Survival outcomes and prognostic factors for first-line abiraterone acetate or enzalutamide in patients with metastatic castration-resistant prostate cancer

Our retrospective cohort study examined the outcomes of mCRPC patients receiving first-line ARAT, specifically patient survival and PSA treatment response. While the enzalutamide group showed a higher rate of PSA decline (≥ 90% from baseline) compared to the abiraterone group, there was no significant difference in OS between the two groups. These real-world outcomes align with a meta-analysis of published phase III randomized trials in patients with mCRPC in both pre- and post-docetaxel settings demonstrating enzalutamide having a longer time to PSA progression and higher response rate compared to abiraterone, but no difference in OS [10]. Evidence suggests that enzalutamide shows better results than abiraterone with prednisone in terms of PSA progression and response rate but not OS.

In this present study, the median OS for patients who received first-line abiraterone was 42.1 months (IQR 20.6–63.7), which was longer than the OS reported in clinical trials involving first-line treatment with ARAT, 35.5 months with enzalutamide in PREVAIL trial [5] and 34.7 months in COU-AA-302 trial [4]. In the Prostate Cancer Registry data from 16 countries, the median OS for real-world first-line treatment of ARAT was even shorter at 27.1 months [18]. Our cohort’s longer OS may be due to ethnic differences in response to androgen-based treatments for prostate cancer. Prior studies have shown that Asian populations demonstrate greater effectiveness and longer survival times with ADT compared to other ethnic groups [19]. This result could potentially apply to the treatment outcomes with ARAT in mCRPC patients as well and warrants further investigation.

In this study, it was observed that 65% (131/202) patients received only one-line therapy for mCRPC, which is more than the clinical practice reflected in real-world data from the United States, where 51% of patients received only one life-prolonging therapy [20]. One of the primary reasons for this difference is that the effects of ADT and ARAT can persist for a prolonged period, especially among the Taiwanese population as mentioned previously, and Taiwan’s National Health Insurance provides continued reimbursement for these treatments. Furthermore, patients receiving ARAT alone had a higher median age compared to those receiving ARAT-chemotherapy, in both abiraterone and enzalutamide groups (Supplementary Table 1) as well as in patients with 2 poor prognostic factors (Supplementary Table 2). Moreover, patients may be reluctant to undergo chemotherapy due to its side effects and the psychological stress associated with the treatment. This reluctance may be even more pronounced among patients who have already received three to four years of treatment, as aging and physical decline become significant considerations when deciding on next-line therapy. Finally, it should be noted that 24% (49/202) of patients in this study were still using ARAT with an effective response at the time of analysis, indicating that the observation time for this cohort may not have been long enough.

We classified patients who had 2 poor prognostic factors, PSA nadir > 2 ng/mL and TTN < 7 months, as high risk for all-cause mortality. In the subgroup of patients with 2 poor prognostic factors who received abiraterone as first-line ARAT, the overall survival was longer for those who received subsequent chemotherapy. This suggests that subsequent chemotherapy may compensate for the less effective treatment of abiraterone. However, it was important to acknowledge that the patients who received subsequent chemotherapy were substantially younger than those who received an ARAT alone (67 v. 83 years) which may explain the prolonged OS in the subsequent chemotherapy groups, although there were no differences between the two groups in terms of ECOG PS and prevalence of cardiovascular comorbidities. In addition, our study had a relatively small sample size. Thus, our results should be considered hypothesis-generating and require further validation in larger studies. Based on these findings, providers may consider a switch to chemotherapy as soon as possible if a patient is found to have a poor response after receiving abiraterone as first-line treatment for mCRPC, particularly in those with 2 poor prognostic factors. Compared to abiraterone, patients treated with enzalutamide as first-line ARAT showed a more favorable response, resulting in a lower need for second-line therapy due to treatment failure. More importantly, if patients could continue to use ARAT, it indicated that the treatment effect was good. In contrast, patients who required a switch to chemotherapy were associated with a more severe disease pattern and poorer outcome. Consequently, the survival benefit of using chemotherapy in the enzalutamide cohort was challenging to discern.

Past studies have examined the predictive effect of PSA nadir and TTN on disease progression in metastatic prostate cancer patients treated with ADT. PSA nadir higher than 0.2 ng/ml and time to PSA nadir less than 10 months were predictors of disease progression and refractory disease in metastatic prostate cancer [15, 21]. In patients with CRPC, studies also showed the impact of PSA nadir and TTN after ADT treatment on the OS [16] and cancer-specific survival [22]. In terms of TTN, 7 months was reportedly the cut-off time interval for the evaluation of effective treatment, especially the efficacy of conventional ADT in metastatic prostate cancer [23]. This study is the first to evaluate a TTN cut-off point in patients with mCRPC treated with first-line ARAT with a TTN of 7 months also being supported by our ROC curve analysis. In terms of PSA nadir, our ROC curve analysis demonstrated a PSA of 2 ng/mL as a significant cut-off value for stratification.

This study has a few limitations. This was a retrospective analysis and the sample size was relatively small. There might be selection bias or other confounding factors, even though we used adjusted multivariate and IPTW model to reduce the interference of these factors when analyzing prognostic factors. Furthermore, since this study used Asian ethnic groups, the proposed cut-off points need to be carefully interpreted if they are to be applied to other ethnic groups. Therefore, our proposed prognostic factors still need to be validated with other real-world data such as PREMISE, a large European observational study in mCRPC patients treated with enzalutamide [24]. OS data from PREMISE trial has not been published yet. In addition, due to the excellent effect of ARAT and long survival period of the patients in this study, some patients still continue to receive ARAT. The median survival of patients receiving first-line ARAT only is premature. To our knowledge, this study is the first to investigate PSA nadir and TTN as predictors of ARAT effectiveness as first-line treatment in patients with mCRPC. These two well-defined prognostic factors with clear cut-off points can be easily used by clinicians as a reference for evaluating treatment effects and anticipating patient outcomes.