From the follow-up database (N=1,917) of PCa patients in The Affiliated Bozhou Hospital of Anhui Medical University and the First Affiliated Hospital of Xinjiang Medical University, we identified 336 patients with aHSPC based on diagnostic, inclusion, and exclusion criteria.

Clinical and demographic characteristics: age, smoking history, alcohol history, hypertension, and diabetes.

Tumor characteristics: initial PSA, clinical tumor stage (TNM classification), Gleason score, and tumor load.

Testosterone indicators: measured at multiple time points following ADT initiation, average testosterone, and minimum testosterone.

Continuity of ADT: continuous androgen deprivation therapy (CADT) and intermittent androgen deprivation therapy (IADT).

Tumor progression: time to progression (TTP) and TTP to metastatic castration-resistant prostate cancer (mCRPC).

Survival metrics: overall survival (OS), defined as the interval from PCa diagnosis to death or last follow-up.

ADT agents: All patients received luteinizing hormone-releasing hormone (LHRH) agonists (goserelin or leuprolide), with therapeutic equivalence confirmed through clinical validation.

Testosterone measurement: Serum testosterone was quantified using chemiluminescence immunoassay (CLIA). Both institutional laboratories maintained ISO15189 accreditation and certification from the China National Accreditation Service (CNAS).

Advanced prostate cancer (aPCa): regionally or distantly metastatic PCa.

Advanced hormone-sensitive prostate cancer: metastatic PCa demonstrating therapeutic response to ADT.

Castration-resistant prostate cancer (CRPC): castration, serum testosterone T < 50 ng/dl, accompanied by biochemical (and) or imaging progression (10).

PCa progression: transition from aHSPC to CRPC, with TTP calculated from ADT initiation to CRPC diagnosis.

Testosterone stabilization without progression (TSP): sustained testosterone suppression, testosterone ≤20 ng/dl, sustained for >12 consecutive months.

The inclusion criteria were as follows: 1) newly diagnosed aHSPC meeting all conditions: no prior endocrine therapy and demonstrated an initial response to ADT; 2) treatment protocol adherence: CADT or IADT was initiated post-diagnosis; 3) monitoring compliance: three or more serum testosterone measurements during ≥12-month follow-up.

Patients were excluded based on the following criteria: 1) with other malignant tumors or combined severe cardiopulmonary diseases at baseline; 2) with incomplete clinical records; and 3) were lost to follow-up.

This study complied with the principles of the Declaration of Helsinki by examining the clinical information of retrospective patients. This study was approved by the Medical Ethics Committee of The People’s Hospital Bozhou (approval no. BYLS2024-147) and the Medical Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University (approval no. K202504-46). The studies were conducted in accordance with local legislation and institutional requirements. The Ethics Committee Review Board waived the requirement for written informed consent for participation from the participants or the participants’ legal guardians/next of kin. The study did not involve patients’ personal privacy and disease characteristics, and coding anonymized identifiable information, consistent with waiving of informed consent.

SPSS 26.0 and R 4.2.2 software were used for data processing and analysis. A t-test was used for comparisons between groups. Count data were described by rate, and the χ ² test was used for comparisons between groups. Grade data were described by rate, and a rank-sum test was used for comparisons between groups. Survival time was calculated from the start of treatment after diagnosis, and death or still alive status (cutoff date May 2025) was considered as the end data, in months. Based on the results of the univariate analysis, variables with a p-value less than 0.05 were selected for inclusion into the multivariate analysis model. To reduce the impact of potential confounders, propensity score matching was employed. The test level for statistical analysis was α = 0.05.

The mean age was 72.04 ± 8.93 years in the overall group, and the median overall TTP was 18 months. There were no statistically significant differences in the age, diabetes, hypertension, smoking history, and alcoholism history variables between the two groups (p > 0.05), and the baseline characteristics were consistent. The overall median survival was 6.17 years, with a median survival of 7.58 years in the testosterone sustained response group and 3.00 years in the testosterone non-sustained response group. For the survival rates, the testosterone sustained response group had a significant survival advantage over the testosterone non-sustained response group (p < 0.001) (see Table 1 ).

In the Cox regression model, it was found that the risk ratio (hazard ratio, HR) for the testosterone non-sustained response group was 2.21 (95%CI = 1.43–3.40), which indicated that the testosterone non-sustained response group had a 121% higher risk of death compared with the testosterone sustained response group. The difference in survival between the two groups was statistically significant (p < 0.001). The survival curves showed that the probability of survival was significantly higher in the testosterone sustained response group than that in the testosterone non-sustained response group (see Figure 4 ).

The multivariate Cox model for subgroup analysis was adjusted for baseline characteristics. In the subgroup analyses of tumor stage, tumor load, and Gleason score, there was no statistically significant difference between the testosterone sustained response and the testosterone non-sustained response group (p > 0.05). There was a significant difference between the testosterone sustained response and the testosterone non-sustained response group in the subgroup analysis of IADT and CADT (p = 0.014) (see Table 2 ).

The baseline characteristics were balanced by propensity score matching in the testosterone sustained response group and testosterone non-sustained response group. A notable reduction in the standardized mean differences after matching can be observed, indicating that the propensity score matching effectively improved the balance of the covariates between the two groups, which performed 1:2 optimal pair matching. The TTP was significantly longer in the testosterone sustained response group compared with the testosterone non-sustained response group (23.70 ± 14.66 vs. 13.68 ± 7.84 months, p < 0.001) (see Table 3 ).

