In 2012, the U.S. Preventive Services Task Force (USPSTF) recommended prostate-specific antigen (PSA)-based screening for prostate cancer, regardless of age. However, the incidence of metastatic prostate cancer has rapidly increased since 2012 (1). Consequently, the incidence of metastatic castration-resistant prostate cancer (mCRPC) has been increasing. Several therapeutic agents are available for the treatment of mCRPC. However, the prognosis of patients with mCRPC remains poor. First-line treatments for castration-resistant prostate cancer (CRPC) include docetaxel and androgen receptor signaling inhibitors (ARSIs), such as enzalutamide, apalutamide, and abiraterone acetate.

Radium-223 treatment of mCRPC in the ALpharadin in SYMPtomatic Prostate Cancer (ALSYMPCA) trial was shown to reduce the risk of death by 30.5% compared with placebo and significantly improve overall survival (OS) with or without prior docetaxel treatment (2). However, it is unclear for whom or when radium-223 should be administered. In this study, we retrospectively analyzed the efficacy of radium-223 and its prognostic factors in patients with mCRPC.

The patients included in this study received radium-223 at the Kindai University Hospital from July 2016 to June 2022. All datasets are shown in Figure 1. Patients with an unclear Gleason Pattern (GP) or neuroendocrine carcinoma were excluded. Overall, 56 patients with mCRPC who received radium-223 were retrospectively analyzed. The following clinical and pathological data were obtained from medical records: age, PSA level, Eastern Cooperative Oncology Group performance status (ECOG-PS), pathological reports of transrectal ultrasound-guided needle prostate biopsy, follow-up data, and OS. The patients were followed up every three months. The metastatic site was evaluated using computed tomography (CT) and bone scintigraphy. Extent of disease (EOD) score was used to classify bone metastases (3).

This study was approved by the Institutional Ethics Committee of Kindai University (R02-247). The requirement for written informed consent was waived due to the retrospective nature of the study.

Statistical analysis. The results are presented as median values (range). The two groups (GP ≤4 and GP 5) were compared using the Mann–Whitney U-test and the χ²-test. The Kaplan–Meier estimate of the survival curve and log-rank test were used to explore the association between the parameters and patient survival. Multivariate analyses of patient age, PSA levels, GP, ECOG-PS, PSA changes, and alkaline phosphatase (ALP) changes were performed using Cox proportional hazard analyses to predict OS. All statistical analyses were performed using SPSS version 11.02 (SPSS, Chicago, IL, USA) and GraphPad Prism 9 (GraphPad Software, La Jolla, CA, USA). All p-values were two-sided, and statistical significance was set at p<0.05.

Patient characteristics are summarized in Table I. Figure 2A and B shows the OS after the diagnosis of CRPC and from the initiation of radium-223 treatment, respectively. The five-year OS rate in patients after diagnosis of CRPC was 62.2% [95% confidence interval (CI)=27.55-112.45], while the five-year OS rate in patients at the initiation of radium-223 treatment was 21.3% (95%CI=17.21-36.79). The median OS in patients after the diagnosis of CRPC was 75.0 months, whereas the median OS rate in patients at the initiation of radium-223 treatment was 29.0 months. Waterfall plots of PSA response and ALP decline rates are shown in Figure 3A and B, respectively. Six patients (11.1%) had a >50% PSA decline rate and 10 (17.9%) had a >50% ALP decline rate.

Next, we analyzed the prognostic factors for OS after initiating radium-223 treatment in patients treated with radium-223. Table II shows the results of univariate and multivariate Cox proportional hazards analyses for OS prediction. In the univariate analysis, the following were significantly associated with OS: PSA levels before radium-223 treatment (HR=1.00; 95%CI=1.00-1.00; p<0.001), GP 5 (HR=7.03; 95%CI=1.60-30.94; p=0.0099), EOD score (HR=3.05; 95%CI=1.20-7.76; p=0.0192), and ALP decline rate (HR=1.01; 95%CI=1.00-1.01; p=0.0278). The early administration of radium-223 as first- or second-line treatment was not associated with OS compared with the late administration of radium-223 as third-line treatment or later (HR=0.54; 95%CI=0.19-1.52; p=0.2412). In the multivariate analysis model 1, GP 5 (HR=5.48; 95%CI=1.10-27.31; p=0.0381), PSA levels before radium-223 treatment (HR=1.00; 95%CI=1.00-1.00; p=0.0050), and ALP change (HR=1.00; 95%CI=1.00-1.01; p=0.0275) were significantly associated with OS.

