The National Cancer Institute (NCI) has been an active proponent in stimulating research in the area of tumor-host interactions for over 2 decades and issued its first request for applications in 1993 entitled, “The Role of Tumor Microenvironment in Breast and Prostate Cancer.” The tumor microenvironment (TME) has remained an area of high research priority at the NCI with a number of workshops, think tanks, and initiatives related to this area of science. The scientific workshops were organized around specific areas, and the summaries of the workshops and their recommendations were shared with the cancer community at large, through publications.1 The range of topics included (a) specific cancer sites such as the prostate, head and neck, melanoma, and bone metastasis; and (b) TME aspects of cancer biology such as lymphangiogenesis, lung inflammation and neoplasia, epithelial-stromal interactions and tumor progression, proteolysis and cancer, tumor metastasis, inflammation, and imaging.
In 2003, the NCI Division of Cancer Biology sponsored a series of think tanks, with topics that spanned the entire spectrum of cancer biology; the goal being to identify emerging themes, anticipate research needs, and facilitate progress in those areas. Nine think tanks including TME, inflammation, tumor stem cells, cancer etiology, tumor immunology, and systems biology were organized. One of the major overarching themes that emerged was the critical need to fully characterize the microenvironment of normal and tumor tissues. It was also recommended that the NCI establish a network that would facilitate development of resources and make them available to the research community at large. A report summarizing the recommendations of the think tank was published.1
In response to think tank recommendations, the NCI made a concerted effort in supporting a number of research and training initiatives, focusing on tumor-host interactions. Two initiatives, (a) Applications for Competing Supplements to Develop and Use Organotypic Models of Cancer and (b) Competing Supplements for Equipment Request to Support Studies on TME, were issued in 2002 and 2003, respectively. In 2005, the NCI established a new training program, “New NCI-Sponsored Tumor Microenvironment Training Program: Techniques in the Establishment and Manipulation of Organotypic Model Systems.” Approximately 75 NCI-funded investigators spent up to 3 weeks in 1 of 6 identified “expert” laboratories learned to establish organotypic models in their own laboratories.
The NCI held the first workshop on epithelial-stromal interactions and tumor progression 15 years ago,2 with the intent to catalyze studies for understanding the interrelationship between tumor cells and their microenvironment and to apply this knowledge to the control of tumor progression. It was not fully appreciated at the time that tumors are organs, composed of multiple cell types with an intimately connected and interdependent structural framework. Areas discussed included the dramatic influence of the stroma on tumor cells particularly in early stages of tumor progression, the complexity of signal transduction pathways that are triggered by the cellular microenvironment and influence the cell physiology, and the complexity of the stromal milieu, which includes multiple stromal cell types, soluble factors, and insoluble matrix-associated molecules. It was realized that understanding the cross-talk between multiple pathways impacting tumor development and progression is required for advancements in cancer biology and treatment. Challenges faced by the community were spatial and temporal complexity of the stroma, organ specificity, and dynamic heterogeneity. The NCI workshop held in 2000 was 1 of many think tanks that occurred prior to the launch of the NCI Tumor Microenvironment Network (TMEN). Since then, significant progress has been made in understanding the role of the TME in all stages of cancer by TMEN investigators as well as other NCI programs discussed herein.
The TMEN was first established in 2006 with 2 explicit goals: (1) to encourage fundamental research on the TME focusing on human cancer in order to generate a comprehensive understanding of stromal composition in normal and cancer tissues and how the stroma affects tumor initiation, progression, and metastasis; and (2) to foster research in this area by developing necessary resources and infrastructure. Scientific advances derived directly from TMEN support have already had a broad impact on the TME field and cancer research community. In addition, the TMEN has successfully established interdisciplinary collaborative programs, generated critical resources, and enhanced entry of new investigators in this area of research. The TMEN has generated novel information about the dynamic complexity of tumor-host interactions in different organ systems using human tissues and experimental models. Tumor Microenvironment Network research revealed that the TME comprised a complex set of interactions that occurred at the earliest onset of tumor initiation and continue to evolve through progression and metastasis. Recent evidence from TMEN and other investigators suggests that the TME is also affected by chemotherapy and radiotherapy and contributes to the occurrence of therapeutic resistance. More information on the TMEN program is available at http://tmen.nci.nih.gov/.
