Tumour microenvironment characterisation to stratify patients for hyperthermic intraperitoneal chemotherapy in high-grade serous ovarian cancer (OVHIPEC-1).

BACKGROUND: Hyperthermic intraperitoneal chemotherapy (HIPEC) improves survival in patients with Stage III ovarian cancer following interval cytoreductive surgery (CRS). Optimising patient selection is essential to maximise treatment efficacy and avoid overtreatment. This study aimed to identify biomarkers that predict HIPEC benefit by analysing gene signatures and cellular composition of tumours from participants in the OVHIPEC-1 trial. METHODS: Whole-transcriptome RNA sequencing data were retrieved from high-grade serous ovarian cancer (HGSOC) samples from 147 patients obtained during interval CRS. We performed differential gene expression analysis and applied deconvolution methods to estimate cell-type proportions in bulk mRNA data, validated by histological assessment. We tested the interaction between treatment and potential predictors on progression-free survival using Cox proportional hazards models. RESULTS: While differential gene expression analysis did not yield any predictive biomarkers, the cellular composition, as characterised by deconvolution, indicated that the absence of macrophages and the presence of B cells in the tumour microenvironment are potential predictors of HIPEC benefit. The histological assessment confirmed the predictive value of macrophage absence. CONCLUSION: Immune cell composition, in particular macrophages absence, may predict response to HIPEC in HGSOC and these hypothesis-generating findings warrant further investigation. CLINICAL TRIAL REGISTRATION: NCT00426257.

Tumor-Infiltrating Lymphocytes in Patients With Stage I Triple-Negative Breast Cancer Untreated With Chemotherapy.

Importance: The absolute benefit of chemotherapy for all patients with stage I triple-negative breast cancer (TNBC) is unclear, and biomarkers are not currently available for selecting patients with an excellent outcome for whom neoadjuvant or adjuvant chemotherapy may have negligible benefit. High levels of stromal tumor-infiltrating lymphocytes (sTILs) are associated with favorable survival in TNBC, but data solely in stage I TNBC are lacking. Objective: To examine the outcomes of patients of all ages with stage I TNBC solely and who received neither neoadjuvant nor adjuvant chemotherapy, according to centrally reviewed sTIL levels at prespecified cutoffs. Design, Setting, and Participants: This cohort study used the Netherlands Cancer Registry to identify patients diagnosed with stage I TNBC between January 1, 2005, and December 31, 2015, who were not treated with chemotherapy. Only patients who did not receive neoadjuvant and/or adjuvant chemotherapy were selected. The clinical data were matched with their corresponding pathology data provided by the Dutch Pathology Registry. Data analysis was performed between February and October 2023. Main Outcomes and Measures: The primary end point was breast cancer-specific survival (BCSS) at 5, 10, and 15 years for the prespecified sTIL level cutoffs of 30%, 50%, and 75%. Hematoxylin and eosin-stained slides were used for central review of histologic subtype, grade, and lymphovascular invasion. The International Immuno-Oncology Biomarker Working Group guidelines were used to score the sTIL levels; these levels were determined for 1041 patients. Results: A total of 4511 females with stage I TNBC (mean [SD] age at diagnosis, 64.4 [11.1] years; median follow-up, 11.4 [95% CI, 10.9-11.9] years) were included. Most tumors (952 [91.5%]) were invasive carcinomas of nonspecial histologic subtype. Most patients (548 [52.6%]) had pT1cN0 tumors. Median (range) sTIL level was 5% (1%-99%). A total of 775 patients (74.4%) had sTIL levels below 30%, 266 (25.6%) had 30% or greater, 203 (19.5%) had 50% or greater, and 141 (13.5%) had 75% or greater. Patients with pT1abN0 tumors had a more favorable outcome vs patients with pT1cN0 tumors, with a 10-year BCSS of 92% (95% CI, 89%-94%) vs 86% (95% CI, 82%-89%). In the overall cohort, sTIL levels of at least 30% were associated with better BCSS compared with sTIL levels less than 30% (96% and 87%, respectively; hazard ratio [HR], 0.45; 95% CI, 0.26-0.77). High sTIL levels of 50% or greater were associated with a better outcome than low sTIL levels of less than 50% (HR, 0.27; 95% CI, 0.10-0.74) in patients with pT1C tumors, with a 10-year BCSS of 95% increasing to 98% with sTIL levels of 75% or greater. Conclusions and Relevance: Results of this study showed that patients with stage I TNBC and high level of sTILs who did not receive neoadjuvant or adjuvant chemotherapy had excellent 10-year BCSS. The findings further support the role of sTILs as integral biomarkers in prospective clinical trials of therapy optimization for this patient population.

