Abstract
Objective
Locally advanced prostate cancer (clinical stage T3 or higher) causes significant harm to patients, including decreased quality of life, high mortality rates, and economic burden. Current multimodal management for locally advanced prostate cancer often includes radiotherapy combined with long-term androgen deprivation therapy (ADT), with radical prostatectomy (RP) reserved for selected patients. However, RP alone or with ADT has limitations such as high biochemical recurrence rates and high positive surgical margin (PSM) rates. Therefore, this study aims to explore new effective treatment strategies.
Methods
In accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, we systematically searched PubMed, Embase, Web of Science and Cochrane databases up of August 19, 2025. We analyzed data from 10 studies. The main outcomes assessed included biochemical progression-free survival (BPFS), PSM rate, overall survival (OS) and metastasis-free survival (MFS). We pooled risk ratios (RR) with 95% confidence intervals (CI) using fixed- or random-effects models.
Results
The results show that compared with RP alone, the combination of NCHT and RP significantly improved BPFS (RR = 0.60, 95% CI: 0.41-0.87, P = 0.008), despite significant heterogeneity (I² = 90%). In RCT subgroup analysis, the benefit remained significant (RR = 0.73, 95% CI: 0.62–0.86; P = 0.0002; I² = 0%). The NCHT+RP group also showed a significant reduction in PSM rate (RR = 0.49, 95% CI: 0.37-0.65, P < 0.00001). For OS, the pooled analysis showed an improvement (RR = 0.45, 95% CI: 0.24-0.84, P = 0.01); however, this effect was entirely driven by non-randomized studies (RR = 0.55, 95% CI: 0.31–0.99; P = 0.04), while the randomized controlled trials (RCT) subgroup did not confirm a significant benefit (RR = 0.43, 95% CI: 0.10-1.88, P = 0.26). No significant improvement was observed in MFS (RR = 0.82, 95% CI: 0.45-1.49, P = 0.51), although the single large RCT (Eastham et al.) favored NCHT+RP (RR = 0.78, 95% CI: 0.63–0.98; P = 0.03).
Conclusion
In selected patients with locally advanced prostate cancer, NCHT combined with RP may improve BPFS and reduce PSM rates, with the BPFS benefit confirmed in the RCT subgroup. However, the apparent OS benefit was not observed in RCTs and likely reflects bias in observational studies. According to GRADE, the certainty of evidence for survival outcomes is very low. These positive findings should be interpreted with caution, and adequately powered RCT with modern agents are needed.
Keywords: androgen deprivation therapy (ADT), neoadjuvant chemohormonal therapy, prostate cancer, radical prostatectomy, randomized controlled trial
Introduction
Prostate cancer (PCa) remains one of the most prevalent malignancies among men worldwide, with locally advanced disease (clinical stage T3 or higher) posing significant therapeutic challenges (1). These patients often experience reduced quality of life, high mortality rates, and substantial economic burdens due to costly treatments and long-term care (2–4). Current multimodal management for locally advanced PCa most commonly involves radiotherapy combined with long-term androgen deprivation therapy (ADT), with radical prostatectomy (RP) considered for selected patients (5). These approaches are associated with persistent limitations, such as high rates of biochemical recurrence and positive surgical margins (PSM), underscoring the need for more effective treatment strategies.
However, the landscape of neoadjuvant therapy for high-risk prostate cancer is rapidly evolving. Contemporary trials are increasingly focused on intensified androgen receptor pathway inhibitors (ARPIs; e.g., enzalutamide, apalutamide, darolutamide) with or without ADT, which have demonstrated superior pathologic complete response (pCR) rates compared to historical docetaxel-based chemohormonal therapy in early-phase studies (6). The largest randomized controlled trial to date evaluating neoadjuvant docetaxel-based neoadjuvant chemohormonal therapy (NCHT) (CALGB 90203) did not meet its primary endpoint of 3-year biochemical progression-free survival, and current National Comprehensive Cancer Network (NCCN) guidelines do not recommend neoadjuvant ADT or chemotherapy prior to radical prostatectomy outside of clinical trials (7). Therefore, while our meta-analysis focuses on the available evidence for NCHT+RP, its findings must be interpreted within this evolving therapeutic context.
NCHT, which combines ADT with chemotherapy prior to RP, has emerged as a promising approach to downstage tumors and improve surgical outcomes. While some studies suggest potential benefits in biochemical progression-free survival (BPFS), evidence regarding its impact on overall survival (OS) and metastasis-free survival (MFS) remains inconclusive (8, 9). Previous systematic reviews have reported mixed results, partly due to variations in study design, patient populations, and treatment protocols (10).
Given these inconsistent findings and to provide a more comprehensive evaluation, we conducted a systematic review and meta-analysis of both randomized controlled trials (RCTs) and comparative cohort studies. This study aims to synthesize current evidence on the efficacy of NCHT combined with RP versus RP alone, focusing on key oncological outcomes including BPFS, OS, PSM, and MFS. Our findings seek to inform clinical decision-making and highlight directions for future research.
Materials and methods
Protocol and guidance
This systematic review and meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Supplementary Table 1) (11). Our protocol has been registered with the International Prospective Register of Systematic Reviews (PROSPERO, CRD420251123186).
