Search

Assessment of PD-L1 Assay Sensitivity and Specificity.

Background: Lung cancer treatment has become increasingly dependent upon invasive biopsies to profile tumors for personalized therapy. Recently, tumor expression of programmed death-ligand 1 (PD-L1) has gained interest as a potential predictor of response to immunotherapy. Circulating biomarkers present an opportunity for tumor profiling without the risks of invasive procedures. We characterized PD-L1 expression within populations of nucleated cells in the peripheral blood of lung cancer patients in hopes of expanding the role of liquid biopsy in this setting.Methods: Peripheral blood samples from a multi-institutional prospective study of patients with clinical diagnosis of lung cancer were subjected to cytomorphometric and immunohistochemical evaluation using single-cell, automated slide-based, digital pathology. PD-L1 expression was determined by immunofluorescence.Results: PD-L1 expression was detected within peripheral circulating cells associated with malignancy (CCAM) in 26 of 112 (23%) non–small cell lung cancer patients. Two distinct populations of nucleated, nonhematolymphoid, PD-L1–expressing cells were identified; cytokeratin positive (CK+, PD-L1+, CD45−) and cytokeratin negative (CK−, PD-L1+, CD45−) cells, both with cytomorphometric features (size, nuclear-to-cytoplasm ratio) consistent with tumor cells. Patients with >1.1 PD-L1(+) cell/mL (n = 14/112) experienced worse overall survival than patients with ≤1.1 PD-L1(+) cell/mL (2-year OS: 31.2% vs. 78.8%, P = 0.00159). In a Cox model adjusting for stage, high PD-L1(+) cell burden remained a significant predictor of mortality (HR = 3.85; 95% confidence interval, 1.64–9.09; P = 0.002).Conclusions: PD-L1 expression is detectable in two distinct cell populations in the peripheral blood of lung cancer patients and is associated with worse survival.Impact: These findings could represent a step forward in the development of minimally invasive liquid biopsies for the profiling of tumors. Cancer Epidemiol Biomarkers Prev; 26(7); 1139–45. ©2017 AACR.

Cox Proportional Hazards Model of Patients With Confirmed Lung Cancer.

// Bigang Liu 1, * , Shuai Gong 1, 2, * , Qiuhui Li 1, 3 , Xin Chen 1, 3 , John Moore 1 , Mahipal V. Suraneni 1 , Mark D. Badeaux 1 , Collene R. Jeter 1 , Jianjun Shen 1 , Rashid Mehmood 3 , Qingxia Fan 2 and Dean G. Tang 1, 3, 4 1 Department of Molecular Carcinogenesis, University of Texas M.D Anderson Cancer Center, Science Park, Smithville, TX 78957, USA 2 Department of Oncology, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, Henan 450052, China 3 Department of Pharmacology & Therapeutics, Roswell Park Cancer Institute, Buffalo, NY 14263, USA 4 Cancer Stem Cell Institute, Research Center for Translational Medicine, East Hospital, Tongji University School of Medicine, Shanghai 200120, China * These authors have contributed equally to this work Correspondence to: Dean G. Tang, email: Dean.Tang@Roswellpark.org Keywords: NanogP8, Nanog, prostate, prostate cancer, stem cells Received: February 13, 2017 Accepted: March 21, 2017 Published: April 18, 2017 ABSTRACT This project was undertaken to address a critical cancer biology question: Is overexpression of the pluripotency molecule Nanog sufficient to initiate tumor development in a somatic tissue? Nanog1 is critical for the self-renewal and pluripotency of ES cells, and its retrotransposed homolog, NanogP8 is preferentially expressed in somatic cancer cells. Our work has shown that shRNA-mediated knockdown of NanogP8 in prostate, breast, and colon cancer cells inhibits tumor regeneration whereas inducible overexpression of NanogP8 promotes cancer stem cell phenotypes and properties. To address the key unanswered question whether tissue-specific overexpression of NanogP8 is sufficient to promote tumor development in vivo , we generated a NanogP8 transgenic mouse model, in which the ARR 2 PB promoter was used to drive NanogP8 cDNA. Surprisingly, the ARR 2 PB-NanogP8 transgenic mice were viable, developed normally, and did not form spontaneous tumors in >2 years. Also, both wild type and ARR 2 PB-NanogP8 transgenic mice responded similarly to castration and regeneration and castrated ARR 2 PB-NanogP8 transgenic mice also did not develop tumors. By crossing the ARR 2 PB-NanogP8 transgenic mice with ARR 2 PB-Myc (i.e., Hi-Myc) mice, we found that the double transgenic (i.e., ARR 2 PB-NanogP8; Hi-Myc) mice showed similar tumor incidence and histology to the Hi-Myc mice. Interestingly, however, we observed white dots in the ventral lobes of the double transgenic prostates, which were characterized as overgrown ductules/buds featured by crowded atypical Nanog-expressing luminal cells. Taken together, our present work demonstrates that transgenic overexpression of NanogP8 in the mouse prostate is insufficient to initiate tumorigenesis but weakly promotes tumor development in the Hi-Myc mouse model.

