ABSTRACT
Despite promising evidence of the efficacy of the androgen deprivation therapy (ADT) plus apalutamide in metastatic castration‐sensitive prostate cancer (mCSPC), it remains unknown how its effectiveness and safety are shaped by genetic factors, particularly among Asian populations. We evaluated whether baseline circulating tumor deoxyribonucleic acid (ctDNA) and germline variants predict outcomes and tolerability in the phase IV, open label, multicenter CUARTET trial in Japan. We enrolled 101 patients with mCSPC starting apalutamide with ADT; 99 contributed pretreatment plasma for ctDNA profiling across 73 prostate cancer driver genes and 86 provided genome‐wide single nucleotide polymorphism (SNP) data. Primary endpoints were time to castration resistant prostate cancer (CRPC) and overall survival; skin rash incidence was assessed as a key adverse event. Cox models (adjusted for volume of disease) and exploratory machine‐learning–assisted logistic regression were applied. ctDNA presence at baseline was associated with shorter time to CRPC (adjusted hazard ratio [HR] 3.18; 95% confidence interval [CI], 1.17–11.09) and overall survival (adjusted HR for all‐cause mortality 3.95; 95% CI, 1.06–25.63). Among 55 paired samples, ctDNA presence increased from 61.8% pretreatment to 87.2% posttreatment, and alterations in androgen receptor and DNA repair pathways were enriched in patients who discontinued for progression. Any grade skin rash occurred in 59.4% and grade ≥ 3 in 5.9%. An exploratory genome‐wide screen identified 12 SNPs associated with skin rash. This study demonstrated that baseline ctDNA predicts prognosis of Japanese patients with mCSPC receiving apalutamide with ADT and identified potential genetic predictors for treatment‐related skin rash.
Trial Registration: ClinicalTrials.gov identifier: jRCTs071200040
Keywords: apalutamide, circulating tumor DNA, genomic alterations, metastatic castration‐sensitive prostate cancer, single nucleotide polymorphism, skin rash
In patients with mCSPC treated with apalutamide plus ADT, baseline ctDNA predicted earlier progression to CRPC and worse overall survival. Exploratory genome‐wide analysis identified 12 SNPs associated with apalutamide‐related skin rash.
Abbreviations
- ADR
Adverse drug reaction
- ADT
Androgen‐deprivation therapy
- AR
Androgen receptor
- cfDNA
Cell‐free DNA
- CI
Confidence interval
- CRPC
Castration‐resistant prostate cancer
- ctDNA
Circulating tumor DNA
- DNA
Deoxyribonucleic acid
- gDNA
genomic DNA
- HR
Hazard ratio
- HRR
Homologous recombination repair
- mCSPC
Metastatic castration‐sensitive prostate cancer
- OR
Odds ratio
- PD
Progressive disease
- PSA
Prostate‐specific antigen
- SNP
Single nucleotide polymorphisms
- TSG
Tumor suppressor gene
1. Introduction
Metastatic castration‐sensitive prostate cancer (mCSPC) is an advanced, incurable stage of the disease [1]. Although it is initially responsive to standard androgen‐deprivation therapy (ADT), the cancer almost always progresses to a castration‐resistant state (i.e., the terminal phase of the disease), often within a year for patients treated with ADT alone [2]. Apalutamide, an oral nonsteroidal antiandrogen, is a promising treatment for mCSPC when combined with ADT [2, 3]. A pivotal phase III trial, TITAN, demonstrated that the addition of apalutamide to ADT provides meaningful efficacy, improving overall survival and prognosis in a broad global patient population [4, 5]. The benefits of apalutamide were also confirmed as consistent in the Japanese subpopulation analysis of TITAN [6].
Despite its potential efficacy, clinical implementation of apalutamide treatment faces challenges primarily due to treatment discontinuation. For example, 49% (257/524) of the apalutamide plus ADT group in TITAN discontinued the study treatment, of which 53.7% (138/257) discontinued due to progressive disease (PD) and 24.1% (62/257) discontinued due to adverse drug reaction (ADR) over the 44 months of study period [5]. A specific and common ADR is skin rash, which was reported in 29.2% of apalutamide‐treated patients in the overall TITAN analysis and 53.6% in the Japanese subpopulation [5, 6]. Identifying reliable predictors for both PD and major ADRs like skin rash remains crucial for clinicians to select patients most likely to benefit from sustainable apalutamide treatment. Accumulating evidence suggests that genomic alterations and measures of tumor burden, such as circulating tumor deoxyribonucleic acid (ctDNA), are known prognostic indicators of poor outcomes and resistance to androgen receptor (AR)‐targeted therapies in advanced prostate cancer [7]. Investigating these biomarkers may help identify patients at higher risk of discontinuation, allowing for tailored and sustainable treatment strategies. However, it remains unknown how baseline genomic factors (e.g., ctDNA) predict PD and skin rash among patients with mCSPC treated with apalutamide.
To address these critical knowledge gaps, the CUARTET study examined the association between baseline ctDNA alterations and time to PD in Japanese patients with mCSPC receiving apalutamide, hypothesizing that a higher ctDNA burden is associated with a shorter time to PD in this patient group. The study also conducted an exploratory analysis to identify specific single nucleotide polymorphisms (SNPs) that may predict the occurrence of skin rash among patients with mCSPC on apalutamide.
2. Materials and Methods
2.1. Study Design
The CUARTET study was conducted as an open‐label, multicenter, single‐arm interventional Phase IV trial in Japan, specifically designed to explore ctDNA in patients with mCSPC receiving apalutamide treatment. The study was conducted in three phases: a screening phase, a treatment phase, and a follow‐up phase (Figure S1). The screening phase was from acquisition of informed consent to initiation of apalutamide treatment. The pretreatment blood collection for biomarkers took place during the screening phase. The treatment phase began once patients were initiated on apalutamide. Throughout the study, investigators may prescribe any concomitant therapies deemed necessary to provide adequate supportive care. Concomitant therapies may include current radiation therapy, surgery, and bone‐sparing agents. Patients were treated until PD, unacceptable toxicity, or other reasons that suggested a treatment discontinuation. The follow‐up phase started from apalutamide treatment discontinuation until death, lost to follow‐up, consent withdrawal for study participation, or end of the follow‐up phase, whichever occurred first. Patients were followed for safety for 30 days after the last dose of apalutamide (in case a patient discontinued apalutamide) or until the final data cutoff date (in case patient continued to take apalutamide or the final data cutoff was earlier than 30 days since the discontinuation of apalutamide). The posttreatment blood collection for biomarkers was performed within 28 days before the initiation of any subsequent therapy for prostate cancer after apalutamide discontinuation. The blood collection for biomarkers was also performed at the end of follow‐up if patients continued apalutamide treatment after PD or at the end of follow‐up. Total study duration was 4.5 years from the date of first informed consent. Enrollment occurred over up to 1.5 years (Figure 1).