Data sources, search strategy and definitions
A thorough review of existing literature was conducted up to August 19, 2025, utilizing the PubMed, Embase, Web of Science, and Cochrane databases. Our search incorporated terms such as neoadjuvant chemohormonal therapy, radical prostatectomy, prostate cancer, locally advanced prostate cancer, and oncological outcomes. To enhance our search, we also included synonyms related to these key concepts found in titles, abstracts, and keywords across the databases. Detailed search strategies for each database can be found in Supplementary Table 2.
Study selection: inclusion and exclusion criteria
Detailed inclusion criteria were defined according to the PICOS framework (Table 1). Inclusion criteria were: 1) Population: enrolled patients with histologically confirmed locally advanced prostate cancer (clinical stage T3 or higher); 2) Intervention: administered NCHT neoadjuvant chemohormonal therapy (any regimen of ADT combined with chemotherapy) prior to RP; 3) Comparator: compared the intervention with RP alone; 4) Outcomes: reported data on at least one of BPFS, OS, PSM rate, or MFS; 5) Study Design: RCTs or comparative cohort studies. Studies must include a clear definition of the intervention and comparator groups. We excluded studies without a control group, non-comparative designs (e.g., case reports, editorials, conference abstracts), and those lacking relevant outcome data.
Table 1.
PICOS eligibility criteria for inclusion of studies in the meta-analysis.
| P (Population) | I (Intervention) | C (Comparator) | O (Outcomes) | S (Study Design) |
|---|---|---|---|---|
| Patients with locally advanced prostate cancer (clinical stage T3 or higher) | Neoadjuvant chemohormonal therapy (NCHT), consisting of androgen deprivation therapy (ADT) combined with chemotherapy, administered prior to radical prostatectomy (RP) | Radical prostatectomy (RP) alone | Biochemical progression-free survival (BPFS) Positive surgical margin (PSM) rate Overall survival (OS) Metastasis-free survival (MFS) |
Randomized controlled trials (RCTs) Comparative cohort studies |
Data items and extraction process
The flow chart illustrating the selection process of studies and the rationale for exclusions is depicted in Figure 1. Initially, duplicates were eliminated, followed by a review of titles and abstracts to assess their relevance. The full-texts of the articles were then scrutinized for eligibility based on previously established criteria. A standardized, pre-piloted data extraction form was used. Extracted items included: study identifiers (first author, year, country); study design and methodology; participant characteristics; intervention details; comparator details; outcome definitions and measurement timepoints; and results for each outcome. For time-to-event outcomes, we sought to extract hazard ratios (HRs) from multivariate analyses. If these were not reported, we extracted event numbers at the longest available follow-up to calculate risk ratios (RRs), as described in the statistical analysis section. All extraction was performed independently by two authors (X.B.W, C.H) with conflicts resolved by consensus or third-author (H.L) adjudication.
Figure 1.
Flow diagram of study selection according to PRISMA guidelines.
Risk of bias assessment
The Cochrane ROB 2 tool was utilized to evaluate the bias risk in the RCTs we included (12). These assessments were derived from the information presented in each study, categorizing the bias risk as “Low,” “High,” or “ some concern”. Each study was assigned a classification of low-risk, some concern or high-risk based on these evaluations, with explanations provided for each rating (Table 2). The methodological quality of the six included retrospective cohort studies was evaluated using the Newcastle-Ottawa Scale (NOS) (Table 3) (13). Two authors (X.B.W, C.H) conducted independent evaluations of the study quality, and any disagreements were addressed with the assistance of a third author (H.L).
Table 2.
Risk of bias assessment using the Cochrane RoB 2 tool across included studies.
Table 3.
Quality assessment of non-randomized studies using the Newcastle–Ottawa Scale (NOS).
| Study, country | Selection (4) | Comparability (2) | Outcome (3) | Total score | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Representativeness of the exposed cohort | Selection of the non-exposed cohort | Ascertainment of exposure | Demonstration that outcome of interest was not present at start of study | Age comparability | Additional comparability | Assessment of Outcome | Enough follow-up | Adequacy of follow up of Cohorts | ||
| Fujita et al., 2017 (19), Japan | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 9 |
| Silberstein et al., 2015 (15), USA | 0 | 1 | 1 | 1 | 1 | 0 | 1 | 1 | 1 | 7 |
| Yokoyama et al., 2025 (20), Japan | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 9 |
| Sugino et al., 2023 (21), Japan | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 9 |
| Oishi et al., 2023 (22), Japan | 0 | 0 | 0 | 1 | 0 | 0 | 1 | 1 | 0 | 3 |
| Pan et al., 2019 (18), China | 0 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 0 | 7 |
Evidence quality assessment
This research utilized the GRADE (Grading of Recommendations Assessment, Development and Evaluation) framework to assess the quality of evidence concerning the primary outcome measures. The evaluation through GRADE was influenced by five factors that could lead to downgrading: bias in studies, inconsistency of results, indirectness, lack of precision and publication bias. Each outcome’s evidence was categorized into one of four levels: “high”, “medium”, “low” or “very low”. The assessment was conducted independently by two researchers, and the findings were compiled using GRADEpro GDT software (https://gradepro.org). If any disagreements arose, a third researcher was brought in to provide a resolution and facilitate consensus.
