Analiză din lumea reală privind cancerul de prostată hormonosensibil metastatic de novo: experiența pe parcursul a trei ani a unui centru regional din România

1. Introduction

Prostate cancer is the second most frequently diagnosed malignancy among men worldwide, with an estimated 1.2 million new cases and 359,000 deaths annually⁽¹⁾. The clinical spectrum of prostate cancer ranges from localized disease to advanced metastatic forms, and therapeutic strategies have evolved considerably over the recent years, now encompassing active screening and surveillance, surgery, radiotherapy, hormone therapy, chemotherapy, targeted agents and bone health interventions^(2,3). Despite these advances, metastatic hormone-sensitive prostate cancer (mHSPC) remains a significant unmet medical need, as patients with metastatic disease continue to experience poor outcomes and limited survival⁽³⁾.

Since 2015, several clinical trials have transformed the management of mHSPC, demonstrating survival benefits for agents such as docetaxel^(4,5), abiraterone^(5,6), apalutamide⁽⁷⁾, enzalutamide^(8,9) and darolutamide⁽¹⁰⁾, leading to their inclusion into international guidelines^(11,12). More recently, triplet therapy and intensified androgen deprivation therapy (ADT) regimens, including combinations with abiraterone or darolutamide, have shown efficacy and safety in studies such as PEACE-1 and ARASENS, and are now recommended as standard of care^(13-15). Radiotherapy has also emerged as a promising therapeutic pathway for patients with low metastatic burden, as evidenced by the HORRAD and STAMPEDE trials, as well as by real-world analyses^(3,5,16).

However, the populations enrolled in clinical trials often differ substantially from those encountered in routine clinical practice. Trial participants typically have better performance status, fewer comorbidities and better organ function, which may limit the replication of trial findings to the broader patient population^(3,17). In fact, only a small fraction of patients (estimated at 3% in the United States of America) participate in clinical trials, with underrepresentation of elderly, low-income and minority groups⁽¹⁸⁾. Strict eligibility criteria further exclude patients with poor performance status or significant comorbidities, resulting in a gap between trial evidence and real-world practice⁽¹⁷⁾.

Real-world data (RWD) analyses are increasingly recognized as essential for bridging this gap, providing insights into treatment patterns, effectiveness and safety in unselected patient populations^(17,18). RWD can validate and expand upon clinical trial findings, inform regulatory decisions and guide the design of future studies, particularly by highlighting the challenges and limitations of current therapeutic approaches in everyday practice⁽¹⁹⁾. Nevertheless, there remains a lack of published data regarding prostate cancer incidence, treatment patterns and risk factors specific to the Romanian population, hindering efforts to optimize care and inform policy.

In this context, our study aims to address these knowledge gaps by presenting real-world treatment patterns and early outcomes for patients with de novo mHSPC in a Romanian county-level referral hospital. By analyzing unselected patient cohorts and reporting on actual clinical practice, our findings offer a more realistic perspective on the management of mHSPC in Romania, with the potential to inform both local and national therapeutic strategies.

2. Materials and method

2.1. Study design and setting

A single-center, observational, retrospective cohort study was conducted at the Bihor County Emergency Clinical Hospital, Oradea, Romania, including patients diagnosed with de novo metastatic hormone-sensitive prostate cancer between 1 January 2021 and 31 December 2023. The data cutoff for this analysis was 31 December 2023. The study setting is a county-level referral institution providing multidisciplinary oncologic care through collaborative partnerships with urology, radiology, radiation oncology and palliative care services. Data were collected from anonymized files from the database of the hospital.

2.2. Study population and eligibility criteria

Eligible patients were adults (aged >18 years old) with histologically confirmed prostate adenocarcinoma and evidence of metastatic disease at initial diagnosis, as verified by medical imaging, including computed tomo­graphy (CT), magnetic resonance imaging (MRI) and bone scintigraphy when available. Patients presenting with metachronous metastatic disease, defined as relapse following prior localized therapy, were excluded from the analysis.

2.3. Patient characteristics and treatment variables

Baseline patient data included demographic information, Eastern Cooperative Oncology Group (ECOG) performance status, comorbidity profiles, Gleason/International Society of Urological Pathology (ISUP) grade, initial prostate-specific antigen (PSA) levels, TNM staging, sites of metastasis (classified as M1a, M1b, or M1c), and disease volume as defined by the CHAARTED criteria. Treatment-related variables encompassed the initial androgen deprivation therapy (ADT) modality (medical castration, surgical castration not performed), treatment intensification (including androgen receptor pathway inhibitor – ARPI, type and number of docetaxel cycles administered), administration of prostate-direc­ted radiotherapy, and documented reasons for limiting treatment intensification (such as ECOG performance status, recent cardiovascular or cerebrovascular events, patient preference, or reimbursement constraints).

