Search

Increased stiffness in the extracellular matrix (ECM) contributes to tumor progression and metastasis. Therefore, stromal modulating therapies and accompanying biomarkers are being developed to target ECM stiffness. Magnetic resonance (MR) elastography can noninvasively and quantitatively map the viscoelastic properties of tumors

Background: Tumor mutational burden (TMB) has emerged as a new biomarker for response to checkpoint inhibitor immunotherapy. Recently, we described a novel assay to calculate TMB from the circulating tumor DNA (ctDNA) in the blood (bTMB) and demonstrated validation to support its application in a prospective phase III trial in 1L NSCLC comparing the anti-PD-L1 agent, atezolizumab, against standard of care chemotherapy. Herein, we demonstrate clinical utility of the bTMB assay across a retrospective study of 794 NSCLC patients from two randomized clinical trials evaluating progression free survival (PFS) and overall survival (OS) between atezolizumab and chemotherapy. Furthermore, we compare TMB between tissue and blood, and evaluate variant-level concordance between the bTMB assay and the CLIA-validated ctDNA assay, FoundationACT (FACT). Methods: Cell free DNA (cfDNA) from plasma was isolated from retrospective clinical samples. The bTMB assay delivered a count of somatic base substitutions in the ctDNA down to 0.5% allele frequency across 394 genes from as little as 1% tumor content, yielding a bTMB score. Tissue TMB was evaluated from formalin-fixed, paraffin embedded specimens using the FoundationOne assay. Results: Patients with bTMB ≥14 mut/Mb were considered high, and demonstrated significant enrichment for both PFS and OS in the phase II POPLAR study evaluating atezolizumab vs. chemotherapy (PFS HR = 0.57, OS HR = 0.56, n=211). Survival results were validated in the larger phase III study, OAK (PFS HR = 0.65, OS HR = 0.64, n=583). Significant enrichment in PFS was observed between bTMB-high and non-high patients receiving atezolizumab, however this effect was not observed for patients receiving chemotherapy. Amongst 259 patients with blood and tissue TMB, overall and positive percent agreement (PPA) in the TMB categorical result were 81.5% and 63.6%, respectively. When bTMB was restricted to the 62 gene FACT assay, the PPA dropped to 17.0%, suggesting a sufficiently sized panel is required to sensitively identify patients with high TMB. When samples were TMB-high in blood, the median percent overlap of TMB variants between blood and tissue was 70%. Blood samples tested in both the bTMB and FACT assays revealed that 93% of variants were detected in both assays across overlapping regions, with ≥99% of variants detected in both platforms with variant allele frequency ≥1.0%. Conclusions: The bTMB assay is analytically validated for high accuracy and precision, and demonstrates a high degree of variant level concordance with the FACT assay. High bTMB is significantly associated with improved survival to atezolizumab vs. chemotherapy in 2L NSCLC, and the survival benefit is specific to atezolizumab and not prognostic. Finally, TMB between blood and tissue is correlated and largely explained by overlapping variants present in both sample types. Citation Format: David Fabrizio, Daniel Lieber, Christine Malboeuf, Jacob Silterra, Emily White, Michael Coyne, Tina Brennan, Jie Ma, Mark Kennedy, Erica Schleifman, Sarah Paul, Yan Li, David Shames, Craig Cummings, Eric Peters, Marcin Kowanetz, Doron Lipson, Geoff Otto. A blood-based next-generation sequencing assay to determine tumor mutational burden (bTMB) is associated with benefit to an anti-PD-L1 inhibitor, atezolizumab [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 5706.

Malignant tumors are typically associated with altered rigidity relative to normal host tissue. Magnetic resonance elastography (MRE) enables the noninvasive quantitation of the mechanical properties of deep-seated tissue following application of an external vibrational mechanical stress to that tissue. In this preclinical study, we used MRE to quantify (kPa) the elasticity modulus Gd and viscosity modulus Gl of three intracranially implanted glioma and breast metastatic tumor models. In all these brain tumors, we found a notable softness characterized by lower elasticity and viscosity than normal brain parenchyma, enabling their detection on Gd and Gl parametric maps. The most circumscribed tumor (U-87 MG glioma) was the stiffest, whereas the most infiltrative tumor (MDA-MB-231 metastatic breast carcinoma) was the softest. Tumor cell density and microvessel density correlated significantly and positively with elasticity and viscosity, whereas there was no association with the extent of collagen deposition or myelin fiber entrapment. In conclusion, although malignant tumors tend to exhibit increased rigidity, intracranial tumors presented as remarkably softer than normal brain parenchyma. Our findings reinforce the case for MRE use in diagnosing and staging brain malignancies, based on the association of different tumor phenotypes with different mechanical properties. Cancer Res; 75(7); 1216–24. ©2015 AACR.

