Tumor microenvironment | ПРЕЦИЗИОННАЯ ОНКОЛОГИЯ

Tumor microenvironment

Encyclopedia of Cancer. Springer-Verlag Berlin Heidelberg 2015


Cancer (or tumor) stroma


The specific conditions existing in the tumor tissue, the nonmalignant cells and the molecules present in proximity to the tumor cells.


It is now widely accepted that “Although abnormalities of cancer genes (Oncogene; tumor suppressor genes) are essential contributors to cancer, most abnormalities in these genes occur relatively early in the disease process and none of them is known to be associated with the metastatic stage. It is this final stage – the seeding and growth of satellite lesions in other organs – that is ultimately responsible for the great majority of neoplastic deaths.” It is the tumor microenvironment that determines and shapes the malignancy phenotype of cancer cells, in other words its metastatic behavior (Metastasis).

The tumor tissue can be viewed as an ecosystem composed of two compartments being intimately associated with each other. The first compartment constitutes the malignant cells. The second is the tumor microenvironment composed of resident cells such as fibroblasts, endothelial cells (Tumor-Endothelial Cross-talk), and other nonmalignant cells; of infiltrating cells such as lymphocytes or macrophages (Tumor-associated macrophages) and of numerous molecules released by the tumor cells as well as by the nonmalignant cells. These molecules may be in complex with other molecules, for example, in the extracellular matrix. Other molecules such as growth factors, cytokines, chemokines, antibodies, proteases, other types of enzymes, various metabolites, or drugs may be present in soluble form. The microenvironment of many solid tumors may be characterized by hypoxia (Hypoxia and Tumor Physiology; hypoxia inducible factor-1); low extracellular pH and by low glucose concentration. Cellular products released from necrotic tumor cells are also present.

Although the term tumor microenvironment is used most often with respect to solid tumors, other types of malignancies have also their specific microenvironments. The bone marrow serving as a microenvironment for certain leukemias and for multiple myeloma is a case in point.

Stephen Paget, over 100 years ago, is credited with being the first to postulate the important role played by the microenvironment in metastasis formation. The concept of his “Seed and Soil” theory, explaining site specific metastasis in breast cancer, has been supported and confirmed. However, numerous studies published in the last three decades demonstrate very clearly that the “soil” functions also, or even primarily, as an active “educational/inductive” venue in which cancer cells are directed, by interacting with microenvironmental factors, into one of several molecular evolution pathways. In other words, by exerting regulatory functions and selective pressures, the tumor microenvironment determines and shapes the malignancy phenotype of cancer cells.

The tumor microenvironment is an interaction arena between microenvironmental components and tumor cells and between different microenvironmental components. This arena is characterized by four major hallmarks: complex regulatory circuits; a yin-yang (double edged sword) interplay; plethora of vicious cycles; abnormality of its “normal,” nonmalignant compartment.

Complex regulatory circuits

The major function of the tumor microenvironment is a regulatory one. Many genes in the tumor cells and in nontumor cells residing in or infiltrating to the tumor microenvironment are regulated by microenvironmental components.

Several 100 proteins were identified in the microenvironment of breast cancer. An extremely large number of signaling cascades would operate in this microenvironment even if only a small portion of these proteins would interact with tumor cells or with nontumor cells. It is safe to predict that a similar number of proteins will be detected in the microenvironment of other types of solid tumors.

These signaling cascades take part in the regulation of genes in tumor and in nontumor cells thereby shaping the phenotype of cancer cells and drive their progression.

The regulatory power of the microenvironment can be amplified by the agonistic or antagonistic crosstalk (Receptor cross-talk) between different signaling cascades. Furthermore, several signaling cascades in cancer cells are aberrant. This may well increase the number of combinatorial signaling pathways, augment their complexity, and decrease the capacity of physiological feedback mechanisms to confront these malignancy-associated processes. Another factor contributing to the complexity of the interactions taking place in the tumor microenvironment is tumor heterogeneity. It is thus to be expected that different tumor variants, expressing different profiles of signaling receptors (Receptor tyrosine kinases), would respond differentially to microenvironment-derived signals.

