Global Burden of Cancer

Worldwide, cancer is responsible for one in eight deaths. The distribution and types of cancer that occur continue to change, being affected primarily by the following: (1) the growth and aging of populations; (2) the increasing encroachment of modifiable risk factors (e.g., cigarette smoking, Western diet, physical inactivity) in developing countries; and (3) the relatively slower decrease in infection-related cancers.³ By 2020, 70% of all cancer-related deaths will be in developing countries, in which survival rates (20% to 30%) are barely half those of developed countries.⁴ Indeed, 80% to 90% of people diagnosed with cancer in developing countries present with late-stage, terminal cancer. It can therefore be seen that the vast majority of cancer deaths will occur in countries least equipped to handle the burden.

Aging and Cancer

Cancer disproportionately affects people 65 years and older. In the United States, this age group comprises 56% of all newly diagnosed cancer patients and 71% of all cancer deaths.⁵ The median age of death for cancers common to both men and women (including lung, colorectal, pancreas, stomach, and urinary bladder) cancers, ranges from 71 to 77 years. The number of people in this age group will double to 70 million (or one in five people) over the next 25 years, driven by the Baby Boom cohort born between 1946 and 1964. This is a recognized trend throughout the developed world.

With an increasingly older population, the incidence of cancer will increase, thereby increasing the overall cancer burden on society. Additionally, cancer care will be of increasingly greater complexity. Reasons for this include that older people have more comorbidities of greater severity in the setting of declining physiologic reserve, difficulties with access to care, and lack of social support.

Cancer treatment in older adults is less well-studied and it has been shown that the older population is underrepresented in clinical trials.^6–8 There have been a number of reports of the underuse of adjuvant therapy, chemotherapy and radiotherapy, in the aging population. O’Connell and colleagues⁹ have studied the Surveillance, Epidemiology, and End Results (SEER) database (1988–1997), and found that although older patients with colorectal or breast cancer had excellent rates of receiving cancer-directed surgery, rates were variable for many other cancers, including lung, esophagus, stomach, liver, and pancreas cancers. Surgical intervention not being recommended was the most common reason. Clearly, the surgeon must more carefully weigh an individual’s operative risk in the context of the difficulty, length and morbidity of the procedure, with greater consideration for quality of life and functional status, beyond just postoperative mortality and mortality and long-term survival.

Obesity and Cancer

Prevalence of overweight (body mass index [BMI], 25 to 30) and obesity (BMI ≥ 30) in most developed countries, and in urban areas of many less developed countries, has been increasing markedly over the past 25 years. In the United States, approximately one third of the population is now classified as obese. Although obesity has long been recognized as an important cause of diabetes and cardiovascular disease, the relationship between obesity and cancer has received less attention. Epidemiologic studies indicate that adiposity contributes to the increased incidence and/or death from cancers of the colon, breast (in postmenopausal women), endometrium, kidney (renal cell), esophagus (adenocarcinoma), gastric cardia, pancreas, gallbladder, and liver (hepatocellular carcinoma). It has been estimated that 15% to 20% of all cancer deaths in the United States can be attributed to overweight and obesity.¹⁰

The mechanisms whereby obesity increases cancer risk appear to involve the metabolic and endocrine effects of obesity via their alterations in peptide and steroid hormone levels. For example, greater amounts of adipose tissue lead to increased circulating levels of free fatty acids. This, in turn, causes liver, muscle and other tissues to increase their usage of fats for energy production, thereby reducing their need for uptake and metabolism of glucose and eventually leading to hyperglycemia. This functional insulin resistance forces an increase in pancreatic insulin secretion. Epidemiologic and experimental evidence suggests that chronic hyperinsulinemia increases the risk of cancers of the colon and endometrium, and probably other tumors (e.g., those of the pancreas and kidney).

Circulating levels of estrogens are strongly related to adiposity. For cancers of the breast (in postmenopausal women) and endometrium, the effects of overweight and obesity on cancer risk are largely mediated by increased estrogen levels. For patients with breast cancer, adiposity has been associated with poorer survival and increased likelihood of recurrence, an effect that persisted after adjustment for tumor stage and grade, hormone receptor status, and adjuvant therapy.