First 2 authors contributed equally Hepatocellular carcinoma (HCC) is the second most common cause of death from cancer worldwide with extremely poor prognosis. HCC is known to be closely associated with liver injury-induced cirrhosis caused by various etiologies, including HBV infection. The prolonged timeline and heterogeneous nature of HCC adds complexity to dissecting the biology of this disease in humans. While Sorafenib is approved for the first-line treatment of metastatic HCC, most patients rapidly progress on treatment with Sorafenib. Consequently, alternative therapeutic options for HCC are much needed. The identification of molecular subtypes and reliable biomarkers associated with disease evolution is critical in facilitating development of new therapeutic agents in HCC. To understand the manifestation of early molecular events in HCC disease progression in human, we analyzed genome-wide RNA-seq data derived from 100 paired samples consisting of HCC tumors (most with 60-80% tumor content) and adjacent cirrhotic tissues from early stage patients (TNM system: T1N0M0, T2N0M0, and T3N0M0; similar to BCLC stage A and B). Differential expression analysis revealed a cluster of genes that significantly differentiated HCC from cirrhotic tissues and illustrated a widespread deregulation of cell cycle machinery modulated by probable molecular abnormalities represented by Polo-Like Kinase, Checkpoint kinases, G2/M DNA damage checkpoint regulation, DNA damage-induced 14-3-3σ signaling, ATM signaling, and estrogen-mediated S-phase entry. Prominent down-regulation of FXR/LXR/RXR activation was also observed in HCC tumors. Strikingly, the unsupervised hierarchical clustering of both cirrhotic and HCC tissues revealed 3 groups of genes with mRNA expression closely correlating with disease progression stage-wise from cirrhosis to T1, T2, and T3 stages. Specifically, we made a novel observation illustrating the stage-wise activation of Wnt signaling pathway, but de-activation of MAPK pathway. Upon in-depth analysis, our data also suggests that as HCC progresses, translation machinery and embryonic morphogenesis are stimulated, while angiogenesis, negative regulation of apoptosis, and mesenchymal cell differentiation are possibly impinged. In addition, we found that components of processes crucial for activating immune response appear to be impaired as disease progresses from cirrhosis to stage T3. To confirm the aforementioned finding through focused assessment of immune-microenvironment by gene expression, we utilized Fluidigm platform and corroborated the down-regulation of effector T cell signature. In conclusion, data presented provides a holistic depiction of evolution of HCC and the associated tumor immunity, thus paving a way for future detailed subtyping and therapeutics discovery. Citation Format: Bonnie P. Liu, Kwame Okrah, Jeff Cheng, Maipelo Motlhabi, Charlie Sun, Teiko Sumiyoshi, Shoji Ikeda, Hartmut Koeppen, Zineb Mounir, Craig Cummings, Nadia Haque, Garret Hampton, Lukas Amler, Mark Lackner, Shih-Min A. Huang. Comprehensive RNA-seq transcriptome interrogation of paired hepatocellular carcinoma and cirrhosis tissues revealed significant molecular features of disease evolution and modulation of tumor immunity. [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 1837.

