In 2025, the field of gynecologic oncology pathology in China witnessed in-depth innovation and a leap toward precision. Molecular subtyping has fully moved from an academic hotspot to routine clinical practice, the integrated diagnosis system has been continuously improved, and AI-assisted technologies have accelerated their penetration. Key advances, including the exploration of multiple molecular subtypes, breakthroughs in research on cervical cancer unrelated to HPV infection, and panoramic detection of immune and targeted therapy targets, have provided an entirely new pathological foundation for the diagnosis and treatment of gynecologic malignancies. Gynecology and Obstetrics Network specially invited Professor Liu Congrong from the Department of Pathology, Peking University Third Hospital, to sort out the annual progress and future directions from three dimensions: subtype translation, multidisciplinary collaboration, and technological empowerment, so as to provide references for clinical diagnosis and treatment.
I. Clinical Implementation of Molecular Subtyping: A Substantive Breakthrough from "Classification" to "Precision Stratification"
In 2025, the molecular subtyping of endometrial cancer and cervical cancer has moved beyond a mere academic label and entered the practical stage of "precision stratification to guide treatment". The core advances are reflected in the following aspects.
1. Endometrial Cancer: Deepening and Expansion of the Multidimensional Subtyping System
Clinical management of multiple molecular subtypes: For complex subtype cases with concurrent molecular features such as POLE mutation, mismatch repair deficiency (MMRd), and TP53 aberration, targeted research directions have been formed in clinical practice. Interpretation is no longer limited to a single molecular marker, but instead focuses on the impact of the combination mode of different molecular features on prognosis and treatment response.
Precision re-stratification of the NSMP subtype: The No Specific Molecular Profile (NSMP) subtype accounts for 60% of endometrial cancers and 70% of endometrioid carcinomas, and its "homogeneous label" has been broken. In 2025, the industry has focused on in-depth exploration within this subtype, attempting to identify sublayer features that can guide treatment by integrating multidimensional indicators, thus changing its current status as a catch-all classification.
Highlighted core value of integrated subtyping: Molecular subtyping has not replaced morphology and immunohistochemistry (IHC), but has formed an integrated diagnostic model of "molecular features + morphology + immune markers". Clinical practice has confirmed that the organic integration of the three can significantly improve the accuracy of diagnosis and the value of treatment guidance, avoiding misjudgment caused by a single indicator.
Quantitative analysis of the immune microenvironment: Multiplex immunofluorescence technology has become an important supplement to molecular subtyping. By analyzing the distribution, functional status of immune cells and their interaction mode with tumor cells in the tumor immune microenvironment, it provides a new dimension for prognostic assessment. The prognosis of some cases is not only determined by the genetic characteristics of tumor cells, but the hot/cold phenotypes of the immune microenvironment are also critical.
2. Cervical Cancer: Breakthroughs in Etiological Subtyping and Exploration of New Subtypes
Systematic research on HPV-unrelated cervical adenocarcinoma: Previously, 80% of cervical adenocarcinomas were HPV-related. In 2025, significant progress has been made in the research on HPV-unrelated cervical adenocarcinoma (represented by gastric-type adenocarcinoma), including the initial establishment of prognostic stratification criteria and clear identification of diagnostic pitfalls. Relevant research results have been gradually applied in clinical practice to guide precise intervention for high-risk subtypes.
Updated understanding of HPV-unrelated cervical squamous cell carcinoma: The traditional view holds that almost all cervical squamous cell carcinomas are associated with HPV infection. However, the latest 2025 data show that 5%-10% (up to 12% in some studies) of cervical squamous cell carcinomas have no detectable HPV viral genetic material or related traces through sophisticated testing. The pathogenesis, prognostic characteristics and treatment response patterns of this subtype may differ significantly from those of HPV-related subtypes, and it has become a key focus of both pathological and clinical research.
II. Empowerment via Multidisciplinary Collaboration: The Transformation of Pathologists from "Diagnosticians" to "Treatment Navigators"
Faced with the complexity of immunotherapy and targeted drugs, the core trend from 2025 to 2026 is to promote the deep participation of pathologists in treatment decision-making through clinical-pathology-oncology Multidisciplinary Team (MDT) collaboration. The transformation path and core practices are as follows.
1. Establishment of an Integrated Diagnosis System: Breaking the Limitations of a Single Diagnostic Dimension
The diagnosis of difficult cases no longer relies on isolated morphological or IHC results, but integrates multidimensional information including "clinical surgical findings + gross observation + HE microscopic features + IHC profile expression + genetic mutation spectrum". This requires pathologists to have a richer knowledge reserve and cross-domain perspective, and to achieve precise characterization of tumors through core indicators such as driver mutation events.
2. Diversified Expansion of Immunotherapy Target Detection
PD-L1 detection is no longer the sole basis for guiding immunotherapy. In 2025, the clinical community has recognized the synergistic value of multiple biomarkers: MMRd molecular phenotype, high Tumor Mutational Burden (TMB), hot immune microenvironment indicated by multiplex immunofluorescence, and SMARCA4 driver mutations (common in dedifferentiated/undifferentiated endometrial cancer and ovarian small cell carcinoma of the hypercalcemic type) have all been proven to be associated with immunotherapy sensitivity. Pathologists need to provide a comprehensive basis for clinical immunotherapy decision-making through systematic detection.
3. Pathological Support for Targeted Therapy and Antibody-Drug Conjugates
In 2025, Antibody-Drug Conjugates (ADCs) targeting HER2, Claudin-6, and folate receptors have achieved significant efficacy in the treatment of gynecologic malignancies, becoming an important breakthrough in targeted therapy. Pathologists need to accurately detect the expression status of relevant targets to provide core references for the clinical application of ADCs. Meanwhile, they should participate in pathological analysis related to efficacy evaluation and adverse reactions, forming a closed-loop support of "detection - treatment - evaluation".