Self-Sufficiency in Growth Signals

Cells within normal tissues are largely instructed to grow by neighboring cells (paracrine signals) or via systemic (endocrine) signals. Similarly, cell to cell growth signaling also occurs in the vast majority of tumors. The immediate tumor cell environment (the stroma) contains resident nonmalignant cells such as parenchymal cells, epithelial cells, fibroblasts, and endothelial cells. In addition, most tumors are characterized by infiltrating immune cells such as lymphocytes, polymorphonuclear cells, mast cells, and macrophages. In some tumors, these cooperating cells may eventually transform themselves, coevolving with the tumor cells to sustain the growth of the latter. Finally, basement membranes form the extracellular matrix (ECM), which provides a scaffold for proliferation of fibroblast and endothelial cells. Together, tumor cells and stroma produce factors (autocrine and paracrine factors) that in cell-bound, matrix-bound, or soluble form directly or indirectly influence tumor development. Autocrine factors secreted by tumor cells promote growth of tumor cells but may also stimulate neighboring cells. In addition, tumor cells secrete paracrine factors that act on host cells or ECMs, generating a supportive microenvironment. For example, transforming growth factor-β (TGF)-β may induce angiogenesis, production of ECM molecules, and production of other cytokines by fibroblasts and endothelial cells. To state it simply, tumor growth is dependent on the response of tumor cells to paracrine and autocrine factors (Fig. 30-4, such as angiogenesis factors, growth factors, chemokines (polypeptide signaling molecules originally characterized by their ability to induce chemotaxis), cytokines, hormones, enzymes, and cytolytic factors, which may promote or reduce tumor growth (Table 30-3).

FIGURE 30-4 Paracrine and autocrine growth mechanisms. Both stromal cells and infiltrate secrete paracrine factors that affect tumor development. Additionally, tumor cells secrete autocrine as well as paracrine factors that, in turn, act on stromal cells and infiltrating cells.

Table 30-3 Cells and Soluble Factors Affecting Tumor Development*

* The list of cells and soluble factors is not meant to be complete but to illustrate the complexity of factors affecting tumor development.

During the evolution of a tumor, its responsiveness to growth signals changes. Paracrine growth mechanisms are dominant during the early development of tumor. Tumors become resistant to paracrine growth inhibitors and gain responsiveness to paracrine growth promoters. However, autocrine growth mechanisms become more prominent as tumors develop further. The observation that in late-stage tumors, metastatic tumor cells tend to spread more randomly throughout the body suggests that autocrine growth mechanisms may be more dominant than paracrine growth mechanisms. Advanced breast cancers, for example, lose hormone responsiveness. It is even possible for a tumor to grow completely autonomously (acrine state) and be independent of growth factors and inhibitors (Fig. 30-5).

FIGURE 30-5 Changes in contribution of growth mechanisms to tumor development. During tumor progression, the contribution of paracrine growth mechanisms decreases and the tumor becomes more dependent on autocrine growth mechanisms. At later stages, the tumor may even become independent of growth mechanisms (acrine state).

To achieve growth self-sufficiency, growth signaling pathways are altered. This involves the alteration of extracellular growth signals, transmembrane transducers of those signals, or intracellular signaling pathways that translate those signals into action. Growth factor receptors are overexpressed in many cancers. Receptor overexpression may enable the cancer cell to respond to low levels of growth factor that normally would not trigger proliferation. For example, the epidermal growth factor receptor (EGFR) and the Her2/neu receptor are overexpressed in breast and other epithelial cancers. Additionally, gross overexpression of growth factor receptors can elicit growth factor–independent signaling. The latter can also be achieved through structural alteration of receptors, such as truncated versions of the EGFR that lack much of its cytoplasmic domain and are constitutively activated.

Cancer cells can also modulate their stromal environment, including the ECM, through secretion of factors such as basic fibroblast growth factor (bFGF), platelet-derived growth factor, and TGF-β. ECM components, such as collagens, fibronectins, laminins, and vitronectins, may bind to two or more receptors and may also bind other ECM molecules. The matrix molecule-receptor interaction induces signals that influence cell behavior, including entrance into the active cell cycle. Cancer cells can switch the types of ECM receptors (e.g., integrins, heparan sulfate proteoglycans) that they express, favoring ones that transmit progrowth signals.

The third and most complex mechanism for acquisition of self-sufficiency in growth signals stem from changes in intracellular signaling pathways. Many of the oncogenes, such as activating mutations in KRAS, mimic normal growth signaling and induce mitogenic signals without stimulation from upstream regulators.