Currently there are multiple approaches for measuring gene expression from human tissue samples on the market. These methods range from measuring the expression of a few genes using quantitative RT-PCR to measuring every gene in a sample by whole transcriptome RNA sequencing (RNA-Seq). Although many techniques and products for measuring gene expression are available, the highly degraded RNA isolated from formalin-fixed, paraffin embedded (FFPE) tumor tissues still poses a challenge for many of them. FFPE tumor samples are an abundant source of biomarker information and gene expression analysis in these samples has the potential to identify new signatures, biomarkers, or diagnostics that could predict patient response to treatment. Accurately measuring gene expression in a high throughput manner from this sample type has long been a struggle in the field. To determine the ability of RNA-Seq to accurately measure gene expression from FFPE-derived RNA, we utilized a large collection of matched fresh frozen (FF) and FFPE samples from 12 different tumor indications. By comparing the results to the matching FF sample, we were able to determine the accuracy and sensitivity of each platform when using degraded FFPE-derived RNA. The FFPE samples in the collection had a wide range of RNA quality scores (RIN score: 0-6.8, DV200:0-78) representing what is typically isolated from clinical trial samples. We further compared whole transcriptome RNA-Seq with a hybrid capture method, RNA Access, to determine the accuracy and feasibility of using this technology on degraded RNA. RNA Access selects for protein coding transcripts by hybridization, whereas whole transcriptome RNA-seq removes ribosomal RNA, but analyzes everything else i.e. the global transcriptome. All samples were sequenced to an average of 50 or 25 million paired-end reads for whole transcriptome RNA-Seq and RNA Access, respectively and 100 ng of RNA was used as input in both assays. We conclude that both methods can be used for analyzing FFPE tumor samples. A similar dynamic range was observed for both methods and both show similar correlation between FF and FFPE samples within the method (0.76 vs 0.74). The overall concordance between methods was 0.69 and 0.63 for FF and FFPE, respectively. When looking specifically at the detection of 90 low expressing immune related genes, both assays were able to detect >94% with an inter-platform correlation of 0.66 and 0.56 in FF and FFPE, respectively. Whole transcriptome RNA-seq provides the maximum amount of information from clinical samples including anti-sense and non-coding RNA detection while RNA Access can be applied more cost-efficiently when one is mostly interested in the protein coding transcriptome. Citation Format: Erica B. Schleifman, Anna Kiialainen, Andreas Roller, Sabine Bader, Maipelo Motlhabi, Priti Hegde, Ian McCaffery, Garret Hampton, Michael Cannarile, Craig Cummings, Olivia Spleiss, Eric Peters. Whole transcriptome and exome targeted RNA sequencing for FFPE tumor samples from clinical trials. [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 3631.

Accurately measuring the expression of genes in formalin-fixed paraffin embedded (FFPE) tumor tissues has long been a struggle due to the inherent degradation of RNA isolated from these materials. Accurate quantification of gene expression levels in FFPE samples can enable the testing of biomarker hypotheses in the clinic and can potentially be used for patient stratification or selection in clinical trials. Platforms such as Fluidigm, Nanostring and microarrays are currently the high-throughput technologies utilized. Each of these platforms has different drawbacks and challenges such as lack of sensitivity, reproducibility or dynamic range when working with degraded FFPE samples. Currently, whole transcriptome RNA sequencing (RNA-Seq) is the only platform that offers a truly high-throughput, sensitive, and reproducible method to accurately quantify mRNA levels in RNA derived from FFPE samples. RNA-Seq, however has not yet been widely evaluated and adapted for use with degraded FFPE samples. Here we report the direct comparison of whole transcriptome RNA-Seq with two platforms that are currently compatible with FFPE derived RNA, Fluidigm and Nanostring, to determine the accuracy and feasibility of using this technology on degraded RNA. Utilizing a collection of matched fresh frozen (FF) and FFPE samples we analyzed gene expression using all three platforms to allow for a direct evaluation of each technology. By comparing the results to the matching FF sample, we were able to determine the accuracy and sensitivity of each platform when using degraded FFPE derived RNA. By titrating down the input of FFPE RNA into the RNA-Seq library prep we were also able to define the optimal input of FFPE RNA needed to accurately and reproducibly quantify gene expression. This work allows for a systematic comparison of different gene expression platforms for their use with degraded FFPE RNA. Citation Format: Erica B. Schleifman, Maipelo Motlhabi, Craig Cummings, Rin Nakamura, Linda Bosch, Rajesh Patel, An Do, Andrew Watson, Thomas Sandmann, Walter Darbonne, Ian McCaffery, Eric Peters, Rajiv Raja. Comparison of gene expression platforms: RNA-Seq, Fluidigm, and Nanostring. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4901. doi:10.1158/1538-7445.AM2015-4901