PatchSorter: a high throughput deep learning digital pathology tool for object labeling.

The discovery of patterns associated with diagnosis, prognosis, and therapy response in digital pathology images often requires intractable labeling of large quantities of histological objects. Here we release an open-source labeling tool, PatchSorter, which integrates deep learning with an intuitive web interface. Using >100,000 objects, we demonstrate a >7x improvement in labels per second over unaided labeling, with minimal impact on labeling accuracy, thus enabling high-throughput labeling of large datasets.

High expression of phosphoglycerate dehydrogenase predicts poor outcome in patients with high-grade serous ovarian cancer.

INTRODUCTION: High-grade serous ovarian cancer (HGSOC) is characterized by high mortality and prevalent recurrences. This study investigates the prognostic value of phosphoglycerate dehydrogenase (PHGDH) in HGSOC which has been linked to metabolic reprogramming and recurrences in other cancers. METHODS: Data from 306 patients with advanced-stage HGSOC treated between 2008 and 2015 were analyzed. PHGDH expression levels were determined using immunohistochemistry and categorized as "low" or "high." RESULTS: PHGDH-high was associated with higher FIGO stage and increased use of neoadjuvant chemotherapy. Patients with PHGDH-high tumors had significantly worse survival than PHDH-low, even after adjusting for confounding factors.

The Information Technology (IT) Infrastructure of the Multicenter Archipelago of Ovarian Cancer Research Biobank: A Potential Blueprint for Other Biobanks.

Objective: Biobanks play a crucial role in fundamental and translational research by storing valuable biomaterials and data for future analyses. However, the design of their information technology (IT) infrastructures is often customized to specific requirements, thereby lacking the ability to be used for biobanks comprising other (types of) diseases. This results in substantial costs, time, and efforts for each new biobank project. The Dutch multicenter Archipelago of Ovarian Cancer Research (AOCR) biobank has developed an innovative, reusable IT infrastructure capable of adaptation to various biobanks, thereby enabling cost-effective and efficient implementation and management of biobank IT systems. Methods and Results: The AOCR IT infrastructure incorporates preexisting biobank software, mainly managed by Health-RI. The web-based registration tool Ldot is used for secure storage and pseudonymization of patient data. Clinicopathological data are retrieved from the Netherlands Cancer Registry and the Dutch nationwide pathology databank (Palga), both established repositories, reducing administrative workload and ensuring high data quality. Metadata of collected biomaterials are stored in the OpenSpecimen system. For digital pathology research, a hematoxylin and eosin-stained slide from each patient's tumor is digitized and uploaded to Slide Score. Furthermore, adhering to the Findable, Accessible, Interoperable, and Reusable (FAIR) principles, genomic data derived from the AOCR samples are stored in cBioPortal. Conclusion: The IT infrastructure of the AOCR biobank represents a new standard for biobanks, offering flexibility to handle diverse diseases and types of biomaterials. This infrastructure bypasses the need for disease-specific, custom-built software, thereby being cost- and time-effective while ensuring data quality and legislative compliance. The adaptability of this infrastructure highlights its potential to serve as a blueprint for the development of IT infrastructures in both new and existing biobanks.

Does "One Size Fits All"? Rethinking FIGO Depth of Invasion Measurements in Vulvar Cancer.

