Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

Section V - Surgical Oncology

759

Chapter 28 - Tumor Biology and Tumor Markers

Peter S. Goedegebuure Ph.D. Udaya Liyanage M.D. Timothy J. Eberlein M.D.

TUMOR BIOLOGY Tumors are characterized by an uncontrolled growth of transformed cells. Although tumors, also called neoplasms, do not arise spontaneously, once triggered, the proliferation of the tumor is independent of the presence of the stimulus. Tumors can be benign or malignant; the malignant ones are often referred to as cancer and can be further distinguished based on origin. All tumors can be either classified as hematopoietic or nonhematopoietic in origin. Within each category, tumors can be further divided into those of similar histologic origin, for example, tumors derived from the same organ, such as lung cancers and colon cancers, or tumors from the same cell type, such as epithelial cellderived tumors. The development of a tumor from a single transformed cell into metastatic tumor is categorized by stages and grades and involves three phases: initiation, promotion, and progression. The initiation phase is characterized by the series of genetic mutations that occur in sequence. For initiated cells to become tumor cells, exposure to promoting agents or conditions is required (promotion phase). The end of the promotion phase is characterized by the appearance of the first neoplastic cells. Before the appearance of neoplastic cells, the abnormal cells are called preneoplastic or premalignant cells. The progression phase is characterized by invasive growth of the transformed cells and progression of the tumorous lesion into a highly metastatic tumor that may ultimately kill the host. A well-documented example of tumor development is presented in Table 281 .[1] [2] The transformation of melanocytes into malignant melanoma can be divided into five major histopathologically and clinically identifiable steps (see Table 281 ). This chapter discusses the initiation and development of tumors as well as the expression of genotypic and phenotypic markers associated with these processes.
How Do Tumors Arise?

The formation, or genesis, of a tumor is a complex process primarily because tumors are not a single disease. The progression of a single transforming event in a cell toward the formation of a tumor requires multiple mutations that occur in sequence. The whole sequence occurs over a period that may exceed 20 years. Additionally, cancer is multifactorial: the cancer-inducing mutations may be induced by external factors (somatic mutations), such as exposure to physical or chemical carcinogens, or they may be germline mutations that are more likely to result in certain types of cancer. Exposure to physical or chemical carcinogens is responsible for most cancers in industrialized countries and includes smoking, ingesting alcohol or particular foods, and exposure to sunlight and chemicals. Epidemiologic studies have shown strong correlations between several internal and external factors, although the exact sequence of events is still unknown. Smoking is associated with lung cancer, bladder cancer, and cancer of the mouth, pharynx, larynx, and esophagus. Immunosuppression is associated with lymphomas, and x-rays and gamma rays are associated with the development of leukemia. Chemical carcinogens include asbestos and polycyclic aromatic hydrocarbons, which increase the risk for lung cancer; benzene, which may induce leukemia; and aromatic amines, which may induce bladder cancer.
760

Figure 28-1 A genetic model for colorectal tumorigenesis. Tumorigenesis proceeds through a series of genetic alterations involving oncogenes (ras) and tumor suppressor genes (particularly those on chromosomes 5q, 12p, 17p, and 18q). The three stages of adenomas in general represent tumors of increasing size, dysplasia, and villous content. In patients with familial adenomatous polyposis (FAP), a mutation on chromosome 5q (APC gene) is inherited. This alteration may be responsible for the hypoproliferative epithelium present in these patients. Hypomethylation is present in very small adenomas in patients with or without polyposis, and this alteration may lead to aneuploidy, resulting in the loss of suppressor gene alleles. The ras gene mutation appears to occur in one cell of a preexisting small adenoma and, through clonal expansion, produces a larger and more dysplastic tumor. Allelic deletions of chromosome 17p and 18q usually occur at a later stage of tumorigenesis than do deletions of chromosome 5q or ras gene mutations. The order of these changes is not invariant, however, and accumulation of these changes, rather than their order with respect to one another, seems most important. Tumors continue to progress once carcinomas have formed, and the accumulated loss of suppressor genes on additional chromosomes correlates with the ability of the carcinomas to metastasize and cause death. (From Fearon ER: A genetic model for colorectal tumorigenesis. Cell 61:759, 1990.)

TABLE 28-1 -- Stepwise Progression From Melanocyte to Metastatic Melanoma

Step * 1 2 3 4 5 Common melanocytic nevus Dysplastic nevus Radial growth phase of melanoma Vertical growth phase of melanoma Metastatic melanoma

Characteristics

Adapted from Clark WH: A study of tumor progression: The precursor lesions of superficial spreading and nodular melanoma. Hum Pathol 15:1147, 1984; and Clark WH: The biologic forms of malignant melanoma. Hum Pathol 17:443, 1986.
* Common acquired and congenital nevi without cytologic atypia (step 1) may progress into dysplastic nevi with clear atypical histologic and cytologic features (step 2). Most of these lesions are stable, but a few may progress to a malignant melanoma that tends to grow outward along the radius of the plaque (step 3). Within the plaque, a nodule develops of fastgrowing cells that expand in a vertical direction, invading the dermis and elevating the epidermis (step 4). Finally, the tumor metastasizes (step 5).

An example of a germline mutation leading to (colon) carcinoma was proposed by Fearon and Vogelstein.[3] They postulated that the morphologic changes from normal mucosa to early, intermediate, and late adenoma and finally to colorectal carcinoma are associated with certain essential mutations that occur in sequence ( Fig. 281 ). Patients with germline mutations in the tumor suppressor gene adenomatous polyposis coli (APC gene) are prone to develop familial adenomatous polyposis. Tumor suppressor genes can be regarded as negative regulators of growth-promoting signaling pathways. The APC gene product is involved in cellular adhesion and intercellular communication, and both functions are absent after gene mutations. Mutations in the APC gene can frequently be found in adenomas, including small adenomas, suggesting that they are an early event. The second important mutation may occur in the K-ras oncogene, a cell membranebound signal-transduction molecule. Oncogenes are positive regulators of growth-promoting signaling pathways. Mutations in K-ras resulting in continuous activation of ras occur in about half of colorectal cancers and adenomas larger than 1 cm but are found less frequently in small adenomas. Thus, K-ras mutations are early events but occur later than APC gene mutations. Late adenomas and colorectal cancer frequently show loss of the tumor suppressor gene DCC (deleted colon cancer). DCC deletion is, however, uncommon in early adenomas, placing this event after the K-ras mutations. The DCC gene product is a cell surface molecule involved in adhesion. Finally, deletion of the tumor suppressor gene p53 is observed in 75% of colorectal cancers but is infrequently observed in adenomas and is therefore a late event in carcinogenesis. The mutations and deletions of the APC, K-ras, DCC, and p53 genes are the most commonly detected genetic alterations in colon cancer. In general, mutated genes in transformed cells include tumor suppressor genes and oncogenes. Both types of genes play an essential role in the regulation of cell cycle progression. Cell proliferation is a series of tightly controlled biochemical processes divided into an interphase and a mitotic (M) phase ( Fig. 282 ). The interphase is further subdivided into two gap phases (G1 and G2 ), separated by a phase of DNA synthesis (S phase). Central to cell cycle progression are the cyclin-dependent kinases 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. Both tumor suppressors and oncogenes can be part of the same signaling pathway, and some of these pathways are disrupted in virtually every tumor.[4]

761

Stimulation of Tumor Development

Even though tumors are characterized by unlimited growth, tumors are dependent on their environment for growth and development. The immediate tumor environment (the stroma) contains residing nonmalignant cells such as parenchymal cells, epithelial cells, fibroblasts, endothelial cells, and mast cells. In addition, most tumors are characterized by infiltrating immune cells such as lymphocytes, polymorphonuclear cells, and macrophages. Finally, basement membranes form the extracellular matrix (ECM) that forms a scaffold for proliferation of fibroblast and endothelial cells. The ECM consists of multiple different proteins and carbohydrates that together are responsible for the organization of cells into organs. ECM components, such as collagens, fibronectins, laminins, vitronectins, and proteoglycans, interact with cellular receptors, such as integrins and heparan sulfate proteoglycans. Each ECM molecule may bind to two or

Figure 28-2 Schematic overview of the cell cycle. Cell division is governed by cyclin proteins and cyclin-dependent kinases (cdks). After mitosis, a cell can terminally differentiate, enter a quiescent state, or re-enter 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) gene and p53 block G1 to S transition, whereas oncogenes such as cyclin D1 and E2F promote transition.

more receptors and may also bind other ECM molecules. The matrix molecule-receptor interaction leads to an attachment signal, which activates certain genes. It is believed that adhesion to ECM is required for cell growth and differentiation. The adhesion is cell-type specific. The specificity is mediated primarily through integrin receptor specificity. Integrins consist of an alpha subunit and a beta subunit, but a particular beta subunit can dimerize with several different alpha subunits. In addition, alternative splicing of the primary transcripts of several subunits has been described. Additional specificity is mediated through the cytoskeletal organization within the cell and through the array of genes being expressed in cells. Together, tumor cells, ECM, stroma, and infiltrate produce factors (autocrine and paracrine factors) that, in cell-bound, matrix-bound, or soluble form,

directly or indirectly influence tumor development. Proliferating tumors invade neighboring cell populations, thereby breaking down boundaries designed to keep normal tissue architecture. These boundaries consist of basement membranes, ECM communication with neighboring cells, and cellcell communication preventing cells from inappropriately mixing. Tumor cells resistant to the regulatory signals can successfully expand into surrounding tissues. There is evidence that at the molecular level, signaling pathways regulating tumor cell motility are linked to pathways leading to proliferation and survival. For example, integrin-mediated signaling involved in disengagement from ECM (essential for motility) is associated with additional proinvasive signaling and proliferative and antiapoptotic signaling.
Paracrine and Autocrine Growth Mechanisms

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. For example, transforming growth factor (TGF)- may induce angiogenesis, production of ECM molecules, and production of other cytokines by fibroblasts and endothelial cells. Simplified, tumor growth is dependent on the response of tumor cells to paracrine and autocrine factors[5] ( Fig. 283 ). These factors include

Figure 28-3 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. ECM, extracellular matrix; NK, natural killer.

762

angiogenesis factors secreted by tumor cells, growth factors, chemokines (polypeptide signaling molecules originally characterized by their ability to induce chemotaxis), cytokines, hormones, enzymes, cytolytic factors, and so forth, which may promote or reduce tumor growth ( Table 282 ). Even though most paracrine factors promote tumor growth, growth inhibitors have also been found. TGF- may inhibit tumor growth, as was found for kidney cells that secrete TGF-1 , which inhibits formation of tumor metastases in the kidney. In melanoma, interleukin (IL)-6 produced by stromal cells (keratinocytes, endothelial cells, and fibroblasts), or infiltrate (monocytes and macrophages) may inhibit the growth of early lateral growth phase melanoma cells, but more progressed melanoma cells have lost responsiveness to IL-6. TABLE 28-2 -- Cells and Soluble Factors Affecting Tumor Development * Cells STROMA Parenchymal cells Endothelial cells Fibroblasts Mast cells Extracellular matrix Keratinocytes INFILTRATE T lymphocytes B lymphocytes NK cells NKT cells Macrophages/monocytes Dendritic cells Polymorphonuclear cells Platelets TUMOR Chemokines, cytokines, angiogenesis factors, degradative enzymes, growth (inhibitory) factors Cytokines, chemokines, cytolytic factors, angiogenesis factors, growth (inhibitory) factors, degradative enzymes, cytostatic factors, antibodies Growth factors, growth inhibitors, nutritional factors, hormones, degradative enzymes, cytokines, angiogenesis factors Soluble Factors

* 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, changes in growth and other properties occur. Paracrine growth mechanisms are dominant during tumor initiation and tumor promotion. Early stages of metastasis are characterized by a preferential outgrowth at restricted sites, suggesting that paracrine growth mechanisms are essential. However, autocrine growth mechanisms become more prominent in later stages. The observation that during tumor progression tumor cells tend to spread more randomly through the body suggests that autocrine growth mechanisms may be more dominant than paracrine growth mechanisms. Tumors become resistant to paracrine growth inhibitors and lose responsiveness to paracrine growth promoters.

Progressing breast cancers, for example, lose hormone responsiveness. It is even possible for a tumor to grow completely autonomous (acrine state) and to be independent of growth factors and inhibitors ( Fig. 284 ).
Role of Inflammation in Cancer

In the past, investigators viewed infiltration of tumors by immune cells as a positive sign, indicative of an antitumor response. More recent work, however, has demonstrated that infiltration in general has no proven prognostic value. Virchow observed in 1863 that tumors are characterized by a leukocyte infiltrate and suggested that cancers arise at sites of chronic inflammation. Recent data have confirmed that inflammation associated with chronic infections is a cofactor in carcinogenesis. Mechanistically, the cells and factors involved in tumor growth are similar to those involved in wound healing.[6] Both processes are characterized by a complex network involving activation and migration of leukocytes to the site of damage and involvement of the ECM. However, in wound healing proliferation of cells and inflammation subside when the tissue is successfully regenerated, but tumor cells maintain their proliferative capacity and as such are wounds that do not heal. The tumor infiltrate may include macrophages, dendritic cells, neutrophils, eosinophils, mast cells, and lymphocytes (see Table 282 ). [6] Macrophages generally

Figure 28-4 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).

763

comprise the majority of the leukocyte infiltrate, recruited by chemokines such as members of the macrophage inflammatory protein family secreted by tumor cells. The tumor-associated macrophages can produce proteolytic enzymes that degrade ECM, growth factors, and angiogenic factors such as vascular endothelial growth factor (VEGF) and prostaglandin E2 . Prostaglandin E2 also stimulates tumor growth and blocks IL-12 production (IL-12 is an inducer of cd4+ T helper 1 [TH ] cell differentiation, which supports a cellular immune response). Just as macrophages, dendritic cells are recruited to sites of inflammation such as tumors. Dendritic cells are essential for uptake and presentation of antigen, and stimulation of naive T lymphocytes. However, their ability to stimulate immune effector cells is often compromised in tumors and instead they may contribute to tumor development through stimulation of immune suppressor cells and suppressor cytokines. Of the tumor-infiltrating lymphocytes (TILs), the vast majority are T lymphocytes expressing the antigenspecific T-cell receptor consisting of an alpha and beta chain. Tumor-derived T cells have received lots of attention because of their potential to recognize tumor antigen and destroy tumor cells ex vivo.[7] On the other hand, however, it has been demonstrated that signaling via the T-cell receptor is partially defective in freshly isolated tumor-derived T cells. It was also demonstrated in vitro that under starvation conditions TILs secrete basic fibroblast growth factor and heparin-binding epidermal growth factor (EGF), two factors that promote angiogenesis.[8] Sustained production by tumor cells of certain cytokines, angiogenesis factors, and chemokines (e.g., IL-8, TGF-, colony-stimulating factor [CSF], and macrophage chemotactic proteins) causes macrophages and other host cells to be continuously attracted and activated to further benefit growth and survival of tumor cells. The examples listed here and in Table 282 focus on only a few factors secreted by only a few cell types that have infiltrated tumors, to illustrate the complexity of the system.
Tumor Progression

The formation of tumor metastases is characterized by 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 (reviewed in Reference 9 ). Like tumor promotion, tumor progression may also be dependent on infiltration. The difference between the growth of tumor at an earlier stage and that at later stages may be that at later stages, more tumor cells have acquired a stimulatory paracrine loop of growth factors as a result of the continuous transformations in tumor cells and the selection for survival. Overexpression of the EGF receptor in breast, lung, and bladder cancers and melanomas, for example, is often associated with poor survival and enhanced metastasis. Within the tumor, certain tumor cells may secrete chemokines that attract inflammatory cells. These inflammatory cells may in turn secrete growth factors, promoting angiogenesis, or cytokines that directly promote tumor growth. As mentioned previously, at later stages of tumor development, tumor cells may start secreting these factors themselves
Intravasation and Extravasation

In addition to adhesion to ECM, detachment from ECM is required for migration of cells. Detachment, like attachment, is an active process that requires signal transduction. Examples of ECM molecules that mediate detachment are hyaluran, which binds to hyaluran-binding cellular receptors such as CD44, and tenascin, which neutralizes adhesion to fibronectin; and certain proteoglycans. Tumor cells have acquired the ability to modify ECM-cell interactions to permit detachment from the primary tumor, intravasation into the circulation, attachment to endothelial cells throughout the body, and extravasation through the endothelial basement membrane. In addition to the attachment-mediating and detachment-mediating ECM molecules, tumor

cells modulate ECM-bound proteases that regulate ECM turnover. It is clear from the variety of processes that tumor progression may involve an increased expression of proteases, decreased expression of protease inhibitors, and enhanced expression of certain ECM-receptor interactions, whereas other matrix-receptor interactions are decreased. The latter may promote detachment of tumor cells. For example, in colorectal cancer, deficiencies in expression of collagen at the edges of the tumor may facilitate the invasiveness of this type of tumor. Expression of the protease tenascin is increased 10-fold in invasive breast carcinoma compared with normal breast tissue, and matrix metalloproteases are overexpressed in melanoma, invasive breast carcinoma, and invasive squamous cell carcinoma. In addition, aberrant expression of cellular receptors may contribute to extravasation and homing in tissues that highly express the ligand for the aberrant receptor. Because most receptors for ECM molecules are integrins, alterations in integrin expression are most frequent. Expression of 4 6 integrin is reduced in breast carcinoma, whereas expression of 4 1 , which binds fibronectin, correlates with progression of melanoma. There is evidence that activation and overexpression of oncogenes, such as N-myc, results in downregulation of integrins such as 2 1 , which binds collagen and laminin, and 3 1 , which binds fibronectin in neuroblastomas. However, there is no clear pattern for the effects of alterations in ECM molecule and receptor expression. Nonetheless, all alterations eventually result in tumor progression. In this context, it is important to realize that most cancers are clonal in origin. Both during tumor promotion and tumor progression, additional genetic mutations occur in tumor cells. These result in the formation of subpopulations of tumor cells that may or may not grow out, dependent on natural selection processes, and cause heterogeneity within a tumor. The clonal and genetic diversity of tumors permits adhesion and detachment from the same matrix. Some tumor cells within a primary tumor may have the correct genotype
764

and phenotype to permit detachment from the surrounding tissue and intravasate blood vessels or lymphatic vessels. Likewise, 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 expression of particular ECM receptors may thrive because this provides an advantage over other clones. The continuous evolution of a tumor has far-reaching consequences for antitumor therapies (discussed later).
Outgrowth at Preferred Sites

Invasion and metastatic spread of tumor cells do not appear to be random processes. Paget observed in 1889 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), where conditions supported tumor growth, hence the so-called seed-and-soil hypothesis. Since then, additional studies have confirmed this hypothesis. 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. Molecular analysis has provided three major theories to explain preferential outgrowth of tumor cells. The first theory, the growth factor theory, proposes that tumor cells in the blood or lymphatics invade organs at pretty much the same 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. The second theory, the adhesion theory, proposes that endothelial cells lining the blood vessels in certain organs express adhesion molecules that bind tumor cells and permit intravasation. The 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 was recently obtained for breast cancer cells preferentially metastasizing in bone marrow, liver, lymph nodes, and lung. These organs were found to secrete CXCL12, which is the ligand for the chemokine receptor, CXCR4, enriched on breast cancer cells compared to 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 to 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. Since 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.
Immune Surveillance

If tumors grow uncontrolled, what causes tumor cell death? In addition to physical constraints and lack of nutrients, the immune system is capable of eradication of tumor cells. In the early 1900s, it was proposed that the frequency of cancerous transformations would be very high if it were not for the defense system of the host. This concept was later substantiated in the 1950s and 1960s and the term immunosurveillance was introduced by Burnet in 1970.[10] Burnet hypothesized that the development of T-lymphocytemediated immunity during evolution was specific for elimination of transformed cells. He further proposed that there is a continuous surveillance of the body for transformed cells, hence the term immunosurveillance. During the subsequent years, experiments in immunosuppressed and immunodeficient mice demonstrated that T-cellmediated immunity provides protection against virally induced tumors. However, no conclusive evidence was obtained for immune surveillance of cancer. More recent discoveries have made it clear that the earlier studies were limited by incomplete knowledge of the models. When tested more accurately in more appropriate mouse models, evidence for immune surveillance of cancer was obtained: immunodeficient mice were significantly more susceptible to formation of chemically induced tumors and spontaneous tumors than immunocompetent mice.[11] This suggests that the unmanipulated immune system is capable of recognizing and eliminating primary tumors. Does immune surveillance of cancer exist in humans? Evaluation of long-term studies in transplant patients who were immunosuppressed and patients with immunodeficiencies showed an increased incidence of virally induced tumors such as non-Hodgkins lymphoma, Kaposis sarcoma, and carcinoma of the genitourinary and anogenital regions. However, they also showed a higher incidence of tumors with no apparent viral etiology such as malignant melanoma, lung cancer, pancreatic cancer, colon cancer, and kidney cancer. In addition, it was found that the presence of lymphocytes in such tumors is positively correlated to increased patient survival, especially in malignant melanoma. This finding should not be confused with the earlier mentioned correlation between inflammation and origin of cancer that applies mostly to tumors caused by chronic infections. The data from mouse and human studies combined suggest that immune surveillance of cancer does exist, mediated through immune cells and soluble factors. It also suggests that continuous pressure of the immune system in an immunocompetent host determines to a great degree if and how tumors evolve, a process called immunoediting.[12]
Immune Effector Cells

The immune system can be divided into two arms: natural or innate immunity and acquired or adaptive immunity.
765

TABLE 28-3 -- Immune Effector Cells Effector Cell INNATE Macrophage/monocyte/dendritic cell NK cell Neutrophil NKT cell ACQUIRED CD8 T lymphocyte CD4 T lymphocyte B lymphocyte Direct tumor cell lysis Regulate function of other immune cells, e.g., CD8 T cells (TH 1) and B cells (TH 2) Secretion of tumor-specific antibodies Phagocytosis, processing and presentation of tumor antigens to T cells Direct tumor cell lysis, ADCC Direct tumor cell lysis, ADCC Direct tumor cell lysis Primary Effector Function

ADCC, antibody-dependent cell-mediated cytotoxicity. The innate immune system often works hand in hand with the acquired immune system. Of the two arms of the immune system, acquired immunity (mediated through B and T lymphocytes) may be the most potent against tumors. A number of immune effector cells have the ability to eradicate tumor cells directly or indirectly ( Table 283 ).
Innate Immune Cells

Innate immune cells, such as macrophages, dendritic cells, natural killer (NK) cells, NKT cells, and polymorphonuclear cells, characteristically do not need sensitization to respond to an immunogen. Macrophages and neutrophils are generally not cytotoxic to tumor cells unless activated by bacterial products. Activated macrophages and dendritic cells may exert direct antitumor activity through members of the tumor necrosis factor (TNF) family such as TNF, Fas ligand, and TNF-related apoptosis inducible ligand (TRAIL), or through oxyradicals such as nitric oxide. More important is perhaps the ability of macrophages and neutrophils to phagocytose and destroy antibody-coated tumor cells. In addition, macrophages and, more so, dendritic cells can process and present tumor antigens to T cells, thereby bridging innate and adaptive immunity. In this process, cytokines (primarily IL-12) and chemokines secreted by macrophages and dendritic cells are essential, too. NK cells are a distinct population of lymphocytes that can efficiently lyse tumor cells. Inhibitory and stimulatory receptors regulate the activation of NK cells. Inhibitory receptors specifically bind certain major histocompatibility complex (MHC) class I molecules; downregulation or loss of expression of these MHC molecules removes the brake and makes tumor cells susceptible to NK-mediated lysis. Stimulatory receptors on NK cells recognize ligands encoded by genes that are selectively expressed or upregulated in tumor cells or virally infected cells. Some of these receptors such as NKG2D are also expressed in macrophages, and CD8 and T cells (see acquired immune cells), and binding of a ligand induces secretion of interferon (IFN)- (NK cells) or TNF- and nitric oxide (macrophages). Finally, NKT cells resemble NK cells and T cells in phenotype and functional characteristics, hence their name. The population of NKT cells is heterogeneous and its exact function in vivo is unknown. NKT cells recognize antigen presented by MHC or MHC-like molecules through an antigenspecific T-cell receptor but at the same time express NK-cell receptors. Cross-linking of the T-cell receptor induces rapid secretion of large amounts of cytokines such as IL-4 and IFN-. Since IL-4 and IFN- generally induce opposing effects, controversy exists about the exact role of NKT cells in tumor surveillance.
Acquired Immunity

Acquired immunity is mediated by antigen-specific effector cells that recognize tumor antigens. Tumor-specific antigens have been discovered that are either unique or shared by other tumors and can be recognized by antibodies or T cells. The role of B cells and antibodies in antitumor responses is poorly understood. Antibodies specific for tumor antigens such as Her2/neu are often present in serum of patients but generally without correlation to clinical status. Possible mechanisms of tumor cell lysis through antibodies involve complex formation between antibody and tumor cells, followed by opsonization by macrophages (phagocytosis). The presence of complement may enhance this process. Alternatively, the Fc part of the antibody may be bound by Fc-receptorpositive immune cells, such as NK cells, macrophages, and neutrophils. This process is called antibody-dependent cellular cytotoxicity. T cells, in contrast to antibodies, may directly lyse tumor cells. Spontaneous regression of solid tumors has sporadically been observed and may be mediated by tumor-specific T cells.[7] [13] T-cellmediated lysis is dependent on recognition of tumor antigens by the T-cell receptor on the T cell on which the lytic machinery of T cells is activated. The T-cell receptor is the antigen receptor expressed on every T cell. The T-cell receptor is encoded by genetically uniquely rearranged gene fragments, yielding a large repertoire of T-cell receptor molecules with a different antigen specificity.[14] The T-cell receptor molecule is designed to bind to molecules of the MHC. The
766

Figure 28-5 Schematic representation of the T-cellmediated immune response to tumor cells. cd4+ T cells recognize tumor antigen presented by MHC class II molecules on antigenpresenting cells (APC). The activation of cd4+ T cells is dependent on costimulation through CD28. Activated cd4+ T cells secrete cytokines that regulate, among other things, activation of cd8+ T cells. Resting cd8+ T cells recognize MHC class I molecules on APC-presenting tumor antigen. When costimulated through CD28, CD8 T cells become activated and are capable of recognizing tumor cells. Lysed tumor cells are processed by antigen-presenting cell (APC).

