Association of artificial intelligence-powered and manual quantification of programmed death-ligand 1 (PD-L1) expression with outcomes in patients treated with nivolumab ± ipilimumab.

Assessment of programmed death ligand 1 (PD-L1) expression by immunohistochemistry (IHC) has emerged as an important predictive biomarker across multiple tumor types. However, manual quantitation of PD-L1 positivity can be difficult and leads to substantial inter-observer variability. Although the development of artificial intelligence (AI) algorithms may mitigate some of the challenges associated with manual assessment and improve the accuracy of PD-L1 expression scoring, use of AI-based approaches to oncology biomarker scoring and drug development has been sparse, primarily due to the lack of large-scale clinical validation studies across multiple cohorts and tumor types. We developed AI-powered algorithms to evaluate PD-L1 expression on tumor cells by IHC and compared it with manual IHC scoring in urothelial carcinoma, non-small cell lung cancer, melanoma, and squamous cell carcinoma of the head and neck (prospectively determined during the phase II and III CheckMate clinical trials). 1,746 slides were retrospectively analyzed, the largest investigation of digital pathology algorithms on clinical trial datasets performed to date. AI-powered quantification of PD-L1 expression on tumor cells identified more PD-L1-positive samples compared with manual scoring at cutoffs of ≥1% and ≥5% in most tumor types. Additionally, similar improvements in response and survival were observed in patients identified as PD-L1-positive compared with PD-L1-negative using both AI-powered and manual methods, while improved associations with survival were observed in patients with certain tumor types identified as PD-L1-positive using AI-powered scoring only. Our study demonstrates the potential for implementation of digital pathology-based methods in future clinical practice to identify more patients who would benefit from treatment with immuno-oncology therapy compared with current guidelines using manual assessment.

Human-interpretable image features derived from densely mapped cancer pathology slides predict diverse molecular phenotypes.

Computational methods have made substantial progress in improving the accuracy and throughput of pathology workflows for diagnostic, prognostic, and genomic prediction. Still, lack of interpretability remains a significant barrier to clinical integration. We present an approach for predicting clinically-relevant molecular phenotypes from whole-slide histopathology images using human-interpretable image features (HIFs). Our method leverages >1.6 million annotations from board-certified pathologists across >5700 samples to train deep learning models for cell and tissue classification that can exhaustively map whole-slide images at two and four micron-resolution. Cell- and tissue-type model outputs are combined into 607 HIFs that quantify specific and biologically-relevant characteristics across five cancer types. We demonstrate that these HIFs correlate with well-known markers of the tumor microenvironment and can predict diverse molecular signatures (AUROC 0.601-0.864), including expression of four immune checkpoint proteins and homologous recombination deficiency, with performance comparable to 'black-box' methods. Our HIF-based approach provides a comprehensive, quantitative, and interpretable window into the composition and spatial architecture of the tumor microenvironment.

A Machine Learning Approach Enables Quantitative Measurement of Liver Histology and Disease Monitoring in NASH.

BACKGROUND AND AIMS: Manual histological assessment is currently the accepted standard for diagnosing and monitoring disease progression in NASH, but is limited by variability in interpretation and insensitivity to change. Thus, there is a critical need for improved tools to assess liver pathology in order to risk stratify NASH patients and monitor treatment response. APPROACH AND RESULTS: Here, we describe a machine learning (ML)-based approach to liver histology assessment, which accurately characterizes disease severity and heterogeneity, and sensitively quantifies treatment response in NASH. We use samples from three randomized controlled trials to build and then validate deep convolutional neural networks to measure key histological features in NASH, including steatosis, inflammation, hepatocellular ballooning, and fibrosis. The ML-based predictions showed strong correlations with expert pathologists and were prognostic of progression to cirrhosis and liver-related clinical events. We developed a heterogeneity-sensitive metric of fibrosis response, the Deep Learning Treatment Assessment Liver Fibrosis score, which measured antifibrotic treatment effects that went undetected by manual pathological staging and was concordant with histological disease progression. CONCLUSIONS: Our ML method has shown reproducibility and sensitivity and was prognostic for disease progression, demonstrating the power of ML to advance our understanding of disease heterogeneity in NASH, risk stratify affected patients, and facilitate the development of therapies.