Depth of invasion (DOI) is an important diagnostic parameter in patients with vulvar carcinoma, where a cutoff value of 1 mm largely determines the tumor stage and the need for groin surgery. DOI measurement should be reproducible and straightforward. In light of the new recommendation on how to measure DOI in the International Federation of Gynecology and Obstetrics (FIGO) staging system 2021, an exploratory study was conducted on the current practice of DOI measurement in vulvar cancer. In this study of 26 selected cases, 10 pathologists with high exposure to vulvar cancer cases in daily practice assessed both the conventional (FIGO 2009) and alternative (FIGO 2021) DOI methods for applicability and preference. In this set of cases, the DOI measurement according to FIGO 2009 was generally considered easier to apply than the measurement according to FIGO 2021, with applicability being rated as "easy to reasonable" in 76.9% versus 38.5% of cases, respectively ( P =0.005). The preferred method was FIGO 2009 or tumor thickness in 14 cases and FIGO 2021 in 6 cases. No invasion was preferred in 1 case. For the remaining 5 cases, half of the pathologists opted for the FIGO 2009 method and half for the FIGO 2021 method. Although the FIGO 2009 method proved to be more readily applicable in most of the cases studied, the method may differ for each case. There may not be a "one size fits all" solution for all cases of vulvar cancer.

Evaluating the effectiveness of pre-operative diagnosis of ovarian cancer using minimally invasive liquid biopsies by combining serum human epididymis protein 4 and cell-free DNA in patients with an ovarian mass.

OBJECTIVE: To assess the feasibility of scalable, objective, and minimally invasive liquid biopsy-derived biomarkers such as cell-free DNA copy number profiles, human epididymis protein 4 (HE4), and cancer antigen 125 (CA125) for pre-operative risk assessment of early-stage ovarian cancer in a clinically representative and diagnostically challenging population and to compare the performance of these biomarkers with the Risk of Malignancy Index (RMI). METHODS: In this case-control study, we included 100 patients with an ovarian mass clinically suspected to be early-stage ovarian cancer. Of these 100 patients, 50 were confirmed to have a malignant mass (cases) and 50 had a benign mass (controls). Using WisecondorX, an algorithm used extensively in non-invasive prenatal testing, we calculated the benign-calibrated copy number profile abnormality score. This score represents how different a sample is from benign controls based on copy number profiles. We combined this score with HE4 serum concentration to separate cases and controls. RESULTS: Combining the benign-calibrated copy number profile abnormality score with HE4, we obtained a model with a significantly higher sensitivity (42% vs 0%; p<0.002) at 99% specificity as compared with the RMI that is currently employed in clinical practice. Investigating performance in subgroups, we observed especially large differences in the advanced stage and non-high-grade serous ovarian cancer groups. CONCLUSION: This study demonstrates that cell-free DNA can be successfully employed to perform pre-operative risk of malignancy assessment for ovarian masses; however, results warrant validation in a more extensive clinical study.

PROACTING: predicting pathological complete response to neoadjuvant chemotherapy in breast cancer from routine diagnostic histopathology biopsies with deep learning.

BACKGROUND: Invasive breast cancer patients are increasingly being treated with neoadjuvant chemotherapy; however, only a fraction of the patients respond to it completely. To prevent overtreatment, there is an urgent need for biomarkers to predict treatment response before administering the therapy. METHODS: In this retrospective study, we developed hypothesis-driven interpretable biomarkers based on deep learning, to predict the pathological complete response (pCR, i.e., the absence of tumor cells in the surgical resection specimens) to neoadjuvant chemotherapy solely using digital pathology H&E images of pre-treatment breast biopsies. Our approach consists of two steps: First, we use deep learning to characterize aspects of the tumor micro-environment by detecting mitoses and segmenting tissue into several morphology compartments including tumor, lymphocytes and stroma. Second, we derive computational biomarkers from the segmentation and detection output to encode slide-level relationships of components of the tumor microenvironment, such as tumor and mitoses, stroma, and tumor infiltrating lymphocytes (TILs). RESULTS: We developed and evaluated our method on slides from n = 721 patients from three European medical centers with triple-negative and Luminal B breast cancers and performed external independent validation on n = 126 patients from a public dataset. We report the predictive value of the investigated biomarkers for predicting pCR with areas under the receiver operating characteristic curve between 0.66 and 0.88 across the tested cohorts. CONCLUSION: The proposed computational biomarkers predict pCR, but will require more evaluation and finetuning for clinical application. Our results further corroborate the potential role of deep learning to automate TILs quantification, and their predictive value in breast cancer neoadjuvant treatment planning, along with automated mitoses quantification. We made our method publicly available to extract segmentation-based biomarkers for research purposes.

