ABSTRACT
Introduction
In real‐world practice in Japan, standard treatment for metastatic hormone‐naïve prostate cancer (mHNPC) is androgen deprivation therapy (ADT), administered as monotherapy, combined androgen blockade (CAB), ADT plus androgen receptor pathway inhibitors (ARPIs), or ADT plus docetaxel. In a previous interim analysis of the large‐scale, longitudinal, multicentre, J‐ROCK registry study of real‐world clinical and patient‐reported outcomes, ADT plus ARPI or ADT plus docetaxel was used more frequently than ADT/CAB in patients (aged ≥ 20 years) with newly diagnosed LATITUDE‐criteria high‐risk mHNPC.
Methods
This post hoc analysis of the J‐ROCK study evaluated prostate‐specific antigen (PSA) response, progression‐free survival (PFS), time to castration‐resistant prostate cancer (CRPC), overall survival (OS) and safety in patients with high‐risk mHNPC who received ADT/CAB (cohort 1) or ADT plus ARPI (cohort 2B) in subgroups were defined according to the following known prognostic factors at baseline: age, Gleason Grade Group (GGG), alkaline phosphatase (ALP), hemoglobin (Hb) and lactate dehydrogenase (LDH).
Results
This analysis included 947 evaluable patients (371 in cohort 1 and 576 in cohort 2B). PSA response rates remained similar among age and GGG subgroups in both cohorts, but were reduced in cohort 2B patients with elevated ALP, low Hb, and elevated LDH. Time to CRPC and OS were longer in cohort 2B than in cohort 1, regardless of prognostic factors. Among patients with poor prognosis (older, high GGG, low Hb, elevated LDH), OS decline occurred earlier in cohort 1 versus cohort 2B. A trend towards a plateau in the time to CRPC curve was observed in both cohorts, even in patients with poor prognosis. Safety data were not affected by prognostic factors or treatment.
Conclusions
These findings suggest that novel ADT plus ARPI regimens for LATITUDE‐criteria high‐risk mHNPC may improve real‐world outcomes compared with ADT monotherapy or CAB, particularly among patients with poor prognosis.
Keywords: androgen deprivation therapy, androgen receptor pathway inhibitor, combined androgen blockade, J‐ROCK study, metastatic hormone‐naïve prostate cancer, prognostic factors
1. Introduction
The standard treatment options for metastatic hormone‐naïve prostate cancer (mHNPC) have undergone a paradigm shift over the last decade from conventional treatment, such as androgen deprivation therapy (ADT) as monotherapy or combined androgen blockade (CAB; vintage hormonal therapy) to ADT plus androgen receptor pathway inhibitors (ARPIs; abiraterone acetate plus prednisolone [AAP], enzalutamide, or apalutamide) or docetaxel (doublet therapy) [1, 2, 3]. Triplet therapy with darolutamide plus ADT plus docetaxel was added to the treatment options in Japan in 2023 [4].
The treatment options for metastatic castration‐resistant prostate cancer (mCRPC) in Japan are ARPI and chemotherapy [5, 6], with radium‐223 chloride, AAP plus olaparib, enzalutamide plus talazoparib or talazoparib monotherapy, pembrolizumab, and 177Lu‐vipivotide tetraxetan being available in selected patients. The mainstay of treatment in patients with metastatic hormone‐sensitive prostate cancer (mHSPC) or mCRPC is ARPIs or chemotherapy only. The availability of triplet therapy represents a further treatment option for patients with mHNPC, but more real‐world data on clinical outcomes with vintage, doublet, or triplet therapy is needed to guide physicians and patients in selecting the optimal treatment.
There are no clinical studies comparing the efficacy and safety of triplet (darolutamide plus ADT plus docetaxel) versus doublet (ADT plus ARPI) therapy; the comparator arm in the phase 3 ARASENS study was placebo plus ADT plus docetaxel [4]. Therefore, the trade‐off benefit of combining docetaxel with ADT plus ARPI (triplet therapy) when considering the added risk of adverse events (AEs) is currently unclear [7, 8]. Furthermore, combining docetaxel with ADT plus ARPI in patients with mHSPC may limit treatment options in those who progress after triplet therapy. In real‐world clinical practice, treatment selection should also consider the patient's condition, quality of life, economic burden, family environment, and residence, in addition to treatment effectiveness and safety.
The J‐ROCK study is a large‐scale, longitudinal, multicentre registry trial of real‐world clinical and patient‐reported outcomes among Japanese adult patients (aged ≥ 20 years) with newly diagnosed, LATITUDE criteria [9] high‐risk mHNPC [10]. We have previously reported findings from the first and second interim analyses after 18 months and 3 years of follow‐up, respectively [10, 11]. In the second interim analysis, we reported improved effectiveness among patients who received ADT plus ARPI or docetaxel (cohort 2) compared with those who received vintage therapy with ADT monotherapy or CAB (cohort 1) [10]. Moreover, compared with cohort 2, an early decline in overall survival (OS) was observed among patients in cohort 1.