T-cell receptor on cd8+ T cells binds to MHC class I molecules that are expressed on all nucleated cells, whereas the T-cell receptor on cd4+ T cells generally binds to MHC class II molecules, which are expressed predominantly on cells of the immune system ( Fig. 285 ). The MHC molecules contain a groove in which protein fragments or peptides can be bound.[15] The peptides are derived from either endogenous (binding MHC class I) or exogenous proteins (binding primarily to MHC class II) after processing by proteolytic enzymes. After peptide binding, the MHC-peptide complex is transported to the cell surface for presentation to T cells. Because of the enormous diversity of different T-cell receptor molecules, it is predicted that for every possible MHC-antigen complex (including tumor antigens), one or more complementary T-cell receptor molecules exist. T-cell activation is a two-signal process. In addition to the T-cell receptor binding to the MHC-peptide complex, a second or costimulatory signal is required (see Fig. 285 ). This signal is provided through interaction of certain ligands on the antigen-presenting cell and their receptors (costimulatory molecules) expressed on the T cell. A number of such interactions have been identified, including the T-cell molecules CD28 and CD11a/CD18 (leukocyte functionassociated molecule-1 [LFA-1]), which bind to molecules of the CD80 (B7) family and CD54 (intercellular adhesion molecule-1 [ICAM-1]), respectively.[15] Lack of costimulation induces unresponsiveness or even cell death. When properly activated, T cells undergo cell division (clonal expansion) and become functionally active. Expansion of T cells is also dependent on the availability of growth factors, such as IL-2. Most cd8+ T cells do not produce sufficient amounts of IL-2 to sustain proliferation and are dependent on cd4+ T cells for their IL-2, in particular the TH 1 subset of cd4+ T cells. Whereas cd8+ T cells recognize antigens presented by MHC class I molecules, cd4+ T cells are generally MHC class II restricted. Because most tumors of nonhematopoietic origin do not express MHC class II, cd4+ T cells are dependent on presentation of tumor antigen by professional antigen-presenting cells, such as dendritic cells and macrophages. B cells, on the other hand, are dependent on IL-4 that is produced by cd4+ TH 2 cells instead of TH 1 cells.
Escape From Immune Surveillance

With the availability of both innate and acquired immunity, one wonders why these systems have failed in cancer patients. Extensive research has demonstrated the existence of multiple escape mechanisms that permit tumor cells to escape from elimination by immune effector cells. Only those transformed cells that have a sufficiently distinct
767

TABLE 28-4 -- Tumor Escape Mechanisms Mechanism TUMOR RELATED Tumor is not immunosensitive No expression of tumor-specific antigens No or low expression of MHC molecules No antigen processing/presentation Resistance to immune cell-mediated killing Tumor is not immunogenic Lack of costimulatory molecules Secretion of immunosuppressive factors Characteristics

Shedding of tumor antigens Induction of T-cell tolerance Induction of T-cell apoptosis HOST RELATED Tumor grows too fast for the immune system Inherited or acquired immunodeficiency Treatment- or carcinogen-related immunosuppression Deficiency in antigen presentation by APC No access to tumor Expression of immunodominant antigens on parental tumor cells Age phenotype from normal cells are expected to induce an immune response. At the early stages of tumor development, the tumor cells may not (or may at insufficient levels) express tumor antigens that can be recognized by T cells and/or B cells, and at the same time escape from innate immune effector cells. After subsequent mutations, however, tumor cells may become more distinct from normal cells, but this may still not lead to eradication by immune effectors because of tumor-related issues, host-related issues, or both (summarized in Table 284 ).
Tumor-Related Immune Escape

The tumor cells may not, for example, be immunosensitive because they lack antigens that induce an immune response. With the exception of malignant melanoma, most tumors do not induce a strong T-cell response in vitro. Even though this could be related to many factors, circumstantial evidence suggests that lack of expression of dominant tumor antigens is an important factor. In addition, the expression of certain or all MHC alleles on the tumor cells may be lower or completely lacking, thereby escaping from MHC-restricted lysis by T cells. It is estimated that at least 40% of solid tumors show a loss of one or more MHC alleles. In a growing number of cases, loss of MHC expression has been positively correlated with tumor aggressiveness and metastatic potential.[16] Furthermore, the tumor cells may not be able to process or present tumor antigens, or antigens may be masked or modulated. The latter was shown in leukemia cells in which specific antibody binding led to modulation of the antigen from the cell surface. Finally, tumor cells may be resistant to tumoricidal activity by immune cells. Different from immunosensitivity is immunogenicity. For example, tumor cells may express tumor antigens but still not induce an immune response because the tumor cells lack expression of costimulatory molecules. It has been shown that lack of expression of (CD54) ICAM-1 or (CD58) LFA-3 confers resistance to T-cellmediated lysis. Further, tumor cells may produce immunosuppressive factors such as TGF- and IL-10 that directly or indirectly inhibit T-cell function. Additionally, tumor cells may secrete soluble factors that downregulate T-cell molecules involved in signal transduction, such as the zeta chain of the T-cell receptor/CD3 complex and the tyrosine phosphatases p56lck and p59fyn . Signaling defects were first demonstrated in TILs isolated from renal cell carcinoma but were later demonstrated in many advanced cancers and even autoimmune diseases.[17] Soluble factors secreted by tumor cells may also redirect immune responses, activating immune cells that do not harm tumor cells. For example, secretion of IL-4 may shift the TH 1/TH 2 balance in favor of TH 2, whereas secretion of VEGF may block maturation and function of dendritic cells. Finally, the recently described regulatory CD4 T cells (Treg ) may be upregulated by factors and cells in the tumor environment. Treg are a subset of CD4 T cells that inhibit autoreactive T cells and B cells to prevent autoimmune disease.[18] As antitumor effector cells are in some respects autoreactive since tumor antigens can be nonmutated self-antigens, Treg inhibit antitumor effector cells. It was recently observed that the prevalence of Treg in peripheral blood and the tumor environment of cancer patients was significantly increased compared to that in normal individuals.[19] As another example of lack of immunogenicity, the antigen may be shed (e.g., carcinoma antigen [CA] 125 in ovarian cancer and carcinoembryonic antigen [CEA] in colon cancer), or it may induce T-cell tolerance. Tolerance induction may be related in some cases to the low avidity of the T cell tumor cell interaction. Finally, the tumor cells may induce apoptosis in activated T cells. Apoptosis, or programmed cell death, is a special type of physiologic cell death and should not be mistaken with nonphysiologic accidental cell death or necrosis. In certain tumors, such as lung cancer, pancreatic cancer, melanoma, and others, tumor cells may express the apoptosis-inducing molecule, Fas ligand, on the cell surface that can bind to the Fas receptor (FasR) expressed on activated but
768

not on resting T cells, and thereby induce apoptosis in T cells.[20]

Host-Related Immune Escape

Host-related factors that induce tumor escape include immunodeficiency (either inherited or acquired) that may increase the incidence of virally associated tumors; immunosuppression induced by radiation treatment, chemotherapeutic drugs, or chemical or physical carcinogens; tumor progression that may outpace tumor regression; deficient presentation of tumor antigens by antigen-presenting cells; lack of access of effector cells to the tumor (there is experimental evidence that the stroma can form a barrier); or expression of immunodominant antigens on parental tumor that prevents stimulation by other tumor antigens. The last mentioned is an intriguing phenomenon that was described in the 1980s but has so far not been explained. There appears to be a hierarchy in the immune response to various tumor antigens because the host fails to recognize all antigens simultaneously. The expression of an immunodominant antigen would prohibit the immune system from responding to antigen-negative tumor cells. A less clear failure of the host to mount an antitumor immune response is related to age. Over time, there may be a decrease in effectiveness of the immune response. The incidence of cancer may also rise over time, however, because certain carcinogens may have a long latency period or because exposure to certain carcinogens is accumulated over many years.
Antitumor Therapies

Therapies to treat cancer include conventional therapies such as surgery, chemotherapy, and radiation therapy, and a large variety of alternative treatments ( Table 285 ) most of which are still experimental. Because surgery is extensively discussed in other chapters, it is not discussed further here. All therapies, apart from surgery, have in common that they directly or indirectly target the destruction of tumor cells. Mechanistically the various strategies TABLE 28-5 -- Antitumor Therapies Treatments CONVENTIONAL Surgery Excision of malignancy Characteristics

Chemotherapy Radiation ALTERNATIVE Oncoviral Gene Antiangiogenesis Immune Stem cell

Drug-induced tumor cell lysis Radiation-induced tumor cell death Tumor cell lysis through selectively replicating virus Restore/add or inhibit expression of particular genes in tumor cells Blockade of tumor vasculature Tumor lysis through manipulation of the hosts immune system Graft-versus-malignancy response through stem cell transfer

differ widely. Some strategies attempt to interfere with intracellular signaling pathways to block cell division and/or induce apoptosis (e.g., chemotherapy and radiation), others target growth factors or cell surface receptors for growth factors, the tumor vasculature, or attempt to induce/enhance an antitumor immune response. Rather than discussing all possible therapies, we briefly discuss some of the main strategies in the following sections.
Chemotherapy

In chemotherapy, the goal is to kill rapidly dividing cells with drugs and leave other cells unharmed. Unfortunately, certain normal cells also divide rapidly; these cells include hair cells, bone marrow cells, and epithelial cells lining the digestive tract. Most anticancer drugs cause DNA damage or inhibit DNA replication and transcription (cytostatic drugs). In addition to the direct damage effects, however, certain genes may be triggered that are involved in apoptosis, leading to cell death.[21] Apoptosis plays a pivotal role in homeostasis of human cell proliferation. 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. 286 ). Activation of caspase cascades leads to DNA fragmentation and apoptosis. Induction of apoptosis by anticancer drugs is either death receptor dependent (extrinsic pathway) or 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 the TRAIL, respectively. Binding of the ligands triggers activation of caspase 8, which indirectly leads to release of cytochrome c from mitochondria and eventually apoptosis. 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

Figure 28-6 Apoptotic pathways. Chemotherapeutic drugs can induce apoptosis in tumor cells through a receptor-dependent (1) or receptor-independent (2) pathway. Both pathways induce release of cytochrome c from mitochondria, which triggers activation of various caspases in sequence, ultimately leading to apoptosis.

769

caspase 9, which activates caspase 3, which then leads to apoptosis. Physiologic apoptosis appears to be regulated by multiple genes, but promotion of cell death by cytostatic drugs is usually regulated by a select number of genes. These genes include both tumor suppressor genes such as p53 and oncogenes such as myc, E2F, c-jun, and bcl-2 (B-cell leukemia/lymphoma gene-2).[22] The p53 gene is rapidly upregulated after DNA damage and induces growth arrest by blocking cell cycle progression from G1 to S phase to allow repair of the DNA. If the repair is unsuccessful (as may be the case after treatment with certain drugs), the cells undergo apoptosis.[4] However, the tumor suppressor gene, p53, is mutated in about 55% of tumor types, which may contribute to resistance to apoptosis-inducing drugs.[23] Both myc and c-jun encode transcription factors that play a central role in regulating cell proliferation, differentiation, and apoptosis. The E2F gene family is a family of transcription factors that promote S-phase entry and thus with respect to cell growth appears to have the opposite function as p53. The role of transcription factors in apoptosis may differ depending on cell type and stimulus. In addition, more than 100 transcription factors have been identified and their biologic roles are not yet fully understood. Further enhancing the difficulty of understanding their role in apoptosis is the fact that several transcription factors activate downstream target genes, whereas others downregulate expression in response to DNA damage. The target genes play important roles in the overall outcome of the cells fate. The bcl-2 gene is a member of a family of proteins that induce (bax, bak, bcl-Xs) or inhibit (bcl-2, bcl-Xl, bag1) apoptosis. High levels of bcl-2 are found in tumor cells and may confer resistance to anticancer drugs. Initially, dysregulation of bcl-2 was found in B-cell malignancies, but overexpression has also been noted in solid tumors, including prostate, lung, colon, and stomach cancers. It has become apparent that cell-cycle progression is under tight genetic control. Growth factors or other stimuli activate oncogenes or inactivate tumor suppressor genes to allow progression through the cell cycle. Cells that receive conflicting, excessive, or unbalanced mitogenic signals undergo apoptosis. A similar deregulated cell cycle control and induction of apoptosis is observed after treatment with certain chemotherapeutic drugs. Some anticancer drugs are effective as single agents (e.g., cisplatin for treatment of testicular cancer), but combination therapy with multiple drugs is common practice. The major advantage with combination chemotherapy is the increased chance to overcome drug resistance. Drug resistance may be inherent to the tumor cells or may arise after exposure to anticancer drugs. Certain types of cancers, such as pancreatic, renal, and non-small cell lung

cancer, are known to be poor responders to chemotherapy. In contrast, cancers such as breast, ovarian, and small cell lung cancers and acute leukemias and lymphomas show a high initial response rate, but durable responses are rare because of the acquisition of drug-resistance mechanisms. Cellular resistance mechanisms can be distinguished from extracellular mechanisms. Cellular mechanisms of drug resistance include decreased intracellular drug accumulation resulting from impaired transport through the membrane, or active efflux of drugs (multidrug resistance [MDR] phenotype) through the MDR-1[21] and multidrug resistanceassociated protein (MRP) gene products or other drug transporters.[24] Alternatively, alterations may occur in intracellular drug metabolism, or the target of the drug may change, reducing the effectiveness of the drug. For example, drugs such as 5-fluorouracil and cyclophosphamide are dependent on conversion from an inactive to an active form by drug-activating enzymes. A common feature of drug-resistant cells is decreased activity or loss of drug-activating enzymes. Intracellular drugs can also be detoxified; for example, glutathione-S transferases are implicated in resistance to alkylators, and cytosolic retinoic acidbinding proteins may neutralize retinoic acid used for treatment of promyelocytic leukemia. A different mechanism is resistance resulting from changes in the target of the drug. Because of mutations in the target affecting the binding affinity of the drug, or overexpression of the drug target, enzymes involved in DNA synthesis or DNA multiplication are less sensitive to drugs. Other escape mechanisms include enhanced DNA repair capacity and disruption of the apoptotic pathway normally activated by drugs (e.g., overexpression of bcl-2 or MDM-2 [p53 inhibitor] or mutation of either p53 or the retinoblastoma gene).[23] Finally, insufficient penetration of the drug to and into the tumor may affect treatment outcome unfavorably.
Radiation Therapy

Radiation therapy is used alone or in conjunction with surgery, chemotherapy, or both. About half of cancer patients receive radiation therapy at some time during their treatment. Despite the successes in tumors such as seminomas, carcinomas of the skin, cervix, prostate, anus, and head and neck, it is not entirely clear why certain tumors are preferentially sensitive to radiation and why other tumors are resistant. Cell differentiation, proliferation, and maturation all are potentially affected by ionizing radiation. The effects of ionizing radiation are obvious at different levels of organization. At the cellular level, the most sensitive target for radiation is DNA, in which single-strand or double-strand breaks or cross-links are induced. Alternatively, sugars, nucleotides, or both are damaged. In addition to DNA damage, transcription of certain genes is induced. These genes can be distinguished into early-response genes, which mostly encode transcription factors, and late-response genes, which encode cytokines and growth factors. The physiologic significance of gene activation through radiation is not entirely clear yet. Other than the nucleus, the cell membrane is radiosensitive, and damage involves lipid peroxidation by radiation-induced oxyradicals. Lipid peroxidation alters membrane fluidity and permeability and may affect ion fluxes and membrane-mediated transport processes. Ionizing radiation may also affect cell-cell interactions, which, for yet unknown reasons, occur more frequently in
770

tumors than in normal tissues. To control radiation-induced damage, cells are equipped with DNA repair mechanisms and mechanisms to detoxify oxygencontaining radicals. Radiation has recently been correlated with the induction of apoptosis (e.g., in thymocytes and the parotid gland, but also in tumors).[25] Recent work in the nematode Caenorhabditis elegans, however, suggests that the apoptotic pathway induced through radiation is different from that induced by physiologic triggers. Radiation may also induce delays in cell division through interference with the cell-cycle processes (see Fig 282 ). Most common is G2 -phase arrest, but G1 -phase arrest and S-phase delay have also been noted. Interference with the cell cycle is most likely related to the effect of radiation on tumor suppressor genes and oncogenes encoding cell-cycle regulatory proteins, such as p53 and cdks.[25] For example, radiation induces the G1 checkpoint through stabilization of the p53 protein and through induction of the cdk inhibitor p21. Chemotherapeutic drugs may increase the effectiveness of radiation. More complicated is the evaluation of radiation damage in organs because of the presence of multiple cell types that may be differentially sensitive to radiation. The response of organs to radiation may in the early stages be characterized by a rapidly renewing stem cell population, early manifestation of damage, and complete healing; late stages are characterized by the lack of a separate stem cell population, late manifestation of damage, and irreversible damage. Intermediate stages are characterized by some early reactions and some late reactions. Tumors mostly resemble early-responding tissues, although the distorted physiology, vasculature, necrosis, and out-of-cycle cells make the situation more complex.
Alternative Treatments

The alternative treatment strategies can roughly be divided into five types ranging from entirely experimental such as oncoviral and genetic therapy to standard adjuvant therapy such as bone marrow or stem cell therapy (see Table 285 ). Oncoviral therapy involves injection of replicating virus into cancer patients. Replication of the virus is genetically engineered to be tumor selective, leading to destruction of tumor cells while sparing normal tissue. The most advanced in clinical testing is ONYX-015, an adenovirus that selectively replicates in cells deficient for the tumor suppressor gene p53. [26] In addition to selective targeting, the oncolytic virus may contain a therapeutic gene. Examples are the enzymes thymidine kinase and cytosine deaminase that convert the prodrugs ganciclovir and 5-fluorocytosine, respectively. Alternatively, oncolytic viruses encode immune-stimulatory genes such as cytokine genes. The use of therapeutic genes in oncolytic viruses combines oncoviral therapy and genetic therapy or immune therapy. The transfer of copies of wild-type genes into tumor cells is referred to as gene therapy and aims to restore (e.g., tumor suppressor gene) or add (e.g., immunestimulatory gene or chemoresistance gene) expression of genes. In contrast, the transfer of antisense oligonucleotides or small interfering RNA into tumor cells aims to inhibit expression of a particular gene such as that of an oncogene. Transfer of DNA or RNA into cells is a relatively simple procedure in vitro, and a variety of methods can be used with a high efficiency. Clinical application, however, requires the efficient transfer of potentially billions of tumor cells in a safe manner that is nondestructive to normal tissues. Thus, choice of vector ( Table 286 ), transgene, and method of delivery all are crucial parameters. The use of viral vectors is common because of the ability of pathogenic viruses to transfer genes in an efficient way. Unlike in oncoviral therapy, the viral genes responsible for viral replication are removed, and the transgene is inserted.[27] Different applications may favor one vector over another. Nonviral vectors are either cell based or noncell based (see Table 286 ). Using cell lines, the genetic modification is relatively easy to perform in vitro, and successful modification can be monitored before the vector is used therapeutically. The noncell-based nonviral vectors include liposomes and plasmid DNA. The artificial lipophilic vesicles fuse with the cell membrane, and the contents are released into the host cell. This application is often used to transfer antisense oligonucleotides. The other noncell-based nonviral vector that is gaining interest is plasmid DNA, in particular for DNA vaccines. Direct injection of DNA into muscle cells, especially skeletal and cardiac muscle cells, has proved more effective than was predicted based on in vitro experiments. The immune response resulting from DNA vaccines appears to be mediated through dendritic cells that have taken up the DNA and express the transgene. Nonviral, noncell-based vectors are sometimes preferred because of biosafety concerns: larger DNA sequences can be transferred and the vector has greater stability. In general, the problems with gene therapy are the poor targeting and the low efficiency of gene transfer in vivo. Attempts to block the vasculature of tumors (antiangiogenesis therapy) have shown encouraging results in mice but have thus far been less successful in

cancer patients. The outgrowth of new blood vessels from preexisting ones is essential for tumor growth and metastasis. Angiogenesis is mediated by both stimulatory (VEGF and basic fibroblast growth factor) and inhibitory factors (endostatin, angiostatin). Both naturally occurring angiogenesis inhibitors such as endostatin and angiostatin as well as synthetic agents or antibodies against angiogenesis factors are currently being tested in clinical trials. The primary goal of immune therapy is to restore the imbalance between tumor growth and tumor destruction. TABLE 28-6 -- Vectors Used for Genetic Therapy System Viral Nonviral Cell based Non-cell based Tumor cells, fibroblasts, dendritic cells, T cells Liposomes, plasmid DNA Vector Retrovirus, adenovirus, adeno-associated virus, herpesvirus, pox virus

771

The secondary goal of immune therapy is to induce protective immunity.[7] [28] Immune therapy is a relatively new treatment modality for cancer patients despite sporadic attempts made more than 100 years ago. Strategies may target innate as well as acquired immune components. Bacterial vaccines, such as bacille Calmette-Gurin (BCG) and Corynebacterium parvum stimulate innate immunity and are used as single agents (e.g., BCG is effective against residual bladder cancer) or in combination with specific stimulants. Likewise, cytokines, such as ILs, may boost innate immunity. Systemic administration of IL-2, for example, is used for the treatment of renal cell carcinoma and malignant melanoma. Durable cancer regressions have been induced by treatment with IL-2 alone. Systemic administration, however, usually requires a high dosage, which may explain the associated severe toxicity that sometimes occurs. Certain cytokines such as IFNs may also directly act on tumor cells and inhibit tumor growth.[11] Strategies that target acquired immunity can be divided into active therapeutic immunization, passive therapy with antibodies, and adoptive transfer of immune effector cells ( Table 287 ). Of these, most emphasis is placed on active therapeutic immunization.[29] The large variety of immunization approaches can roughly be grouped into three strategies. First, autologous or histologically matched allogeneic tumor cells genetically modified ex vivo to express a cytokine, a costimulatory molecule, or other immune-stimulatory molecule. Alternatively, autologous dendritic cells modified to express tumor antigen(s) are used. Tumor antigens can be introduced into dendritic cells through coculture or through transduction or transfection. Dendritic cells can also be fused with tumor cells in vitro. Finally, tumor antigen(s) in the form of DNA, peptide, or protein can directly be injected into patients with either an adjuvant or by using viral or nonviral vectors (see Table 286 ). For hematopoietic tumors, immunization can be achieved with anti-idiotypic antibodies, which have the internal image of a tumor antigen. The B-cell receptor, just like the T-cell receptor, consists of a heavy chain and a light chain covalently linked through disulfide bonds. Both the heavy chain and the light chain are genetically encoded by a unique combination of DNA segments (variable, diverse [heavy chain only], and joining regions). The genetic rearrangement provides a unique antigen-binding site determined by the sets of variable domains. This creates unique regions or epitopes within the variable domains that are called idiotopes. The set of idiotopes expressed by a given immunoglobulin molecule is called an idiotype. Because of their uniqueness, antibodies can be generated against idiotypes; these are termed antiidiotype antibodies. Thus, antibodies against tumor-specific antibodies can be used as immunogens to induce an antitumor immune response. Passive antibody therapy focuses on molecules that are either uniquely expressed on tumor cells or overexpressed on tumor cells relative to normal cells. Examples include gangliosides and mucins. By using IgM antibody, complement-mediated cytotoxicity is induced, as well as inflammation and phagocytosis of tumor cells by the reticuloendothelial system. On the other hand, IgG antibodies may trigger antibody-mediated cytotoxicity. A variation of this theme is the use of bispecific antibodies. These antibodies are engineered to have specific binding sites for two different antigens. Consequently, bispecific antibodies can simultaneously bind to immune effector cells and tumor cells and thereby bridge effector cells and tumor cells. By cross-linking receptors such as CD3 or T-cell receptors, the effector mechanism of the immune cells is activated and the tumor cell is lysed. Antibody therapy may be successful against free tumor cells and micrometastases. Although not strictly used to target acquired immunity, there are a number of antibody-based therapies worth mentioning. They involve the use of antibodies directed against growth factor receptors or angiogenesis factors expressed on tumor cells or in the tumor environment. It was recently reported at the 36th Annual Meeting of the American Society of Clinical Oncology that anti-VEGF antibody effectively blocked angiogenesis in metastatic colon cancer patients and significantly prolonged survival. Likewise, the IL-2 receptor on T-cell tumors is targeted by anti-IL-2 receptor antibodies. Recently, it was observed that certain antibodies may activate apoptosis in tumor cells through a yet undetermined mechanism, such as the anti-Her2/neu antibody for treatment of breast cancer. Finally, tumor-specific antibodies are used as vehicles for delivery of toxins or radionuclides to tumors. Adoptive transfer of immune effector cells was made popular in the early 1980s.[28] Initially, a high concentration (1000 IU/mL) of IL-2 was used to generate lymphokine-activated TABLE 28-7 -- Strategies of Immunotherapy Strategy Active therapeutic immunization 1. Gene-modified tumor cells 2. Dendritic cells expressing tumor antigen 3. Direct injection of tumor antigen Attempts to decrease tumorigenicity Passive antibody therapy Adoptive transfer of immune effector cells Attempts to induce tumor cell cytolysis or cytostasis Attempts to increase the number of specific antitumor effector cells in the host Characteristics Attempts to boost a host immune response through:

772

killer (LAK) cells in vitro from peripheral blood mononuclear cells (PBMCs) obtained by leukapheresis. LAK cells and IL-2 were administered to patients. PBMCs were replaced by tumor-derived lymphocytes when it was found that these cells activated with IL-2 were more effective than IL-2

activated PBMCs in animal models. Both strategies, however, showed a fairly similar efficacy in patients with metastatic disease, primarily in patients with metastatic melanoma. Response rates, including complete and partial remissions, ranged from 25% to 35%.[30] These studies illustrate some of the complexities related to immune therapy of cancer. For example, thus far, there are no good in vitro parameters that predict in vivo efficacy of the transferred cells. The number of adoptively transferred cells and the tumor load do not appear to correlate with clinical responses. Likewise, tumorspecific cytolytic activity or tumor-specific cytokine release in vitro does not guarantee a clinical response. The adoptively transferred T cells of clinical responders, however, usually do show a tumor-specific response in vitro. This points to the importance of inducing or selecting for tumor-specific effector cells and raises the question of which strategy should be used. Another question is which immune effector cells should be activated. Increased understanding of how NK and NKT cells recognize their target cells may change the focus from T cells to NK and NKT cells. Because loss of MHC expression is a common phenomenon in many solid tumors, NK cells could be an important component in immune therapy. For many years, controversy has existed about how adoptively transferred T cells induce tumor eradication. The most common thought has been that the transferred tumor-specific T cells find their way back to the tumor and eradicate it. In a mouse sarcoma tumor model, however, evidence was obtained that suggests a different scenario.[31] A proportion of the adoptively transferred T cells indeed traffic back to the tumor, but they induce a successful host antitumor response. The induction is mediated through secretion of inflammatory cytokines, such as IFN- and granulocyte-macrophage CSF, that are secreted in response to reactivation of tumor-specific donor T cells by tumor cells. Strategies aimed at promoting immune responses can potentially be combined with strategies aimed at blocking immune suppression. The documented benefit of cyclophosphamide to eliminate suppressor cells prior to therapy dates back many years, and cyclophosphamide is still presently used. Recently, cd4+ CD25+ regulatory T cells (Treg ) were identified as suppressor cells of autoreactive CD4 and CD8 T cells, which includes tumorreactive T cells.[18] Treg can be depleted in the host through anti-CD25 antibody, and in animal tumor models this has led to significant tumor regression. Finally, stem cell therapy is a standard treatment for hematologic malignancies. Prior to stem cell transfer, high doses of cytotoxic agents with or without total-body irradiation is used to eradicate malignant disease and reduce graft-versus-host disease in case of an allogeneic stem cell transfer. The goal is to induce a graft-versus-malignancy response mediated by donor T cells. This response is more pronounced after allogeneic stem cell transfer than autologous stem cell transfer and is likely mediated by donor CD4 and CD8 T cells that recognize tumor-specific proteins, overexpressed normal proteins in malignant cells, and minor histocompatibility antigens.
Conclusion

Even though scientists have learned a great deal about tumor biology since the early 1990s, a detailed understanding of how tumors arise, develop, and progress into highly metastatic tumors is still lacking. We are still at the stage at which finding the answer to one question raises several new questions. Nonetheless, our increased knowledge will most certainly lead to the design of better and more refined treatment modalities. Additionally, as more is learned about the biology of tumors, more tumor markers will become available that can be successfully applied to the diagnosis and treatment of cancer.
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/301.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

TUMOR MARKERS
What Is a Tumor Marker?