The prognostic value of tumor-stroma ratio and a newly developed computer-aided quantitative analysis of routine H&E slides in high-grade serous ovarian cancer.

Introduction: Tumor-stroma ratio (TSR) is prognostic in multiple cancers, while its role in high-grade serous ovarian cancer (HGSOC) remains unclear. Despite the prognostic insight gained from genetic profiles and tumor-infiltrating lymphocytes (TILs), the prognostic use of histology slides remains limited, while it enables the identification of tumor characteristics via computational pathology reducing scoring time and costs. To address this, this study aimed to assess TSR's prognostic role in HGSOC and its association with TILs. We additionally developed an algorithm, Ovarian-TSR (OTSR), using deep learning for TSR scoring, comparing it to manual scoring. Methods: 340 patients with advanced-stage who underwent primary debulking surgery (PDS) or neo-adjuvant chemotherapy (NACT) with interval debulking (IDS). TSR was assessed in both the most invasive (MI) and whole tumor (WT) regions through manual scoring by pathologists and quantification using OTSR. Patients were categorized as stroma-rich (≥ 50% stroma) or stroma-poor (< 50%). TILs were evaluated via immunohistochemical staining. Results: In PDS, stroma-rich tumors were significantly associated with a more frequent papillary growth pattern (60% vs 34%), while In NACT stroma-rich tumors had a lower Tumor Regression Grading (TRG 4&5, 21% vs 57%) and increased pleural metastasis (25% vs 16%). Stroma-rich patients had significantly shorter overall and progression-free survival compared to stroma-poor (31 versus 45 months; P < 0.0001, and 15 versus 17 months; P = 0.0008, respectively). Combining stromal percentage and TILs led to three distinct survival groups with good (stroma-poor, high TIL), medium (stroma-rich, high TIL, or; stroma-poor, Low TIL), and poor(stroma-rich, low TIL) survival. These survival groups remained significant in CD8 and CD103 in multivariable analysis (Hazard ratio (HR) = 1.42, 95% Confidence-interval (CI) = 1.02-1.99; HR = 1.49, 95% CI = 1.01-2.18, and HR = 1.48, 95% CI = 1.05-2.08; HR = 2.24, 95% CI = 1.55-3.23, respectively). OTSR was able to recapitulate these results and demonstrated high concordance with expert pathologists (correlation = 0.83). Conclusions: TSR is an independent prognostic factor for survival assessment in HGSOC. Stroma-rich tumors have a worse prognosis and, in the case of NACT, a higher likelihood of pleural metastasis. OTSR provides a cost and time-efficient way of determining TSR with high reproducibility and reduced inter-observer variability.

B-cells and regulatory T-cells in the microenvironment of HER2+ breast cancer are associated with decreased survival: a real-world analysis of women with HER2+ metastatic breast cancer.

