Editor's Note: High-risk and locally advanced prostate cancer is characterized by substantial biological heterogeneity, with significant differences in recurrence risk, metastatic potential, and treatment response among patients, making it a major challenge in clinical practice. With the development of novel androgen receptor pathway inhibitors, PSMA PET/CT, multi-omics technologies, and artificial intelligence, overcoming the limitations of traditional risk stratification, accurately identifying patients at truly extremely high risk, and integrating local and systemic therapies to improve long-term outcomes have become important areas of research. Oncology Frontier · UroStream invited Professor Mo Zhang of the Department of Urology, First Hospital of China Medical University, for an in-depth interview on the institution's single-center experience, precision stratification strategies, and future research directions in high-risk and locally advanced prostate cancer.Exploring the Value of Neoadjuvant Therapy to Achieve Better Local Control in High-Risk Patients
Oncology Frontier · UroStream: Thank you very much, Professor Zhang, for accepting our interview and for sharing data from the First Hospital of China Medical University on precision diagnosis and treatment of high-risk and locally advanced prostate cancer. Based on your institution’s clinical cohort, what important findings have emerged regarding the diagnosis and treatment of these patients?
Professor Mo Zhang: Thank you to Oncology Frontier · UroStream for your interest in our research. At this conference, I mainly shared the single-center experience of the First Hospital of China Medical University in the diagnosis and treatment of high-risk and locally advanced prostate cancer.
One important area of our work has been exploring the value of neoadjuvant endocrine therapy in these patients. A considerable proportion of patients received intensified treatment incorporating novel androgen receptor pathway inhibitors (ARPIs). After 3–6 months of preoperative treatment, we observed a marked decline in PSA levels in most patients. Some patients also experienced reductions in tumor burden and showed trends toward clinical or pathological downstaging, creating more favorable conditions for subsequent radical surgery.
Of particular interest was the positive surgical margin (PSM) rate. A positive surgical margin is an important adverse pathological factor following radical prostatectomy and is closely associated with an increased risk of biochemical recurrence. In our clinical cohort, the PSM rate decreased after neoadjuvant treatment, suggesting that this approach may have the potential to improve local tumor control.
However, these findings need to be interpreted objectively and cautiously. On the one hand, different studies have used different treatment durations, drug combinations, and definitions of pathological response. On the other hand, PSA reduction, tumor regression, or improved surgical margin status does not necessarily translate into long-term survival benefits. Previous studies have shown that neoadjuvant treatment with novel ARPIs can produce substantial PSA responses and a certain degree of pathological response, but the overall rate of pathological complete response remains limited. Whether these responses can consistently translate into improvements in metastasis-free survival or overall survival has long required validation through higher-level evidence.
The recently reported phase III PROTEUS study is particularly noteworthy and represents an important development in this field. The study showed that, in patients with high-risk localized or locally advanced prostate cancer undergoing radical prostatectomy, perioperative androgen deprivation therapy combined with apalutamide not only increased the proportion of patients achieving pathological complete response or minimal residual disease outcomes, but also improved metastasis-free survival. These findings provide higher-level evidence regarding whether perioperative systemic therapy can translate into long-term oncological benefit.
Of course, how these clinical trial findings should be translated into routine practice, which patients are most suitable for preoperative or perioperative treatment intensification, and how to determine the optimal drug, treatment duration, and integration with surgery will require further exploration based on guideline updates, long-term follow-up, and real-world evidence.
Integrating Clinical, Imaging, and Molecular Information for Further Stratification of High-Risk Patients
Oncology Frontier · UroStream: In clinical practice, how should clinicians integrate tumor risk stratification, imaging assessment, and pathological characteristics to develop more precise individualized treatment strategies for patients with high-risk and locally advanced prostate cancer?
Professor Mo Zhang: This is also an area that our team is currently focusing on. With the rapid development of digital pathology, radiomics, deep learning, and artificial intelligence, prognostic assessment and risk stratification for high-risk prostate cancer are gradually shifting from reliance on individual indicators toward multidimensional information integration.
