General Information
Conflict of Interest
Request Clinical Trials
Associated Trial(s): What type of data are you looking for?: Individual Participant-Level Data, which includes Full CSR and all supporting documentationRequest Clinical Trials
Data Request Status
Status: OngoingResearch Proposal
Project Title: Prognostic Impact of Low Initial PSA in mHSPC Patients with Extensive Bone Metastases: A Post Hoc Analysis of the TITAN Trial
Scientific Abstract:
Background:
Our retrospective data suggest that patients with metastatic hormone-sensitive prostate cancer (mHSPC) who have extensive bone metastases but relatively low initial PSA levels may have significantly worse overall survival (OS). This apparent paradox--worse survival despite lower initial PSA levels--may reflect a subgroup with low PSA production and aggressive, androgen receptor (AR)-independent tumor biology. It remains unclear whether these patients derive the same survival benefit from AR-targeted therapy. In particular, it is of clinical interest to assess whether the survival benefit of apalutamide plus androgen deprivation therapy (ADT), compared to ADT alone, differs in patients with relatively low PSA levels versus the overall trial population.
Objective:
To evaluate the prognostic impact of relatively low initial PSA in mHSPC patients with extensive bone metastases, and to explore the OS benefit of upfront apalutamide therapy in this subgroup compared to those with higher PSA levels.
Participants and Methods:
Post hoc analysis of the TITAN Phase 3 trial.
Patients stratified by extent of disease (EOD) using bone metastasis data.
PSA levels categorized into quartiles; a cutoff (e.g., <200 ng/mL) will be used for subgroup analysis.
Outcomes:
Primary: Overall survival (OS)
Secondary: Progression-free survival (PFS), PSA response, and second PFS (PFS2), if available
Statistical Analysis:
Kaplan--Meier estimates and Cox models (adjusted for covariates)
Restricted cubic spline modeling to assess PSA--OS relationship
Subgroup analysis: apalutamide + ADT vs. ADT
Brief Project Background and Statement of Project Significance:
Recent retrospective data from our multi-institutional study in Japan suggest that in patients with de novo metastatic hormone-sensitive prostate cancer (mHSPC) and high tumor burden (extent of disease [EOD] >=2), relatively low initial PSA levels (<200 ng/mL) are associated with significantly shorter overall survival (OS). This observation appears paradoxical and may reflect an aggressive tumor phenotype with limited PSA production, possibly due to reduced androgen receptor (AR) dependence. However, the original study was limited by sample size and consisted predominantly of patients treated with abiraterone.
Using the TITAN trial dataset, we aim to validate this hypothesis by examining whether patients with extensive bone metastases and relatively low baseline PSA levels experience worse prognosis, even when treated with apalutamide plus ADT. The robust sample size and uniform treatment protocol of the TITAN trial allow for a well-powered, exploratory subgroup analysis. Initial PSA values will be analyzed by quartiles, and an optimal cutoff will be determined (e.g., 200 ng/mL) to define the low-PSA subgroup.
In our prior study, no significant difference in PFS was observed between low- and high-PSA groups, suggesting that ARSI treatment was effective even in low-PSA cases. However, the TITAN trial includes both ARSI-treated and ADT-only groups, providing a unique opportunity to assess whether the addition of apalutamide offers disproportionately greater OS and PFS benefit in patients with low baseline PSA compared to the overall cohort.
Furthermore, among ARSIs, apalutamide is known to achieve particularly rapid and profound PSA reductions. If our hypothesis is confirmed, it would support the notion that patients with relatively low PSA and high tumor burden should be considered for early, aggressive treatment with apalutamide. This could lead to more nuanced, PSA-guided treatment decisions in mHSPC and contribute to more personalized therapeutic strategies.
References:
Fujimoto S et al. Prognostic Impact of Initial PSA in mHSPC Patients with Extensive Bone Metastases under ARSI. The Prostate (Under Review). Manuscript ID: PROS-25-270.
Chi KN et al. Apalutamide for Metastatic, Castration-Sensitive Prostate Cancer. N Engl J Med. 2019;381:13--24.
Specific Aims of the Project:
The primary aim of this study is to validate the hypothesis that, in patients with metastatic hormone-sensitive prostate cancer (mHSPC) and high tumor burden (extent of disease >=2), relatively low baseline PSA levels are associated with worse overall survival (OS). This finding, based on our recent retrospective study, will be evaluated in a well-powered post hoc analysis using the TITAN trial dataset.