To predict the efficacy of radium-223, only the variables before the initiation of radium-223 treatment were used. Thus, we performed a multivariate analysis of these variables before radium-223 treatment. Multivariate analysis model 2 showed that GP 5 (HR=5.39; 95%CI=1.09-26.70; p=0.0389) and PSA levels before radium-223 treatment (HR=1.00; 95%CI=1.00-1.00; p=0.0073) were significantly associated with OS.

In both models, GP 5 was a poor prognostic factor for OS in patients treated with radium-223. When comparing the characteristics of two distinct groups of patients, 40 with GP 5 and 16 with GP ≤4, there was no significant difference in age at the time of radium-223 treatment, PSA levels before radium-223 treatment, or time to CRPC between patients with GP 5 and GP ≤4. There were significant differences in ECOG-PS and EOD scores between the two groups in which patients received radium-223. Figure 4 shows the Kaplan–Meier curves of the OS of GP 5 and GP ≤4. The five-year OS rate in patients with GP 5 was 0% (95%CI=0.00-0.00) and 75.0% (95%CI=17.21-36.79) in patients with GP ≤4. The median OS with GP ≤4 was 27.0 months. A significant difference in OS was observed between the GP 5 and GP ≤4 groups (log-rank test, p=0.0029).

In this study, we retrospectively analyzed the prognosis of patients with mCRPC treated with radium-223 and found that GP 5 was a poor prognostic factor in CRPC patients treated with radium-223.

The ALSYMPCA trial, which evaluated the efficacy of radium-223 in patients with advanced prostate cancer and bone metastases, demonstrated an increase in OS regardless of prior docetaxel therapy (2). In 2018, the European Medicines Agency (EMA) issued a statement recommending that radium-223 be restricted to third-line treatments or cases where other treatments are not applicable (4). Furthermore, one report advised against the concurrent use of abiraterone acetate and radium-223, citing concerns related to OS and the risk of fractures (4). Baucknht et al. developed a BIO-Ra score to categorize patients into three risk groups and predict outcomes in patients receiving radium-223 treatment (5,6). The BIO-Ra score consists of neutrophil-to-lymphocyte ratio (NLR: <3.1 vs. ≥3.1), ECOG-PS (0-1 vs. 2-3), number of bone metastases (<6 vs. 6-20 vs. ≥20), and ALP (<220 vs. ≥220). The median OS of the high-risk group was significantly lower than those of the low- and moderate-risk groups. Han-Chung Yang et al. reported a comparison between the Radium-223 treatment groups of less than 4 cycles and those receiving 5-6 cycles. They reported that both the median PSA progression-free survival and OS were prolonged in the group receiving ≥5 cycles compared to the group receiving less than ≤4 cycles (7). Matsumoto et al. reported that the presence or absence of anemia (Hb <12) and bone pain before treatment were prognostic factors associated with OS (8). Additionally, a distinction in OS was noted between radium-223 treatments comprising ≤2 cycles and those with ≥3 cycles, indicating that higher treatment completion rates associated with extended OS. Okabe et al. reported that the pre-DOC group exhibited significantly better outcomes compared to the post-DOC group in terms of the radium-223 treatment completion rate of 6 cycles, progression-free survival (PFS), and cause-specific survival (CSS) (9). The findings suggest the potential benefit of treating patients before disease progression, or in other words, administering treatment at an earlier stage. However, in the present study, there was no difference in survival between patients who received radium-223 earlier versus later. When examining the patients’ backgrounds between the two groups, it was noted that those who received radium-223 as the first- and/or second-line treatment had a shorter time to CRPC and poorer ECOG-PS than those who received it as third-line therapy or beyond (data not shown). In other words, the treatment could have been administered to patients with advanced disease. Nakashima et al. reported that serum ALP and BAP levels decreased in response to treatment, and regarding OS, the group with a BSI value of <2 showed a significant improvement compared to those with a BSI value of ≥2 (10). In the present study, we investigated the association between EOD score and OS, but no significant difference was observed. EOD score is defined on a 4-point scale (1-4), which may be problematic because it equally categorizes large and small metastatic lesions.

