Current laboratory testing practices for mismatch repair deficiency and microsatellite instability testing: A survey-based review of current laboratory practices.

OBJECTIVES: To describe mismatch repair (MMR) and microsatellite instability (MSI) testing practices in laboratories using the College of American Pathologists (CAP) MSI/MMR proficiency testing programs prior to the 2022 publication of the MSI/MMR practice guidelines copublished by CAP and the Association of Molecular Pathology (AMP). METHODS: Data from supplemental questionnaires provided with the 2020-B MSI/MMR programs to 542 laboratories across different practice settings were reviewed. Questionnaires contained 21 questions regarding the type of testing performed, specimen/tumor types used for testing, and clinical practices for checkpoint blockade therapy. RESULTS: Domestic laboratories test for MSI/MMR more often than international laboratories (P = .04) and academic hospitals/medical centers test more frequently than nonhospital sites/clinics (P = .03). The most commonly used testing modality is immunohistochemistry, followed by polymerase chain reaction, then next-generation sequencing. Most laboratories (72.6%; 347/478) reported awareness of the use of immune checkpoint inhibitor therapy for patients with high MSI or MMR-deficient results. CONCLUSIONS: The results demonstrate the state of MMR and MSI testing in laboratories prior to the publication of the CAP/AMP best practice guidelines, highlighting differences between various laboratory types. The findings indicate the importance of consensus guidelines and provide a baseline for comparison after their implementation.

Validation and Implementation of a Somatic-Only Tumor Exome for Routine Clinical Application.

Next-generation sequencing-based genomic testing is standard of care for tumor workflows. However, its application across different institutions continues to be challenging given the diversity of needs and resource availability among different institutions globally. Moreover, the use of a variety of different panels, including those from a few individual genes to those involving hundreds of genes, results in a relatively skewed distribution of care for patients. It is imperative to obtain a higher level of standardization without having to be restricted to specific kits or requiring repeated validations, which are generally expensive. We show the validation and clinical implementation of the DH-CancerSeq assay, a tumor-only whole-exome-based sequencing assay with integrated informatics, while providing similar input requirements, sensitivity, and specificity to a previously validated targeted gene panel and maintaining similar turnaround times for patient care.

Deep Learning for Grading Endometrial Cancer.

Endometrial cancer is the fourth most common cancer in women in the United States, with a lifetime risk of approximately 2.8%. Precise histologic evaluation and molecular classification of endometrial cancer are important for effective patient management and determining the best treatment options. This study introduces EndoNet, which uses convolutional neural networks for extracting histologic features and a vision transformer for aggregating these features and classifying slides into high- and low-grade cases. The model was trained on 929 digitized hematoxylin and eosin-stained whole-slide images of endometrial cancer from hysterectomy cases at Dartmouth-Health. It classifies these slides into low-grade (endometrioid grades 1 and 2) and high-grade (endometrioid carcinoma International Federation of Gynecology and Obstetrics grade 3, uterine serous carcinoma, or carcinosarcoma) categories. EndoNet was evaluated on an internal test set of 110 patients and an external test set of 100 patients from The Cancer Genome Atlas database. The model achieved a weighted average F1 score of 0.91 (95% CI, 0.86 to 0.95) and an area under the curve of 0.95 (95% CI, 0.89 to 0.99) on the internal test, and 0.86 (95% CI, 0.80 to 0.94) for F1 score and 0.86 (95% CI, 0.75 to 0.93) for area under the curve on the external test. Pending further validation, EndoNet has the potential to support pathologists without the need of manual annotations in classifying the grades of gynecologic pathology tumors.

Does ROSE matter? Evaluation of final diagnostic, PD-L1 immunohistochemistry, and molecular testing yields of CT-guided lung biopsies performed before and after the onset of the COVID-19 pandemic.

