Working Group in November 2020 did advocate a proactive endorsement of somatic testing in PCa patients where PARPi are accessible [23]. The EAU addressed genomic testing in its April 2021 guideline, recommending germline BRCA1/2 testing in PCa patients depending on their family and medical history, and somatic testing for HRRm in all metastatic PCa patients, ideally before the initiation of first-line mCRPC therapy [1]. By 2023, the ESMO updated its guidelines to broadly support germline and somatic testing in metastatic PCa, better harmonizing with other major guidelines. Note that the Dutch PCa guideline currently contains no recommendations regarding genomic testing [24]. In our study, genomic profiling identified HRRm in 22.9% of cases, markedly lower than the 31% and 39% recently reported in Europe and the US, respectively [21]. BRCAm was detected in 11.8% of the tested cohort, representing a higher prevalence compared to the 8.3% reported in a recent large-scale study conducted in the US [25]. This may be due to missing data on panel types in our cohort, as some patients were likely tested with BRCA-focused rather than broader HRR panels. Our data revealed selection bias in patients selected for genomic testing. Eligibility for olaparib therapy may be influenced by patient-specific factors (e.g., bone marrow function, performance status), irrespective of BRCAm status. However, our registry highlighted significant variability in testing rates across hospitals, reflecting systematic disparities in access to advanced diagnostics. This may stem from differences in healthcare providers' perspectives, patient selection criteria, and population characteristics [21]. Importantly, the overrepresentation of genomic testing in specific centers might skew the representativeness of RW data. We observed an mOS of 13.4 months in the total cohort of BRCAm mCRPC patients treated with olaparib monotherapy following progression on ARPI in SOC. To date, there are only two RW studies that examined olaparib in this setting. Xie et al. observed a comparable mOS of 12.0 months in 86 patients with DNA damage repair–deficient mCRPC (of whom 73% harbored BRCA1/2 or ATM mutations) [26]. Additionally, an abstract from the 2024 American Society of Clinical Oncology Genitourinary Cancers Symposium reported a longer mOS of 16.6 months in 144 HRRm mCRPC patients [27]. These limited RW data show outcomes broadly similar to our findings but remain shorter than the mOS of 19.1 months demonstrated in the PROfound trial’s cohort A (HRRm mCRPC) [8], which is likely due, at least in part, to the inherent patient selection in randomized clinical trials. Our median PFS of 8.0 months is consistent with findings in RW studies, varying between 5.3 and 8.0 months among those with HRRm [15–17, 26], and is also similar to the 7.4 months reported in the PROfound trial [8]. Our data revealed beneficial outcomes in the taxane-naïve subgroup versus post-taxane subgroup, both in PFS (21.1 vs. 4.6 months, p < 0.001) and mOS (28.7 vs. 10.5 months, p = 0.003). However, due to significant imbalance between these subgroups, the observed benefit is likely biased by confounding factors. Supporting literature from the PROfound trial demonstrated a greater rPFS benefit of olaparib versus the control group (ARPI retreatment) in taxane-naïve patients (14.6 vs. 3.7 months, hazard ratio [HR] 0.20, 95% CI 0.12–0.34) than in post-taxane patients (9.0 vs. 1.9 months, HR 0.13, 95% CI 0.05–0.31) [28]. Regarding other PARPi, the 2023 TRITON-3 study evaluated rucaparib versus SOC in the taxane-naïve setting following progression on ARPI in BRCAm mCRPC, and showed that rucaparib versus docetaxel significantly improved rPFS (11.2 vs. 8.3 months, HR 0.53, 95% CI 0.37–0.77) [29]. In the earlier disease setting, the recently presented AMPLITUDE trial in HRRm metastatic castration-sensitive prostate cancer showed that niraparib plus AAP versus placebo plus AAP significantly improved rPFS and showed a trend toward an OS benefit at the first interim analysis [30]. These findings support that earlier initiation of PARPi is associated with improved patient outcomes. Nevertheless, neither the PROfound nor the TRITON-3 trial reported a significant OS benefit, emphasizing the importance of selecting appropriate control arms and study endpoints. Our study has several limitations. Despite screening a large cohort of mCRPC patients, only a limited subset received olaparib. We encountered high rates of missing values in some baseline and outcome variables and significant differences in baseline characteristics between the subgroups. As the PCWG3 and PPP criteria