Insensitivity to Antigrowth Signals

Cell division is an ordered, tightly regulated process involving stimulatory and inhibitory signals. Thus, in addition to acquiring stimulatory growth signals, tumor cells need to overcome or neutralize growth inhibitory signals. These signals include soluble growth inhibitors and immobilized inhibitors embedded in the ECM and on the surfaces of neighboring cells. Similar to many of the stimulatory signals, the growth inhibitory signals are transduced by transmembrane receptors coupled to intracellular signaling pathways that target genes regulating the cell cycle. The cell cycle can be divided into an interphase and a mitotic (M) phase (Fig. 30-6).¹¹ The interphase is further subdivided into two gap phases (G₁ and G₂), separated by a phase of DNA synthesis (S phase). The two gap phases involve crucial regulatory events that prepare the cell for DNA replication and mitosis.

FIGURE 30-6 Schematic overview of the cell cycle. Cell division is governed by cyclin proteins and CDKs. After mitosis, a cell can terminally differentiate, enter a quiescent state, or reenter the cell cycle. A critical point in the cell cycle control is the transition from G₁ to S. After passing this checkpoint, the cell is committed to division. Tumor suppressor genes such as the retinoblastoma (Rb) and p53 gene block G₁ to S transition, whereas oncogenes such as cyclin D1 and E2F promote transition.

Central to cell cycle progression are the cyclin-dependent kinases (CDKs) that bind to the cyclin proteins. These proteins are regulated by numerous other proteins, including tumor suppressors and oncogenes that induce stimulatory or inhibitory signals. Antigrowth signals can block cell division by two distinct mechanisms. Cells may be forced to exit the cell cycle into a quiescent (G₀) state (Fig. 30-6).

Alternatively, cells may be induced to enter a postmitotic state, usually associated with terminal differentiation. Much of the signaling pathways that enable normal cells to respond to antigrowth signals are associated with the cell cycle block, specifically with the components governing the restriction point in the G₁ phase of the cell cycle. The restriction point marks the point between early and late G₁ phase passage that represents an irreversible commitment to undergo one cell division. Cells monitor their external environment during this period and, on the basis of sensed signals, decide whether to proliferate, be quiescent, or enter into a postmitotic state. At the molecular level, many and perhaps all antiproliferative signals involve the retinoblastoma protein (pRb) and its two family members, p107 and p130.¹¹ pRb is a key negative regulator at the restriction point. In quiescent cells, pRb is hypophosphorylated and blocks cell division by binding E2F transcription factors that control the expression of many genes essential for progression from the G₁ into the S phase (see Fig. 30-6). In contrast, growth stimulatory signals induce phosphorylation of pRb that does not bind E2F factors and is considered functionally inactive. Similarly, disruption of the pRb pathway liberates E2Fs and thus allows cell proliferation, rendering cells insensitive to antigrowth factors that normally operate along this pathway to block advance through the G₁ phase of the cell cycle. For example, TGF-β prevents the phosphorylation of pRb that inactivates pRb and thereby blocks advance through G₁. In some tumors, such as breast, colon, liver, and pancreatic cancers, TGF-β responsiveness is lost through downregulation of TGF-β receptors or through the expression of mutant, dysfunctional receptors. In others, such as colon, lung, and liver cancers, the cytoplasmic Smad4 protein, which transduces signals from ligand-activated TGF-β receptors to downstream targets, may be eliminated through mutation of its encoding gene. Alternatively, in cervical carcinomas induced by human papillomavirus (HPV), the viral oncoprotein E7 binds pRb and thereby induces dissociation of E2F and subsequent transcription of genes necessary for cell cycle progression. In addition, cancer cells can also turn off the expression of integrins and other cell adhesion molecules that send antigrowth signals. In summary, the antigrowth signaling pathways converging onto Rb and the cell cycle are disrupted in most human cancers.

Cyclin-CDK complexes, essential for cell cycle progression, are regulated by two families of cyclin-CDK inhibitors in normal cells. However, in tumor cells, these regulatory proteins, such as the p16 member of the INK4 family, are frequently deleted, allowing tumor cells to bypass cell cycle arrest.

In addition to avoiding antigrowth signals, tumor cells may also avoid terminal differentiation, such as through overexpression of the oncogene c-myc, which encodes a transcription factor regulating the expression of cyclins and CDKs, or through upregulation of ID (inhibitor of DNA-binding and differentiation) family members. Similarly, during human colon carcinogenesis, inactivation of the APC/β-catenin pathway serves to block the egress of enterocytes in the colonic crypts into a differentiated postmitotic state.