INTRODUCTION: The onset of the COVID-19 pandemic, with urgent implementation of safety protocols limiting the number of on-site personnel, essentially terminated the use of rapid on-site evaluation (ROSE) for computed tomography (CT)--guided lung biopsies at our institution. The diminished use of ROSE during the pandemic prompted us to reevaluate the potential value of ROSE for CT-guided lung biopsies. MATERIALS AND METHODS: We retrospectively identified all CT-guided lung biopsies from 2017 to 2022. Associations between the use of ROSE, adequate diagnostic and ancillary testing (programmed death-ligand 1 immunohistochemistry and next-generation sequencing) outcomes, and other factors such as the number of passes performed and lesion size, were evaluated. RESULTS: Nine hundred twelve CT-guided lung biopsies were performed from 2017 to 2022; 171 (19%) utilized ROSE. The use of ROSE had been steadily decreasing prior to the pandemic but was essentially eliminated with the onset of the pandemic. By univariable analysis, the employment of ROSE was more likely to be associated with an adequate final diagnosis (odds ratio = 2.14, 95% confidence interval: [1.24-3.70], P = 0.006) and successful molecular testing (odds ratio = 2.16, 95% confidence interval: [1.11-4.21], P = 0.024). However, those associations were not present in multivariable analyses that incorporated the number of passes performed or lesion size. There were no differences in diagnostic adequacy or ancillary testing yields when comparing the periods 2017-2019 and 2020-2022, despite declining use of ROSE. CONCLUSIONS: If ROSE is not requested for CT-guided lung biopsies, proceduralists should err on the side of performing more, rather than fewer, passes, particularly for smaller lesions.

A Pineal Parenchymal Tumor of Intermediate Differentiation in an Octogenarian Contains a Rare KBTBD4 Insertion.

Pineal parenchymal tumors are rare central nervous system tumors that pose diagnostic challenges for surgical pathologists. Due to their paucity, their clinicopathologic features are still being defined. We report an 86-year-old woman with a remote history of breast lobular carcinoma who presented with a 2-month neurologic history that included gait instability, blurry vision, and headaches. Magnetic resonance imaging revealed a lobular, heterogeneously enhancing pineal region mass compressing the aqueduct of Sylvius. A biopsy performed concomitant with endoscopic third ventriculostomy consisted of small sheets of cells with eosinophilic to clear cytoplasm, multipolar processes, and ovoid nuclei with stippled chromatin. Whole exome sequencing revealed a small in-frame insertion (duplication) in exon 4 of KBTBD4 (c.931_939dup, p.P311_R313dup/ p.R313_M314insPRR), which has very recently been reported in 2 pineal parenchymal tumors of intermediate differentiation (PPTID). Additionally, variants of uncertain significance in CEBPA (c.863G > C, p.R288P) and MYC (c.655T > C, p.S219P) were identified. Although PPTID is considered a disease of young adulthood, review of 2 institutional cohorts of patients with pineal region tumors revealed that 25% of individuals with PPTID were over 65 years of age. In conclusion, PPTID should be considered in the differential diagnosis of pineal region tumors in older adults.

Recommendations for Cell-Free DNA Assay Validations: A Joint Consensus Recommendation of the Association for Molecular Pathology and College of American Pathologists.

Diagnosing, selecting therapy for, and monitoring cancer in patients using a minimally invasive blood test represents a significant advance in precision medicine. Wide variability exists in how circulating tumor DNA (ctDNA) assays are developed, validated, and reported in the literature, which hinders clinical adoption and may negatively impact patient care. Standardization is needed for factors affecting ctDNA assay performance and reporting, including pre-analytical variables, analytical considerations, and elements of laboratory assay reporting. The Association for Molecular Pathology Clinical Practice Committee's Liquid Biopsy Working Group (LBxWG), including organizational representation from the American Society of Clinical Oncology and the College of American Pathologists, has undertaken a full-text data extraction of 1228 ctDNA publications that describe assays performed in patients with lymphoma and solid tumor malignancies. With an emphasis on clinical assay validation, the LBxWG has developed a set of 13 best practice consensus recommendations for validating, reporting, and publishing clinical ctDNA assays. Recommendations include reporting key pre-analytical considerations and assay performance metrics; this analysis demonstrates these elements are inconsistently included in publications. The LBxWG recommendations are intended to assist clinical laboratories with validating and reporting ctDNA assays and to ensure high-quality data are included in publications. It is expected that these recommendations will need to be updated as the body of literature continues to mature.