A secondary aim is to determine whether the magnitude of clinical benefit from adding apalutamide to ADT is greater in patients with low PSA levels compared to those with higher PSA, and compared to the overall study population. Quartile-based stratification of PSA levels will be used initially to define cutoffs for subgroup analysis.
We hypothesize that patients with low PSA levels but high disease burden may have more aggressive, potentially AR-independent tumor biology, making early, potent AR pathway inhibition particularly beneficial. If confirmed, this could support clinical practice.
Additional endpoints include progression-free survival (PFS), PSA response, and second progression-free survival (PFS2), depending on data availability.
Study Design: Individual trial analysis
What is the purpose of the analysis being proposed? Please select all that apply.: New research question to examine treatment effectiveness on secondary endpoints and/or within subgroup populations Confirm or validate previously conducted research on treatment effectiveness Participant-level data meta-analysis Meta-analysis using only data from the YODA Project
Software Used: R, STATA
Data Source and Inclusion/Exclusion Criteria to be used to define the patient sample for your study:
We will use individual participant-level data from the TITAN Phase 3 randomized clinical trial (NCT02489318), which compared apalutamide plus androgen deprivation therapy (ADT) versus ADT alone in patients with metastatic hormone-sensitive prostate cancer (mHSPC).
Inclusion criteria:
Patients diagnosed with de novo mHSPC
Baseline PSA value available
Bone metastasis data available (e.g., lesion counts or variables that allow estimation of extent of disease [EOD])
Randomized to either apalutamide + ADT or placebo + ADT
Exclusion criteria:
Missing baseline PSA or bone metastasis information
Prior systemic treatment for prostate cancer before trial enrollment
Patients with visceral-only metastases, if bone status is unknown
Data analysis will be conducted within the secure YODA platform using available statistical tools (R, Stata, RStudio). No additional datasets from outside the YODA Project will be used.
We will initially stratify patients by PSA quartiles to explore prognostic trends. A subsequent analysis will use a cutoff (e.g., 200 ng/mL) to define a "relatively low PSA" group. Baseline characteristics and outcomes (OS, PFS, PFS2 if available) will be compared across PSA-defined subgroups, with subgroup comparisons also made between treatment arms. Additional descriptive analysis of clinical and laboratory characteristics (e.g., Gleason pattern, LDH, ALP) will be performed to better characterize the low-PSA/high-EOD subgroup.
Primary and Secondary Outcome Measure(s) and how they will be categorized/defined for your study:
Primary Outcome Measure:
Overall survival (OS), defined as the time from randomization to death from any cause. Patients who are alive at the time of analysis will be censored at their last known date of follow-up. OS will be compared between PSA-defined subgroups, particularly focusing on the low PSA (<200 ng/mL) and high extent of disease (EOD >=2) population. Additional analysis will examine the OS difference between treatment arms (apalutamide + ADT vs. ADT alone) within these PSA subgroups.
Secondary Outcome Measures:
Radiographic progression-free survival (rPFS): defined as the time from randomization to radiographic disease progression or death, whichever occurs first, consistent with the original TITAN protocol.
Second progression-free survival (PFS2): defined as the time from randomization to progression on subsequent therapy or death, if available in the dataset.
PSA response rate: proportion of patients who achieve >=50% reduction in PSA from baseline, and time to PSA nadir.
Time to CRPC (castration-resistant prostate cancer): if defined and available in the dataset.
Where available, these outcomes will be evaluated both in the overall cohort and stratified by PSA quartiles or dichotomized groups (e.g., <200 vs >=200 ng/mL). We will also explore interaction effects between PSA level and treatment group with respect to each outcome.
The endpoints selected mirror those previously reported in the TITAN trial and allow for clinically meaningful validation of subgroup-specific outcomes relevant to real-world treatment decision-making.
Main Predictor/Independent Variable and how it will be categorized/defined for your study:
The main independent variable in this study is the baseline prostate-specific antigen (PSA) level measured at trial enrollment. PSA will be treated as both a continuous and categorical variable.
First, PSA will be analyzed as a continuous predictor using restricted cubic spline regression to assess its non-linear relationship with overall survival (OS). Second, PSA will be stratified into quartiles for exploratory subgroup comparisons. Based on findings from our prior retrospective study, we will then dichotomize PSA using a clinically relevant threshold (e.g., <200 ng/mL vs. >=200 ng/mL) to define a "relatively low PSA" group.