BACKGROUND: Despite major improvements in treatment of HER2-positive metastatic breast cancer (MBC), only few patients achieve complete remission and remain progression free for a prolonged time. The tumor immune microenvironment plays an important role in the response to treatment in HER2-positive breast cancer and could contain valuable prognostic information. Detailed information on the cancer-immune cell interactions in HER2-positive MBC is however still lacking. By characterizing the tumor immune microenvironment in patients with HER2-positive MBC, we aimed to get a better understanding why overall survival (OS) differs so widely and which alternative treatment approaches may improve outcome. METHODS: We included all patients with HER2-positive MBC who were treated with trastuzumab-based palliative therapy in the Netherlands Cancer Institute between 2000 and 2014 and for whom pre-treatment tissue from the primary tumor or from metastases was available. Infiltrating immune cells and their spatial relationships to one another and to tumor cells were characterized by immunohistochemistry and multiplex immunofluorescence. We also evaluated immune signatures and other key pathways using next-generation RNA-sequencing data. With nine years median follow-up from initial diagnosis of MBC, we investigated the association between tumor and immune characteristics and outcome. RESULTS: A total of 124 patients with 147 samples were included and evaluated. The different technologies showed high correlations between each other. T-cells were less prevalent in metastases compared to primary tumors, whereas B-cells and regulatory T-cells (Tregs) were comparable between primary tumors and metastases. Stromal tumor-infiltrating lymphocytes in general were not associated with OS. The infiltration of B-cells and Tregs in the primary tumor was associated with unfavorable OS. Four signatures classifying the extracellular matrix of primary tumors showed differential survival in the population as a whole. CONCLUSIONS: In a real-world cohort of 124 patients with HER2-positive MBC, B-cells, and Tregs in primary tumors are associated with unfavorable survival. With this paper, we provide a comprehensive insight in the tumor immune microenvironment that could guide further research into development of novel immunomodulatory strategies.

Characterization of the Tumor Microenvironment of De Novo Oligometastatic Breast Cancer in a Nationwide Cohort.

PURPOSE: Oligometastatic breast cancer (OMBC) has a more favorable outcome than widespread metastatic breast cancer. Some patients with OMBC achieve long-term remission if treated with multimodality therapy, including systemic and locally ablative therapies. However, not all patients with OMBC benefit from such treatment, while all experience toxicity. To explore biomarkers identifying patients with OMBC and potential long-term survival, we compared tumor-immune characteristics of patients with OMBC and long-term versus shorter-term survival. MATERIALS AND METHODS: We collected tumor tissue of 97 patients with de novo OMBC (≤5 metastases) via the Dutch nationwide cancer and pathology registries using a case-control design. Long-term survivors (LTS) were defined as patients alive ≥10 years since OMBC diagnosis. Fifty-five LTS and 42 shorter-term survivors (STS) were included. Median follow-up was 15 years (IQR, 14-16). Tumor characteristics and infiltrating immune cells were assessed by immunohistochemistry and next-generation RNA-sequencing. Association of the resulting 52 biomarkers with long-term survival was assessed using logistic regression. Associations with survival within LTS were assessed using Cox-proportional hazards modeling. P values were adjusted for multiple hypothesis testing. RESULTS: Most patients had estrogen receptor (ER)-positive OMBC (n = 86; 89%) and 23 (24%) had human epidermal growth factor receptor 2-positive disease. ER positivity in primary tumors distinguished LTS from STS. In addition, extracellular matrix (ECM)2-low and ECM4-high distinguished between long-term and shorter-term survival. Immune levels in the primary tumor did not associate with LTS. However, within the LTS subset, higher immune levels associated with improved progression-free survival. CONCLUSION: We identified tumor and ECM features in the primary tumor of patients with de novo OMBC that were associated with long-term survival. Our data should be validated in other patients with OMBC before they can be used in clinical practice.

CMTM6 shapes antitumor T cell response through modulating protein expression of CD58 and PD-L1.

The dysregulated expression of immune checkpoint molecules enables cancer cells to evade immune destruction. While blockade of inhibitory immune checkpoints like PD-L1 forms the basis of current cancer immunotherapies, a deficiency in costimulatory signals can render these therapies futile. CD58, a costimulatory ligand, plays a crucial role in antitumor immune responses, but the mechanisms controlling its expression remain unclear. Using two systematic approaches, we reveal that CMTM6 positively regulates CD58 expression. Notably, CMTM6 interacts with both CD58 and PD-L1, maintaining the expression of these two immune checkpoint ligands with opposing functions. Functionally, the presence of CMTM6 and CD58 on tumor cells significantly affects T cell-tumor interactions and response to PD-L1-PD-1 blockade. Collectively, these findings provide fundamental insights into CD58 regulation, uncover a shared regulator of stimulatory and inhibitory immune checkpoints, and highlight the importance of tumor-intrinsic CMTM6 and CD58 expression in antitumor immune responses.