In most clinical studies, a Gleason score (GS) ≥8 was categorized as high-risk (11-13). The higher the GS, the more likely it is that the prostate cancer will progress rapidly. A GP 5 is pathologically different from GP 4 (14). GP 5 is characterized by a complete loss of glandular lumina, whereas GP 4 is characterized by fused glands or cribriform patterns (15). As the microscopic characteristics of GP 4 and 5 differ, the aggressiveness and responsiveness to therapeutic agents for GS 9, including GP 5, may differ from those for GS 8. GS is a prognostic factor for mCRPC (16). Previous studies suggested that the presence of GP 5 tumors in biopsy and surgical specimens is a prognostic factor after local and systemic treatments (13,17). Although little is known about the molecular alterations in these poorly differentiated tumors, Velho et al. characterized GP 5 prostate cancers using whole prostatectomy specimens (15). They reported a 4-fold increased risk of metastasis, and TP53 mutations, PTEN loss, and ERG expression were associated with biochemical failure and the risk of metastasis (15). Immunohistochemical analysis showed that in 22% of GP 5 patients with mutations in the homologous recombination pathway, 12% had mismatch repair deficiency, and 50% had PTEN loss. In the present study, radium-223 treatment for mCRPC, including GP 5, resulted in a poor prognosis. However, radium-223 treatment for mCRPC containing GP 5 may have a better prognosis than for mCRPC patients with GP 5 who did not receive radium-223, as the present study did not compare the radium-223-treated and radium-223-untreated groups. Patients with mCRPC GP 5 may benefit from treatment with PARP inhibitors, immuno-oncological drugs, and AKT inhibitors.

Study limitations. The study was based on a limited dataset and conducted exclusively with Japanese participants at a single facility. Therefore, regional factors must be considered when interpreting these results. Furthermore, this was a retrospective study. A prospective randomized trial is preferable for a more comprehensive understanding of treatment sequences.

Patients with mCRPC with GP 5 and high PSA levels before radium-223 treatment had a poor prognosis compared with patients with GP ≤4 and low PSA levels. In contrast, first- and second-line radium-223 treatments were not associated with OS. Further studies are necessary to determine the optimal timing of radium-223.

The Authors have no conflicts of interest to declare in relation to this study.

Conceptualization: Mitsuhisa Nishimoto and Kazutoshi Fujita; Methodology: Mitsuhisa Nishimoto and Kazutoshi Fujita; Software and Validation: Mitsuhisa Nishimoto and Kazutoshi Fujita; Formal Analysis: Mitsuhisa Nishimoto, Kazutoshi Fujita and Saizo Fujimoto; Investigation: Mitsuhisa Nishimoto, Kazutoshi Fujita, Aritoshi Ri, Saizo Fujimoto, Yasuo Oguma, Shingo Toyoda, Mamoru Hashimoto, Takashi Kikuchi, Shogo Adomi, Yoshitaka Saito, Yasunori Mori, Takafumi Minami, Masahiro Nozawa, Kazuhiro Yoshimura, Makoto Hosono and Hirotsugu Uemura; Resources: Mitsuhisa Nishimoto, Kazutoshi Fujita; Data Curation: Mitsuhisa Nishimoto, Kazutoshi Fujita; Writing – Original Draft Preparation: Mitsuhisa Nishimoto; Writing – Review & Editing: Kazutoshi Fujita; Visualization: Mitushisa Nishimoto and Kazutoshi Fujita; Supervision: Makoto Hosono and Hirotsugu Uemura; Project Administration: Makoto Hosono and Hirotsugu Uemura.

This study received no funding.