Many previous studies have extracted the prognostic factors of doublet therapy in patients with mHNPC [2, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22]. The objective of this post hoc analysis of 3‐year data from the second interim analysis of the J‐ROCK study was to clarify the unmet needs of current high‐risk mHNPC treatment by analyzing the outcomes of patients with poor prognosis after vintage and doublet therapy, using known prognostic factors [2, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22], and to explore whether there is an early decline in the OS curve in patients with known prognostic factors. We performed subgroup analyses of the effectiveness and safety of treatment using baseline prognostic factors with few missing data, including age [12, 23], Gleason Grade Group (GGG) [13, 14, 15, 16, 17, 18, 20, 21], alkaline phosphatase (ALP) [13, 14, 18, 20, 21], hemoglobin (Hb) [2, 13, 19, 22], and lactate dehydrogenase (LDH) [16, 20, 21]. Considering the mechanisms of action of the drugs used in two cohorts [10], these analyses included patients from cohort 1 (ADT monotherapy or CAB) and cohort 2B (ADT plus ARPI); cohort 2 A (ADT plus docetaxel), which had a small number of patients, was excluded [10].
2. Materials and Methods
2.1. Study Design
Full details of the J‐ROCK registry study (ClinicalTrials. gov identifier: NCT04034095; University Hospital Medical Information Network‐Clinical Trial Registration [UMIN‐CTR] identifier: UMIN000037127) have been reported previously [10, 11]. Based on the treatment received in the first 3 months after diagnosis, patients were allocated to either cohort 1 (received ADT alone or CAB), cohort 2A (received ADT plus docetaxel), or cohort 2B (received ADT plus AAP, enzalutamide, or apalutamide). Patients who initiated treatment with a cohort 1 regimen but later received a cohort 2A or 2B regimen at least once within the first 3 months after diagnosis were allocated to cohort 2A or 2B.
2.2. Patients, Subgroups and Outcomes
Outcomes, including imaging and PSA levels, were collected according to the routine clinical practice of each physician. The outcomes for these subgroup analyses were prostate‐specific antigen (PSA) response rates, progression‐free survival (PFS), time to castration‐resistant prostate cancer (CRPC), OS, and safety [10]. PSA response rates included the proportion of patients who achieved PSA ≤ 0.2 ng/mL within 3 months after treatment initiation. PFS was defined as the duration from diagnosis of high‐risk mHNPC to either radiographic progression, clinical progression, or death, whichever occurred first. Time to CRPC was defined as the time from diagnosis of high‐risk mHNPC to diagnosis of CRPC. OS was defined as the duration from diagnosis of high‐risk mHNPC to death from any cause. Subgroups were defined according to the following cut‐off values: 75 years for age, 4 for GGG, 113 U/L for ALP, 13.1 g/dL for Hb, and 222 U/L for LDH. These cut‐off values were chosen because: age > 75 years is generally considered the definition of “elderly” in Japan [24]; patients with a GGG of 5 are known to have poor prognosis compared with those with a GGG < 4 [13, 14, 15, 16, 17, 18, 20, 21]; and typical standard values for Japanese real‐world settings were used for ALP (113 U/L), Hb (13.1 g/dL), and LDH (222 U/L). Patients with missing data for these baseline prognostic factors were excluded from this analysis.
2.3. Statistical Analysis
Descriptive statistics were used to evaluate patient demographics, baseline characteristics with continuous variables presented as median (range), and binary outcome endpoint (PSA responses and safety outcome) with categorical variables presented as the number and proportion of patients. PFS, time to CRPC, and OS were estimated using the Kaplan–Meier method and compared between cohorts. Kaplan–Meier curves were not adjusted. Comparisons were estimated using multivariate Cox regression models, with adjusted hazard ratios (aHRs) and 95% confidence intervals (95% CIs) calculated, using the number of bone metastases, visceral metastases, Gleason score, age, PSA, comorbidities (cardiovascular disorders, respiratory disorders, renal disorders, hepatic disorders, neurological disorders, diabetes, and other clinically important disorders), LDH, ALP, Hb, SSE, and metastasis stage at first diagnosis of PC (M1 vs. M0) as confounding variables. Patients with missing data on prognostic factors were excluded from each analysis. All statistical analyses were performed using SAS version 9.4 software (SAS Institute, Cary, NC, USA).
3. Results
3.1. Study Population
At the time of the second interim analysis cut‐off date, 983 patients were enrolled in the J‐ROCK study from 77 sites in Japan. The second interim analysis included 974 evaluable patients (371 in cohort 1, 27 in cohort 2 A, and 576 in cohort 2B) and had a median observation period of 22.7 months [10]. Patient demographics and baseline characteristics of cohort 1 and cohort 2B are shown in Table 1, and those in each cohort according to the different prognostic factor subgroups are shown in Supporting Information: Tables S1–S5.
Table 1.
Patient demographic and baseline clinical characteristics in cohorts 1 and 2B.