FIGURE 1.
Graphical abstract of the CUARTET study. In patients with mCSPC treated with apalutamide plus ADT, baseline ctDNA predicted earlier progression to CRPC and worse overall survival. Exploratory genome‐wide analysis identified 12 SNPs associated with apalutamide‐related skin rash.
The CUARTET study included Japanese men aged ≥ 20 years with a documented diagnosis of metastatic prostate cancer (histologically or cytologically confirmed adenocarcinoma) that was castration naïve or castration sensitive. For patients with de novo mCSPC, prior systemic ADT or combined androgen blockade before registration was permitted if the total duration was < 6 months. ADT included luteinizing hormone‐releasing hormone agonist, luteinizing hormone‐releasing hormone antagonist, or bilateral orchiectomy. Combined androgen blockade was defined as ADT in combination with first generation nonsteroidal AR antagonists, such as bicalutamide. For patients with recurrent metastatic disease, prior neoadjuvant and/or adjuvant hormonal therapy, defined as ADT or combined androgen blockade administered before and/or after definitive local treatment (radical prostatectomy or radiotherapy), was permitted, with a maximum cumulative duration of 36 months. Patients were also excluded if they had received any prior therapy with abiraterone acetate, docetaxel, enzalutamide, apalutamide, or darolutamide, or had known allergies or contraindications to apalutamide or ADT. Of the 102 eligible patents, 101 patients signed written informed consent for the safety outcomes. Of these, 99 patients agreed to the assessment of ctDNA (including 73 prostate cancer driver genes) and 86 agreed to the assessment of SNPs (Figure 2).
FIGURE 2.
Patient flow diagram. Flowchart of study enrollment and analytic population definitions. Of 102 eligible patients, 101 consented and were included in the safety analysis set. After exclusions for missing ctDNA or SNP measurements, 99 patients were included in the effectiveness analysis and 86 in the skin‐rash analysis. SNPs, single nucleotide polymorphisms.
The study protocol was reviewed and approved by the Clinical Research Network Fukuoka Certified Review Board (certification ID: CRB7180004) on September 2, 2020. The study was registered with the Japan Registry of Clinical Trials (jRCT registration ID: jRCTs071200040) on October 9, 2020. The study was conducted in accordance with the principles of the Declaration of Helsinki and applicable local regulations. Written informed consent was obtained from all patients prior to enrollment and before any study‐related procedures were performed.
2.2. Measurements
Genetic Predictors: We measured ctDNA for effectiveness analyses and SNPs for skin rash analyses, using 20 mL peripheral blood samples. ctDNA was measured at pre‐treatment (screening phase) and at one of three predefined post‐treatment time points: (i) within 28 days before initiation of subsequent therapy in patients who had PD and switched treatment; (ii) treatment change in patients who discontinued apalutamide due to any reasons other than PD; or (iii) at the end of follow‐up in patients without PD. Genomic alterations were profiled in plasma ctDNA using a targeted panel of 73 prostate cancer driver genes, and ctDNA presence (genomic alteration positivity) was defined as detection of ≥ 1 somatic alteration in this panel (Doc. S1) [8]. Plasma‐derived cell‐free DNA (cfDNA) was extracted using the QIAamp Circulating Nucleic Acid Kit (50) on a QIAvac24 Plus vacuum manifold. For matched germline controls, genomic DNA (gDNA) was extracted from buffy coat using the Maxwell RSC Buffy Coat DNA Kit on the Maxwell RSC automated nucleic acid extraction instrument. For next‐generation sequencing‐based mutation profiling, libraries were prepared from plasma cfDNA and buffy‐coat gDNA using SureSelectXT HS Reagents together with a custom panel targeting the predefined set of 73 prostate cancer driver genes. After library amplification, target regions were enriched and sequenced on a NextSeq platform. Sequencing data were processed and variants called using DRAGEN Software [9]. Germline variants were called from the buffy‐coat gDNA. Matched tumor–normal analysis was then performed, designating cfDNA as the tumor and buffy‐coat gDNA as the normal, to produce a somatic variant list. We created a binary indicator for the presence of genomic alteration in any of these 73 prostate cancer driver genes. SNPs were evaluated at screening phase using Infinium Asian Screening Array from Illumina K.K and the quality check was performed by Riken Genesis Co. Ltd. If SNP genotyping could not be completed at screening phase, SNPs were genotyped from the blood sample obtained for the second ctDNA analysis (i.e., the post‐baseline ctDNA draw).
Outcomes: Our outcomes included disease progression metrics: overall survival (time to all‐cause mortality), time to a CRPC state, and treatment continuation status (on treatment, discontinuation due to PD, or discontinuation without PD). CRPC was defined according to European Association of Urology guidelines 2019. Prostate‐specific antigen (PSA) progression was assessed using two nadir PSA thresholds: ≤ 0.2 ng/mL, in accordance with the Prostate Cancer Clinical Trials Working Group 3 criteria, and ≤ 0.02 ng/mL, based on recent emerging evidence linking ultralow PSA levels with clinical outcomes [10, 11, 12]. We also assessed the incidence of skin rash outcomes of all grades or Grade ≥ 3 during the follow‐up period.
Baseline Demographics and Clinical Parameters: Baseline patient characteristics were assessed as covariates. These covariates included age, Gleason score at diagnosis, Eastern Cooperative Oncology Group score, metastatic volume defined by CHAARTED criteria [13], site of metastasis (bone, visceral, distant lymph node), previous radiation therapy, previous surgery, and previous hormone therapy (types included bilateral orchiectomy, luteinizing hormone‐releasing hormone agonist/antagonist, or bicalutamide).
2.3. Statistical Analysis
Three types of population were assessed for the analyses based on the availability of data: the safety analysis set, the effective analysis set, and the skin rash analysis set (Figure 1). We compared baseline characteristics of the safety analysis set stratified by the existence of any baseline ctDNA genomic alteration. The incidence of ADRs, without ctDNA assessment, was also evaluated and reported in this population.