Tumor markers are molecules that can be detected in blood, body fluids, or tissue of a host with underlying cancer. It is not unusual for tumor cells to secrete or shed molecules or parts of a molecule. Occasionally, these molecules can be useful tumor markers, as evidenced by the monitoring of prostate-specific antigen (PSA) or prostate serum acid phosphatase (PSAP) levels in serum.[32] Elevated levels of PSA or PSAP are indicative of prostate cancer. PSA is a proteolytic enzyme that is secreted by the prostate into the glandular lumina, where it is contained and excreted with prostatic fluid. Cancer in the prostate can cause PSA to leak out of the gland into the blood. Note that other disease forms in the prostate, such as benign hyperplasia of the prostate and prostatitis, may also induce leakage of PSA. PSAP is another proteolytic enzyme that may leak into the blood, but it is associated with a more advanced disease stage. Likewise, elevated levels of CA 125 and mucin (MUC)-1 are associated with disease progression of ovarian and breast carcinoma, respectively.[33] Classically, tumor markers are macromolecules such as proteins, carbohydrates, and DNA released by the tumor and detected in blood. For example, CEA is a modified protein secreted by multiple tumor types such as colorectal, pancreatic, breast, and lung cancers.[34] However, tumor markers may also be molecules released by the normal host tissue in response to the invasion of cancer. The first tumor marker described acid phosphatase was such a normal host enzyme secreted by bone in response to invading prostate carcinoma. The ideal tumor marker should have several characteristics to be useful clinically. First, a simple test should detect a tumor marker long before the malignancy spreads widely. In many occasions, tumors are diagnosed by microscopic examination after obtaining a biopsy specimen of the tumor. However, by the time a tumor becomes
773

clinically apparent enough to prompt a biopsy examination, the tumor has already spread widely. This can be avoided by using a tumor marker that can be detected long before a biopsy is possible, allowing the clinician to alter the natural history of the tumor in the host. Second, testing for a tumor marker should be sensitive; that is, the test for the tumor marker should be positive in all patients with a particular cancer (no false-negative results). Third, the test for the tumor marker should be specific, meaning that the test should be positive for the marker only when the patient has that particular cancer (no false-positive results). Fourth, the level of the tumor marker detected in the biologic sample should correlate with the size of the tumor. Fifth, the test for the tumor marker should be cheap enough to allow use of the marker in screening of large numbers of individuals. The sensitivity of a tumor marker is determined by three interrelated factors: (1) the ability of the assay to reliably detect the tumor marker in a given biological sample; (2) the critical tumor mass that should be present before the marker can be detected; and (3) the percentage of tumors that expresses the particular tumor marker being tested. Development of highly sensitive immunoassays has decreased the threshold concentration of detection as well as the size of the tumor needed to achieve that concentration. However, discovering tumor markers that are uniformly expressed by all tumors with the same classification has been more of a problem. The specificity of a tumor marker is also dependent on three factors: (1) the fidelity with which the assay detects only the tumor marker; (2) presence of other tumors that test positive for the same marker; and (3) other nonmalignant conditions in which the tumor marker is elevated. The ideal tumor marker with all the desired characteristics has not been found yet. Despite advances in molecular biologic techniques and discovery of many new markers, all known tumor markers have characteristics that limit their clinical utility. Clinical applications of tumor markers in screening, diagnosis, prognosis, evaluation of success of therapy, and monitoring for recurrence are limited by problems in sensitivity and specificity. However, a marker that is not ideal for one application may be useful in another application. For example, a marker such as CEA may not be ideal for screening for colorectal cancer in large populations of people, but it may be useful in detecting recurrence in a patient who has undergone resection of the primary tumor. Examples of tumor markers with their clinical utility and potential limitations are listed in Table 288 .
Molecular Basis of Tumor Markers

Tumor markers are a result of genetic alterations that occur in tumor cells that directly or indirectly affect the gene expression pattern of the tumor cells or the surrounding tissue. Molecular changes that occur in a cell that leads to the expression of a tumor marker can stem from incorporation of viral genes into the tumor cell, chromosomal translocations in the tumor cell, changes in the DNA methylation pattern, or any other genetic change such as point mutations and deletions. Two categories of abnormalities can be distinguished based on their importance in oncogenesis: primary abnormalities that are essential at the tumor initiation stage, and secondary abnormalities that occur later and are thought to play a role during tumor progression. The primary abnormalities, such as the Philadelphia chromosome,[35] are specific for a particular tumor, whereas the secondary abnormalities are less specific. Primary abnormalities are mostly detected in leukemias, non-Hodgkins lymphoma, and mesenchymal solid tumors and are of diagnostic value. In general, the genetic changes are reciprocal translocations leading to disruption of genes controlling growth, differentiation, and apoptosis. In epithelial solid tumors, the high incidence of chromosomal rearrangements with gross aneuploidy complicates karyotyping. In addition, these tumors are usually studied at later stages of development, at which time the tumor cells have undergone many additional genetic changes. Nonetheless, several primary abnormalities have been proposed in lung carcinoma, nonpapillary renal cell carcinoma, transitional cell carcinomas of the bladder, and adenocarcinoma of the prostate. Most chromosomal aberrations in these tumors involve deletions, unbalanced translocations, and loss of entire chromosomes. This may suggest that loss or inactivation of tumor suppressor genes plays an important role in these tumors.
Viral Genes

Tumors induced by viruses such as the Epstein-Barr virus (EBV) implicated in Burkitts lymphoma, human papillomavirus (HPV) linked to cervical cancer, hepatitis B virus (HBV) associated with hepatocellular carcinoma, and human T-lymphotropic virus type 1 (HTLV-1) associated with T-cell leukemia may express viral gene products that are easily detectable because they are not expressed in normal cells. The viral DNA may be inserted into the host genome or exist in the cytoplasm, but in both cases, the cellular mechanisms of the host are manipulated to serve the virus. Papillomavirus DNA, for example, encodes genes of which the product inactivates the p53 and the Rb genes. In addition, viral genes are transcribed that are involved in viral replication.
Chromosomal Translocations

Chromosomal translocations occur frequently and often nonrandomly. Because of the breakpoint and annealing at a different location, hybrid proteins may be expressed that could be used as markers.[36] Good examples of hybrid proteins with diagnostic value are the Philadelphia chromosome in chronic myeloid leukemia (CML) (90% of cases) and the EGF receptor in glioblastomas. The Philadelphia chromosome is a shortened version of chromosome 22 that resulted from the reciprocal translocation of part of chromosomes 9 and 22, resulting in a longer chromosome 9 and shorter chromosome 22.[35] The EGF receptor, which is expressed in about 40% of gliomas, has an internal deletion resulting in a characteristic fusion protein that contains
774

TABLE 28-8 -- Tumor Markers in Clinical Application Tumor Marker CEA Type of Molecule Glycoprotein Tumor Type Stomach, liver, pancreas, breast, and colorectal cancer Pancreas cancer Type of Application Prognosis and monitoring therapeutic response Nonmalignant Conditions with Elevated Levels Hepatitis, cirrhosis, jaundice, COPD, inflammatory bowel disease, smoking

CA 199 AFP PSA -HCG

Mucin-type glycoprotein Oncofetal glycoprotein Oncofetal glycoprotein Trophoblastic protein Ovarian cell surface protein Membranes of breast cancer cells Prostate cellular protein

Diagnosis, prognosis, and monitoring therapeutic response

Hepatitis, cirrhosis, sclerosing cholangitis, and extrahepatic biliary stasis Hepatitis, cirrhosis, pregnancy, inflammatory bowel disease BPH, prostatic massage or biopsy Pregnancy

Liver and testicular Diagnosis, prognosis, and monitoring cancer therapeutic response Prostate cancer Screening, diagnosis, prognosis, and monitoring therapeutic response

Choriocarcinoma, Diagnosis, prognosis, and monitoring hydatidiform mole, therapeutic response and invasive mole Ovarian cancer Breast cancer Prostate cancer Prognosis and monitoring therapeutic response Prognosis Prognosis and monitoring therapeutic response Diagnosis

CA 125 CA 153 Prostatic acid phosphatase

Pregnancy, endometriosis, menstruation, jaundice, and pancreatitis Cirrhosis, hepatitis, and benign breast disease BPH, dermatologic disorders

5Peptide metabolite Carcinoid Hydroxyindoleacetic of indoleacetic acid acid Calcitonin Metanephrine Hormone Catecholamine metabolite Medullary thyroid cancer

Diagnosis, prognosis, and monitoring therapeutic response

Pheochromocytoma Diagnosis, prognosis, and monitoring therapeutic response

CEA, carcinoembryonic antigen; CA, carcinoma antigen; AFP, -fetoprotein; PSA, prostate-specific antigen; -HCG, beta-human chorionic gonadotropin; COPD, chronic obstructive pulmonary disease; BPH, benign prostatic hypertrophy. one or more new amino acids at the fusion point.[37] The fusion protein still has the ability to bind EGF and may in fact be more stable than the wild-type EGF receptor, resulting in overexpression. The overexpression may give glioma cells an advantage over normal cells when the amount of available EGF is limiting.
DNA Methylation

Altered DNA methylation patterns are also common in cancer cells and may be responsible for expression of some types of tumor antigens such as oncofetal antigens. Methylation of DNA at restricted cytosine phosphate guanine (CpG) sequences is a mechanism of regulating expression of certain genes. Genes that are expressed in fetal development are later silenced in adult life by methylating areas known as CpG islands near the promoter regions.[38] Methylation of CpG sequences in promoter regions blocks directly or indirectly the binding of transcription factors and thereby blocks transcription. When the distribution of methylation sites was evaluated in tumors, it was found that areas of hypomethylation exist as well as areas of hypermethylation. Hypomethylation may increase gene expression and may also decrease chromosome stability and lead to allelic loss. Hypermethylation, on the other hand, could suppress gene transcription. With the isolation and characterization of oncogenes and tumor suppressor genes, another feature of DNA methylation became apparent. Point mutations in those genes are frequently associated with CpG sites. The methylation of cytosine is thought to increase the chance for a C-to-T or C-to-A transition by 12-fold to 30-fold. Thus, 5-methylcytosine is considered an endogenous mutagen that contributes to about 30% of all point mutations despite the fact that 5-methylcytosine constitutes only 1% of human DNA. For example, when all human tumors are considered together, one third of all inactivating point mutations in the tumor suppressor gene p53 occur at methylation sites. Note that the point mutations related to methylation are not restricted to p53 only but rather occur in many genes.

775 Point Mutations and Deletions

Gene mutations have been identified in various human cancers, and the list of mutated genes includes tumor suppressor genes (BRCA-1, BRCA-2, APC, p53, nm23), oncogenes (ret, -catenin, ras, bcl), and genes involved in cell growth (cdk4). Other genetic changes affect the transcription level of genes: transcription of previously silent genes may be induced; the transcription level may be enhanced over the normal level, leading to overexpression of the gene product; or transcription may be reduced or entirely blocked. Mutations occur at various locations in genes, but often hot spots are present. These hot spots are gene restricted and not specific for the type of tumor. Nonetheless, certain of these mutations have diagnostic value, such as mutated ras, p53, and others, because they confirm the malignant state of a cell.

Detection Methods of Tumor Markers

Immunologic Detection

Immunologic techniques have long been used to identify existing specific tumor markers and to discover new markers. The discovery of monoclonal antibodies and the generation of antibody-producing hybridomas[39] have greatly expanded the repertoire of clinicians and pathologists in the detection of malignant cells. More recently, the immunologic discoveries of tumor-specific T cells further expanded the available techniques to identify new tumor markers. Tumor-specific T cells are currently not used for detection purposes since the generation and maintenance is too labor intensive. However, the use of tumor-specific T cells, in particular cytotoxic T lymphocytes, in vitro has clearly helped to identify tumor antigens in a variety of tumors and has formed the rationale for immune-based therapies (see Tumor Biology section). In general, monoclonal antibodies that bind specifically to epitopes on tumor markers are used for detection of tumor cells in biopsy specimens, blood cells, or other body fluids. The first and still most commonly used method for the initial diagnosis of cancer is through microscopic evaluation of tissue sections or cytologic preparations. Routinely, tissue is fixed in formalin to preserve morphologic structures and is embedded in paraffin or plastic. The fixation process, however, may negatively affect the three-dimensional structure of antigens recognized by antibodies. An alternative method involves the snap-freezing of tissue, which does not include a fixation step and does not denature antigens. The downside of this method, however, is that the morphology is poorly preserved. To use antibodies in paraffin sections, second-generation antibodies are developed that are selected to recognize antigens that are not destroyed by the fixation process. Antibodies that recognize interesting epitopes in cryostat sections are used to identify the gene encoding the antigen. Subsequently, animals are immunized with the recombinant protein or protein fragments, and hybridomas are generated. The hybridomas are tested for secretion of antibodies that are reactive with the antigen of interest in paraffin sections. Second-generation antibodies have been made, for example, against the cell cycle dependent nuclear proliferation marker, Ki67, expressed in many cancers. Instead of immunohistology, flow cytometric analysis of antibody-stained cells is usually performed on hematopoietic tumors, mainly to facilitate classification. This method permits evaluation of the percentage of positive cells in a population of cells in suspension. In situ hybridization can be used to detect specific antigens in paraffin-embedded tissues. Specific oligonucleotide probes are incubated with the tissue slide to permit hybridization with the messenger RNA of the antigen of interest. The probe is labeled with a fluorescent or other dye to permit detection. Finally, enzyme-linked immunosorbent assays or radioimmunoassays are antibody-based detection methods to quantify the level of a particular marker in blood or other body fluids. Both procedures take less than a day and can detect quantities in the nanogram to picogram range (10[6] to 10[9] g). These types of assays are used for routine measurement of known markers, such as CA 153 in serum from breast cancer patients.
Cytogenetic Analysis

Another screening method used primarily to support the diagnosis of hematopoietic cancers is cytogenetic analysis. Chromosome aberrations can be the result of incorrect separation of chromosome pairs during mitosis (numerical chromosome change). One of the daughter cells may end up with an additional copy of the chromosome, whereas the other daughter cell is one copy short. Structural chromosome changes are initiated with DNA damage and incorrect DNA repair, leading to abnormal reconfiguration of broken chromosome ends. There is evidence that cancer-associated chromosomal aberrations are not random, although exact mechanisms are not fully understood. The first specific chromosome abnormality was observed in 1960 in patients with CML. The abnormality involves the earlier mentioned, unusually small chromosome 22.[35] Since the discovery was made in Philadelphia, the chromosome was named Philadelphia chromosome. It was subsequently found that chromosomal aberrations in hematopoietic cancers are restricted to a few chromosomes with an otherwise normal diploid karyotype. Cytogenetics, especially in solid tumors, is used to identify possible locations of new tumor suppressor genes. By comparing numerous chromosomal aberrations in tumors of a particular histologic type, certain abnormalities may show up with an elevated frequency, which may indicate a common loss of a tumor suppressor gene. This mapping strategy has helped significantly in identification of the BRCA-1 and BRCA-2 genes in breast carcinoma.[40]
Genetic Analysis

In addition to immunologic techniques, molecular biologic techniques have been developed for detection of tumor markers. One such technique is based on gene
776

analysis for detection of mutations in known molecules suspected of carrying a mutation. Generally, messenger RNA is isolated from the tumor cells and reverse transcribed into complementary DNA. The complementary DNA is incubated with an oligonucleotide specific for the tumor marker of interest, and the DNA is amplified by polymerase chain reaction (PCR), introduced in 1985.[41] Because of the specific primer, only the gene of interest is amplified by PCR. The PCR product is subsequently analyzed using a variety of techniques, such as analysis by gel electrophoresis, Southern blotting, and sequencing. Mutations in oncogenes (ras) and tumor suppressor genes (p53), are easily detected with these procedures. The PCR amplification method of genetic material can also be performed on formalin-fixed, paraffin-embedded tissues.
Clinical Applications of Tumor Markers
Diagnosis

Routinely, the first diagnosis of tumor is performed by the pathologist based on histologic and cellular characteristics (morphologic markers). Histopathologic classification is essential because not all tumors are equal, and they require differential treatment. In addition, within a particular histologic type, treatment strategies vary depending on the grade and stage of the tumor. Immunohistology using antibodies to detect specific markers is helpful for classification of the problem cases, such as apparently undifferentiated tumors. These tumors include large cell tumors, such as anaplastic carcinoma, and round cell tumors, such as Ewings sarcoma. Antibodies reactive to components of the intermediate-sized filaments can be helpful to classify tumors because their expression is tissue specific and is often preserved in malignancies. The intermediate-sized filaments are part of the intracellular matrix and comprise cytokeratins (carcinoma), vimentin (lymphoma, sarcoma, melanoma), desmin (myosarcoma), neurofilaments (neuroblastoma), and glial fibrillary acidic protein (astrocytoma). Within the more than 18 different keratins, cell-typespecific combinations occur, and antibodies against individual keratins may help in subclassification. Likewise, antibodies have been generated against a number of cell-lineage associated markers, such as CD45 (common leukocyte antigen), PSA and PSAP (prostate epithelium), G250 (renal cell carcinoma), gp100 (melanoma), and polymorphic epithelial mucins (secretorial epithelia). Antibodies reactive with leukocyte-associated antigens are used for the subclassification of lymphomas and leukemias. The list of clusters of differentiation, or CD, numbers contains more than 250 different antigens and keeps growing.[42] A similar, albeit much smaller, list of clusters has been prepared for differentiating antigens in lung cancer.

Prognosis

Prognosis is related to treatment. Morphologic criteria and the extent of metastatic disease are important prognostic parameters. Thus, factors that determine tumor growth, such as expression of particular growth factor receptors, the rate at which the tumor is growing, metastatic potential mediated through expression of certain receptors for ECM molecules, and sensitivity to therapy, all are important for prognosis. A marker indicative of tumor growth is the cell-cycledependent proliferation marker Ki-67. A high proportion of Ki-67positive cells is inversely correlated with survival. Likewise, overexpression of Her2/neu and the EGF receptor on breast cancer cells and high levels of MUC-1 in the serum of breast cancer patients is associated with poor prognosis. A high expression of integrin molecules (involved in attachment to ECM molecules) is related to poor prognosis. Many factors are found to be involved in tumor development, and therefore many are related to prognosis.
Monitoring Efficacy of Therapy

Tumor markers such as CEA, PSA, and CA 125 are easily detectable in serum. Their expression levels in serum correlate with tumor volume, which means that levels that are significantly higher than average baseline levels indicate the presence of malignancy. By evaluating levels pretherapy and post-therapy, the clinician can monitor the efficacy of therapy and detect recurrences. For example, a patient with very high CEA level will experience a drastic decrease in that level immediately after the successful removal of his or her colon cancer. Thereafter, any significant increase in that level in the years after the operation may signify recurrent colon cancer or metachronous cancer, and the patients physician can have a higher index of suspicion to prompt a cancer detection test such as colonoscopy early.
Guiding Choice of Therapy

Some tumor markers that can be easily evaluated through immunohistology may provide clinically important information, such as the evaluation of hormone receptor status in breast carcinoma. Estrogen receptor status has been for years the basis of selecting patients for potential therapy. Estrogen receptorpositive patients (especially postmenopausal patients) have been treated with tamoxifen. In contrast, estrogen receptornegative patients (especially premenopausal patients) have been treated with chemotherapy. Although physicians rely on a number of markers, such as CEA, PSA, and fetoprotein (AFP), to treat patients, they do not necessarily select treatment on the basis of the level of expression.
Prevention of Cancer

Examples of markers that may help prevent tumor development are the genes with inheritable mutations such as BRCA-1, BRCA-2, ret (associated with multiple endocrine neoplasia),[43] and APC, or other genes that are commonly mutated in the initial stages of tumor development. Mutations in these genes are likely to cause cancer at some stage in life. Early screening for the known gene mutations (i.e., before the development of cancer) may permit corrective
777

therapy, such as prophylactic surgery. Another such treatment may be gene therapy in which an intact copy of the mutated gene is transferred into cells with the defective gene.
Some Tumor Markers Most Frequently Used in Surgical Oncology

CEA is among frequently used markers in surgical oncology and is discussed here as a representative example (see also Table 288 ). CEA is an oncofetal protein with a molecular weight of 200,000 kD that is expressed in embryonic tissue. In adults, CEA is found in the mucous membranes of stomach, small intestine, and the biliary tree as a component of the glycocalix. Malignant epithelial type tumors such as colon cancer, lung cancer, and others may express higher levels of CEA on the cell surface and release CEA into blood. CEA in the normal cell is thought to function as a cell adhesion molecule. For cancer cells, CEA may offer a survival advantage by allowing adhesion into other cells and allowing metastasis. It may also offer some refuge from immune attack of the host due to the large glycosylated nature of the molecule creating a physical barrier between the immune effector cells and the cancer cell. CEA is detected in serum by commercially available immunoassay kits that very specifically measure the level of CEA. However, the measured level can vary depending on the brand of kit used, although the measured level is reproducible within the brand and the laboratory. In general, normal individuals have less than 2.5 ng/mL. Benign conditions listed in Table 288 can raise this level to higher than 10 ng/mL. CEA is cleared from serum by hepatocytes and Kupffer cells. Any benign liver condition that affects hepatocellular function or causes cholestasis can decrease clearance of CEA, resulting in higher plasma levels. Serum CEA level is not a useful test for screening for colorectal cancer for several reasons. In large studies, only about 40% patients with localized, potentially curable cancers had serum CEA levels higher than 5 ng/mL (sensitivity of 40%). This results in a large false-negative rate. If one uses the same 5-ng/mL level as cutoff for a normal CEA level, the test is 95% specific. However, due to the very low prevalence of colorectal cancer in large populations (1/1000), that 95% specificity still yields a large frequency of false-positive tests. Although CEA is not useful in screening large populations, the level of CEA in colorectal cancer patients is of prognostic value. In general, the higher the preoperative CEA level, the poorer the rate of long-term survival. Patients with normal preoperative CEA levels have lower recurrence rate than patients with elevated CEA levels. If the CEA level does not decrease to normal levels after resection of the cancer, the risk of recurrence is higher. Related to CEA are CA 125, CA 153, and CA 199, serum markers for detection of ovarian cancer, breast cancer, and pancreas cancer, respectively. All four molecules are high-molecular-weight molecules expressed on the respective tumors but not on normal tissues. Their expression levels correlate with tumor volume and as such they are useful indicators to monitor response to therapy and recurrence. As mentioned earlier, estrogen and progesterone receptor status is routinely determined on breast biopsies to determine choice of treatment. Additionally, Her2/neu status is evaluated using specific antibodies. Patients with disseminated Her2/neu-positive tumors are potentially eligible for treatment with the anti-Her2/neu antibody, Herceptin. PSA and AFP are the standard markers for detection of prostate cancer and hepatocellular carcinoma, respectively, and for monitoring the response to therapy. Finally, neuron-specific enolase (NSE), CEA, and squamous cell carcinoma (SCC) antigen are commonly used markers for lung cancer. NSE detected in serum helps to support a diagnosis of small cell lung carcinoma and a postoperative decrease of serum level of NSE is the first sign of curative resection. SCC antibody measurement in serum of lung cancer patients is an aid to histologic analysis: Patients with an SCC antibody level higher than 2 g/L have a 95% probability of having nonsmall cell lung cancer and 80% probability of having a squamous tumor. In spite of the usefulness of many markers as outlined earlier, caution should be taken with the interpretation of the observed values. In addition to

malignancy, several benign conditions may increase the value of serum markers. For example, many acute and chronic conditions of the biliary tract as well as cancers of the pancreas, colon, and stomach result in increased levels of CA 199. Likewise, the level of AFP is elevated in serum during pregnancy and cancers such as hepatocellular carcinoma and nonseminomatous testicular cancer. Nonmalignant conditions such as inflammatory bowel disease, hepatitis, and cirrhosis can also give rise to elevated AFP levels. Finally, benign conditions such as pregnancy, menstruation, endometriosis, pelvic inflammatory disease, hepatitis, cirrhosis, and renal failure may increase CA 125 levels.
Conclusion

Many tumor markers exist, and most are diagnostic markers. There is still a need for additional markers, however, especially those that indicate the presence of a tumor at very early stages. The completion of the sequencing of the human genome may increase the number and quality of available markers for all possible applications. The availability of markers for early diagnosis can be combined with more refined treatment modalities (e.g., gene therapy, vaccine therapy, and others) or with finding new indications for old treatments (e.g., prophylactic surgery).
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/302.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

Selected References
Burnet FM: The concept of immunological surveillance. Prog Exp Tumor Res 13:1, 1970. An introduction to the concept of immunologic surveillance.