Evasion of Cell Death

The growth of tumors is determined by the ability of tumor cells to proliferate, offset by cell death. Most, if not all, types of tumors are characterized by defects in cell death–signaling pathways and are resistant to cell death. Cell death in tumors is caused primarily by programmed cell death, or apoptosis, which is the most common and well-defined form of cell death.¹² Apoptosis is a physiologic cell suicide program essential for embryonic development, functioning of the immune system, and maintenance of tissue homeostasis. Apoptosis is characterized by disruption of membranes and chromosomal degradation in a matter of hours. The general apoptosis signaling pathway involves the release of cytochrome c from mitochondria that activates various caspases (a family of at least 10 proteases) in sequence (Fig. 30-7).

FIGURE 30-7 Apoptotic pathways. Extracellular and intracellular stresses can induce apoptosis in tumor cells. Extracellular triggering can occur through a receptor-dependent (1) or receptor-independent (2) pathway. Both pathways induce the release of cytochrome c from mitochondria, which triggers the activation of various caspases in sequence, ultimately leading to apoptosis.

Activation of caspase cascades leads to DNA fragmentation and apoptosis. Induction of apoptosis is death receptor–dependent (extrinsic pathway) or death receptor–independent (intrinsic pathway). The two best understood death receptor pathways include the Fas receptor and death receptor (DR)-5 that bind the extracellular Fas ligand and TRAIL, respectively. Binding of the ligands triggers the activation of caspase 8 and promotes the cascade of procaspase activation, leading to the release of cytochrome c from mitochondria and eventually apoptosis. The intrinsic pathway is triggered by various extracellular and intracellular stresses, such as growth factor withdrawal, hypoxia, DNA damage, and oncogene induction. Receptor-independent pathways involve translocation of proapoptotic molecules from the cytoplasm to the mitochondria, causing mitochondrial damage and release of cytochrome c. Cytochrome c is directly involved in the activation of caspase 9, which activates caspase 3, which then leads to apoptosis.

The concept that apoptosis forms a constraint to cancer was first put forth in 1972, when massive apoptosis was observed in cells populating rapidly growing, hormone-dependent tumors following hormone withdrawal.¹³ The discovery of the bcl-2 oncogene as having antiapoptotic activity led to the investigation of apoptosis in cancer at the molecular level.¹⁴ Bcl-2 promotes the formation of B cell lymphomas through a chromosomal translocation linking the bcl-2 gene to an immunoglobulin locus, which results in the constitutive activation of bcl2, driving lymphocyte survival. Further research has demonstrated that altering components of the apoptotic machinery allows a cell to resist death signals, providing it with a selective growth advantage. For example, functional inactivation of the tumor suppressor p53 is observed in more than 50% of human cancers. p53 is a key regulator of apoptosis by sensing DNA damage that cannot be repaired and subsequent activation of the apoptotic pathway. Other abnormalities, such as hypoxia and oncogene overexpression, are also channeled, in part via p53 to the apoptotic machinery, and fail to elicit apoptosis when p53 function is lost. Also, alterations in cell survival pathways can suppress or alter apoptosis. For example, the PI3 kinase-AKT pathway, which transmits antiapoptotic survival signals, is likely involved in inhibiting apoptosis in many human tumors. This signaling pathway can be activated by extracellular factors such as insulin-like growth factor 1 (IGF-1), IGF-2, or interleukin-3 (IL-3), intracellular signals from Ras, or loss of the pTEN tumor suppressor that negatively regulates the PI3 kinase–AKT pathway. A final example is the discovery of a nonsignaling decoy receptor for Fas ligand in a high fraction of lung and colon carcinoma cell lines. Expression of this decoy receptor dilutes the death signal mediated through Fas.

Nonapoptotic types of cell death include necrosis, autophagy, and mitotic catastrophe. Necrosis is normally induced by pathophysiologic conditions such as infection, inflammation, or ischemia. Necrosis is characterized by unregulated cell destruction. Autophagy, on the other hand, is characterized by the proteolysis of long-lived proteins and organelle components in lysosomes.¹⁵ Cells that undergo excessive autophagy undergo apoptosis. Autophagy is triggered by growth factor withdrawal, hypoxia, DNA damage, and differentiation and developmental triggers.¹² Finally, aberrant mitosis caused by failure of the G₂ checkpoint to block mitosis when DNA is damaged, can lead to cell death, known as mitotic catastrophe. The signaling pathways involved in these types of nonapoptotic cell death are less well defined compared with those that regulate apoptosis, but it is clear that defects in nonapoptotic cell death pathways have been linked to cancer. For example, amplification of the MDM2 oncogene, which negatively regulates expression of p53, results in inadequate expression of p53 and thereby loss of tumor suppressor function. Another example is the deletion of the autophagy-regulating gene becklin-1 in high percentages of ovarian, breast, and prostate cancers. In addition to cell death, cells can undergo permanent growth arrest, known as senescence, when repair of damaged DNA fails. Senescent cells lose their clonogenicity, but defects in the senescent program contribute to tumor development.