We investigated the associations between ctDNA presence and outcomes in patients who had ctDNA measurements, defined as the effectiveness analysis set. Kaplan–Meier curves were estimated to assess the association between ctDNA presence at baseline and time to CRPC and all‐cause mortality. Censoring was defined as the assignment of a last observable date for patients who did not experience the event of interest by the time of analysis. Specific censoring dates varied by endpoint but included: the last assessment date showing no progression for patients without events, the date of last tumor assessment for those lost to follow‐up or withdrawing, the study treatment start date when no assessments existed, or the last assessment before subsequent therapy initiation. We estimated hazard ratios (HRs) for the association between ctDNA presence at baseline and time to CRPC and all‐cause mortality using univariate and multivariable Cox proportional hazards models. The 95% confidence intervals (CIs) were estimated using the profile likelihood method. The multivariable Cox proportional hazards model analyses were adjusted for metastatic volume. As a supplementary analysis, we also conducted multivariable logistic regression for the association between ctDNA presence at baseline and nadir PSA ≤ 0.2 ng/mL at the end of the follow‐up as a measure of effectiveness of the apalutamide treatment, adjusting for metastatic volume and log baseline PSA. Missing baseline covariate data were not imputed. Multivariable models were restricted to covariates with complete data; variables with missing values were summarized descriptively but were not included as covariates in the Cox proportional hazards or logistic regression models. Changes in genetic alterations in each of the individual prostate cancer driver genes pre‐ vs. post‐apalutamide treatment in relation to patients' treatment discontinuation status were compared in patients who had the second ctDNA measurements at appropriate timing defined above.
Finally, of the total of 404,508 SNPs assessed in the blood sample data, we identified potentially important SNPs that predicted skin rash of any grade by (1) selecting 100 SNPs with the largest absolute values of log odds ratio for the association with skin rash via genome‐wide analysis with logistic regression of each SNP on skin rash, adjusting for age, hormone therapy types, Gleason score at diagnosis, serum albumin, and bone metastasis, and (2) independently selecting top 100 SNPs with the best predictive performance based on the area under the curve metric based on two machine learning algorithms (point‐wise linear model called B3 and elastic‐net) via 5‐fold cross‐validation [14]. We determined 12 SNPs that were included in both selection criteria as important genetic predictors of skin rash. We then fit logistic regression models for each of these 12 SNPs, adjusting for age, types of castration, Gleason score at diagnosis, serum albumin, and bone metastasis, and estimated odds ratios (ORs) for the association between the selected SNPs and skin rash incidence. These analyses were conducted in SAS version 9.4 (SAS Institute, Cary, NC, USA), PLINK version 1.9 (Broad Institute, Cambridge, MA, USA), and Python version 3.7.4 (Python Software Foundation, Wilmington, DE, USA).
3. Results
Baseline demographic and clinical characteristics for the entire analysis set and the effectiveness analysis set, stratified by baseline ctDNA status based on 73 prostate cancer driver genes, were presented (Table 1). Median follow‐up was 31.4 months (range, 0.0–48.8) in the safety analysis set; it was 39.0 months (range, 0.0–48.8) for patients with baseline ctDNA and 42.8 months (range, 9.5–48.8) for those with complete ctDNA at study end. Of the 99 patients for effectiveness analysis, 60 (60.6%) were ctDNA‐present at baseline while 39 (39.4%) were ctDNA‐absent. The groups were comparable in median age (73.0 years in the ctDNA‐present group vs. 71.0 years in the ctDNA‐absent group) and Gleason score at diagnosis; however, patients in the ctDNA‐present group had higher baseline PSA levels (median 270.5 vs. 217.5 ng/mL). Patients in the ctDNA‐present group also had higher rates of aggressive metastatic disease, with diffuse bone metastases present in 53.7% compared with 19.4% of patients in the ctDNA‐absent group. Similarly, visceral metastases were more frequent in patients in the ctDNA‐present group (27.4% vs. 20.5%).
TABLE 1.
Demographics and baseline characteristics safety analysis set.
| Safety analysis set | Effectiveness analysis set | |||
|---|---|---|---|---|
| Overall | ctDNA‐absent | ctDNA‐present | ||
| Category | N = 101 | N = 99 | N = 39 | N = 60 |
| Agea (years) | ||||
| Median (range) | 72 (52, 85) | 72 (52, 85) | 71 (60, 84) | 73 (52, 85) |
| Gleason score at diagnosis, n (%) | ||||
| ≤ 7 | 2 (2.0) | 2 (2.0) | 2 (5.1) | 0 (0.0) |
| 8 | 30 (29.7) | 30 (30.3) | 11 (28.2) | 19 (30.6) |
| 9 | 58 (57.4) | 56 (56.6) | 20 (51.3) | 38 (61.3) |
| 10 | 11 (10.9) | 11 (11.1) | 6 (15.4) | 5 (8.1) |
| ECOG PS, n (%) | ||||
| 0 | 71 (81.6) | 69 (81.2) | 27 (84.4) | 44 (80.0) |
| 1 | 13 (14.9) | 13 (15.3) | 3 (9.4) | 10 (18.2) |
| 2 | 3 (3.4) | 3 (3.5) | 2 (6.3) | 1 (1.8) |
| Missing | 14 | 14 | 7 | 7 |
| PSA value at diagnosis (ng/mL) | ||||
| n | 100 | 98 | 38 | 62 |
| Missing | 1 | 1 | 1 | 0 |
| Median (range) | 266.5 (5.2, 14305.6) | 270.5 (5.2, 14305.6) | 217.5 (5.7, 12128.4) | 270.5 (5.2, 14305.6) |
| Site of metastasis, n (%) | 83 (83.8) | |||
| Bone metastasis | 85 (84.2) | 83 (83.8) | 31 (79.5) | 54 (87.1) |
| 1–3 | 29 (34.1) | 28 (33.7) | 15 (48.4) | 14 (25.9) |
| 4–10 | 14 (16.5) | 13 (15.7) | 8 (25.8) | 6 (11.1) |
| 11–20 | 2 (2.4) | 2 (2.4) | 1 (3.2) | 1 (1.9) |
| 21 < | 5 (5.9) | 5 (6.0) | 1 (3.2) | 4 (7.4) |
| Diffuse bone metastasis | 35 (41.2) | 35 (42.2) | 6 (19.4) | 29 (53.7) |
| Missing | 0 | 25 (25.3) | 0 | 0 |
| 22 (88.0) | ||||
| Visceral metastasisc | 25 (24.8) | 2 (8.0) | 8 (20.5) | 17 (27.4) |
| Lung | 22 (88.0) | 3 (12.0) | 6 (75.0) | 16 (94.1) |
| Liver | 2 (8.0) | 38 (38.4) | 0 (0.0) | 2 (11.8) |
| Other | 3 (12.0) | 5 (5.1) | 2 (25.0) | 1 (5.9) |
| Distant lymph node metastasis | 38 (37.6) | 9 (9.1) | 14 (35.9) | 24 (38.7) |
| Other | 5 (5.0) | 90 (90.9) | 0 (0.0) | 5 (8.1) |
Note: Demographics and baseline characteristics for the safety and effectiveness analysis sets are shown. The distributions of variables in the effectiveness analysis set, stratified by baseline ctDNA status, are also reported. The presence of ctDNA was defined as detection of ≥ 1 somatic alteration in 73 prostate cancer driver genes. Categorial variables are expressed as numbers (percentage), and continuous variables are presented as medians (ranges).