III. Amid the Wave of AI Technology: Upholding the Core Value and Upgrading the Capabilities of Pathologists
With the popularization of AI-assisted diagnosis and digital pathology, the reduction of repetitive work has become an industry trend, and discussions on the role positioning of AI and the core competitiveness of pathologists have become the focus. The key consensus is as follows.
1. Core Positioning of AI: An Efficient "Assistant" Rather Than a "Competitor"
The advantages of AI focus on the precise execution of stereotyped, single tasks, such as lymphocyte counting, cell spacing measurement, and stratified counting of the Ki-67 index. Its computing power and pixel-level recognition ability far exceed those of humans, which can significantly improve work efficiency and data accuracy. However, the limitations of AI are also clear: it lacks the ability to comprehensively interpret clinical scenarios. For example, AI cannot independently judge the differences in clinical significance of a high Ki-67 index in the late proliferative phase of the endometrium, suspected STIC (serous tubal intraepithelial carcinoma) mucosa, and high-grade HGSC (high-grade serous carcinoma). In addition, AI does not possess the "wisdom" of cross-domain knowledge integration and assimilation, and can only perform tasks within existing frameworks.
2. The Essence of Clinical Diagnostic Intuition: Externalization of Tacit Knowledge and Humanistic Temperature
Clinical diagnostic intuition is not "subjective speculation", but the rapid externalization of tacit knowledge accumulated by pathologists over a long period of time, including massive case experience and cross-disciplinary cognition. It is a high integration of logical reasoning and experience precipitation. AI cannot replicate this core capability, nor can it replace the "humanistic temperature" of medicine. Clinical guidelines are only applicable to 80%-90% of routine cases. For 10%-20% of special cases (such as the fertility preservation demands of young patients with endometrial cancer), the optimal plan needs to be formulated through doctor-patient interaction by combining personalized factors such as the patient's age, economic status, and treatment compliance. This process is inseparable from human fraternity and empathy, a gap that AI will never cross.
3. Directions for Capacity Upgrading of Pathologists
Faced with the impact of AI, pathologists should avoid becoming "cookbook atlas doctors", but transform into "AI trainers + clinical decision-makers + innovation leaders". On the one hand, they can use AI to improve work efficiency and focus their energy on core links such as the diagnosis of complex cases, participation in treatment planning, and knowledge innovation. On the other hand, they need to strengthen their creativity and participate in the update of disease classification standards (pathological and disease classifications are significantly adjusted every 2-10 years, and AI relies on humans to redefine frameworks for training), expanding the boundaries of medical knowledge through continuous exploration. At the same time, they should uphold humanistic care and integrate the temperature of the medical core into the entire process of diagnosis and treatment, which is the irreplaceable core value of pathologists.
Summary and Outlook
The field of gynecologic oncology pathology has achieved a key leap from "academic molecular subtyping" to "clinical precision treatment". Multidisciplinary collaboration and the integration of AI technology will become the dual engines driving the development of the industry. The role of pathologists has also transformed from a mere "diagnostic judge" to a "navigator" throughout the entire diagnosis and treatment process. Their core competitiveness is no longer the ability to perform repetitive work, but the comprehensive quality of integrated analysis, innovation leadership and humanistic care. In the future, with the continuous progress of technology and the constant update of concepts, "precision + temperature + innovation" will become the core development direction of the gynecologic oncology pathology field, bringing better diagnosis and treatment experience and survival benefits to patients.
Expert Profile
Professor Liu Congrong
Affiliation: Peking University Third HospitalTitles: Director of the Department of Pathology, Peking University Third Hospital; Deputy Director of the Department of Pathology, Peking University; Director of the Peking University Pathology Center; Deputy Director of the Oncology Discipline Group, Peking University Health Science Center; Chief Physician, Professor, Doctoral Supervisor
Academic Positions:
Chairperson, Pathology Professional Committee, Chinese Maternal and Child Health Association
Chairperson, Pathology Professional Committee, Chinese Medical Women's Association
Member, 14th Pathology Committee, Chinese Medical Association; Leader, Female Reproductive Group
Member, Pathology Committee, Chinese Medical Doctor Association
Standing Committee Member, Pathology Branch, Beijing Medical Association; Leader, Women and Children Group
Member, International Society of Gynecological Pathologists (ISGyp)
Editorial Board Member, 5th and 6th WHO Classification of Tumours of Female Genital Organs; Associate Editor-in-Chief, Chinese Journal of Diagnostic Pathology; Associate Editor-in-Chief, Chinese Edition of International Journal of Gynecological Pathology; Editorial Board Member, Chinese Journal of Clinical Obstetrics and Gynecology and Progress in Modern Gynecology and Obstetrics
Research Focus and Achievements:Professor Liu Congrong is mainly engaged in the diagnostic and basic research of gynecological pathology. She has presided over the National Key R&D Program, three General Programs of the National Natural Science Foundation of China, the Beijing Natural Science Foundation, and the Startup Foundation for Returned Overseas Chinese Scholars. She has participated in a number of large-scale national and municipal scientific research projects. She has published more than 130 papers in domestic and foreign journals, including Protein & Cell, Modern Pathology, The Journal of Pathology and Gynecologic Oncology, and has compiled and translated 23 professional works. She was awarded the Top 5% Highly Cited Scholar by CNKI in 2024.
Editor-in-Charge: Lily