Characterize Tumor Heterogeneity Research Articles

Radiomics for Classification of Lung Cancer Histological Subtypes Based on Nonenhanced Computed Tomography

To evaluate the performance of using radiomics method to classify lung cancer histological subtypes based on nonenhanced computed tomography images. 278 patients with pathologically confirmed lung cancer were collected, including 181 nonsmall cell lung cancer (NSCLC) and 97 small cell lung cancers (SCLC) patients. Among the NSCLC patients, 88 patients were adenocarcinomas (AD) and 93 patients were squamous cell carcinomas (SCC). In total, 1695 quantitative radiomic features (QRF) were calculated from the primary lung cancer tumor in each patient. To build radiomic classification model based on the extracted QRFs, several machine-learning algorithms were applied sequentially. First, unsupervised hierarchical clustering was used to exclude highly correlated QRFs; second, the minimum Redundancy Maximum Relevance feature selection algorithm was employed to select informative and nonredundant QRFs; finally, the Incremental Forward Search and Support Vector Machine classification algorithms were used to combine the selected QRFs and build the model. In our work, to study the phenotypic differences among lung cancer histological subtypes, four classification models were built. They were models of SCLC vs NSCLC, SCLC vs AD, SCLC vs SCC, and AD vs SCC. The performance of the classification models was evaluated by the area under the receiver operating characteristic curve (AUC) estimated by three-fold cross-validation. The AUC (95% confidence interval) for the model of SCLC vs NSCLC was 0.741(0.678, 0.795). For the models of SCLC vs AD and SCLC vs SCC, the AUCs were 0.822(0.755, 0.875) and 0.665(0.583, 0.738), respectively. The AUC for the model of AD vs SCC was 0.655(0.570, 0.731). Several QRFs ("Law_15," "LoG_Uniformity," "GLCM_Contrast," and "Compactness Factor") that characterize tumor heterogeneity and shape were selected as the significant features to build the models. Our results show that phenotypic differences exist among different lung cancer histological subtypes on nonenhanced computed tomography image.

BayCount: A Bayesian decomposition method for inferring tumor heterogeneity using RNA-Seq counts

Tumors are heterogeneous. A tumor sample usually consists of a set of subclones with distinct transcriptional profiles and potentially different degrees of aggressiveness and responses to drugs. Understanding tumor heterogeneity is therefore critical for precise cancer prognosis and treatment. In this paper we introduce BayCount—a Bayesian decomposition method to infer tumor heterogeneity with highly over-dispersed RNA sequencing count data. Using negative binomial factor analysis, BayCount takes into account both the between-sample and gene-specific random effects on raw counts of sequencing reads mapped to each gene. For the posterior inference, we develop an efficient compound Poisson-based blocked Gibbs sampler. Simulation studies show that BayCount is able to accurately estimate the subclonal inference, including the number of subclones, the proportions of these subclones in each tumor sample, and the gene expression profiles in each subclone. For real world data examples, we apply BayCount to The Cancer Genome Atlas lung cancer and kidney cancer RNA sequencing count data and obtain biologically interpretable results. Our method represents the first effort in characterizing tumor heterogeneity using RNA sequencing count data that simultaneously removes the need of normalizing the counts, achieves statistical robustness, and obtains biologically/clinically meaningful insights. The R package BayCount implementing our model and algorithm is available for download.

LIFEx: A Freeware for Radiomic Feature Calculation in Multimodality Imaging to Accelerate Advances in the Characterization of Tumor Heterogeneity

Textural and shape analysis is gaining considerable interest in medical imaging, particularly to identify parameters characterizing tumor heterogeneity and to feed radiomic models. Here, we present a free, multiplatform, and easy-to-use freeware called LIFEx, which enables the calculation of conventional, histogram-based, textural, and shape features from PET, SPECT, MR, CT, and US images, or from any combination of imaging modalities. The application does not require any programming skills and was developed for medical imaging professionals. The goal is that independent and multicenter evidence of the usefulness and limitations of radiomic features for characterization of tumor heterogeneity and subsequent patient management can be gathered. Many options are offered for interactive textural index calculation and for increasing the reproducibility among centers. The software already benefits from a large user community (more than 800 registered users), and interactions within that community are part of the development strategy.Significance: This study presents a user-friendly, multi-platform freeware to extract radiomic features from PET, SPECT, MR, CT, and US images, or any combination of imaging modalities. Cancer Res; 78(16); 4786-9. ©2018 AACR.

Epigenetic Biomarkers for Risk Assessment of Particulate Matter Associated Lung Cancer.