Abbreviations: ECOG PS, Eastern Cooperative Oncology Group performance status; PSA, prostate‐specific antigen.
Kaplan–Meier curves showed the association of baseline ctDNA evidence with (A) time to CRPC and (B) overall survival during the follow‐up period (Figure 3). Patients with baseline ctDNA had faster progression to both CRPC and all‐cause mortality compared with those without ctDNA. In the multivariable analyses adjusting for metastatic volume (Table 2), the presence of ctDNA at baseline was significantly associated with increased hazard of disease progression (adjusted HR = 3.18; 95% CI: 1.17–11.09) and all‐cause mortality (adjusted HR = 3.95; 95% CI: 1.06–25.63). In contrast, our multivariable logistic regression found no evidence of nadir PSA ≤ 0.2 ng/mL achievement by ctDNA among those with mCSPC on apalutamide (adjusted OR = 1.93; 95% CI: 0.74–5.25).
FIGURE 3.
Baseline ctDNA and time to outcomes. Kaplan–Meier curves depicting the association of baseline ctDNA status based on 73 prostate cancer driver genes (present vs. absent at pretreatment baseline) with (A) time to CRPC onset and (B) overall survival. Time zero is study baseline (apalutamide initiation). CRPC, castration‐resistant prostate cancer.
TABLE 2.
Association of baseline ctDNA with progression to CRPC and overall survival.
| Factor | Point estimate | 95% CI | ||
|---|---|---|---|---|
| Lower | Upper | |||
| Adjusted hazard ratio estimates for time to CRPC | ||||
| ctDNA‐present at baseline | Presence vs. absence | 3.18 | 1.17 | 11.09 |
| Metastatic volume | High volume vs. low volume | 1.84 | 0.79 | 4.64 |
| Adjusted hazard ratio estimates for overall survival | ||||
| ctDNA‐present at baseline | Presence vs. absence | 3.95 | 1.06 | 25.63 |
| Metastatic volume | High volume vs. low volume | 1.81 | 0.63 | 5.90 |
| Adjusted odds ratio estimates for Nadir PSA ≤ 0.2 ng/mL achievement rate | ||||
| ctDNA‐present at baseline | Presence vs. absence | 1.93 | 0.74 | 5.25 |
| Metastatic volume | High volume vs. low volume | 0.58 | 0.22 | 1.49 |
| log(PSA) (SD) | 1SD | 0.51 | 0.30 | 0.84 |
Note: First and second rows show the results of Cox proportional hazards models estimating HRs and 95% confidence intervals for the association between baseline ctDNA status based on 73 prostate cancer driver genes (presence vs. absence) and time to CRPC and overall survival. Multivariable models adjust for metastatic volume. The third row shows the results of the logistic regression model for nadir PSA ≤ 0.2 ng/mL at the study end. The multivariable model adjusted for metastatic volume and log baseline PSA.
Abbreviations: CI, confidence interval; CRPC, castration‐resistant prostate cancer; HR, hazard ratio; mCSPC, metastatic castration‐sensitive prostate cancer; OS, overall survival; SD, standard deviation.
The prevalences of ctDNA and genomic alterations of selected prostate cancer driver genes before and after apalutamide treatment were presented, stratified by treatment discontinuation status at study end (Table 3). Among 55 patients with matched pre‐ and post‐treatment blood samples, the overall prevalence of detectable ctDNA increased from 61.8% at baseline to 87.2% after treatment. Post‐treatment prevalence was comparable across the treatment discontinuation groups: 84.6% in discontinuation for PD, 85.7% in discontinuation for reasons other than PD, and 89.3% for the on‐treatment group. When assessing specific driver genes, certain alterations were more frequent among patients who discontinued treatment due to PD. For example, AR alterations were observed in 27.3% of patients in the PD discontinuation group, compared with 0%–4% in the other groups. Tumor suppressor gene (TSG) alterations were also more common in the PD discontinuation group (RB1, 18.2%; TP53, 27.3%) than in the other two groups (0%–8%). Similarly, alterations in DNA repair–related genes (e.g., FANC/A/C/D2/E/F/G) were identified in 45.5% of the PD discontinuation group, compared with 0.0% of the non‐PD discontinuation group and 20.0% of the on‐treatment group. The overall distribution of driver gene alterations is shown as a heatmap (Figure S2). Patients who progressed on apalutamide had more frequent post‐treatment genomic alterations compared with those who did not progress. The proportion of patients who achieved a nadir PSA ≤ 0.02 ng/mL was higher in the on‐treatment group (71.4%) than in the group that discontinued due to PD (15.4%) and the group that discontinued for reasons other than PD (50.0%) (Table S1).
TABLE 3.