In addition to PSA, treatment group (apalutamide + ADT vs. ADT alone) will be used as an independent variable, particularly for interaction analysis. We will examine whether the survival benefit of apalutamide is more pronounced in the low PSA subgroup compared to higher PSA levels or the overall study population.
An interaction term between PSA level (as continuous or dichotomized) and treatment arm will be included in multivariable Cox regression models to formally test effect modification. Other independent variables considered for adjustment include age, performance status, extent of disease, visceral metastases, and Gleason score, depending on data availability.
The combined use of PSA and treatment arm as independent variables will help elucidate the prognostic and predictive value of PSA in the context of ARSI treatment in mHSPC.
Other Variables of Interest that will be used in your analysis and how they will be categorized/defined for your study:
The following variables will be used for multivariable risk adjustment and descriptive analysis:
Extent of disease (EOD): defined based on the number and distribution of bone metastases. If an exact EOD score is not available, surrogate indicators such as lesion count or predefined categories (e.g., high vs. low volume disease) will be used.
Age: analyzed as a continuous variable and also categorized into clinically relevant groups (e.g., <70 vs. >=70 years).
ECOG Performance Status: used as an ordinal or binary variable (e.g., 0--1 vs. >=2) depending on distribution.
TNM Classification: clinical T stage (e.g., T1--2 vs. T3--4) and N stage (node-negative vs. node-positive) will be used to capture primary tumor burden and lymphatic spread.
Visceral metastasis: binary variable indicating the presence or absence of non-bone metastases (e.g., lung or liver).
Gleason score or ISUP grade group: categorized into low (<=7 [3+4]), intermediate (7 [4+3]), and high (8--10), if available.
Hemoglobin, LDH, and ALP levels: treated as continuous variables and possibly dichotomized using institutional upper limits of normal or median values. These will be included to explore their association with low-PSA/high-EOD phenotypes and to adjust for tumor aggressiveness.
Treatment arm: included as a binary variable (apalutamide + ADT vs. placebo + ADT), used in subgroup and interaction analyses.
These variables will be incorporated into Cox proportional hazards models and subgroup analyses to control for potential confounders and characterize the clinical profile of patients with relatively low PSA and high metastatic burden.
Statistical Analysis Plan:
Descriptive statistics will be used to summarize baseline characteristics stratified by quartiles of baseline PSA level and by treatment arm (apalutamide + ADT vs. placebo + ADT). Continuous variables will be reported as means with standard deviations or medians with interquartile ranges, depending on distribution. Categorical variables will be summarized using frequencies and proportions.
Baseline PSA will be treated as both a continuous and categorical variable. Initially, PSA will be modeled continuously using restricted cubic spline (RCS) regression to explore potential non-linear associations with overall survival (OS). Then, patients will be stratified by PSA quartiles to describe trends in clinical outcomes across the distribution. Based on prior retrospective research, a dichotomized variable (<200 vs. >=200 ng/mL) will be used to define a "relatively low PSA" subgroup for primary survival analysis.
The primary outcome is overall survival (OS), and secondary outcomes include radiographic progression-free survival (rPFS), PSA response (e.g., >=50% reduction), time to PSA nadir, and second progression-free survival (PFS2), if available. Time-to-event endpoints will be analyzed using the Kaplan-Meier method with log-rank tests. Multivariable Cox proportional hazards models will be employed to assess associations between baseline PSA (continuous and categorical) and survival outcomes, adjusting for relevant covariates: age, ECOG performance status, extent of disease (EOD), bone metastasis count, visceral metastases, Gleason score, TNM stage, LDH and ALP (if available), and treatment arm.
Interaction analyses will be performed to determine whether the effect of apalutamide on OS differs across PSA subgroups. This includes adding an interaction term between PSA (continuous or dichotomized) and treatment arm in Cox models. Subgroup analyses will be conducted in patients with high metastatic burden (e.g., EOD >=2) to evaluate whether apalutamide provides a greater survival benefit in those with relatively low PSA levels.
PSA response curves and time to nadir will be visualized and compared between treatment arms within PSA strata. Proportional hazards assumptions will be tested using Schoenfeld residuals. Model discrimination and calibration may be assessed with Harrell's C-index and calibration plots as appropriate.