RNA Sequencing for Solid Tumor Fusion Gene Detection: Proficiency Testing Practice and Performance Comparison.

CONTEXT.­: Next-generation sequencing-based approaches using RNA have increasingly been used by clinical laboratories for the detection of fusion genes, intragenic rearrangements, and exon-skipping events. Correspondingly, the College of American Pathologists (CAP) has advanced RNA sequencing proficiency testing (PT) to ensure optimal performance of these assays. OBJECTIVE.­: To report on laboratory performance and practices of RNA sequencing for the detection of fusion genes, intragenic rearrangements, and exon-skipping events using CAP PT data from 8 mailings (2018-A through 2021-B). DESIGN.­: CAP PT RNA sequencing program results from 153 laboratories across 24 proficiency test specimens, interrogating 22 distinct engineered fusion transcripts, were analyzed for correct identification of the fusion event, associated performance variables, and laboratory practices. RESULTS.­: Overall, the 4-year program detection rate (sensitivity) was 95.5% (1486 of 1556 results). False-negative rates were 3.6% (53 of 1463) and 18.3% (17 of 93) for fusion gene and intragenic rearrangement/exon-skipping events, respectively. Only 19 false-positive results were reported among the 8 PT mailings, and most were likely the result of preanalytical or postanalytical errors. There were no practice characteristics (eg, instrumentation, sequencing method) significantly associated with the fusion detection results. CONCLUSIONS.­: These data reveal a high overall sensitivity and specificity for fusion gene detection by participating laboratories using clinical RNA sequencing. Performance was comparable across all laboratories, regardless of methodology. The fraction of false-negative results for intragenic rearrangement/exon-skipping events was greater than that for the chimeric fusion genes. False-negative results could not be attributed to any specific practice characteristics.

Cutaneous myoepithelioma with EWSR1 gene rearrangement and its differentials: A diagnostic challenge.

Cutaneous myoepithelioma is a rare benign soft tissue neoplasm of myoepithelial cells involving the skin and subcutis. These tumors can be diagnostically challenging. The plasticity of myoepithelial cells leads to wide variability in the cytomorphology, immunophenotype, and genetic features of myoepithelioma. Their protean presentations may mimic malignant neoplasms. Therefore, distinction from malignancy is essential. Herein, we report a case of cutaneous myoepithelioma presenting similarly to Ewing sarcoma, with small round blue cells and an EWSR1 rearrangement. Our case highlights the important morphologic, immunohistochemical, and cytogenetic features of this benign basaloid cutaneous tumor.

Fluorescence molecular optomic signatures improve identification of tumors in head and neck specimens.

Background: Fluorescence molecular imaging using ABY-029, an epidermal growth factor receptor (EGFR)-targeted, synthetic Affibody peptide labeled with a near-infrared fluorophore, is under investigation for surgical guidance during head and neck squamous cell carcinoma (HNSCC) resection. However, tumor-to-normal tissue contrast is confounded by intrinsic physiological limitations of heterogeneous EGFR expression and non-specific agent uptake. Objective: In this preliminary study, radiomic analysis was applied to optical ABY-029 fluorescence image data for HNSCC tissue classification through an approach termed "optomics." Optomics was employed to improve tumor identification by leveraging textural pattern differences in EGFR expression conveyed by fluorescence. The study objective was to compare the performance of conventional fluorescence intensity thresholding and optomics for binary classification of malignant vs. non-malignant HNSCC tissues. Materials and Methods: Fluorescence image data collected through a Phase 0 clinical trial of ABY-029 involved a total of 20,073 sub-image patches (size of 1.8 × 1.8 mm2) extracted from 24 bread-loafed slices of HNSCC surgical resections originating from 12 patients who were stratified into three dose groups (30, 90, and 171 nanomoles). Each dose group was randomly partitioned on the specimen-level 75%/25% into training/testing sets, then all training and testing sets were aggregated. A total of 1,472 standardized radiomic features were extracted from each patch and evaluated by minimum redundancy maximum relevance feature selection, and 25 top-ranked features were used to train a support vector machine (SVM) classifier. Predictive performance of the SVM classifier was compared to fluorescence intensity thresholding for classifying testing set image patches with histologically confirmed malignancy status. Results: Optomics provided consistent improvement in prediction accuracy and false positive rate (FPR) and similar false negative rate (FNR) on all testing set slices, irrespective of dose, compared to fluorescence intensity thresholding (mean accuracies of 89% vs. 81%, P = 0.0072; mean FPRs of 12% vs. 21%, P = 0.0035; and mean FNRs of 13% vs. 17%, P = 0.35). Conclusions: Optomics outperformed conventional fluorescence intensity thresholding for tumor identification using sub-image patches as the unit of analysis. Optomics mitigate diagnostic uncertainties introduced through physiological variability, imaging agent dose, and inter-specimen biases of fluorescence molecular imaging by probing textural image information. This preliminary study provides a proof-of-concept that applying radiomics to fluorescence molecular imaging data offers a promising image analysis technique for cancer detection in fluorescence-guided surgery.