In the Pearson’s correlation analysis, the testosterone minimum and the testosterone minimum time (r=0.245, p < 0.001), the testosterone minimum and TTP (r = −0.238, p < 0.001), the average testosterone and the testosterone minimum time (r=0.228, p < 0.001), and the average testosterone and TTP (r = −0.220, p < 0.001), the above two parameters were weakly correlated with each other. There was a very high positive correlation between testosterone minimum and average testosterone (r=0.844, p < 0.001) (see Table 4 ).

The multivariate analysis revealed significant associations between several clinical characteristics and the testosterone stabilization without progression outcome. PCa patients with visceral metastasis had a lower likelihood of testosterone stabilization without progression outcome compared to those with no visceral metastasis (adjusted OR = 0.45, 95%CI = 0.21–0.98, p = 0.043). Similarly, a higher tumor load was associated with reduced odds of testosterone stabilization without progression outcome (adjusted OR = 0.53, 95%CI = 0.33–0.85, p = 0.008) (see Table 5 ).

The multivariate analysis revealed several associations with mortality. Patients with diabetes exhibited a non-significantly higher risk of mortality compared to those without diabetes (adjusted HR = 1.406, 95%CI = 0.929–2.127, p = 0.107). In contrast, those with testosterone sustained response had a significantly lower mortality risk compared to those with non-sustained response (adjusted HR = 0.605, 95%CI = 0.369–0.990, p = 0.045). A higher minimum testosterone level was significantly associated with increased mortality (adjusted HR = 1.358, 95%CI = 1.116–1.654, p = 0.002), whereas average testosterone levels showed no significant association (adjusted HR = 0.910, 95%CI = 0.776–1.068, p = 0.248) (see Table 6 ).

Sustained testosterone suppression represents a paradigm shift in assessing the efficacy of ADT by emphasizing the cumulative duration of castrate testosterone levels (<20 ng/dl) rather than relying solely on nadir values as a single indicator. This approach recognizes that, although sustained testosterone suppression suggests a response to ADT, long-term castration may paradoxically accelerate hypogonadism and compound metabolic toxicity. The clinical utility of the sustained testosterone suppression framework lies in its capacity to synergize with PSA kinetics for the early identification of patients at high risk, enabling timely therapeutic intensification. By quantifying the cumulative testosterone exposure, this standard provides an evidence-based approach to ADT optimization, particularly valuable for metastatic HSPC patients with low tumor load concerned about hypogonadal effects. Prognostically, sustained testosterone suppression for a long time (12 months) is associated with a good prognosis (5-year survival, >60%); on the other hand, no-sustained testosterone suppression for a long time suggests early drug resistance, rapid progression to metastatic castration resistance, and the need for intensive therapy or combination therapy. Persistent testosterone fluctuations may reflect residual PCa or adrenal compensatory androgens, serving as early warning signs of impending castration resistance. However, the hormonal regulatory mechanisms of PCa and their interactions with the metabolic microenvironment require further elucidation. In particular, the interactions between the androgen–thyroid hormone signaling pathways and the biological links between prostatic fat volume and ADT responsiveness warrant further investigation (46, 47). These mechanisms often occur in complex clinical settings characterized by multiple hormonal imbalances, metabolic abnormalities, and concurrent drug therapies. A deeper understanding of these interaction mechanisms and elucidation of the crosstalk between these pathways may provide a theoretical basis for the development of combined PCa targeting strategies or personalized treatments.

The sustained testosterone suppression represents advancement in PCa management by systematically characterizing the dynamic relationship between longitudinal testosterone fluctuations and tumor progression. Our findings demonstrate that sustained testosterone suppression provides critical insights into ADT response assessment and prognostic prediction in advanced PCa. Importantly, this approach facilitates timely initiation of combination therapies (triplet or quadruplet regimens) at optimal dosages to achieve rapid, profound, and sustained testosterone suppression—a strategy associated with improved survival outcomes. By transcending the limitations of non-rigorous and non-serial testosterone monitoring, sustained testosterone suppression offers a refined tool for personalized therapeutic decision-making in advanced PCa management.

This investigation presents three key contributions to the management of aHSPC: firstly, we introduced the concept of sustained testosterone suppression. This paradigm shift advances testosterone management strategies by integrating longitudinal hormonal patterns rather than relying on isolated measurements. Secondly, while grounded in physiological principles and validated through real-world data, the model acknowledges inherent limitations in clinical translation. Practical constraints including the monitoring frequency, the sample size restrictions, and the heterogeneous treatment regimens (incorporating chemotherapy, radiotherapy, and targeted immunotherapy) may affect generalizability. However, these real-world conditions precisely enhance the clinical relevance of the model by demonstrating utility in complex treatment environments using accessible monitoring techniques. Thirdly, sustained testosterone suppression addresses critical gaps in current practice by quantifying the relationship between sustained castration duration and testosterone fluctuations, parameters previously overlooked despite their prognostic significance for tumor progression. Sustained testosterone suppression during ADT reflects therapeutic efficacy and great tumoral outcomes. In summary, sustained testosterone suppression is an advanced concept beyond the traditional binary (yes/no to castration) and represents a more refined and dynamic management and evaluation paradigm in ADT treatment for PCa, which is closely associated with improved long-term prognosis of patients with aHSPC.