Development and validation of a deep learning-based microsatellite instability predictor from prostate cancer whole-slide images.

Microsatellite instability-high (MSI-H) is a tumor-agnostic biomarker for immune checkpoint inhibitor therapy. However, MSI status is not routinely tested in prostate cancer, in part due to low prevalence and assay cost. As such, prediction of MSI status from hematoxylin and eosin (H&E) stained whole-slide images (WSIs) could identify prostate cancer patients most likely to benefit from confirmatory testing to evaluate their eligibility for immunotherapy and need for Lynch syndrome testing. Prostate biopsies and surgical resections from prostate cancer patients referred to our institution were analyzed. MSI status was determined by next-generation sequencing. Patients sequenced before a cutoff date formed an algorithm development set (n = 4015, MSI-H 1.8%) and a paired validation set (n = 173, MSI-H 19.7%) that consisted of two serial sections from each sample, one stained and scanned internally and the other at an external site. Patients sequenced after the cutoff date formed a temporally independent validation set (n = 1350, MSI-H 2.3%). Attention-based multiple instance learning models were trained to predict MSI-H from H&E WSIs. The predictor achieved area under the receiver operating characteristic curve values of 0.78 (95% CI [0.69-0.86]), 0.72 (95% CI [0.63-0.81]), and 0.72 (95% CI [0.62-0.82]) on the internally prepared, externally prepared, and temporal validation sets, respectively, showing effective predictability and generalization to both external staining/scanning processes and temporally independent samples. While MSI-H status is significantly correlated with Gleason score, the model remained predictive within each Gleason score subgroup.

Prediction of MET Overexpression in Lung Adenocarcinoma from Hematoxylin and Eosin Images.

Mesenchymal epithelial transition (MET) protein overexpression is a targetable event in non-small cell lung cancer and is the subject of active drug development. Challenges in identifying patients for these therapies include lack of access to validated testing, such as standardized immunohistochemistry assessment, and consumption of valuable tissue for a single gene/protein assay. Development of prescreening algorithms using routinely available digitized hematoxylin and eosin (H&E)-stained slides to predict MET overexpression could promote testing for those who will benefit most. Recent literature reports a positive correlation between MET protein overexpression and RNA expression. In this work, a large database of matched H&E slides and RNA expression data were leveraged to train a weakly supervised model to predict MET RNA overexpression directly from H&E images. This model was evaluated on an independent holdout test set of 300 overexpressed and 289 normal patients, demonstrating a receiver operating characteristic area under curve of 0.70 (95th percentile interval: 0.66 to 0.74) with stable performance characteristics across different patient clinical variables and robust to synthetic noise on the test set. These results suggest that H&E-based predictive models could be useful to prioritize patients for confirmatory testing of MET protein or MET gene expression status.

Artificial intelligence-augmented histopathologic review using image analysis to optimize DNA yield from formalin-fixed paraffin-embedded slides.

To achieve minimum DNA input requirements for next-generation sequencing (NGS), pathologists visually estimate macrodissection and slide count decisions. Unfortunately, misestimation may cause tissue waste and increased laboratory costs. We developed an artificial intelligence (AI)-augmented smart pathology review system (SmartPath) to empower pathologists with quantitative metrics for accurately determining tissue extraction parameters. SmartPath uses two deep learning architectures, a U-Net based network for cell segmentation and a multi-field-of-view convolutional network for tumor area segmentation, to extract features from digitized H&E-stained formalin-fixed paraffin-embedded slides. From the segmented tumor area, SmartPath suggests a macrodissection area. To predict DNA yield per slide, the extracted features from within the macrodissection area are correlated with known DNA yields to fit a regularized linear model (R = 0.85). Then, a pathologist-defined target yield divided by the predicted DNA yield per slide gives the number of slides to scrape. Following model development, an internal validation trial was conducted within the Tempus Labs molecular sequencing laboratory. We evaluated our system on 501 clinical colorectal cancer slides, where half received SmartPath-augmented review and half traditional pathologist review. The SmartPath cohort had 25% more DNA yields within a desired target range of 100-2000 ng. The number of extraction attempts was statistically unchanged between cohorts. The SmartPath system recommended fewer slides to scrape for large tissue sections, saving tissue in these cases. Conversely, SmartPath recommended more slides to scrape for samples with scant tissue sections, especially those with degraded DNA, helping prevent costly re-extraction due to insufficient extraction yield. A statistical analysis was performed to measure the impact of covariates on the results, offering insights on how to improve future applications of SmartPath. With these improvements, AI-augmented histopathologic review has the potential to decrease tissue waste, sequencing time, and laboratory costs by optimizing DNA yields, especially for samples with scant tissue and/or degraded DNA.