Approximately 15% of men with newly diagnosed prostate cancer have high risk features which increase the risk of recurrence and metastasis. Better predictive biomarkers could allow for earlier detection of biochemical recurrence and change surveillance and adjuvant treatment paradigms. Circulating tumor cells are thought to represent the earliest form of metastases. However, their role as biomarkers in men with high risk, localized prostate cancer is not well defined.Two to 5 months after prostatectomy we obtained blood samples from 37 patients with high risk, localized prostate cancer, defined as stage T3a or higher, Gleason score 8 or greater, or prostate specific antigen 20 ng/ml or greater. Circulating tumor cells were enumerated using a commercial platform. Matched tumor and single circulating tumor cell sequencing was performed.Circulating tumor cells were detected in 30 of 37 samples (81.1%) with a median of 2.4 circulating tumor cells per ml (range 0 to 22.9). Patients with detectable circulating tumor cells showed a trend toward shorter recurrence time (p=0.12). All patients with biochemical recurrence had detectable circulating tumor cells. Androgen receptor over expression was detected in 7 of 37 patients (18.9%). Patients with biochemical recurrence had more circulating tumor cell copy number aberrations (p=0.027). Matched tumor tissue and single circulating tumor cell sequencing revealed heterogeneity.We noted a high incidence of circulating tumor cell detection after radical prostatectomy and shorter time to biochemical recurrence in men with a higher circulating tumor cell burden and more circulating tumor cell copy number aberrations. Genomic alterations consistent with established copy number aberrations in prostate cancer were detectable in circulating tumor cells but often discordant with cells analyzed in bulk from primary lesions. With further testing in appropriately powered cohorts early circulating tumor cell detection could be an informative biomarker to assist with adjuvant treatment decisions.