Statistical analysis
The analysis of data was performed utilizing Review Manager 5.3 software provided by the Cochrane Collaboration. For time-to-event outcomes, (BPFS, OS, MFS), we planned to extract hazard ratios (HRs) with 95% confidence intervals from multivariate analyses as the preferred effect measure. However, several included studies did not report HRs or provide Kaplan–Meier curves with sufficient detail to allow extraction of individual patient data or calculation of HRs using established methods. Therefore, as a pragmatic approach, we pooled risk ratios (RRs) at the longest available follow-up for each study. We acknowledge that this approach does not account for time-to-event nature of the data, censoring, or differential follow-up durations, and may either underestimate or overestimate treatment effects. This limitation is addressed in the discussion. Statistical heterogeneity was assessed using the Cochran Q test and I² statistic, with I² ≥ 50% indicating substantial heterogeneity. We explored potential sources of heterogeneity through prespecified subgroup analyses (by study design, geographic region, and chemotherapy regimen) and sensitivity analyses using the leave-one-out method. Given the small number of RCTs for OS and MFS, subgroup results should be interpreted with caution due to limited statistical power. Publication bias was assessed using funnel plots, Egger’s test, and trim-and-fill analysis, although the power of these tests is limited with fewer than 10 studies.
Results
Study selection and characteristics
A systematic search across PubMed, EMBASE, Web of Science, and Cochrane yielded 2,075 records. After removing 635 duplicates, 1,440 studies underwent title and abstract screening, excluding 1,332 records. Of 108 full-text articles assessed, 95 were excluded due to: absence of a control group (n = 15), incomplete data (n = 8), irrelevant outcomes (n = 11), conference abstracts (n = 8), or non-original research (n = 53). Ultimately, 13 studies were included for qualitative synthesis, and 10 studies (comprising 4 randomized controlled trials and 6 retrospective cohort studies, with a total of 2,761 patients) were eligible for quantitative meta-analysis (Figure 1). Three studies were excluded from quantitative synthesis due to unmergeable data. Ultimately, the final selection consisted of four RCTs and six retrospective cohorts, with two conducted in USA (14, 15), one in Russia (16), two in China (17, 18), and five in Japan (19–23). Ten studies compared NCHT plus RP versus RP alone (Table 4).
Table 4.
Baseline characteristics and interventions included in the meta-analysis.
| Author, Year | Country | Study design | Interventions | Sample size | Age (years) | Follow−up (months) |
|||
|---|---|---|---|---|---|---|---|---|---|
| NCHT+RP | RP | NCHT+RP | RP | NCHT+RP | RP | ||||
| Eastham et al., 2020 (14) | USA | RCT | Docetaxel + androgen deprivation + RP vs. RP alone | 391 | 397 | 62 (IQR:40 - 78) | 63 (IQR:33 - 84) | 73.2 (IQR:0 -145.2) | 73.2 (IQR:0 -145.2) |
| Narita et al., 2018 (23) | Japan | RCT | Docetaxel + estramustine + androgen blockade + RP vs. RP alone | 56 | 56 | 71 (IQR:69 - 75) | 72 (IQR:69 - 75) | 14.0 (IQR:6.0 -25.5) | 25.6 (IQR:11.3 -47.2) |
| Qu et al., 2021 (17) | China | RCT | Docetaxel + androgen deprivation + RP vs. RP alone | 42 | 54 | 64 (IQR:57 – 75) | 67 (IQR:60 – 75) | NA | NA |
| Berkut et al., 2020 (16) | Russia | RCT | Docetaxel + androgen deprivation + RP vs. RP alone | 36 | 35 | 62.88 ± 7.37 | 63.77 ± 7.23 | 32.78 ± 16.49 | 42.69 ± 22.34 |
| Fujita et al., 2017 (19) | Japan | Retrospective | Estramustine + androgen deprivation + RP vs. RP alone | 436 | 177 | 68 (IQR: 64 –7 2) | 68 (IQR:64 – 71) | 48.8 (IQR:28.4 –77.8) | 111 (IQR:83.7 –138) |
| Silberstein et al., 2015 (15) | USA | Retrospective | Paclitaxel + carboplatin + estramustine + RP vs. RP alone | 34 | 123 | 56.0 (IQR:51.0 - 61.0) | 60.8 (IQR:55.5 -64.9) | 157.2 | 147.6 |
| Sugino et al., 2023 (21) | Japan | Retrospective | Androgen deprivation + tegafur-Uracil + RP vs. RP alone | 101 | 126 | 70.0 (IQR:67.0 – 73.0) | 72.0 (IQR:68.0 –74.0) | 17.5 (IQR:6.2 –35.2) | 12.(IQR:5.0 –29.0) |
| Oishi et al., 2023 (22) | Japan | Retrospective | Estramustine+ androgen deprivation + RP vs. RP alone | 302 | 150 | 68 (IQR:65 - 72) | 70 IQR:(66 - 72) | 80 (IQR:56 – 98) | 41 (IQR:27 – 58) |
| Yokoyama et al., 2025 (20) | Japan | Retrospective | Androgen deprivation + tegafur-Uracil + RP vs. RP alone | 139 | 139 | 72 (IQR:69 – 75) | 71 (IQR:69 – 75) | 14.0 (IQR:6.0 –25.5) | 25.6 (IQR:11.3 –47.2) |
| Pan et al., 2019 (18) | China | Retrospective | Docetaxel + androgen deprivation + RP vs. RP alone | 60 | 44 | 65 (IQR:46 - 78) | 69 (IQR:57 - 78) | 12.5 | 22.8 |
This meta-analysis encompassed nine studies that assessed the impact of NCHT in conjunction with RP on BPFS. The NCHT+RP group demonstrated significantly improved BPFS versus RP alone in the overall analysis (RR = 0.60, 95% CI: 0.41– 0.87; P = 0.008; I2 = 90%) (Figure 2A). However, due to substantial heterogeneity and inclusion of observational studies, the RCT-only subgroup provides the most reliable estimate (RR = 0.73, 95% 95% CI: 0.62 - 0.86; P = 0.0002; I2 = 0%) (Figure 2B). The non-RCTs showed benefit persisted but with high heterogeneity (RR = 0.53, 95% CI: 0.29 – 0.97; P = 0.04; I2 = 93%) (Figure 2C).