Highly Multiplexed Fluorescence in Situ Hybridization for in Situ Genomics.

The quantification of changes in gene copy number is critical to our understanding of tumor biology and for the clinical management of cancer patients. DNA fluorescence in situ hybridization is the gold standard method to detect copy number alterations, but it is limited by the number of genes one can quantify simultaneously. To increase the throughput of this informative technique, a fluorescent bar-code system for the unique labeling of dozens of genes and an automated image analysis algorithm that enabled their simultaneous hybridization for the quantification of gene copy numbers were devised. We demonstrate the reliability of this multiplex approach on normal human lymphocytes, metaphase spreads of transformed cell lines, and cultured circulating tumor cells. It also opens the door to the development of gene panels for more comprehensive analysis of copy number changes in tissue, including the study of heterogeneity and of high-throughput clinical assays that could provide rapid quantification of gene copy numbers in samples with limited cellularity, such as circulating tumor cells.

The organization of leukotriene biosynthesis on the nuclear envelope revealed by single molecule localization microscopy and computational analyses.

The initial steps in the synthesis of leukotrienes are the translocation of 5-lipoxygenase (5-LO) to the nuclear envelope and its subsequent association with its scaffold protein 5-lipoxygenase-activating protein (FLAP). A major gap in our understanding of this process is the knowledge of how the organization of 5-LO and FLAP on the nuclear envelope regulates leukotriene synthesis. We combined single molecule localization microscopy with Clus-DoC cluster analysis, and also a novel unbiased cluster analysis to analyze changes in the relationships between 5-LO and FLAP in response to activation of RBL-2H3 cells to generate leukotriene C4. We identified the time-dependent reorganization of both 5-LO and FLAP into higher-order assemblies or clusters in response to cell activation via the IgE receptor. Clus-DoC analysis identified a subset of these clusters with a high degree of interaction between 5-LO and FLAP that specifically correlates with the time course of LTC4 synthesis, strongly suggesting their role in the initiation of leukotriene biosynthesis.

The RNA-binding protein SFPQ orchestrates an RNA regulon to promote axon viability.

To achieve accurate spatiotemporal patterns of gene expression, RNA-binding proteins (RBPs) guide nuclear processing, intracellular trafficking and local translation of target mRNAs. In neurons, RBPs direct transport of target mRNAs to sites of translation in remote axons and dendrites. However, it is not known whether an individual RBP coordinately regulates multiple mRNAs within these morphologically complex cells. Here we identify SFPQ (splicing factor, poly-glutamine rich) as an RBP that binds and regulates multiple mRNAs in dorsal root ganglion sensory neurons and thereby promotes neurotrophin-dependent axonal viability. SFPQ acts in nuclei, cytoplasm and axons to regulate functionally related mRNAs essential for axon survival. Notably, SFPQ is required for coassembly of LaminB2 (Lmnb2) and Bclw (Bcl2l2) mRNAs in RNA granules and for axonal trafficking of these mRNAs. Together these data demonstrate that SFPQ orchestrates spatial gene expression of a newly identified RNA regulon essential for axonal viability.

ERK-MAPK drives lamellipodia protrusion by activating the WAVE2 regulatory complex.

Cell movement begins with a leading edge protrusion, which is stabilized by nascent adhesions and retracted by mature adhesions. The ERK-MAPK (extracellular signal-regulated kinase-mitogen-activated protein kinase) localizes to protrusions and adhesions, but how it regulates motility is not understood. We demonstrate that ERK controls protrusion initiation and protrusion speed. Lamellipodial protrusions are generated via the WRC (WAVE2 regulatory complex), which activates the Arp2/3 actin nucleator for actin assembly. The WRC must be phosphorylated to be activated, but the sites and kinases that regulate its intermolecular changes and membrane recruitment are unknown. We show that ERK colocalizes with the WRC at lamellipodial leading edges and directly phosphorylates two WRC components: WAVE2 and Abi1. The phosphorylations are required for functional WRC interaction with Arp2/3 and actin during cell protrusion. Thus, ERK coordinates adhesion disassembly with WRC activation and actin polymerization to promote productive leading edge advancement during cell migration.