Utility of Retrospective Molecular Analysis in Diagnostically Challenging Mesenchymal Neoplasms.

Introduction: Molecular analysis plays a growing role in the diagnosis of mesenchymal neoplasms. The aim of this study was to retrospectively apply broad, multiplex molecular assays (a solid tumor targeted next-generation sequencing [NGS]) assay and single nucleotide polymorphism [SNP] microarray) to selected tumors, exploring the current utility and limitations. Methods: We searched our database (2010-2020) for diagnostically challenging mesenchymal neoplasms. After histologic review of available slides, tissue blocks were selected for NGS, SNP microarray, or both. DNA and RNA were extracted using the AllPrep DNA/RNA FFPE Kit Protocol on the QIAcube instrument. The NGS platform used was the TruSight Tumor 170 (TST-170). For SNP array, copy number variant (CNV) analysis was performed using the OncoScanTM CNV Plus Assay. Results: DNA/RNA was successfully extracted from 50% of tumors (n = 10/20). Specimens not successfully extracted included 6 core biopsies, 3 incisional biopsies, and 1 resection; 4 were decalcified (3 hydrochloric acid, 1 ethylenediaminetetraacetic acid). Higher tumor proportion and number of tumor cells were parameters positively associated with sufficient DNA/RNA extraction whereas necrosis and decalcification were negatively associated with sufficient extraction. Molecular testing helped reach a definitive diagnosis in 50% of tumors (n = 5/10). Conclusions: Although the overall utility of this approach is limited, these molecular panels can be helpful in detecting a specific "driver" alteration.

Current Laboratory Testing Practices for Assessment of ERBB2/HER2 in Endometrial Serous Carcinoma and Colorectal Carcinoma.

CONTEXT.­: Therapy targeted at human epidermal growth factor receptor 2 (HER2; also known as ERBB2) was used initially for breast and gastroesophageal carcinoma and has more recently been adopted for endometrial serous carcinoma (ESC) and colorectal carcinoma (CRC). There is evidence that predictive biomarker testing algorithms for HER2 must be tumor type specific and that an algorithm validated for one tumor type cannot be applied to another. OBJECTIVE.­: To describe current laboratory practices for HER2 assessment in ESC and CRC. DESIGN.­: We surveyed laboratories participating in the 2021 College of American Pathologists (CAP) HER2 immunohistochemistry proficiency testing program. RESULTS.­: The survey was distributed to 1548 laboratories and returned by 1195, of which 83.5% (998) were in the United States. For ESC, 24.0% (287) of laboratories reported performing in-house testing for HER2 by immunohistochemical staining and/or in situ hybridization; of these, 44.3% (127) performed it reflexively on all cases of ESC. The most common criterion for evaluating HER2 was the American Society of Clinical Oncology/CAP 2018 guideline for breast carcinoma (69.0%; 194 of 281), whereas only 16.0% (45) of laboratories used guidelines specific to ESC. For CRC, 20.2% (239 of 1185) of laboratories performed in-house HER2 testing, and 82.0% of these (196) did the test only at the clinician's request. A plurality (49.4%; 115 of 233) used gastroesophageal cancer guidelines when scoring CRC, 30.0% (70) used the CRC scoring system from the HERACLES trial, and 16.3% (38) used the American Society of Clinical Oncology/CAP 2018 guideline for breast carcinoma. CONCLUSIONS.­: Laboratories vary in their approach to HER2 testing in ESC and CRC. Most laboratories did not report using tumor type-specific recommendations for HER2 interpretation. The lack of standardization could present a challenge to evidence-based practice when considering targeted therapy for these diseases.