Figure 2.
Forest plots for BPFS. (A) Overall analysis of NCHT+RP vs. RP alone. (B) Subgroup analysis of RCTs. (C) Subgroup analysis of non-randomized studies.
We conducted an analysis of overall survival (OS). For overall survival (OS), the RCT subgroup did not demonstrate a statistically significant benefit with NCHT+RP (RR = 0.43, 95% CI: 0.10–1.88; P = 0.26; I² = 53%) (Figure 3C), based on three RCTs with limited events and low statistical power. In contrast, the non-randomized studies showed a significant improvement (RR = 0.55, 95% CI: 0.31–0.99; P = 0.04; I² = 36%) (Figure 3B), but these estimates are susceptible to selection bias. The overall pooled analysis, combining both designs, yielded a nominally significant result (RR = 0.45, 95% CI: 0.24–0.84; P = 0.01; I² = 61%) (Figure 3A); however, this finding is driven entirely by the non-randomized studies and should be interpreted with caution given the lack of confirmation in RCTs and the methodological limitations of pooling disparate designs.
Figure 3.
Forest plots for OS. (A) Overall analysis of NCHT+RP vs. RP alone. (B) Subgroup analysis of non-randomized studies. (C) Subgroup analysis of RCTs.
For positive surgical margin (PSM) rates, the RCT subgroup confirmed a significant reduction with NCHT+RP (RR = 0.52, 95% CI: 0.32–0.84; P = 0.007; I² = 79%) (Figure 4B). The non-randomized studies also showed a significant benefit (RR = 0.45, 95% CI: 0.35–0.58; P < 0.00001; I² = 47%) (Figure 4C), with lower heterogeneity. The overall pooled analysis, consistent with the subgroups, demonstrated a significant reduction in PSM (RR = 0.49, 95% CI: 0.37–0.65; P < 0.00001; I² = 65%) (Figure 4A), although moderate heterogeneity remained, largely attributable to differences in study populations and chemotherapy regimens.
Figure 4.
Forest plots for PSM rates. (A) Overall analysis. (B) Subgroup analysis of RCTs. (C) Subgroup analysis of non-randomized studies.
For metastasis-free survival (MFS), the single large RCT (Eastham et al.) reported a nominally significant benefit in an exploratory analysis favoring NCHT+RP (RR = 0.78, 95% CI: 0.63–0.98; P = 0.03) (Figure 5B). In contrast, the two non-randomized studies showed no significant difference (RR = 0.78, 95% CI: 0.19–3.15; P = 0.73) with high heterogeneity (I² = 87%) (Figure 5C). The overall pooled analysis, combining all three studies, did not reveal a significant benefit (RR = 0.82, 95% CI: 0.45–1.49; P = 0.51; I² = 77%) (Figure 5A), largely due to the imprecision and inconsistency contributed by the observational data. Exploratory regional subgroup analysis suggested a significant benefit in Japanese cohort studies (RR = 0.37, 95% CI: 0.16–0.86; P = 0.02) (Figure 5D) but not in US studies (RR = 1.04, 95% CI: 0.55–1.96; P = 0.91) (Figure 5E); however, these analyses are based on single studies and should be interpreted with caution.
Figure 5.
Forest plots for MFS. (A) Overall analysis. (B) Subgroup analysis of the RCT by Eastham et al. (C) Subgroup analysis of non-RCTs. (D) Subgroup analysis of Japanese studies. (E) Subgroup analysis of US studies.
Sensitivity analysis
To evaluate how bias risk affects the overall findings, we conducted sensitivity analyses (Figure 6). The Baujat plot identified Pan et al., 2019 and Silberstein et al., 2015 as primary contributors to overall heterogeneity. The studies with minimal impact on heterogeneity (Eastham et al., 2020, Yokoyama et al., 2025, Narita et al., 2018) clustered near the origin, indicating robust alignment with the pooled estimate. Notably, Oishi et al., 2023 contributed moderately to heterogeneity.
Figure 6.
Baujat plot for sensitivity analysis, identifying studies contributing to heterogeneity.