778

Davis MM: T-cell antigen receptor genes and T-cell recognition. Nature 334:395, 1988. A review on structural features and genetic organization of the T-cell receptor in the context of antigen recognition. Fearon ER: A genetic model for colorectal tumorigenesis. Cell 61:759, 1990. The first genetic model for tumorigenesis. Kohler G: Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256:495, 1975. An introduction to hybridoma technology. Saiki RK: Enzymatic amplification of beta-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 230:1350, 1985. An introduction to polymerase chain reaction technology.

Copyright 2004 Elsevier Inc. All rights reserved.

www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/303.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

References
1. Clark 2. Clark

WH Jr, Elder DE, Guerry DT, et al: A study of tumor progression: The precursor lesions of superficial spreading and nodular melanoma. Hum Pathol 15:11471165, 1984. WH Jr, Elder DE, Van Horn M: The biologic forms of malignant melanoma. Hum Pathol 17:443450, 1986. ER, Vogelstein B: A genetic model for colorectal tumorigenesis. Cell 61:759767, 1990. K: Disruption of oncogene/tumor suppressor networks during human carcinogenesis. Cancer Invest 20:7181, 2002.

3. Fearon

4. Munger

GL: Autocrine and paracrine growth mechanisms in cancer progression and metastasis. In Bertino JR (ed): Encyclopedia of Cancer, vol 1, 2nd ed. San Diego, Academic Press, 2002, pp 165177.
6. Coussens

5. Nicolson

LM, Werb Z: Inflammation and cancer. Nature 420:860867, 2002. SA: Progress in human tumour immunology and immunotherapy. Nature 411:380384, 2001.

7. Rosenberg 8. Peoples

GE, Blotnick S, Takahashi K, et al: T lymphocytes that infiltrate tumors and atherosclerotic plaques produce heparin-binding epidermal growth factorlike growth factor and basic fibroblast growth factor: A potential pathologic role. Proc Natl Acad Sci U S A 92:65476551, 1995.

9. Culp LA: Extracellular matrix and matrix receptors: Alterations during tumor progression. In Bertino JR (ed): Encyclopedia of Cancer, vol 2, 2nd ed. San Diego, Academic Press, 2002, pp 215233. 10. Burnet

FM: The concept of immunological surveillance. Prog Exp Tumor Res 13:127, 1970. V, Ikeda H, Bruce AT, et al: IFN-gamma and lymphocytes prevent primary tumour development and shape tumour immunogenicity. Nature 410:11071111, 2001.

11. Shankaran 12. Dunn

GP, Bruce AT, Ikeda H, et al: Cancer immunoediting: From immunosurveillance to tumor escape. Nat Immunol 3:991998, 2002. SA: A new era for cancer immunotherapy based on the genes that encode cancer antigens. Immunity 10:281287, 1999.

13. Rosenberg 14. Davis 15. van

MM, Bjorkman PJ: T-cell antigen receptor genes and T-cell recognition. Nature 334:395402, 1988.

der Merwe PA, Davis SJ: Molecular interactions mediating T-cell antigen recognition. Annu Rev Immunol 21:659684, 2003. F, Ruiz-Cabello F, Cabrera T, et al: Implications for immunosurveillance of altered HLA class I phenotypes in human tumours. Immunol Today 18:8995, 1997.

16. Garrido 17. Finke

J, Ferrone S, Frey A, et al: Where have all the T cells gone? Mechanisms of immune evasion by tumors. Immunol Today 20:158160, 1999. EM: Regulatory T cells in autoimmmunity. Annu Rev Immunol 18:423449, 2000.

18. Shevach

19. Liyanage

UK, Moore TT, Joo HG, et al: Prevalence of regulatory T cells is increased in peripheral blood and tumor microenvironment of patients with pancreas or breast adenocarcinoma. J Immunol 169:27562761, 2002. J, Bennett MW, OSullivan GC, et al: The Fas counterattack: Cancer as a site of immune privilege. Immunol Today 20:4652, 1999.

20. OConnell 21. Tsuruo

T, Naito M, Tomida A, et al: Molecular targeting therapy of cancer: Drug resistance, apoptosis, and survival signal. Cancer Sci 94:1521, 2003.

22. Brantley-Finley C, Lyle CS, Du L, et al: The JNK, ERK, and p53 pathways play distinct roles in apoptosis mediated by the antitumor agents vinblastine, doxorubicin, and etoposide. Biochem Pharmacol 66:459469, 2003. 23. Brown

JM, Wouters BG: Apoptosis, p53, and tumor cell sensitivity to anticancer agents. Cancer Res 59:13911399, 1999. A: Cancer multidrug resistance. Nat Biotechnol 17:9495, 1999. B, Joseph B, Orrenius S: Tumor radiosensitivity and apoptosis. Exp Cell Res 248:1017, 1999.

24. Persidis

25. Zhivotovsky 26. Biederer 27. Wu

C, Ries S, Brandts CH, et al: Replicationselective viruses for cancer therapy. J Mol Med 80:163175, 2002.

N, Ataai MM: Production of viral vectors for gene therapy applications. Curr Opin Biotechnol 11:205208, 2000. SA: Adoptive immunotherapy for cancer. Sci Am 262:6269, 1990. N: Overview of therapeutic vaccination approaches for cancer. Semin Oncol 30:18, 2003.

28. Rosenberg 29. Berinstein

30. Goedegebuure

PS, Douville LM, Li H, et al: Adoptive immunotherapy with tumor-infiltrating lymphocytes and interleukin-2 in patients with metastatic malignant melanoma and renal cell carcinoma: A pilot study. J Clin Oncol 13:19391949, 1995. M, Goedegebuure PS, Burger UL, et al: Successful adoptive cellular immunotherapy is dependent on induction of a host immune response triggered by cytokine (IFNgamma and granulocyte/macrophage colony-stimulating factor) producing donor tumor-infiltrating lymphocytes. J Immunol 160:334344, 1998.

31. Nagoshi

32. Sauvageot

J, Epstein JI: Immunoreactivity for prostatespecific antigen and prostatic acid phosphatase in adenocarcinoma of the prostate: Relation to progression following radical prostatectomy. Prostate 34:2933, 1998. Bruijn HWA, ten Hoor KA, Boonstra H, et al: Cancerassociated antigen CA 195 in patients with mucinous ovarian tumours: A comparative analysis with CEA, TATI, and CA 125 in serum specimens and cyst fluids. Tumour Biol 14:105115, 1993.

33. de

34. Mitchell

EP: Role of carcinoembryonic antigen in the management of advanced colorectal cancer. Semin Oncol 25:1220, 1998. E, Lifshitz B, Gale RP, et al: Fused transcript of abl and bcr genes in chronic myelogenous leukaemia. Nature 315:550554, 1985.

35. Shtivelman 36. Rabbitts

TH: Chromosomal translocations in human cancer. Nature 372:143149, 1994.

779

37. Humphrey

PA, Gangarosa LM, Wong AJ, et al: Deletion-mutant epidermal growth factor receptor in human gliomas: Effects of type II mutation on receptor function. Biochem Biophys Res Commun 178:14131420, 1991. PA: The DNA methylation paradox. Trends Genet 15:3437, 1999. G, Milstein C: Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256: 495497, 1975.

38. Jones

39. Kohler 40. Marx 41. Saiki

J: A second breast cancer susceptibility gene is found. Science 271:3031, 1996.

RK, Scharf S, Faloona F, et al: Enzymatic amplification of beta-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 230:1350 1354, 1985.
42. Zola H, Swart B, Boumsell L, et al: Human leucocyte differentiation Antigen nomenclature: Update on CD nomenclature. Report of IUIS/WHO Subcommittee. J Immunol Methods 275:18, 2003. 43. Ponder

BA: The phenotypes associated with ret mutations in the multiple endocrine neoplasia type 2 syndrome. Cancer Res 59:1736s-1742s, 1999.

Copyright 2004 Elsevier Inc. All rights reserved.

www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/304.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

781

Chapter 29 - Melanoma and Cutaneous Malignancies

Marshall M. Urist M.D. Seng-jaw Soong Ph.D.

Skin cancers account for more than 40% of all malignancies in the United States, and the incidence continues to rise. This increase is attributed to environmental exposure, principally sunlight. The majority of skin cancers are basal cell carcinoma (BCC), squamous cell carcinoma (SCC), and melanoma, which account for more than 95% of the total. This chapter focuses on these three major types and briefly discusses the identification and management of less common cutaneous malignancies.
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/305.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

MELANOMA Melanocytes are cells of neural crest origin that migrate during fetal development to multiple sites in the body, principally the skin. Positioned along the basement membrane at the dermoepidermal junction, these cells are exposed to carcinogenic stimuli that result in malignant transformation to become melanoma. This event is relatively rare compared to the transformation rate for the neighboring basal keratinocytes that become SCC and BCC. Melanoma accounts for only 4% to 5% of all skin cancers but causes the majority of deaths from skin malignancies. It is the eighth most common cancer in the United States, and the incidence is rising faster than any other type of cancer. It is estimated that there will be 54,200 new cases diagnosed and 7600 deaths from melanoma in 2003.[1] The lifetime probability of developing melanoma is 1 in 57 for males and 1 in 81 for females. In whites, the 5-year relative survival rate has risen from 80% for 19741976 to 89% for 19921998.[2] This impressive progress has resulted from increased public awareness and education programs.
Epidemiology and Etiology

The incidence and outcome of melanoma are related to multiple factors. Melanoma is principally a disease of whites, particularly those of Celtic ancestry. The disease occurs much less commonly in Asian and black populations. It is estimated that melanoma occurs 20 times more often in whites than in blacks. The reason for this difference is unknown. The disease occurs slightly more often in men than women, and the prognosis is slightly better for women when other prognostic factors are taken into account. The anatomic distribution of melanoma varies between the two genders. Melanomas arise more commonly on the lower extremity in women and more often on the trunk and head and neck in men. These differences in distribution are not accounted for by sun exposure alone. Melanoma can occur at any age from birth to advanced age. The median age of diagnosis is in the range of 45 to 55 years. Tumors rarely develop before the age of puberty; however, there is a significant incidence in the 3rd and 4th decades of life.

782

It is well established that exposure to sunlight increases the risk of developing melanoma in susceptible populations. This is specifically attributed to solar ultraviolet (UV) radiation.[3] UVA and UVB cause different patterns of effect in the skin; however, both are considered to be carcinogenic. UVB induces the effects of sunburn and increases melanin production and is the most carcinogenic part of the UV spectrum. UVA has a deeper level of penetration resulting in dermal connective tissue damage, loss of elasticity, and skin wrinkling. It is not clear whether it is the total amount of UV exposure or the pattern in which individuals receive UV irradiation that leads to development of melanoma. It is reported that people incurring severe burns in childhood appear to be at higher risk for development of melanoma years later. In contrast, those who receive exposure on a regular basis may not be at as high a risk. There is also a role for skin type since individuals who tan easily are not at as high a risk for the development of melanoma even with prolonged exposure. The highest risk population appears to be those individuals with a fair complexion who receive intermittent doses of radiation resulting in severe sunburns. Additional factors that increase the risk for development of melanoma include dysplastic nevus (DN) syndrome, xeroderma pigmentosum, a history of nonmelanoma skin cancer (NMSC), and a family history of melanoma. The risk of melanoma increases with age; however, the role of aging is not clear. With increasing age there is more opportunity for the initiation of new tumors, either through exposure to carcinogens (UV irradiation) or through the decreasing ability of individual cells to repair DNA damage.[4]
Precursor Lesions and Risk Factors

Congenital nevi, DNs, Spitz nevi, and familial patterns all raise the risk of developing melanoma. Individuals with congenital nevi have an increased risk that is proportional to the size and number of nevi. Small congenital nevi represent a low risk and are therefore observed unless local changes appear. Giant congenital nevi are rare (1 in 20,000 newborns) and carry an increased risk for development of melanoma within the nevi ( Fig. 291 ). This lifetime risk has been estimated to be in the range of 5% to 8%, which has led some authors to recommend complete excision if possible. At minimum, these patients should be examined regularly throughout life. In general, a DN is a large (6- to 15-mm) pigmented flat skin lesion with indistinct margins and variable color. This simple definition belies the difficulty in making the diagnosis because precise criteria may vary both clinically and histologically. DNs may occur sporadically or in a familial pattern. Individuals with DNs and a family history of melanoma have an extremely high risk of developing melanoma. Patients with DN syndrome (B-K mole syndrome, familial atypical molemalignant melanoma syndrome) have multiple nevi (>100) that present a great challenge to the patient and physician. There is a lack of consensus regarding the management of DNs since it may

Figure 29-1 Giant congenital nevus of the trunk with a melanoma (arrow) arising on the lower back.

be difficult to follow changes in multiple nevi over time and most DNs will not become melanoma. Excision of all nevi might be reassuring for the patient and physician, but this still requires surveillance for the appearance of new lesions. In most instances, physician examinations are scheduled at 3to 6-month intervals in addition to monthly patient self-examinations. When multiple nevi are present, reference photographs provide an excellent way to compare the appearance of nevi over time.[5] Spitz nevi (juvenile melanoma, spindle cell melanoma, epithelioid cell melanoma) are rapidly growing, pink or brown, benign skin lesions arising most often in children and adolescents, although adult skin lesions may also have spitzoid features. They may be difficult to distinguish histologically from melanoma. Consultation with an experienced pathologist is often required to accurately diagnose these lesions. Complete local excision is the treatment of choice. In borderline cases, it may be necessary to excise the areas as a melanoma to ensure adequate treatment.[6] Sentinel lymph node (SLN) biopsy has been proposed as a mechanism to clarify the malignant potential in indeterminate cases.[7] If the diagnosis of melanoma arising within a Spitz nevus is made, the treatment should be based on the same criteria as other types of melanoma.
Familial Melanoma

Approximately 5% to 10% of melanoma patients have a family history of the disease. Compared to patients with sporadic melanoma, the age of onset is earlier, the incidence
783

of DNs is higher, and multiple primary melanomas are more common. Chromosome mapping studies have shown evidence of linkage and heterogeneity to chromosomes 1p and 9p. Chromosome 1p contributes to both sporadic and familial melanoma, whereas 9p contributes more to sporadic melanoma alone. All reported kindreds are white.
Clinical Features
Cutaneous Melanoma

Melanoma commonly presents as a changing pigmented skin lesion. Patients typically describe a flat lesion that spreads over the surface of the skin and later becomes elevated. If the lesion is allowed to progress, itching, bleeding and ulceration will follow. In some instances, melanomas arise in preexisting nevi; however, the majority arise de novo. The most important aspect of the history is change. Even experienced clinicians may not recognize a melanoma; therefore, physicians should have a low threshold for performing a diagnostic biopsy on any changing lesion. There are several other pigmented benign skin lesions that can mimic the appearance of melanoma: nevi (congenital and acquired), blue nevus, solar lentigo, keratosis, hemangioma, and pyogenic granuloma. The common features of melanoma are summarized in the mnemonic ABCDE: Asymmetrical outline, changing irregular Borders, variation in Color, Diameter greater than 6 mm, and Elevation. In early melanoma, the changes may be limited to two or three features. Not all melanomas are pigmented. These amelanotic lesions appear as raised papules that can be pink, red, purple, or of normal skin colors. Their atypical appearance frequently leads to a delay in diagnosis and therefore a poorer prognosis. Desmoplastic melanoma is a specific type of amelanotic melanoma that commonly arises on the face and can be associated with a lentigo maligna melanoma (LMM). Desmoplastic melanomas exhibit neurotropism, which is also a poor prognostic factor.[8] In summary, any changing skin lesion should be carefully evaluated, and clinicians should have a low threshold to perform a diagnostic biopsy. This is especially true in individuals who have multiple pigmented lesions, a history of atypical or dysplastic lesions, or a family history of skin cancer.
Unknown Primary Melanoma

Melanoma may present as a nodal or distant metastasis as the first evidence of the disease. This occurs in less than 2% of all melanoma cases and less than 5% of all patients who present with metastatic melanoma.[9] Under these circumstances, the prognosis may be as good or better than if the primary were present. A thorough search for the primary lesion should include a histologic review of all previously removed skin lesions; questions regarding skin lesions that resolved without treatment; and inspection of areas that may have been missed in initial examination, including the scalp, external auditory canal, oral and nasal mucosa, nail beds, genitalia, anal canal, perianal skin, and the eye. In the case of a lymph node metastasis, a completion

regional lymph node dissection is performed on the assumption that it is a regional node and therefore represents stage III, rather than stage IV, disease. [10] The patient should then be evaluated for adjuvant therapy, especially for participation in investigational protocols. For metastases at other sites see Surgical Considerations for Metastases, later in this chapter.
Noncutaneous Melanoma

In embryogenesis, melanocytes arise in the neural crest area and migrate to many sites other than the skin. Fewer than 10% of melanomas arise in these areas, which include the eye, mucosal surfaces, and unknown primary sites. A review from the National Cancer Database from 1985 to 1994 reports on a population of more than 80,000 melanomas: Whereas 91% were found in the skin, 5.2% occurred in the eye, 1.3% on mucosal surfaces, and 2.2% were considered of unknown primary origin.[9] Although melanoma has been reported to arise from many tissues and organs throughout the body, there is often the possibility that these are actually metastases from an unknown primary site on the skin. One exception may be in the esophagus, where melanocytic atypia and melanoma in situ have been shown to occur. Ocular melanoma is the most common malignancy arising in the eye. Within the eye, melanocytes are found in the retina and uveal tract (iris, ciliary body, and choroids). The options for treatment are photocoagulation, partial resection, radiation, or enucleation. Although ocular and cutaneous melanomas have several common histologic features, their clinical course is quite different. Ocular melanoma rarely metastasizes to lymph nodes since the uveal tract has no lymphatic vessels. The most common site of distant metastases is the liver, and this may be the presenting site of disease for patients with retinal melanoma. The most common sites of origin for melanomas arising on the mucous membranes are the head and neck (oral cavity, oropharynx, nasopharynx, and paranasal sinuses), anal canal, rectum, and the female genitalia. Compared to melanomas arising on the skin, mucosal melanomas are more advanced and have a uniformly poor prognosis. These tumors should be excised to negative margins. Extensive local resections do not affect survival, although locoregional control may be improved.[11] In general, lymph node dissections are not indicated unless patients have clinically evident lymphadenopathy. The one exception is for patients with vulvar melanomasSLN biopsy is now being performed for this group of patients.[12] The overall prognosis for patients with mucosal melanomas is poor, with less than 10% of patients surviving 5 years.
Clinical Management
Choice of Biopsy

The clinical management of melanoma begins with an accurate diagnosis. The classic signs of melanoma include
784

a skin lesion with changing characteristics such as irregular borders, varying degrees of pigmentation, an irregular surface, bleeding, itching, and ulceration. The decision to perform a biopsy is frequently based on clinical experience; however, even senior dermatologists and surgeons may underdiagnose melanoma. Concern on the part of either the physician or the patient is a valid indication for the biopsy. The specific method for the biopsy depends on the size of the lesion and anatomic location. Regardless of the method, biopsies should be full thickness into the subcutaneous tissues. For small lesions, an excisional biopsy is commonly performed that includes a narrow (1- to 2-mm) margin of surrounding skin. The biopsy area should not be enlarged to permit a better cosmetic appearance since this may lead to an unnecessary expansion of the final wide excision. Although shaved biopsies are commonly performed for benign-appearing lesions, this technique should not be used when melanoma is suspected. A shave biopsy may lead to a pathology report showing extension of the tumor to the deep margins of excision. Under these circumstances, the most important prognostic factor, tumor thickness, will not be accurate. This may lead to incorrect decisions regarding wide local excision (WLE), SLN biopsy, and adjuvant therapy. The clinical appearance of melanoma may be deceptively benign. This is why the use of cautery or cryoablation may lead to a delay in the diagnosis of melanoma. If a skin lesion appears at the site of a previously cauterized or frozen skin lesion, excisional biopsy and histopathologic analysis are mandatory. The technique of a surgical biopsy is straightforward. The biopsy removes the full thickness of skin, taking a layer of the underlying fatty tissue and all the visible tumor. Care should be taken not to crush or otherwise traumatize the specimen since this may hinder the histologic interpretation. These biopsies are performed under local anesthesia in the office or outpatient setting. The wound is closed in one or two layers in an orientation that is consistent with a possible wider excision. There may be circumstances in which complete surgical excision is not appropriate. This may be because of a large primary lesion or proximity of the lesion to important structures such as the eye, nose, or ear. Under these circumstances, a punch biopsy or excision of a segment of the lesion is appropriate. Again, this should be a full-thickness biopsy and include a margin of adjacent normal skin if possible. When a biopsy of a large lesion is performed, at least one punch should be placed through the most elevated portion to accurately classify the thickness. When performing a diagnostic biopsy, orientation of the biopsy closure may affect the options for closure of the WLE. For this reason, biopsy excisions on the extremities should be closed longitudinally to maximize the possibility for a primary wound closure and lower the use of skin grafts. Larger tissue defects may be closed with local rotational/advancement skin flaps or a skin graft.
Histologic Features of Cutaneous Melanoma

Histologically, melanoma is divided into four major types based on growth pattern and location. These forms are LMM, superficial spreading melanoma (SSM), acral lentiginous melanoma (ALM), and nodular melanoma (NM). Melanomas arise as proliferations of melanocytes in the basal layer of the skin. As they multiply, these cells expand radially in the epidermis and superficial dermal layer. With time, the growth begins in a vertical direction as the skin lesion may become palpable. NMs are an exception to this pattern wherein the vertical growth phase is present from an early point in tumor development. It is the vertical growth phase that more than any other histologic parameter of the primary tumor determines prognosis. LMM (10%) has distinctive clinical and histologic features. It occurs most commonly in older individuals with sun-damaged skin and presents as a flat, darkly pigmented lesion with irregular borders and a history of slow development. It is not uncommon to see LMMs that are several centimeters in diameter owing to the patients inability to detect slow progress of the lesion. ( Fig. 292 ). Overall, the prognosis of LMM is better than for the other histopathologic types; however, this is primarily related to the superficial nature of these lesions. The most common histologic type is SSM (70%). It is not necessarily associated with sun-exposed skin. As the name SSM suggests, these lesions initially appear as a flat, pigmented lesion growing in the radial pattern ( Fig. 293 ). If left in place, the lesion begins to thicken as it develops a vertical growth phase. ALM (5%) is classified principally by its anatomic site of origin, although it does have a characteristic histologic appearance. These tumors are confined to the subungual

Figure 29-2 Lentigo maligna melanoma covering the left side of the neck. Despite a long history of growth, the tumor thickness remained less than 1 mm.

785

Figure 29-3 Superficial spreading melanoma. The 2-cm-diameter melanoma developed over a 2-year period.

areas and the glabrous skin of the palms and soles ( Fig. 294 ). This is the most common histologic variant arising in blacks, and the diagnosis is often delayed because of the common appearance of irregular benign pigmentation on the surfaces. For this reason, the overall prognosis is poor. Melanomas arising in subungual areas also are frequently ignored because of their similar appearance to subungual hematomas secondary to trauma. The histologic appearance of ALMs is similar to melanomas arising on the mucous membranes. NMs (15%) develop a vertical growth pattern early in their history and may be devoid of junctional changes ( Fig. 295 ). Melanomas in this group have the worst prognosis based on a higher average tumor thickness. Historically, the classification of melanoma into various histologic types had a role in clinical management. With improved understanding of prognostic factors, management is based primarily on thickness and ulceration.
Prognostic Factors

Until the 1960s, invasive melanoma was considered to be a high-risk disease that required an extensive local excision for all tumors. In 1969, Clark and associates[13] described a classification of melanoma based on the extent of tumor invasion relative to the anatomic layers of the skin and showed that this level of invasion was related to survival ( Fig. 296 ). Level 1 tumors are limited to the epidermis, are in situ, and theoretically have no risk of metastasis. Level 2 lesions extend into the papillary dermis and also have an excellent prognosis. Clark level 3 tumors fill the papillary dermis and are associated with a significant risk of tumor metastases. Extension into the reticular dermis defines a Clark 4 lesion and growth into the subcutaneous fat characterizes a Clark level 5, both of which signify a high mortality risk. In some cases, determining the Clark level was found to be difficult, and readings of the same slides could differ between pathologists

Figure 29-4 Acral lentiginous melanoma. The extensively pigmented areas on the sole were predominately melanoma in situ. The single invasive area was ulcerated.