The current TMEN (2011–2016) is focusing on many emerging themes including characterization of the stem cell niche; the origin and functional relevance of tumor stroma in tumor progression, metastasis, therapeutic resistance, and recurrence; the role of epigenetic changes in TME in conferring invasive phenotype; delineating the role of immune cells in cancer initiation, progression, and metastasis; aberrant metabolism in tumor and stroma; microbiome and tumor metastasis; characterization of the extracellular matrix (ECM) and its role in the establishment of metastatic niches; functional relevance of tumor dormancy; and targeting tumor stroma. At present, the field of TME research is an area with enormous excitement as new possibilities to target TME are emerging. These advances have stimulated many cancer biologists to incorporate TME studies into their pursuit of cancer biology and therapeutics. Progress is being made in the understanding of the TME, and a few research advances are highlighted in this review.
The NCI program on Molecular and Cellular Characterization of Screen-Detected Lesions (2015–2020) aims to address a critical unmet need in the management of malignancies, which is minimally invasive methods that predict whether a screen-detected lesion is indolent—requiring only careful monitoring—or progressive, thus requiring appropriate intervention. Comprehensive cellular and molecular characterizations of early lesions are needed, including the tumor cell and its microenvironment and identification of features that distinguish indolent from aggressive and/or progressive lesions. This is particularly true for cancers of the breast, prostate, lung, pancreas, and skin. The use of enabling approaches and technologies will allow determination of cellular and molecular phenotypes of early lesions and assessment of the degree to which the behavior of these lesions is predictable or stochastic and will allow better predictions of the fate of early lesions.
The NCI has supported early stages of technology development primarily via the Innovative Molecular Analysis Technologies (IMAT) Program since 1998. Technologies relevant to studying the TME that were developed with IMAT support include functionalization of quantum dots for biomedical research, novel biosensors, and culture platforms to study cellular mechanical properties.3 The IMAT program continues to support innovative technologies developed by independent laboratories and small businesses. Other NCI programs that support TME-related research include those described below.
The Integrative Cancer Biology Program (2004–2015) and the upcoming Cancer Systems Biology Consortium (2016–2021) support a deeper understanding of cancers as integrated systems of genes, networks, and intercellular interactions.
Since 2004, the NCI Cancer Nanotechnology Program has supported rapid advancement and characterization of new nanotechnology discoveries, increasing their transformation into clinically relevant cancer applications. Cancer nanotechnologies are aimed at targeting tumor cells or the TME.4
The National Institutes of Health Common Fund Program has supported new technologies and, in collaboration with the National Center for Advancing Translational Science Tissue Chip for Drug Screening Program (since 2009), supported development of bioengineered devices to mimic various human tissues to better understand drug toxicity and efficacy.5 Efforts are underway to adapt the tissue chip technologies to applications in studying cancer biology.
The Physical Sciences in Oncology Initiative (2009–2014) supports the integration of physical sciences perspectives and approaches in cancer research.6 Centers funded through this program focused on the (1) physics of cancer, (2) evolutionary theory of cancer, (3) information transfer, and (4) deconvoluting the complexity of cancer. In the second phase (2015–2020), areas of research emphasis include the physical dynamics of cancer and spatiotemporal organization and information transfer in cancer. Small vignettes of scientific areas pursued under this program and described here include matrix stiffness and hypoxia as well as a shift from the traditional 2- to 3-dimensional model systems.