Quality assessment and publication bias
Among the RCTs, the risk of bias assessment using the ROB 2 tool ranged from “Low” (Berkut et al., Qu et al., Narita et al.) to “High” (Eastham et al.) (Figure 7A). The methodological quality of the non-randomized studies, assessed by the NOS, varied (scores 3–9) (Table 3). The study by Oishi et al. (score=3) was identified as a major source of heterogeneity. Funnel plots indicated the absence of significant publication bias (Figure 7B). The trim-and-fill confirmed robust results, reinforcing the reliability of our results (Figure 7C).
Figure 7.
Risk of bias and publication bias assessment. (A) Risk of bias summary for included RCTs (Cochrane ROB 2 tool). (B) Funnel plot for assessment of publication bias. (C) Funnel plot after trim-and-fill analysis.
GRADE evidence quality assessment
We used the GRADE (Grading of Recommendations Assessment, Development and Evaluation) framework to assess the certainty of evidence for each primary outcome (Table 5). For BPFS and PSM rate, the evidence was rated as Low certainty, downgraded due to serious risk of bias (inclusion of observational studies with potential confounding) and serious inconsistency (substantial statistical heterogeneity). For OS and MFS, the evidence was rated as Very Low certainty, further downgraded for imprecision (wide confidence intervals crossing the line of no effect, and few events in RCTs) and indirectness (variations in chemotherapy regimens and patient populations). This very low certainty indicates that the true effect of NCHT on survival outcomes is likely to be substantially different from the estimated effect.
Table 5.
Summary of findings table with GRADE ratings for critical outcomes.
| Outcomes | GRADE rating | Explanation |
|---|---|---|
| Biochemical progression-free survival (BPFS) | Low | Downgraded due to serious risk of bias (inclusion of observational studies with potential confounding) and serious inconsistency (substantial statistical heterogeneity, I² = 90%). |
| PSM rate | Low | Downgraded due to serious risk of bias and serious inconsistency (moderate to high heterogeneity, I² = 65%). |
| OS | Very low | Downgraded due to serious risk of bias, serious inconsistency (I² = 61%), imprecision (wide confidence intervals crossing no effect, few events in RCTs), and indirectness (variations in chemotherapy regimens and patient populations). |
| MFS | Very low | Downgraded due to serious risk of bias, serious inconsistency (I² = 77%), imprecision, and indirectness. The true effect is likely substantially different from the estimate. |
BPF, biochemical progression-free survival; PSM, positive surgical margin rate; OS, Overall survival; MFS, Metastasis-free survival.
Discussion
This meta-analysis demonstrates that NCHT combined with RP offers significant advantages over RP alone in improving BPFS and reducing PSM rates in patients with locally advanced prostate cancer. However, the benefits for MFS were not statistically significant overall, and the improvement in OS was observed primarily in the non-randomized study subgroup. Previous studies have indicated that NCHT may help improve BPFS, but controversy remains regarding its impact on OS and MFS (14). The findings of this meta-analysis provide a nuanced understanding of the efficacy of NCHT. Our results, demonstrating a significant improvement in BPFS and a reduction in PSM rates, are consistent with some previous systematic reviews (24–26), but contrast with others that found no overall survival benefit (29). This updated synthesis incorporates recent trials and highlights the ongoing uncertainty. This consistency underscores the robustness of the local disease control benefit offered by NCHT. The biological plausibility of this benefit is well-supported: the synergistic effect of androgen deprivation and chemotherapy potently induces tumor cell apoptosis and reduces tumor volume, facilitating a more complete surgical resection with negative margins (27). This mechanistic pathway directly translates into the observed reduction in biochemical recurrence, a well-established surrogate endpoint for local treatment efficacy (28).
An important methodological consideration is our decision to pool RCTs and non-randomized studies. While we conducted prespecified subgroup analyses by study design, the overall pooled estimates (especially for OS) are heavily influenced by observational data, which are susceptible to selection bias and confounding. This is clearly illustrated by the OS results: the overall significant effect disappears in the RCT subgroup. Therefore, the main conclusions of this meta-analysis should be based primarily on the RCT evidence, and the non-randomized results should be viewed as hypothesis-generating only.
Despite these positive signals, the largest RCT included in our analysis—CALGB 90203 (Eastham et al.)—reported that neoadjuvant docetaxel-based NCHT did not significantly improve its primary endpoint of 3-year biochemical progression-free survival, and only modest benefits were observed in secondary outcomes (14). This discrepancy highlights the limitations of docetaxel-based regimens and underscores the need for more potent neoadjuvant strategies. Indeed, the field has now shifted toward evaluating ARPI-based combinations, which have shown promising pCR rates in phase II trials (29). Our meta-analysis, therefore, represents a synthesis of the older generation of chemohormonal therapy; its conclusions should not be extrapolated to modern ARPI-containing regimens, which require dedicated randomized trials.
However, our results regarding survival outcomes introduce a critical layer of complexity to the existing literature. The significant improvement in OS in the overall analysis, which was primarily driven by non-randomized studies, stands in contrast to some earlier meta-analyses that found no OS benefit (30). This discrepancy may be attributable to the inclusion of more recent studies with longer follow-up durations or differing patient risk profiles. Importantly, while the RCT subgroup did not demonstrate a statistically significant OS improvement (RR = 0.43, 95% CI: 0.10–1.88; P = 0.26), the point estimate suggesting a substantial 57% reduction in mortality risk highlights a crucial limitation: the current body of RCT evidence is likely underpowered to detect a survival difference. This underscores the imperative for larger, well-designed randomized trials with sufficient statistical power.