A predictive tool for identification of SARS-CoV-2 PCR-negative emergency department patients using routine test results.

BACKGROUND: Testing for COVID-19 remains limited in the United States and across the world. Poor allocation of limited testing resources leads to misutilization of health system resources, which complementary rapid testing tools could ameliorate. OBJECTIVE: To predict SARS-CoV-2 PCR positivity based on complete blood count components and patient sex. STUDY DESIGN: A retrospective case-control design for collection of data and a logistic regression prediction model was used. Participants were emergency department patients > 18 years old who had concurrent complete blood counts and SARS-CoV-2 PCR testing. 33 confirmed SARS-CoV-2 PCR positive and 357 negative patients at Stanford Health Care were used for model training. Validation cohorts consisted of emergency department patients > 18 years old who had concurrent complete blood counts and SARS-CoV-2 PCR testing in Northern California (41 PCR positive, 495 PCR negative), Seattle, Washington (40 PCR positive, 306 PCR negative), Chicago, Illinois (245 PCR positive, 1015 PCR negative), and South Korea (9 PCR positive, 236 PCR negative). RESULTS: A decision support tool that utilizes components of complete blood count and patient sex for prediction of SARS-CoV-2 PCR positivity demonstrated a C-statistic of 78 %, an optimized sensitivity of 93 %, and generalizability to other emergency department populations. By restricting PCR testing to predicted positive patients in a hypothetical scenario of 1000 patients requiring testing but testing resources limited to 60 % of patients, this tool would allow a 33 % increase in properly allocated resources. CONCLUSIONS: A prediction tool based on complete blood count results can better allocate SARS-CoV-2 testing and other health care resources such as personal protective equipment during a pandemic surge.

Pharma-Oncogenomics in the Era of Personal Genomics: A Quick Guide to Online Resources and Tools.

The past two decades have seen unprecedented advances in the field of oncogenomics. The ongoing characterization of neoplastic tissues through genomic techniques has transformed many aspects of cancer research, diagnosis, and treatment. However, identifying sequence variants with biological and clinical significance is a challenging endeavor. In order to accomplish this task, variants must be annotated and interpreted using various online resources. Data on protein structure, functional prediction, variant frequency in relevant populations, and multiple other factors have been compiled in useful databases for this purpose. Thus, understanding the available online resources for the annotation and interpretation of sequence variants is critical to aid molecular pathologists and researchers working in this space.

A Survey of Somatic Mutations in 41 Genes in a Cohort of T-Cell Lymphomas Identifies Frequent Mutations in Genes Involved in Epigenetic Modification.

Here, we utilize a high throughput sequencing panel that covers several genes known to be recurrently mutated in certain T-cell lymphoma subtypes as well as genes frequently mutated in other hematolymphoid malignancies, including myeloid neoplasms. This panel was applied to formalin-fixed, paraffin-embedded tissue from 84 biopsies from 78 patients selected for this study. The biopsies included ones a with a diagnosis of T-cell lymphoma (n=79), including peripheral T-cell lymphoma not otherwise specified (PTCL-NOS; n=26) and angioimmunoblastic T-cell lymphoma (AITL; n=13), as well as 5 cases of atypical T-cell proliferations. KMT2C and KMT2D, which code for proteins involved in histone modifications, were the 2 most frequently mutated genes in our cohort and were altered across a range T-cell lymphomas. Mutations in TET2 and DNMT3A, which are involved in regulating DNA methylation, were also found in a variety of T-cell lymphoma categories. The RHOA G17V mutation that is frequently found in AITL was identified 5 of 13 (40%) cases of AITL and in 3 of 26 (12%) cases of PTCL-NOS, but not in biopsies involved by other T-cell proliferations. Our study adds to the already significant evidence from other investigators that, among T-cell lymphomas, the RHOA G17V variant is specific for AITL and PTCL-NOS. In contrast, variants in epigenetic modifier genes do not appear to be particularly specific for T-cell lymphoma subcategories evaluated in our study.