Particulate matter directly emitted into the air by sources such as combustion processes and windblown dust, or formed in the atmosphere by transformation of emitted gases are the major contributors to air pollution that triggers a diverse array of human pathologies including lung cancer. The mortality in lung cancer is usually high as the disease is not symptomatic at its early treatable stage. Moreover, available methods for screening are costly and mainly rely on imaging techniques which lack sufficient sensitivity and specificity. Despite progress in the identification of biomarkers, gene mutation based approaches still face formidable challenges as the disease evolves from a complex interplay between environment and host. Therefore, identification of an epigenomic signature might be useful for early diagnosis with the potential to reduce the environmental-associated disease burden. The review discusses the utility of epigenomic signature in identification and management of the environmental-associated lung cancers. Non-invasive 'liquid biopsy' based epigenomic screening has recently emerged as a methodology which has potential to characterize tumor heterogeneity at initial stages. Epigenetic signatures (methylated DNA, miRNA, and post transcriptionally modified histones) known to reflect the vital cellular changes, circulate at higher levels in the individuals with lung cancer. These circulating biological entities are reported to be closely associated with the clinical outcome of lung cancer patients and thus strongly stand as the probable candidate to identify disease at an early stage and monitor treatment response, thereby, benefiting patients and improving their lives. However, for effective implementation of the strategy as "point-of-care" test for screening population-at-risk will require exhaustive clinical validation.

Abstract 1013: Analyzing infiltrating B cell populations in the tumor microenvironment using single cell transcriptomics

Abstract Understanding the complex interactions between malignant, stromal and infiltrating lymphocyte cell types within the tumor microenvironment is a critical component of personalized cancer treatments. We describe an approach that couples single-cell transcriptional profiling of tumors with high-resolution receptor profiling of tumor infiltrating lymphocytes (TILs). Using a fully integrated, droplet-based system for 5' single cell RNA sequencing (scRNA-seq), we simultaneously profiled the transcriptome and immune repertoire of the same cells from a primary colorectal cancer (CRC) tumor and a primary non-small cell lung cancer (NSCLC) tumor. Each tumor varied in type and proportion of its cellular components, noticeably in the proportion of TILs. The tumor cells in the CRC were mostly epithelial in nature, with the TILs comprising 21% of the total cells sequenced. This lymphocyte population consisted of both T (5% CD4+, 3% CD8+) and, interestingly, B cells (5% CD19+), with a number of plasma B cells (IGH high, CD138+) also identified. The NSCLC tumor had a high immune cell infiltrate, with lymphocytes comprising 43% of cells (19% T, 23% B, and 5% plasma B cells) indicating a robust adaptive immune response. To examine these cells further, CD45+ cells were subject to scRNA-seq. When compared to the unenriched sample, the enriched cell population had a similar B to T cell ratio with a reduction in the proportion of plasma B cells, likely due to exclusion of mature plasma B cells by the sort (47% T (44% CD4+, 29% CD8+), 42% B, and 2% plasma B cells). To fully characterize the B cell infiltrate we applied targeted 5' scRNA-seq to obtain full length, paired, B cell receptor heavy and light chain sequences. Despite the presence of plasma B cells in the CRC tumor, limited clonal expansion was observed. In the NSCLC tumor, clonal expansion of B cells was observed in the unenriched tumor sample but these clonotypes were not present in the CD45+ population, suggesting the clonally expanded cells are in the mature plasma B cell population. We also applied a targeted scRNA-seq approach to obtain full length paired T cell receptor alpha and beta sequences, however, no clear T cell expansion was observed in either sample. These findings emphasize the importance of examining receptor sequences rather than relying on the presence of B or T cells alone to determine if a robust immune response is being mounted. The presence of tumor-infiltrating B lymphocytes has been linked to a favorable clinical outcome in some types of cancers. Identification of these cells, their subsets and high resolution profiling of their receptor sequences using scRNA-seq allows a novel insight into the adaptive immune response in tumors. This technology will allow better characterization of tumor heterogeneity and the adaptive immune response to the tumor microenvironment and will serve as a foundation for future research into tumor immunology and immunotherapy. Citation Format: Sarah E. Taylor, Stephane Boutet, Valeria Giangarrá, Grace X. Zheng, Alvaro M. Barrio, Luz Montesclaros, Josephine Lee, Samuel Marrs, Kevin J. Wu, Paul Ryvkin, Tarjei Mikkelsen, Deanna M. Church. Analyzing infiltrating B cell populations in the tumor microenvironment using single cell transcriptomics [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1013.

Abstract 2180: Investigating lung adenocarcinoma tumor heterogeneity with single-cell mass cytometry