// Xin Liu 1, * , Xin Chen 1, * , Kiera Rycaj 1, * , Hsueh-Ping Chao 1, 3 , Qu Deng 1, 3 , Collene Jeter 1 , Can Liu 1 , Sofia Honorio 1 , Hangwen Li 1 , Tammy Davis 1 , Mahipal Suraneni 1 , Brian Laffin 1 , Jichao Qin 1 , Qiuhui Li 1 , Tao Yang 2 , Pamela Whitney 1 , Jianjun Shen 1 , Jiaoti Huang 4 , Dean G. Tang 1, 2, 3, 5 1 Department of Epigenetics and Molecular Carcinogenesis, University of Texas MD Anderson Cancer Center, Science Park, Smithville, TX 78957, USA 2 Cancer Stem Cell Institute, Research Center for Translational Medicine, East Hospital, Tongji University School of Medicine, Shanghai 200120, China 3 Program in Molecular Carcinogenesis, University of Texas Graduate School of Biomedical Sciences (GSBS), Houston, TX 77030, USA 4 Department of Pathology and Laboratory Medicine, David Geffen School of Medicine, UCLA, Los Angeles, CA 90095, USA 5 Centers for Cancer Epigenetics, Stem Cell and Developmental Biology, RNA Interference and Non-Coding RNAs, and Molecular Carcinogenesis, University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA * These authors have contributed equally to this work Correspondence to: Dean G. Tang, e-mail: dtang@mdanderson.org Keywords: prostate cancer, cancer stem cells, stem cells, differentiation, heterogeneity Received: April 15, 2015 Accepted: June 12, 2015 Published: June 24, 2015 ABSTRACT Human cancers are heterogeneous containing stem-like cancer cells operationally defined as cancer stem cells (CSCs) that possess great tumor-initiating and long-term tumor-propagating properties. In this study, we systematically dissect the phenotypic, functional and tumorigenic heterogeneity in human prostate cancer (PCa) using xenograft models and >70 patient tumor samples. In the first part, we further investigate the PSA −/lo PCa cell population, which we have recently shown to harbor self-renewing long-term tumor-propagating cells and present several novel findings. We show that discordant AR and PSA expression in both untreated and castration-resistant PCa (CRPC) results in AR + PSA + , AR + PSA − , AR − PSA − , and AR − PSA + subtypes of PCa cells that manifest differential sensitivities to therapeutics. We further demonstrate that castration leads to a great enrichment of PSA −/lo PCa cells in both xenograft tumors and CRPC samples and systemic androgen levels dynamically regulate the relative abundance of PSA + versus PSA −/lo PCa cells that impacts the kinetics of tumor growth. We also present evidence that the PSA −/lo PCa cells possess distinct epigenetic profiles. As the PSA −/lo PCa cell population is heterogeneous, in the second part, we employ two PSA − (Du145 and PC3) and two PSA + (LAPC9 and LAPC4) PCa models as well as patient tumor cells to further dissect the clonogenic and tumorigenic subsets. We report that different PCa models possess distinct tumorigenic subpopulations that both commonly and uniquely express important signaling pathways that could represent therapeutic targets. Our results have important implications in understanding PCa cell heterogeneity, response to clinical therapeutics, and cellular mechanisms underlying CRPC.

Prostate cancer (PCa) is heterogeneous and contains both differentiated and undifferentiated tumor cells, but the relative functional contribution of these two cell populations remains unclear. Here we report distinct molecular, cellular, and tumor-propagating properties of PCa cells that express high (PSA(+)) and low (PSA(-/lo)) levels of the differentiation marker PSA. PSA(-/lo) PCa cells are quiescent and refractory to stresses including androgen deprivation, exhibit high clonogenic potential, and possess long-term tumor-propagating capacity. They preferentially express stem cell genes and can undergo asymmetric cell division to generate PSA(+) cells. Importantly, PSA(-/lo) PCa cells can initiate robust tumor development and resist androgen ablation in castrated hosts, and they harbor highly tumorigenic castration-resistant PCa cells that can be prospectively enriched using ALDH(+)CD44(+)α2β1(+) phenotype. In contrast, PSA(+) PCa cells possess more limited tumor-propagating capacity, undergo symmetric division, and are sensitive to castration. Altogether, our study suggests that PSA(-/lo) cells may represent a critical source of castration-resistant PCa cells.

15-Lipoxygenase 2 (15-LOX2), a lipid-peroxidizing enzyme, is mainly expressed in the luminal compartment of the normal human prostate, and is often decreased or lost in prostate cancer. Previous studies from our lab implicate 15-LOX2 as a functional tumor suppressor. To better understand the biological role of 15-LOX2 in vivo, we generated prostate-specific 15-LOX2 transgenic mice using the ARR2PB promoter. Unexpectedly, transgenic expression of 15-LOX2 or 15-LOX2sv-b, a splice variant that lacks arachidonic acid-metabolizing activity, resulted in age-dependent prostatic hyperplasia and enlargement of the prostate. Prostatic hyperplasia induced by both 15-LOX2 and 15-LOX2sv-b was associated with an increase in luminal and Ki-67(+) cells; however, 15-LOX2-transgenic prostates also showed a prominent increase in basal cells. Microarray analysis revealed distinct gene expression profiles that could help explain the prostate phenotypes. Strikingly, 15-LOX2, but not 15-LOX2sv-b, transgenic prostate showed upregulation of several well-known stem or progenitor cell molecules including Sca-1, Trop2, p63, Nkx3.1 and Psca. Prostatic hyperplasia caused by both 15-LOX2 and 15-LOX2sv-b did not progress to prostatic intraprostate neoplasia or carcinoma and, mechanistically, prostate lobes (especially those of 15-LOX2 mice) showed a dramatic increase in senescent cells as revealed by increased SA-betagal, p27(Kip1) and heterochromatin protein 1gamma staining. Collectively, our results suggest that 15-LOX2 expression in mouse prostate leads to hyperplasia and also induces cell senescence, which may, in turn, function as a barrier to tumor development.