Similarly, the neutral overall result for MFS (RR = 0.82, P = 0.51) contrasts with the hypothesis that effective local control should delay metastasis. This divergence from pre-clinical expectations could be explained by several factors. First, the heterogenous use and timing of salvage therapies upon biochemical recurrence across studies may confound the true MFS effect of NCHT (31). Second, it is possible that NCHT, while effective locally, may not adequately eradicate micrometastatic disease present at diagnosis in a subset of high-risk patients, necessitating more effective systemic adjuvant strategies (18, 20, 23). The striking regional heterogeneity we observed—with significant benefit in Japanese cohorts but none in US studies—further complicates the picture. This may not merely reflect differences in clinical practice but could also hint at underlying ethnic variations in tumor biology or pharmacogenomics affecting drug response.
The substantial heterogeneity observed, particularly for BPFS (I² = 90%), warrants careful consideration of its sources and impact on our conclusions. Our sensitivity analysis identified Pan et al. (2019) and Silberstein et al. (2015) as major contributors, likely due to differences in chemotherapy backbone, patient risk profiles, and follow-up duration. In contrast, studies with more homogeneous designs (e.g., Eastham et al., Yokoyama et al.) clustered near the origin in the Baujat plot, indicating greater consistency with the pooled estimate. Despite these exploratory analyses, residual heterogeneity remained unexplained, which contributed to the downgrading of evidence certainty to low or very low for all outcomes in the GRADE assessment. This highlights the need for future studies with standardized protocols and harmonized outcome definitions.
According to the GRADE framework, the certainty of evidence for BPFS and PSM was rated as low, primarily due to serious risk of bias and inconsistency. For OS and MFS, the evidence was rated as very low, further downgraded for imprecision and indirectness. These ratings reflect the limited confidence we can place in the pooled effect estimates and reinforce the exploratory nature of our findings. Clinicians should interpret these results with caution when considering NCHT in routine practice.
Moreover, the results of this study have significant implications for clinical practice, especially in formulating treatment plans for high-risk locally advanced PCa patients. The study suggests that NCHT may serve as an effective treatment option, improving patient prognosis and reducing the risk of biochemical recurrence. These findings not only highlight the potential application value of NCHT in radical treatment but also provide a basis for developing more personalized treatment strategies in the future. However, it is imperative to note that current clinical guidelines (e.g., NCCN) do not recommend neoadjuvant ADT or chemotherapy prior to RP outside of clinical trials. The use of NCHT should remain investigational, and patients should be counseled about the availability of ongoing trials with novel ARPI-based therapies.
Several limitations of this study should be acknowledged. First, methodological constraints include the pooling of RCTs and observational studies, which increases heterogeneity and potential bias. Second, the use of risk ratios rather than hazard ratios for time-to-event outcomes is a significant limitation. Because we used event counts at varying follow-up times, our analyses do not fully incorporate the timing of events, censoring, or differences in follow-up duration between studies. This could bias the effect estimates and reduces the reliability of our survival conclusions. Third, substantial heterogeneity was observed, particularly in the BPFS analysis, likely stemming from variations in study design, patient characteristics, and treatment protocols. Fourth, the number of RCTs was relatively small (n=4) (14, 16, 17, 23), which may limit the robustness and generalizability of the findings. Potential unmeasured confounding factors in the non-randomized studies could also influence the outcomes. Additionally, the geographical concentration of studies (mainly Japan and the US) may limit the broad applicability of our results. Finally, while we performed tests for publication bias, the limited number of included studies (n=10) renders funnel plots and statistical tests such as Egger’s test underpowered to reliably detect bias, which is an inherent limitation of our review.
Conclusions
In summary, this meta-analysis suggests that for locally advanced PCa, NCHT (docetaxel-based chemohormonal therapy) combined with RP may improve BPFS and reduce PSM rates compared to RP alone, although the certainty of this evidence is low. The apparent OS benefit was not robust in RCT subgroup analyses and may be influenced by confounding in observational studies. According to GRADE, the evidence for survival outcomes is very low certainty, indicating that the true effects may differ substantially from the estimates. Given the modest benefits observed and the evolution of neoadjuvant therapy towards more effective ARPI-based regimens, these findings should be considered hypothesis-generating and primarily of historical context. Well-powered, methodologically rigorous randomized controlled trials evaluating modern combination therapies (e.g., ARPI+ADT) with standardized protocols and long-term follow-up are urgently needed to redefine the role of neoadjuvant treatment in this setting.
Funding Statement
The author(s) declared that financial support was received for this work and/or its publication. This research was funded by the Medical Research Project of Sichuan Medical Association (S2024047).