Abstract The lack of accuracy in predicting behavior of early detected lung adenocarcinoma (ADC) presents a major challenge to patients and their providers. Indolent tumors may be overdiagnosed and overtreated. While imaging tools may contribute to better prediction of tumor behavior, unveiling tumor heterogeneity of ADCs at the single-cell level will elucidate mechanisms of tumor progression. Mass cytometry allows us to profile tumor heterogeneity and identify cell populations driven by the activation of specific signaling pathways. We report the development of an antibody panel for mass cytometry and preliminary experiments in cancer cell lines and in ADC. ADCs were collected at the time of surgery, dissociated into suspension and cryopreserved within 24 hours. A mass cytometry antibody panel was developed, including commercially available and in-house metal conjugated antibodies. Biaxial gating or unsupervised analysis approaches SPADE and viSNE were used to compare major populations of cells. Our comprehensive antibody panel includes markers for cellular lineage (14) (immune, epithelial, mesenchymal, fibroblasts), cancer cells (3), signaling pathways (15) and quality control markers (2). The in-house conjugated antibodies (9/34) were titrated in concentrations going from 0 to 1 ug/mL using cell lines with known expression of the marker of interest. A panel of previously validated antibodies was used to gate cell populations. Signal medians of positive and negative populations and standard deviations of negative population were used to determine the stain index and the signal/noise ratio. Based on titration curves, optimal concentrations of antibody were selected for use. In the tumor studied, four cell populations were identified: endothelial cells, fibroblasts, epithelial tumor cells and leukocytes. Basal kinase activity was detected in cancer cells and infiltrating leukocytes, both exhibiting p-STAT5 activation, and cancer cells showing high p-AKT activation. Using our Marker Enrichment Modeling algorithm, we identified 3 differentially enriched subpopulations of tumor cells, as well as different leukocyte populations. These preliminary data proved mass cytometry and the panel developed are suitable tools to characterize tumor heterogeneity in lung ADC. The work in progress studying indolent and aggressive ADCs promises to identify detailed phenotypes and respective signaling pathway activation that may ultimately allow us to better predict tumor behavior and integrate this knowledge in the clinical workflow. Supported by CA196415. Citation Format: Maria-Fernanda Senosain-Ortega, Yong Zou, Deon B. Doxie, Caroline E. Roe, Jonathan M. Lehman, Jonathan M. Irish, Pierre P. Massion. Investigating lung adenocarcinoma tumor heterogeneity with single-cell mass cytometry [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 2180.

Characterization of tumor heterogeneity by latent haplotypes: a sequential Monte Carlo approach.

Tumor samples obtained from a single cancer patient spatially or temporally often consist of varying cell populations, each harboring distinct mutations that uniquely characterize its genome. Thus, in any given samples of a tumor having more than two haplotypes, defined as a scaffold of single nucleotide variants (SNVs) on the same homologous genome, is evidence of heterogeneity because humans are diploid and we would therefore only observe up to two haplotypes if all cells in a tumor sample were genetically homogeneous. We characterize tumor heterogeneity by latent haplotypes and present state-space formulation of the feature allocation model for estimating the haplotypes and their proportions in the tumor samples. We develop an efficient sequential Monte Carlo (SMC) algorithm that estimates the states and the parameters of our proposed state-space model, which are equivalently the haplotypes and their proportions in the tumor samples. The sequential algorithm produces more accurate estimates of the model parameters when compared with existing methods. Also, because our algorithm processes the variant allele frequency (VAF) of a locus as the observation at a single time-step, VAF from newly sequenced candidate SNVs from next-generation sequencing (NGS) can be analyzed to improve existing estimates without re-analyzing the previous datasets, a feature that existing solutions do not possess.

Abstract No. 478 Shearwave elastography for characterization of tumor heterogeneity in the rabbit liver VX2 tumor

To evaluate the diagnostic performance of shear wave elastography (SWE) imaging for differentiating normal liver from tumor and characterizing tumor heterogeneity in the rabbit liver VX2 cancer model. Nine rabbits were implanted with VX2 tumor in liver and imaged with conventional US and SWE 2-3 weeks after tumor implantation. Depending on tumor size, up to 12 ROI were placed in tumor and up to 8 ROI in normal liver. The mean, standard deviation and ratio of SWS values (m/s) and elastic modulus (SWEM, KPa) on SWE were recorded. Additionally, 2D elastic images were classified into four color patterns (red, blue, green and yellow). The area under the receiver operating characteristic (AUROC) curve analysis was performed to evaluate the diagnostic performance of SWE in differentiating tumor versus normal liver. There were statistically significant differences between tumor rim and tumor core (mean tumor rim SWEM = 32.69, SD = 18.64 KPa, mean tumor core SWEM = 13.21, SD = 2.73 KPa, p = 0.001), total tumor and liver (mean liver = 7.96, SD = 1.28 KPa, p<0.0001), tumor core and liver (p<0.001). Compared with other quantitative SWE parameters, mean values expressed in KPa had the highest AUROC value (AUROC = 0.90), with corresponding cut-off value of 17.8 KPa for differentiating tumor from liver, sensitivity of 85.1%, specificity of 96.6%, accuracy of 94.2%, PPV of 86.9%, and NPV of 96.2%. SWE with elastic modulus values can successfully differentiate tumor rim vs. tumor core vs. normal liver in the rabbit model of liver cancer.

Abstract A182: Apobec-mutagenesis drives mutational heterogeneity while copy number alterations drive transcriptomic and proteogenomic heterogeneity in metastatic lung adenocarcinoma and thymic carcinoma

