Prediction of DNA methylation-based tumor types from histopathology in central nervous system tumors with deep learning.

Precision in the diagnosis of diverse central nervous system (CNS) tumor types is crucial for optimal treatment. DNA methylation profiles, which capture the methylation status of thousands of individual CpG sites, are state-of-the-art data-driven means to enhance diagnostic accuracy but are also time consuming and not widely available. Here, to address these limitations, we developed Deep lEarning from histoPathoLOgy and methYlation (DEPLOY), a deep learning model that classifies CNS tumors to ten major categories from histopathology. DEPLOY integrates three distinct components: the first classifies CNS tumors directly from slide images ('direct model'), the second initially generates predictions for DNA methylation beta values, which are subsequently used for tumor classification ('indirect model'), and the third classifies tumor types directly from routinely available patient demographics. First, we find that DEPLOY accurately predicts beta values from histopathology images. Second, using a ten-class model trained on an internal dataset of 1,796 patients, we predict the tumor categories in three independent external test datasets including 2,156 patients, achieving an overall accuracy of 95% and balanced accuracy of 91% on samples that are predicted with high confidence. These results showcase the potential future use of DEPLOY to assist pathologists in diagnosing CNS tumors within a clinically relevant short time frame.

Chromosome 7 to the rescue: overcoming chromosome 10 loss in gliomas.

The co-occurrence of chromosome 10 loss and chromosome 7 gain in gliomas is the most frequent loss-gain co-aneuploidy pair in human cancers, a phenomenon that has been investigated without resolution since the late 1980s. Expanding beyond previous gene-centric studies, we investigate the co-occurrence in a genome-wide manner taking an evolutionary perspective. First, by mining large tumor aneuploidy data, we predict that the more likely order is 10 loss followed by 7 gain. Second, by analyzing extensive genomic and transcriptomic data from both patients and cell lines, we find that this co-occurrence can be explained by functional rescue interactions that are highly enriched on 7, which can possibly compensate for any detrimental consequences arising from the loss of 10. Finally, by analyzing transcriptomic data from normal, non-cancerous, human brain tissues, we provide a plausible reason why this co-occurrence happens preferentially in cancers originating in certain regions of the brain.

Optimizing cancer immunotherapy response prediction by tumor aneuploidy score and fraction of copy number alterations.

Identifying patients that are likely to respond to cancer immunotherapy is an important, yet highly challenging clinical need. Using 3139 patients across 17 different cancer types, we comprehensively studied the ability of two common copy-number alteration (CNA) scores-the tumor aneuploidy score (AS) and the fraction of genome single nucleotide polymorphism encompassed by copy-number alterations (FGA)-to predict survival following immunotherapy in both pan-cancer and individual cancer types. First, we show that choice of cutoff during CNA calling significantly influences the predictive power of AS and FGA for patient survival following immunotherapy. Remarkably, by using proper cutoff during CNA calling, AS and FGA can predict pan-cancer survival following immunotherapy for both high-TMB and low-TMB patients. However, at the individual cancer level, our data suggest that the use of AS and FGA for predicting immunotherapy response is currently limited to only a few cancer types. Therefore, larger sample sizes are needed to evaluate the clinical utility of these measures for patient stratification in other cancer types. Finally, we propose a simple, non-parameterized, elbow-point-based method to help determine the cutoff used for calling CNAs.

MYC Induces Immunotherapy and IFNγ Resistance Through Downregulation of JAK2.

Immunotherapy has revolutionized the treatment of advanced melanoma. Because the pathways mediating resistance to immunotherapy are largely unknown, we conducted transcriptome profiling of preimmunotherapy tumor biopsies from patients with melanoma that received PD-1 blockade or adoptive cell therapy with tumor-infiltrating lymphocytes. We identified two melanoma-intrinsic, mutually exclusive gene programs, which were controlled by IFNγ and MYC, and the association with immunotherapy outcome. MYC-overexpressing melanoma cells exhibited lower IFNγ responsiveness, which was linked with JAK2 downregulation. Luciferase activity assays, under the control of JAK2 promoter, demonstrated reduced activity in MYC-overexpressing cells, which was partly reversible upon mutagenesis of a MYC E-box binding site in the JAK2 promoter. Moreover, silencing of MYC or its cofactor MAX with siRNA increased JAK2 expression and IFNγ responsiveness of melanomas, while concomitantly enhancing the effector functions of T cells coincubated with MYC-overexpressing cells. Thus, we propose that MYC plays a pivotal role in immunotherapy resistance through downregulation of JAK2.

Clinically oriented prediction of patient response to targeted and immunotherapies from the tumor transcriptome.

BACKGROUND: Precision oncology is gradually advancing into mainstream clinical practice, demonstrating significant survival benefits. However, eligibility and response rates remain limited in many cases, calling for better predictive biomarkers. METHODS: We present ENLIGHT, a transcriptomics-based computational approach that identifies clinically relevant genetic interactions and uses them to predict a patient's response to a variety of therapies in multiple cancer types without training on previous treatment response data. We study ENLIGHT in two translationally oriented scenarios: personalized oncology (PO), aimed at prioritizing treatments for a single patient, and clinical trial design (CTD), selecting the most likely responders in a patient cohort. FINDINGS: Evaluating ENLIGHT's performance on 21 blinded clinical trial datasets in the PO setting, we show that it can effectively predict a patient's treatment response across multiple therapies and cancer types. Its prediction accuracy is better than previously published transcriptomics-based signatures and is comparable with that of supervised predictors developed for specific indications and drugs. In combination with the interferon-γ signature, ENLIGHT achieves an odds ratio larger than 4 in predicting response to immune checkpoint therapy. In the CTD scenario, ENLIGHT can potentially enhance clinical trial success for immunotherapies and other monoclonal antibodies by excluding non-responders while overall achieving more than 90% of the response rate attainable under an optimal exclusion strategy. CONCLUSIONS: ENLIGHT demonstrably enhances the ability to predict therapeutic response across multiple cancer types from the bulk tumor transcriptome. FUNDING: This research was supported in part by the Intramural Research Program, NIH and by the Israeli Innovation Authority.

A cell-type-specific atlas of the inner ear transcriptional response to acoustic trauma.

Noise-induced hearing loss (NIHL) results from a complex interplay of damage to the sensory cells of the inner ear, dysfunction of its lateral wall, axonal retraction of type 1C spiral ganglion neurons, and activation of the immune response. We use RiboTag and single-cell RNA sequencing to survey the cell-type-specific molecular landscape of the mouse inner ear before and after noise trauma. We identify induction of the transcription factors STAT3 and IRF7 and immune-related genes across all cell-types. Yet, cell-type-specific transcriptomic changes dominate the response. The ATF3/ATF4 stress-response pathway is robustly induced in the type 1A noise-resilient neurons, potassium transport genes are downregulated in the lateral wall, mRNA metabolism genes are downregulated in outer hair cells, and deafness-associated genes are downregulated in most cell types. This transcriptomic resource is available via the Gene Expression Analysis Resource (gEAR; https://umgear.org/NIHL) and provides a blueprint for the rational development of drugs to prevent and treat NIHL.