have been primarily created for prospective use, biased results can be expected as there are no prespecified intervals of imaging/laboratory testing in retrospective analyses. Taking all of this into consideration, outcome differences between our subgroups should be interpreted with caution. Nevertheless, our research fills a gap in knowledge as RW data on utilization and outcomes of olaparib treatment for mCRPC are scarce. Our findings highlight the need for prospective validation of olaparib treatment per treatment phase – specifically with and without prior taxane exposure. Also, we suggest examining the barriers for genomic testing. Conclusions This RW analysis of 1996 mCRPC patients diagnosed in the Netherlands between 2016 and 2021 and followed until January 2024 highlights substantial variability in genomic testing and olaparib utilization across medical centers. Despite the very small number of eligible BRCAm patients, most received olaparib treatment, and taxane-naïve patients seemed to benefit most. However, OS was shorter compared to outcomes reported in prospective research. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (PDF 92 KB) Funding This research was funded by the CAPRI foundation, registered at the Dutch Chamber of Commerce in December 2019. This nonprofit organization aims to serve as the data source on metastatic prostate cancer in the Netherlands. The foundation is sponsored by the following pharmaceutical companies: AAA Novartis, Astellas, Bayer, and MSD/AstraZeneca. The pharmaceutical companies had no role in the design and conduct of the study, collection of the data, management of the data, analysis, interpretation of the data, and the preparation, review, or approval of the manuscript. Declarations Conflict of interest D. Bosch received speaker honoraria from MSD. A.M. Bergman received research grants and honoraria from Astellas, Sanofi, Bayer, and Janssen. M.J. van der Doelen received research grants from Bayer and Janssen and received speaker fees from Astellas. A.J.M. van den Eertwegh received study grants from Bristol-Myers Squibb and Idera and travel fees from Ipsen, and is a member of the advisory boards of Bristol-Myers Squibb, MSD Oncology, Ipsen, Pierre Fabre, and Janssen. A.C.M. van de Luijtgaarden received honoraria from Astellas, Bayer, Eisai, Ipsen, Janssen, MSD/AstraZeneca, AAA Novartis, and Pfizer, and is a member of the advisory board for the Dutch Prostate Cancer Foundation. N. Mehra reports grants from Janssen-Cilag, Astellas Pharma, AstraZeneca, Bristol Myers Squibb Foundation, and MSD Oncology and personal and institutional fees for consulting or advisory role from Janssen-Cilag, Astellas Pharma, AstraZeneca, MSD Oncology, Bayer, and Pfizer. R.J.A. van Moorselaar received speaker fees from or is a member of the advisory boards of Astellas, AstraZeneca, Bayer, BMS, Curium, Janssen and Pantarhei Bioscience. W.L. Vervenne received honoraria from Servier Nederland Farma. P.F.A. Mulders received research funds from Astellas, AstraZeneca, Telix, and Byondis. M.C.P. Kuppen received research grants from AAA Novartis, Astellas, Bayer, and MSD/AstraZeneca. I.M. van Oort is a member of the advisory boards for MSD/AstraZeneca, Pfizer, Astellas, and Bayer; received a research grant from AAA Novartis; and received research grants and personal fees from Astellas, Bayer, Janssen Pharmaceuticals, MSD/AstraZeneca, and Pfizer. The other authors declare no conflicts of interest. Ethical Approval The study was approved by the Medical Research Ethics Committee Oost-Nederland (CMO 2020-6573) and all participating institutions. Consent to Participate Written informed consent was obtained from living patients by the treating physician, except for those lost to follow-up. Consent for Publication Not applicable. Availability of Code, Data, and Material The custom code (SPSS syntaxes) and datasets generated and/or analyzed during the current study are not publicly available due to license agreements between the CAPRI foundation, participating hospitals, and sponsoring parties, but are available from the corresponding author on reasonable request and with permission of the CAPRI foundation. Authors' Contributions DB and KJMV performed the acquisition and (statistical) analysis of data. DB interpretated the data and drafted the manuscript. MCPK and IMO obtained funding, performed administrative support, and provided supervision. All authors contributed to the conception and design and critical revision of the manuscript. All authors read and approved the final manuscript. References Associated Data Supplementary Materials Supplementary file1 (PDF 92 KB)