Abstract Background: Molecular characterization of metastatic tumors to investigate intra- and intertumor heterogeneity has proven more difficult than early stage disease as opportunities for surgical removal are limited. Rapid/warm autopsy programs that are conducted soon after death minimize tissue ischemia, thereby enabling procurement of high-quality viable tissue. To our knowledge, integrated proteo-genomics studies to characterize tumor heterogeneity of metastatic lung adenocarcinoma and thymic carcinoma are lacking. Methods: We sought to characterize and decipher mechanisms of intra- and intertumor heterogeneity of Stage IV lung adenocarcinoma and thymic carcinoma through whole exome sequencing, RNA sequencing, copy number estimation, and mass spectrometry-based proteomics of 38 tumors from five rapid/warm autopsy patients. The autopsy procedure was initiated between 2-4 hours of death. Results: One lung adenocarcinoma (RA003) and one thymic carcinoma (RA006) patient had high mutation burden and extreme mutational (SNV) heterogeneity. This was likely driven by APOBEC-mutagenesis. We constructed phylogenetic trees showing truncal, shared, and private SNVs and copy number alterations (CNAs). Interestingly, we discovered a pattern of recurrent, parallel but opposite arm level CNAs in different metastatic branches within a subset of patients (RA000, RA004, and RA006). In contrast, RA003 had only truncal CNAs and the least heterogeneity at the CNA level. Assessment of intra- and intertumor heterogeneity at the transcriptomic and proteomic levels revealed high concordance in gene expression and protein abundance within tumors from patient RA003 compared to patient RA004, potentially reflecting the downstream effect of CNA vs. APOBEC-induced mutations. Further analysis of cis vs. trans effects of CNA on gene expression and protein abundance in all the metastatic tumors revealed many trans-acting hot spots and limited cis effects, which is contrary to copy number effects on gene expression in primary lung adenocarcinoma tumors. Trans-acting hot spots differed between two major metastatic lineage of patient RA004, suggesting copy number changes may be responsible for high intermetastatic proteomic heterogeneity in this patient. Conclusions: Our results demonstrate ongoing mutational and chromosomal instability throughout the metastatic process of lung adenocarcinoma and thymic carcinoma. APOBEC mutagenesis was likely the primary driver of SNV heterogeneity. However, intertumor heterogeneity of metastatic disease at the level of the transcriptome and proteome occurred in association with CNA rather than mutational heterogeneity. APOBEC mutagenesis and CNA are complementary mechanisms of tumor evolution and heterogeneity affecting tumor metastasis. Citation Format: Nitin Roper, James Gao, Rouf Banday, Tapan Maity, Xu Zhang, Abhilash Venugopalan, Rajesh Patidar, Sivasish Sindiri, Constance Cultraro, Alexandr Goncearenco, Anna Panchenko, Romi Biswas, Jing Wang, Stephen Hewitt, David Kleiner, Bing Zhang, Javed Khan, Ludmila Prokunina-Olsson, Udayan Guha. Apobec-mutagenesis drives mutational heterogeneity while copy number alterations drive transcriptomic and proteogenomic heterogeneity in metastatic lung adenocarcinoma and thymic carcinoma [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2017 Oct 26-30; Philadelphia, PA. Philadelphia (PA): AACR; Mol Cancer Ther 2018;17(1 Suppl):Abstract nr A182.

Potential Management of Circulating Tumor DNA as a Biomarker in Triple-Negative Breast Cancer.

As a specific subtype of breast cancer, Triple-negative breast cancer (TNBC) is associated with worse prognosis and higher tumor aggressiveness than HER2-amplified or hormone receptor positive breast cancers. Circulating tumor DNA (ctDNA), as a non-invasive “liquid biopsy”, is an emerging original blood-based biomarker for early breast cancer diagnosis, monitoring treatment response, and determining prognosis. In TNBC patients, ctDNA has an inherent tendency to characterize tumor heterogeneity and metastasis-specific mutations providing a key alternative to tumor tissue profiling. Several studies have already demonstrated the potential of ctDNA in TNBC patients from early to advanced stages of the disease including diagnosis, therapy decisions and assessment of prognosis. This review provides a critical brief summary of the evidence that gives credence to the utility of ctDNA as a biomarker for its role into clinical management in TNBC.

Ultrasound molecular imaging of the secreted tumor marker Netrin-1 in multiple breast cancer models

With regard to biomarker expression, tumors are heterogeneous media displaying intra- and interpatient variability. Ultrasound molecular imaging (USMI) is a newly developed tool that can characterize heterogeneous tumors. Here, we focus on Netrin-1-overexpressing breast cancer. A newly developed targeted therapy interfering with Netrin-1 requires patient stratification. We developed Netrin-1-targeted microbubbles (MBs) and assessed USMI for identification of Netrin-1-positive breast cancer. In vitro flow chamber assays showed specific binding of anti-Netrin-1-MBs to human and murine breast cancer cell lines overexpressing Netrin-1. In vivo, in Netrin-1-positive transgenic or implanted mouse breast cancer models, USMI showed significantly increased signals with Netrin-1-targeted MBs, compared with isotype control MBs and targeted MBs after blocking of Netrin-1:—44.8%, 32.0%, 33.7% intensity for MBsNetrin-1, MBsIsotype, or after blocking, resp., in MMTV tumors;—28.2%, 13.7%, 13.4%, resp., in SKBR7 tumors;—60.9%, 47.8%, 47.9%, resp., in 4T1 tumors. In Netrin-1-negative tumors and in normal glands, targeted and control MBs showed no significant differences. Immunohistochemistry confirmed the expression of Netrin-1 in endothelial cells. In conclusion, USMI allowed the characterization of tumor heterogeneity and the differentiation between Netrin-1-positive and -negative tumors, and has the potential to become a companion diagnostic for breast cancer patient stratification.