Normal human prostate (NHP) epithelial cells undergo senescence in vitro and in vivo, but the underlying molecular mechanisms remain obscure. Here we show that the senescence of primary NHP cells, which are immunophenotyped as intermediate basal-like cells expressing progenitor cell markers CD44, α2β1, p63, hTERT, and CK5/CK18, involves loss of telomerase expression, up-regulation of p16, and activation of p53. Using genetically defined manipulations of these three signaling pathways, we show that p16 is the primary determinant of the NHP cell proliferative capacity and that hTERT is required for unlimited proliferative life span. Hence, suppression of p16 significantly extends NHP cell life span, but both p16 inhibition and hTERT are required to immortalize NHP cells. Importantly, immortalized NHP cells retain expression of most progenitor markers, demonstrate gene expression profiles characteristic of proliferating progenitor cells, and possess multilineage differentiation potential generating functional prostatic glands. Our studies shed important light on the molecular mechanisms regulating the proliferative life span of NHP progenitor cells.

Purpose Androgen receptor (AR) is frequently detected in breast cancers, and AR-targeted therapies are showing activity in AR-positive (AR+) breast cancer. However, the role of AR in breast cancers is still not fully elucidated and the biology of AR in breast cancer remains incompletely understood. Circulating tumor cells (CTCs) can serve as prognostic and diagnostic tools, prompting us to measure AR protein expression and conduct genomic analyses on CTCs in patients with metastatic breast cancer. Methods Blood samples from patients with metastatic breast cancer were deposited on glass slides, subjected to nuclear staining with DAPI, and reacted with fluorescent-labeled antibodies to detect CD45, cytokeratin (CK), and biomarkers of interest (AR, estrogen receptor [ER], and HER2) on all nucleated cells. The stained slides were scanned and enumerated by non-enrichment-based non-biased approach independent of cell surface epithelial cell adhesion molecule (EpCAM) using the Epic Sciences CTC platform. Data were analyzed using established digital pathology algorithms. Results Of 68 patients, 51 (75%) had at least 1 CTC, and 49 of these 51 (96%) had hormone-receptor-positive (HR+)/HER2-negative primary tumors. AR was expressed in CK+ CTCs in 10 patients. Of these 10 patients, 3 also had ER expression in CK+ CTCs. Single cell genomic analysis of 78 CTCs from 1 of these 3 patients identified three distinct copy number patterns. AR+ cells had a lower frequency of chromosomal changes than ER+ and HER2+ cells. Conclusions CTC enumeration and analysis using no enrichment or selection provides a non-biased approach to detect AR expression and chromosomal aberrations in CTCs in patients with metastatic breast cancer. The heterogeneity of intrapatient AR expression in CTCs leads to the new hypothesis that patients with AR+ CTCs have heterogeneous disease with multiple drivers. Further studies are warranted to investigate the clinical applicability of AR+ CTCs and their heterogeneity.