Evaluation of tumor microenvironment and biomarkers of immune checkpoint inhibitor response in metastatic renal cell carcinoma.

BACKGROUND: Immunotherapy combinations including ipilimumab and nivolumab are now the standard of care for untreated metastatic renal cell carcinoma (mRCC). Biomarkers of response are lacking to predict patients who will have a favorable or unfavorable response to immunotherapy. This study aimed to use the OmniSeq transcriptome-based platform to develop biomarkers of response to immunotherapy. METHODS: Two cohorts of patients were retrospectively collected. These included an investigational cohort of patients with mRCC treated with immune checkpoint inhibitor therapy from five institutions, and a subsequent validation cohort of patients with mRCC treated with combination ipilimumab and nivolumab from two institutions (Duke Cancer Institute and Cleveland Clinic Taussig Cancer Center). Tissue-based RNA sequencing was performed using the OmniSeq Immune Report Card on banked specimens to identify gene signatures and immune checkpoints associated with differential clinical outcomes. A 5-gene expression panel was developed based on the investigational cohort and was subsequently evaluated in the validation cohort. Clinical outcomes including progression-free survival (PFS) and overall survival (OS) were extracted by retrospective chart review. Objective response rate (ORR) was assessed by Response Evaluation Criteria in Solid Tumors (RECIST) V.1.1. RESULTS: The initial investigation cohort identified 86 patients with mRCC who received nivolumab (80%, 69/86), ipilimumab/nivolumab (14%, 12/86), or pembrolizumab (6%, 5/86). A gene expression score was created using the top five genes found in responders versus non-responders (FOXP3, CCR4, KLRK1, ITK, TIGIT). The ORR in patients with high gene expression (GEhigh) on the 5-gene panel was 29% (14/48), compared with low gene expression (GElow) 3% (1/38, χ2 p=0.001). The validation cohort was comprised of 62 patients who received ipilimumab/nivolumab. There was no difference between GEhigh and GElow in terms of ORR (44% vs 38.5%), PFS (HR 1.5, 95% CI 0.58 to 3.89), or OS (HR 0.96, 95% CI 0.51 to 1.83). Similarly, no differences in ORR, PFS or OS were observed when patients were stratified by tumor mutational burden (high=top 20%), PD-L1 (programmed death-ligand 1) expression by immunohistochemistry or RNA expression, or CTLA-4 (cytotoxic T-lymphocytes-associated protein 4) RNA expression. The International Metastatic RCC Database Consortium (IMDC) risk score was prognostic for OS but not PFS. CONCLUSION: A 5-gene panel that was associated with improved ORR in a predominantly nivolumab monotherapy population of patients with mRCC was not predictive for radiographic response, PFS, or OS among patients with mRCC treated with ipilimumab and nivolumab.

Predicting oncogene mutations of lung cancer using deep learning and histopathologic features on whole-slide images.