Limitless Replication Potential

Acquired disruption of cell to cell signaling by itself does not ensure expansive tumor growth on its own. This is caused by the intrinsic programmed decline in replication potential that limits the multiplication of normal somatic cells. This program must be disrupted in order for a clone of cells to develop into a macroscopic tumor. Normal cells have a finite replicative potential. Once a cell population has progressed through a certain number of doublings, they stop growing, a process termed senescence.

With the exception of stem cells, activated lymphocytes, and germline cells, normal cells have a limited replicative potential. Stem cells give rise to progenitor cells that can progress through a certain number of doublings, with an increasing degree of differentiation. Fully differentiated cells do not have replicative potential. The number of doublings is controlled by telomeres, the ends of chromosomes that are composed of several thousand repeats of a short 6-bp sequence element.¹⁶ Telomeres prevent end to end chromosomal fusion. However, each DNA replication is associated with a loss of 50 to 100 base pairs of telomeric DNA from the ends of every chromosome. The progressive shortening of telomeres through successive cycles of replication eventually causes them to lose their ability to protect the ends of chromosomal DNA. When the critical length is bridged, the unprotected chromosomal ends participate in end to end chromosomal fusions, yielding a karyotype disarray that almost inevitably results in the death of the affected cell. Telomeric attrition is negated by the enzyme telomerase, which elongates telomeric DNA. Telomerase activity is high during embryonic development and in certain cell populations, such as adult stem cells. However, many tumors are characterized by elevated telomerase activity. Alternatively, telomeres are maintained through recombination-based interchromosomal exchanges of sequence information. Thus, by maintaining a telomere length above a critical threshold, the tumor cells have unlimited proliferative potential and are considered immortal.

Recently, evidence has been obtained for the existence of cancer stem cells or cancer-initiating cells that give rise to tissue-specific progenitor cells and phenotypically diverse cancer cells with limited replicative potential.¹⁷ Unlike mature, terminally differentiated tissue cells, cells with the capacity for self-renewal would live long enough for the stepwise accumulation of genetic mutations over time. Small populations of putative cancer stem cells have now been identified in many of the common cancers based on their ability to replicate, whereas most cancer cells have no or limited ability to proliferate. The most likely origin of cancer stem cells is normal adult stem cells that replace short-lived mature cells in tissues such as skin, gut, and blood. When normal stem cells divide, one of the daughter cells inherits stem cell capabilities, whereas the other cell is launched along the differentiation pathway. In cancer stem cells, the genes regulating self-renewal, such as Bmi-1, are overexpressed, thereby suppressing the default pathway of differentiation.

Sustained Angiogenesis

Based on the observation that many individuals who died of non–cancer-related causes had in situ tumors at the time of autopsy, physicians and scientists have concluded these microscopic tumors are in a dormant state. Tumor dormancy occurs because the body blocks the tumor from recruiting its own blood supply to provide tumor cells with the required oxygen and nutrients. The growth of new blood vessels, angiogenesis, is a highly regulated process to ensure supply to all cells in an organ. Surprisingly, the microscopic tumors lack the ability to induce angiogenesis, and only an estimated 1 in 600 acquire angiogenic activity. Research pioneered by Judah Folkman has demonstrated that naturally occurring endogenous angiogenesis inhibitors prevent tumors from expanding.¹⁸ The angiogenesis inhibitors keep the tumors in check by counterbalancing the angiogenic signals. These signals are mediated by soluble factors and their receptors on endothelial cells, as well as integrins and adhesion molecules mediating cell-matrix and cell-cell interactions. Angiogenic activity is induced by growth factors such as vascular endothelial growth factor (VEGF), basic and acidic fibroblast growth factor (FGF), and platelet-derived growth factor. Each binds to transmembrane tyrosine kinase receptors displayed primarily by endothelial cells connected to intracellular signaling pathways. Angiogenesis inhibitors are associated with specific tissues or circulate in the blood. The first inhibitor, interferon-α (IFN-α), was reported in 1980, and an additional 26 endogenous inhibitors have been identified since then. These include thrombospondin, tumstatin, canstatin, endostatin, and angiostatin.