Abstract 3935: Dissecting small cell lung carcinoma heterogeneity and chemotherapy resistance with mass cytometry

Abstract Introduction: Small cell lung cancer (SCLC) is a high grade neuroendocrine carcinoma of the lung responsible for up to 25% of lung cancer deaths. Treatment in SCLC has not changed significantly in the last 20 years. SCLC initially responds well to chemotherapy, but inevitably recurs. Characterization of tumor heterogeneity and changes in SCLC cell signaling and phenotypes after chemotherapy could yield new insights and therapeutic options. Mass cytometry uses metal labeled antibodies to profile expression and phosphorylation of more than 40 proteins in single cells and offers the opportunity to identify new subpopulations including potential cancer stem cell populations as well as targets for novel therapies in SCLC. Methods: Nude mice with SCLC patient derived xenografts (PDXs) were treated with one cycle of carboplatin/etoposide or saline injection. Tumors were harvested at ~2000mm3, disaggregated, and cryopreserved. PDX samples were stained with a 22 marker panel and an intercalator dye to identify nucleated cells. This panel measured phospho-signaling, neuroendocrine, immune, and mesenchymal cell markers, and functional markers including ki67 and cleaved caspase 3. ViSNE analysis and biaxial gating were used to identify major subpopulations of interest. Results: PDX tumors released viable tumor and stromal cells suitable for cryopreservation and mass cytometry. ACK buffer and enzymatic dissociation yielded the best quality cells by depleting red blood cells. Mouse cells, including leukocytes, were excluded using mouse MHC1 gating and iridium intercalator was used to identify nucleated cells. Single cell protein expression and phosphorylation was analyzed using viSNE and yielded at least 9 distinct subpopulations based on density islands with neuroendocrine (CD56+) and non-neuroendocrine (CD56-) populations. Chemotherapy treated cells had dramatic changes in subpopulation distribution compared to matched mock treated tumor. This included 2-3 fold expansion of SOX2+, CD117+, and pSTAT3+ populations with chemotherapy treatment. A small CD44+ tumor subpopulation identified in the chemotherapy treated cells was not present in the matched mock treated tumor suggesting a potential chemotherapy resistant/ stem- like subpopulation. Kinase activity showed stable p-AKT overall, but increased p-S6 in the chemotherapy treated cells. Conclusions: Mass cytometry was able to identify multiple neuroendocrine and non-neuroendocrine cell populations from SCLC PDXs and characterize their signaling. Chemotherapy treated PDX had differential subpopulation distribution with enrichment of multiple stem-like signaling factors. This work demonstrates the utility of mass cytometry and viSNE as novel techniques to identify subpopulations associated with chemotherapy resistance for future targeting and demonstrates the feasibility of this technique for characterizing signaling heterogeneity in human SCLC tumors. Citation Format: Jonathan M. Lehman, Nalin Leelatian, Bradford Harris, Megan Hoeksema, Zou Yong, Deon B. Doxie, Jonathan M. Irish, Pierre P. Massion. Dissecting small cell lung carcinoma heterogeneity and chemotherapy resistance with mass cytometry [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3935. doi:10.1158/1538-7445.AM2017-3935

Abstract 2911: Single cell mass cytometry analysis of human lung adenocarcinoma

Abstract Introduction: Lung cancer is the leading cause of cancer-related mortality in the world. Lung adenocarcinoma is the most common subtype. Tumor heterogeneity among adenocarcinomas presents a challenge in the management of the disease. Understanding heterogeneity may have implications in understanding the biological processes driving progression. Single cell platforms like mass cytometry offer an opportunity to profile tumor heterogeneity and to identify populations of driven by the activation of signaling pathways. Methods: Adenocarcinomas were collected at the time of surgery and dissociated into suspension and cryopreserved. Mass cytometry analysis of tumors and adenocarcinoma cell lines was performed with a 30 marker antibody panel and rhodium intercalator dye to identify dead cells. The antibody panel included markers to characterize identity, cell cycle status, and signaling events. Biaxial gating or unsupervised analysis approaches SPADE and viSNE were used to compare major populations of cells. Results: Validation of mass cytometry panels was performed with adenocarcinoma cell lines and human tumors. Adenocarcinoma cell lines PC9 (lung), A549 (lung), H520 (squamous) and SW620 (colon) exhibited phenotypically distinct cells between samples by expression of cytokeratin, CK7, and EGFR. Furthermore, PD-L1 expression was expressed on subsets within cell lines. Disaggregation of tumors was optimized with dissociation that included collagenase and DNase to release viable cells within 1 hour. Detection of viable cells was optimized with rhodium intercalator viability dye and histone H3 to identify nucleated cells. Using the 30 marker panel and viSNE analysis four populations were identified within tumors as infiltrating leukocytes, endothelial cells, fibroblast, and epithelial tumor cells. Epithelial cells were found to exhibit cellular heterogeneity between patients based on the expression of cytokeratin, CK7, TTF1, EGFR, vimentin, CD44, and MET. Preliminary signaling data identified basal kinase activity active in infiltrating leukocytes and cancer cells. Cancer cells and infiltrating immune cells had basal p-STAT5 activation, and cancer cells had high basal p-AKT implicated in dysregulated cell growth. Conclusions: Populations of infiltrating stromal cells and epithelial (cancer) cells were identified in lung adenocarcinoma. The single cell phenotyping from tumors was consistent with the profile found in two lung adenocarcinoma cell lines. Preliminary differences in basal signaling pathways responsible for growth were observed within adenocarcinoma cell populations. This work demonstrates the feasibility of mass cytometry to identify and characterize tumor heterogeneity. Work is underway to define phenotypes within the epithelial and leukocyte populations that could predict tumor behavior and immune response within the microenvironment. This work is supported by NCI CA196405 to PPM. Citation Format: Deon B. Doxie, Jonathan M. Lehman, Yong Zou, Maria S. Ortega, Caroline E. Maier, Jonathan M. Irish, Pierre P. Massion. Single cell mass cytometry analysis of human lung adenocarcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2911. doi:10.1158/1538-7445.AM2017-2911