Lung cancer is a leading cause of death in both men and women globally. The recent development of tumor molecular profiling has opened opportunities for targeted therapies for lung adenocarcinoma (LUAD) patients. However, the lack of access to molecular profiling or cost and turnaround time associated with it could hinder oncologists' willingness to order frequent molecular tests, limiting potential benefits from precision medicine. In this study, we developed a weakly supervised deep learning model for predicting somatic mutations of LUAD patients based on formalin-fixed paraffin-embedded (FFPE) whole-slide images (WSIs) using LUAD subtypes-related histological features and recent advances in computer vision. Our study was performed on a total of 747 hematoxylin and eosin (H&E) stained FFPE LUAD WSIs and the genetic mutation data of 232 patients who were treated at Dartmouth-Hitchcock Medical Center (DHMC). We developed our convolutional neural network-based models to analyze whole slides and predict five major genetic mutations, i.e., BRAF, EGFR, KRAS, STK11, and TP53. We additionally used 111 cases from the LUAD dataset of the CPTAC-3 study for external validation. Our model achieved an AUROC of 0.799 (95% CI: 0.686-0.904) and 0.686 (95% CI: 0.620-0.752) for predicting EGFR genetic mutations on the DHMC and CPTAC-3 test sets, respectively. Predicting TP53 genetic mutations also showed promising outcomes. Our results demonstrated that H&E stained FFPE LUAD whole slides could be utilized to predict oncogene mutations, such as EGFR, indicating that somatic mutations could present subtle morphological characteristics in histology slides, where deep learning-based feature extractors can learn such latent information.

Primary Pulmonary Round Cell Sarcomas: Multiple Potential Pitfalls for the Pathologist.

Primary sarcomas of the lung are extremely uncommon. A diverse group of round cell sarcomas has been reported to originate in this location, including Ewing sarcoma, desmoplastic small round cell tumor, rhabdomyosarcoma, and poorly differentiated synovial sarcoma. The rarity of these tumors presents a potential pitfall; without careful study, they may easily be misidentified as the significantly more common poorly differentiated carcinoma. While histomorphology is a key aspect of correctly identifying a sarcoma, ancillary testing has become increasingly important in making a definitive diagnosis, as more and more recurrent genetic alterations are discovered and new entities are defined. We present three cases of primary round cell sarcomas of the lung that proved diagnostically challenging, describe the features and ancillary testing that led to the correct diagnoses, and discuss classic and evolving entities among sarcomas with round cell morphology.

Molecular Pathology Education: A Suggested Framework for Primary Care Resident Training in Genomic Medicine: A Report of the Association for Molecular Pathology Training and Education Committee.

Developments in genomics are profoundly influencing medical practice. With increasing use of genetic and genomic testing across every aspect of the health care continuum, patients and their families are increasingly turning to primary care physicians (PCPs) for discussion and advice regarding tests, implications, and results. Yet, with the rapid growth of information, technology, and applications, PCPs are finding it challenging to fill the gaps in knowledge and support the growing needs of their patients. A critical component in expanding PCP genomic literacy lies in the education of physicians in training and in practice. Although a framework for developing physician competencies in genomics has already been developed, the Association for Molecular Pathology is uniquely situated to actively utilize the skills of its members to engage and support PCPs in this effort. This report provides an overview and a suggested basic teaching framework, which can be used by molecular professionals in their individual institutions as a starting point for educational outreach.

Using molecular testing to improve the management of thyroid nodules with indeterminate cytology: an institutional experience with review of molecular alterations.

