Background: Traditionally, mutational burden and mutational signatures have been assessed by tumor-normal pair DNA sequencing. The requirement of having both normal and tumor samples is not always feasible from a clinical perspective, and led us to investigate the efficacy of using RNA sequencing of only the tumor sample to determine the mutational burden and signatures, and subsequently molecular cause of the cancer. The potential advantages include reducing the cost of testing, and simultaneously providing information on the gene expression profile and gene fusions present in the tumor. Results: In this study, we devised supervised and unsupervised learning methods to determine mutational signatures from tumor RNA-seq data. As applications, we applied the methods to a training set of 587 TCGA uterine cancer RNA-seq samples, and examined in an independent testing set of 521 TCGA colorectal cancer RNA-seq samples. Both diseases are known associated with microsatellite instable high (MSI-H) and driver defects in DNA polymerase ɛ (POLɛ). From RNA-seq called variants, we found majority (> 95%) are likely germline variants, leading to C > T enriched germline variants (dbSNP) widely applicable in tumor and normal RNA-seq samples. We found significant associations between RNA-derived mutational burdens and MSI/POLɛ status, and insignificant relationship between RNA-seq total coverage and derived mutational burdens. Additionally we found that over 80% of variants could be explained by using the COSMIC mutational signature-5, -6 and -10, which are implicated in natural aging, MSI-H, and POLɛ, respectively. For classifying tumor type, within UCEC we achieved a recall of 0.56 and 0.78, and specificity of 0.66 and 0.99 for MSI and POLɛ respectively. By applying learnt RNA signatures from UCEC to COAD, we were able to improve our classification of both MSI and POLɛ. Conclusions: Taken together, our work provides a novel method to detect RNA-seq derived mutational signatures with effective procedures to remove likely germline variants. It can leads to accurate classification of underlying driving mechanisms of DNA damage deficiency.

Background and Aims: Cholangiocarcinoma (CCA) is one of the most lethal cancers, characterized by molecular heterogeneity and treatment resistance. To uncover new biological signals and therapeutic opportunities, we employed multiomic characterization combined with machine learning. Approach and Results: We profiled all anatomical CCA subtypes using whole exome sequencing, mRNA sequencing, and proteome/phosphoproteome analysis. Integrative dimensional reduction revealed RNA, protein, and phosphoprotein features driving tumor heterogeneity, enabling clustering. Machine learning algorithms identified molecular features for each cluster and mapped external datasets and patient-derived xenograft (PDX) models onto these clusters. Kinase enrichment analysis highlighted targetable kinases active in each cluster. In vivo validation was performed in cluster-specific PDX models using the selective TNK1 inhibitor, TP-5801. We identified 3 molecular clusters with distinct pathway characterization: immunomodulatory (cluster 1), metabolic (cluster 2), and gene regulation/cellular fate (cluster 3). Cluster assignment was independent of anatomic subtype but correlated with overall survival following curative-intent resection. We also identified multiomic features and pathways linked to overall survival and lymph node metastases, crucial for patient treatment selection. Kinase enrichment analysis pinpointed TNK1 as a highly active kinase in the metabolic cluster. Treatment with TP-5801 significantly reduced tumor growth in a metabolic PDX model, but not in models representing the other clusters. Combining internal data with publicly available datasets, we identified the immunomodulatory cluster as most responsive to gemcitabine/cisplatin therapy, confirmed in vivo using cluster-specific PDX models. Conclusions: Integrated multiomic characterization provides translational insights by defining unique molecular subtypes associated both with therapeutic response and overall clinical outcomes. This approach identified TNK1 as a previously unrecognized therapeutic target in a defined subset of CCA tumors.