Multiplexed detection of viral infections using rapid in situ RNA analysis on a chip.

Viral infections are a major cause of human disease, but many require molecular assays for conclusive diagnosis. Current assays typically rely on RT-PCR or ELISA; however, these tests often have limited speed, sensitivity or specificity. Here, we demonstrate that rapid RNA FISH is a viable alternative method that could improve upon these limitations. We describe a platform beginning with software to generate RNA FISH probes both for distinguishing related strains of virus (even those different by a single base) and for capturing large numbers of strains simultaneously. Next, we present a simple fluidic device for reliably performing RNA FISH assays in an automated fashion. Finally, we describe an automated image processing pipeline to robustly identify uninfected and infected samples. Together, our results establish RNA FISH as a methodology with potential for viral point-of-care diagnostics.

Diacylglycerol kinase ζ limits the generation of natural regulatory T cells.

Natural regulatory T (nT(reg)) cells are important for maintaining tolerance to self- and foreign antigens, and they are thought to develop from thymocytes that receive strong T cell receptor (TCR)-mediated signals in the thymus. TCR engagement leads to the activation of phospholipase C-γ1, which generates the lipid second messenger diacylglycerol (DAG) from phosphatidylinositol 4,5-bisphosphate. We used mice that lack the ζ isoform of DAG kinase (DGKζ), which metabolizes DAG to terminate its signaling, to enhance TCR-mediated signaling and identify critical signaling events in nT(reg) cell development. Loss of DGKζ resulted in increased numbers of thymic CD25(+)Foxp3(-)CD4(+) nT(reg) cell precursors and Foxp3(+)CD4(+) nT(reg) cells in a cell-autonomous manner. DGKζ-deficient T cells exhibited increased nuclear translocation of the nuclear factor κB subunit c-Rel, as well as enhanced extracellular signal-regulated kinase (ERK) phosphorylation in response to TCR stimulation, suggesting that these downstream pathways may contribute to nT(reg) cell development. Indeed, reducing c-Rel abundance or blocking ERK phosphorylation abrogated the increased generation of nTreg cells by DGKζ-deficient thymocytes. The extent of ERK phosphorylation correlated with TCR-mediated acquisition of Foxp3 in immature thymocytes in vitro. Furthermore, the development of nT(reg) cells was augmented in mice in which ERK activation was selectively enhanced in T cells. Together, these data suggest that DGKζ regulates the development of nT(reg) cells by limiting the extent of activation of the ERK and c-Rel signaling pathways.

The ζ isoform of diacylglycerol kinase plays a predominant role in regulatory T cell development and TCR-mediated ras signaling.

Diacylglycerol (DAG) is a critical second messenger that mediates T cell receptor (TCR)-stimulated signaling. The abundance of DAG is reduced by the diacylglycerol kinases (DGKs), which catalyze the conversion of DAG to phosphatidic acid (PA) and thus inhibit DAG-mediated signaling. In T cells, the predominant DGK isoforms are DGKα and DGKζ, and deletion of the genes encoding either isoform enhances DAG-mediated signaling. We found that DGKζ, but not DGKα, suppressed the development of natural regulatory T (T(reg)) cells and predominantly mediated Ras and Akt signaling downstream of the TCR. The differential functions of DGKα and DGKζ were not attributable to differences in protein abundance in T cells or in their localization to the contact sites between T cells and antigen-presenting cells. RasGRP1, a key DAG-mediated activator of Ras signaling, associated to a greater extent with DGKζ than with DGKα; however, in silico modeling of TCR-stimulated Ras activation suggested that a difference in RasGRP1 binding affinity was not sufficient to cause differences in the functions of each DGK isoform. Rather, the model suggested that a greater catalytic rate for DGKζ than for DGKα might lead to DGKζ exhibiting increased suppression of Ras-mediated signals compared to DGKα. Consistent with this notion, experimental studies demonstrated that DGKζ was more effective than DGKα at catalyzing the metabolism of DAG to PA after TCR stimulation. The enhanced effective enzymatic production of PA by DGKζ is therefore one possible mechanism underlying the dominant functions of DGKζ in modulating T(reg) cell development.

Enhanced effector responses in activated CD8+ T cells deficient in diacylglycerol kinases.

