A platform-independent AI tumor lineage and site (ATLAS) classifier.

Histopathologic diagnosis and classification of cancer plays a critical role in guiding treatment. Advances in next-generation sequencing have ushered in new complementary molecular frameworks. However, existing approaches do not independently assess both site-of-origin (e.g. prostate) and lineage (e.g. adenocarcinoma) and have minimal validation in metastatic disease, where classification is more difficult. Utilizing gradient-boosted machine learning, we developed ATLAS, a pair of separate AI Tumor Lineage and Site-of-origin models from RNA expression data on 8249 tumor samples. We assessed performance independently in 10,376 total tumor samples, including 1490 metastatic samples, achieving an accuracy of 91.4% for cancer site-of-origin and 97.1% for cancer lineage. High confidence predictions (encompassing the majority of cases) were accurate 98-99% of the time in both localized and remarkably even in metastatic samples. We also identified emergent properties of our lineage scores for tumor types on which the model was never trained (zero-shot learning). Adenocarcinoma/sarcoma lineage scores differentiated epithelioid from biphasic/sarcomatoid mesothelioma. Also, predicted lineage de-differentiation identified neuroendocrine/small cell tumors and was associated with poor outcomes across tumor types. Our platform-independent single-sample approach can be easily translated to existing RNA-seq platforms. ATLAS can complement and guide traditional histopathologic assessment in challenging situations and tumors of unknown primary.

Machine Learning & Molecular Radiation Tumor Biomarkers.

Developing radiation tumor biomarkers that can guide personalized radiotherapy clinical decision making is a critical goal in the effort towards precision cancer medicine. High-throughput molecular assays paired with modern computational techniques have the potential to identify individual tumor-specific signatures and create tools that can help understand heterogenous patient outcomes in response to radiotherapy, allowing clinicians to fully benefit from the technological advances in molecular profiling and computational biology including machine learning. However, the increasingly complex nature of the data generated from high-throughput and "omics" assays require careful selection of analytical strategies. Furthermore, the power of modern machine learning techniques to detect subtle data patterns comes with special considerations to ensure that the results are generalizable. Herein, we review the computational framework of tumor biomarker development and describe commonly used machine learning approaches and how they are applied for radiation biomarker development using molecular data, as well as challenges and emerging research trends.

Intrinsic Molecular Subtypes of Metastatic Castration-Resistant Prostate Cancer.

PURPOSE: Although numerous biology-driven subtypes have been described previously in metastatic castration-resistant prostate cancer (mCRPC), unsupervised molecular subtyping based on gene expression has been less studied, especially using large cohorts. Thus, we sought to identify the intrinsic molecular subtypes of mCRPC and assess molecular and clinical correlates in the largest combined cohort of mCRPC samples with gene expression data available to date. EXPERIMENTAL DESIGN: We combined and batch-effect corrected gene expression data from four mCRPC cohorts from the Fred Hutchinson Cancer Research Center (N = 157), a small-cell neuroendocrine (NE) prostate cancer (SCNC)-enriched cohort from Weill Cornell Medicine (N = 49), and cohorts from the Stand Up 2 Cancer/Prostate Cancer Foundation East Coast Dream Team (N = 266) and the West Coast Dream Team (N = 162). RESULTS: Hierarchical clustering of RNA-sequencing data from these 634 mCRPC samples identified two distinct adenocarcinoma subtypes, one of which (adeno-immune) was characterized by higher gene expression of immune pathways, higher CIBERSORTx immune scores, diminished ASI benefit, and non-lymph node metastasis tropism compared with an adeno-classic subtype. We also identified two distinct subtypes with enrichment for an NE phenotype, including an NE-liver subgroup characterized by liver metastasis tropism, PTEN loss, and APC and SPOP mutations compared with an NE-classic subgroup. CONCLUSIONS: Our results emphasize the heterogeneity of mCRPC beyond currently accepted molecular phenotypes, and suggest that future studies should consider incorporating transcriptome-wide profiling to better understand how these differences impact treatment responses and outcomes.

Identification of phenocopies improves prediction of targeted therapy response over DNA mutations alone.