| Cohort 1a(N = 371) | Cohort 2Bb(N = 576) | |
|---|---|---|
| Age, years | 75.0 (47.0–94.0) | 72.0 (45.0–92.0) |
| PSA, ng/mL | 288.115 (1.00–24565.10) | 349.000 (2.06–21016.32) |
| Gleason score, n (%) | ||
| < 7 | 0 | 1 (0.2) |
| 7 | 2 (0.5) | 7 (1.2) |
| 8 | 143 (38.5) | 201 (34.9) |
| 9 | 186 (50.1) | 276 (47.9) |
| 10 | 26 (7.0) | 76 (13.2) |
| Unknown | 14 (3.8) | 15 (2.6) |
| Tumor stage, n (%) | ||
| Tx | 6 (1.6) | 14 (2.4) |
| T0 | 0 | 0 |
| T1 | 1 (0.3) | 5 (0.9) |
| T2 | 42 (11.3) | 53 (9.2) |
| T3 | 184 (49.6) | 314 (54.5) |
| T4 | 125 (33.7) | 180 (31.3) |
| Unknown | 13 (3.5) | 10 (1.7) |
| Node stage, n (%) | ||
| Nx | 3 (0.8) | 6 (1.0) |
| N0 | 141 (38.0) | 171 (29.7) |
| N1–N3 | 216 (58.2) | 383 (66.5) |
| Unknown | 11 (3.0) | 16 (2.8) |
| Metastasis stage, n (%) | ||
| Mx | 1 (0.3) | 2 (0.3) |
| M0 | 6 (1.6) | 9 (1.6) |
| M1 | 348 (93.8) | 542 (94.1) |
| Unknown | 16 (4.3) | 23 (4.0) |
| Clinical stage, n (%) | ||
| I | 0 | 0 |
| II | 1 (0.3) | 5 (0.9) |
| III | 3 (0.8) | 1 (0.2) |
| IV | 355 (95.7) | 551 (95.7) |
| Unknown | 12 (3.2) | 19 (3.3) |
| Treatment history, n (%) | 5 (1.3) | 10 (1.7) |
| Radiation therapy | 3 (0.8) | 5 (0.9) |
| Surgery | 4 (1.1) | 9 (1.6) |
| Systemic therapy | 2 (0.5) | 1 (0.2) |
Note: Values are median (range) unless stated otherwise. Patients with missing data on the aforementioned baseline prognostic factors were excluded from the analysis.
Abbreviations: ADT, androgen deprivation therapy; ARPI, androgen receptor pathway inhibitor; CAB, combined androgen blockade; PSA, prostate‐specific antigen.
Treated with ADT monotherapy or CAB.
Treated with ADT plus ARPI.
3.2. PSA Response
The proportion of patients who achieved a ≥ 90% PSA decline or a PSA of ≤ 0.2 ng/mL within 3 months tended to be higher in cohort 2B than in cohort 1 (Supporting Information: Table S6). In the prognostic factor subgroups, PSA response rates in cohorts 1 and 2B remained similar regardless of baseline age or GGG, whereas the proportion of patients in cohort 2B who achieved a PSA of ≤ 0.2 ng/mL within 3 months was approximately 10% lower in those with elevated ALP, low Hb and elevated LDH compared with patients without these prognostic factors (Table 2).
Table 2.
Prostate‐specific antigen response outcomes within 3 months of treatment initiation in patients with metastatic mHNPC in cohorts 1 and 2B in subgroups defined by prognostic factors.
| n (%) | Cohort 1a | Cohort 2Bb | ||
|---|---|---|---|---|
| Age, years | < 75 (n = 178) | ≥ 75 (n = 193) | < 75 (n = 373) | ≥ 75 (n = 202) |
| Achieved ≥ 90% PSA decline | 124 (69.7) | 133 (68.9) | 327 (87.7) | 176 (87.1) |
| Achieved PSA ≤ 0.2 ng/mL | 17 (9.6) | 15 (7.8) | 117 (31.4) | 54 (26.7) |
| GGG | ≤ 4 ( n = 147) | 5 ( n = 221) | ≤ 4 ( n = 211) | 5 ( n = 361) |
| Achieved ≥ 90% PSA decline | 99 (67.3) | 158 (71.5) | 187 (88.6) | 312 (86.4) |
| Achieved PSA ≤ 0.2 ng/mL | 12 (8.2) | 20 (9.0) | 60 (28.4) | 109 (30.2) |
| ALP, U/L | < 113 ( n = 123) | ≥ 113 ( n = 221) | < 113 ( n = 185) | ≥ 113 ( n = 352) |
| Achieved ≥ 90% PSA decline | 79 (64.2) | 160 (72.4) | 165 (89.2) | 310 (88.1) |
| Achieved PSA ≤ 0.2 ng/mL | 11 (8.9) | 16 (7.2) | 64 (34.6) | 95 (27.0) |
| Hb, g/dL | ≥ 13.1 ( n = 166) | < 13.1 ( n = 188) | ≥ 13.1 ( n = 306) | < 13.1 ( n = 243) |
| Achieved ≥ 90% PSA decline | 116 (69.9) | 130 (69.1) | 272 (88.9) | 212 (87.2) |
| Achieved PSA ≤ 0.2 ng/mL | 17 (10.2) | 13 (6.9) | 105 (34.3) | 60 (24.7) |
| LDH, U/L | < 222 ( n = 222) | ≥ 222 ( n = 116) | < 222 ( n = 363) | ≥ 222 ( n = 171) |
| Achieved ≥ 90% PSA decline | 161 (72.5) | 74 (63.8) | 315 (86.8) | 156 (91.2) |
| Achieved PSA ≤ 0.2 ng/mL | 19 (8.6) | 8 (6.9) | 120 (33.1) | 36 (21.1) |
Abbreviations: ADT, androgen deprivation therapy; ALP, alkaline phosphatase; ARPI, androgen receptor pathway inhibitor; CAB, combined androgen blockade; GGG, Gleason Grade Group; Hb, hemoglobin; LDH, lactate dehydrogenase; PSA, prostate‐specific antigen.
Treated with ADT monotherapy or CAB.
Treated with ADT plus ARPI.