Background: Expression of PD-L1 on tumor tissue is associated with improved response to PD-1 and PD-L1 checkpoint inhibitors. Additionally, expression of PD-L1 on infiltrating T cells is associated with immune exhaustion. Currently, PD-L1 analysis as well as measures of infiltrating T cells are examined in surgically removed or biopsied samples, usually taken long before clinical decision points. Additionally, given tumor heterogeneity, metastatic lesions are likely to be under-sampled. We sought to examine expression of PD-L1 on circulating tumor cells (CTCs) and leukocytes cell populations with a non-invasive liquid biopsy. Examining dynamic biomarker changes in longitudinal samples could enable the development of novel diagnostic tools for response prediction and pharmacodynamics studies related to immunotherapy. Methods: Blood samples from cancer patients were received and shipped to Epic Sciences. Contrived blood samples were also developed utilizing spike-in of cancer cell line controls into healthy blood samples. RBCs were lysed, and nucleated cells plated onto glass slides utilizing the Epic Sciences process. Slides are stained with an IF cocktail (CK, DAPI, CD45, PD-L1) and scanned. Approximately 3 million nucleated cells are examined through advanced Digital Pathology pipelines to detect rare CTCs (CK+/-, cancer morphology, CD45-) and leukocytes (CK-, CD45+, leukocyte morphology) and analyze for PD-L1 expression. Results: A limit of detection of 1 cell/mL of blood was analytically validated. Using Epic’s rare cell detection platform, we could detect a wide range of PD-L1+ leukocyte counts in a subset of peripheral blood of cancer patients, observed as low as 0.003% (110 of 3,192,263) of total leukocytes. PD-L1+ CTCs were observed in the presence of both high and low counts of PD-L1+ leukocyte populations. Conclusions: Epic Sciences CTC platform’s low limit of detection coupled with ability to archive patient blood samples allowed precise quantification of PD-L1 expression on CTCs and leucocytes retrospectively. In addition, we have previously demonstrated the platform’s ability to detect leucocyte subtypes such as CD3, CD4 and CD8 cells, which will allow us to further characterize PD-L1 expression in T-lymphocytes and other immune cell types. Development of a liquid biopsy based platform that is capable of simultaneously measuring immune biomarkers in CTCs as well as leucocytes will allow real time assessment of response to immune checkpoints inhibitors. Citation Format: Adam Jendrisak, Jiyun Byun, Mahipal Suraneni, David Lu, Rachel Krupa, Sarah Orr, Ryon Graf, Yipeng Wang, Mark Landers, Ryan Dittamore. Quantification of rare PD-L1 expressing leukocytes and CTCs in peripheral blood of cancer patients [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 683. doi:10.1158/1538-7445.AM2017-683

110 Background: Approximately 15% of men with newly diagnosed prostate cancer (PCa) have high-risk features, many of these patients will recur despite definitive therapy. Better predictive biomarkers could allow for earlier detection of recurrence and change surveillance paradigms. The role of circulating tumor cells (CTCs) as biomarkers in this context is not well defined. Here, we evaluate the ability to detect CTCs from men with high risk, localized PCa after radical prostatectomy (RP) and correlate their presence with prospective clinical data. Methods: Blood samples from 31 patients with high risk, localized PCa were obtained 2-4 months post RP and sent to Epic Sciences on an IRB approved protocol. Nucleated cells were subjected to immunofluorescent (IF) staining for cytokeratin (CK), CD45, and AR N-terminus. CTCs were identified by fluorescent scanners using algorithmic analysis. Cytokeratin expressing (CK+) CTCs were enumerated and subsequently analyzed for AR expression and individually sequenced for copy number variation (CNV) and large scale transitions (LST, a surrogate of genomic instability). Patients were followed prospectively for biochemical recurrence, defined as detectable PSA. Progression free survival (PFS) was calculated using Kaplan-Meier and Cox proportional hazards. Results: CTCs were detected in 87.1% (27/31) samples with an average of 5.6 CTCs/ml (range: 0 – 22.87) detected per patient. AR expression was detected in 12.9% (4/31) of patients. Ninety-nine CTCs from 14 patients were amenable to LST and CNV analyses. 10.1% (10/99) CTCs from 7 patients exhibited higher ( > = 6) LSTs than control WBCs (95% WBCs had LST < 6). Copy number alterations were detected in CTCs in commonly mutated genes in PCa, including AR, MYC, and TP53 amplification and deletions in PTEN and RB1. Patients with higher CTC burdens exhibited a trend toward shorter PFS (hazard ratio: 1.65; 95% confidence interval: 0.7-3.86; p: 0.13). Conclusions: There was a high incidence of CTC detection after RP in this population using a novel platform. We observed a trend toward shorter PFS in those with higher CTC burden. Genomic alterations were detectable in CTCs and consistent with established CNAs in PCa.