Refined imaging strategies that could improve diagnosis and management of children with brain malignancies are urgently required. Pediatric brain malignancies possess distinct underlying biologies that discriminate them from adult tumors, even within a common neuropathological diagnosis such as glioblastoma (GBM). Magnetic resonance elastography (MRE) exploits the ability of MRI to visualize the propagation of shear waves resulting from vibrations applied to the cranium to map and quantify (kPa) brain tissue elasticity (Gd) and viscosity (Gl) in vivo. The altered viscoelastic properties of tumors, combined with the sensitivity of MRE for differences in tumor microstructure, establishes MRE as an attractive modality for the detection and differential diagnosis of brain malignancies. Here we aimed to determine whether the potential of MRE in the neuroradiological management of patients with brain malignancies could be extended to the pediatric population. The viscoelastic properties of two pediatric brain tumor models that faithfully emulate high risk childhood disease were investigated: i) Orthotopic D-212 MG pediatric supratentorial giant cell GBM (H3F3A wildtype, 11 year old patient) xenografts propagated in NCr nu/nu mice and ii) GTML/Trp53KI/KI transgenic mice that spontaneously develop aggressive, MYCN driven, p53 depleted, medulloblastomas. D-212 MG and GTML/Trp53KI/KI tumors were less elastic (lower Gd) and viscous (lower Gl), and therefore softer, than the surrounding brain tissue. Both tumor types were also significantly less elastic (D-212 MG Gd = 3.9±0.2; GTML/Trp53KI/KI Gd = 3.5±0.1) than the soft thalamic parenchyma in non-tumor-bearing mice (Gd = 5.9±0.2; p = 0.001 and p = 0.02, respectively, Mann-Whitney), in addition to being less viscous. Interestingly, GTML/Trp53KI/KI tumors demonstrated a bimodal distribution of Gd, which reflected the more marked transition between their relatively stiffer rim and the softer core compared with D-212 MG tumors. We demonstrate that two representative models of major high risk pediatric brain malignancies share the unique softness previously observed in adult brain tumor models, allowing their detection by MRE. This supports observations that pediatric GBMs are morphologically indistinguishable from adult GBMs. Although median Gd values were not sufficient to discriminate between the tumor types, the marked bimodal distribution of Gd in the GTML/Trp53KI/KI tumors was not apparent in D-212 MG GBM tumors. Whilst ongoing histopathological investigation into growth patterns, vascular, cellular and extracellular networks will aid elucidation of the pathological determinants of the bimodal signature, to date unique to GTML/Trp53KI/KI, these data reinforce the potential of MRE for the detection and differential diagnosis of pediatric brain malignancies based on their mechanical properties. Citation Format: Jin Li, Jessica K.R. Boult, Maria Vinci, Sergey Popov, Karen Barker, Zai Ahmad, Yann Jamin, Craig Cummings, Suzanne A. Eccles, Jeffrey C. Bamber, Ralph Sinkus, Louis Chesler, Chris Jones, Simon P. Robinson. In vivo magnetic resonance elastography in pediatric brain tumor models. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 1488. doi:10.1158/1538-7445.AM2015-1488

While treatment with anti-CTLA-4 antibody can induce clinical responses in advanced cancer patients, its effects on the breadth of the T cell response is unknown. We used a next-generation sequencing-based method to assess T cell repertoire diversity in 46 patients with metastatic castration resistant prostate cancer or metastatic melanoma. Peripheral blood mononuclear cells were obtained from patients prior to and during treatment with anti-CTLA-4 antibody. mRNA was amplified using locus-specific primer sets for T cell receptor (TCR) beta, and the amplified product was sequenced. Sequence reads were used to quantitate absolute TCR frequencies using standardized clonotype determination algorithms with normalization by spiked reference TCR sequences. Following clonotype quantitation, repertoire differences between serial samples were assessed by the Morisita index, a statistical measure of population dispersion. 97 paired samples were assessed, of which 46 (47%) had increases and 22 (23%) had decreases in TCR diversity by more than 2-fold. By comparison, none of 9 untreated sample pairs underwent more than a 2-fold change in diversity (P = 0.005, Fisher's exact test, two tailed). TCR repertoire differences between monthly samples were markedly higher than for time-matched controls. After the first treatment, median Morisita index between samples was 0.197 for treated samples versus 0.039 for untreated (P = 0.0005, Mann-Whitney U test). The median number of clones that significantly changed in abundance was 421 for treated versus 45 for controls. In patients with multiple time points, this rapid clonotype evolution continued through treatment. Whereas the number of clonotypes that increased with treatment was not associated with clinical outcome, improved overall survival was associated with maintenance of high frequency clones present at baseline. In contrast the highest frequency clonotypes fell with treatment in patients with shorter overall survival. Stably maintained clonotypes included T cells possessing high avidity TCR such as CMV-reactive T cells. Together, these results suggest that CTLA-4 blockade induces T-cell repertoire evolution and diversification. Moreover, improved clinical outcomes are associated with less clonotype loss, consistent with the maintenance of high frequency TCR clonotypes during treatment. These clones may represent the presence of pre-existing high avidity T cells that may be relevant in the anti-tumor response. Citation Format: Lawrence H. Fong, Edward Cha, Mark Klinger, Yafei Hou, Craig Cummings, Antoni Ribas, Malek Faham. T-cell repertoire turnover induced by anti-CTLA-4 antibody treatment in cancer patients. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 5017. doi:10.1158/1538-7445.AM2014-5017