Evidence for the importance of inducing and sustaining angiogenesis in tumors is overwhelming. For example, the switch of dormant human tumors into fast-growing tumors in immunocompromised mice is associated with an angiogenesis gene signature. Most telling are the results of clinical studies with the anti-VEGF antibody, bevacizumab (Avastin), the first angiogenesis inhibitor approved by the U.S. Food and Drug Administration (FDA) for the treatment of colon cancer. Bevacizumab significantly prolongs the survival of patients with advanced cancer. Similarly, a dominant-interfering version of the VEGF receptor 2 has been shown to impair neovascularization and growth of subcutaneous tumors in mice.

The ability to induce and sustain angiogenesis seems to be acquired in a discrete step (or steps) during tumor development via a switch to the angiogenic phenotype.^18,19 Tumors appear to activate the angiogenic switch by changing the balance between total angiogenic stimulation and total angiogenic inhibition.²⁰ This usually occurs when the angiogenesis stimulators overwhelm the angiogenesis inhibitors. In some tumors, these changes may be linked. It is likely that such disruption in the angiogenic balance is under control of the genetic makeup of the individual tumor cell and its microenvironment. Angiogenesis inducers and inhibitors may be genetically controlled by tumor suppressor genes such as p53, whereas oncogenes such as ras may downregulate transcription of endogenous inhibitors or activate inducers. For example, bcl2 activation leads to significantly increased expression of VEGF and angiogenesis. Another dimension of regulation is through proteases, which can control the bioavailability of angiogenic activators and inhibitors. Thus, a variety of proteases can release bFGF stored in the ECM, whereas plasmin, a proangiogenic component of the clotting system, can cleave itself into an angiogenesis inhibitor form called angiostatin. Another angiogenesis inhibitor, endostatin, is an internal fragment of the basement membrane collagen XVIII. Finally, hypoxia and other metabolic stressors, mechanical stress from proliferating cells, or inflammatory immune responses can trigger angiogenesis. The coordinated expression of proangiogenic and antiangiogenic signaling molecules and their modulation by proteolysis appear to reflect the complex homeostatic regulation of normal tissue angiogenesis and vascular integrity. Different types of tumors use distinct molecular strategies to activate the angiogenic switch.

Key in the formation of new blood vessels are endothelial cells via the production or expression of angiogenesis-promoting factors. These factors include proinflammatory cytokines such as IL-6, VEGF, and hematopoietic growth factors such as colony-stimulating factors that recruit and activate bone marrow–derived progenitor cells. Among the progenitor cells are myeloid precursors that further promote the proinflammatory responses at the tumor and actively contribute to angiogenesis by producing matrix metalloprotease-9 (MMP-9), a critical regulator of tumor angiogenesis through the induced release of VEGF. Bone marrow–derived endothelial precursors foster tumor blood vessel assembly.

Tissue Invasion and Metastasis

Progressing tumors give rise to distant metastases that are the cause of 90% of human cancer deaths. The formation of tumor metastases is characterized by the detachment of some tumor cells from the primary tumor and infiltration into the bloodstream or lymphatics (intravasation). The reciprocal process occurs at other locations in the body (extravasation). Both intravasation and extravasation are characterized by changes in ECMs and their interactions with tumor cells. The cell-cell and cell-matrix interactions are mediated through cell adhesion molecules (CAMs), primarily by members of the immunoglobulin and calcium-dependent cadherin families,²¹ hyaluronan receptor CD44, selectins, and integrins,²² which link cells to ECM substrates. Studies have shown that the molecules mediating adhesion are also capable of signal transduction. As such, changes in expression of adhesion molecules will alter signaling pathways and, conversely, signaling molecules can directly affect the function of adhesion molecules in tumor cells.

Epithelial (E)-cadherin is the prototype cadherin responsible for cell polarity and organization of epithelium. E-cadherin function is lost in most epithelial tumors during progression to tumor malignancy, and may in fact be a prerequisite for tumor cell invasion and metastasis formation. In normal cells, extracellular domains of E-cadherin on opposing cells couple and form cell-cell junctions (Fig. 30-8). The cytoplasmic cell-adhesion complex is linked to the actin cytoskeleton through catenins (α, β, γ). Mechanisms that include mutational inactivation of the E-cadherin or β-catenin genes, transcriptional repression, or proteases of the extracellular cadherin domain induce loss of E-cadherin function.²¹ This prevents catenins from binding and leads to their accumulation in the cytoplasm. Inactivation of nonsequestered β- and γ-catenin is dependent on the presence of the tumor suppressor gene APC and an inactive Wnt signaling pathway (Fig. 30-8). However, when APC function is lost, as is the case in many colon cancers or in case of Wnt activation, β-catenin is not degraded but instead translocates to the nucleus; here transcription is activated of genes involved in cell proliferation and tumor progression, such as c-myc, cyclin D1, CD44, and others.