The Yin-yang (double edged sword) interplay in the tumor microenvironment

The cross-talk between tumor cells and microenvironmental factors may result in diametrically opposed effects which could either enhance or block tumor formation or progression. There are several examples of such a yin-yang interaction. The activity of transforming growth factor-beta (TGFβ) is an example for a microenvironmental molecule manifesting a “love–hate relationship” with tumor cells. Whereas TGFβ is a potent inhibitor of normal mammary epithelial cells, it enhances tumor cell invasion and metastasis of advanced breast cancer cells (Epithelial Tumors). Moreover, cancer cells may secrete TGFβ which augments angiogenesis and is capable of suppressing antitumor immune responses of the host. On the other side of the coin, it was recently demonstrated that the progression of pancreatic and of intestinal tumors is enhanced by the inactivation of the TGF signaling cascade.

Another prominent example for yin-yang interplay in the tumor microenvironment is inflammation versus protective tumor immunity. Cells and molecules of the immune system may, under certain circumstances, inhibit tumor growth and under different circumstances promote it.

Vicious cycles in the tumor microenvironment

A vicious cycle may be described as an input event that drives and amplifies other events which, in turn, promote tumor progression. Among such activities, the input event may also augment itself (positive feedback). A well studied vicious cycle in the tumor microenvironment is the cross-talk between osteoblasts, osteoclasts and other microenvironmental factors on the one hand and breast, prostate (Prostate cancer clinical oncology) and lung (Lung cancer) tumor cells on the other hand. This cross-talk promotes bone metastasis (Bone tropism). Tumor-derived molecules such as cytokines cause either an osteoblastic or an osteolytic response. Such molecules feedback on the tumor and on various cells in the microenvironment causing the release of factors driving tumor progression.

Another example of a vicious cycle in the tumor microenvironment is the cross-talk between chemokines (Chemoattraction; chemokine) such as CCL2 and CCL5 secreted from mammary tumors of mice or from breast cancer cells of humans and cytokines such as TNFα secreted from macrophages infiltrating into these tumors. The tumor-derived chemokines attract monocytes to the microenvironment. These monocytes differentiate into macrophages which secrete TNFα. This cytokine up regulates the secretion of CCL2 and CCL5 from the tumor cells. CCL2 and CCL5, in turn, promote the secretion of TNFα from the tumor-associated macrophages. In this vicious cycle, the tumor cells and the macrophages promote each other’s ability to express and secrete pro malignancy factors.

Hypoxia (Hypoxia and tumor physiology; hypoxia inducible factor-1) characterizes the microenvironment of solid tumors. Hypoxia-induced changes in the proteome (Proteomics) may lead to either impairment of tumor growth and spread or, alternatively, to tumor propagation and progression. In the later case, a vicious cycle is created in which tumor cells surviving and propagating under hypoxia will aggravate the state of tumor hypoxia which in turn promotes genomic instability and further progression.

The nontumor cells in the tumor microenvironment may express a different phenotype than their counterparts at distant sites

The conditions in the tumor microenvironment, for example, hypoxia, may induce or promote genetic instability and cause mutations and alterations in gene expression profiles of cancer cells (Genetic polymorphisms). It is not unlikely that such conditions may induce genetic alterations also in nontumor cells present in the microenvironment. The question, if the phenotype and functions of nontumor cells in the tumor microenvironment are similar or different from those of their counterparts in normal microenvironments, is by and large open. Several studies clearly demonstrate that at least some of the nontumor cells in the tumor microenvironment may not represent faithfully the characteristics of their counterparts in other sites of the body. Cancer-associated fibroblasts and endothelial cells are two prominent examples that illustrate the abnormality of tumor-associated nontumor cells.