3.3. PFS, Time to CRPC and OS
The Kaplan–Meier curves of PFS, time to CRPC, and OS for cohorts 1 and 2B are shown in Supporting Information: Figure S1. Regardless of prognostic factors, PFS (Figure 1), time to CRPC (Figure 2), and OS (Figure 3) tended to be longer in cohort 2B when compared with cohort 1. PFS events occurred throughout the observation period (Figure 1), while the Kaplan–Meier curve of the time to CRPC tended to plateau over time, even in subgroups with poor prognostic factors (Figure 2). Despite the divergence in PFS and time to CRPC between cohorts, there was no difference in OS between cohorts 1 and 2B in the GGG ≤ 4 (Figure 3B) or Hb > 13.1 g/dL (Figure 3D) subgroups due to the short observation period. When OS was assessed in prognostic subgroups defined by age, ALP, Hb, and LDH, there was a relatively early separation of Kaplan–Meier curves between cohorts (Figures 3A,C–E). The aHRs for the time to CRPC and OS in cohort 1 versus cohort 2B tended to be smaller among patients with poor prognostic factors than in those without poor prognostic factors (Figures 2 and 3), indicating that cohort 2B treatment may provide improved outcomes over cohort 1, particularly among patients with poor prognosis.
Figure 1.
Kaplan–Meier curves of PFS in patients with high‐risk metastatic mHNPC in cohorts 1a and 2Bb across prognostic factor subgroups defined by (A) age, (B) Gleason Grade Group, (C) alkaline phosphatase level, (D) Hb level, or (E) lactate dehydrogenase level. In each graph, the 24‐month event‐free rate is indicated by the black dotted lines and percentage value in blue (cohort 1) or red (cohort 2B). ADT, androgen deprivation therapy; aHR, adjusted hazard ratio; ALP, alkaline phosphatase; ARPI, androgen receptor pathway inhibitor; CAB, combined androgen blockade; CI, confidence interval; GGG, Gleason Grade Group; Hb, hemoglobin; LDH, lactate dehydrogenase; mHNPC, metastatic hormone‐naïve prostate cancer; mPFS, median progression‐free survival; NR, not reached. aTreated with ADT monotherapy or CAB. bTreated with ADT plus ARPI. [Color figure can be viewed at wileyonlinelibrary.com]
Figure 2.
Kaplan–Meier curves of time to CRPC in patients with high‐risk metastatic mHNPC in cohorts 1a and 2Bb across prognostic factor subgroups defined by (A) age, (B) Gleason Grade Group, (C) alkaline phosphatase level, (D) Hb level, and (E) lactate dehydrogenase level. In each graph, the 24‐month event‐free rate is indicated by the black dotted lines and percentage value in blue (cohort 1) or red (cohort 2B). ADT, androgen deprivation therapy; aHR, adjusted hazard ratio; ALP, alkaline phosphatase; ARPI, androgen receptor pathway inhibitor; CAB, combined androgen blockade; CI, confidence interval; GGG, Gleason Grade Group; Hb, hemoglobin; LDH, actate dehydrogenase; mHNPC, metastatic hormone‐naïve prostate cancer; mTTCR, median time to castration‐resistant prostate cancer; NR, not reached. aTreated with ADT monotherapy or CAB. bTreated with ADT plus ARPI. [Color figure can be viewed at wileyonlinelibrary.com]
Figure 3.
Kaplan–Meier curves of overall survival in patients with high‐risk metastatic mHNPC in cohorts 1a and 2Bb across prognostic factor subgroups defined by (A) age, (B) Gleason Grade Group, (C) alkaline phosphatase level, (D) Hb level, and (E) lactate dehydrogenase level. In each graph, the 24‐month event‐free rate is indicated by the black dotted lines and percentage value in blue (cohort 1) or red (cohort 2B). ADT, androgen deprivation therapy; aHR, adjusted hazard ratio; ALP, alkaline phosphatase; ARPI, androgen receptor pathway inhibitor; CAB, combined androgen blockade; CI, confidence interval; GGG, Gleason Grade Group; Hb, hemoglobin; LDH, lactate dehydrogenase; mHNPC, metastatic hormone‐naïve prostate cancer; mOS, median overall survival. aTreated with ADT monotherapy or CAB. bTreated with ADT plus ARPI. [Color figure can be viewed at wileyonlinelibrary.com]
3.4. Safety
Safety data were similar across subgroups defined by age (Table 3) or other evaluated prognostic subgroups (Supporting Information: Tables S7–S10). AEs leading to discontinuation, dose reduction, or dose interruption were reported in a higher proportion of patients in cohort 2B versus cohort 1, regardless of the prognostic subgroup.
Table 3.
Summary of adverse events by study cohort, overall, and in subgroups defined by age.
| n (%) | Cohort 1a | Cohort 2Bb | ||||
|---|---|---|---|---|---|---|
| Age, years | Overall (n = 371) | < 75 years (n = 178) | ≥ 75 years (n = 193) | Overall (n = 576) | < 75 years (n = 373) | ≥ 75 years (n = 202) |
| ADRSIs | ||||||
| Any | 16 (4.3) | 6 (3.4) | 10 (5.2) | 110 (19.1) | 67 (18.0) | 43 (21.3) |
| Grade 3/4 | 3 (0.8) | 0 | 3 (1.6) | 19 (3.3) | 13 (3.5) | 6 (3.0) |
| Serious TRAEs | 0 | 0 | 0 | 19 (3.3) | 13 (3.5) | 6 (3.0) |
| AEs leading to death | 7 (1.9) | 2 (1.1) | 5 (2.6) | 9 (1.6) | 4 (1.1) | 5 (2.5) |
| AEs leading to discontinuation | 16 (4.3) | 7 (3.9) | 9 (4.7) | 82 (14.2) | 50 (13.4) | 32 (15.8) |
| AEs leading to dose reduction | 0 | 0 | 0 | 44 (7.6) | 32 (8.6) | 12 (5.9) |
| AEs leading to dose interruption | 2 (0.5) | 1 (0.6) | 1 (0.5) | 55 (9.5) | 33 (8.8) | 22 (10.9) |
Abbreviations: ADT, androgen deprivation therapy; AE, adverse event; ADRSIs, adverse drug reactions of special interest; ARPI, androgen receptor pathway inhibitor; CAB, combined androgen blockade; TRAE, treatment‐related adverse event.