Footnotes
Edited by: Rui Miguel Gil Da Costa, Federal University of Maranhão, Brazil
Reviewed by: Zeyad Schwen, Cleveland Clinic, United States
Kumar Madhavan, All India Institute of Medical Sciences, Bhopal, India
Author contributions
HL: Conceptualization, Methodology, Writing – original draft, Writing – review & editing, Funding acquisition. XW: Data curation, Formal analysis, Investigation, Writing – original draft. CH: Software, Validation, Visualization, Writing – review & editing.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fonc.2026.1693222/full#supplementary-material
References
- 1. Chien TM, Lu YM, Geng JH, Huang TY, Ke HL, Huang CN, et al. Predictors of positive bone metastasis in newly diagnosed prostate cancer patients. Asian Pac J Cancer Prev. (2016) 17:1187–91. doi: 10.7314/apjcp.2016.17.3.1187. PMID: [DOI] [PubMed] [Google Scholar]
- 2. Voerman B, Fischer M, Visser A, Garssen B, van Andel G, Bensing J. Health-related quality of life in Dutch men with prostate cancer. J Psychosoc Oncol. (2006) 24:49–64. doi: 10.1300/J077v24n02_04. PMID: [DOI] [PubMed] [Google Scholar]
- 3. Berenguer CV, Pereira F, Câmara JS, Pereira JAM. Underlying features of prostate cancer-statistics, risk factors, and emerging methods for its diagnosis. Curr Oncol. (2023) 30:2300–21. doi: 10.3390/curroncol30020178. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Smith-Palmer J, Takizawa C, Valentine W. Literature review of the burden of prostate cancer in Germany, France, the United Kingdom and Canada. BMC Urol. (2019) 19:19. doi: 10.1186/s12894-019-0448-6. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Sandler HM, Mirhadi AJ. Radical radiotherapy for prostate cancer is the 'only way to go'. Oncol (Williston Park). (2009) 23:840–3. [PubMed] [Google Scholar]
- 6. Attard G, Murphy L, Clarke NW, Cross W, Jones RJ, Parker CC, et al. Abiraterone acetate and prednisolone with or without enzalutamide for high-risk non-metastatic prostate cancer: a meta-analysis of primary results from two randomised controlled phase 3 trials of the STAMPEDE platform protocol. Lancet. (2022) 399:447–60. doi: 10.1016/S0140-6736(21)02437-5. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. EK, Eastham JA. Chemotherapy and novel therapeutics before radical prostatectomy for high-risk clinically localized prostate cancer. Urol Oncol. (2015) 33:217–25. doi: 10.1016/j.urolonc.2014.11.020. PMID: [DOI] [PubMed] [Google Scholar]
- 8. Bhindi B, Lokeshwar SD, Klaassen Z, Klotz L, Wallis CJD. Systematic review and meta-analysis of trials evaluating the role of adjuvant radiation after radical prostatectomy for prostate cancer: implications for early salvage. Can Urol Assoc J. (2020) 14:330–6. doi: 10.5489/cuaj.6440. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Liang Z, Lin S, Lai H, Li L, Wu J, Zhang H, et al. Efficacy and safety of salvage radiotherapy combined with endocrine therapy in patients with biochemical recurrence after radical prostatectomy: a systematic review and meta-analysis of randomized controlled trials. Front Oncol. (2022) 12:1093759. doi: 10.3389/fonc.2022.1093759. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Liu W, Yao Y, Liu X, Liu Y, Zhang GM. Neoadjuvant hormone therapy for patients with high-risk prostate cancer: a systematic review and meta-analysis. Asian J Androl. (2021) 23:429–36. doi: 10.4103/aja.aja_96_20. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, et al. The Prisma statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. Bmj. (2009) 339:b2700. doi: 10.1136/bmj.b2700. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Higgins JP, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. Bmj. (2011) 343:d5928. doi: 10.1136/bmj.d5928. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol. (2010) 25:603–5. doi: 10.1007/s10654-010-9491-z. PMID: [DOI] [PubMed] [Google Scholar]
- 14. Eastham JA, Heller G, Halabi S, Monk JP, Beltran H, Gleave M, et al. Cancer and Leukemia Group B 90203 (Alliance): radical prostatectomy with or without neoadjuvant chemohormonal therapy in localized, high-risk prostate cancer. J Clin Oncol. (2020) 38:3042–50. doi: 10.1200/JCO.20.00315. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Silberstein JL, Poon SA, Sjoberg DD, Maschino AC, Vickers AJ, Bernie A, et al. Long term oncologic outcomes of a Phase II trial of neoadjuvant chemohormonal therapy followed by radical prostatectomy for patients with clinically localized, high-risk prostate cancer. BJU Int. (2015) 116:50–6. doi: 10.1111/bju.12676. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Berkut MV, Artemjeva AS, Reva SA, Tolmachev SS, Petrov SB, Nosov AK. Oncological results of neoadjuvant chemohormonal therapy in patients with high and very high-risk prostate cancer. Cancer Urol. (2020) 16:54–63. doi: 10.17650/1726-9776-2020-16-1-54-63 [DOI] [Google Scholar]