As our understanding of the factors which affect efficacy of a targeted therapy increases, the reliance on histopathological analysis of a biomarker has also increased. This is often due to the necessity to weigh the critical factors of a target or biomarker protein in tissue and cellular context. Currently, histopathologic assessment of tumors which aims to project patient clinical outcome utilize drug target response factors, such as relative expression of the drug target of the drug target or a resistance mechanism in the target (tumor) cells or the tumor microenvironment (TME ). Multiple studies have also shown that TME factors such as inflammatory cell content can be prognostic and predictive. For example, adding the Immunoscore assessment to the TNM classification systems improves the accuracy of disease prognosis. This correlation has led to the concept of predictive “immunoprofiling”, which uses an individual's immune system profile to predict that patient's response to immunomodulating antibody therapy. For these reasons, it is critical to evaluate the drug target or biomarker in conjunction with TME features when defining a patient selection strategy. Critical factors such as tissue and/or cellular compartmentalization and tumor heterogeneity direct the interpretation of these measures and their predictive value. To address the need for careful, contextual interpretation of histopathological evaluations, Flagship has built CellMap™ image analysis algorithms to directly measure heterogeneity and the TME components in a whole tissue section. These approaches allow biomarker interpretation in the complex context of spatial, architectural, and morphological information to aid histological definition and quantification. In this study, we utilized a cohort of specimens from 20 colorectal cancer (CRC) patients and immunologically stained them in order to visualize c-Met, as a characteristic and biologically relevant therapy target; and CD3+ and CD8+ to visualize the inflammatory cell environment. Using these immunohistochemical markers as a prototype for a simultaneous evaluation of a molecular target and the TME, we characterized the biomarker and inflammatory content in both the tumor and stroma using our CellMap™ image analysis algorithms. This provided detailed accounting of the molecular target profile (tumor vs stroma; membrane vs cytoplasm vs nucleus), and the “immunoprofile”, which allowed us to quantitatively describe and associate these features relative to each other in the context of heterogeneity. This data demonstrated discrete patterns of association between c-met and the TME, serving as a potentially critical measure which reflects a biological process relevant to disease outcome. These studies demonstrate novel tools which can assess both the prognostic and predictive value of key measurements which reflect complex tumor biology. Citation Format: Joseph S. Krueger, Brian Laffin, Holger Lange, Anthony Milici, Mirza Peljto, Eric Neeley, Mahipal Suraneni, David Young. Evaluating the contribution of heterogeneity and the tumor microenvironment in companion diagnostic approaches. [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 4981. doi:10.1158/1538-7445.AM2014-4981