Treated with ADT monotherapy or CAB.
Treated with ADT plus ARPI.
4. Discussion
In this post hoc subgroup analysis of 3‐year data from the J‐ROCK study, increased age, high GGG, elevated ALP, low Hb, and elevated LDH were poor prognostic factors in patients with LATITUDE criteria high‐risk mHNPC, regardless of treatment cohort. Although treatment response in patients with poor prognosis was reduced compared with that observed in those without prognostic factors, patients in cohort 2B (ADT plus ARPI) had improved outcomes compared with cohort 1 (ADT monotherapy or CAB), even in the presence of poor prognostic factors.
Our analysis included five known prognostic factors for mHNPC. In addition, Eastern Cooperative Oncology Group performance status (ECOG PS) [2, 17], the presence of liver metastases [2, 16, 21], and C‐reactive protein (CRP) levels have previously been reported as prognostic factors [20]. These factors were not included in our analysis: ECOG PS data were missing for most patients; the number of patients with liver metastases was very low; the presence of cancer‐related symptoms and albumin were not analyzed; and the collection of CRP data was not planned. Although the outcomes used to identify prognostic factors in previous studies varied, such as PFS [12, 13, 15, 17], PSA‐PFS [20], time to CRPC [3, 21], PFS2 [14], and OS [2, 16, 18, 20, 22, 25], the findings of our analysis suggested prognostic factors in mHNPC were relatively consistent, regardless of the clinical outcome.
Previous studies of PSA response by prognostic factors in patients with high‐risk mHNPC have not been reported. The proportion of patients who achieved a ≥ 90% decline in PSA within 3 months showed no difference among prognostic subgroups defined by age and GGG. In contrast, in cohort 2B, abnormal ALP, Hb, and LDH were associated with lower proportions of patients with PSA ≤ 0.2 ng/mL.
In terms of PFS, our interim analysis observed a relatively gradual downward gradient in the Kaplan–Meier curve among cohort 2B subgroups without poor prognostic factors, whereas these curves had steeper gradients in subgroups with poor prognostic factors in both cohorts 1 and 2B. Okamoto et al. previously evaluated PFS outcomes among patients with mHSPC with or without anemia who were treated with ADT monotherapy/CAB or upfront AAP [15]. The PFS Kaplan–Meier curve in ADT/CAB‐treated patients and AAP‐treated patients with anemia continued to decline until approximately 20 months of follow‐up [15], similar to that observed in the low Hb subgroup of our analysis. Takahara et al. also reported PFS outcomes in patients treated with AAP with or without low Hb, high GGG, and poor ECOG PS [17]. PFS Kaplan–Meier curves showed steady declining gradient in those with these prognostic factors [17]. These findings were also consistent with the trends observed in the current subgroup analysis.
When we examined the Kaplan–Meier curves of time to CRPC, the curves plateaued over time, regardless of treatment cohort or prognostic factors, a different trend to that observed with PFS, which showed a downward gradient. This suggests that a certain proportion of patients with high‐risk mHNPC may not progress to mCRPC, regardless of treatment or the presence of poor prognostic factors, and also indicates that hormone therapy is sensitive and durable in Japanese patients with high‐risk mHNPC. Our subgroup analysis findings are consistent with those of a previous retrospective study in older patients [23], which also reported a plateau in the time to CRPC Kaplan–Meier curve. Of note, this trend has not been previously reported for the other prognostic factors.
The OS Kaplan–Meier curves in our analysis showed trend towards an early decline among patients in cohort 1 with poor prognostic factors at around 6 months of follow‐up, whereas OS remained stable until around 12 months of follow‐up among patients in cohort 2B with poor prognostic factors. As a result, the OS Kaplan–Meier curves tended to spread between the two cohorts relatively early for patients with prognostic factors, and relatively late for those without prognostic factors. Miura et al. previously reported that the OS Kaplan–Meier curves for conventional therapy (ADT plus bicalutamide or ADT monotherapy) versus ADT plus upfront intensive therapy (AAP or docetaxel) tended to spread after around 12 months in older patients with high‐tumor burden metastatic castration‐sensitive prostate cancer (mCSPC) [23]. Likewise, Ueda et al. reported that the spread between the OS Kaplan–Meier curves between the bicalutamide and apalutamide arms in the high GGG subgroup occurred early in the observation period among patients with mHSPC [25].
Despite the differences in PFS and time to CRPC between cohorts 1 and 2B in our analysis for the low GGG subgroup, there were no differences in OS between cohorts for this subgroup. Consistent with our findings, Ueda et al. reported that the OS Kaplan–Meier curves for bicalutamide and apalutamide showed no tendency towards spreading in the low GGG subgroup, while a spread between OS curves occurred early in the observation period in the high GGG subgroup [25]. In the current subgroup analysis, the spread between cohorts of the PFS and time to CRPC Kaplan–Meier curves occurred relatively early, regardless of prognostic factors, whereas the spread between OS curves varied according to the presence of poor prognostic factors. Of note, in patients with poor prognostic factors, such as older age, high GGG, low Hb, and elevated LDH, the spread between the OS curves occurred within the first 6 months of treatment.
