Hao Chen

Ovarian Gynandroblastoma: A Rare Sex Cord-Stromal Tumor with Unique Morphological, Genetic, and Clinical Features—A Series of Two Patients with Clinicopathological and Molecular Analysis, and a Review of the Literature

Background Gynandroblastoma is an exceptionally rare ovarian sex cord-stromal tumor containing both male and female components. Due to its rarity, molecular data are limited, and its clinical behavior remains incompletely understood. Methods We retrospectively analyzed 2 tumors of gynandroblastoma diagnosed at 2 teaching hospitals. Clinicopathological data, including patient demographics, tumor characteristics, and outcomes, were collected. Immunohistochemistry (IHC) and next-generation sequencing (NGS) were performed on both tumors to evaluate for the presence of FOXL2 (Forkhead Box L2) and/or DICER1 (dicer 1, ribonuclease III) mutations. Results Both tumors were FIGO stage IA . Microscopically, tumor #1 contained well-differentiated Sertoli cell tumor (SCT) and adult-type granulosa cell tumor (AGCT) components, while tumor #2 comprised moderately differentiated Sertoli-Leydig cell tumor (SLCT) and AGCT components. Immunophenotypically, both tumors were positive for the sex cord-stromal markers calretinin and inhibin. NGS revealed the absence of FOXL2 c.402C>G (p.Cys134Trp) in both tumors. DICER1 mutations were absent in both tumors. Both patients remained disease-free during follow-up periods of 6 and 62 months, respectively. Conclusion Gynandroblastoma remains a distinct and rare entity with characteristic histopathological and molecular features. Our findings align with previous studies, emphasizing the rarity of FOXL2 mutation and the absence of DICER1 mutations in tumors with an AGCT component, regardless of the male sex cord component. Further studies are needed to better understand its molecular pathogenesis and long-term outcomes.

Ovarian Mesonephric-Like Adenocarcinoma Arising Within a Serous Borderline Tumor: Insights into the Implications of Cell Origin—A Case Report and Review of the Literature

Background Mesonephric-like adenocarcinoma (MLA) is a rare and aggressive tumor found in the uterus and ovaries, characterized by unique morphological and immunophenotypic features similar to mesonephric carcinoma. It has been suggested that some tumors of MLA may originate from Müllerian-type lesions. Tumor Presentation A 54-year-old woman presented with abdominal discomfort and adnexal mass. Laboratory results revealed elevated carcinoembryonic antigen and CA-125 levels. Surgical findings showed ovarian MLA arising within a serous borderline tumor (SBT) with metastasis to lymph nodes, classified as FIGO stage IIIA2. Pathological Findings The SBT component exhibited typical papillary architecture, while the MLA component showed admixture of various architectural patterns, with intraluminal eosinophilic secretions. Immunohistochemistry revealed positivity for GATA3, TTF-1, and CD10, and negativity for ER and WT1. Molecular analysis identified a KRAS p.G12D mutation, typical of MLA, and a SMAD4 mutation. Conclusion This tumor represents the ninth documented example of ovarian carcinoma with distinct components of MLA and low-grade serous neoplasm, SBT or low-grade serous carcinoma, providing valuable insights into the clinicopathological features and molecular characteristics of MLA, and contributing to the understanding of its origin and clinical behavior.

Mesonephric-Like Adenocarcinomas, an Underdiagnosed Rare Type of Gynecological Malignancy Associated with Aggressive Clinical Behaviors: A Series of 4 Patients from Single Institution

Mesonephric-like adenocarcinoma (MLA) is a newly recognized type of carcinoma that occurs in the uterus and ovaries. MLA exhibits distinct morphological and immunophenotypical features similar to those of mesonephric carcinoma of the cervix or vagina, with the majority of reported tumors containing KRAS mutations. MLA is exceedingly rare and appears to be associated with aggressive clinical behavior. Literature regarding the clinical behavior of MLAs remains limited. Here, we report four additional MLA patients with detailed morphological and immunophenotypical analysis, as well as clinicopathological and follow-up information. Additionally, molecular study was conducted on the ovarian MLA tumor and one uterine MLA tumor. Our findings support the observation that MLA is associated with an aggressive clinical course and poor clinical outcomes. Out of four tumors presented, 2 (50%) were at an advanced clinical stage, 3 (75%) presented with metastasis (including 2 (50%) with lung metastasis), and 2 (50%) patients died of the disease. MLA is often misdiagnosed as other types of gynecological carcinoma, as indicated in the literature and evidenced in our study, where 2 (50%) were initially misdiagnosed as low-grade endometrioid carcinoma in biopsy. Therefore, an integrated approach, including careful evaluation of the morphological and immunophenotypical features of the tumor, preferably with molecular confirmation, is critical for accurate diagnosis of this aggressive malignancy.