RadioLOGIC, a healthcare model for processing electronic health records and decision-making in breast disease.

Digital health data used in diagnostics, patient care, and oncology research continue to accumulate exponentially. Most medical information, and particularly radiology results, are stored in free-text format, and the potential of these data remains untapped. In this study, a radiological repomics-driven model incorporating medical token cognition (RadioLOGIC) is proposed to extract repomics (report omics) features from unstructured electronic health records and to assess human health and predict pathological outcome via transfer learning. The average accuracy and F1-weighted score for the extraction of repomics features using RadioLOGIC are 0.934 and 0.934, respectively, and 0.906 and 0.903 for the prediction of breast imaging-reporting and data system scores. The areas under the receiver operating characteristic curve for the prediction of pathological outcome without and with transfer learning are 0.912 and 0.945, respectively. RadioLOGIC outperforms cohort models in the capability to extract features and also reveals promise for checking clinical diagnoses directly from electronic health records.

Immune landscape of breast tumors with low and intermediate estrogen receptor expression.

Immune checkpoint blockade (ICB) is currently approved for patients with triple-negative breast cancer (TNBC), whereas responses to ICB are also observed in a small subgroup of Estrogen Receptor (ER)-positive breast cancer. The cut-off for ER-positivity (≥1%) is based on likelihood of endocrine treatment response, but ER-positive breast cancer represents a very heterogeneous group. This raises the question whether selection based on ER-negativity should be revisited to select patients for ICB treatment in the context of clinical trials. Stromal tumor-infiltrating lymphocytes (sTILs) and other immune parameters are higher in TNBC compared to ER-positive breast cancer, but it is unknown whether lower ER levels are associated with more inflamed tumor microenvironments (TME). We collected a consecutive series of primary tumors from 173 HER2-negative breast cancer patients, enriched for tumors with ER expression between 1 and 99% and found levels of stromal TILs, CD8 + T cells, and PD-L1 positivity in breast tumors with ER 1-9% and ER 10-50% to be comparable to tumors with ER 0%. Expression of immune-related gene signatures in tumors with ER 1-9% and ER 10-50% was comparable to ER 0%, and higher than in tumors with ER 51-99% and ER 100%. Our results suggest that the immune landscape of ER low tumors (1-9%) and ER intermediate tumors (10-50%) mimic that of primary TNBC.

PD-L1 blockade in combination with carboplatin as immune induction in metastatic lobular breast cancer: the GELATO trial.

Invasive lobular breast cancer (ILC) is the second most common histological breast cancer subtype, but ILC-specific trials are lacking. Translational research revealed an immune-related ILC subset, and in mouse ILC models, synergy between immune checkpoint blockade and platinum was observed. In the phase II GELATO trial ( NCT03147040 ), patients with metastatic ILC were treated with weekly carboplatin (area under the curve 1.5 mg ml-1 min-1) as immune induction for 12 weeks and atezolizumab (PD-L1 blockade; triweekly) from the third week until progression. Four of 23 evaluable patients had a partial response (17%), and 2 had stable disease, resulting in a clinical benefit rate of 26%. From these six patients, four had triple-negative ILC (TN-ILC). We observed higher CD8+ T cell infiltration, immune checkpoint expression and exhausted T cells after treatment. With this GELATO trial, we show that ILC-specific clinical trials are feasible and demonstrate promising antitumor activity of atezolizumab with carboplatin, particularly for TN-ILC, and provide insights for the design of highly needed ILC-specific trials.

Predicting breast cancer types on and beyond molecular level in a multi-modal fashion.