The most prominent characteristic of high-risk prostate cancer is its substantial heterogeneity. Even among patients classified within the same “high-risk” category, disease burden and tumor biology can differ dramatically. For example, although patients with PSA levels slightly above 20 ng/mL and those with PSA levels exceeding 100 ng/mL may both be categorized as high risk under certain conventional criteria, their potential metastatic risk, treatment sensitivity, and long-term prognosis are likely to differ considerably. Therefore, high-risk patients should not simply be regarded as a biologically homogeneous population.
Traditional risk stratification mainly relies on PSA level, clinical T stage, ISUP Grade Group, and the number of positive biopsy cores. These parameters remain fundamental to clinical decision-making. However, for high-risk and locally advanced disease, traditional parameters alone often cannot fully reflect the tumor’s true biological characteristics.
Our research strategy is to build upon conventional clinicopathological indicators by further integrating imaging information from multiparametric MRI and PSMA PET/CT, digital pathology and pathomics features, and tumor-associated molecular biomarkers such as PSCA, gradually establishing a multimodal, high-dimensional data framework. By applying artificial intelligence and machine-learning methods to analyze relationships across these different dimensions, we hope to achieve more refined internal stratification of high-risk patients.
It is important to emphasize that the value of artificial intelligence is not to replace physicians in making decisions. Rather, it can help clinicians extract features from complex and large-scale datasets that may be difficult to identify through conventional assessment and establish more objective risk-prediction models. A truly clinically valuable model must not only perform well in a training cohort, but also undergo validation in independent cohorts, external centers, and ultimately prospective studies to demonstrate clear incremental value over existing clinical models.
Our goal is to further identify patients with truly clinically meaningful extremely high-risk disease from the broader population currently defined as having high-risk or locally advanced prostate cancer. For these patients, more comprehensive systemic evaluation, more aggressive multimodal treatment, and closer postoperative monitoring may be considered, with timely treatment adjustment when early signs of progression emerge. Conversely, patients whose recurrence risk is relatively low after comprehensive assessment should avoid unnecessary overtreatment.
Therefore, the core of precision stratification is not simply to increase the number of tests, but to integrate multidimensional information to answer three critical questions: How high is the patient’s true recurrence risk? Which treatment is the patient most likely to benefit from? And what intensity of treatment is appropriate?
Building a Multimodal Treatment System for High-Risk Prostate Cancer Based on the Characteristics of Chinese Patients
Oncology Frontier · UroStream: What research directions and potential breakthroughs do you consider most important for high-risk or locally advanced prostate cancer in the future?
Professor Mo Zhang: First, we need to recognize that the disease spectrum and stage at diagnosis among Chinese patients with prostate cancer differ to some extent from those observed in some Western countries. Prostate cancer screening is relatively well established in Europe and the United States, allowing many patients to be diagnosed at an earlier stage. In China, however, a considerable proportion of patients still present with high-risk, locally advanced, or even metastatic disease at initial diagnosis.
This characteristic is particularly prominent in some large referral centers in China, especially those in northern regions. At our center, for example, approximately 70% of the nearly 1,500 patients who underwent radical prostatectomy for prostate cancer over the past three years were classified as high risk. Developing an evidence-based treatment system for this large and highly heterogeneous patient population is an important challenge for Chinese urologists.
I believe four areas deserve particular attention in the future.
First, systematic integration of local and systemic therapies. High-risk and locally advanced prostate cancer generally cannot be managed with a single treatment modality. As robotic and laparoscopic techniques continue to mature, radical prostatectomy has become an important treatment option for carefully selected patients, but surgery represents only one component of multimodal treatment. For other patients, radiotherapy combined with long-term androgen deprivation therapy, with the addition of novel endocrine therapy when appropriate, is also an important curative-intent treatment pathway.