Establishing reagent specificity during immunohistochemistry (IHC) based biomarker or companion diagnostic (CDx) assays is challenging. Antibody specificity is dictated in part by recognition of 3D confirmation of the target binding, target activity, and and/or epitope's post-translation modifications. Fixation effects pose additional challenges to epitope recognition during IHC assay. For these reasons, different antibodies against the same target biomarker may demonstrate diversity in prevalence, range, and staining patterns over identical specimens. Thus, determining reagent specificity is a critical part of IHC and CDx assay development. Interpretation is further complicated by the pattern of biomarker expression in specific cell types (e.g. tumor v. stroma) or cell compartments (e.g. membrane v. cytosol). These factors may be critical to associate the drug's mechanism of action with efficacy. In the companion diagnostic (CDx) setting, the mechanism of the drug, epitope recognition, and staining features used to interpret and quantify the biomarker to predict patient response requires an evidence-based approach, where all features of an IHC assay are considered and tied empirically to patient response to the drug. In this study, we demonstrate these complexities in gastric cancer specimens, by comparing IHC assays using two antibodies that recognize either intracellular or extracellular domains of c-Met receptor (SP44/ C-term and EP154Y/ N-term) in the context of the ligand for c-Met, Hepatocyte Growth Factor (HGF). Therapeutic antibodies targeting c-Met [such as MetMab® (OA-5D5/ Roche)]; or HGF directly [such as Rilotumumab (AMG 102/ Amgen)], have directly linked c-Met protein expression as revealed by IHC to patient response. Thus, we hypothesized that a link between c-Met protein expression and HGF should be discernible. To test this hypothesis, we used image analysis approaches to determine the staining features of each IHC assay in comparison to each other. Surprisingly, we found little concordance between the two c-Met antibodies in evaluating c-Met and HGF expression. We found distinct populations of c-Met expressing vs HGF expressing specimens, whose HGF-c-Met association differed with the c-Met assay used. We examined the tissue and cell compartmentalization, and identified staining features of each reagent which would aid or impede in interpretation strategies for understanding the relationship between c-Met and HGF. These results suggest that the epitope-specific features of each c-Met antibody determines the relationship with HGF expression, and how quantitative image analysis endpoints can be used to make critical decisions during the development of an IHC companion diagnostic. Note: This abstract was not presented at the meeting. Citation Format: Joseph S. Krueger, Brian Laffin, Holger Lange, Eric Neeley, Mirza Peljto, Mohamed Salama, Mahipal Suraneni, David Young. Evaluation of immunohistochemistry assays against c-Met and HGF to guide companion diagnostic decisions. [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 2842. doi:10.1158/1538-7445.AM2014-2842

In recent years, the tumor microenvironment (TME) has been identified as an important factor influencing the growth and metastasis of the tumor. Multiple studies have shown that adding the “Immunoscore” assessment to the AJCC/UICC-TNM classification system improves the accuracy of outcome prediction, and in some cases outperforms traditional TNM staging. In many instances these studies have been performed utilizing 2-3 independent readers to manually quantify the cells, performed on selective high-powered fields or TMA cores rather than the entire specimen, resulting in variations in counts when different high powered fields (HPFs) or cores are chosen. A key method to increase the throughput and to decrease the variability is to utilize whole slide imaging and computerized image analysis to provide leukocyte counts. An image analysis algorithm which can automatically differentiate tumor from stroma would allow rapid quantification of endpoints in each tissue compartment across the whole specimen. To address these issues, Flagship Biosciences has designed proprietary CellMapTM image analysis algorithm tools to develop an Immunoscore-like paradigm for colorectal cancers (CRCs) to potentially provide new and more accurate TME information to aid in interpretation. Utilizing whole slide imaging (WSI) approaches, CellMap™ allows the quantitation of leukocyte populations (e.g., CD3+, CD8+, FoxP3+) automatically across whole tissue sections. Using this algorithm, leukocyte populations were quantified in sections that have been either singly or dually labeled for inflammatory markers. Our preliminary studies indicate that Immunoscore-like scoring paradigm should be established both in tumor areas and in adjacent stroma to provide the most complete information on the biology of the tumor. We compared the use of this approach in tissue microarray (TMA) cores which generally sample areas of dense tumor mass, and compared automated WSI to manual HPF approaches. Accuracy of the algorithm was demonstrated by comparing data from manual counts to algorithm derived counts using high-powered fields. These data support using CellMap™ in the prospective or retrospective assessment of leukocyte subpopulations in whole slides of clinical samples. This approach will diminish variability in counting, expand the types of endpoints determined, and improve the statistical value of these determinations, thereby facilitating robust TME measurements with clinical value. Note: This abstract was not presented at the meeting. Citation Format: Joseph S. Krueger, Brian Laffin, Holger Lange, Anthony Milici, Eric Neeley, Mirza Peljto, Mahipal Suraneni, David Young. Using whole slide digital image analysis to quantify leukocyte populations in tumor sections. [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 2548. doi:10.1158/1538-7445.AM2014-2548