Genomic alterations in prostate cancer driver genes at pre‐ and post‐treatment timepoints of apalutamide.
| Baseline measurements | Study end | |||
|---|---|---|---|---|
| (n = 55) | Discontinuation for PD (n = 13) | Discontinuation for reasons other than PD (n = 14) | On treatment (n = 28) | |
| ctDNA evidence | 34 (61.8%) | 11 (84.6%) | 12 (85.7%) | 25 (89.3%) |
| Prostate cancer driver gene | ||||
| AR | 2/34 (5.9%) | 3/11 (27.3%) | 0/12 (0.0%) | 1/25 (4.0%) |
| ERG | 1/34 (2.9%) | 0/11 (0.0%) | 0/12 (0.0%) | 1/25 (4.0%) |
| NKX3‐1 | 2/34 (5.9%) | 0/11 (0.0%) | 2/12 (16.7%) | 4/25 (16.0%) |
| TMPRSS2 | 3/34 (8.8%) | 1/11 (9.1%) | 0/12 (0.0%) | 1/25 (4.0%) |
| AKT1/2/3 | 1/34 (2.9%) | 2/11 (18.2%) | 0/12 (0.0%) | 6/25 (24.0%) |
| PIK3CA/CB/R1 | 4/34 (11.8%) | 2/11 (18.2%) | 0/12 (0.0%) | 2/25 (8.0%) |
| PTEN | 5/34 (14.7%) | 1/11 (9.1%) | 0/12 (0.0%) | 1/25 (4.0%) |
| ATM | 4/34 (11.8%) | 1/11 (9.1%) | 0/12 (0.0%) | 1/25 (4.0%) |
| BRCA1 | 1/34 (2.9%) | 2/11 (18.2%) | 0/12 (0.0%) | 0/25 (0.0%) |
| BRCA2 | 2/34 (5.9%) | 2/11 (18.2%) | 0/12 (0.0%) | 0/25 (0.0%) |
| CDK12 | 2/34 (5.9%) | 1/11 (9.1%) | 0/12 (0.0%) | 0/25 (0.0%) |
| ERCC1‐5 | 9/34 (26.5%) | 3/11 (27.3%) | 1/12 (8.3%) | 2/25 (8.0%) |
| FANC/A/C/D2/E/F/G | 14/34 (41.1%) | 5/11 (45.5%) | 0/12 (0.0%) | 5/25 (20.0%) |
| MSH2/6 | 6/34 (17.6%) | 2/11 (18.2%) | 0/12 (0.0%) | 0/25 (0.0%) |
| RAD51B/C | 2/34 (5.9%) | 3/11 (27.3%) | 0/12 (0.0%) | 1/25 (4.0%) |
| MYC | 6/34 (17.6%) | 2/11 (18.2%) | 0/12 (0.0%) | 1/25 (4.0%) |
| RB1 | 3/34 (8.8%) | 2/11 (18.2%) | 0/12 (0.0%) | 2/25 (8.0%) |
| TP53 | 7/34 (20.6%) | 3/11 (27.3%) | 0/12 (0.0%) | 1/25 (4.0%) |
| SPOP | 1/34 (2.9%) | 2/11 (18.2%) | 0/12 (0.0%) | 0/25 (0.0%) |
| APC | 4/34 (11.8%) | 2/11 (18.2%) | 0/12 (0.0%) | 1/25 (4.0%) |
| CHD1 | 4/34 (11.8%) | 2/11 (18.2%) | 0/12 (0.0%) | 1/25 (4.0%) |
| MET | 3/34 (8.8%) | 1/11 (9.1%) | 0/12 (0.0%) | 1/25 (4.0%) |
Note: Prevalence of detectable ctDNA and alterations in selected prostate cancer driver genes before and after apalutamide, stratified by treatment status at study end (discontinuation due to PD, discontinuation for reasons other than PD, or remaining on treatment). Analyses include patients with matched pre‐ and post‐treatment blood samples (n = 55).
Abbreviation: PD, progressive disease.
The effectiveness outcome (achievement of nadir PSA ≤ 0.2 ng/mL study and ≤ 0.02 ng/mL at any time during follow‐up) and progression outcomes (proportions of patients free from progression and overall survival at 12, 24, 36, and 48 months from baseline) were summarized according to baseline ctDNA status (ctDNA‐present vs. ctDNA‐absent; Figure S3). While the proportion achieving a nadir PSA ≤ 0.2 ng/mL was similar between groups, progression outcomes differed markedly. Specifically, patients who were ctDNA‐present at baseline showed a lower proportion achieving a nadir PSA ≤ 0.02 ng/mL, experienced more rapid disease progression, and had higher cumulative rates of CRPC and all‐cause mortality over time, compared with those without ctDNA.
Of the 101 patients in the safety analysis set, 77 (76.2%) experienced ADRs (Table S2). Skin rash of any grade occurred in 60 (59.4%) patients after treatment with apalutamide, although grade ≥ 3 skin rash occurred in only 6 patients (5.9%). A total of 28 (27.7%) patients managed skin rash by dose reduction, while 17 (16.8%) patients experienced permanent treatment discontinuation (Table S3). The distribution of subsequent therapies among those who discontinued apalutamide during follow‐up was summarized; 40 (39.6%) patients received any subsequent therapy (Table S4).
The machine learning analysis identified 12 SNPs associated with all grade skin rash. Table 4 shows the OR estimates for the association between each of these SNPs and skin rash. Eight SNPs were associated with lower odds of developing skin rash, while four SNPs showed increased odds of skin rash.
TABLE 4.
Clinical importance of identified SNPs; All grade skin rash.
| SNP ID | Adjusted OR (95% CI) | Related gene*1 | SNP location*1 |
|---|---|---|---|
| rs2236648 | 0.017 (0.001–0.209) | STIP1 |
Intron (Enhancer) |
| rs1011219 | 0.015 (0.001–0.197) | STIP1 |
Intron (Promoter of FERMT3) |
| kgp6952983 | 0.008 (0.000–0.351) | Unknown | — |
| rs79950190 | 110.924 (3.195–3850.540) | ESRRB | Intron |
| rs7583604 | 0.015 (0.001–0.319) | Unknown | Intergenic variant |
| rs59355030 | 0.018 (0.001–0.321) | FARS2 | Intron |
| kgp5674490 | 0.021 (0.001–0.312) | CETP | Intron |
| rs4901444 | 50.125 (2.996–838.726) | lncRNA | Intron |
| rs149252025 | 53.392 (2.808–1015.057) | Intergenic region | Intergenic region |
| rs11071459 | 92.777 (4.939–1742.892) | GCNT3 (glucosaminyl [N‐acetyl] transferase 3, mucin type) | Intron |
| rs174489 | 0.019 (0.001–0.432) | CNTN6 | Intron |
| kgp22175839 | 0.014 (0.001–0.219) | CAMK4 | Intron |
Note: ORs and 95% CIs for SNPs identified by machine‐learning screening as associated with skin rash of any grade during apalutamide treatment (12 SNPs total). Positive associations (OR > 1) indicate increased odds of skin rash, and negative associations (OR < 1) indicate decreased odds.
Abbreviations: CI, confidence interval; OR, odds ratio; SNP, single‐nucleotide polymorphism.