Investigation of selective glucocorticoid receptor modulation in high-grade serous ovarian cancer PDX models

In ovarian cancer (OvCa), tumor cell high glucocorticoid receptor (GR) has been associated with poor patient prognosis. In vitro, GR activation inhibits chemotherapy-induced OvCa cell death in association with transcriptional upregulation of genes encoding anti-apoptotic proteins. A recent randomized phase II study demonstrated improvement in progression-free survival (PFS) for heavily pre-treated OvCa patients randomized to receive therapy with a selective GR modulator (SGRM) plus chemotherapy compared to chemotherapy alone. We hypothesized that SGRM therapy would improve carboplatin response in OvCa patient-derived xenograft (PDX). Six high-grade serous (HGS) OvCa PDX models expressing GR mRNA ( One of the 6 GR-positive PDX models showed a significant improvement in PFS with the addition of a SGRM. Interestingly, the single model with an improved PFS was least carboplatin sensitive. Possible explanations for the modest SGRM activity include the high carboplatin sensitivity of 5 of the PDX tumors and the potential that SGRMs activate the tumor invasive immune cells in patients (absent from immunocompromised mice). The level of tumor GR protein expression alone appears insufficient for predicting SGRM response. The significant improvement in PFS shown in 1 of the 6 models after treatment with a SGRM plus chemotherapy underscores the need to determine predictive biomarkers for SGRM therapy in HGS OvCa and to better identify patient subgroups that are most likely to benefit from adding GR modulation to chemotherapy.

Nucleolar Localization of the RNA Helicase DDX21 Predicts Survival Outcomes in Gynecologic Cancers

Abstract Cancer cells with DNA repair defects (e.g., BRCA1/2 mutant cells) are vulnerable to PARP inhibitors (PARPi) due to induction of synthetic lethality. However, recent clinical evidence has shown that PARPi can prevent the growth of some cancers irrespective of their BRCA1/2 status, suggesting alternative mechanisms of action. We previously discovered one such mechanism in breast cancer involving DDX21, an RNA helicase that localizes to the nucleoli of cells and is a target of PARP1. We have now extended this observation in endometrial and ovarian cancers and provided links to patient outcomes. When PARP1-mediated ADPRylation of DDX21 is inhibited by niraparib, DDX21 is mislocalized to the nucleoplasm resulting in decreased rDNA transcription, which leads to a reduction in ribosome biogenesis, protein translation, and ultimately endometrial and ovarian cancer cell growth. High PARP1 expression was associated with high nucleolar localization of DDX21 in both cancers. High nucleolar DDX21 negatively correlated with calculated IC50s for niraparib. By studying endometrial cancer patient samples, we were able to show that high DDX21 nucleolar localization was significantly associated with decreased survival. Our study suggests that the use of PARPi as a cancer therapeutic can be expanded to further types of cancers and that DDX21 localization can potentially be used as a prognostic factor and as a biomarker for response to PARPi. Significance: Currently, there are no reliable biomarkers for response to PARPi outside of homologous recombination deficiency. Herein we present a unique potential biomarker, with clear functional understanding of the molecular mechanism by which DDX21 nucleolar localization can predict response to PARPi.

Utility of a PAX2, PTEN, and β-catenin Panel in the Diagnosis of Atypical Hyperplasia/Endometrioid Intraepithelial Neoplasia in Endometrial Polyps

The diagnosis of atypical hyperplasia/endometrioid intraepithelial neoplasm (AH/EIN) within endometrial polyps (EMPs) often poses a diagnostic conundrum. Our previous studies demonstrated that a panel of immunohistochemical (IHC) markers consisting of PAX2, PTEN, and β-catenin can be effectively utilized for the identification of AH/EIN. A total of 105 AH/EIN within EMP were analyzed using the 3-marker panel. We also evaluated these cases for the presence of morules. Benign EMP (n=90) and AH/EIN unassociated with polyp (n=111) served as controls. Aberrant expression of PAX2, PTEN, or β-catenin was observed in AH/EIN in EMP in 64.8%, 39.0%, and 61.9% of cases, respectively. At least 1 IHC marker was abnormal in 92.4% of cases. Overall, 60% of AH/EIN in EMP demonstrated abnormal results for≥2 IHC markers. The prevalence of PAX2 aberrancy was significantly lower in AH/EIN in EMP than in nonpolyp AH/EIN (64.8% vs. 81.1%, P=0.007), but higher than in benign EMP (64.8% vs. 14.4%, P<0.00001). The prevalence of β-catenin aberrancy was significantly higher in AH/EIN in EMP than in nonpolyp AH/EIN (61.9% vs. 47.7%, P=0.037). All control benign EMP demonstrated normal expression of PTEN and β-catenin. Morules were present in 38.1% of AH/EIN in EMP versus 24.3% in nonpolyp AH/EIN, and absent in benign EMP. A strong positive association was found between β-catenin and morules (Φ=0.64). Overall, 90% cases of atypical polypoid adenomyoma (n=6) and mucinous papillary proliferation (n=4) showed IHC marker aberrancy. In conclusion, the 3-marker IHC panel (PAX2, PTEN, and β-catenin) is (1) a useful tool in the diagnosis of AH/EIN in EMP; (2) PAX2 loss should be interpreted with caution and in combination with morphology and other markers.