Figure 29-5 Nodular melanoma. This raised lesion had no radial growth phase.

(especially in the level 3 to 4 range). In 1970, Breslow described a more straightforward system based on measuring the vertical thickness of the tumor in millimeters.[14] This was found to be accurately reproducible between pathologists, and there was an excellent correlation with
786

Figure 29-6 Schematic representation of Clarks levels of tumor penetration in relation to the normal layers of the skin.

Figure 29-7 Observed (solid squares) and predicted (solid line) 10-year mortality rate for patients with clinically localized melanoma. This is based on a mathematical model derived from the American Joint Committee on Cancer melanoma database of 15,230 patients. (From Balch CM, Soong S-j, Gerschenwald JE, et al: Prognostic factors analysis of 17,600 melanoma patients: Validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol 19:36223634, 2001.)

5-year survival. Prognosis worsens with increasing thickness as a continuous logarithmic function without stair-step areas or natural breakpoints ( Fig. 297 ). The mortality rate begins to plateau at about 8 mm and never reaches 100%. Comparison of the two systems showed that the Clark level added little to the prognosis as determined by the Breslow thickness. In melanoma, analysis of factors contributing to prognosis has led to a remarkably accurate prediction of outcome. Balch and collaborators from 13 institutions [15] collected a complete data set on 17,600 patients to form the American Joint Committee on Cancer (AJCC) Melanoma Database. In an analysis of 13,581 patients with localized melanoma, they defined the relative contribution of multiple known prognostic factors including age, gender, level, site, thickness, and ulceration. The findings of their multifactorial analysis are summarized in Table 291 . As in all previous studies, tumor thickness was found to be the strongest predictor of outcome. Melanoma thickness is also associated with an increasing risk of local recurrence, regional metastases, distant metastases, and survival. The findings were the same at participating institutions in Australia, Europe, and North America. Based on the findings just discussed, a complete pathologic report of a cutaneous melanoma should include the following: Breslow thickness, presence or absence of ulceration, Clark level, status of the surgical margins, histologic type, presence or absence of satellitosis, and presence or absence of regression. The report may also describe tumor-infiltrating lymphocytes, lymphovascular invasion, vertical growth phase, neurotropism, and mitotic rate.
Staging

Staging for melanoma uses the tumor-node-metastasis (TNM) system of classification as defined by the AJCC staging system for cutaneous melanoma

(Box 291 (Box Not Available) ). The 6th edition (2002) of the AJCC Cancer Staging Manual contains important changes from the 1977 edition.[16] These changes are based on the in-depth prognostic factors analysis cited earlier.[15] Changes in the 2002 AJCC staging system for cutaneous melanoma include the following:
787

TABLE 29-1 -- Cox Regression Analysis for 13,581 Patients With Melanoma Without Evidence of Nodal or Distant Metastases Variable Thickness Ulceration Age Site Level Gender DF 1 1 1 1 1 1 Chi-Square Value 244.3 189.5 45.6 41.0 32.7 15.1 P <0.00001 <0.00001 <0.00001 <0.00001 <0.00001 0.001 Risk Ratio 1.558 1.901 1.101 1.338 1.214 0.836 95% CL 1.4731.647 1.7351.083 1.0711.132 1.2241.463 1.1361.297 0.7640.915

DF, degrees of freedom; CL, confidence limits. Revised from Balch CM, Soong S-j, Gerschenwald JE, et al: Prognostic factors analysis of 17,600 melanoma patients: Validation of the American Joint Committee on Cancer Melanoma Staging System. J Clin Oncol 19:36223634, 2001.

Box 29-1. American Joint Committee on Cancer TNM Melanoma Classification2002 (Not Available)

788

Clarks level of invasion is now used only for defining T1 melanomas. The thickness categories are now simplified to change at 1.0, 2.0, and 4.0 mm. Ulceration is included in all T stages and classified as (a) without ulceration and (b) with ulceration. Satellite metastases are now included in the N category. Thick melanomas larger than 4 mm are now staged as IIc. The dimensions of lymph nodes are no longer included. N stage is determined by number of positive lymph nodes. N staging includes the size of metastases within the node to account for microscopic disease found at SLN biopsy. Lung metastases are defined as a separate category of M1 disease because they are associated with a longer survival than other visceral metastases. 10. SLN biopsy results are included in the staging ( Table 292 ). As is true for other malignancies, the patient presenting with melanoma should undergo a systematic evaluation for metastatic disease. This begins with a history focused on constitutional, central nervous system, pulmonary, gastrointestinal, and soft tissue symptoms. A standard physical examination includes a detailed inspection and palpation of the skin and subcutaneous tissues to TABLE 29-2 -- American Joint Committe on Cancer Melanoma Stage Classification Grouping Pathologic Stage 0 IA IB IIA IIB IIC IIIA IIIB Tis T1a T1b T2a T2b T3a T3b T4a T4b T14a T14a T14b T14b T14a T14a T14a/b IIIC T14b Tumor N0 N0 N0 N0 N0 N0 N0 N0 N0 N1a N2a N1a N2a N1b N2b N2c N1b Node Metastasis M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0

1. 2. 3. 4. 5. 6. 7. 8. 9.

T14b Any T IV Any T

N2b N3 Any N

M0 M0 M1

detect satellites, in-transit metastases, other primary tumors, and lymph node enlargement. When present, all symptoms and signs of metastasis require further radiologic evaluation. Clinical stages 0 and I patients do not require any further tests. Stages II and III patients may have a chest radiograph and serum lactate dehydrogenase level determined; however, these are rarely abnormal in the asymptomatic patient. This 2002 version of the AJCC staging system provides excellent separation of prognostic groups by stage as is shown in Figure 298. The presence of ulceration indicates a significantly worse prognosis and can result in a change in stage ( Table 293 ). The new system also provides useful substaging within the N categories based on the number of positive lymph nodes and the presence or absence of ulceration as is shown in Table 294 .
Surgical Management of the Primary Lesion

The fundamental principle in the management of primary melanoma is to resect the tumor and minimize the risk of local recurrence. Historically, the resection of the primary site and surrounding skin was based on recommendations proffered by William Sampson Handley in 1907. From observations made on the autopsy of a single patient with locally advanced melanoma, he recommended WLE and regional lymph node dissection. This became the standard treatment except in anatomic locations where major adjacent structures (especially head and neck) were spared. With the insightful contributions of Wallace Clark[13] and Alexander Breslow[14] in the late 1960s, the natural history of melanoma became better understood. From many retrospective studies, it was clear that the risk of local recurrence and overall survival rates were related to tumor thickness. Four randomized studies were carried out to test whether narrow margins of excision could achieve the same results as wide margins. The current guidelines for WLE ( Table 295 ) are based on these studies. The first trial, published in 1991, was the World Health Organization Melanoma Study[17] comparing WLE using a 1cm margin versus a 3-cm margin in patients with primary tumors less than 2 mm in thickness.[18] This trial included 612 patients, and all local recurrences occurred in the group of patients undergoing a 1-cm radius of excision for tumors measuring 1.1 to 2 mm in thickness. The overall survival for all major groups and subgroups showed no differences. These findings confirm that melanomas measuring 1 mm or less in diameter can be resected with a 1-cm margin with a low subsequent risk for local recurrence. Melanomas between 1 and 2 mm in thickness have an equally low risk of local recurrence when a 2-cm margin is used. These margins may be lowered to 1 cm when this change facilitates primary closure of the wound. A narrower margin may result in a small increase in the number of patients who develop local recurrence but no difference in overall survival rates. The Melanoma Intergroup Trial compared margins of 2 versus 4 cm for patients whose tumors measured 1 to 4 mm in thickness.[19] This prospective trial randomized 462 patients with melanomas of the trunk or proximal
789

Figure 29-8 Fifteen-year survival curves for the melanoma staging system, comparing localized melanoma (stages I and II), regional metastases (stage III), and distant metastases (stage IV). The numbers in parentheses are the numbers of patients from the American Joint Committee on Cancer melanoma staging database used to calculate the survival rates. The differences between the curves are highly significant (P < 0.0001). (From Balch CM, Buzaid AC, Soong S-j, et al: Final version of the American Joint Committee on Cancer Staging System for cutaneous melanoma. J Clin Oncol 19:36353648, 2001.)

TABLE 29-3 -- Ten-Year Survival Rates for Stages I and II Melanomas Stage IA IB IIA IIB IIC Tumor Ulceration No Yes No Yes No Yes No Yes T-Stage T1a T1b T2a T2b T3a T3b T4a T4b Approximate 10-Year Survival (%) 90 80 80 65 65 50 55 35

TABLE 29-4 -- Five-Year Survival Rates for Stage III Melanoma Patients

Stage IIIA IIIB

Tumor Ulceration No No Yes Yes No No

N-Stage N1a N2a N1a N2a N1b N2b N1b N2b N3

Approximate 5-Year Survival (%) 70 60 55 50 55 45 30 25 30

IIIC

Yes Yes Yes or No

extremities to receive either a 2- or 4-cm radius of excision. After a median follow-up of 10 years the incidence of local recurrence was the same for both groups (2.1% vs. 2.6%). The factor that most closely correlated with the appearance of a local recurrence was primary tumor ulceration. The conclusion from this trial was that all patients with tumors 1 to 4 mm in thickness should undergo WLE with a 2-cm margin. TABLE 29-5 -- Recommended Margins for Surgical Resection of Primary Melanoma Tumor Thickness (mm) In situ <1.0 12 >2.0
* Recommended margins may be adjusted to accommodate anatomic or cosmetic circumstances.

Margin Radius (cm) * 0.5 1.0 2.0 2.0

The Swedish Melanoma Trial[20] and the French Melanoma Trial[21] compared 2-cm versus 5-cm margins for tumors less than 2-mm thickness. Both trials show no significant differences in disease-free or overall survival between treatment groups. When melanomas are greater than 4 mm in thickness, recommendations for management are based on retrospective analyses in which there does not appear to be any advantage to extending the resection beyond 2 cm.[22] The members of the Melanoma Committee of the National Comprehensive Cancer Network, a consortium of oncologists from National Cancer Institutedesignated cancer centers, annually update their consensus-based guidelines for the treatment of cancer. The guidelines for the management of primary melanoma are summarized in Figure 299 . (The complete guidelines algorithm is available on-line). The operative procedure of WLE is often performed under local anesthesia using intravenous sedation if necessary. In most cases, the margin for a WLE is measured from the edges of the biopsy scar. This again emphasizes the importance of a minimal excision for the original biopsy to limit the size of the final resection. The incision is made through the skin and subcutaneous tissues to the level of the superficial fascia. The specimen is oriented for
790

Figure 29-9 Management algorithm for primary cutaneous melanoma derived from the National Comprehensive Cancer Network guidelines (www.nccn.org). LDH, lactate dehydrogenase.

the pathologist and submitted for permanent section. In many cases, the resulting wound can be closed by elevation and advancement of skin edges or the use of local skin flaps. Skin grafts may be required on the hands, feet, and distal extremities. Tumors arising in proximity to structures such as the nose, eye, and ear may require a compromise of the conventional margins to avoid deformities or disabilities. Subungual melanomas are treated with amputation of the distal digit to provide 1 cm of margin from the tumor. For fingers, this commonly involves only the distal phalanx; ray amputations

are unnecessary. In all cases, resection should reach histologically normal margins. Most mucosal melanomas are locally extensive before becoming symptomatic. Oral cavity melanomas are an exception to this rule because they may be discovered during routine dental examinations. The diagnosis can be delayed in this area because of similarities to amalgam stains. Tumors should be resected with histologically clear margins; however, there is no evidence that WLE increases the chance for cure. Anorectal melanomas are excised to clear margins. For extensive tumors, abdominoperineal resection may be necessary. Abdominoperineal resection reduces the incidence of local and regional recurrence but does not result in an improvement in overall survival. Radiation therapy has been recommended for patients with head and neck melanomas as well as mucosal melanomas in the pelvic region. Retrospective analyses suggest that this will reduce the incidence of local recurrence.[23]
Melanoma and Pregnancy

Early reports suggested an adverse relationship between pregnancy and outcome in patients with melanoma. This was reinforced by the finding of estrogen receptors in some melanoma tumors. More recent comprehensive analyses have not confirmed any differences in the course of the disease in gravid versus nongravid patients when all relevant prognostic factors are taken in account. Unfortunately, recommendations derived from these early reports included early termination of pregnancy when the diagnosis was made and delaying pregnancy for 2 years after treatment for melanoma. The data do support these recommendations. The decision regarding pregnancy is no different in melanoma than other malignancies. These decisions should be made between the patient and her physicians after an in-depth discussion of prognosis and options for treatment.
Management of Regional Lymph Nodes

After WLE of the primary tumor, the most common sites of first recurrence are regional (lymph nodes, in-transit metastases, and local recurrences). Nodal metastases generally appear in the basin(s) draining from the primary site. This is a predictable pattern for extremity melanomas; however, truncal and head and neck melanomas may drain to more than one site. The lines of drainage for truncal melanomas are divided by the midline and the line of Sappey that extends from the umbilicus across the iliac crest and around to the spine at the level
791

of L2. The sequence of recurrence led surgeons to conclude that resection of nodal basins containing occult metastases could provide an increase in survival. This procedure, termed elective lymph node dissection (ELND), was commonly practiced but was often accompanied by significant morbidity including lymphedema, muscle weakness, and restricted range of motion. As prognostic factors became better understood, it was postulated that patients with thin tumors (<1-mm thickness) would have a low risk of metastases at any site and patients with thick tumors (>4-mm thickness) had a high risk of distant as well as regional metastases. In contrast, patients with intermediate-thickness melanoma (1 to 4 mm) would have an elevated risk of nodal metastases without a high risk of distant disease. The intermediate-thickness group formed the population of patients who would, in theory, benefit from ELND. Early retrospective analyses supported this hypothesis and provided the rationale for prospective, randomized trials comparing WLE alone versus WLE with ELND. Subsequent larger retrospective series reported no benefit.[24] [25] Four phase III prospective, randomized trials have failed to provide convincing evidence to support ELND. Two early trials were criticized for being underpowered and uncontrolled for important prognostic factors. The Intergroup Melanoma Trial accessioned 740 patients with well-balanced treatment groups but did not show a survival benefit for the ELND population. In subgroup analysis, however, patients with tumors 1 to 2 mm in thickness were found to have a survival benefit.[26] The Intergroup investigators defend the criticism of subgroup analysis by reason of the detailed stratification used in their trial. A fourth trial, the World Health Organization Melanoma Programme Trial of melanomas arising on the trunk, randomized 240 patients with tumor greater than 1.5 mm in thickness to WLE or WLE plus ELND. The 5-year survival rates were 51.3% and 61.7% (P = 0.09), respectively.[17] In this trial, patients with occult nodal metastases did have a statistically significantly longer 5-year survival, suggesting that this may be true for SLN biopsy patients (see later). The development of the SLN concept ended one debate over ELND, changed clinical management, and opened a new series of questions about the tumor biology of melanoma. In the mid-1970s, Dr. Donald Morton and colleagues described a radionuclide mapping technique to define the lymphatic drainage area from the primary site on the skin. This was designed to answer a perplexing problem for surgeons planning elective lymphadenectomy to resect occult metastases, that is, the at-risk lymph nodes. Primary sites, especially on the trunk and head and neck, could potentially drain to multiple lymphatic basins. This technique used technetium 99mlabeled colloid, injected intradermally at the primary site, to flow through lymphatic vessels and was taken up in regional nodes. This simple outpatient procedure identified the lymphatic basin(s) to be resected. More than 15 years later, Dr. Mortons group utilized blue dye injected intradermally at the primary site to show that the first blue node in the regional lymphatic basin(s) was the node that would contain a metastasis if any tumor were present. This node was termed the SLN. The theory was tested by performing the sentinel node biopsy in conjunction with a complete regional dissection.[27] Early reports showed several important findings, as follows: 1. Using a combination of isotope lymphatic mapping, an intraoperative hand-held gamma probe, and intraoperative blue dye, the SLN could be identified in more than 95% of cases. 2. There was great anatomic variation resulting in drainage to multiple or uncommon sites.[28] [29] [30] 3. A detailed pathologic analysis of the node using step sections enabled detection of micrometastases that could be missed by the standard techniques.[31] 4. In most cases a positive sentinel node was the only positive node.[32] 5. There were no prognostic factors that accurately identified a subpopulation of SLN-positive patients not requiring completion lymphadenectomy.
[33] [34]

6. When regional nodal metastases appear after a negative SLN biopsy, in the majority of cases micrometastases can be found in the original SLN by further histologic sectioning and examination. Additional studies have confirmed that the hottest lymph node (most radioactive) is not always the positive sentinel node. For this reason it is recommended that all nodes with radioactive counts greater than 10% of the hottest node be resected for analysis.[35] [36] The details of the SLN biopsy process requires close communication between all members of the team (radiologist, pathologist, and surgeon). The lymphoscintigram can be scheduled on the afternoon before or the day of the operative procedure. Multiple intradermal injections of Tc-99 sulfur colloid (total dose 1 mCi) are made at the perimeter of the biopsy scar. All regional nodebearing areas are scanned under the gamma camera and the sites of uptake are labeled on the lymphoscintigram (and on the patients skin if necessary) ( Fig. 2910 ). In the operating room, a hand-held gamma probe is used to precisely localize the most radioactive areas. Prior to the skin prep, isosulfan blue (lymphozurin) dye is injected intradermally at the biopsy margins. During the procedure, a 2- to 3-cm incision is made over the previously identified area and the dissection is performed by blunt dissection until a dye-colored lymphatic vessel is seen ( Fig. 2911 ). This vessel is traced down to the blue node that is removed. Using a combination of inspection and the gamma probe, the wound is examined for all blue and/or hot nodes. The wound is also palpated because nodes obliterated with tumor may not take up blue dye or radioisotope. SLN biopsy has rapidly become the standard of care for patients with tumors greater than 1 mm in thickness to accurately stage the disease and provide guidance for treatment planning (see Fig. 299 ). [37] In a retrospective analysis, SLN biopsy appeared to provide a survival benefit compared to WLE

alone.[38] The therapeutic benefit is now being evaluated in the Multicenter Selective Lymphadenectomy Trial (MSLT) in which patients with melanomas greater than 1 mm in thickness were randomized to undergo WLE alone or WLE plus SLN biopsy. Patients with positive lymph nodes then underwent completion lymph node dissection. Accessions to this trial
792

Figure 29-10 Lymphoscintigram showing the lateral view of a patient with a primary melanoma of the back. Note the three parallel lymphatic vascular pathways leading to the axillary sentinel lymph nodes.

Figure 29-11 Operative view of a sentinel lymph node stained with isosulfan blue dye. Note two parallel afferent vessels leading the sentinel node.

were closed at the end of March 2002 after 2001 patients were enrolled. The primary outcome is melanoma-related death with secondary outcomes being disease-free survival, and local, regional, and distant recurrence rates. The trial was designed to answer three questions. 1. What is the false-negative rate for SLN biopsy? 2. Is there a survival benefit for patients with micrometastases diagnosed by SLN biopsy? 3. Is there a therapeutic benefit for all patients who undergo SLN biopsy compared to those who have WLE alone? In so doing, it addresses one of the most fundamental questions in oncology: Is there a therapeutic benefit from the diagnosis and treatment of early lymphatic metastases? If a benefit is observed, it challenges the modern theory of Fisher and others who maintain that cancer is systemic from the onset of the metastatic process. If the early detection and resection of metastatic disease makes no difference in the long-term outcome, SLN biopsy will remain an important procedure for the purpose of establishing a prognosis and identifying patients who are candidates for subsequent systemic therapy. The MSLT investigators are planning a second trial to test the value of completion lymph node dissection in patients who are SLN positive. A second large trial currently examining the value of SLN biopsy is the Sunbelt Melanoma Trial. In this ongoing study, all patients with melanoma greater than 1 mm in thickness undergo SLN biopsy. Patients whose SLN is positive for metastasis by hemotoxylin/eosin, immunohistochemistry (S100 and HMB-45), or reverse transcriptase polymerase chain reaction (tyronaise, MAGE1, MART3, gp100) may participate in further randomization of surgery (completion lymph node dissection) and/or adjuvant interferon therapy. The role of interferon in stage III patients with melanoma remains controversial. The Sunbelt Melanoma Trial will determine if interferon has a beneficial role to play in the treatment of stage III patients with a minimal burden of metastatic disease.
Monitoring of Patients After Surgical Therapy

After the primary treatment of melanoma, the pattern of recurrences is predictable based on the same factors used to estimate survival (tumor thickness, ulceration, and lymph node status). The risk of the first metastasis being at a distant site increases with thick primary tumors and resected regional positive nodes. Follow-up examinations should focus on the detection of treatable metastases. The most common sites of initial recurrence are local and regional. Patients should be informed about the common symptoms and signs of recurrence so that they can report important changes arising between scheduled examinations. These include local swelling, itching, new lesions in and beneath the skin, enlargement of lymph nodes, central nervous system changes, and pulmonary and gastrointestinal symptoms. The physical examination is the most important aspect of the return visit. A complete skin examination is performed with inspection and palpation of the primary site and skin surfaces leading to regional nodal basins. In-transit metastases may be palpable but not visible. The follow-up examination schedule should reflect the risk of developing a recurrence. Initially patients are seen at 3- to 6-month intervals until they have reached the 3-year anniversary. By this time 75% of patients who would ever develop a metastasis would have had that event occur. Annual examinations are scheduled thereafter. Patients with early melanoma, stage IA, are followed without radiologic or laboratory studies. For asymptomatic patients, a chest radiograph and serum lactate dehydrogenase tests may be performed at 6- to 12-month
793

intervals, although there is no evidence that the routine use of these follow-up tests results in a survival benefit. The routine use of screening computed tomographic (CT), magnetic resonance imaging, or positron emission tomographic (PET) scans has not been shown to be cost effective[39] and remains a

subject of investigation.[40] In stage III patients, PET scan results change treatment decisions in up to 20% of cases. Scans and other tests may be required for patients participating in clinical protocols.
Surgical Considerations for Metastases

Approximately 80% of patients who develop melanoma are cured of their disease. Recurrent disease appears locally, regionally, or systemically or in a combination of these sites.
Regional Nodal Recurrence

Regional nodal metastases are the most common site of first recurrence in patients who undergo WLE alone. When patients develop palpable lymph nodes, the most rapid form of diagnosis is through the use of fine-needle aspiration (FNA) performed during the office visit. If positive, complete resection of the nodal basin will control regional disease in a large proportion of patients. If the FNA is negative or insufficient, an excisional biopsy should be performed to verify the diagnosis. If nodes are positive, unfortunately the long-term survival is low. Even with a single palpable nodal metastasis, the 5-year survival rate is 40% to 50% (see Table 294 ). Prior to complete regional lymphadenectomy, a full metastatic work-up is performed. This includes CT scans of the head, chest, abdomen and pelvis, although these scans are normal in the majority of patients who are otherwise asymptomatic. The risk of further locoregional recurrence after complete lymph node basin dissection is increased in the presence of multiple positive nodes, especially those containing extracapsular extension. Postoperative irradiation of the involved areas has been advocated in some centers as a way to further reduce recurrences; however, this has not been tested in a prospective, randomized trial.
Local and Regional Recurrences

True local recurrence (N2c, stage III) is defined as tumor appearing in the skin or subcutaneous tissues within a 5-cm radius of the primary wide excision site ( Fig 2912 ). The factors that predict local recurrence are the same as those predicting overall survival. Local recurrence risk has been reported to be 0.2% for primary tumors less than 0.76 mm, 2% for those 0.76 to 1.49 mm, 6% for lesions 1.5 to 3.99 mm, and 13% for thick melanomas greater than 4 mm. Local recurrence is a poor prognostic sign: less than 20% of patients survive long term after local recurrence. The treatment of local recurrence is surgical resection. This should be performed to reach histologically clear margins. WLE guidelines for primary tumors do not apply for local recurrences.

Figure 29-12 Multiple local recurrences growing around the scar on the leg of a patient who had undergone wide local excision 4 years earlier.