Abstract 2000: Targeted characterization of tumor heterogeneity through RNA-Seq analysis of phenotypically defined subpopulations

Abstract Cancer progression and metastases have been linked to the complex heterogeneity found within many solid tumors. In order to understand the complexities within solid tumors there is a need for in-depth characterization of the tumor and the surrounding microenvironment, including the identification and measurement of the subpopulations found within the diseased tissue. Here we describe a method to interrogate the heterogeneity within patient derived xenograft (PDX) derived solid tumors through tissue dissociation, single cell flow sorting, and expression analysis of the resulting subpopulations. Cell sorting enables the identification and capture of phenotypically defined subpopulations within the tumor microenvironment for additional downstream analysis. Whole transcriptome RNA-Seq was performed on samples collected throughout the workflow to characterize the expression changes in cells during the different steps of the process. Expression differences were measured before and after tissue dissociation, before and after incubation, and before and after cell sorting. Expression differences were also measured between phenotypically defined subpopulations sorted by the CD133 and CD49f surface markers. Several significant differentially expressed gene combinations between samples were observed. These differences involved unique pathway characteristics between the different subpopulations sorted based on expression of CD133 and CD49f markers. In contrast, changes induced by the workflow process alone were restricted to a small and well-defined gene subset of immediate early genes (IEGs). Within the workflow, these changes were restricted to the tumor dissociation step and no changes were detected after incubating and/or cell sorting the dissociated cells. Our results indicate that solid tumor dissociation followed by flow cytometric analysis and sorting enables the interrogation of phenotypically defined subpopulations found within solid tumors. Since workflow induced changes are minimal and defined, these methods reveal biological differences between subpopulations within and between cancers. Differential expression of numerous surface marker genes were identified by our method allowing discovery of other candidate surface markers that may extend the subpopulation structure. Once verified these additional surface markers could serve to elucidate the dynamic changes cancer tissue undergoes during progression as well as in response to selective pressures such as drug therapy. Citation Format: Warren Porter, Eileen Snowden, Friedrich Hahn, Mitchell Ferguson, Frances Tong, Shannon Dillmore, Rainer Blaesius. Targeted characterization of tumor heterogeneity through RNA-Seq analysis of phenotypically defined subpopulations [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2000. doi:10.1158/1538-7445.AM2017-2000

Abstract 2866: Volumetric optoacoustic imaging of tumor cell death using a targeted imaging agent

Abstract Multispectral optoacoustic tomography (MSOT) generates high-resolution cross-sectional images in less than a second. MEDI3039, is a newly described agonist of the TNF-related apoptosis-inducing ligand receptor2 (TRAILR2), which has been shown in preclinical studies to preferentially induce cell death in cancer versus normal cells1. We show here that MSOT, when used with a near infra-red (NIR) fluorophore-labelled protein domain of Synaptotagmin-I (C2Am-750, ~15kDa) that binds to phosphatidylserine (PS) exposed by apoptotic and necrotic cells2, can be used to image MEDI3039-induced cell death within the entire tumor region. Non-specific probe retention was assessed using a site-directed mutant (iC2Am-680 or iC2Am-750). PS-specific binding of C2Am-750 to MEDI3039-treated TRAIL-sensitive Colo205 and TRAIL-resistant HT-29 human colon adenocarcinoma cells in vitro was evaluated by flow cytometry and confocal microscopy. The capability of C2Am-750 to detect cell death in vivo was assessed in mice bearing Colo205 or HT-29 xenografts, following a single dose (0.4 mg/kg, i.v.) of MEDI3039. All mice then received an i.v. injection of a 1:1 mixture of 0.1 µmole/kg C2Am-750 and iC2Am-680 16 h post treatment, followed by fluorescence imaging (FLI) measurements using an IVIS 200 camera at 0 and 3 h post probe injection. FLI measurements in treated Colo205 xenografts showed significantly increased retention of C2Am-750 uptake versus size-matched untreated tumors (6.0-fold, P value &amp;lt;0.0001), but there was no significant difference between treated and untreated HT-29 xenografts. iC2Am-680 retention was negligible in all groups. Next, volumetric MSOT measurements of C2Am-750 and iC2Am-750 (0.2 µmole/kg) retention in MEDI3039-treated Colo205 tumors were made using an iTheraMedical inVision 256 system at different wavelengths (660-900nm) and time points (3-24 h). C2Am-750 signal was markedly increased in tumors at 3 h post probe injection. Maximal C2Am-750 signal appeared in the center of treated tumors, whereas there was negligible iC2Am-750 retention. Subsequent histological analyses showed that C2Am-750 signal was strongly correlated with both cleaved caspase-3 and TUNEL staining of dead cells. We have shown that real-time volumetric MSOT measurements with a NIR fluorophore-labelled C2Am, can be used to detect early tumor cell death within the entire tumor region following TRAILR2 agonist treatment. This study shows that this imaging technique can be used to characterize tumor heterogeneity and to determine the most appropriate therapy at an early stage post drug administration. 1. Swers, J.S. et al. Multivalent scaffold proteins as superagonists of TRAIL receptor 2-induced apoptosis. Molecular cancer therapeutics 12, 1235-1244 (2013). 2. Alam, I.S. et al. Comparison of the C2A domain of synaptotagmin-I and annexin-V as probes for detecting cell death. Bioconjugate chemistry 21, 884-891 (2010). Citation Format: Bangwen Xie, Michal Tomaszewski, André A. Neves, Stefanie R. Mullins, David Tice, Richard Sainson, Sarah Bohndiek, Robert W. Wilkinson, Kevin M. Brindle. Volumetric optoacoustic imaging of tumor cell death using a targeted imaging agent [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2866. doi:10.1158/1538-7445.AM2017-2866

