Development of model for identifying homologous recombination deficiency (HRD) status of ovarian cancer with deep learning on whole slide images DOI Creative Commons
Ke Zhang, Yinghe Qiu,

Songwei Feng

и другие.

Journal of Translational Medicine, Год журнала: 2025, Номер 23(1)

Опубликована: Март 4, 2025

Homologous recombination deficiency (HRD) refers to the dysfunction of homologous repair (HRR) at cellular level. The assessment HRD status has important significance for formulation treatment plans, efficacy evaluation, and prognosis prediction patients with ovarian cancer. This study aimed construct a deep learning-based classifier identifying tumor regions from whole slide images (WSIs) stratify cancer (OC). learning models were trained on 205 H&E-stained sections which contained patients, 64 found have while 141 had proficiency (HRP) two institutions Memorial Sloan Kettering Cancer Center (MSKCC) Zhongda Hospital, Southeast University. framework includes identification by UNet + subtypes construction. Referring EasyEnsemble, we classified HRP into three distributed subsets. These subsets combined establish new training groups subsequent model integrated single named Ensemble Model. algorithm segmented 81.8% accuracy, 85.9% recall, 83.8% dice score 68.3% IoU. AUC Model was 0.769 (Precision = 0.800, Recall 0.727, F1-score 0.762) in study. most discriminative features between comprised S_mean_dln_obtuse_ratio, S_mean_dln_acute_ratio mean_Graph_T-S_Betweenness_normed. constructed enables accurate discrimination non-tumor tissues as well

Язык: Английский

Unraveling the complexity of HRD assessment in ovarian cancer by combining genomic and functional approaches: translational analyses of MITO16-MaNGO-OV-2 trial DOI Creative Commons
Benedetta Pellegrino, Ettore Capoluongo, Marina Bagnoli

и другие.

ESMO Open, Год журнала: 2025, Номер 10(1), С. 104091 - 104091

Опубликована: Янв. 1, 2025

Ovarian cancer (OvC) constitutes significant management challenges primarily due to its late-stage diagnosis and the development of resistance chemotherapy. The standard treatment regimen typically includes carboplatin paclitaxel, with addition poly (ADP-ribose) polymerase inhibitors for patients high-grade serous ovarian (HGSOC) harboring BRCA1/2 mutations. However, variability in responses suggests need investigate factors beyond mutations, such as DNA repair mechanisms epigenetic alterations. Notably, homologous recombination deficiency (HRD) is observed an additional 20% HGSOC cases, indicating a broader spectrum defects. Existing commercial HRD assays have certain limitations, prompting global effort develop new genomic functional tests through academic research.

Язык: Английский

Процитировано

1
Niraparib restricts intraperitoneal metastases of ovarian cancer by eliciting CD36-dependent ferroptosis DOI Creative Commons

Ning Jin,

Yiyu Qian,

Xiao-Fei Jiao

и другие.

Redox Biology, Год журнала: 2025, Номер 80, С. 103528 - 103528

Опубликована: Фев. 3, 2025

Ovarian cancer (OC) is prone to peritoneum or omentum dissemination, thus giving rise the formidable challenge of unresectable surgery and a dismal survival rate. Although niraparib holds pivotal role in maintenance treatment OC, its effect on suppressing metastases during primary intervention remains enigmatic. Recently, we initiated prospective clinical study (NCT04507841) order evaluate therapeutic efficacy neoadjuvant monotherapy for advanced OC with homologous recombination deficiency. An analysis patient tumor burden before after showed remarkable vulnerability intraperitoneal exposure. This killing capacity was closely associated accumulation fatty acids within abdomen, which confirmed by increased susceptibility cells presence acids. In context abundant acids, elevated intracellular levels lipid peroxidation, leading subsequent cell ferroptosis p53 BRCA-independent manner. Notably, under exposure, critical acid transporter CD36 dramatically upregulated tumors, facilitating excessive uptake Pharmacological inhibition either impaired anti-tumor activity both vitro murine ID8 models. Our findings demonstrate as novel mechanism underlying regression induced niraparib, thereby offering tantalizing prospects frontline application this agent management OC.

Язык: Английский

Процитировано

1
Current HRD assays in ovarian cancer: differences, pitfalls, limitations, and novel approaches DOI Creative Commons
Federica Guffanti,

Ilaria Mengoli,

Giovanna Damia

и другие.

Frontiers in Oncology, Год журнала: 2024, Номер 14

Опубликована: Авг. 16, 2024

Ovarian carcinoma (OC) still represents an insidious and fatal malignancy, few significant results have been obtained in the last two decades to improve patient survival. Novel targeted therapies such as poly (ADP-ribose) polymerase inhibitors (PARPi) successfully introduced clinical management of OC, but not all patients will benefit, drug resistance almost inevitably occurs. The identification who are likely respond PARPi-based relies on homologous recombination deficiency (HRD) tests, this condition is associated with response these treatments. This review summarizes genomic functional HRD assays currently used practice those under evaluation, implications testing their current pitfalls limitations. Special emphasis be placed development use machine learning artificial intelligence technologies novel strategies overcome limitations tests for a better-personalized treatment outcomes.

Язык: Английский

Процитировано

6
Identification of the novel BRCA1 c.2463_2464delTA mutation in two high grade serous ovarian cancer sisters and potential dosage effects implications: a case report DOI
Valentina Lombardo, Giuseppa Scandurra,

Angela Pettinato

и другие.

Molecular Biology Reports, Год журнала: 2025, Номер 52(1)

Опубликована: Янв. 7, 2025

Язык: Английский

Процитировано

0
Development of model for identifying homologous recombination deficiency (HRD) status of ovarian cancer with deep learning on whole slide images DOI Creative Commons
Ke Zhang, Yinghe Qiu,

Songwei Feng

и другие.

Journal of Translational Medicine, Год журнала: 2025, Номер 23(1)

Опубликована: Март 4, 2025

Homologous recombination deficiency (HRD) refers to the dysfunction of homologous repair (HRR) at cellular level. The assessment HRD status has important significance for formulation treatment plans, efficacy evaluation, and prognosis prediction patients with ovarian cancer. This study aimed construct a deep learning-based classifier identifying tumor regions from whole slide images (WSIs) stratify cancer (OC). learning models were trained on 205 H&E-stained sections which contained patients, 64 found have while 141 had proficiency (HRP) two institutions Memorial Sloan Kettering Cancer Center (MSKCC) Zhongda Hospital, Southeast University. framework includes identification by UNet + subtypes construction. Referring EasyEnsemble, we classified HRP into three distributed subsets. These subsets combined establish new training groups subsequent model integrated single named Ensemble Model. algorithm segmented 81.8% accuracy, 85.9% recall, 83.8% dice score 68.3% IoU. AUC Model was 0.769 (Precision = 0.800, Recall 0.727, F1-score 0.762) in study. most discriminative features between comprised S_mean_dln_obtuse_ratio, S_mean_dln_acute_ratio mean_Graph_T-S_Betweenness_normed. constructed enables accurate discrimination non-tumor tissues as well

Язык: Английский

Процитировано

0