NEW MODELS AND TECHNOLOGY IN TME
Three-dimensional (3D) culture models of solid tumors made it feasible to study tumors in the context of their microenvironment.7 Such models closely mimic cell polarity, tissue morphogenesis, and differentiation. Many observations made using 3D models have clinical relevance and mimic what is observed in mouse models and human cancer including signaling pathways, cancer cell migration and invasion, hypoxia, metabolism, and invasive tumor signatures.8–10
Recent emphasis has been placed in generating improved 3D models or organoids that contain not only human tumor cells but also stromal cells (e.g., fibroblasts, immune cells, and endothelial cells). These patient-derived organoids are amenable to genetic and pharmacologic studies and for evaluating intratumoral heterogeneity as well as response to therapeutic interventions. Such experimental perturbations will establish whether organoids can be predictive models in cancer biology and guide treatment strategies.11–13
The mechanical properties of the microenvironment or matrix stiffness have been identified as an important factor in promoting loss of polarity, epithelial-to-mesenchymal transition (EMT), invasion, and metastasis. Utilization of polyacrylamide gels with specific moduli of elasticity has become widespread to study how matrix stiffness affects cell behavior (e.g., cancer cell migration, induction of EMT-inducing transcription factors, invasion, and metastasis).14–17
Microfluidics, when combined with biomimetic tissue-engineered technologies, offer unique opportunities to design models that encompass vascular and lymphatic systems. Dynamic assessment of cellular migration as well as luminal and interstitial pressure and fluid flow can be accomplished using microfluidic technologies, and the devices can be utilized to examine early or late steps of cancer progression as well as for drug discovery.18–20
Single-cell analysis has become important as an approach to understanding tumor heterogeneity with respect to genomics, growth, metastatic potential, and treatment response.21,22 As an example, precise measurements of single-cell mass and density with high spatial resolution have been made possible by a sophisticated suspended microchannel resonator system, which translates mass changes into resonance frequency.23 This novel technology has enabled a better understanding of the ability of metastatic cells to invade through tight intercellular junctions and the cellular physical changes that occur in metastasis.24
Several reviews in this volume elegantly and effectively capture progress in the areas of angiogenesis, endothelial metabolism, antivascular therapy, cancer metastasis including metastasis to specific organs, therapeutic aspects related to chemotherapy/immunotherapy, and biomarkers. A brief summary of additional TME-related research areas follows.
When disseminated tumor cells reach a metastatic site, they can form a metastatic lesion or, alternatively, become dormant. Dormant cells can be reactivated at a later time, frequently leading to incurable disease. Evidence suggests that signals from the microenvironment contribute to the regulation of tumor dormancy.25 Disseminated tumor cells in the bone hematopoietic niche in prostate cancer have been shown to contain a repertoire of ligands and receptors that enable them to lie dormant in the bone marrow hematopoietic stem cell niche.26,27 In human squamous carcinoma cells, signaling pathways that control tumor dormancy have been recently identified.28,29 Similarly, elegant forward genetic screens in mice have been used to identify genes or microRNAs that drive metastatic reactivation of dormant tumor cells.30
TUMOR MICROENVIRONMENT AND EMT
Tumor cells and associated stroma have dynamic and reciprocal interactions, and it is increasingly appreciated that the microenvironment can influence EMT, a process in which epithelial cells transdifferentiate to a more invasive, mesenchymal phenotype.31 Epithelial-to-mesenchymal transition also confers “stemness” to tumor cells, and the role of tumor-associated macrophage–mediated signaling in this process has been recently defined.32 Virtually all stromal cells have been implicated in the induction of EMT in carcinoma cells via secretion of different molecules including growth factors, inflammatory cytokines, or proteases. Epithelial-to-mesenchymal transition–inducing stromal cells include macrophages, CD8 T cells and neutrophils,33–35 and carcinoma-associated fibroblasts.36,37 Collectively, these findings on tumor-stromal interactions suggest that targeting cells in the microenvironment that induce EMT could be a means to inhibit tumor progression.
Recently, nerve cells have been identified as a component of the tumor stroma and shown to mediate tumor-neuronal cell interactions resulting in perineural invasion38, tumor cell growth and nerve growth induction.39 Perineural invasion has now been shown in many human cancers.38
It is now widely accepted that many cells, including tumor cells, shed extracellular vesicles (EVs) of varying sizes from smaller exosomes (40–100 nm) to larger microvesicles (300–1000 nm). Since their discovery in 1983, EVs have been shown to modulate normal physiology and diseases including cancer and are now considered an important mechanism of paracrine signaling and intercellular communication. Recent efforts have been made to understand the biogenesis of EVs, their cargo contents, and their potential to change noncancer epithelial cells into cancer cells in vivo. Extracellular vesicles may provide useful biomarkers for monitoring cancer patients, especially given that EVs can be readily isolated from biological fluids.40 Breast cancer–associated exosomes have been shown to contain specific pre-miRNAs along with a multiprotein complex to process pre-miRNA to mature miRNA that can mediate silencing of target cell mRNA and drastically alter the transcriptome, thereby promoting tumor formation in nontransformed epithelial cells.41
Tumor cell–derived exosomes have recently been shown to contain double-stranded genomic DNA spanning all chromosomes and oncogenic mutations in human pancreatic cancer,42 human chronic myeloid leukemia, colorectal carcinoma, and murine melanoma cell lines.43 In the human pancreatic cancer study, the investigators provided evidence that exosomes contain greater than 10-kb fragments of double-stranded genomic DNA and that mutations can be detected in KRAS and p53. Serum exosomes can be effectively exploited to determine response to therapy and to assess therapeutic resistance.40
EXTRACELLULAR MATRIX (MATRISOME)
A major component of the local TME, or tumor niche, is the ECM, which is often deregulated in cancer and can directly promote cellular transformation and metastasis. Importantly, ECM anomalies also deregulate behavior of stromal cells, facilitate tumor-associated angiogenesis and inflammation, and lead to generation of a tumorigenic microenvironment. A TMEN program at Massachusetts Institute of Technology (MIT) initiated a comprehensive study to characterize and predict genes encoding the “matrisome,” that is, the ensemble of ECM and ECM-associated proteins. The MIT group has established a Web site to facilitate access to this information (http://matrisomeproject.mit.edu). Proteomic analysis of the matrisome in human primary and metastatic tumors indicates that both tumor and stromal cells contribute to the tumor matrix and that the ECM composition of primary tumors varies with metastatic potential.44,45 This promising area of research paves the way to develop new therapeutic interventions targeting the tumor niche.