FIGURE 30-8 Loss of E-cadherin permits tumor progression. Functional loss of E-cadherin to sequester β-catenin leads to the accumulation of β-catenin in the cytoplasm. Similarly, Wnt signaling inactivates GSK-3β, which leads to the stabilization of β-catenin instead of its degradation. Also, loss of APC function may result in the accumulation of β-catenin in the cytoplasm. This leads to the translocation of β-catenin to the nucleus, where it binds the T cell–specific transcription factor/lymphoid enhancer factor-1 (TCF/LEF-1), inducing a genetic program that leads to tumor progression. α, α-Catenin; APC, adenomatous polyposis coli; β, β-catenin; Frz, frizzled (transmembrane receptor for Wnt growth factors); DSH, disheveled; GSK-3β, glycogen synthase kinase 3β.

Changes in expression of CAMs in the immunoglobulin superfamily also appear to play critical roles in the processes of invasion and metastasis.^2,21 Neuronal (N)-CAM, for example, undergoes a switch in expression from a highly adhesive isoform to poorly adhesive (or even repulsive) forms in Wilms’ tumor, neuroblastoma, and small cell lung cancer. In invasive pancreatic cancer and colorectal cancers, the overall expression of N-CAM is reduced.

Selectins are a family of transmembrane molecules consisting of E-, L- (leukocyte), and P (platelet)-selectin, that normally mediate blood cell–endothelial cell interactions. However, alterations in the expression level of selectins and/or their ligands, such as the E- and L-selectin ligand, CD44, have been associated with increased invasiveness and poor survival in several malignancies (e.g., breast cancer, colorectal cancer).

Changes in integrin expression are also evident in invasive and metastatic cells. For invading and metastasizing cells to be successful, they need to adapt to changing tissue microenvironments. This is accomplished through shifts in the spectrum of integrin α and β subunits displayed on the cell surface by the migrating cells. The large extracellular domain of integrins can bind to ECM molecules such as collagens, laminin, and fibronectin, ligands associated with vascular and coagulation physiology, such as thrombospondin and factor X, or with other CAMs. Each integrin molecule consists of an α subunit and β subunit, but a particular β subunit can dimerize with several different α subunits. These novel permutations result in different integrin subtypes—24 combinations have now been described—having distinct substrate preferences. Also, integrins may exhibit different specificities when expressed on different cell types. Thus carcinoma cells facilitate invasion by shifting their expression of integrins from those that favor ECM present in normal epithelium to other integrins that preferentially bind the degraded stromal components produced by extracellular proteases.² For example, expression of α₄β₁, which binds fibronectin, correlates with the progression of melanoma. The changes are incompletely understood because of the large number of distinct integrin genes, by the even larger number of heterodimeric receptors resulting from combinatorial expression of various α and β receptor subunits, and by increasing evidence of complex signals emitted by the cytoplasmic domains of these receptors. Changes in integrin expression may also be essential for the expansion of the tumor stem cell compartment by inhibiting differentiation or apoptosis.²³

The second general parameter of the invasive and metastatic capability involves extracellular proteases that regulate ECM turnover. It has become clear that tumor progression may involve an increased expression of proteases, decreased expression of protease inhibitors, and inactive zymogen forms of proteases that are converted into active enzymes. Expression of the protease tenascin, which neutralizes adhesion to fibronectin, is increased 10-fold in invasive breast carcinoma compared with normal breast tissue. MMPs are overexpressed in melanoma, invasive breast carcinoma, and invasive squamous cell carcinoma. Matrix-degrading proteases are characteristically associated with the cell surface by synthesis with a transmembrane domain, binding to specific protease receptors, or association with integrins. It is possible that docking of active proteases on the cell surface can facilitate invasion by cancer cells into nearby stroma, across blood vessel walls, and through normal epithelium cell layers. Nevertheless, it is difficult to ascribe unambiguously the functions of particular proteases solely to this capability, given their evident roles in other hallmark capabilities, including angiogenesis and growth signaling, which in turn contribute directly or indirectly to their invasive or metastatic capability. A further complexity derives from the many cell types involved in protease expression and display, including stromal and inflammatory cells.