- 17. Qu F, Gu Y, Xue M, He M, Zhou F, Wang G, et al. Impact of therapy on cancer metabolism in high‐risk localized prostate cancer treated with neoadjuvant docetaxel and androgen deprivation therapy. Prostate. (2021) 81:560–71. doi: 10.1002/pros.24134. PMID: [DOI] [PubMed] [Google Scholar]
- 18. Pan J, Chi C, Qian H, Zhu Y, Shao X, Shao J, et al. Neoadjuvant chemohormonal therapy combined with radical prostatectomy and extended PLND for very high risk locally advanced prostate cancer: a retrospective comparative study. Urol Oncol. (2019) 37:991–8. doi: 10.1016/j.urolonc.2019.07.009. PMID: [DOI] [PubMed] [Google Scholar]
- 19. Fujita N, Koie T, Ohyama C, Tanaka Y, Soma O, Matsumoto T, et al. Overall survival of high-risk prostate cancer patients who received neoadjuvant chemohormonal therapy followed by radical prostatectomy at a single institution. Int J Clin Oncol. (2017) 22:1087–93. doi: 10.1007/s10147-017-1160-8. PMID: [DOI] [PubMed] [Google Scholar]
- 20. Yokoyama Y, Kawase M, Ebara S, Tatenuma T, Sasaki T, Ikehata Y, et al. Efficacy and safety of neoadjuvant chemohormonal therapy for high-risk prostate cancer treated with robot-assisted laparoscopic radical prostatectomy: a propensity score-matched analysis (the MSUG94 group). Int Urol Nephrol. (2025) 57:809–16. doi: 10.1007/s11255-024-04268-2. PMID: [DOI] [PubMed] [Google Scholar]
- 21. Sugino F, Nakane K, Kawase M, Ueda S, Tomioka M, Takeuchi Y, et al. Efficacy and safety of neoadjuvant luteinizing hormone-releasing hormone antagonist and tegafur-uracil chemohormonal therapy for high-risk prostate cancer. Life (Basel). (2023) 13:1072. doi: 10.3390/life13051072. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Oishi T, Hatakeyama S, Tabata R, Fujimori D, Fukuda M, Shinozaki T, et al. Comparison of neoadjuvant chemohormonal therapy vs. extended pelvic lymph-node dissection in high-risk prostate cancer treated with robot-assisted radical prostatectomy. Sci Rep. (2023) 13:3436. doi: 10.1038/s41598-023-30627-7. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Narita S, Nara T, Kanda S, Numakura K, Saito M, Inoue T, et al. Radical prostatectomy with and without neoadjuvant chemohormonal pretreatment for high-risk localized prostate cancer: a comparative propensity score matched analysis. Clin Genitourin Cancer. (2019) 17:e113–22. doi: 10.1016/j.clgc.2018.09.019. PMID: [DOI] [PubMed] [Google Scholar]
- 24. Bolla M, van Poppel H, Collette L, van Cangh P, Vekemans K, Da Pozzo L, et al. Postoperative radiotherapy after radical prostatectomy: a randomised controlled trial (EORTC trial 22911). Lancet. (2005) 366:572–8. doi: 10.1016/S0140-6736(05)67101-2. PMID: [DOI] [PubMed] [Google Scholar]
- 25. Collette L, van Poppel H, Bolla M, van Cangh P, Vekemans K, Da Pozzo L, et al. Patients at high risk of progression after radical prostatectomy: do they all benefit from immediate post-operative irradiation? (EORTC trial 22911). Eur J Cancer. (2005) 41:2662–72. doi: 10.1016/j.ejca.2005.06.024. PMID: [DOI] [PubMed] [Google Scholar]
- 26. Hudson E, Kynaston H, Varma M, Carter A, Staffurth J, Barber J, et al. Radiotherapy after radical prostatectomy for adenocarcinoma of the prostate: a UK institutional experience and review of published studies. Clin Oncol (R Coll Radiol). (2008) 20:353–7. doi: 10.1016/j.clon.2008.02.008. PMID: [DOI] [PubMed] [Google Scholar]
- 27. Santer FR, Erb HH, Oh SJ, Handle F, Feiersinger GE, Luef B, et al. Mechanistic rationale for MCL1 inhibition during androgen deprivation therapy. Oncotarget. (2015) 6:6105–22. doi: 10.18632/oncotarget.3368. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Bria E, Cuppone F, Giannarelli D, Milella M, Ruggeri EM, Sperduti I, et al. Does hormone treatment added to radiotherapy improve outcome in locally advanced prostate cancer?: meta-analysis of randomized trials. Cancer. (2009) 115:3446–56. doi: 10.1002/cncr.24392. PMID: [DOI] [PubMed] [Google Scholar]
- 29. Jiang J, Wang Y, Bai J, Yang G, Wang H, Ding H, et al. Efficacy and safety of triple or dual therapies for metastatic hormone-sensitive prostate cancer: a systematic review and Bayesian network meta-analysis. Future Oncol. (2026) 22:567–82. doi: 10.2217/FON-2022-1114. PMID: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Petrelli F, Vavassori I, Coinu A, Borgonovo K, Sarti E, Barni S. Radical prostatectomy or radiotherapy in high-risk prostate cancer: a systematic review and metaanalysis. Clin Genitourin Cancer. (2014) 12:215–24. doi: 10.1016/j.clgc.2014.01.010. PMID: [DOI] [PubMed] [Google Scholar]
- 31. Yang YJ, Lee YS, Tae JH, Choi J, Kim JH, Yang EJ, et al. Salvage lymphadenectomy or radiation therapy in prostate cancer patients with biochemical recurrence and PET positive lymph nodes after radical prostatectomy: a systematic review and pooled analysis. Eur J Surg Oncol. (2024) 50:108704. doi: 10.1016/j.ejso.2024.108704. PMID: [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.