2.4. Study outcomes

The study was designed with three key objectives. The primary objective was to characterize real-world first-line treatment patterns for patients with de novo metastatic hormone-sensitive prostate cancer at a Romanian county-level referral hospital, encompassing therapies initiated at diagnosis and during follow-up, and stratified by clinical subgroups such as disease volume and risk category. The secondary objective was to evaluate early clinical outcomes associated with first-line treatment selection, including time to castration resistance (TTCR), prostate-specific antigen response rates, and imaging-based assessments where available. The third exploratory objective was to delineate diagnostic pathways and baseline patient characteristics pertinent to treatment decision-making, including disease volume, risk stratification and performance status.

 2.5. Statistical analysis

Descriptive statistics were utilized to summarize patient demographics, disease characteristics and treatment patterns. Medians were reported where estimable; otherwise, event counts and follow-up status were detailed. Time-to-event outcomes were analyzed using institutional Kaplan-Meier survival estimates. Exploratory Cox proportional hazards models were employed to assess the potential prognostic impact of variables such as Gleason grade, metastatic category (M category), disease volume, comorbidity burden and treatment intensification. Given the limited sample size, these analyses are considered hypothesis-generating and should be interpreted with caution due to potential confounding. All descriptive statistical analyses were conducted using Microsoft Excel spreadsheet software. Kaplan-Meier survival analyses were performed using R software (R Foundation for Statistical Computing, Vienna, Austria).

2.6. Ethical statement

Due to the retrospective and observational, non-interventional design of the study, the requirement for patient informed consent was waived. All study procedures were conducted in accordance with the ethical standards set forth in the Declaration of Helsinki.

Results

3.1. Patient characteristics

A total of 286 individuals were diagnosed with prostate cancer at the regional center between January 2021 and December 2023, of whom 88 presented with de novo mHSPC. The median age among the de novo mHSPC cohort was 72 years old (range: 55-91). ECOG performance status was distributed as follows: eight patients with a score of 0; 25 with a score of 1; 18 with a score of 2; 34 with a score of 3; three patients with a score of 4; 11 patients had a recent major cardiovascular event. The median baseline PSA level was 126 ng/mL (data available for 87 patients; range: 5.4-2400 ng/mL).

3.2. Tumor characteristics

Tumor aggressiveness was evaluated using the Gleason grading system, with scores available for 83 patients. Of these, six patients had a Gleason score of 6, 25 had a score of 7, 22 had a score of 8, 22 had a score of 9, and eight patients had a score of 10. At baseline, bone metastases were identified in 68 patients, 12 exhibited extra-pelvic lymph node involvement, and eight had visceral metastases. According to CHAARTED criteria, approximately 70% of patients were classified as having high-volume disease, while 58% met high-risk criteria as defined by the LATITUDE trial. The summarized patient data can be found in Table 1.

3.3. Treatment patterns and outcomes

The median interval from biopsy to initial oncology consultation was 26.5 days (range: 8-91 days), and the median time from de novo mHSPC diagnosis to the initiation of therapy was one day (range: 1-80 days). All patients with de novo mHSPC received androgen deprivation therapy (one of the following: goserelin, triptorelin, leuprorelin) – 60 (68.2%) as monotherapy, six (6.81%) in combination with docetaxel (mean number of docetaxel cycles was 5.7), five (5.7%) with abiraterone, 13 (14.8%) with apalutamide, and four (4.5%) with the triple combination of ADT, docetaxel and abiraterone. Eight patients (9.1%) from the total cohort of 88 de novo mHSPC cases with low-volume disease underwent local radiotherapy, seven received ADT monotherapy, and one received ADT plus apalutamide. No surgical castration was performed, and first-generation antiandrogens (bicalutamide, flutamide) were not used. Median progression-free survival (mPFS) and TTCR for each treatment line can be found in Table 2. Six patients were lost to follow-up, and 20 had not progressed to the castration-resistant stage at the time of analysis; at the time of the data cutoff, 19 patients were deceased, all of them being on the ADT monotherapy treatment line. The median PSA level at six months post-treatment initiation was 0.465 ng/mL (data available for 62 patients; range: 0-442 ng/mL). The mPSA values at six months for each therapy line can be found in Table 3.

3.4. Kaplan-Meier analysis results

Kaplan-Meier survival analysis was conducted to evaluate the time to the development of metastatic castration-resistant prostate cancer (mCRPC) in the study cohort, with results stratified by several clinical and pathological variables.