When developing an immunohistochemistry (IHC) based biomarker or companion diagnostic assay, establishing reagent specificity is very challenging. Because antibodies recognize three dimensional epitopes, epitope recognition may be based on a specific confirmation (activated, receptor occupied, etc) or biological state (glycosylation, cell surface vs cytoplasmic) of the target protein. Furthermore, recognition of the epitope in Formalin-Fixed, Paraffin Embedded (FFPE) specimens imposes additional challenges due to the fixation effects. For these reasons, different antibodies against the same target biomarker may demonstrate different prevalence, range, and staining patterns in the same specimens. Determining reagent specificity is a critical part of IHC assay development; but the typical approaches utilized (such as Western Blotting, etc) do not directly prove specificity in the FFPE setting. Additionally, the presence or absence of the protein in a particular tissue (tumor vs stroma) or cell (membrane/ cytosol/ nuclear) compartment may be critical to associate the drugs mechanism of action with efficacy. In the companion diagnostic (CDx) setting, the mechanism of the drug, epitope recognition, and staining features used to interpret and quantify the biomarker to predict patient response requires an evidence-based approach, where all features of an IHC assay are considered and tied empirically to patient response to the drug. In this study, we demonstrate these complexities in gastric cancer specimens, by comparing IHC assays using two antibodies against the intracellular vs extracellular domains of c-met (SP44/ C-term and EP154Y/ N-term) in the context of the ligand for c-met, Hepatocyte Growth Factor (HGF). Therapeutic antibodies targeting c-met [such as MetMab® (OA-5D5/ Roche)]; or HGF directly [such as Rilotumumab (AMG 102/ Amgen)], have linked c-met protein expression by IHC directly to patient response. Thus, we hypothesized that a link between c-met protein expression and HGF should be discernible. To test this hypothesis, we used image analysis approaches to determine the staining features of each IHC assay in relation to each other. Surprisingly, we found little concordance between the two c-met antibodies in evaluating c-met and HGF expression. We found distinct populations of c-met expressing vs HGF expressing specimens, whose HGF-c-met association differed with the c-met assay used. We examined the tissue and cell compartmentalization, and identified staining features of each reagent which would aid or impede in interpretation strategies for understanding the relationship between c-met and HGF. These results suggest that the epitope-specific features of each c-met antibody determines the relationship with HGF expression, and how quantitative image analysis endpoints can be used to make critical decisions during the development of an IHC companion diagnostic. Citation Information: Mol Cancer Ther 2013;12(11 Suppl):C34. Citation Format: Joseph S. Krueger, Brian Laffin, Mirza Peljto, Mahipal Suraneni, Holger Lange, David Young. Using quantitative image analysis of a putative immunohistochemistry assay against C-met to guide companion diagnostic decisions. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2013 Oct 19-23; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2013;12(11 Suppl):Abstract nr C34.

Prostate cancer (PCa) is a heterogeneous malignancy containing different types of tumor cells. The cellular basis underlying PCa cell heterogeneity and functional importance of different PCa cell populations in maintaining tumor homeostasis and mediating castration-resistant progression remain poorly understood. By whole-genome microarray analysis, time-lapse videomicroscopy, serial tumor transplantations, and other functional assays, here we report the distinct molecular, cell biological, and tumor-propagating properties of PCa cells that express high (i.e., PSA+) and low (PSA−/lo) levels of PSA. PSA−/lo PCa cells are relatively quiescent and resistant to multiple stresses, exhibit high clonogenic potential, and possess long-term tumor-propagating capacity in intact male mice. They preferentially express stem cell-associated genes and epigenetic profiles and can generate PSA+ cells by either asymmetric or symmetric cell division. Of great clinic significance, PSA−/lo PCa cells can initiate robust tumor development in fully castrated hosts and survive experimental androgen-deprivation therapy. In contrast, PSA+ PCa cells, despite being highly tumorigenic in androgen-proficient hosts, possess more limited tumor-propagating capacity, mainly undergo symmetric division, and are sensitive to castration. Our data together suggest that the two populations of PCa cells appear to play differential roles in tumor maintenance and PSA−/lo cells may represent an important source of castration-resistant PCa cells. Our findings have important implications in understanding PCa cell heterogeneity, tumor response to the mainstay antiandrogen therapies, and emergence of castration-resistant PCa. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 476. doi:10.1158/1538-7445.AM2011-476