Chronic lymphocytic leukemia (CLL) is characterized by multiple copy number alterations (CNAs) and somatic mutations that are central to disease prognosis, risk stratification, and mechanisms of therapy resistance. Fluorescence in situ hybridization (FISH) panels are widely used in clinical applications as the gold standard for screening prognostic chromosomal abnormalities in CLL. DNA sequencing is an alternative approach to identifying CNAs but is not an established method for clinical CNA screening. We sequenced DNA from 509 individuals with CLL or monoclonal B-cell lymphocytosis (MBL), the precursor to CLL, using a targeted sequencing panel of 59 recurrently mutated genes in CLL and additional amplicons across regions affected by clinically relevant CNAs [i.e., del(17p), del(11q), del(13q), and trisomy 12]. We used the PatternCNV algorithm to call CNA and compared the concordance of calling clinically relevant CNAs by targeted sequencing to that of FISH. We found a high accuracy of calling CNAs via sequencing compared to FISH. With FISH as the gold standard, the specificity of targeted sequencing was >95%, sensitivity was >86%, positive predictive value was >90%, and negative predictive value was >84% across the clinically relevant CNAs. Using targeted sequencing, we were also able to identify other common CLL-associated CNAs, including del(6q), del(14q), and gain 8q, as well as complex karyotype, defined as the presence of 3 or more chromosomal abnormalities, in 26 patients. In a single and cost-effective assay that can be performed on stored DNA samples, targeted sequencing can simultaneously detect CNAs, somatic mutations, and complex karyotypes, which are all important prognostic features in CLL.

Background: Glioblastoma (GBM) remains a devastating cancer with limited treatment options, largely due to its heterogeneity. While supramaximal resection has recently provided survival benefits, therapeutic profiling of different tumor compartments, particularly its infiltrative edge remains largely unexplored. Methods: Here, we leveraged magnetic resonance imaging (MRI)-guided multi-sampling, collecting 2 cores and 2 margins per case, to query GBM heterogeneity. Whole-exome and RNA-seq with drug testing in two patient-derived 3D models were used to reveal similarities and differences in genomic and transcriptomic makeups, cellular compositions, and drug responses across cores and margins. Bioinformatics interrogations further identified response biomarkers. Results: Mutation analysis showed that oncogenes exhibited a higher degree of spatial heterogeneity than tumor suppressor genes, regardless of MRI status. While the mesenchymal transcriptional subtype with extracellular matrix remodeling, stress response, and immune programs were preferentially enriched in enhancing cores, proneural tumors with neurological processes favored non-enhancing margins. Using a 15-drug GBM-targeted panel, ERK (ulixertinib) and PI3K pathway (paxalisib, CC-115) inhibitors showed preferential efficacy in enhancing cores and non-enhancing margins, respectively. The anti-apoptosis, pan-Bcl2 agent navitoclax and the epigenetic drug trotabresib represented the most effective, tumor-wide monotherapies. Importantly, drug combinations generally outperformed single agents across all regions. Conclusions: This work demonstrates the regional heterogeneity of therapeutic vulnerabilities in GBM ex vivo, showing various drugs with tumor-wide or MRI-enhancement informed activity. These findings offer preclinical bases of numerous monotherapies and drug combinations for future clinical trial design.

Background: Traditionally, mutational burden and mutational signatures have been assessed by tumor-normal pair DNA sequencing. The requirement of having both normal and tumor samples is not always feasible from a clinical perspective, and led us to investigate the efficacy of using RNA sequencing of only the tumor sample to determine the mutational burden and signatures, and subsequently molecular cause of the cancer. The potential advantages include reducing the cost of testing, and simultaneously providing information on the gene expression profile and gene fusions present in the tumor. Results: In this study, we devised supervised and unsupervised learning methods to determine mutational signatures from tumor RNA-seq data. As applications, we applied the methods to a training set of 587 TCGA uterine cancer RNA-seq samples, and examined in an independent testing set of 521 TCGA colorectal cancer RNA-seq samples. Both diseases are known associated with microsatellite instable high (MSI-H) and driver defects in DNA polymerase ɛ (POLɛ).From RNA-seq called variants, we found majority (>95%) are likely germline variants, leading to C>T enriched germline variants (dbSNP) widely applicable in tumor and normal RNA-seq samples. We found significant associations between RNA-derived mutational burdens and MSI/POLɛ status, and insignificant relationship between RNA-seq total coverage and derived mutational burdens. Additionally we found that over 80% of variants could be explained by using the COSMIC mutational signature-5, -6 and -10, which are implicated in natural aging, MSI-H, and POLɛ, respectively. For classifying tumor type, within UCEC we achieved a recall of 0.56 and 0.78, and specificity of 0.66 and 0.99 for MSI and POLɛ respectively. By applying learnt RNA signatures from UCEC to COAD, we were able to improve our classification of both MSI and POLɛ. Conclusions: Taken together, our work provides a novel method to detect RNA-seq derived mutational signatures with effective procedures to remove likely germline variants. It can leads to accurate classification of underlying driving mechanisms of DNA damage deficiency.