Computational Methods for Characterizing Cancer Mutational Heterogeneity.

Advances in DNA sequencing technologies have allowed the characterization of somatic mutations in a large number of cancer genomes at an unprecedented level of detail, revealing the extreme genetic heterogeneity of cancer at two different levels: inter-tumor, with different patients of the same cancer type presenting different collections of somatic mutations, and intra-tumor, with different clones coexisting within the same tumor. Both inter-tumor and intra-tumor heterogeneity have crucial implications for clinical practices. Here, we review computational methods that use somatic alterations measured through next-generation DNA sequencing technologies for characterizing tumor heterogeneity and its association with clinical variables. We first review computational methods for studying inter-tumor heterogeneity, focusing on methods that attempt to summarize cancer heterogeneity by discovering pathways that are commonly mutated across different patients of the same cancer type. We then review computational methods for characterizing intra-tumor heterogeneity using information from bulk sequencing data or from single cell sequencing data. Finally, we present some of the recent computational methodologies that have been proposed to identify and assess the association between inter- or intra-tumor heterogeneity with clinical variables.

Understanding Changes in Tumor Texture Indices in PET: A Comparison Between Visual Assessment and Index Values in Simulated and Patient Data

The use of texture indices to characterize tumor heterogeneity from PET images is being increasingly investigated in retrospective studies, yet the interpretation of PET-derived texture index values has not been thoroughly reported. Furthermore, the calculation of texture indices lacks a standardized methodology, making it difficult to compare published results. To allow for texture index value interpretation, we investigated the changes in value of 6 texture indices computed from simulated and real patient data. Methods: Ten sphere models mimicking different activity distribution patterns and the 18F-FDG PET images from 54 patients with breast cancer were used. For each volume of interest, 6 texture indices were measured. The values of texture indices and how they changed as a function of the activity distribution were assessed and compared with the visual assessment of tumor heterogeneity. Results: Using the sphere models and real tumors, we identified 2 sets of texture indices reflecting different types of uptake heterogeneity. Set 1 included homogeneity, entropy, short-run emphasis, and long-run emphasis, all of which were sensitive to the presence of uptake heterogeneity but did not distinguish between hyper- and hyposignal within an otherwise uniform activity distribution. Set 2 comprised high-gray-level-zone emphasis and low-gray-level-zone emphasis, which were mostly sensitive to the average uptake rather than to the uptake local heterogeneity. Four of 6 texture indices significantly differed between homogeneous and heterogeneous lesions as defined by 2 nuclear medicine physicians (P < 0.05). All texture index values were sensitive to voxel size (variations up to 85.8% for the most homogeneous sphere models) and edge effects (variations up to 29.1%). Conclusion: Unlike a previous report, our study found that variations in texture indices were intuitive in the sphere models and real tumors: the most homogeneous uptake distribution exhibited the highest homogeneity and lowest entropy. Two families of texture index reflecting different types of uptake patterns were identified. Variability in texture index values as a function of voxel size and inclusion of tumor edges was demonstrated, calling for a standardized calculation methodology. This study provides guidance for nuclear medicine physicians in interpreting texture indices in future studies and clinical practice.