All analyses will be conducted within the secure YODA platform using R or Stata. A two-sided p-value <0.05 will be considered statistically significant. No adjustment for multiplicity will be performed given the exploratory nature of subgroup and interaction analyses.
Narrative Summary: Our prior study suggested that in mHSPC patients with high disease burden (EOD >=2), relatively low initial PSA (<200 ng/mL) was associated with worse overall survival, possibly due to AR-independent tumor biology. However, most patients received abiraterone and sample size was limited. Using the TITAN trial dataset, we aim to validate this hypothesis in a broader cohort, focusing on the apalutamide arm. We will also assess whether the benefit of ADT + APA over ADT alone is more pronounced in patients with low PSA, who may respond well despite poor prognosis. As apalutamide induces rapid, deep PSA responses, such patients may especially benefit from intensified therapy. This analysis may help optimize treatment selection in aggressive mHSPC.
Project Timeline:
Project start date: August 1, 2025 (upon approval and data access)
Data familiarization and preprocessing: August 2025
Primary statistical analysis completion: October 2025
Secondary/subgroup analyses and figure preparation: November--December 2025
Manuscript drafting and internal review: January 2026
Manuscript submission to peer-reviewed journal: February 2026
Revisions and resubmission (if needed): March--May 2026
Final report submission to the YODA Project: June 2026
The entire project is expected to be completed within the 12-month data use period. If needed, a request for extension will be submitted. We plan to publish the results in a peer-reviewed oncology or urology journal and will report findings back to YODA regardless of publication outcome.
Dissemination Plan:
The results of this study will be disseminated through submission of a full-length original manuscript to a peer-reviewed journal in the fields of oncology or urology. Target journals include European Urology, Journal of Clinical Oncology, or The Prostate, depending on the scope and strength of the findings. The manuscript will focus on the prognostic impact of baseline PSA in patients with extensive bone metastases and the differential treatment effect of apalutamide in this subgroup.
In addition, a summary report of key findings will be submitted to the YODA Project as required, and the team will consider publicly accessible formats for knowledge dissemination (e.g., preprints, institutional repositories) to promote transparency and accelerate scientific exchange.
The target audience includes academic researchers, clinicians, and guideline developers who focus on personalized treatment strategies for metastatic prostate cancer.
Bibliography:
- Fujimoto S, Fujita K, et al. Prognostic Impact of Initial PSA in mHSPC Patients with Extensive Bone Metastases under ARSI. The Prostate. Submitted May 6, 2025. Manuscript ID: PROS-25-270
This retrospective study assessed the prognostic significance of initial prostate-specific antigen (iPSA) levels in 276 patients with high-risk metastatic hormone-sensitive prostate cancer (mHSPC) and extensive bone metastases (extent of disease score >=2) treated with androgen receptor signaling inhibitors (ARSI) as first-line therapy. Patients were stratified into quartiles based on iPSA levels, with a cutoff of 200 ng/mL for subgroup analysis. Kaplan-Meier and Cox regression analyses revealed a significant association between iPSA quartiles and overall survival (OS), with the lowest survival observed in the lowest iPSA group (<200 ng/mL). Restricted cubic spline analysis indicated an inverted J-shaped relationship between iPSA and OS, with the lowest hazard ratio at 1664 ng/mL. Patients with iPSA <200 ng/mL had significantly shorter OS (p = 0.015), while progression-free survival (PFS) did not differ significantly (p = 0.869). These findings suggest that initial PSA levels are associated with prognosis in high-risk mHSPC with extensive bone metastases.
Figures from this manuscript has been provided as Supplementary Material.
- Chi KN, Agarwal N, Bjartell A, et al. Apalutamide for Metastatic, Castration-Sensitive Prostate Cancer. New England Journal of Medicine. 2019;381(1):13--24. doi:10.1056/NEJMoa1903307
- Mahal BA, Yang DD, Spratt DE, et al. Clinical and Genomic Characterization of Low--PSA, High-Grade Prostate Cancer. European Urology. 2018;74(2):146--154. doi:10.1016/j.eururo.2018.02.014
- Sweeney CJ, Chen YH, Carducci M, et al. Chemohormonal Therapy in Metastatic Hormone-Sensitive Prostate Cancer. New England Journal of Medicine. 2015;373(8):737--746. doi:10.1056/NEJMoa1503747
Supplementary Material: supple.pdf