4. Discussion
This open‐label, multicenter, single‐arm interventional phase IV trial in Japan examined genetic predictors of prognosis and skin rash among individuals with mCSPC receiving apalutamide. The key findings were fourfold. First, the efficacy and safety profile of apalutamide observed in the phase III trial were confirmed in the present real‐world data study. Second, the presence of baseline ctDNA was strongly associated with higher rates of disease progression (CRPC onset) and all‐cause mortality. Third, the relationship between genomic alterations and treatment discontinuation due to progression varied across the 73 prostate cancer driver genes analyzed, with alterations in the AR pathway, TSG, and DNA repair genes being particularly associated with PD. Fourth, we observed an increase in the prevalence of genomic alterations between baseline and post‐treatment, especially among patients who experienced PD. Additionally, 12 SNPs were identified as being associated with the risk of developing skin rash.
In this real‐world cohort, effectiveness and safety outcomes appeared broadly consistent with those reported in the phase III TITAN trial, particularly in the Japanese subgroup, although cross‐trial comparisons should be interpreted with caution. In the TITAN Japanese subgroup, more than 80% of patients in the intervention arm were alive and approximately 70% were free from progression at 48 months of follow‐up [6]. These results were replicable in our overall cohort; 82.2% were alive and 67.9% were progression‐free at 48 months after study initiation. Although patients with baseline ctDNA tended to have lower overall survival and shorter time to CRPC than those without ctDNA, more than 70% were alive and over 50% remained progression‐free at 48 months, suggesting generally acceptable effectiveness across subgroups. Safety findings were also comparable: skin rash occurred in 53.6% of patients in the phase III trial and in 59.4% of our cohort. Although all‐grade skin rash was common, grade ≥ 3 events occurred in 5.9% of patients, suggesting that most reactions were manageable. Taken together, these observations indicate that outcomes with apalutamide in routine clinical practice may approximate those seen in the phase III setting, while recognizing the limitations of our non‐randomized design and potential differences in patient characteristics and follow‐up.
In the current study, baseline ctDNA detectability was associated with higher rates of disease progression and all‐cause mortality. Because ctDNA detectability is closely linked to tumor burden and aggressive metastatic patterns, this association may partly reflect underlying disease severity rather than a fully independent biological prognostic effect. Although we adjusted for metastatic volume, the limited number of overall survival events precluded more extensive multivariable adjustment. Therefore, while baseline ctDNA may provide additional prognostic information beyond conventional clinical factors, its incremental prognostic value could not be definitively established in this cohort and should be validated in larger studies with sufficient events and comprehensive covariate adjustment.
Baseline ctDNA status was not associated with the proportion of patients who achieved a nadir PSA ≤ 0.2 ng/mL in our study; however, patients who were ctDNA‐absent at baseline had a higher rate of achieving a nadir PSA ≤ 0.02 ng/mL than those who were ctDNA‐present at baseline (Figure S3). Patients who achieve a nadir PSA ≤ 0.02 ng/mL (ultra‐low PSA) are reported to have the most favorable long‐term outcomes compared with those whose nadir PSA is > 0.02 to ≤ 0.2 ng/mL or > 0.2 ng/mL, in both clinical trials and real‐world settings. In addition, patients who remained progression‐free through study end more frequently achieved a nadir PSA ≤ 0.02 ng/mL and had a lower frequency of genomic alterations in specific driver genes. Taken together, these findings suggest that baseline ctDNA status may serve as a sensitive biomarker for identifying patients most likely to achieve an ultra‐low PSA and derive the greatest benefit from apalutamide in the mCSPC setting.
Genomic alterations in specific pathways showed distinct associations with treatment outcomes. For example, previous studies have linked ctDNA to resistance to AR‐targeted therapies in advanced prostate cancer [7, 15, 16, 17, 18, 19, 20]. AR pathway inhibition exerts strong selective pressure for AR‐driven resistance. In a phase III randomized clinical trial, patients randomized to early treatment intensification (ADT with enzalutamide and abiraterone acetate) had a higher prevalence of plasma AR alterations at first progression than those on ADT alone (48% vs. 25%), and the presence of AR alterations at progression was associated with inferior overall survival, supporting the clinical relevance of the AR signal in the PD subgroup [15]. We also observed that TSG alterations were enriched among patients who discontinued due to PD (RB1 18.2%; TP53 27.3%), a pattern concordant with emerging data that defects in RB1, TP53, and PTEN mark an aggressive‐variant biology, associate with neuroendocrine differentiation, and portend shorter time to CRPC and worse survival in mCSPC—particularly when two TSGs are concurrently low or altered [16]. Although our dataset did not show a significant association for PTEN, a STAMPEDE transcriptomic analysis reported that a PTEN‐inactivity classifier identified tumors with poorer outcomes on ADT or ADT plus abiraterone acetate but a relative overall‐survival benefit from adding docetaxel [17]. Consistent with this, a recent real‐world cohort integrating “aggressive‐variant” TSG status (≥ 1 alteration in TP53/RB1/PTEN–PI3K/AKT) with clinical variables reported significantly shorter progression‐free survival and overall survival for AVPC‐TSG–altered mCSPC and little or no progression‐free survival gain from AR pathway inhibitor intensification, whereas AVPC‐TSG–wild‐type tumors derived clear benefit [18]. Similarly, alterations in homologous‐recombination repair (HRR) genes, particularly BRCA1/2 but also non‐BRCA HRR genes such as FANCA, have been associated with worse outcomes in mCSPC irrespective of treatment with AR pathway inhibitors or docetaxel [19]. In the CAPTURE/ACHILLES cohort, HRR alterations were observed in 28.6% of patients and were linked to significantly shorter rPFS and overall survival across both high‐ and low‐volume disease subgroups, mirroring the more adverse alteration profile we see in patients stopping for PD [20]. These findings suggest that additional therapeutic strategies may be required for patients who harbor or develop specific genomic alterations.
Our machine learning genome‐wide analysis identified 12 SNPs associated with the risk of skin rash development, which may provide insights into apalutamide‐related cutaneous reactions. For instance, the T allele of rs2236648 in STIP1 was significantly associated with a lower risk of rash. Given that STIP1 encodes a protein involved in cellular stress responses and inflammatory pathways—processes relevant to hypersensitivity reactions—this association may offer mechanistic clues for apalutamide‐induced skin rash [21]. Further studies are warranted to validate these findings and to clarify the genomic determinants of rash in patients receiving apalutamide.