TME IN THERAPEUTIC RESISTANCE AND RECURRENCE
Recent evidence suggests that therapeutic resistance in cancer is due not only to tumor cells but also to stromal cells of the TME. Genotoxic cancer therapies induce alterations in the stroma, including the production of secreted cytokines and growth factors that promote cancer progression, EMT, stemness, and therapeutic resistance.46 That all stromal cells are not tumor promoting was shown recently where deletion of fibroblasts via genetic ablation or inhibition of sonic hedgehog signaling led to early lethality in mouse models of pancreatic cancer.47,48 Gene expression analysis of human colorectal cancer revealed a molecular mesenchymal subtype that confers therapeutic resistance and poor clinical outcomes. Interestingly, the gene expression pattern of the stroma was found to be more predictive than that of epithelial cells.49,50
Current literature on TME has elucidated many novel mechanisms of migration, invasion, and intravasation/extravasation and has largely focused on heterogeneity of primary tumors. Some areas that need more focused attention are metastasis, technologies, and integration of cancer biology with technologies.
Studies over the last 2 decades have largely focused on the TME of the primary tumor. Because metastasis confers morbidity and mortality, it is important to understand the TME at the metastatic site, including the multiple cell types, ECM characteristics, and tumor-stromal-ECM interactions. Currently, there is a paucity of data in this area. Integration of cancer biology, including the physical aspects of cancer biology, and physical sciences (e.g., engineering, physics), mathematical models, and computational biology will allow further investigation of tumor complexity and the heterogeneity of stromal and tumor cells. Understanding the intricacies and heterogeneity of tumor-stroma interactions will be key for the application of effective targeting strategies.
Advancement of enabling technologies for cancer research has allowed for in vitro studies of cancer phenomena from single cells to multicellular, organotypic cultures. Furthermore, physical science–based studies of forces, diffusion/transport of molecules and energy, electrical potentials, and thermodynamic stability have been enabled by new technologies. Application of these advanced technologies could allow studies of how the physical interactions affect cancerous and normal states at all length scales and contribute to the spatiotemporal complexity of cancer and its treatment. Integration of engineered devices with physics, molecular biology, and imaging techniques will allow for new insights into cancer biology and treatment with respect to the TME.
1. Mohla S. Tumor microenvironment
. J Cell Biochem
. 2007; 101: 801–804.
2. Matrisian LM, Cunha GR, Mohla S. Epithelial-stromal interactions and tumor progression: meeting summary and future directions. Cancer Res
. 2001; 61: 3844–3846.
3. Dickherber A, Sorg B, Divi R, et al. Guest editorial: funding for innovative cancer-relevant technology
development. Lab Chip
. 2014; 14: 3445–3446.
4. Morris SA, Farrell D, Grodzinski P. Nanotechnologies in cancer treatment and diagnosis. J Natl Compr Canc Netw
. 2014; 12: 1727–1733.
5. Fabre KM, Livingston C, Tagleand DA. Organs-on-chips (microphysiological systems): tools to expedite efficacy and toxicity testing in human tissue. Exp Biol Med (Maywood)
. 2014; 239: 1073–1077.
6. Michor F, Liphardt J, Ferrari M, et al. What does physics have to do with cancer? Nat Rev Cancer
. 2011; 11: 657–670.