Amputation for extensive local regional recurrence is seldom indicated. These patients have a high risk of having other distant metastases and therefore long-term disease-free survival is not achieved by resection. Occasionally patients have indolent locoregional disease for which amputation is indicated only after other attempts at locoregional tumor control have been unsuccessful. Patients with multiple recurrences on the limbs may be candidates for isolated hyperthermic limb perfusion (IHLP). This technique, introduced in the 1950s, uses cannulation of the principal extremity artery and vein, a tourniquet, and hyperthermic perfusion (40C) with L-phenylalanine mustard. Interleukin-2, tumor necrosis factor, and multiple other chemotherapeutic agents have also been utilized. Response rates exceed 80%, and complete responses are seen in 10% to 15% of patients. Unfortunately many of these complete responses are short-lived. Reperfusion of extremities can be performed for patients who have an excellent initial response.[41] Based on encouraging results using therapeutic IHLP, a randomized trial was designed to test the value of prophylactic perfusion in patients with high risk (>1.5-mm thickness) melanoma. More than 800 patients participated in this trial comparing WLE to WLE with IHLP. After more than 6 years median follow-up, there was no improvement in overall survival, although the number of in-transit metastases were reduced from 6.6% to 3.3%.[42] Therefore, IHLP is recommended only for patients with established multiple in-transit metastases.
794

TABLE 29-6 -- One-Year Survival Rates for Patients With Distant Metastases Stage M1a M1b M1c
Distant Metastases

Metastatic Site(s) Skin, subcutaneous tissues, lymph nodes Lung Other visceral sites

Approximate 1-Year Survival (%) 60 55 40

The most common sites of initial distant metastases are in the brain, lung, and liver and less commonly in the skin, bone, and other gastrointestinal tract sites. The prognosis varies significantly with the site of first metastases ( Table 296 ). In the majority of cases, metastases appear at multiple sites simultaneously. Under these circumstances, systemic therapy is indicated for palliation. Occasionally patients develop metastases that are apparently isolated to a single site. These patients should be evaluated for surgical resection because the long-term disease-free survival rate after metastasectomy

is reported in the range of 10% to 20%.[43] Patients being considered for resection of visceral metastases should undergo complete staging, including CT and PET scans. In general, the prognosis for metastases to distant sites is related to the number of metastases and the disease-free interval between primary therapy and recurrent disease. Highly selected patients may undergo excision of multiple intra-abdominal metastases with a favorable outcome.
[44]

For patients with isolated lung metastases, a period of observation (which may include chemotherapy or investigational protocols) has been recommended to determine if additional metastases are to appear in a short period (4 to 6 weeks). Pulmonary resection is then indicated for patients with no evidence of further recurrence. Melanoma is one of the most frequent tumors that metastasizes to the gastrointestinal tract. These metastases are commonly intramural lesions that may grow to form an intussusception leading to obstruction. Patients who undergo resection of visceral metastases from occult primary sites have been shown to have a better survival than when the primary site has been identified.[45] Isolated metastases may appear in the adrenal glands, which, if stable, are also appropriately treated by resection. Symptomatic skeletal metastases can be effectively palliated with radiation. Metastases resulting in fractures of weight-bearing bones require internal fixation prior to radiation. Melanoma patients also present with central nervous system metastases that are commonly multiple lesions at the time of diagnosis. At the time of autopsy, the majority of patients have central nervous system metastases. When single brain metastases cause symptoms, long-term favorable results have been obtained through surgical resection followed by radiation. The most successful form of radiation therapy is a stereotactic program (gamma knife).[46]
Systemic Treatment for Melanoma

Most of the increase in the incidence of melanoma comprises thin melanomas with an excellent prognosis. Unfortunately, the number of deaths from melanoma is also rising. Although melanoma has been reported to metastasize to almost any tissue site, the most common areas are lung, liver, bone, and brain. The most commonly used drug for systemic therapy is dacarbazine (DTIC) which has a response rate of 15% to 30%; however, complete responses are rare. A large number of clinical trials have investigated combinations of chemotherapy in an attempt to improve response rates and prolong survival. A doubling of the response rate has been observed with CVD (cisplatin, vinblastine, dacarbazine) combined with interferon-alfa or interleukin-2 or a combination of these two biologicals. Unfortunately, the increases in survival have been either nonsignificant or less than 6 months. A randomized trial comparing CVD to DTIC indicated a doubling of the response rate and no effect on overall survival. The combination of CVD + interferon + interleukin-2 (frequently called biochemotherapy) has a response rate of 50% and a complete response rate of 15%; however, several trials of this combination have not resulted in a significant prolongation in survival.[47] Stage IV patients are also candidates for investigational protocols using immunotherapy.[48] [49] It is postulated that stage IV patients who can undergo resection of all detectable disease will be the group of patients who have the best chance of responding to immunotherapy.
Adjuvant Systemic Therapy

Adjuvant systemic therapy has proven to be a distinct advance in the treatment of common cancers such as those arising in the breast and colon. Clinical investigators have been attempting to identify an effective adjuvant therapy for melanoma for more than 40 years and yet there is no treatment regimen that has shown a conclusive benefit. In the mid-1990s the U.S. Food and Drug Administration approved interferon-alfa-2b for adjuvant therapy in treating patients with nodal metastases or thick melanomas in whom the expected survival rate is less than 50%. This approval was based on results from a single trial showing a significant increase in disease-free and overall survival. Subsequent randomized trials of interferon therapy have failed to confirm the initial observation. An updated analyses of randomized trials[50] [51] in addition to a meta-analysis[52] do not show a consistent benefit for interferon adjuvant therapy. At the present time, stages IIc and III patients should be evaluated for and invited to participate in randomized clinical trials of adjuvant therapy, when available.
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/306.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

CUTANEOUS MALIGNANCIES: NONMELANOMA SKIN CANCER SCC and BCC are the most common types of malignant neoplasms in the world. Just as in melanoma, the incidence of these cancers is rising each year. The current
795

predictions are that one in five Americans will develop this disease during his or her lifetime. Fortunately mortality rates for NMSCs are falling, and this is attributed to early detection and effective treatment. Patients who develop any type of skin cancer should have long-term periodic surveillance. After the initial diagnosis of BCC or SCC, the risk of developing an additional skin cancer is estimated to be 35% in 3 years and 50% in 5 years. In addition, there is a risk of developing other common malignancies such as lung cancer.
Squamous Cell Carcinoma Epidemiology and Etiology

By some estimates more than 1 million people develop NMSC annually; however, accurate statistics are problematic for a disease that is often treated without a histologic diagnosis. Although BCC is the most common type of NMSC, SCC has a higher mortality rate. As is true with other types of skin cancer, the incidence of SCC is increasing. There is a disproportionate increasing risk for women compared to men. The causes of SCC include the following: sunlight, susceptible phenotype, and compromise of immunity, in addition to environmental conditions and diseases. Sunlight is thought to be the major causative factor because most SCCs occur on the sun-exposed surfaces of the head and neck. In susceptible individuals (fair skin, blonde hair, blue eyes), increasing sun exposure carries a growing risk to develop SCC. Individuals with dark complexions have a lower risk even with prolonged sun exposure. Specifically, UVB is thought to be the form of UV radiation causing this disease. Most of the evidence for UV radiation comes from population-based studies in Australia where individuals of Celtic origin moved to a geographic area resulting in higher sun exposure. The pattern of skin cancer appearing in this population indicated that exposure to UV radiation earlier in life was a major risk factor since individuals who moved to Australia after adolescence had a lower incidence of skin cancer than those who moved in childhood. The risk of skin cancer increases with occupational or recreational sun exposure, advancing age, and proximity to the equator. The amount of sun exposure is also proportional to the incidence of precursor skin changes to a SCC, namely nevi, atrophy, and actinic keratosis. It is postulated that UV radiation affects the skin in two ways that result in an increased incidence of SCC: There is a direct carcinogenic effect on frequently dividing keratinocytes in the basilar layer of the epidermis. Unrepaired mutations result in tumor promotion and growth. The second mechanism relates to the depression of the cutaneous immune surveillance response that in turn inhibits tumor rejection. The P53 tumor suppressor gene is mutated in more than 90% of SCCs.[4] Occupational and environmental exposure to arsenic, organic hydrocarbon, ionizing radiation, and cigarette smoke all have been associated with the increasing risk for SCCs. Genetic disorders including xeroderma pigmentosum and albinism are associated with increased risk of many types of skin cancer. Chronic conditions of the skin such as burn scars (Marjolins ulcer), draining sinuses, infections, and ulcers can predate the development of SCCs. Previously healed wounds that break down or chronic wounds that will not heal should be biopsied for the presence of SCC. Impaired immunity, especially cell-mediated immunity, is a well-established cause of SCCs of the skin. The largest population of chronically immunosuppressed patients are those undergoing organ transplantation ( Fig. 2913 ). Immunosuppressive drugs such as eosothyoprin, cyclosporine, and prednisone have been linked to a greater than 50% increase in the risk of SCC. Both the intensity of immunosuppression and the duration of therapy are associated with the risk of development of malignancies. After 10 years of immunosuppression, 10% of patients develop malignancies, and this increases to 40% risk after 20 years.[53] The conditions associated with acquired impaired cell-mediated immunity including lymphomas, leukemias, and autoimmune diseases all increase the risk of development of SCCs. Human papillomavirus, an infection associated with immunosuppression, is proposed as a causative factor of SCCs. Most SCCs begin with a proliferation of keratin cells in the basal layer of epidermis that appear as red or pink areas, clinically termed actinic keratoses (solar keratoses). Local symptoms may wax and wane over a period of many months. Lesions are scaling with an uneven surface and an erythematous base. Individual lesions are usually less than 1 cm in diameter and appear in chronically sun-damaged skin. The diagnosis is both clinical and histologic since actinic keratoses have many features in common with SCC in situ microscopically. The overall risk of malignant conversion to an invasive SCC is low and estimated to be in the range of 1 in 1000 per lesion per year.[54] When the reddened area begins to develop a plaque-like thickening, it is termed Bowens disease, which appears histologically as SCC in situ and may vary from small lesions less than 1 cm to large areas of the anogluteal region. Invasive SCCs are palpable scaling lesions that become ulcerated centrally and have elevated edges ( Fig. 2914 ).

Figure 29-13 Squamous cell carcinoma appearing as areas of thickened, red, scaling skin.

796

Figure 29-14 Multiple squamous cell carcinomas on the upper extremity of a patient 11 years after kidney transplantation.

These may be confused with keratoacanthoma, a benign lesion that can also thicken and ulcerate. Biopsy may be required to differentiate between these two conditions. Most SCCs can be treated locally with excellent results. (see treatment options). Recurrence is associated with tumor size, degree of differentiation, depth of invasion, perineural involvement, immune status of the patient, and anatomic site. Local recurrence is associated with increased risk for regional and distant metastases. The first site of metastases is usually in regional lymph nodes.
Basal Cell Carcinoma

In contrast to SCCs and actinic keratoses, there is no precursor skin lesion for BCCs. These lesions may have an appearance that varies from nodules in the skin to a large nonhealing sore with drainage and crusting. In comparison to SCCs, they have a slow growth rate, which can lead to a delay in diagnosis. BCCs grow in distinct patterns described as nodular, pigmented, cystic, and superficial. The nodular growth pattern is characterized by a well-defined, elevated lesion with a waxy appearance ( Fig. 2915 ). As the lesion grows, it develops pearly opalescent nodules along the margins. A central depression with umbilication is a classic sign. Distinct blood vessels (telangiectasia) may be seen across the surface of the tumor mass. Although most BCCs are pink or skin colored, they may also have shades of brown or black pigmentation, thereby mimicking a benign mole or melanoma. Cystic BCCs are less common but have a distinctive appearance. Their surface is translucent and may appear blue or gray and be confused with a blue

Figure 29-15 Nodular basal cell carcinoma.

Figure 29-16 Basal cell carcinoma simulating the appearance of psoriasis.

nevus. Superficial BCCs (20%) are more macular than other growth patterns and may extend over the surface of the skin in a multicentric pattern ( Fig. 2916 ). The center can ulcerate and the margins become ill defined. These lesions may appear very similar to psoriasis, tenia, or eczema. They may also be multiple pink or red, small slightly elevated lesions that pepper the skin. This is a more aggressive growth pattern that is associated with extension well beyond visible changes in the skin surface and can penetrate deep into the underlying subdermis. The white scarring varieties of this growth pattern are termed morpheaform. BCCs commonly infiltrate locally but rarely metastasize. Metastases are associated with advanced patient age and neglected large lesions. The primary site has often been resected on multiple occasions before metastases appear. The median survival time for patients with metastatic disease is less than 1 year.
Treatment Options for Squamous and Basal Cell Carcinoma

NMSC is staged by different criteria than melanoma. The T stage is determined by the largest diameter of the lesion
797

Box 29-2. American Joint Committee on Cancer System for Classification and Staging of Carcinomas of the Skin2002

Primary Tumor (T) TX T0 Tis T1 T2 T3 T4 Primary tumor cannot be assessed No evidence of primary tumor Carcinoma in situ Tumor 2 cm in greatest dimension Tumor >2 cm, but not >5 cm, in greatest dimension Tumor >5 cm in greatest dimension Tumor invades deep extradermal structures (i.e., cartilage, skeletal muscle, or bone) Regional lymph nodes cannot be assessed No regional lymph node metastasis Regional lymph node metastasis Distant metastasis cannot be assessed No distant metastasis Distant metastasis Tis T1 T2 T3 Stage III Stage IV T4 Any T Any T N0 N0 N0 N0 N0 N1 Any N M0 M0 M0 M0 M0 M0 M1

Regional Lymph Nodes (N) NX N0 N1 MX M0 M1 Stage Grouping Stage 0 Stage I Stage II

Distant Metastasis (M)

on the skin surface and invasion of extradermal structures ( Box 292 ).[16] The overall favorable prognosis and the fact that many patients develop multiple primary skin cancers make this staging system less useful in planning treatment compared to the melanoma staging system. Actinic keratoses and the precursor lesions of SCC are most often treated with cryotherapy; however, alternate treatments include topical 5-fluorouracil, electrodesiccation and curettage, CO2 laser, dermabrasion, and chemical peel. A tissue biopsy is indicated when the actinic keratosis is raised or recurrent after topical therapy. Since multiple techniques are available, the strategy for surgical treatment of SCCs and BCCs begins with an assessment for high-risk factors ( Table 297 ). Considerations include size, location, primary versus recurrent, histology, and individual patient factors. All appropriate options should be reviewed with the patient in addition to making a specific recommendation. Surgical resection techniques include a histopathologic analysis to define the margins of resection. In contrast, field therapies treat a generalized area but do not define the status of margins. These approaches include radiation therapy, cryosurgery, curettage, and electrodesiccation. TABLE 29-7 -- Nonmelanoma Skin Cancer: Risk Factors for Local Recurrence Based on Characteristics of the Primary Tumor Factor Location Trunk and extremities Forehead and neck Central face Borders Incidence Immunosuppression Prior radiation therapy/chronic inflammation Rapid growth rate Neurologic symptoms Differentiation Perineural/vascular invasion <20 mm <10 mm <6 mm Well defined Primary Negative Negative Negative Negative Well Negative 20 mm 10 mm 6 mm Poorly defined Recurrent Positive Positive Positive Positive Moderate or poorly Positive Low Risk High Risk

Modified from National Comprehensive Cancer Network Practice Guidelines in Oncology. www.nccn.org. Standard surgical excision is the preferred treatment for the majority of SCCs and BCCs.[55] This procedure is usually performed under local anesthesia. The margin for resection is not as well defined as in the treatment of melanoma. A minimum acceptable margin is one that is found to be histologically free of carcinoma. This commonly involves a 3- to 4-mm area of normal-appearing skin. The risk of local recurrence is less when wider margins are obtained, especially in the presence of micronodular, infiltrative, and morpheaform histologic patterns. Using these methods, a local cure rate should be greater than 90%. An alternative surgical approach is the use of Mohs micrographic excision (MME) in which there is a high rate of local tumor

control with the use of horizontal frozen sections. The high success rate of MME is attributed to examination of a greater proportion of the margin of excision in addition to mapping the precise location of any margins found to be positive. Excisions in positive areas continue until clear margins are obtained. MME is ideal under high-risk conditions and for anatomic areas where it is important to preserve as much tissue as possible such as around the eye, nose, mouth, and ear.[56] Although field therapy techniques (cryotherapy, topical fluorouracil, electrodesiccation) do not histologically define the margins of treatment, they may still be effective in local tumor control. Cryotherapy is best suited for small superficial lesions and can be expected to have local control rates of greater than 90%. Treated areas may heal slowly by secondary intention and leave pale scars. Radiation therapy is highly effective in the treatment of BCC and SCC, especially for preserving wide areas of skin in the head and neck region. Radiation is also useful in treating areas that are at high risk for recurrence after extensive surgical excision.

798

Uncommon Cutaneous Malignancies

Among the hundreds of specific types of skin conditions and tumors, there are four uncommon skin malignancies that are important for the general surgeon to understand and be prepared to manage. Cutaneous angiosarcoma is a rare, aggressive soft tissue sarcoma derived from blood or lymphatic endothelium. It is most often seen on the face and scalp of older white men. Angiosarcoma has also been observed as a consequence of chronic lymphedema following axillary dissection for breast cancer (Stewart-Treves syndrome). Angiosarcoma may also arise in irradiated tissues after intervals of 10 to 20 years. The typical presentation is a flat, painless, often pruritic macule or plaque with a red, blue, or purple color that develops into a mass and ulcerates if left in place. Histologically they are high grade and often multifocal with skip areas of normal-appearing skin. Compared to other sarcomas, there is a high incidence of lymph node metastases (15%). Treatment consists of resection with histologically negative margins and radiation therapy to the involved field. Lymph node dissection is indicated if adenopathy appears before distant metastases are identified. There is no consensus about the role of adjuvant chemotherapy. The 5-year survival rate is less than 40%. Dermatofibrosarcoma protuberans is a low-grade sarcoma arising from dermal fibroblasts. The lesion appears as a smooth nodule in or immediately beneath the skin (trunk 40% and head/neck 40%) in mid-adult life. Due to their slow growth, lesions are commonly 1 to 2 cm at diagnosis. The external appearance belies the true character because tumor cells frequently invade the underlying soft tissues, leading to incomplete excision and local recurrence. Treatment consists of WLE with 3- to 4-cm margins. Specimen orientation and pathologic analysis of margins are required. Distant metastases are uncommon and are preceded by two or more local recurrences. Radiation therapy has been used effectively after resection of recurrences. Extramammary Pagets disease (EMPD) is a rare form of adenocarcinoma arising from apocrine glands of the skin most commonly in the perianal area, vulva, and scrotum. The clinical appearance is that of an erythematous plaque but may also be white or depigmented with crusts and scaling. The size is variable from less than 1 cm to an entire area in the anogenital region. Since EMPD can share many clinical characteristics in common with eczema, bacterial and fungal infections, and nonspecific dermatitis, the diagnosis is often made by biopsy of lesions not responding to standard therapies. In the majority of cases EMPD is confined to the epidermis and is well controlled with excision. When invasion of the deeper structures appears, the disease becomes increasingly difficult to control and the mortality rate increases to about 50%. Since EMPD is also associated with an increased risk of simultaneous internal malignancies of the genitourinary and gastrointestinal tracts (40%), a complete work-up should include a survey of these locations. The standard treatment is surgical resection extending to histologically clear margins. This may require multiple procedures because the histologic changes are best seen on permanent section. Patients require close clinical follow-up because local recurrences are common.[57] Radiation therapy has been reported to reduce local recurrences after excision. Kaposis sarcoma, a low-grade soft tissue malignancy, arises from lymphatic vascular endothelial cells in the skin. The incidence is rising because it is most often seen in patients with acquired immunodeficiency syndrome (AIDS) and other immunosuppressed states such as organ transplantation. In patients with human immunodeficiency virus, human herpesvirus-8 (HHV-8) has been identified as the causative agent. There is also a classic variant seen on the lower extremities of older men of Eastern European and Mediterranean descent. The clinical picture is variable, beginning as asymptomatic purple to brown bruises and progressing to spots, plaques, or nodules on both lower extremities. Local symptoms appear late as the tumors become advanced. In AIDS patients, skin changes respond best to aggressive antiretroviral therapy. Symptomatic skin lesions can be treated with radiation therapy, intralesional injection of chemotherapeutic agents, cryotherapy, or excision. Merkel cell carcinoma, derived from neuroendocrine cells, is histologically indistinguishable from small cell carcinoma arising in the lung or any other site. The initial work-up should include a chest radiograph to rule out a pulmonary primary. From any site of origin, a small cell carcinoma is a highly malignant tumor with a propensity to spread locally and regionally to nodes and distant sites. In the skin it presents as a rapidly growing redblue nodule most frequently in the head and neck area of elderly individuals. The diagnosis is confirmed on biopsy, and the primary treatment is WLE (2 to 3 cm) with histologically confirmed negative margins. SLN biopsy has been used successfully to identify patients with occult regional lymphatic metastases (10% to 30%); however, there is no evidence that patients benefit other than by improved regional tumor control. Involved-field radiation has been shown to reduce the local recurrence rate, and some reports have suggested a survival benefit; however, all studies are too small and uncontrolled to draw definitive conclusions.[58] Although metastases may be responsive to chemotherapy, there is little evidence to support adjuvant systemic therapy. Overall, the prognosis is poor, with variable mortality rates of 55% to 79%.[59] There are many other forms of cutaneous malignancies and cutaneous conditions associated with malignancy. These are beyond the scope of this chapter; however, the important principles in the management of these entities, as follows, are the same as reviewed earlier: 1. 2. 3. 4. Clinicians should have a low threshold for biopsy of new or changing skin lesions. The diagnosis is made by biopsy and histologic analysis. If appropriate, surgical excision should be performed with histologically defined negative margins. Further treatment and follow-up schedules will be determined by the specific diagnosis.
www.mdconsult.com

Copyright 2004 Elsevier Inc. All rights reserved.

Bookmark URL: /das/book/view/41779730-2/1235/307.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

799

Selected References
Allen PJ, Coit DG: The surgical management of metastatic melanoma. Ann Surg Oncol 9:762770, 2002. In selected patients, surgical resection of distant melanoma metastases will result in long-term disease-free survival. This is a comprehensive review of indications and results. Balch CM, Soong S-j, Gerschenwald JE et al: Prognostic factors analysis of 17,600 melanoma patients: Validation of the American Joint Committee on Cancer Melanoma Staging System. J Clin Oncol 19:36223634, 2001. This detailed analysis of a large melanoma database provides the basis for the new 2002 melanoma staging system. Balch CM, Soong S-j, Smith T, et al: Long-term results of a prospective surgical trial comparing 2-cm versus 4-cm excision margins for 740 patients with 14-mm melanomas. Ann Surg Oncol 8:101108, 2001. This paper summarizes the results of the first randomized trial, balanced for all important prognostic factors, to test the hypothesis that elective lymph node dissection has a therapeutic benefit. No advantage was found. Chang AE, Karnell LH, Menck HR: The National Cancer Data Base report on cutaneous and noncutaneous melanoma: A summary of 84,836 cases from the past decade. Cancer 83:16641678, 1998. This patterns of care study from the National Cancer Data Base summarizes the results of the current practice patterns in the United States. Feldman AL, Alexander HR, Bartlett DL, et al: Management of extremity recurrences after complete responses to isolated limb perfusion in patients with melanoma. Ann Surg Oncol 6:562567, 1999. With standardization of techniques, isolated limb perfusion has become an effective tool in the management of metastatic melanoma in the small number of patients with disease limited to an extremity. This is an excellent summary with complete references. Hersey P: Adjuvant therapy for high-risk primary and resected metastatic melanoma. Int Med J 33:3343, 2003. This is a well-written summary of all interferon-alfa-2 trials and discussion of the controversy. National Comprehensive Cancer Network (NCCN) Practice Guidelines in Oncology. www.nccn.org. The NCCN updates these on-line consensus-based guidelines annually or more often whenever major clinical information becomes available to change practice recommendations. Guidelines are described for all major types of malignancies. A CD-ROM can be ordered free of charge from NCCN. Rosenberg SA: Progress in the development of immunotherapy for the treatment of patients with cancer. J Int Med 250:462475, 2001. The Rosenberg laboratory is a major center for immunotherapy research for the treatment of melanoma. Thompson, JF, Shaw HM: The prognosis of patients with thick primary melanomas: Is regional lymph node status relevant, and does removing positive regional nodes influence outcome [editorial]? Ann Surg Oncol 9:719722, 2002. Drs. Thompson and Shaw from the Sydney Melanoma Unit present a well-balanced discussion about the role of lymphatic metastases in patients with a high risk of disease.

Copyright 2004 Elsevier Inc. All rights reserved.

www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/308.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

References
1. Jemal 2. Rigel

A, Murray T, Samuels A, et al: Cancer Statistics, 2003. CA Cancer J Clin 53:526, 2003. DS, Carucci JA: Malignant melanoma: Prevention, early detection, and treatment in the 21st century. CA Cancer J Clin 50:215236, 2000. BA, Eller MS, Geller AC, et al: The pathogenesis of melanoma induced by ultraviolet radiation. N Engl J Med 340:13411347, 1999.

3. Gilchrest 4. Brash

DE, Ziegler A, Jonason AS, et al: Sunlight and sunburn in human skin cancer: p53, apoptosis, and tumor promotion. J Investig Dermatol Symp Proc 1:136142, 1996. MA, Fraser MC, Goldstein AM, et al: A natural history of melanomas and dysplastic nevi: An atlas of lesions in melanoma-prone families. Cancer 94:31923209, 2002. ME, Boyer JD, Stashower ME, et al: The surgical management of Spitz nevi. Dermatol Surg 28:10651069, 2002.

5. Tucker

6. Murphy 7. Su

LD, Fullen DR, Sondak VK, et al: Sentinel lymph node biopsy for patients with problematic spitzoid melanocytic lesions. Cancer 97:499507, 2003. MJ, Crotty KA, Thompson JF, et al: Desmoplastic and desmoplastic neurotropic melanoma: Experience with 280 patients. Cancer 83:11281135, 1998.

8. Quinn 9. Chang

AE, Karnell LH, Menck HR: The National Cancer Data Base report on cutaneous and noncutaneous melanoma: A summary of 84,836 cases from the past decade. Cancer 83:16641678, 1998. B, Stroebel W, Ellwanger U, et al: Metastatic melanoma of unknown primary origin shows prognostic similarities to regional metastatic melanoma: Cancer 80:60

10. Schlagenhauff

65, 1997.
11. Thibault

C, Sagar P, Nivatvongs S, et al: Anorectal melanomaan incurable disease? Dis Colon Rectum 40:661668, 1997.

12. Abramova L, Parekh J, Irvin WP Jr, et al: Sentinel node biopsy in vulvar and vaginal melanoma: Presentation of six cases and a literature review. Ann Surg Oncol 9:840846, 2002. 13. Clark

WH Jr, From L, Bernadino EA, et al: The histogenesis and biologic behavior of primary human malignant melanomas of the skin. Cancer Res 29:705727, 1969. A: Thickness, cross-sectional areas and depth of invasion in the prognosis of cutaneous melanoma. Ann Surg 172:902908, 1970.

14. Breslow 15. Balch

CM, Soong S-j, Gerschenwald JE, et al: Prognostic factors analysis of 17,600 melanoma patients: Validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol 19:36223634, 2001. FL, Page DL, Fleming ID, et al (eds): AJCC Cancer Staging Manual, 6th ed. New York, Springer-Verlag, 2002.