INTRODUCTION: Molecular testing has helped clinicians and cytopathologists to further categorize indeterminate thyroid fine needle aspiration (FNA) specimens. The purpose of the present study was to evaluate the accuracy of commercially available molecular tests, review their effects on patient treatment, and correlate the molecular alterations with the histologic findings. MATERIALS AND METHODS: A pathology laboratory information system search identified thyroid FNAs performed at our institution between January 1, 2015 and June 30, 2020. The results of surgical follow-up and ancillary molecular testing were collected. We evaluated the accuracy of these tests and whether they could reduce the number of surgeries performed. RESULTS: Our laboratory information system search identified 510 cases reported as atypia of undetermined significance, 94 as suspicious for follicular neoplasm, and 44 as suspicious for follicular neoplasm, Hurthle cell type. Of the specimens, 343 had no ancillary molecular testing, 146 were sent for ThyGenX/ThyraMIR, and 136 were sent for ThyroSeq. Of the patients without molecular testing, 50.4% had undergone follow-up surgery compared with 30.8% after ThyGenX/ThyraMIR and 38.2% after ThyroSeq testing, resulting in 38.9% and 24.2% fewer surgeries and an odds ratio of 0.04 (95% confidence interval, 0.00-0.33) and 0.14 (95% confidence interval, 0.01-0.95), respectively. For ThyGenX/ThyraMIR testing, the risk of malignancy for high and moderate to high risk alterations was 80%, 28.6% for moderate and low to moderate risk alterations, and 23.1% for low risk alterations. For ThyroSeq, the risk of malignancy was 87.5% for high risk alterations, 36.8% for intermediate to high risk alterations, 27.3% for intermediate risk alterations, and 0% for low risk alterations. The areas under the curve for ThyGenX/ThyraMIR and ThyroSeq testing were 0.65 and 0.85, respectively. CONCLUSIONS: These findings suggest that, at our institution, both ThygenX/ThyraMIR and ThyroSeq can be used to effectively stratify cytology specimens based on the risk of malignancy and reduce the number of surgeries performed at our institution.

Resolution-based distillation for efficient histology image classification.

Developing deep learning models to analyze histology images has been computationally challenging, as the massive size of the images causes excessive strain on all parts of the computing pipeline. This paper proposes a novel deep learning-based methodology for improving the computational efficiency of histology image classification. The proposed approach is robust when used with images that have reduced input resolution, and it can be trained effectively with limited labeled data. Moreover, our approach operates at either the tissue- or slide-level, removing the need for laborious patch-level labeling. Our method uses knowledge distillation to transfer knowledge from a teacher model pre-trained at high resolution to a student model trained on the same images at a considerably lower resolution. Also, to address the lack of large-scale labeled histology image datasets, we perform the knowledge distillation in a self-supervised fashion. We evaluate our approach on three distinct histology image datasets associated with celiac disease, lung adenocarcinoma, and renal cell carcinoma. Our results on these datasets demonstrate that a combination of knowledge distillation and self-supervision allows the student model to approach and, in some cases, surpass the teacher model's classification accuracy while being much more computationally efficient. Additionally, we observe an increase in student classification performance as the size of the unlabeled dataset increases, indicating that there is potential for this method to scale further with additional unlabeled data. Our model outperforms the high-resolution teacher model for celiac disease in accuracy, F1-score, precision, and recall while requiring 4 times fewer computations. For lung adenocarcinoma, our results at 1.25× magnification are within 1.5% of the results for the teacher model at 10× magnification, with a reduction in computational cost by a factor of 64. Our model on renal cell carcinoma at 1.25× magnification performs within 1% of the teacher model at 5× magnification while requiring 16 times fewer computations. Furthermore, our celiac disease outcomes benefit from additional performance scaling with the use of more unlabeled data. In the case of 0.625× magnification, using unlabeled data improves accuracy by 4% over the tissue-level baseline. Therefore, our approach can improve the feasibility of deep learning solutions for digital pathology on standard computational hardware and infrastructures.

Endometrial Cancer Suppresses CD8+ T Cell-Mediated Cytotoxicity in Postmenopausal Women.

Endometrial cancer is the most common gynecological cancer. To investigate how it suppresses host immune function, we isolated CD8+ T cells from endometrial endometroid carcinomas and adjacent non-cancerous endometrium and determined if the tumor environment regulates cytotoxic capacity. Endometrial carcinomas had increased numbers of CD8+ T cells compared to adjacent non-cancerous endometrium. Tumor CD8+ T cells expressed significantly less granzyme A (GZA), B (GZB), and PD-1 than those in adjacent non-cancerous tissues and also had significantly lower cytotoxic killing of allogeneic target cells. CD103-CD8+ T cells, but not CD103+CD8+ T cells, from both adjacent and tumor tissue were primarily responsible for killing of allogeneic target cells. Secretions recovered from endometrial carcinoma tissues suppressed CD8+ cytotoxic killing and lowered perforin, GZB and PD-1 expression relative to non-tumor CD8+ T cells. Furthermore, tumor secretions contained significantly higher levels of immunosuppressive cytokines including TGFß than non-tumor tissues. Thus, the tumor microenvironment suppresses cytotoxic killing by CD8+ T cells via the secretion of immunosuppressive cytokines leading to decreased expression of intracellular cytolytic molecules. These studies demonstrate the complexity of CD8+ T cell regulation within the endometrial tumor microenvironment and provide a foundation of information essential for the development of therapeutic strategies for gynecological cancers.