The safety data did not appear to be affected by the presence or absence of prognostic factors. This is the first report of safety data with conventional therapy or ADT plus ARPI for patients with high‐risk mHNPC by prognostic factor subgroup. Older patients tended to receive ADT monotherapy or CAB rather than ADT plus ARPI [10]. Although not evaluated, conventional therapy was most likely chosen to minimize the risk of toxicity in these patients. The incidence of safety signals was similar in the age subgroups for cohort 2B, suggesting ADT plus ARPI treatment may not be avoided because of patient age.
Previously, the only real‐world clinical data of triplet therapy in patients with mHNPC were from studies conducted in Austria [26] and Japan [7]. Both of these studies reported initial PSA response and safety data only; PFS, time to CRPC, and OS have not yet been reported. Therefore, further real‐world studies are needed to determine whether outcomes (especially OS) can be improved by triplet therapy in clinical practice among patients with poor prognostic factors (age, GGG, and LDH), who were relatively poor treatment responders in cohort 2B in this analysis.
The current exploratory post hoc analysis has several limitations. First, as the J‐ROCK study is a nonrandomized registry study, unknown sources of bias relating to the patient characteristics cannot be ruled out. Differences in patient baseline characteristics (e.g., performance status) between cohorts were not included in the Cox regression model, so interpretation of the comparative effectiveness and safety data should be made with caution. In addition, there were no data on triplet therapy because the registration period was before the approval of triplet therapy in Japan; therefore, current real‐world treatment practices may differ from those reported in this study. Assessment of time to next treatment or exposure‐normalized rates of AEs were also not included in this exploratory analysis. The prognostic factors chosen for our analysis were based on previous studies, and were not directly derived by analysis of the current dataset; therefore, it is unclear whether these factors were also independent prognostic factors in this dataset. It should also be noted that sensitivity analyses regarding the selection of cut‐off values for the prognostic markers were not conducted. Finally, as the J‐ROCK study was conducted only in Japan, Japanese real‐world practice and insurance reimbursement may have influenced our results, which may not be fully generalizable to other regions and countries.
5. Conclusions
In patients with high‐risk mHNPC, ADT plus ARPI therapy (cohort 2B) was associated with longer time to CRPC and OS compared with ADT monotherapy or CAB (cohort 1), even in the presence of poor prognostic factors. Patients in cohort 1 with poor prognostic factors (older age, high GGG, low Hb, and elevated LDH) showed a trend towards earlier decline in OS than those in cohort 2B, which may indicate that caution is required when cohort 1 treatment is selected for patients with poor prognostic factors. The time to CRPC curve also showed a trend towards a plateau, irrespective of prognostic factors or treatment arm. Based on these findings, novel hormonal regimens for Japanese patients with high‐risk mHNPC, such as ADT plus ARPI, may be associated with improved real‐world outcomes compared with ADT monotherapy or CAB, particularly among those with poor prognostic factors. Treatment selection should consider these outcomes in patients with various prognostic factors and safety, as well as the patient's condition, quality of life, economic burden, family environment, and residence.
Author Contributions
Study design (conceptualization, methodology): Atsushi Mizokami, Hideaki Miyake, Hiroji Uemura, Hirotsugu Uemura, Sachie Okazaki, Eri Adachi, Ryo Yano, Yohei Tajima, Kiyohide Fujimoto, and Hideyasu Matsuyama. Data access (data curation, software): Hiroaki Tsuchiya. Analysis (investigation, formal analysis): Atsushi Mizokami, Rikiya Matsumoto, Hideaki Miyake, Hiroji Uemura, Hirotsugu Uemura, Satoru Kawakami, Kazuyoshi Nakamura, Shigekatsu Maekawa, Hiroaki Tsuchiya, Kiyohide Fujimoto, and Hideyasu Matsuyama. Writing, reviewing, and decision to submit: All authors.
Ethics Statement
All procedures in the J‐ROCK study were performed in accordance with the ethical principles and related regulations outlined in the Declaration of Helsinki. The study protocol and its amendment were approved by the central ethics committee of each participating site. All participating patients were informed of the observational nature of the study and provided written consent for their data to be collected and analyzed before data collection was started. If approved by the independent ethics committees or institutional review boards, the informed consent requirement was waived for participants who had died.