Background: Cholangiocarcinoma is an aggressive malignancy with limited systemic therapeutics. We have recently identified LCK, a non-receptor tyrosine kinase, as a potential therapeutic target in select subtypes of cholangiocarcinoma. Proteomic profiling in other cancer types has proven beneficial in predicting response to systemic therapies. Herein, we present proteomic profiling of cholangiocarcinoma can predict sensitivity to a novel LCK inhibitor in patient derived xenograft models. Methods: Proteins from 28 unique cholangiocarcinoma patient derived xenograft models (PDX) were isolated, labeled with tandem mass tags (TMT) and then analyzed by liquid chromatography and mass spectroscopy (LC/MS). Phosphopeptides were separately evaluated both by Immobilized Metal Affinity Chromatography and pY-enrichment techniques. Following normalization, the global proteome, serine/threonine phosphoproteome, and tyrosine phosphoproteome were analyzed and a proteomic signature was created for each tumor. These signatures were compared to PDX models with known response profiles to NTRC0652-0, an LCK inhibitor (1 resistant and 1 sensitive). The top 200 differentially abundant proteins (log2FC 2.6) were then utilized to develop a sensitive and a resistant signature. Correlation indices were then calculated across the 28 PDX model library. An additional predicted resistant and predicted sensitive model was expanded. Tumor bearing mice were randomized based on tumor volume to either treatment with NTRC (30mg/kg PO daily), n=5, or vehicle, n=4, (10% Kolliphor, 10% DMSO, 80% Water, 2 mol equivalents HCl). Tumors were measured weekly with calipers and then weighed at the end of the study. Results: In the top 200 differentially abundant proteins that represented the proteomic signature there was over representation of proteins with decreased abundance. There were only 12 proteins with higher abundance in the sensitive model as compared to the resistant. Utilizing the signature developed from the global proteomics data, nine tumors were predicted to be sensitive while 17 were predicted to be resistant to NTRC. PDX115 is a an intrahepatic CCA PDX known to contain an IDH1 mutation and was predicted to be resistant by modeling. The model was resistant to single agent NTRC with no demonstrable effect on tumor growth in vivo. PDX175 is a distal CCA PDX and was predicted to be sensitive by modeling. In vivo growth was inhibited by NTRC as compared to vehicle treated mice following two weeks of treatment (983% vs 435%, p=0.03). Additionally, final tumor weight was significantly lower in NTRC treated mice (0.651 g vs 0.317 g, p=0.04). Conclusion: Cholangiocarcinoma proteomic signatures can be used to predict therapeutic response to a novel LCK inhibitor in preclinical models. Whether or not this could be applied to other treatment modalities in cholangiocarcinoma remains to be determined. Citation Format: Ryan Watkins, Roberto Alva-Ruiz, Caitlin Conboy, Dong-Gi Mun, Erik Jessen, Diep Vu, Jos de Man, Rogier Buijsman, Akhilesh Pandey, Mitesh Borad, Gregory Gores, Rory Smoot. Proteomic profiling of cholangiocarcinoma predicts response to a novel small molecule inhibitor in preclinical models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3094.