Recent clinical trials have shown promise in the use of chimeric antigen receptor (CAR)-transduced T cells; however, augmentation of their activity may broaden their clinical use and improve their efficacy. We hypothesized that because CAR action requires proteins essential for T-cell receptor (TCR) signal transduction, deletion of negative regulators of these signaling pathways would enhance CAR signaling and effector T-cell function. We tested CAR activity and function in T cells that lacked one or both isoforms of diacylglycerol kinase (dgk) expressed highly in T cells, dgkα and dgkζ, enzymes that metabolize the second messenger diacylglycerol (DAG) and limit Ras/ERK activation. We found that primary murine T cells transduced with CARs specific for the human tumor antigen mesothelin showed greatly enhanced cytokine production and cytotoxicity when cocultured with a murine mesothelioma line that stably expresses mesothelin. In addition, we found that dgk-deficient CAR-transduced T cells were more effective in limiting the growth of implanted tumors, both concurrent with and after establishment of tumor. Consistent with our studies in mice, pharmacologic inhibition of dgks also augments function of primary human T cells transduced with CARs. These results suggest that deletion of negative regulators of TCR signaling enhances the activity and function of CAR-expressing T cells and identify dgks as potential targets for improving the clinical potential of CARs.

Diacylglycerol kinases: regulated controllers of T cell activation, function, and development.

Diacylglycerol kinases (DGKs) are a diverse family of enzymes that catalyze the conversion of diacylglycerol (DAG), a crucial second messenger of receptor-mediated signaling, to phosphatidic acid (PA). Both DAG and PA are bioactive molecules that regulate a wide set of intracellular signaling proteins involved in innate and adaptive immunity. Clear evidence points to a critical role for DGKs in modulating T cell activation, function, and development. More recently, studies have elucidated factors that control DGK function, suggesting an added complexity to how DGKs act during signaling. This review summarizes the available knowledge of the function and regulation of DGK isoforms in signal transduction with a particular focus on T lymphocytes.

AKT1 and AKT2 maintain hematopoietic stem cell function by regulating reactive oxygen species.

Although AKT is essential for multiple cellular functions, the role of this kinase family in hematopoietic stem cells (HSCs) is unknown. Thus, we analyzed HSC function in mice deficient in the 2 isoforms most highly expressed in the hematopoietic compartment, AKT1 and AKT2. Although loss of either isoform had only a minimal effect on HSC function, AKT1/2 double-deficient HSCs competed poorly against wild-type cells in the development of myeloid and lymphoid cells in in vivo reconstitution assays. Serial transplantations revealed an essential role for AKT1 and AKT2 in the maintenance of long-term HSCs (LT-HSCs). AKT1/2 double-deficient LT-HSCs were found to persist in the G(0) phase of the cell cycle, suggesting that the long-term functional defects are caused by increased quiescence. Furthermore, we found that the intracellular content of reactive oxygen species (ROS) is dependent on AKT because double-deficient HSCs demonstrate decreased ROS. The importance of maintaining ROS for HSC differentiation was shown by a rescue of the differentiation defect after pharmacologically increasing ROS levels in double-deficient HSCs. These data implicate AKT1 and AKT2 as critical regulators of LT-HSC function and suggest that defective ROS homeostasis may contribute to failed hematopoiesis.

Lytic granule loading of CD8+ T cells is required for HIV-infected cell elimination associated with immune control.

Virus-specific CD8+ T cells probably mediate control over HIV replication in rare individuals, termed long-term nonprogressors (LTNPs) or elite controllers. Despite extensive investigation, the mechanisms responsible for this control remain incompletely understood. We observed that HIV-specific CD8+ T cells of LTNPs persisted at higher frequencies than those of treated progressors with equally low amounts of HIV. Measured on a per-cell basis, HIV-specific CD8+ T cells of LTNPs efficiently eliminated primary autologous HIV-infected CD4+ T cells. This function required lytic granule loading of effectors and delivery of granzyme B to target cells. Defective cytotoxicity of progressor effectors could be restored after treatment with phorbol ester and calcium ionophore. These results establish an effector function and mechanism that clearly segregate with immunologic control of HIV. They also demonstrate that lytic granule contents of memory cells are a critical determinant of cytotoxicity that must be induced for maximal per-cell killing capacity.