16. Greene

N, Morabito A, Santinami M, et al: Immediate or delayed dissection of regional nodes in patients with melanoma of the trunk: A randomized trial. WHO Melanoma Programme. Lancet 351:793796, 1998.

17. Cascinelli

800

18. Veronesi 19. Balch

U, Cascinelli N: Narrow excision (1-cm margin): A safe procedure for thin cutaneous melanoma. Arch Surg 126:438441, 1991.

CM, Soong S-j, Smith T, et al: Long-term results of a prospective surgical trial comparing 2-cm versus 4-cm excision margins for 740 patients with 14-mm melanomas. Ann Surg Oncol 8:101108, 2001.

20. Cohn-Cedermark

G, Rutquist LE, Andersson R, et al: Long-term results of a randomized study by the Swedish Melanoma Study Group on 2-cm versus 5-cm resection margins for patients with cutaneous melanoma with tumor thickness 0.82.0 mm. Cancer 89:14951501, 2000. P, Thomas A, Vuillermin E, et al: Wide versus narrow excision in thin (<2 mm) stage I primary cutaneous melanoma: Long-term results of a French multicentric prospective randomized trial of 319 patients [abstract]. Proc Am Assoc Clin Oncol 12:387, 1993. KM, Sussman JJ, Gershenwald JE, et al: Surgical margins and prognostic factors in patients with thick (>4 mm) primary melanoma. Ann Surg Oncol 5:322328, 1998.

21. Banzet

22. Heaton 23. Irvin

WP Jr, Bliss SA, Rice LW, et al: Malignant melanoma of the vagina and locoregional control: Radical surgery revisited. Gynecol Oncol 71:476480, 1998. CL Jr, Stidham KR, Ricci WM, et al: Surgical management of regional lymph nodes in patients with melanoma: Experience with 4682 patients. Ann Surg 219:120130,

24. Slingluff

1994.
25. Coates AS, Ingvar CI, Peterson-Schaefer K, et al: Elective lymph node dissection in patients with primary melanoma of the trunk and limbs treated at the Sydney Melanoma Unit from 1960 to 1991. J Am Coll Surg 180:402409, 1995. 26. Balch CM, Soong S-j, Ross MI, et al: Long-term results of a multi-institutional randomized trial comparing prognostic factors and surgical results for intermediate thickness melanomas (1.04.0 mm). Intergroup Melanoma Surgical Trial. Ann Surg Oncol 7:8797, 2000. 27. Kelley

MC, Ollila DW, Morton DL: Lymphatic mapping and sentinel lymphadenectomy for melanoma. Semin Surg Oncol 14:283290, 1998. JF, Uren RF, Shaw HM, et al: Location of sentinel lymph nodes in patients with cutaneous melanoma: New insights into lymphatic anatomy. J Am Coll Surg 189:195

28. Thompson

206, 1999.
29. Schmalbach

CE, Nussenbaum F, Rees RS, et al: Reliability of sentinel lymph node mapping with biopsy for head and neck cutaneous melanoma. Arch Otolaryngol Head Neck Surg 129:6165, 2003.

30. Chao 31. Clary

C, Wong SL, Edwards MJ, et al: Sentinel lymph node biopsy for head and neck melanomas. Ann Surg Oncol 10:2126, 2003.

BM, Brady MS, Lewis JJ, et al: Sentinel lymph node biopsy in the management of patients with primary cutaneous melanoma: Review of a large single-institutional experience with an emphasis on recurrence. Ann Surg 233:250258, 2001. C, Wong SL, Ross MI, et al: Patterns of early recurrence after sentinel lymph node biopsy for melanoma. Am J Surg 184:520524, 2002. KM, Wong SL, Edwards MJ, et al: Frequency of nonsentinel lymph node metastasis in melanoma. Ann Surg Oncol 9:137141, 2002.

32. Chao

33. McMasters 34. Reeves

ME, Delgado R, Busam KJ, et al: Prediction of nonsentinel lymph node status in melanoma. Ann Surg Oncol 10:2731, 2003. KM, Reintgen DS, Ross MI, et al: Sentinel lymph node biopsy for melanoma: How many radioactive nodes should be removed? Ann Surg Oncol 8:192197, 2001. KM, Reintgen DS, Ross MI, et al: Sentinel lymph node biopsy for melanoma: Controversy despite widespread agreement. J Clin Oncol 19:28512855, 2001.

35. McMasters 36. McMasters 37. Thompson

JF, Shaw HM: The prognosis of patients with thick primary melanomas: Is regional lymph node status relevant, and does removing positive regional nodes influence outcome [editorial]? Ann Surg Oncol 9:719722, 2002. S, Soong SJ, Ross MI, et al: Improved staging of node-negative patients with intermediate to thick melanomas (>1 mm) with the use of lymphatic mapping and sentinel lymph node biopsy. Ann Surg Oncol 8:766770, 2001. SM, Carroll LA, Johnson DL, et al: Positron emission tomography is superior to computed tomography for metastatic detection in melanoma patients. Ann Surg Oncol 9:646653, 2002. RS, Hill AD, Skehan SJ, et al: Positron emission tomography for staging and management of malignant melanoma. Br J Surg 89:389396, 2002.

38. Dessureault

39. Swetter

40. Prichard

41. Feldman AL, Alexander HR, Bartlett DL, et al: Management of extremity recurrences after complete responses to isolated limb perfusion in patients with melanoma. Ann Surg Oncol 6:562567, 1999. 42. Koops

HS, Vaglini M, Suciu S, et al: Prophylactic isolated limb perfusion for localized, high-risk limb melanoma: Results of a multicenter randomized phase III trial, European Organization for Research and Treatment of Cancer Malignant Melanoma Cooperative Group Protocol-18832, the World Health Organization Melanoma Program Trial-15, and the North American Perfusion Group Southwest Oncology Group-8593. J Clin Oncol 16:29062912, 1998.
43. Allen 44. Wood 45. Vijuk

PJ, Coit DG: The surgical management of metastatic melanoma. Ann Surg Oncol 9:762770, 2002. TF, DiFronzo LA, Rose DM, et al: Does complete resection of melanoma metastatic to solid intra-abdominal organs improve survival? Ann Surg Oncol 8:658662, 2001. G, Coates AS: Survival of patients with visceral metastatic melanoma from an occult primary lesion: A retrospective match cohort study. Ann Oncol 9:419422, 1998. JG, Margolin K: The treatment of brain metastases from malignant melanoma. Semin Oncol 29:518524, 2002.

46. Douglas 47. Eton

O, Legha SS, Bedikian AY, et al: Sequential biochemotherapy versus chemotherapy for metastatic melanoma: results from a phase III randomized trial. J Clin Oncol 20:20452052, 2002. ME, Wunderlich JR, Robbins PF, et al: Cancer regression and autoimmunity in patients after clonal repopulation with antitumor lymphocytes. Science 298:850854, 2002. SA: Progress in the development of immunotherapy for the treatment of patients with cancer. J Int Med 250:462475, 2001.

48. Dudley

49. Rosenberg 50. Hersey 51. Lens

P: Adjuvant therapy for high-risk primary and resected metastatic melanoma. Int Med J 33:3343, 2003.

MB, Dawes M: Interferon-alfa therapy for malignant melanoma: A systematic review of randomized controlled trials. J Clin Oncol 20:18181825, 2002. K, Hancock B, Fore M, et al: Interferon- as adjuvant therapy for melanoma: A meta-analysis of the randomized trials. Proc Am Soc Clin Oncol 20:1394, 2001.

52. Wheatley

53. Jensen P, Hansen S, Moller B, Leivestad T, et al: Skin cancer in kidney and heart transplant recipients with different long-term immunosuppressive therapy regimens. J Am Acad Dermatol 40:177186, 1999. 54. Fu

W, Cockerell CJ: The actinic (solar) keratosis. Arch Dermatol 139:6670, 2003. Comprehensive Cancer Network Practice Guidelines in Oncology. www.nccn.org.

55. National

801

56. Kuijpers DI, Thissen MR, Neumann MH: Basal cell carcinoma: Treatment options and prognosis, a scientific approach to a common malignancy. Am J Clin Dermatol 3:247259, 2002. 57. Pierie JP, Choudry U, Muzikansky A, et al: Prognosis and management of extramammary Pagets disease and the association with secondary malignancies. J Am Coll Surg 196:4550, 2003. 58. Medina-Franco H, Urist MM, Fiveash J, Heslin MJ, et al: Multimodality treatment of Merkel cell carcinoma: Case series and literature review of 1024 cases. Ann Surg Oncol 8:204208, 2001. 59. Goessling

W, McKee PH, Mayer RJ: Merkel cell carcinoma. J Clin Oncol 20:588598, 2002.

Copyright 2004 Elsevier Inc. All rights reserved.

www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/309.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

803

Chapter 30 - Soft Tissue Sarcomas and Bone Tumors

Soft Tissue Sarcomas
Samuel Singer M.D. Murray F. Brennan M.D. Soft tissue sarcomas are rare and unusual neoplasms, accounting for about 1% of adult human cancers and 15% of pediatric malignancies. However, sarcomas continue to carry biologic and clinical interest and significance disproportionate to their clinical frequency because of the often clearly defined molecular genetic basis and the challenges they pose in diagnosis and management. Although these tumors may develop in any anatomic site, approximately 50% occur in the extremities, followed in order of frequency by the abdominal cavity/retroperitoneum, trunk/thoracic region, and head and neck.[1] This section focuses on the biology and management of soft tissue sarcomas among adults (>16 years old). PREDISPOSING FACTORS AND MOLECULAR GENETICS In most patients, no specific etiologic agent is found. Multiple predisposing factors have been identified ( Box 301 ). Genetic syndromes such as neurofibromatosis, familial adenomatous polyposis, and the Li-Fraumeni syndrome have all been shown to be associated with the development of soft tissue sarcoma.[2] [3] [4] Ionizing radiation and lymphedema are well-established but uncommon antecedents to the development of soft tissue sarcoma.[5] [6] The association with trauma is uncertain as a true causal factor. Chemical carcinogens have also been widely implicated, but the data to support their association are not well founded.[7] [8] [9] Genetic alterations that play a role in the development of soft tissue sarcoma segregate into two major types. The first type consists of sarcomas with specific genetic alterations that include simple karyotypes, including fusion genes due to reciprocal translocations and specific point mutations such as KIT mutations in gastrointestinal stromal tumors and APC/-catenin mutations in desmoid tumors. The second type consists of sarcomas with nonspecific genetic alterations and typically complex unbalanced karyotypes, representing numerous genetic losses and gains. A significant subset of soft tissue sarcoma, as well as most types of adipocytic tumors, are characterized by specific chromosomal aberrations, most commonly reciprocal translocations, which can be diagnostically[10] [11] [12] [13] and occasionally prognostically useful ( Table 301 ). [14] [15] The fusion gene translocations include 11 different gene fusions involving the EWS gene or EWS family members (TLS, TAF2N) found in five different sarcomas and 10 other types of fusions in seven other sarcoma types.[16] If conventional cytogenetics is not available, molecular genetic techniques (e.g., reverse transcription polymerase chain reaction and fluorescence in situ hybridization) are very useful as diagnostic adjuncts. In addition, investigation of molecular changes of genes at the sites of chromosomal alterations has led to the identification of novel genes and the characterization of their mechanisms of deregulation. The tumor suppressor genes best studied in sarcomas are p53 and RB1. Inactivation of both genes is involved in the tumorigenesis of several sarcomas.
[12] [17] [18]

804

Box 30-1. Predisposing Factors for Sarcomas Genetic Predisposition Neurofibromatosis (von Recklinghausens disease) Li-Fraumeni syndrome Retinoblastoma Gardners syndrome (familial adenomatous polyposis) Radiation Exposure Ortho- and megavoltage therapeutic radiation Lymphedema Postsurgical Postirradiation Parasitic infection (filariasis) Trauma Post parturition

Extremity Chemical 2,3,7,8-Tetrachlorodibenzodioxin (TCDD) Polyvinyl chloride Hemachromatosis Arsenic

The relevance of the p53 gene to sarcoma tumorigenesis is underscored by the frequent occurrence of soft tissue sarcomas in the Li-Fraumeni syndrome; all families studied have p53 germline mutations. The major mechanisms of p53 pathway inactivation in sarcomas include p53 point mutations, homozygous deletion of CDKN2A, which encodes both p14ARF and p16, and MDM2 amplification. In sarcomas with specific reciprocal translocations,p53 pathway alteration is a rare event, but when present it is a strong prognostic factor, associated with significantly decreased survival in synovial sarcoma,[19] [20] myxoid liposarcoma,[21] and Ewings sarcoma/peripheral neuroectodermal tumor (PNET).[22] [23] Decreased survival in Ewings sarcoma/ PNET was associated with deletion of CDKN2A, representing a type of p53 pathway alteration through loss of the CDKN2A alternative product p14ARF.[24] [25] In contrast, in sarcomas with nonspecific genetic alterations and complex karyotypes, p53 pathway alteration is more common and has weaker prognostic value, often requiring large numbers of patients to achieve statistical significance, as demonstrated in several studies of mixed adult soft tissue sarcoma.[26] [27] Its high prevalence in this class of sarcomas may account for its limited ability to define distinct clinical prognostic subsets in these tumors. In addition to serving as very specific and powerful diagnostic markers, fusion genes resulting from translocations encode chimeric proteins that are important determinants of tumor biology, acting as abnormal transcription factors that alter the transcription of multiple TABLE 30-1 -- Cytogenetic and Molecular Abnormalities in Sarcomas Histologic Type Synovial sarcoma Myxoid/round cell liposarcoma Ewings sarcoma Cytogenetic Changes t(X;18)(p11.2;q11.2) t(12;16)(q13;q11) t(12;22)(q13;q1112) t(11;22)(q24;q12) t(21;22)(q22;q12) t(7;22)(p22;q12) t(17;22)(q12;q12) t(2;22)(q33;q12) Alveolar rhabdomyosarcoma Extraskeletal myxoid chondrosarcoma Dermatofibrosarcoma protuberans Desmoplastic small round cell tumor Clear cell sarcoma Infantile fibrosarcoma Alveolar soft part sarcoma Atypical lipomatous tumor/well-differentiated liposarcoma Leiomyosarcoma Malignant fibrous histiocytoma Malignant peripheral nerve sheath tumor t(2;13)(q35;q14) t(1;13)(p36;q14) t(9;22)(q22;q12) t(17;22)(q22;q13) t(11;22)(p13;q12) t(12;22)(q13;q12) t(12;15)(p13;q25) 17q25 rearrangement 12q rings and giant markers complex complex t(11;22) (q24;q11.212) Gene Rearrangement/Molecular Abnormality SYT-SSX1 fusion SYT-SSX2 fusion CHOP-TLS fusion CHOP-EWS fusion FLI1-EWS fusion ERG-EWS fusion ETV1-EWS fusion EIAF-EWS fusion FEV-EWS fusion PAX3-FKHR fusion PAX7-FKHR fusion TEC-EWS fusion PDGFB-COL1A1 fusion WT1-EWS fusion ATF1-EWS fusion ETV6-NTRK3 fusion Unknown HMGI-C, CDK4, and MDM2 amplification RB1 point mutations or deletions p53 point mutations or deletions

805

downstream genes and pathways.[28] The structure of these chimeric proteins play a prominent role in the pathogenesis of sarcoma, as evidenced by the impact of relatively minor cytogenetic variability, as a result of variant molecular breakpoints, on tumor phenotype and clinical behavior.[15] [29] [30] [31] A recent analysis of synovial sarcoma has clearly identified a characteristic SYT-SSX fusion gene resulting from the chromosomal translocation t(x;18) (p11;q11) detectable in almost all synovial sarcomas. Translocation fuses the SYT gene from chromosome 18 to either of two highly homologous genes at Xp11, SSX1, or SSX2. SYT-SSX1 and SYT-SSX2 are thought to function in aberrant transcriptional regulation. Recent analysis has suggested that these fusion products may influence outcome. It does appear that all biphasic synovial sarcomas have an SYT-SSX1 fusion transcript, and tumors that were positive for SYT-SSX2 were monophasic. Conversely, monophasic sarcomas may have either transcript.[14]
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/310.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

PATHOLOGIC EVALUATION There are more than 50 histologic subtypes, many of which are associated with distinctive clinical, therapeutic, or prognostic features. Detailed descriptions of the histopathologic classification and guidelines for the histologic reporting of soft tissue sarcoma have been published elsewhere.[32] To summarize, the most commonly found are liposarcoma, malignant fibrous histiocytoma (MFH), and leiomyosarcoma ( Fig. 301 ). Histopathology is anatomic site dependent: the common subtypes in the extremity are liposarcoma or MFH; in the retroperitoneal location liposarcoma and leiomyosarcoma are the most common histiotypes, whereas in the visceral location, gastrointestinal stromal tumors are found almost exclusively ( Fig. 302 ). Age is also a factor in histopathology. In childhood, embryonal rhabdomyosarcoma is most common; synovial sarcoma is more likely to be seen in young adults (<35 years old); and there is an even distribution of liposarcoma and MFH as the predominant types in the older population ( Fig. 30 3 ). Sarcoma histiotype is generally an important determinant of prognosis and a predictor of distinctive patterns of behavior, because none of the existing grading systems is ideal and applicable to all tumor types. Biologic behavior is currently best predicted based on histologic type, histologic grade, tumor size, and depth. Although many published series have combined all the histologic types of sarcoma, the significance of such subtyping is exemplified by liposarcoma in which the five subsets (well differentiated, dedifferentiated, myxoid, round cell, and pleomorphic) have totally different biologies and patterns of behavior.[32] A further clear demonstration is the importance of myogenic differentiation in pleomorphic sarcomas, which is associated with a substantially increased risk of metastasis.[33] In a postoperative

Figure 30-1 Histopathologic subtype distribution of 5069 patients with soft tissue sarcoma treated at Memorial Sloan-Kettering Cancer Center from July 1, 1982, through June 30, 2002. These data include extremity, trunk, visceral, and retroperitoneal tumors. MFH, malignant fibrous histiocytoma; MPNT, malignant peripheral nerve tumor.

Figure 30-2 Site-specific histopathologic subtype distribution of 3336 patients with soft tissue sarcoma treated at Memorial Sloan-Kettering Cancer Center from July 1, 1982, through June 30, 2002. MFH, malignant fibrous histiocytoma; MPNT, malignant peripheral nerve tumor.

806

Figure 30-3 Distribution by age and diagnosis for patients with fibrosarcoma (n = 541), leiomyosarcoma (n = 884), liposarcoma (n = 924), and malignant fibrous histiocytoma (n = 922) seen at Memorial Sloan-Kettering Cancer Center from July 1, 1982, through June 30, 2002.

nomogram based on a database of 2136 adult patients from MSKCC, histologic type was found to be one of the most important predictors of sarcomaspecific death, with malignant peripheral nerve sheath tumors having the highest risk of mortality.[34]
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/311.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

CLINICAL EVALUATION AND DIAGNOSIS Patients with extremity sarcoma usually present with a painless mass, although pain is noted at presentation in up to 33% of patients. Delay in diagnosis is common, with the most common differential diagnosis for extremity and trunk lesions being a hematoma or a pulled muscle. Physical examination should include assessment of the size of the mass and its relationship to neurovascular and bony structures. Generally, in an adult, any soft tissue mass that is symptomatic or enlarging, any mass that is larger than 5 cm, or any new mass that persists beyond 4 weeks should be sampled.[35] Biopsy technique is important. For most soft tissue masses, an incisional or core biopsy is usually preferred. Ideally, the initial diagnostic procedure should be performed at a center where the patient will be treated. This facilitates proper placement of the biopsy site (or incision) and also avoids the complications and diagnostic difficulties that can arise if such biopsy samples are handled infrequently. Limb masses are generally best sampled through a longitudinal incision so that the entire biopsy tract can be excised at the time of definitive resection. The incision should be centered over the mass in its most superficial location. No tissue flap should be raised, and meticulous hemostasis should be ensured to prevent cellular dissemination by hematoma. Excisional biopsy is recommended only for small cutaneous or subcutaneous tumors, usually smaller than 3 cm, in which a wide re-excision (if required) is usually straightforward. Fine-needle aspiration biopsy has a limited role in diagnosing extremity soft tissue tumors but may be of value in the documentation of recurrence. An analysis of 164 soft tissue masses for the value of Tru-cut biopsy suggests that 83% of specimens obtained at initial biopsy are adequate for diagnosis. Of the adequate biopsy specimens, 95% correlated with the final resection diagnosis for malignancy, 88% for histologic grade, and 75% for histologic subtype, respectively. Tru-cut biopsy can be then advocated as the first step in the diagnostic armamentarium. The ease of performance, low cost, and low complication rate make this technique attractive. Should tissue be inadequate
807

or there be any indecision, then open linearly placed incisional biopsy is indicated. Biopsy should be indicated only if the actual treatment will be altered by a definitive diagnosis.[36] Tumor type and grade are correctly identified in most patients. Patients with intra-abdominal or retroperitoneal sarcomas often experience nonspecific abdominal discomfort and gastrointestinal symptoms before diagnosis.[37] The diagnosis is usually suspected on finding a soft tissue mass on abdominal computed tomography (CT) or magnetic resonance imaging (MRI). Fine-needle aspiration biopsy or CT-guided core biopsy has a limited role in the routine diagnostic evaluation of these patients. Needle or core biopsy is indicated if abdominal lymphoma is strongly suspected as part of the differential diagnosis. In most patients, exploratory laparotomy should be performed and the diagnosis made at operation, unless the patients tumor is clearly unresectable or the patient will be undergoing preoperative investigational treatment. In a recent analysis of 500 patients with retroperitoneal soft tissue sarcoma,[38] median survival was 72 months for patients with primary presentation, 28 months for those with local recurrence, and 10 months for those presenting with metastasis. For all patients with primary or locally recurrent tumors, complete resection of low-grade tumors was the more favorable factor.
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/312.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

EVALUATION OF EXTENT OF DISEASE All patients require a thorough history and physical examination. MRI examination is the usually preferred procedure for imaging extremity soft tissue masses.[39] [40] MRI enhances the contrast between tumor and adjacent structures and provides excellent three-dimensional definition of fascial planes. The relative value of MRI over CT has been the subject of a national study. The results of that study show there was no statistically significant difference between CT and MR imaging in determining tumor involvement of muscle, bone, joints, or neurovascular structures. The combined interpretation of CT and MR images did not significantly improve accuracy.[41] Once the diagnosis and grade are known, evaluation for sites of potential metastasis can be performed. Lymph node metastases occur in less than 3% of adult soft tissue sarcoma.[42] For extremity lesions, the lung is the principal site for metastasis of high-grade lesions[43] ; for visceral lesions, the liver is the principal site.[44] Thus, patients with low-grade extremity lesions require a chest radiograph, and the majority of those with high-grade lesions require a chest CT. Patients with visceral lesions should have their livers imaged as part of the initial abdominal CT or MRI. The authors do not generally perform angiography because this adds little that will change the management strategy.
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/313.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

STAGING Current staging systems focus on histologic grade of the tumor, size of the primary tumor, and presence or absence of metastasis.[45] [46] [47] Histologic grade is a major prognostic determinant and is based on degree of mitosis, cellularity, presence of necrosis, differentiation, and stromal content. Various grading systems exist, all of which should be considered as categories in a histologic spectrum. For therapeutic planning, the broad categories of low (I or II) and high (III or IV) grade suffice. Clearly such arbitrary decisions may be difficult, but they facilitate practical management of the patient. Lowgrade lesions are assumed to have a low (<15%) risk of subsequent metastasis, and high-grade lesions have a high (>50%) risk of subsequent metastasis. Size has historically been considered a less important determinant of biologic behavior, but large lesions can be associated with late recurrence. Unequivocal characterization of grade is difficult in large lesions, especially in tumors that can reach 2 or 3 kg. Conversely, very small, high-grade lesions less than 5 cm in maximal diameter have limited risk for metastatic disease if treated appropriately at the first encounter. The current staging systems for soft tissue sarcoma updated in 1992 and 1997 have been further updated more recently ( Table 302 ). The new staging system[48] addresses the issue of depth and size as independent variables. However, the staging systems continue to undergo evolution. Analysis of the primary extremity soft tissue sarcomas seen at Memorial SloanKettering Hospital from July 1, 1982 to June 30, 2002 suggests that the probability of metastasis by stage is better TABLE 30-2 -- Current Staging Systems for Soft Tissue Sarcoma G, HISTOLOGIC GRADE GX G1 G2 G3 G4 T, PRIMARY TUMOR SITE TX T0 T1 T1a T1b T2 T2a T2b N, REGIONAL NODES NX N0 N1 M, DISTANT METASTASIS MX M0 M1 STAGING GROUPING Stage I Stage II Stage III Stage IV G1-2 G3-4 G3-4 Any G Any G T1a, 1b, 2a, 2b T1a, 1b, 2a T2b Any T Any T N0 N0 N0 N1 N0 M0 M0 M0 M0 M1 Presence of distant metastasis cannot be assessed No distant metastasis Distant metastasis present Regional nodes cannot be assessed No regional lymph node metastasis Regional lymph node metastasis Primary size cannot be assessed No evidence of primary tumor Tumor less than 5 cm Superficial tumor Deep tumor Tumor 5 cm or greater Superficial tumor Deep tumor Grade cannot be assessed Well differentiated Moderately differentiated Poorly differentiated Undifferentiated

From Greene F, Page D, Fleming I, et al (eds): AJCC Cancer Staging Manual, 6th ed. Heidelberg, Springer-Verlag, 2002.