Therefore, the clinical question should not simply be “surgery or radiotherapy.” Instead, a comprehensive treatment plan should be developed according to the patient’s age, life expectancy, tumor burden, clinical stage, lymph-node status, comorbidities, and personal preferences. Current guidelines similarly emphasize multimodal management for high-risk localized and locally advanced prostate cancer, with radical surgery serving as one component of comprehensive treatment in appropriately selected patients.
For patients undergoing radical prostatectomy, postoperative management should also take into account final pathological stage, ISUP Grade Group, surgical margin and lymph-node status, as well as whether PSA remains detectable and the PSA doubling time. These factors should be considered together when determining whether to pursue close monitoring, adjuvant therapy, or early salvage treatment. A single adverse pathological factor should not automatically lead to the same postoperative treatment strategy for every patient.
Second, further expansion of advanced imaging technologies. Multiparametric MRI can provide important information regarding the local extent of the tumor, extraprostatic extension, seminal vesicle invasion, and the risk of neurovascular bundle involvement. PSMA PET/CT improves the detection of lymph-node and distant metastases and can identify low-volume lesions that may be missed by conventional CT and bone scans.
The proPSMA study showed that PSMA PET/CT had significantly greater accuracy than conventional imaging for initial staging of high-risk prostate cancer and could influence clinical treatment planning. Current European Association of Urology guidelines recommend PSMA PET/CT for staging high-risk localized and locally advanced prostate cancer when available.
However, detecting more small-volume lesions on imaging does not automatically mean that all patients should receive intensified treatment. Future studies need to determine whether changing treatment strategies based on PSMA PET/CT restaging can actually improve metastasis-free survival and overall survival. The value of advanced imaging needs to progress from simply “seeing more clearly” to “guiding treatment and improving outcomes.”
Third, development of multimodal risk-stratification models. Future risk-assessment models should integrate clinical parameters, imaging characteristics, pathological information, and molecular biomarkers rather than relying on a single indicator. Through standardized database development, harmonized imaging and pathological data collection, and multicenter external validation, we may be able to develop predictive models that better reflect the characteristics of Chinese patients.
Such models should not only predict biochemical recurrence, but also address metastasis-free survival, cancer-specific survival, and patient-reported outcomes. Ultimately, they should guide treatment selection—for example, helping determine which patients are suitable for local therapy alone, which require perioperative treatment intensification, and which should receive early postoperative radiotherapy or systemic therapy.
Fourth, precision molecular classification and dynamic monitoring based on biomarkers. High-risk prostate cancer is not a single molecular entity. Some patients harbor alterations in homologous recombination repair (HRR) pathway genes, while others have PTEN loss, abnormalities in the PI3K–AKT pathway, or other highly aggressive molecular features. These alterations may be associated with recurrence risk and treatment sensitivity and could become important factors for selecting targeted or intensified therapies in the future.
In addition to tissue and genetic testing, liquid-biopsy technologies such as circulating tumor DNA (ctDNA) also warrant attention. In the future, ctDNA may help assess minimal residual disease after surgery and provide an early indication of recurrence risk through dynamic changes. However, ctDNA levels are generally low in patients with localized prostate cancer, and detection sensitivity, standardization, and clinically meaningful thresholds still require further investigation. Therefore, at present, ctDNA is better regarded as an important research tool and cannot replace standardized PSA monitoring and imaging assessment.
Overall, the diagnosis and treatment of high-risk and locally advanced prostate cancer are moving from conventional risk grouping toward refined stratification based on clinical, imaging, pathological, and molecular characteristics. Treatment strategies are likewise evolving from single-modality local treatment toward comprehensive management integrating local and systemic therapies.
The most meaningful breakthroughs in the future will not simply come from new drugs or imaging technologies. More importantly, we need to determine which patients require treatment intensification, when treatment should begin, and how treatment should be dynamically adjusted according to treatment response. Only by achieving precise matching between the patient, treatment, and timing can we truly improve long-term survival and quality of life for patients with high-risk prostate cancer.
Professor Mo Zhang