Accurately determining the molecular subtypes of breast cancer is important for the prognosis of breast cancer patients and can guide treatment selection. In this study, we develop a deep learning-based model for predicting the molecular subtypes of breast cancer directly from the diagnostic mammography and ultrasound images. Multi-modal deep learning with intra- and inter-modality attention modules (MDL-IIA) is proposed to extract important relations between mammography and ultrasound for this task. MDL-IIA leads to the best diagnostic performance compared to other cohort models in predicting 4-category molecular subtypes with Matthews correlation coefficient (MCC) of 0.837 (95% confidence interval [CI]: 0.803, 0.870). The MDL-IIA model can also discriminate between Luminal and Non-Luminal disease with an area under the receiver operating characteristic curve of 0.929 (95% CI: 0.903, 0.951). These results significantly outperform clinicians' predictions based on radiographic imaging. Beyond molecular-level test, based on gene-level ground truth, our method can bypass the inherent uncertainty from immunohistochemistry test. This work thus provides a noninvasive method to predict the molecular subtypes of breast cancer, potentially guiding treatment selection for breast cancer patients and providing decision support for clinicians.

Homologous Recombination Deficiency and Cyclin E1 Amplification Are Correlated with Immune Cell Infiltration and Survival in High-Grade Serous Ovarian Cancer.

BACKGROUND: How molecular profiles are associated with tumor microenvironment (TME) in high-grade serous ovarian cancer (HGSOC) is incompletely understood. Therefore, we analyzed the TME and molecular profiles of HGSOC and assessed their associations with overall survival (OS). METHODS: Patients with advanced-stage HGSOC treated in three Dutch hospitals between 2008-2015 were included. Patient data were collected from medical records. BRCA1/2 mutation, BRCA1 promotor methylation analyses, and copy number variations were used to define molecular profiles. Immune cells were assessed with immunohistochemical staining. RESULTS: 348 patients were categorized as BRCA mutation (BRCAm) (BRCAm or promotor methylation) (30%), non-BRCA mutated HRD (19%), Cyclin E1 (CCNE1)-amplification (13%), non-BRCAmut HRD and CCNE1-amplification (double classifier) (20%), and no specific molecular profile (NSMP) (18%). BRCAm showed highest immune cell densities and CCNE1-amplification lowest. BRCAm showed the most favorable OS (52.5 months), compared to non-BRCAmut HRD (41.0 months), CCNE1-amplification (28.0 months), double classifier (27.8 months), and NSMP (35.4 months). Higher immune cell densities showed a favorable OS compared to lower, also within the profiles. CD8+, CD20+, and CD103+ cells remained associated with OS in multivariable analysis. CONCLUSIONS: Molecular profiles and TME are associated with OS. TME differs per profile, with higher immune cell densities showing a favorable OS, even within the profiles. HGSOC does not reflect one entity but comprises different entities based on molecular profiles and TME.

HRD-related morphology discovery in breast cancer by controlling for confounding factors.

Lazard et al.1 predict homologous recombination deficiency from hematoxylin and eosin-stained slides of breast cancer tissue using deep learning. By controlling for technical artifacts on a curated dataset, the model puts forward novel HRD-related morphologies in luminal breast cancers.

Deep learning for fully-automated nuclear pleomorphism scoring in breast cancer.

To guide the choice of treatment, every new breast cancer is assessed for aggressiveness (i.e., graded) by an experienced histopathologist. Typically, this tumor grade consists of three components, one of which is the nuclear pleomorphism score (the extent of abnormalities in the overall appearance of tumor nuclei). The degree of nuclear pleomorphism is subjectively classified from 1 to 3, where a score of 1 most closely resembles epithelial cells of normal breast epithelium and 3 shows the greatest abnormalities. Establishing numerical criteria for grading nuclear pleomorphism is challenging, and inter-observer agreement is poor. Therefore, we studied the use of deep learning to develop fully automated nuclear pleomorphism scoring in breast cancer. The reference standard used for training the algorithm consisted of the collective knowledge of an international panel of 10 pathologists on a curated set of regions of interest covering the entire spectrum of tumor morphology in breast cancer. To fully exploit the information provided by the pathologists, a first-of-its-kind deep regression model was trained to yield a continuous scoring rather than limiting the pleomorphism scoring to the standard three-tiered system. Our approach preserves the continuum of nuclear pleomorphism without necessitating a large data set with explicit annotations of tumor nuclei. Once translated to the traditional system, our approach achieves top pathologist-level performance in multiple experiments on regions of interest and whole-slide images, compared to a panel of 10 and 4 pathologists, respectively.