Conflicts of Interest
Atsushi Mizokami has received speaking and lecture fees from Johnson & Johnson. Hideaki Miyake has received consulting or advisory fees from Astellas Pharma and Johnson & Johnson, and speaking and lecture fees from Johnson & Johnson. Hiroji Uemura has received funding grants from Takeda Pharmaceutical; speaking and lecture fees from Bayer, Johnson & Johnson, and Takeda Pharmaceutical; and travel reimbursement from Astellas Pharma, Bayer, Daiichi Sankyo, Johnson & Johnson, Kyowa Kirin, Sanofi, and Takeda Pharmaceutical. Hirotsugu Uemura has received funding grants from Astellas Pharma, AstraZeneca, Chugai, Johnson & Johnson, Kissei, MSD, Ono Pharmaceutical, Osaka Urology Research Foundation, Pfizer and Takeda Pharmaceutical; consulting or advisory fees from Bayer, BMS and Ono Pharmaceutical; and speaking and lecture fees from Bayer, BMS, Johnson & Johnson, MSD, Ono Pharmaceutical, Pfizer, Sanofi and Takeda Pharmaceutical. Kazuyoshi Nakamura has received funding grants from AstraZeneca, Eli Lilly & Company, Johnson & Johnson, and Tsumura; and speaking and lecture fees from Astellas Pharma, AstraZeneca, Bayer, Johnson & Johnson, and Takeda Pharmaceutical. Hiroaki Tsuchiya, Sachie Okazaki, Eri Adachi, and Yohei Tajima are an employee of Johnson & Johnson. Ryo Yano is a contractor for Johnson & Johnson and an employee of CMIC Inizio Co. Kiyohide Fujimoto has board membership and is receiving funding grants, and speaking and lecture fees from Johnson & Johnson. Hideyasu Matsuyama has received funding grants from Bayer and Johnson & Johnson; consulting or advisory fees from Johnson & Johnson; and speaking and lecture fees from Merck Biopharma Co. Ltd. and Johnson & Johnson. The remaining authors have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Supporting information
Mizokami_supporting information.
Acknowledgments
The authors would like to thank IQVIA Japan for assistance with study operations, patient monitoring, data management, and statistical analysis. We also thank Nicola Ryan, who wrote the outline of this manuscript on behalf of Springer Health+, and Sarah Greig, PhD, CMPP, and Simone Tait, CMPP, of Springer Health+, who wrote the first and subsequent drafts. Medical writing assistance was funded by Johnson & Johnson. This study and development of the manuscript were supported by Johnson & Johnson.
Data Availability Statement
The authors confirm that all data supporting the findings of this study are available within the article and/or its supporting materials.
References
- 1. Raval A. D., Chen S., Littleton N., Constantinovici N., and Goebell P. J., “Real‐World Use of Androgen‐Deprivation Therapy Intensification for Metastatic Hormone‐Sensitive Prostate Cancer: A Systematic Review,” BJU International 135, no. 3 (2025): 408–421. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Suzuki K., Hara T., Watanabe H., et al., “Rapid and Deep Prostate‐Specific Antigen Decline Is a Prognostic Marker in Metastatic Hormone‐Sensitive Prostate Cancer: A Real‐World Multi‐Intuitional Analysis,” Prostate 85, no. 5 (2025): 448–455. [DOI] [PubMed] [Google Scholar]
- 3. Urabe F., Muramoto K., Yanagisawa T., et al., “Changes in the Treatment Landscape of Metastatic Hormone‐Sensitive Prostate Cancer Following Approval of Upfront Androgen Receptor Signaling Inhibitors: A Multicenter Study,” International Journal of Urology 31, no. 11 (2024): 1248–1255. [DOI] [PubMed] [Google Scholar]
- 4. Smith M. R., Hussain M., Saad F., et al., “Darolutamide and Survival in Metastatic, Hormone‐Sensitive Prostate Cancer,” New England Journal of Medicine 386, no. 12 (2022): 1132–1142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Yamamoto Y., Fujimoto S., Hashimoto M., et al., “Real‐World Outcomes in Patients With Metastatic Castration‐Resistant Prostate Cancer Beyond Progression After Upfront Androgen Receptor Signaling Inhibitor,” International Journal of Clinical Oncology 29, no. 12 (2024): 1946–1958. [DOI] [PubMed] [Google Scholar]
- 6. Yamamoto Y., Nishimoto M., Akashi Y., et al., “Impact of Novel Agents on Patient Characteristics, Treatment Patterns, and Outcomes in Patients With Metastatic Castration‐Resistant Prostate Cancer,” Anticancer Research 44, no. 7 (2024): 3155–3161. [DOI] [PubMed] [Google Scholar]
- 7. Urabe F., Imai Y., Goto Y., et al., “Real‐World Evidence of Triplet Therapy Efficacy in Patients With Metastatic Castration‐Sensitive Prostate Cancer: A Japanese Multicenter Study,” Japanese Journal of Clinical Oncology 54, no. 11 (2024): 1208–1213. [DOI] [PubMed] [Google Scholar]
- 8. Chen W., Yoshida S., Miura N., et al., “Efficacy of Docetaxel Addition to Next‐Generation Androgen Receptor‐Axis‐Targeted Therapies and Androgen Deprivation Therapy in Metastatic Hormone‐Sensitive Prostate Cancer: A Tumor Volume‐Specific Analysis,” International Journal of Urology 32, no. 4 (2025): 361–370. [DOI] [PubMed] [Google Scholar]
- 9. Fizazi K., Tran N., Fein L., et al., “Abiraterone Plus Prednisone in Metastatic, Castration‐Sensitive Prostate Cancer,” New England Journal of Medicine 377, no. 4 (2017): 352–360. [DOI] [PubMed] [Google Scholar]