Integrative analysis of protein-coding and non-coding RNAs identifies clinically relevant subtypes of clear cell renal cell carcinoma

Protein-coding genes and non-coding RNAs cooperate mutually in cells. Integrative analysis of protein-coding and non-coding RNAs may facilitate characterizing tumor heterogeneity. We introduced integrated consensus clustering (ICC) method to integrate mRNA, miRNA and lncRNA expression profiles of 431 primary clear cell renal cell carcinomas (ccRCCs). We identified one RCC subgroup easily misdiagnosed as ccRCC in clinic and four robust ccRCC subtypes associated with distinct clinicopathologic and molecular features. In subtype R1, AMPK signaling pathway is significantly upregulated, which may improve the oncologic-metabolic shift and partially account for its best prognosis. Subtype R2 has more chromosomal abnormities, higher expression of cell cycle genes and less expression of genes in various metabolism pathways, which may explain its more aggressive characteristic and the worst prognosis. Moreover, much more miRNAs and lncRNAs are significantly upregulated in R2 and R4 respectively, suggesting more important roles of miRNAs in R2 and lncRNAs in R4. Triple-color co-expression network analysis identified 28 differentially expressed modules, indicating the importance of cooperative regulation of mRNAs, miRNAs and lncRNAs in ccRCC. This study establishes an integrated transcriptomic classification which may contribute to understanding the heterogeneity and implicating the treatment of ccRCC.

Abstract 2710: Model evolution technique as a novel concept for characterization of tumor heterogeneity in dynamic contrast enhanced MRI studies

Abstract Introduction: Many studies have shown that tumor vascular network and the assortment of tumorous cells inside and on the periphery of solid tumors are spatially heterogeneous. Variation in cell packing density (VCPD), hypoxia, acidosis, and elevated interstitial fluid pressure (IFP) are main characteristic features of solid tumors. Elevated IFP and VCPD in solid tumors can be generally relevant to the pathological structures at the cellular level that is fundamental to understanding the chance of response to treatment and recurrence. Therefore, non-invasive quantification of tumor heterogeneity for the same types of tumors can play an important role in diagnosis and treatment planning. Hypothesis: In this pilot study, using Nested Model (NM) selection technique, Model Evolution (ME) concept is framed and introduced to quantify the evolutions of 3 different physiologically NM that are derived from standard Tofts model, throughout the course of Dynamic Contrast Enhanced (DCE) Magnetic Resonance Imaging (MRI) experiment. Using ME technique for pharmacokinetic (PK) modeling and DCE-MRI data analysis, a heterogeneity measure is formulated and introduced based on the evolutionary profile of the estimated extra-cellular extra-vascular (ve) volume. We hypothesized that the ME profiles in the course of DCE-MRI experiment, highly depend on the inward diffusion and outward convection of contrast agent concentration and contain abundant information for describing the compartmentalization and heterogeneity levels of solid tumors. Material and Methods: 24 athymic Nude rats with U251n rat tumor model of cerebral tumor were studied. Look-Locker T1 mapping and DCE-MRI experiments (Dual Gradient Echo, 150 image sets at 4.0 sec intervals over 10 min: matrix = 128×64, five 2.0 mm slices, NE = 2, NA = 1, TE/TE/TR = 2.0/4.0/40ms with bolus intravenous injection of the Magnevist at 0.25 mmol/kg) acquired at 7T field strengths. In each animal, in-vivo measurement of tumor IFP was done right after the DCE-MRI experiment using a wick-in-needle technique. The ME technique was applied on DCE-MR data of 24 U251n rat tumors to characterize the heterogeneity of each tumor and then the results were compared to their known in vivo measure of IFPs. Results and Conclusions: Results of this pilot study clearly attest that the evolutionary profile of ve can be used to characterize the heterogeneity level of solid tumors. The ME results imply that as the slop of the evolutionary profile increases, the IFP of tumor increases. Also, the latency of the profile during the course of MR experiment can reliably explain the tumor compartmentalization and their elevated IFP. This pilot study confirms that the ME concept can make a paradigm shift in non-invasive quantification of tumor heterogeneity from DCE-MRI studies Citation Format: Hassan Bagher-Ebadian, Azimeh NV Dehkordi, Rasha Alamgharibi, David Nathanson, Hamid Soltanian-Zadeh, Arbab S. Ali, Stephen Brown, Meser M. Ali, Tom Mikkelsen, James R. Ewing. Model evolution technique as a novel concept for characterization of tumor heterogeneity in dynamic contrast enhanced MRI studies. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 2710.

Multiscale Texture Analysis: From 18F-FDG PET Images to Histologic Images

Three mice bearing orthotopically implanted mammary tumors derived from transgenic MMTV-PyMT mice were scanned with 18F-FDG PET/CT. The tumors were then sliced, and the slices were imaged with autoradiography and stained with hematoxylin and eosin. Six texture indices derived from the PET images, autoradiographic images, and histologic images were compared for their ability to capture heterogeneity on different scales. The PET-derived indices correlated significantly with the autoradiography-derived ones (R = 0.57-0.85), but the values differed in magnitude. The histology-derived indices correlated poorly with the autoradiography- and PET-derived ones (R = 0.06-0.54). All indices were slightly to moderately influenced by the difference in voxel size and spatial resolution in the autoradiographic images. The autoradiography-derived indices differed significantly (P < 0.05) between regions with a high density of cells and regions with a low density and between regions with different spatial arrangements of cells. Heterogeneity derived in vivo from PET images accurately reflects the heterogeneity of tracer uptake derived ex vivo from autoradiographic images. Various tumor-cell densities and spatial cell distributions seen on histologic images can be distinguished using texture indices derived from autoradiographic images despite the difference in voxel size and spatial resolution. Yet, tumor texture derived from PET images only coarsely reflects the spatial distribution and density of tumor cells.