Endocervical Adenocarcinoma With Micropapillary Component, a Distinct Histological Subtype Associated With Aggressive Behavior and Poor Prognosis: A Clinicopathologic Study of 10 Patients

Objective. Endocervical adenocarcinoma (ECA) with a micropapillary component is considered to be associated with aggressive behavior and poor prognosis. So far, only limited studies investigated this histological subtype of ECA in the literature. In this study, we present a clinicopathological analysis of 10 such tumors. Methods. We retrieved ten ECAs with a micropapillary component between January 2014 and July 2023 and analyzed their clinicopathologic features. Results. At diagnosis, nine tumors (90%) were of advanced clinical stage (FIGO stage ≥ IIA), nine tumors (90%) exhibited deep stromal invasion (≥2/3 of the cervix), and three tumors (30%) showed parametrial involvement. Lymphovascular invasion was evident in all tumors (100%), and lymph node metastasis was found in eight tumors (80%). Among the 10 patients, six were alive without disease (60%), one had a recurrent/later metastatic tumor (10%), and three died from the disease (30%). Furthermore, eight tumors were positive for PD-L1 expression (80%), while only one tumor showed HER2 overexpression (10%), and one tumor exhibited p53 mutant-type staining (10%). Conclusion. Endocervical adenocarcinomas with micropapillary components are associated with aggressive clinical behavior and a poor prognosis, which can be found in various conventional histological types of ECAs regardless of the HPV status. The high prevalence of PD-L1 expression suggests that patients with micropapillary ECAs may be good candidates for PD-L1 inhibitor treatment.

HMIL: Hierarchical Multi-Instance Learning for Fine-Grained Whole Slide Image Classification

Fine-grained classification of whole slide images (WSIs) is essential in precision oncology, enabling precise cancer diagnosis and personalized treatment strategies. The core of this task involves distinguishing subtle morphological variations within the same broad category of gigapixel-resolution images, which presents a significant challenge. While the multi-instance learning (MIL) paradigm alleviates the computational burden of WSIs, existing MIL methods often overlook hierarchical label correlations, treating fine-grained classification as a flat multi-class classification task. To overcome these limitations, we introduce a novel hierarchical multi-instance learning (HMIL) framework. By facilitating on the hierarchical alignment of inherent relationships between different hierarchy of labels at instance and bag level, our approach provides a more structured and informative learning process. Specifically, HMIL incorporates a class-wise attention mechanism that aligns hierarchical information at both the instance and bag levels. Furthermore, we introduce supervised contrastive learning to enhance the discriminative capability for fine-grained classification and a curriculum-based dynamic weighting module to adaptively balance the hierarchical feature during training. Extensive experiments on our large-scale cytology cervical cancer (CCC) dataset and two public histology datasets, BRACS and PANDA, demonstrate the state-of-the-art class-wise and overall performance of our HMIL framework. Our source code is available at https://github.com/ChengJin-git/HMIL.

Development and validation of an interpretable model integrating multimodal information for improving ovarian cancer diagnosis

AbstractOvarian cancer, a group of heterogeneous diseases, presents with extensive characteristics with the highest mortality among gynecological malignancies. Accurate and early diagnosis of ovarian cancer is of great significance. Here, we present OvcaFinder, an interpretable model constructed from ultrasound images-based deep learning (DL) predictions, Ovarian–Adnexal Reporting and Data System scores from radiologists, and routine clinical variables. OvcaFinder outperforms the clinical model and the DL model with area under the curves (AUCs) of 0.978, and 0.947 in the internal and external test datasets, respectively. OvcaFinder assistance led to improved AUCs of radiologists and inter-reader agreement. The average AUCs were improved from 0.927 to 0.977 and from 0.904 to 0.941, and the false positive rates were decreased by 13.4% and 8.3% in the internal and external test datasets, respectively. This highlights the potential of OvcaFinder to improve the diagnostic accuracy, and consistency of radiologists in identifying ovarian cancer.

Molecular Imaging of Ovarian Follicles and Tumors With Near‐Infrared II Bioconjugates

AbstractFollicular tracking is typically conducted using ultrasound technology, but its effectiveness is constrained by limited resolution. High‐resolution imaging of deep tissues can be accomplished using luminescence imaging in the near‐infrared II window (NIR‐II, 1000–1700 nm); however, the contrast agents that are used lack specificity. Here, it is reported that the FDA‐approved indocyanine green (ICG)‐conjugated recombinant human chorionic gonadotropin (hCG) protein can target early follicles with long‐term effectiveness. A novel high‐resolution NIR‐II imaging approach is developed for monitoring follicular development as well as ovulation using multi‐color imaging of ovarian vessels with a combination of non‐overlapping downconversion nanoparticles (DCNPs). The results showed that the ability to monitor early follicles of around 50 µm in diameter exceeded the spatial and temporal resolution of ultrasound or MRI without the reproductive damage associated with computed tomography radiation, and this enabled the clinical identification of the follicular reserve in patients with infertility diseases such as polycystic ovary syndrome (PCOS). In addition, NIR‐II imaging clearly targeted ovarian tumors and showed micro‐metastatic lesions, thus providing a new tool for monitoring tumors in vivo and guiding surgical resection.