7. Weigelt B, Ghajar CM, Bisselland MJ. The need for complex 3D culture models to unravel novel pathways and identify accurate biomarkers in breast cancer. Adv Drug Deliv Rev
. 2014; 69–70: 42–51.
8. Dietze EC, Sistrunk C, Miranda-Carboni G, et al. Triple-negative breast cancer in African-American women: disparities versus biology. Nat Rev Cancer
. 2015; 15: 248–254.
9. Gilkes DM, Semenza GL, Wirtzand D. Hypoxia and the extracellular matrix
: drivers of tumour metastasis. Nat Rev Cancer
. 2014; 14: 430–439.
10. Tung JC, Barnes JM, Desai SR, et al. Tumor mechanics
and metabolic dysfunction. Free Radic Biol Med
. 2015; 79: 269–280.
11. Gao D, Vela I, Sboner A, et al. Organoid cultures derived from patients with advanced prostate cancer. Cell
. 2014; 159: 176–187.
12. Karthaus WR, Iaquinta PJ, Drost J, et al. Identification of multipotent luminal progenitor cells in human prostate organoid cultures. Cell
. 2014; 159: 163–175.
13. van de Wetering M, Francies HE, Francis JM, et al. Prospective derivation of a living organoid biobank of colorectal cancer patients. Cell
. 2015; 161: 933–945.
14. Butcher DT, Alliston T, Weaverand VM. A tense situation: forcing tumour progression. Nat Rev Cancer
. 2009; 9: 108–122.
15. Fernandez-Sanchez ME, Barbier S, Whitehead J, et al. Mechanical induction of the tumorigenic beta-catenin pathway by tumour growth pressure [published online ahead of print May 11, 2015]. Nature
16. Sottnik JL, Dai J, Zhang H, et al. Tumor-induced pressure in the bone microenvironment causes osteocytes to promote the growth of prostate cancer bone metastases. Cancer Res
. 2015; 75: 2151–2158.
17. Wei SC, Fattet L, Tsai JH, et al. Matrix stiffness drives epithelial-mesenchymal transition and tumour metastasis through a TWIST1-G3BP2 mechanotransduction pathway. Nat Cell Biol
. 2015; 17: 678–688.
18. Abaci HE, Shuler ML. Human-on-a-chip design strategies and principles for physiologically based pharmacokinetics/pharmacodynamics modeling. Integr Biol (Camb)
. 2015; 7: 383–391.
19. Schuessler TK, Chan XY, Chen HJ, et al. Biomimetic tissue-engineered systems for advancing cancer research: NCI Strategic Workshop report. Cancer Res
. 2014; 74: 5359–5363.
20. Sung KE, Beebeand DJ. Microfluidic 3D models of cancer. Adv Drug Deliv Rev
. 2014; 79–80: 68–78.
21. Irish JM, Doxie DB. High-dimensional single-cell cancer biology. Curr Top Microbiol Immunol
. 2014; 377: 1–21.
22. Bendall SC, Nolanand GP. From single cells to deep phenotypes in cancer. Nat Biotechnol
. 2012; 30: 639–647.
23. Burg TP, Godin M, Knudsen SM, et al. Weighing of biomolecules, single cells and single nanoparticles in fluid. Nature
. 2007; 446: 1066–1069.
24. Byun S, Son S, Amodei D, et al. Characterizing deformability and surface friction of cancer cells. Proc Natl Acad Sci U S A
. 2013; 110: 7580–7585.
25. Yumoto K, Eber MR, Berry JE, et al. Molecular pathways: niches in metastatic dormancy. Clin Cancer Res
. 2014; 20: 3384–3389.
26. Shiozawa Y, Pedersen EA, Havens AM, et al. Human prostate cancer metastases target the hematopoietic stem cell niche to establish footholds in mouse bone marrow. J Clin Invest
. 2011; 121: 1298–1312.
27. Taichman RS, Patel LR, Bedenis R, et al. GAS6 receptor status is associated with dormancy and bone metastatic tumor formation. PLoS One
. 2013; 8. e61873.
28. Adam AP, George A, Schewe D, et al. Computational identification of a p38SAPK-regulated transcription factor network required for tumor cell quiescence. Cancer Res
. 2009; 69: 5664–5672.
