Real‐world first‐line systemic therapy patterns in metastatic castration‐resistant prostate cancer

Abstract Introduction Several systemic therapies have demonstrated a survival advantage in metastatic castration resistant prostate cancer (mCRPC). Access to these medications varies significantly worldwide. In Australia until recently, patients must have received docetaxel first, unless unsuitable for chemotherapy, despite no evidence suggesting superiority over androgen receptor signalling inhibitors (ARSIs). Our study investigated real‐world systemic treatment patterns in Australian patients with mCRPC. Methods The electronic CRPC Australian Database (ePAD) was interrogated to identify mCRPC patients. Clinicopathological features, treatment and outcome data, stratified by first‐line systemic therapies, were extracted. Comparisons between groups utilised Kruskal–Wallis tests and Chi‐Square analyses. Time‐to‐event data were calculated using Kaplan–Meier methods and groups compared using log‐rank tests. Factors influencing overall survival (OS) and time to treatment failure (TTF) were analysed through Cox proportional hazards regression models. Results We identified 578 patients who received first‐line systemic therapy for mCRPC. Enzalutamide (ENZ) was most commonly prescribed (n = 240, 41%), followed by docetaxel (DOC, n = 164, 28%) and abiraterone (AA, n = 100, 17%). Patients receiving ENZ or AA were older (79, 78.5 years respectively) compared with DOC (71 years, p = 0.001) and less likely to have ECOG performance status 0 (45%, 44%, 59% in ENZ, AA and DOC groups respectively p < 0.0001). Median TTF was significantly higher in those receiving ENZ (12.4 months) and AA (11.9 months) compared to DOC (8.3 months, p < 0.001). PSA50 response rates and OS were not statistically different. Time to developing CRPC > 12 months was independently associated with longer TTF (HR 0.67, p < 0.001) and OS (HR 0.49, p = 0.002). Conclusion In our real‐world population, ENZ and AA were common first‐line systemic therapy choices, particularly among older patients and those with poorer performance status. Patients receiving ENZ and AA demonstrated superior TTF compared to DOC, while OS was not statistically different. Our findings highlight the important role of ARSIs, given the variability of access worldwide.

International guidelines support the choice of any approved first-line systemic therapy, given the absence of validated predictive biomarkers. 7,8 In Australia however, the reimbursement of androgen receptor signaling inhibitors (ARSIs) was until recently, limited to patients previously treated with docetaxel, or those who are deemed unsuitable for chemotherapy. Our study aimed to examine real-world patterns of care among Australian patients with mCRPC, focusing on first-line treatment choices, toxicity and outcomes.

| METHODS
The electronic CRPC Australian Database (ePAD) was interrogated to identify patients who received systemic therapy for mCRPC. The ePAD multi-site clinical registry prospectively collects data including clinicopathologic details, treatment choices, response, toxicity and long-term outcomes for consecutive patients with CRPC. Nineteen Australian healthcare sites including private and public settings in both metropolitan and regional locations were included in the registry at the time of data extraction. Data are entered into the password-   Table 1 displays baseline characteristics for each treatment group, based on first-line therapy choice.

| Patient characteristics
Patients who received DOC or trial/other therapies were significantly younger (median age 71 years) compared to those treated with ENZ (79 years) and AA (78.5 years; p = 0.001) and were more likely to have an ECOG performance status of 0 (72% vs. 59% vs. 45% vs. 44% in trial/other, DOC, ENZ and AA groups respectively, p < 0.0001).

| Disease characteristics
There were no significant differences between treatment groups with regards to Gleason score, PSA doubling time (PSADT) or the presence of visceral metastases. Overall, 73 patients (13%) had received upfront docetaxel in the hormone sensitive setting. Only 4 (2%) in the DOC group received prior upfront chemotherapy compared to those treated with AA, ENZ and trial/other treatments (19%, 15%, 11% respectively; p < 0.001). Median pretreatment PSA was lowest in the trial/other group (5.0 ng/ml) compared to the AA, ENZ or DOC groups (16.2, 15.85 and 15.0 ng/ml; p = 0.001). The proportion of patients who had undergone prior 'watchful waiting', surgery or radiotherapy before initiation of systemic treatment was similar between groups.
Overall survival was not significantly different between groups (p = 0.77; Figure 2B).

| Multivariate analyses
Univariate and multivariate analyses were performed to assess the effect of individual variables on TTF and OS ( include PSADT, Gleason score and the presence of visceral metastases. 16,17 However, these did not appear to influence the choice of systemic therapy and were not independently associated with either TTF or OS in our cohort. When comparing our results with pivotal trials, median TTF with AA (11.9 months) and ENZ (12.4 months) was lower than the median radiographic progression-free survival (PFS) reported in the COU-AA-302 (16.5 months) and PREVAIL (20 months) studies. 4,18 PSA50 response rates were also lower in our cohort, likely reflecting differences between the trial and real-world populations. In our Australian cohort, this may reflect the limited reimbursement of ARSIs to patients deemed unsuitable for chemotherapy at the time of analysis. Similarly, recently reported real-world data relating to ENZ demonstrated lower PSA50 response rates in chemotherapy-naïve patients compared to the PREVAIL study. 19 Furthermore, another European real-world study demonstrated a lower PFS in those receiving AA of 10.8 months, likely related to the older age, higher rates of comorbidities and visceral metastases within their cohort compared to the COU-AA-302 study. 20 In patients receiving DOC however, the median TTF in our cohort (8.3 months) was higher than the median PFS in the FIRSTANA study (5.3 months), with similar PSA50 response rates than the pivotal TAX 327 study. 6,21 Importantly, the median age in our cohort was slightly higher than in the prior trials, which also involved narrow eligibility criteria, excluding patients with abnormal cardiac function or significant medical conditions. This highlights the importance of recognising differences between real-world patients compared with clinical trial populations and the challenges in extrapolating results in the real-world setting. In fact, in our cohort, patients who were enrolled in clinical trials were younger, had fewer comorbidities and better performance status. It is also likely that they commenced treatment earlier in their disease trajectory, reflected by a lower median pretreatment PSA.
The higher use of ENZ compared to AA in our study contradicts observations from previous international registries. [12][13][14] This likely reflects the differences in access or cost between ARSIs in many countries. For example, within several European countries the cost of abiraterone to patients is lower than that of enzalutamide due to differences in reimbursement regulations. 2 Whereas in Australia these drugs are both reimbursed equally. A phase II randomised trial comparing AA and ENZ in the first-line setting demonstrated higher PSA30 response rates with ENZ (82 vs. 68%) but no differences in time to PSA progression or overall survival. 22 Furthermore, inferior patient-reported outcomes relating to quality of life, fatigue and cognition have been reported with ENZ. 23 Our study similarly demonstrated lower rates of significant adverse events with AA compared to ENZ and treatment choice reflected the known toxicity profiles.
Patients receiving ENZ were more likely to have a history of ischaemic heart disease and more patients treated with AA had prior stroke history.
Subsequent therapies and potential cross-resistance are also significant considerations when making initial treatment choices.
For example, ARSIs have demonstrated survival benefits in chemotherapy-naïve and post-docetaxel settings. However, the use of a second ARSI following progression is associated with modest response rates. 24 Importantly, there are conflicting data regarding the effect of prior ARSIs on docetaxel efficacy with some studies demonstrating inferior PSA50 response rates, 25 while others have demonstrated no significant difference. 26 The high use of ARSIs in our cohort will allow the comparison of different treatment sequences and potential cross-resistance with longer follow-up.