DNA mutations in specific genes can confer preferential benefit from drugs targeting those genes. However, other molecular perturbations can "phenocopy" pathogenic mutations, but would not be identified using standard clinical sequencing, leading to missed opportunities for other patients to benefit from targeted treatments. We hypothesized that RNA phenocopy signatures of key cancer driver gene mutations could improve our ability to predict response to targeted therapies, despite not being directly trained on drug response. To test this, we built gene expression signatures in tissue samples for specific mutations and found that phenocopy signatures broadly increased accuracy of drug response predictions in-vitro compared to DNA mutation alone, and identified additional cancer cell lines that respond well with a positive/negative predictive value on par or better than DNA mutations. We further validated our results across four clinical cohorts. Our results suggest that routine RNA sequencing of tumors to identify phenocopies in addition to standard targeted DNA sequencing would improve our ability to accurately select patients for targeted therapies in the clinic.

Longitudinal Molecular Profiling of Circulating Tumor Cells in Metastatic Renal Cell Carcinoma.

PURPOSE: Liquid biopsies in metastatic renal cell carcinoma (mRCC) provide a unique approach to understand the molecular basis of treatment response and resistance. This is particularly important in the context of immunotherapies, which target key immune-tumor interactions. Unlike metastatic tissue biopsies, serial real-time profiling of mRCC is feasible with our noninvasive circulating tumor cell (CTC) approach. METHODS: We collected 457 longitudinal liquid biopsies from 104 patients with mRCC enrolled in one of two studies, either a prospective cohort or a phase II multicenter adaptive immunotherapy trial. Using a novel CTC capture and automated microscopy platform, we profiled CTC enumeration and expression of HLA I and programmed cell death-ligand 1 (PD-L1). Given their diametric immunological roles, we focused on the HLA I to PD-L1 ratio (HP ratio). RESULTS: Patients with radiographic responses showed significantly lower CTC abundances throughout treatment. Furthermore, patients whose CTC enumeration trajectory was in the highest quartile (> 0.12 CTCs/mL annually) had shorter overall survival (median 17.0 months v 21.1 months, P < .001). Throughout treatment, the HP ratio decreased in patients receiving immunotherapy but not in patients receiving tyrosine kinase inhibitors. Patients with an HP ratio trajectory in the highest quartile (≥ 0.0012 annually) displayed significantly shorter overall survival (median 18.4 months v 21.2 months, P = .003). CONCLUSION: In the first large longitudinal CTC study in mRCC to date to our knowledge, we identified the prognostic importance of CTC enumeration and the change over time of both CTC enumeration and the HP ratio. These insights into changes in both tumor burden and the molecular profile of tumor cells in response to different treatments provide potential biomarkers to predict and monitor response to immunotherapy in mRCC.

Predicting cancer drug TARGETS - TreAtment Response Generalized Elastic-neT Signatures.

We are now in an era of molecular medicine, where specific DNA alterations can be used to identify patients who will respond to specific drugs. However, there are only a handful of clinically used predictive biomarkers in oncology. Herein, we describe an approach utilizing in vitro DNA and RNA sequencing and drug response data to create TreAtment Response Generalized Elastic-neT Signatures (TARGETS). We trained TARGETS drug response models using Elastic-Net regression in the publicly available Genomics of Drug Sensitivity in Cancer (GDSC) database. Models were then validated on additional in-vitro data from the Cancer Cell Line Encyclopedia (CCLE), and on clinical samples from The Cancer Genome Atlas (TCGA) and Stand Up to Cancer/Prostate Cancer Foundation West Coast Prostate Cancer Dream Team (WCDT). First, we demonstrated that all TARGETS models successfully predicted treatment response in the separate in-vitro CCLE treatment response dataset. Next, we evaluated all FDA-approved biomarker-based cancer drug indications in TCGA and demonstrated that TARGETS predictions were concordant with established clinical indications. Finally, we performed independent clinical validation in the WCDT and found that the TARGETS AR signaling inhibitors (ARSI) signature successfully predicted clinical treatment response in metastatic castration-resistant prostate cancer with a statistically significant interaction between the TARGETS score and PSA response (p = 0.0252). TARGETS represents a pan-cancer, platform-independent approach to predict response to oncologic therapies and could be used as a tool to better select patients for existing therapies as well as identify new indications for testing in prospective clinical trials.