When patients were grouped according to Gleason score categories (≤6; 7; ≥8), the analysis demonstrated a statistically significant difference in mCRPC-free survival among the groups. Patients with a Gleason score ≥8 exhibited the shortest median mCRPC-free survival, followed by those with a score of 7, while patients with a score ≤6 had the longest survival without progression to mCRPC (p=0.04). The distribution of events and the separation of survival curves indicated a clear association between higher Gleason scores and earlier onset of castration resistance (Figure 1).

When analyzing mCRPC-free survival by metastasis type (M1A, M1B, M1C), the Kaplan-Meier (KM) curves showed no statistically significant difference between the groups (log-rank p=0.65). Although prior boxplot summaries suggested a possible gradient (M1C > M1B > M1A) in time to progression, this visual difference was not supported by time-to-event statistics once censoring and follow-up time were properly accounted for in the KM model.

Six patients with M1B disease were excluded from the survival analysis because they were lost to follow-up before any evaluable event time (i.e., mcrpc_time missing). These censored cases could not contribute to the survival estimate, explaining the slightly lower M1B sample size (63 instead of 69). Boxplots summarized observed times only, while KM analysis correctly accounted for censored follow-up, providing a more accurate estimate of mCRPC-free survival probability.

Thus, even though the raw progression times appear shorter in M1C and M1B on simple plots, the KM log-rank test indicates that these differences are not statistically significant in this dataset (Figure 2).

The cohort was also analyzed based on metastatic disease volume at baseline, classified as low-volume or high-volume disease. Kaplan-Meier curves revealed that individuals with high-volume metastatic disease experienced a more rapid progression to mCRPC compared to those with low-volume disease. The difference in mCRPC-free survival between these two groups was statistically significant (p<0.001), with the high-volume group showing a steeper decline in survival probability over time (Figure 3).

Treatment regimens were compared by grouping patients who received ADT alone and those who received intensified regimens, which included ADT in combination with docetaxel and/or androgen receptor pathway inhibitors. The Kaplan-Meier analysis showed that the group treated with intensified therapy had a significantly longer duration of mCRPC-free survival compared to the ADT monotherapy group (p=0.038). The survival curves for these groups diverged early after treatment initiation and remained separated throughout the follow-up period. PSA response at six months post-treatment initiation was evaluated as a prognostic factor (Figure 4).

Patients with initial PSA≥20 ng/mL showed a trend toward shorter mCRPC-free survival compared with those below 20, but this did not reach statistical significance (p=0.093). This suggests that, while higher baseline PSA may be associated with earlier progression, the current sample size may limit the power to detect a significant effect (Figure 5).

Patients were stratified into three categories based on PSA levels at six months: ≤0.2 ng/mL, 0.2-4 ng/mL, and ≥4 ng/mL. The analysis demonstrated that patients with PSA levels ≤0.2 ng/mL at six months had the longest mCRPC-free survival, whereas those with PSA levels ≥4 ng/mL had the shortest survival (p<0.001). The intermediate group (0.2-4 ng/mL) showed survival outcomes between the other two groups. The survival curves were clearly separated and remained distinct throughout the observation period (Figure 6).

Finally, mCRPC-free survival was assessed according to LATITUDE risk classification, which categorizes patients as high-risk or low-risk based on specific clinical criteria. Patients classified as high-risk had a significantly shorter time to mCRPC compared to those in the low-risk group (p=0.002). The survival analysis showed a distinct difference in the timing of progression between these risk categories (Figure 7).

Across all subgroup analyses, the log-rank test was used to assess statistical significance, and the results consistently indicated that higher Gleason score, high-volume metastatic disease, absence of treatment intensification, higher PSA at six months and high-risk LATITUDE classification were each associated with reduced mCRPC-free survival.

Discussion

4.1. Overview

This single-center, retrospective cohort study describes real-world first-line management and early outcomes for patients with de novo mHSPC in a Romanian county referral hospital. The cohort was characterized by older age, higher disease burden and substantial comorbidity, reflecting a population that differs from those typically enrolled in phase III clinical trials. These findings are consistent with international real-world data, where delayed presentation to medical consultations and system-level constraints often limit the adoption of guideline-recommended treatment intensification^{(11,12,19-21)}.

4.2. Treatment patterns and barriers

A considerable proportion of patients received ADT monotherapy. This was influenced by several factors:

• Patient-level factors – frailty, high ECOG performance status, recent cardiovascular or cerebrovascular events limited eligibility for intensified therapy.
• System-level constraints – reimbursement issues and pandemic-related disruptions restricted access to novel agents and combination regimens^(22,23).
• Disease biology and timing – advanced disease volume, high PSA levels and visceral involvement at presentation reduced the opportunity for clinical stabilization and treatment intensification.

These barriers are consistent with reports from other real-world settings, where older age, comorbidities and access limitations are associated with lower rates of treatment intensification^(24-26).