29. Sosa MS, Parikh F, Maia AG, et al. NR2F1 controls tumour cell dormancy via SOX9- and RARbeta-driven quiescence programmes. Nat Commun
. 2015; 6: 6170.
30. Gao H, Chakraborty G, Lee-Lim AP, et al. Forward genetic screens in mice uncover mediators and suppressors of metastatic reactivation. Proc Natl Acad Sci U S A
. 2014; 111: 16532–16537.
31. Thiery JP, Acloque H, Huang RY, et al. Epithelial-mesenchymal transitions in development and disease. Cell
. 2009; 139: 871–890.
32. Lu H, Clauser KR, Tam WL, et al. A breast cancer stem cell niche supported by juxtacrine signalling from monocytes and macrophages. Nat Cell Biol
. 2014; 16: 1105–1117.
33. Bonde AK, Tischler V, Kumar S, et al. Intratumoral macrophages contribute to epithelial-mesenchymal transition in solid tumors. BMC Cancer
. 2012; 12: 35.
34. Grosse-Steffen T, Giese T, Giese N, et al. Epithelial-to-mesenchymal transition
in pancreatic ductal adenocarcinoma and pancreatic tumor cell lines: the role of neutrophils and neutrophil-derived elastase. Clin Dev Immunol
. 2012; 2012: 720768.
35. Reiman JM, Knutson KL, Radisky DC. Immune promotion of epithelial-mesenchymal transition and generation of breast cancer stem cells. Cancer Res
. 2010; 70: 3005–3008.
36. Biddle A, Mackenzie IC. Cancer stem cells and EMT in carcinoma [published online ahead of print February 3, 2012]. Cancer Metastasis Rev
37. Sigurdsson V, Hilmarsdottir B, Sigmundsdottir H, et al. Endothelial induced EMT in breast epithelial cells with stem cell properties. PLoS One
. 2011; 6: e23833.
38. Liebig C, Ayala G, Wilks JA, et al. Perineural invasion in cancer: a review of the literature. Cancer
. 2009; 115: 3379–3391.
39. Olar A, He D, Florentin D, et al. Biological correlates of prostate cancer perineural invasion diameter. Hum Pathol
. 2014; 45: 1365–1369.
40. Carvalho J, Oliveira C. Extracellular vesicles—powerful markers of cancer evolution. Front Immunol
. 2014; 5: 685.
41. Melo SA, Sugimoto H, O’Connell JT, et al. Cancer exosomes perform cell-independent microRNA biogenesis and promote tumorigenesis. Cancer Cell
. 2014; 26: 707–721.
42. Kahlert C, Melo SA, Protopopov A, et al. Identification of double-stranded genomic DNA spanning all chromosomes with mutated KRAS and p53 DNA in the serum exosomes of patients with pancreatic cancer. J Biol Chem
. 2014; 289: 3869–3875.
43. Thakur BK, Zhang H, Becker A, et al. Double-stranded DNA in exosomes: a novel biomarker in cancer detection. Cell Res
. 2014; 24: 766–769.
44. Naba A, Clauser KR, Lamar JM, et al. Extracellular matrix
signatures of human mammary carcinoma identify novel metastasis promoters. Elife
. 2014; 3: e01308.
45. Naba A, Clauser KR, Whittaker CA, et al. Extracellular matrix
signatures of human primary metastatic colon cancers and their metastases to liver. BMC Cancer
. 2014; 14: 518.
46. Sun Y, Campisi J, Higano C, et al. Treatment-induced damage to the tumor microenvironment
promotes prostate cancer therapy resistance through WNT16B. Nat Med
. 2012; 18: 1359–1368.
47. Ozdemir BC, Pentcheva-Hoang T, Carstens JL, et al. Depletion of carcinoma-associated fibroblasts and fibrosis induces immunosuppression and accelerates pancreas cancer with reduced survival. Cancer Cell
. 2014; 25: 719–734.
48. Rhim AD, Oberstein PE, Thomas DH, et al. Stromal elements act to restrain, rather than support, pancreatic ductal adenocarcinoma. Cancer Cell
. 2014; 25: 735–747.
49. Calon A, Lonardo E, Berenguer-Llergo A, et al. Stromal gene expression defines poor-prognosis subtypes in colorectal cancer. Nat Genet
. 2015; 47: 320–329.
50. Isella C, Terrasi A, Bellomo SE, et al. Stromal contribution to the colorectal cancer transcriptome. Nat Genet
. 2015; 47: 312–319.