- 10. Miyake H., Matsumoto R., Fujimoto K., et al., “Clinical Outcomes of Patients With High‐Risk Metastatic Hormone‐Naïve Prostate Cancer: A 3‐year Interim Analysis of the Observational J‐ROCK Study,” European Urology Oncology 7, no. 3 (2024): 625–632. [DOI] [PubMed] [Google Scholar]
- 11. Uemura H., Matsumoto R., Mizokami A., et al., “Treatment Strategies and Outcomes in a Long‐Term Registry Study of Patients With High‐Risk Metastatic Hormone‐Naïve Prostate Cancer in Japan: An Interim Analysis of the J‐ROCK Study,” International Journal of Urology 29, no. 9 (2022): 1061–1070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Briones J., Khan M., Sidhu A. K., Zhang L., Smoragiewicz M., and Emmenegger U., “Population‐Based Study of Docetaxel or Abiraterone Effectiveness and Predictive Markers of Progression Free Survival in Metastatic Castration‐Sensitive Prostate Cancer,” Frontiers in Oncology 11 (2021): 658331. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Naiki T., Takahara K., Ito T., et al., “Comparison of Clinical Outcomes Between Androgen Deprivation Therapy With Up‐Front Abiraterone and Bicalutamide for Japanese Patients With LATITUDE High‐Risk Prostate Cancer in a Real‐World Retrospective Analysis,” International Journal of Clinical Oncology 27, no. 3 (2022): 592–601. [DOI] [PubMed] [Google Scholar]
- 14. Narita S., Kimura T., Hatakeyama S., et al., “Real‐World Outcomes and Risk Stratification in Patients With Metastatic Castration‐Sensitive Prostate Cancer Treated With Upfront Abiraterone Acetate and Docetaxel,” International Journal of Clinical Oncology 27, no. 9 (2022): 1477–1486. [DOI] [PubMed] [Google Scholar]
- 15. Okamoto T., Noro D., Hatakeyama S., et al., “Impact of Pretreatment Anemia on Upfront Abiraterone Acetate Therapy for Metastatic Hormone‐Sensitive Prostate Cancer: A Multicenter Retrospective Study,” BMC Cancer 21, no. 1 (2021): 605. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Suzuki K., Shiraishi Y., Furukawa J., et al., “Clinical Outcomes and Risk Stratification in Patients With Metastatic Hormone‐Sensitive Prostate Cancer Treated With New‐Generation Androgen Receptor Signaling Inhibitors,” Clinical Genitourinary Cancer 22, no. 5 (2024): 102140. [DOI] [PubMed] [Google Scholar]
- 17. Takahara K., Naiki T., Ito T., et al., “Useful Predictors of Progression‐Free Survival for Japanese Patients With LATITUDE‐High‐Risk Metastatic Castration‐Sensitive Prostate Cancer Who Received Upfront Abiraterone Acetate,” International Journal of Urology 29, no. 3 (2022): 229–234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Ueda T., Fujita K., Nishimoto M., et al., “Predictive Factors for the Efficacy of Abiraterone Acetate Therapy in High‐Risk Metastatic Hormone‐Sensitive Prostate Cancer Patients,” World Journal of Urology 40, no. 12 (2022): 2939–2946. [DOI] [PubMed] [Google Scholar]
- 19. Ueda T., Shiraishi T., Ito S., et al., “Abiraterone Acetate Versus Bicalutamide in Combination With Gonadotropin Releasing Hormone Antagonist Therapy for High Risk Metastatic Hormone Sensitive Prostate Cancer,” Scientific Reports 11, no. 1 (2021): 10094. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Watanabe H., Nakane K., Takahara K., et al., “Prognostic Outcomes in Japanese Patients With Metastatic Castration‐Sensitive Prostate Cancer: Comparative Assessments Between Conventional Androgen Deprivation Therapy (ADT) and ADT With Novel Androgen Receptor Signal Inhibitor,” International Journal of Urology 31, no. 9 (2024): 986–993. [DOI] [PubMed] [Google Scholar]
- 21. Yanagisawa T., Kimura T., Mori K., et al., “Abiraterone Acetate Versus Nonsteroidal Antiandrogen With Androgen Deprivation Therapy for High‐Risk Metastatic Hormone‐Sensitive Prostate Cancer,” Prostate 82, no. 1 (2022): 3–12. [DOI] [PubMed] [Google Scholar]
- 22. Yamada Y., Sato K., Sakamoto S., et al., “Characterization of PSA Dynamics and Oncological Outcomes in Patients With Metastatic Hormone‐Sensitive Prostate Cancer Treated With Androgen Receptor Signaling Inhibitors,” International Journal of Clinical Oncology 30, no. 3 (2025): 539–550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Miura Y., Hatakeyama S., Narita S., et al., “Effect of Upfront Intensive Therapy on Oncological Outcomes in Older Patients With High Tumor Burden Metastatic Castration‐Sensitive Prostate Cancer: A Multicenter Retrospective Study,” Prostate 82, no. 13 (2022): 1304–1312. [DOI] [PubMed] [Google Scholar]
- 24. Ogawa N. and Matsukura R., “Ageing in Japan: The Health and Wealth of Older Persons,” 2005, https://www.un.org/development/desa/pd/sites/www.un.org.development.desa.pd/files/unpd_egm_200508_09_ogawa.pdf.
- 25. Ueda T., Shiraishi T., Miyashita M., et al., “Apalutamide Versus Bicalutamide in Combination With Androgen Deprivation Therapy for Metastatic Hormone Sensitive Prostate Cancer,” Scientific Reports 14, no. 1 (2024): 705. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Kafka M., Giannini G., Artamonova N., et al., “Real‐World Evidence of Triplet Therapy in Metastatic Hormone‐Sensitive Prostate Cancer: An Austrian Multicenter Study,” Clinical Genitourinary Cancer 22, no. 2 (2024): 458–466.e1.E451. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Mizokami_supporting information.
Data Availability Statement
The authors confirm that all data supporting the findings of this study are available within the article and/or its supporting materials.