The activation of extracellular proteases and altered binding specificities of cadherins, CAMs, selectins, and integrins are clearly central to the acquisition of invasiveness and metastatic potential. The clonal and genetic diversity of tumors permits adhesion and detachment from the same matrix. Some tumor cells in a primary tumor may have the correct genotype and phenotype to permit detachment from the surrounding tissue and entry into blood vessels or lymphatic vessels. Similarly, extravasation may be mediated by a few tumor cells that express the required receptors for certain ECM molecules. In general, those mutations that confer escape from homeostatic control mechanisms in the host or that give the tumor cell a growth advantage over others are favorably selected. Thus, tumor clones that best complement the environment with the expression of particular ECM receptors may thrive because this provides an advantage over other clones. However, the regulatory pathways and molecular mechanisms that govern these changes are incompletely understood, and appear to differ from one tissue environment to another.

Outgrowth at Preferred Sites

Invasion and metastatic spread of tumor cells do not appear to be random processes. In 1889, Paget observed that breast carcinoma often metastasized to the liver, lungs, bone, adrenals, or brain. He hypothesized that tumor cells (the seed) would grow only in selective environments (the soil), in which conditions supported tumor growth—hence, the so-called seed-and-soil hypothesis. Since then, additional studies have confirmed this. For example, malignant melanoma metastasizes to the brain, but ocular malignant melanoma frequently metastasizes to the liver. Prostate cancer metastasizes to the bone and colon carcinoma to the liver.

Although metastatic spread is in part determined by circulation patterns, the retention of disseminated tumor cells in distant organs suggests the existence of specific molecular interactions. Molecular analysis has provided several theories to explain the preferential outgrowth of tumor cells. One theory, the growth factor theory, proposes that tumor cells in the blood or lymphatics invade organs at a similar frequency, but only those that find favorable growth factors multiply. Transferrins, for example, are iron-transferring ferroproteins required for cell growth that have additional mitogenic properties beyond their iron-transporting function. Increased concentrations of transferrin are found in lung, bone, and the brain and are associated with elevated levels of transferrin receptors on metastasizing tumor cells.

Another theory, the adhesion theory, proposes that endothelial cells lining the blood vessels in certain organs express adhesion molecules that bind tumor cells and permit extravasation. A third theory is that chemokines secreted by the target organ can enter the circulation and selectively attract tumor cells that express receptors for the chemokines. Evidence for the importance of chemokines in tumor progression has been obtained for breast cancer cells preferentially metastasizing in bone marrow, liver, lymph nodes, and lung.²⁴ These organs were found to secrete CXCL12, which is the ligand for the chemokine receptor, CXCR4, enriched on breast cancer cells compared with normal breast epithelial cells. A similar phenomenon was observed for melanoma cells that were found to express elevated levels of the receptors CXCR4, CCR7, and CCR10 compared with normal melanocytes. Lymph nodes, lung, liver, bone marrow, and skin express the highest levels of the ligands for these receptors and are the preferred sites for metastatic spread of melanomas. Because chemokines are now known to affect angiogenesis and expression of cytokines, adhesion molecules, and proteases, in addition to inducing migration, it appears that chemokines and their receptors play an essential role in the successful outgrowth of tumors at preferential sites.

Detailed analyses of primary tumors have indicated that gene functions mediating metastatic activities are present early in the disease. These functions result from genetic or epigenetic alterations. The genes can be grouped into classes such as metastasis-initiating genes, which control invasion, angiogenesis, circulation, and bone marrow mobilization. Similarly, metastasis progression genes control extravasation, survival, and reinitiation, whereas metastasis virulence genes regulate organ-specific colonization. These intrinsic properties of the tumor, together with its cellular origin, determine the organ specificity and temporal course of metastasis formation.

Central to the mechanisms dictating metastatic predisposition are bone marrow–derived progenitor cells expressing the VEGF receptor 1 (VEGFR1) and VLA-4, which are prompted by the primary tumor to establish premetastatic niches before arrival of metastatic tumor cells.²⁵ Tumor-secreted humoral factors induce fibronectin (a VLA-4 ligand) expression on fibroblasts and fibroblast-like cells in specific distant organs. Simultaneously, the VEGFR1⁺, VLA-4⁺ cells leave the bone marrow and migrate to the premetastatic niche, where they form cellular clusters that permit the development of metastases.

Immunosurveillance and Immunoediting