This study has several limitations. First, some analyses had small sample sizes, limiting precision and precluding definitive conclusions about the prognostic value of baseline ctDNA. Although we observed consistent associations between ctDNA and both overall survival and time to CRPC in univariable and multivariable analyses, variables that were not included in the Cox proportional hazard models potentially alter these associations. Second, although the multivariable Cox models were adjusted for metastatic volume, residual confounding may remain. We considered including additional baseline covariates (e.g., baseline PSA, metastatic sites, ECOG performance status, and prior duration of ADT) in the multivariable Cox models; however, the effectiveness analysis set was modest (n = 99) and the number of overall survival events was limited, so adding multiple covariates could lead to unstable hazard‐ratio estimates. In addition, baseline PSA level and metastatic sites were not significant in univariable models, and ECOG performance status and prior duration of ADT were missing for 14 and 36 patients, respectively, further limiting robust multivariable adjustment. Third, the prospective design with close physician monitoring may have attenuated the severity of skin rash. Dose titration of apalutamide—including dose reductions and temporary interruptions—may have prevented progression to severe events, and therefore our findings may not fully reflect routine practice in real‐world clinical settings. Fourth, the exploratory genomic analyses (including the machine‐learning SNP findings) involved multiple comparisons and lacked external validation, raising the risk of overfitting and false‐positive associations; these results should be considered hypothesis‐generating. Future larger studies are needed to address these limitations and to validate genetic predictors and their clinical utility.
This study demonstrated the prognostic significance of baseline ctDNA in Japanese patients with mCSPC receiving apalutamide and identified potential genetic predictors for treatment‐related skin rash. These findings suggest opportunities for precision medicine in prostate cancer treatment, using both genomic profiling for prognosis and genetic testing for adverse event prediction. Further research is warranted to validate these biomarkers in larger and more diverse patient populations.
Author Contributions
Masaki Shiota: conceptualization, methodology, data curation, investigation, supervision, writing – review and editing. Taku Naiki: conceptualization, investigation, writing – review and editing. Koji Hatano: conceptualization, investigation, writing – review and editing. Akira Yokomizo: conceptualization, investigation, writing – review and editing. Hiroaki Tsuchiya: formal analysis, methodology, project administration, resources, software, supervision, validation, visualization, writing – original draft, writing – review and editing. Yoko Takahashi: project administration, writing – review and editing. Taku Nakayama: conceptualization, writing – original draft, writing – review and editing. Masaki Yoshida: conceptualization, formal analysis, methodology, writing – original draft, writing – review and editing. Hirotsugu Uemura: conceptualization, data curation, investigation, supervision, writing – review and editing.
Funding
The CUARTET study is funded by Janssen pharmaceutical K.K. (Grant Number: 56021927PCR4013).
Ethics Statement
The study protocol was reviewed and approved by the Clinical Research Network Fukuoka Certified Review Board (certification ID: CRB7180004) on September 2, 2020. The study was conducted in accordance with the principles of the Declaration of Helsinki and applicable local regulations. Written informed consent was obtained from all participants prior to enrollment and before any study‐related procedures were performed.
Conflicts of Interest
Masaki Shiota has received research funding or honorarium from Astellas Pharma Inc., AstraZeneca K.K., Bayer Yakuhin Ltd., Janssen pharmaceutical K.K. and Sanofi K.K. Koji Hatano has received research funding, honorarium or advisory roles from Accuray Inc., Astellas Pharma Inc., AstraZeneca K.K., Bayer Yakuhin Ltd., Chugai Pharmaceutical Co. Ltd., Janssen pharmaceutical K.K., Kissei Pharmaceutical Co. Ltd., PDRadiopharma Inc. and Takeda Pharmaceutical Company Limited. Akira Yokomizo has received honorarium, consulting or advisory roles from Astellas Pharma Inc., Bayer Yakuhin Ltd., Ferring Pharmaceuticals Co. Ltd., Janssen pharmaceutical K.K. and Takeda Pharmaceutical Company Limited. Hirotsugu Uemura has received research funding, honorarium, consulting or advisory roles from AstraZeneca K.K., Bristol Myers Squibb K.K., Janssen pharmaceutical K.K., MSD K.K., Ono Pharmaceutical CO. LTD. and Takeda Pharmaceutical Company Limited. Taku Naiki has received honorarium from Astellas Pharma Inc., Janssen pharmaceutical K.K. and Takeda Pharmaceutical Company Limited. Hiroaki Tsuchiya, Yoko Takahashi, Taku Nakayama and Masaki Yoshida are employees of Janssen pharmaceutical K.K.
Supporting information
Figure S1: Study design of the CUARTET study.
Figure S2: Heatmap of ctDNA comparing baseline status.
Figure S3: Prevalences of nadir PSA ≤ 0.2 ng/mL and ≤ 0.02 ng/mL achievement, CRPC‐free survival, and overall survival.
Table S1: Achievement of nadir PSA ≤ 0.02 ng/mL stratified by treatment status at study end in patients with paired samples.
Table S2: Overview of Adverse Drug Reaction in the Safety Analysis Set.
Table S3: Treatment‐Related Skin Rash Events and Treatment Adjustments after Apalutamide.
Table S4: Distribution of subsequent therapy.
Doc. S1. ctDNA gene panel measurement and analysis.
Acknowledgments
We thank A2 Healthcare Corporation, Riken Genesis Co. Ltd., and Hitachi Ltd. for their cooperation in conducting the study. We also thank MeDiCU Inc. for medical writing support in drafting and editing the manuscript, under the direction of the authors.
Data Availability Statement
The data that support the findings of this study are available from Johnson & Johnson under license. Access requires approval by Johnson & Johnson and is not assured. The authors cannot provide the data themselves.