Cancer-associated fibroblasts have genetic changes both at the DNA level as well as at the expression level. For example, fibroblasts in human carcinomas have tumor suppressor gene mutations. DNA microarrays (Microarray (cDNA) technology) identified over 100 genes differentially expressed by prostate carcinoma (Prostate cancer, clinical oncology)-derived fibroblasts and by systemically derived ones. These alterations may well manifest themselves by altered functions of such tumor-associated cells. Similar findings were also reported for endothelial cells. Cytogenetic abnormalities have been shown to occur in tumor endothelium and such cells may also express proteins that are not expressed by endothelial cells of the corresponding normal tissue.

Site specific metastasis and the metastatic microenvironment

Paget already realized that there is a predilection of tumors to metastasize to specific organ sites and that the metastatic capacity of a certain tumor is not restricted to a single organ site. Each tumor type has therefore several different potential metastatic microenvironments.

Since the tumor and its microenvironment regulate and shape each other’s phenotype, it is to be expected that the metastases arising in one organ site be different from metastases derived from the very same tumor developing in a different organ site. It is also assumed that different reciprocal signaling cascades take place between metastases and nontumor microenvironmental cells in different metastatic microenvironments. These assumptions are indeed supported by experimental evidence.

Dormant micrometastasis and microenvironmental control

Many organs including those of healthy people harbor solitary tumor cells or very small clusters of such cells. These cells do not proliferate either due to a balance between proliferation and apoptosis or due to cell cycle arrest.

There is strong experimental support that such cells are precursors for metastasis and that microenvironmental control mechanisms keep these micrometastatic cells under check, i.e., in a state of dormancy. It is not unlikely that breakdown of these control mechanisms is responsible for the awakening of dormant micrometastases and their progression towards frank metastasis.

The tumor microenvironment as target in cancer therapy

Tumor autonomous factors as well as the tumor microenvironment cooperate in the formation of primary tumors and of metastasis. It is therefore not surprising that cells and molecules that originate in the microenvironment and that drive tumor progression do serve as candidates for cancer therapy.

Moreover the fact that nontumor cells in the tumor microenvironment express a different phenotype than that expressed by the corresponding cells in the normal microenvironment (see above) might also be exploited in cancer therapy modalities, targeting these differentially expressed molecules. Numerous clinical trials targeting various components of the tumor microenvironment are in progress.


Baglole CJ, Ray DM, Bernstein SH et al (2006) More than structural cells, fibroblasts create and orchestrate the tumor microenvironment. Immunol Invest 35:297–325

Bierie B, Moses HL (2006) Tumour microenvironment: TGFβ: the molecular Jekyll and Hyde of cancer. Nat Rev Cancer 6:506–520

Fidler IJ (2002) Critical determinants of metastasis. Semin Cancer Biol 12:89–96

Goss PE, Chambers AF (2010) Does tumour dormancy offer a therapeutic targetNat Rev Cancer 10:871–877

Klein-Goldberg A, Maman S, Witz IP (2014). The role played by the microenvironment in site-specific metastasis. Cancer Lett 352:54–58

Maman S, Witz IP (2013) The metastatic microenvironment. In: Shurin MR et al (eds) The tumor microenvironment. Springer Science+Business Media, pp 15–38

Vogelstein B, Kinzler KW (2004) Cancer genes and the pathways they control. Nat Med 10:789–799

Witz IP, Levy-Nissenbaum O (2006) The tumor microenvironment in the post-PAGET era. Cancer Lett 42:1–10

Witz IP (2008) Tumor-microenvironment interactions: dangerous liaisons. Adv Cancer Res 100:203–229

Witz IP (2008) The selectin-selectin ligand axis in tumor progression. Cancer Metastasis Rev 27:19–30

Witz IP (2008) Yin-yang activities and vicious cycles in the tumor microenvironment. Cancer Res 68:9–13

Witz IP (2009) The tumor microenvironment: the making of a paradigm. Cancer Microenviron Suppl 1:9–17