Clinicopathologic and Genomic Characterization of Inflammatory Myofibroblastic Tumors of the Head and Neck: Highlighting a Novel Fusion and Potential Diagnostic Pitfall.

Inflammatory myofibroblastic tumor (IMT) is a distinctive fibroblastic and myofibroblastic spindle cell neoplasm with an accompanying inflammatory cell infiltrate and frequent receptor tyrosine kinase activation at the molecular level. The tumor may recur and rarely metastasizes. IMT is rare in the head and neck region, and limited information is available about its clinicopathologic and molecular characteristics in these subsites. Therefore, we analyzed a cohort of head and neck IMTs through a multi-institutional approach. Fourteen cases were included in the provisional cohort, but 1 was excluded after molecular analysis prompted reclassification. Patients in the final cohort included 7 males and 6 females, with a mean age of 26.5 years. Tumors were located in the larynx (n=7), oral cavity (n=3), pharynx (n=2), and mastoid (n=1). Histologically, all tumors showed neoplastic spindle cells in storiform to fascicular patterns with associated chronic inflammation, but the morphologic spectrum was wide, as is characteristic of IMT in other sites. An underlying fusion gene event was identified in 92% (n=11/12) of cases and an additional case was ALK-positive by IHC but could not be evaluated molecularly. ALK represented the driver in all but 1 case. Rearrangement of ALK, fused with the TIMP3 gene (n=6) was most commonly detected, followed by 1 case each of the following fusion gene partnerships: TPM3-ALK, KIF5B-ALK, CARS-ALK, THBS1-ALK, and a novel alteration, SLC12A2-ROS1. The excluded case was reclassified as spindle cell rhabdomyosarcoma after detection of a FUS-TFCP2 rearrangement and retrospective immunohistochemical confirmation of rhabdomyoblastic differentiation, illustrating an important diagnostic pitfall. Two IMT patients received targeted therapy with crizotinib, with a demonstrated radiographic response. One tumor recurred but none metastasized. These results add to the growing body of evidence that kinase fusions can be identified in the majority of IMTs and that molecular analysis can lead to increased diagnostic accuracy and broadened therapeutic options for patients.

Sensitivity of cells to ATR and CHK1 inhibitors requires hyperactivation of CDK2 rather than endogenous replication stress or ATM dysfunction.

DNA damage activates cell cycle checkpoint proteins ATR and CHK1 to arrest cell cycle progression, providing time for repair and recovery. Consequently, inhibitors of ATR (ATRi) and CHK1 (CHK1i) enhance damage-induced cell death. Intriguingly, both CHK1i and ATRi alone elicit cytotoxicity in some cell lines. Sensitivity has been attributed to endogenous replications stress, but many more cell lines are sensitive to ATRi than CHK1i. Endogenous activation of the DNA damage response also did not correlate with drug sensitivity. Sensitivity correlated with the appearance of γH2AX, a marker of DNA damage, but without phosphorylation of mitotic markers, contradicting suggestions that the damage is due to premature mitosis. Sensitivity to ATRi has been associated with ATM mutations, but dysfunction in ATM signaling did not correlate with sensitivity. CHK1i and ATRi circumvent replication stress by reactivating stalled replicons, a process requiring a low threshold activity of CDK2. In contrast, γH2AX induced by single agent ATRi and CHK1i requires a high threshold activity CDK2. Hence, phosphorylation of different CDK2 substrates is required for cytotoxicity induced by replication stress plus ATRi/CHK1i as compared to their single agent activity. In summary, sensitivity to ATRi and CHK1i as single agents is elicited by premature hyper-activation of CDK2.