References
- 1. Kyriakopoulos C. E., Chen Y. H., Carducci M. A., et al., “Chemohormonal Therapy in Metastatic Hormone‐Sensitive Prostate Cancer: Long‐Term Survival Analysis of the Randomized Phase III E3805 CHAARTED Trial,” Journal of Clinical Oncology 36, no. 11 (2018): 1080–1087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Hussain M., Fizazi K., Shore N. D., et al., “Metastatic Hormone‐Sensitive Prostate Cancer and Combination Treatment Outcomes: A Review,” JAMA Oncology 10, no. 6 (2024): 807–820. [DOI] [PubMed] [Google Scholar]
- 3. Clegg N. J., Wongvipat J., Joseph J. D., et al., “ARN‐509: A Novel Antiandrogen for Prostate Cancer Treatment,” Cancer Research 72, no. 6 (2012): 1494–1503. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Chi K. N., Agarwal N., Bjartell A., et al., “Apalutamide for Metastatic, Castration‐Sensitive Prostate Cancer,” New England Journal of Medicine 381, no. 1 (2019): 13–24. [DOI] [PubMed] [Google Scholar]
- 5. Chi K. N., Chowdhury S., Bjartell A., et al., “Apalutamide in Patients With Metastatic Castration‐Sensitive Prostate Cancer: Final Survival Analysis of the Randomized, Double‐Blind, Phase III TITAN Study,” Journal of Clinical Oncology 39, no. 20 (2021): 2294–2303. [DOI] [PubMed] [Google Scholar]
- 6. Uemura H., Arai G., Uemura H., et al., “Safety and Efficacy of Apalutamide in Japanese Patients With Metastatic Castration‐Sensitive Prostate Cancer Receiving Androgen Deprivation Therapy: Final Report for the Japanese Subpopulation Analysis of the Randomized, Placebo‐Controlled, Phase III TITAN Study,” International Journal of Urology 29, no. 6 (2022): 533–540. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Davies A., Azad A. A., and Kwan E. M., “Genomic Alterations to Guide Treatment Selection in Metastatic Prostate Cancer,” Critical Reviews in Oncogenesis 27, no. 1 (2022): 61–80. [DOI] [PubMed] [Google Scholar]
- 8. Vandekerkhove G., Struss W. J., Annala M., et al., “Circulating Tumor DNA Abundance and Potential Utility in De Novo Metastatic Prostate Cancer,” European Urology 75 (2019): 667–675. [DOI] [PubMed] [Google Scholar]
- 9. Masood D., Ren L., Nguyen C., et al., “Evaluation of Somatic Copy Number Variation Detection by NGS Technologies and Bioinformatics Tools on a Hyper‐Diploid Cancer Genome,” Genome Biology 25 (2024): 163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Scher H. I., Morris M. J., Stadler W. M., et al., “Prostate Cancer Clinical Trials Working Group 3. Trial Design and Objectives for Castration‐Resistant Prostate Cancer: Updated Recommendations From the Prostate Cancer Clinical Trials Working Group 3,” Journal of Clinical Oncology 34, no. 12 (2016): 1402–1418. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Wenzel M., Steuber T., Siech C., et al., “Ultralow Prostate‐Specific Antigen Nadir After Apalutamide: Outcomes in Synchronous Versus Metachronous Metastatic Hormone‐Sensitive Prostate Cancer,” European Urology Open Science 3, no. 77 (2025): 19–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Merseburger A. S., Agarwal N., Bjartell A., et al., “Targeted Investigational Treatment Analysis of Novel Anti‐Androgen (TITAN) Study: Ultralow Prostate‐Specific Antigen Decline With Apalutamide Plus Androgen‐Deprivation Therapy,” BJU International 134, no. 6 (2024): 982–991. [DOI] [PubMed] [Google Scholar]
- 13. Sweeney C. J., Chen Y. H., Carducci M., et al., “Chemohormonal Therapy in Metastatic Hormone‐Sensitive Prostate Cancer,” New England Journal of Medicine 373 (2015): 737–746. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Shiota M., Nemoto S., Ikegami R., et al., “Predictive Model of Castration Resistance in Advanced Prostate Cancer by Machine Learning Using Genetic and Clinical Data: KYUCOG‐1401‐A Study,” BJC Report 2, no. 1 (2024): 69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Leone G., Orlando F., Dutey‐Magni P., et al., “STAMPEDE Collaborators. Plasma AR Alterations and Timing of Intensified Hormone Treatment for Prostate Cancer: The STAMPEDE Phase 3 Randomized Clinical Trial,” JAMA Oncologia 11, no. 4 (2025): 426–429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Garcia de Herreros M., Jiménez N., Rodríguez‐Carunchio L., et al., “Prognostic Expression Signature of RB1, PTEN, and TP53 Genes in Patients With Metastatic Hormone‐Sensitive Prostate Cancer Treated With Androgen Receptor Pathway Inhibitors,” European Urology Open Science 70 (2024): 86–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Grist E., Dutey‐Magni P., Parry M. A., et al., “Tumor Transcriptome‐Wide Expression Classifiers Predict Treatment Sensitivity in Advanced Prostate Cancers,” Cell S0092‐8674, no. 25 (2025): 864–865. [DOI] [PubMed] [Google Scholar]
- 18. Pedrani M., Salfi G., Merler S., et al., “Integrating Aggressive‐Variant Prostate Cancer‐Associated Tumor Suppressor Gene Status With Clinical Variables to Refine Prognosis and Predict Androgen Receptor Pathway Inhibitor Response in Metastatic Hormone‐Sensitive Setting,” International Journal of Molecular Sciences 26, no. 11 (2025): 5309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Hage Chehade C., Graf R. P., Ozay Z. I., et al., “Association of Homologous Recombination Repair Alterations With Outcomes in Patients With Metastatic Hormone‐Sensitive Prostate Cancer,” Clinical Cancer Research 31 (2025): 4101–4109, 10.1158/1078-0432.CCR-25-1178. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Olmos D., Lorente D., Jambrina A., et al., “BRCA1/2 and Homologous Recombination Repair Alterations in High‐ and Low‐Volume Metastatic Hormone‐Sensitive Prostate Cancer: Prevalence and Impact on Outcomes,” Annals of Oncology 36, no. 10 (2025): 1190–1202. [DOI] [PubMed] [Google Scholar]
- 21. Salhi M., Lahmar O., Salah M. O., et al., “GLCCI1 and STIP1 Variants Are Associated With Asthma Susceptibility and Inhaled Corticosteroid Response in a Tunisian Population,” Journal of Asthma 58 (2021): 197–206. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1: Study design of the CUARTET study.
Figure S2: Heatmap of ctDNA comparing baseline status.
Figure S3: Prevalences of nadir PSA ≤ 0.2 ng/mL and ≤ 0.02 ng/mL achievement, CRPC‐free survival, and overall survival.
Table S1: Achievement of nadir PSA ≤ 0.02 ng/mL stratified by treatment status at study end in patients with paired samples.
Table S2: Overview of Adverse Drug Reaction in the Safety Analysis Set.
Table S3: Treatment‐Related Skin Rash Events and Treatment Adjustments after Apalutamide.
Table S4: Distribution of subsequent therapy.
Doc. S1. ctDNA gene panel measurement and analysis.
Data Availability Statement
The data that support the findings of this study are available from Johnson & Johnson under license. Access requires approval by Johnson & Johnson and is not assured. The authors cannot provide the data themselves.