 4.3. Prognostic factors and treatment outcomes

Several independent risk factors for earlier progression to mCRPC were identified, including high Gleason grade (8-10), high-volume disease and major comorbidities. These findings align with previous studies demonstrating the adverse prognostic impact of these features^(20,27,28). When comparing mCRPC-free survival across different metastasis types (M1A, M1B, M1C), our analysis did not reveal any statistically significant differences between the groups (log-rank p=0.65). While initial boxplot visualizations suggested a possible gradient in progression times (with M1C appearing to progress faster than M1B and M1A), these apparent differences were not confirmed by the Kaplan-Meier survival analysis, which properly accounted for censored data and variable follow-up durations. The exclusion of six M1B patients who were lost to follow-up before any evaluable event further underscores the importance of using time-to-event methods for accurate survival estimates. Overall, although raw data plots may hint at shorter progression times in certain subgroups, the more robust KM analysis indicates that these differences are not statistically significant in our current dataset. Treatment intensification with ADT plus an ARPI and/or docetaxel was associated with longer time to castration resistance, consistent with results from randomized trials such as CHAARTED, STAMPEDE, LATITUDE, TITAN, ARCHES, ENZAMET and triplet therapy^(5-10,20,28). Although subgroup sizes were limited and follow-up is ongoing, the observed benefit supports the use of intensified regimens in eligible patients.

4.4. PSA levels as a prognostic marker

PSA levels at six months post-treatment initiation were a strong prognostic indicator. Patients achieving a PSA≤0.2 ng/mL had significantly longer mCRPC-free survival, while those with PSA≥4 ng/mL experienced earlier progression. This supports the use of early PSA response as a practical marker for prognosis and treatment monitoring^(29-32).

Although patients with an initial PSA of 20 ng/mL or higher appeared to experience earlier progression to mCRPC, this trend did not achieve statistical significance in our analysis (log-rank p=0.093). This observation suggests that a higher baseline PSA may be linked to a greater risk of earlier disease progression; however, the lack of statistical significance indicates that our study may not have had sufficient sample size or power to confirm this association definitively. These findings highlight the need for further research with larger cohorts to clarify the prognostic value of baseline PSA in predicting mCRPC-free survival.

4.5. Clinical implications

The findings highlight the importance of:

• Prompt staging and early supportive optimiza­tion to improve eligibility for treatment intensification.
• Standardized documentation of disease volume and risk classification to guide therapy decisions.
• Expanding access to ARPIs beyond the current narrow criteria to allow more patients to benefit from evidence-based doublet therapy^(33-35).
• Integrating prostate-directed radiotherapy for low-volume disease in coordination with systemic therapy^(36,37).

4.6. Limitations

This study is limited by its retrospective, single-center design and small subgroup sizes, particularly for triplet therapy. Imaging modalities varied, which may have affected staging accuracy. The adverse events were not systematically recorded, and confounding by indication is possible, as fitter patients were more likely to receive intensified therapy. These analyses should be considered hypothesis-generating.

4.7. Future directions

Further research should focus on multicenter registry collaborations with standardized data collection to enable more robust comparisons of treatment strategies. Incorporating genomic profiling and evaluating care pathways that support treatment intensification after functional recovery may improve access to effective therapies. Public health initiatives to promote earlier detection could reduce the proportion of de novo metastatic presentations and improve outcomes.

5. Conclusions

This retrospective, single-center study provides a real-world perspective on the management of de novo metastatic hormone-sensitive prostate cancer in a Romanian county referral hospital. The findings highlight that a substantial proportion of patients present with advanced disease, significant comorbidities and limited functional status, which often restricts the use of treatment intensification strategies recommended by international guidelines. Despite these challenges, intensified therapy with androgen receptor pathway inhibitors and/or docetaxel was associated with improved time to castration resistance, consistent with evidence from randomized clinical trials.

Early PSA response at six months emerged as a strong prognostic marker, supporting its use in routine monitoring and risk stratification. The study also identified high Gleason grade, high-volume disease, visceral metastases and comorbidity burden as key predictors of earlier progression to castration-resistant disease. These results underscore the need for prompt diagnosis, comprehensive supportive care and standardized treatment pathways to optimize outcomes for patients with de novo mHSPC.

Expanding access to novel therapies and integrating multidisciplinary care are important steps toward improving real-world outcomes. Future research should focus on multicenter collaborations, standardized data collection and strategies to increase early detection and equitable access to evidence-based treatments.   

Acknowledgements. The authors would like to thank AstraZeneca for providing support with the statistical analyses included in this manuscript.

Corresponding author: Lajos Szilard Bereczki E-mail: bereczki.szilard77@gmail.com

Conflict of interest: none declared.

Financial support: none declared.

This work is permanently accessible online free of charge and published under the CC-BY licence.