808

discriminated in the new AJCC 2002 staging system ( Table 303 ). The large, low-grade, and deep tumors in stage 2A of the AJCC 1997 system are now considered as stage 1 disease in the AJCC 2002 staging system. Figure 304 shows the excellent discrimination by stage for distant recurrence-free survival using the AJCC 2002 system. It should be important to emphasize under staging that staging systems (1) apply to risk of metastasis or diseasespecific TABLE 30-3 -- Primary Extremity Soft Tissue Sarcoma: Distant Metastases by Stage (n = 1410) * Total (n) OLD AJCC STAGING SYSTEM (1992) 1A 1B 136 252 2 (1%) 31 (12%) Distant Metastases (%)

3A 3B AJCC STAGING SYSTEM (1997) 1A 1B 2A 2B 2C 3A 3B AJCC STAGING SYSTEM (2002) 1 2 3

362 660 136 28 224 362 33 302 325 388 395 627

72 (20%) 274 (42%) 2 (1%) 3 (11%) 28 (13%) 72 (20%) 13 (40%) 105 (35%) 156 (48%) 33 (9%) 85 (22%) 261 (42%)

* From Memorial Sloan-Kettering Cancer Center, July 1, 1982-June 30, 2002. Excludes desmoid and dermatofibrosarcoma protuberans.

survival or overall survival and (2) are almost exclusively confined to extremity lesions. There is as yet no adequate staging system for retroperitoneal and visceral lesions.
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/314.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

MANAGEMENT Algorithms for management of soft tissue sarcomas are shown in Figures 305 and 306 . Surgical excision remains the dominant modality of curative therapy. Whenever possible, function- and limb-sparing procedures should be performed. As long as the entire tumor is removed, less radical procedures have not been demonstrated to adversely affect local recurrence or outcome.[49] The surgical objective should be complete removal of the tumor with negative margins and maximal preservation of function. When possible, tumors should be excised with 1 to 2 cm of normal tissue, because of the propensity for local, unappreciated spread. Conversely, deliberate sacrifice of major neurovascular structures can generally be avoided, provided the surgeon pays meticulous attention to dissection.[49] [50] Adjuvant radiation has been demonstrated to improve local control. This includes using either brachytherapy for high-grade lesions or external-beam radiation therapy for large (>5 cm) high- or low-grade lesions.[51] [52] For subcutaneous or intramuscular high-grade sarcoma smaller than 5 cm, or any size low-grade sarcoma, surgery alone should be considered if adequate wide excision with a good 1- to 2-cm cuff of surrounding fat and muscle can be achieved. If the excision margin is close, particularly with extramuscular involvement, or if a local recurrence would result in the sacrifice of a major neurovascular bundle or

Figure 30-4 Distant recurrence-free survival for patients with primary extremity soft tissue sarcoma (n = 1410) by American Joint Committee on Cancer 2002 stage seen at Memorial Sloan-Kettering Cancer Center from July 1, 1982, through June 30, 2002 (excludes desmoid and dermatofibrosarcoma protuberans).

809

Figure 30-5 Algorithm for management of primary (with no metastases) extremity or trunk soft tissue sarcoma, using a biologic rationale (i.e., size and grade of tumor). CT, computed tomography; MRI, magnetic resonance imaging; EBRT, external beam radiation therapy; BRT, brachytherapy.

amputation, then adjuvant radiation therapy should be added to the surgical resection to reduce the probability of local failure.[51] However, irrespective of grade, we believe that postoperative irradiation is probably used more than is strictly necessary. In fact, several studies have shown that a significant subset of subcutaneous and intramuscular sarcomas can be treated by wide excision alone, with a local recurrence rate of 5% to 10%.[53] [54] The value of chemotherapy depends on the histologic type of sarcoma. Neoadjuvant chemotherapy is usually indicated for the treatment of Ewings sarcoma (PNET) and rhabdomyosarcoma, because of the high risk of microscopic metastasis at diagnosis and high response rate seen to such therapy.[55] [56] The potential for cure is inversely proportional to the volume and spread of disease. For other histologic subtypes of sarcoma the role of chemotherapy remains controversial. Adjuvant chemotherapy has had no measurable impact on overall survival, with a small 10% to 15% improvement in disease-free survival.[57] [58] Thus, adjuvant chemotherapy for soft tissue sarcoma should be regarded as investigational and is rarely indicated, except in a clinical trial. The preoperative use of neoadjuvant combination chemotherapy (usually with doxorubicin [Adriamycin] and ifosfamide) may be justified in carefully selected high-risk patients with large, high-grade tumors.
810

Figure 30-6 Algorithm for management of primary retroperitoneal or visceral soft tissue sarcoma. Fine-needle aspiration biopsy is not routinely used. CT, computed tomography; XRT, x-ray therapy.

In retroperitoneal and visceral lesions, operation remains the dominant method of therapy,[38] [59] with the most important prognostic factors for survival being completeness of resection and grade. Despite an aggressive surgical approach local control is still a major problem, and multifocal, unresectable tumors recur in many patients, particularly those with liposarcoma. The role of radiation therapy for retroperitoneal sarcoma is not well defined and is in need of further investigation.[60] In theory, preoperative or postoperative irradiation to this site is desirable, but in reality it is often not possible to deliver full-dose radiation therapy (60 to 66 Gy) to areas at risk because the dose is limited by the large treatment volume required and sensitivities of adjacent normal tissues, such as bowel, kidney, liver, and spinal cord. Brachytherapy or intraoperative radiation therapy at the time of surgical resection may be used to treat a localized area at high risk of microscopic or gross residual disease when further surgical excision is not possible. However, care must be taken to avoid excessive morbidity and even increase in mortality that may result from aggressive brachytherapy, particularly when combined with external-beam radiation therapy.[61] Trials of preoperative radiation therapy with or without intraoperative irradiation are underway.
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/315.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

TREATMENT OF RECURRENT DISEASE Despite optimum multimodality limb-sparing treatment for extremity soft tissue sarcoma, a significant number of patients continue to develop distant metastasis. In a recent analysis of 994 patients with primary extremity soft tissue sarcomas and a median follow-up of 33 months, distant metastasis developed in 230 patients (23%). Median survival after the development of metastasis was 11.6 months, and in 73% of these patients the lungs were the first metastatic site. Analysis of this by multivariate analysis suggests that extent of metastatic disease and the length of the disease-free interval, the presence of a preceding local recurrence, and older age were all significant predictors of post-metastasis survival.[62] Local extremity recurrence presents as a nodular mass or series of nodules arising in the surgical scar. Patients with retroperitoneal recurrence usually present with nonspecific symptoms, often only after the lesion has reached a substantial size. After work-up to determine the extent of disease, patients with isolated local recurrence should undergo reresection. When re-resection can be performed, two thirds of these patients experience long-term survival benefit.[63] [64] Adjuvant radiation therapy should be administered after re-operation on the extremity, if feasible, dependent on the method and extent of previous radiation. For extremity lesions, the most common site of metastasis is the lung. It is the only site of recurrence in approximately half of all patients.[65] Extrapulmonary metastases are relatively uncommon and usually occur as a late manifestation of widely disseminated disease. Patients whose primary tumors are controlled or controllable, who have no extrathoracic disease, who are medically fit for thoracotomy, and in whom complete resection of all lung disease appears possible should undergo thoracotomy with the intent of resecting all disease.[62] [65] Patients with unresectable pulmonary metastases or extrapulmonary metastatic sarcoma in more than a single site have a uniformly poor prognosis and are best treated with systemic chemotherapy. The role of chemotherapy in advanced sarcoma is controversial[66] and, at present, the treatment of metastatic sarcoma represents palliative, not curative, therapy. Current active drugs that have significant response rates include doxorubicin, ifosfamide, and dacarbazine (DTIC), but none has had a major impact on long-term
811

survival.[67] The combination of mesna, ifosfamide, doxorubicin, and dacarbazine (MAID) has been shown to have a 47% response rate and a 10% complete response rate.[68] Randomized prospective clinical trials on combination chemotherapy regimens such as MAID and other ifosfamidedoxorubicin combinations with cytokine support have been shown to yield statistically improved rates of antitumor response.[67] [69] However, these do not translate into improvements in survival and come at the cost of increased toxicity and a decrease in quality of life. Given the limitations and toxicities associated with cytotoxic chemotherapy, emphasis has been to develop novel drugs against rational drug targets such as the KIT receptor tyrosine kinase, which is constitutively activated in most gastrointestinal stromal tumors (GISTs). GISTs are mesenchymal neoplasms showing differentiation toward the interstitial cells of Cajal and are typically characterized by the expression of the receptor tyrosine kinase KIT (CD117).[70] Recent studies have established that activating mutations of KIT are present in up to 92% of GISTs and likely play a key role in the development of these tumors.[71] [72] Along with mitotic activity, histologic subtype, and size, the type and the location of KIT mutation are prognostic for survival in patients with GIST.[73] Imatinib is a competitive inhibitor of BCR-ABL, KIT, PDGFR tyrosine kinases. [74] In preclinical studies, imatinib was active against mutant isoforms of KIT commonly found in GIST.[75] A recently completed phase II trial has shown substantial response rates as well as clinical benefit of imatinib in patients with advanced and metastatic GIST,[76] a group typically highly resistant to convention doxorubicin/ifosfamidebased chemotherapy. A total of 147 patients were randomly assigned to receive 400 mg or 600 mg of imatinib daily. Overall, 79 patients (53.7%) had a partial response, 41 patients (27.9%) had stable disease, and, for technical reasons, response could not be evaluated in 7 patients (4.8%). No patient had a complete response to the treatment. The median duration of response had not been reached after a median follow-up of 24 weeks after the onset of response. Early resistance to imatinib was noted in 20 patients (13.6%). Therapy was well tolerated, although mild-to-moderate edema, diarrhea, and fatigue were common. Gastrointestinal or intra-abdominal hemorrhage occurred in approximately 5% of patients. There were no significant differences in toxic effects or response between the two doses. Thus, inhibition of the KIT signal-transduction pathway is a promising treatment for gastrointestinal stromal tumors. Trials are presently underway to evaluate the efficacy of adjuvant imatinib therapy for patients with primary GIST larger than 2.5 cm.
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/316.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

PROGNOSTIC FACTORS AND RESULTS A prospective collected series of over 1000 patients has characterized the risk factors for outcome in patients with extremity soft tissue sarcoma.[77] Overall 5-year survival for this cohort of patients was 76% with a median follow-up time of 4 years. Significant independent adverse prognostic factors are outlined in Table 304 . The important prognostic factors for local recurrence were age greater than 50, recurrent disease at the time of presentation, microscopically positive surgical margins, and the histologic subtypes fibrosarcoma and malignant peripheral nerve tumor. For distant recurrence, large tumor size, deep location, high histologic grade, recurrent disease at presentation, and leiomyosarcoma and nonliposarcoma histology were all independent adverse prognostic factors, as was depth. For disease-specific survival, large tumor size, high grade, deep location, recurrent disease at presentation, histologic subtypes leiomyosarcoma and malignant peripheral TABLE 30-4 -- Multivariate Analysis of Prognostic Factors for Outcome, in 1041 Patients with Extremity Soft Tissue Sarcoma Managed at a Single Institution Local Recurrence Age > 50 Recurrent presentation Size > 10 cm Size > 5 cm Deep location Grade high Histology: fibrosarcoma Not liposarcoma Leiomyosarcoma Malignant peripheral nerve tumor Margin positive 0.001 0.0001 0.006 0.003 0.024 0.012 0.008 0.011 0.001 0.0001 0.015 0.03 0.0001 0.0007 0.0001 0.0002 0.0001 0.003 0.0001 Distant Recurrence Disease-Specific Survival

Modified from Pisters P, Leung D, Woodruff J, et al: Analysis of prognostic factors in 1041 patients with localized soft tissue sarcomas of the extremity. J Clin Oncol 14:1679, 1996.

812

Figure 30-7 Disease-specific survival for patients with retroperitoneal/intra-abdominal soft tissue sarcoma grouped by presentation status. Of the 775 patients, 439 (56%) had primary disease, 159 (21%) had local recurrence, and 177 (23%) had metastasis. Median survival was 86 months for those with primary disease, 23 months for those with local recurrence, and 12 months for those with metastasis. Patients were seen at Memorial Sloan-Kettering Cancer Center from July 1, 1982, to June 30, 2002.

nerve tumor, and microscopically positive margins were all adverse prognostic factors. This emphasizes that there are numerous independent adverse prognostic factors for distant recurrence and disease-specific survival and that these are clearly different from those involved in local recurrence. For retroperitoneal soft tissue sarcoma, an analysis of 500 patients with retroperitoneal soft tissue sarcoma treated and followed at a single institution has been reported.[38] Two hundred seventy-eight of these patients presented with a primary tumor, and 422, or 44%, presented with recurrent disease with a median follow-up of 28 months (1 to 172 months with 40-month follow-up for all survivors); this suggests a median survival of 72 months for patients with primary disease, 28 months for those with local recurrence and 10 months for those presenting with metastasis. Patients with locally recurrent tumors, unresectable disease, or incomplete resection and high-grade tumors all had a diminished survival time. Both for primary and locally recurrent tumors, the ability to completely resect the tumor was a predominant factor in outcome. After complete resection, the presence of a low-grade

tumor was a favorable factor for outcome. Disease-specific survival is illustrated in Figure 307 . It was of value to re-resect patients who experienced recurrence, the complete resectability rate falling with progressive recurrence with few tumors ever being able to be completely resected after a third or subsequent recurrence. An analysis of 200 patients identified as having GISTs morphologically but not by expression of KIT by immunohistochemistry has suggested that complete surgical resection is the only factor that makes a significant outcome benefit for the patient.[44] Size but not microscopic margin appears to be a factor in predicting survival. A more recent analysis of 49 patients with GIST,[73] all confirmed by significant immunohistochemical expression of KIT, has demonstrated the prognostic importance of both mitotic activity and GIST mutation type in predicting sarcomaspecific survival in a group of patients with GISTs before the availability of imatinib therapy. In contradistinction to retroperitoneal sarcomas, recurrence occurs equally both locally (throughout peritoneal cavity) and systemically in the liver. Unfortunately, almost half of all patients will present with metastasis, usually to the liver. If complete surgical resection of gross disease in patients with a primary presentation can be achieved, then these patients will have a 5-year actuarial survival of 54%.
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/317.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

LONG-TERM FOLLOW-UP It is essential to emphasize that long-term follow-up for all patients with soft tissue sarcoma is important. A recent analysis of long-term follow-up for patients followed more than 5 years showed that approximately 9% of patients who were disease free at the end of 5 years would go on to have further recurrence of the primary extremity sarcoma.[78]
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/318.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

SUMMARY Soft tissue sarcomas are relatively rare, with an annual incidence of 6000 to 6500 in the United States. Primary therapy is predicated on surgical resection with an adequate
813

margin of normal tissue. For high-risk patients, local control is improved with postoperative adjuvant irradiation. Local recurrence rates vary, depending on anatomic site. In extremity lesions, one third of patients develop locally recurrent disease, with a median disease-free interval of 18 months. Treatment results for localized extremity local recurrence may approach those for primary disease. Isolated pulmonary metastases may be resected with 20% to 30% 3-year survival rates after complete resection. In patients with retroperitoneal and visceral sarcoma, complete resection remains the dominant factor in outcome. As opposed to extremity sites, local recurrence in this site is a common cause of death. Patients with unresectable pulmonary metastases or extrapulmonary metastatic sarcoma have a uniformly poor prognosis and are best treated with systemic chemotherapy.
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/319.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

Selected References
Baldini EH, Goldberg J, Jenner C, et al: Long-term outcomes after function-sparing surgery without radiotherapy for soft tissue sarcoma of the extremities and trunk. J Clin Oncol 17:32523259, 1999. This study suggests that there may be a select subset of patients with soft tissue sarcoma in whom carefully performed function-sparing surgery may serve as definitive therapy and in whom adjuvant radiotherapy may not be necessary. Brennan MF, Lewis JJ: Diagnosis and Management of Soft Tissue Sarcoma. London, Martin Dunitz, 2002. Lewis JJ, Brennan MF: Soft tissue sarcomas. Curr Probl Surg 33:817880, 1996. Singer S, Demetri GD, Baldini EH, Fletcher CDM: Management of soft-tissue sarcomas: An overview and update. Lancet Oncol 1:7585, 2000. These reviews summarize the subject in a single monograph/book. Demetri GD, von Mehren M, Blanke CD, et al: Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med 347:472480, 2002. Singer S, Rubin BP, Lux ML, et al: Prognostic value of KIT mutation type, mitotic activity, and histologic subtype in gastrointestinal stromal tumors. J Clin Oncol 20:38983905, 2002. These studies demonstrate the importance of KIT activation and mutations in GIST pathogenesis and rationale and application of KIT tyrosine kinase inhibitors for targeted therapy of GIST. Ladanyi M, Bridge JA: Contribution of molecular genetic data to the classification of sarcomas. Hum Pathol 31:532538, 2000. This thorough review details the significant progress made in recent years in characterizing chromosomal changes associated with soft tissue sarcomas. In addition, recent molecular analyses of several sarcoma-associated translocations and the identification of novel genes and mechanisms of dysregulation are discussed. The role of cytogenetics and molecular changes, in the context of diagnosis and future investigation, is discussed. Lewis JJ, Leung D, Woodruff JM, Brennan MF: Retroperitoneal soft tissue sarcoma: Analysis of 500 patients treated and followed at a single institution. Ann Surg 228:355365, 1998. This manuscript provides an extensive description of outcome for patients with retroperitoneal sarcoma. Pisters P, Leung D, Woodruff J, et al: Analysis of prognostic factors in 1041 patients with localized soft tissue sarcomas of the extremity. J Clin Oncol 14:16791689, 1996. This manuscript provides data on prognostic factors for extremity soft tissue sarcoma, from a large single institution series. Pisters PWT, Harrison LB, Leung DH, et al: Long-term results of a prospective randomized trial evaluating the role of adjuvant brachytherapy in soft tissue sarcoma. J Clin Oncol 14:859868, 1996. Yang JC, Chang AE, Baker AR, et al: Randomized prospective study of the benefit of adjuvant radiation therapy in the treatment of soft tissue sarcomas of the extremity. J Clin Oncol 16:197203, 1998. These studies confirm the benefit of adjuvant radiation therapy in patients with completely resected localized extremity sarcoma.

Copyright 2004 Elsevier Inc. All rights reserved.

www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/320.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

References
1. Lewis

JJ, Brennan MF: Soft tissue sarcomas. Curr Probl Surg 33:817, 1996. D, Wright E, Nguyen D: Gene for von Recklinghausen neurofibromatosis is in the pericentromeric region of chromosome 17. Science 236:1100, 1987.

2. Barken 3. Li

FP, Fraumeni JF: Soft-tissue sarcomas, breast cancer, and other neoplasms: A familial syndrome? Ann Intern Med 71:747, 1969. SA, Mulvihill JJ, Nielsen A: Long-term follow-up of von Recklinghausen neurofibromatosis: Survival and malignant neoplasms. N Engl J Med 314:1010, 1986.

4. Sorensen 5. Brady

MS, Gaynor JJ, Brennan MF: Radiation-associated sarcoma of bone and soft tissue. Arch Surg 127:1379, 1992. MF, Lewis JJ: Diagnosis and Management of Soft Tissue Sarcoma. London, Martin Dunitz, 2002. PA, Zocchetti C, Guercilena S, et al: Dioxin exposure and cancer risk: A 15-year mortality study after the Seveso accident. Epidemiology 8:646, 1997.

6. Brennan 7. Bertazzi 8. Dich

J, Zahm SH, Hanberg A, Adami HO: Pesticides and cancer. Cancer Causes Control 8:420, 1997.

9. Mundt

KA, Dell LD, Austin RP, et al: Historical cohort study of 10,109 men in the North American vinyl chloride industry, 194272: Update of cancer mortality to 31 December 1995. Occup Environ Med 57:774, 2000. JA: Cytogenetics of soft tissue tumors. Cancer Treat Res 91:9, 1997. CD: Soft tissue tumours: The impact of cytogenetics and molecular genetics. Verh Dtsch Ges Pathol 81:318, 1997.

10. Fletcher 11. Fletcher

12. Sreekantaiah

C, Ladanyi M, Rodriguez E, Chaganti RS: Chromosomal aberrations in soft tissue tumors: Relevance to diagnosis, classification, and molecular mechanisms. Am J Pathol 144:1121, 1994.

13. Meis-Kindblom JM, Sjogren H, Kindblom LG, et al: Cytogenetic and molecular genetic analyses of liposarcoma and its soft tissue simulators: Recognition of new variants and differential diagnosis. Virchows Arch 439:141, 2001.

815

14. Kawai

A, Woodruff J, Healey JH, et al: SYT-SSX gene fusion as a determinant of morphology and prognosis in synovial sarcoma. N Engl J Med 338:153, 1998.

15. Ladanyi M, Antonescu CR, Leung DH, et al: Impact of SYT-SSX fusion type on the clinical behavior of synovial sarcoma: A multi-institutional retrospective study of 243 patients. Cancer Res 62:135, 2002. 16. Bennicelli 17. Cance 18. Dei

JL, Barr FG: Chromosomal translocations and sarcomas. Curr Opin Oncol 14:412, 2002.

WG, Brennan MF, Dudas ME, et al: Altered expression of the retinoblastoma gene product in human sarcomas. N Engl J Med 323:1457, 1990.

Tos AP, Doglioni C, Piccinin S, et al: Molecular abnormalities of the p53 pathway in dedifferentiated liposarcoma. J Pathol 181:8, 1997.

19. Antonescu

CR, Leung DH, Dudas M, et al: Alterations of cell cycle regulators in localized synovial sarcoma: A multifactorial study with prognostic implications. Am J Pathol 156:977, 2000. Y, Sakamoto A, Satio T, et al: Molecular abnormalities of p53, MDM2, and H-ras in synovial sarcoma. Mod Pathol 13:994, 2000.

20. Oda

21. Antonescu CR, Tschernyavsky SJ, Decuseara R, et al: Prognostic impact of P53 status, TLS-CHOP fusion transcript structure, and histological grade in myxoid liposarcoma: A molecular and clinicopathologic study of 82 cases. Clin Cancer Res 7:3977, 2001. 22. Abudu

A, Mangham DC, Reynolds GM, et al: Overexpression of p53 protein in primary Ewings sarcoma of bone: Relationship to tumour stage, response and prognosis. Br J Cancer 79:1185, 1999. Alava E, Antonescu CR, Panizo A, et al: Prognostic impact of P53 status in Ewing sarcoma. Cancer 89:783, 2000. G, Antonescu CR, de Alava E, et al: Prognostic impact of INK4A deletion in Ewing sarcoma. Cancer 89:793, 2000. JA, Pellin A, Noguera R, et al: Molecular analysis of the 9p21 locus and p53 genes in Ewing family tumors. Lab Invest 81:803, 2001.

23. de

24. Wei

25. Lopez-Guerrero

26. Drobnjak M, Latres E, Pollack D, et al: Prognostic implications of p53 nuclear overexpression and high proliferation index of Ki-67 in adult soft-tissue sarcomas. J Natl Cancer Inst 86:549, 1994. 27. Wurl

P, Meye A, Lautenschlager C, et al: Clinical relevance of pRb and p53 co-overexpression in soft tissue sarcomas. Cancer Lett 139:159, 1999. M, Bridge JA: Contribution of molecular genetic data to the classification of sarcomas. Hum Pathol 31:532, 2000.

28. Ladanyi

29. Zoubek A, Dockhorn-Dworniczak B, Delattre O, et al: Does expression of different EWS chimeric transcripts define clinically distinct risk groups of Ewing tumor patients? J Clin Oncol 14:1245, 1996. 30. de

Alava E, Kawai A, Healey JH, et al: EWS-FLI1 fusion transcript structure is an independent determinant of prognosis in Ewings sarcoma. J Clin Oncol 16:1248, 1998.

31. Sorensen

PH, Lynch JC, Qualman SJ, et al: PAX3-FKHR and PAX7-FKHR gene fusions are prognostic indicators in alveolar rhabdomyosarcoma: A report from the childrens oncology group. J Clin Oncol 20:2672, 2002.

32. Fletcher CD, Unni KK, Mertens F: Pathology and genetics of tumors of soft tissue and bone. In Kleihues P, Sobin LH (eds): World Health Organization Classification of Tumors, vol 1. Lyon, IARC Press, 2002. 33. Brown 34. Kattan 35. Lewis 36. Heslin 37. Lewis 38. Lewis

FM, Fletcher CD: Problems in grading soft tissue sarcomas. Am J Clin Pathol 114(Suppl):S82, 2000. MW, Leung DH, Brennan MF: Postoperative nomogram for 12-year sarcoma-specific death. J Clin Oncol 20:791, 2002.

J, Brennan MF: Soft tissue sarcomas. Curr Probl Surg 33:817, 1996. MJ, Lewis JJ, Woodruff JM, Brennan MF: Core needle biopsy for diagnosis of extremity soft tissue sarcoma. Ann Surg Oncol 4:425, 1997. JJ, Brennan MF: The management of retroperitoneal soft tissue sarcoma. Adv Surg 33:329, 1999. JJ, Leung D, Woodruff JM, Brennan MF: Retroperitoneal soft-tissue sarcoma: Analysis of 500 patients treated and followed at a single institution. Ann Surg 228:355, 1998. DG: Optimal radiologic imaging of soft tissue sarcomas. Semin Surg Oncol 17:2, 1999. DM, Go SD, Healey JH, et al: Soft-tissue sarcoma involving bone or neurovascular structures: MR imaging prognostic factors. Radiology 205:871, 1997.

39. Varma

40. Panicek

41. Panicek DM, Gatsonis C, Rosenthal DI, et al: CT and MR imaging in the local staging of primary malignant musculoskeletal neoplasms: Report of the Radiology Diagnostic Oncology Group. Radiology 202:237, 1997. 42. Fong Y, Coit DG, Woodruff JM, Brennan MF: Lymph node metastasis from soft tissue sarcoma in adults: Analysis of data from a prospective database of 1772 sarcoma patients. Ann Surg 217:72, 1993. 43. Gadd

MA, Casper ES, Woodruff JM, et al: Development and treatment of pulmonary metastases in adult patients with extremity soft tissue sarcoma. Ann Surg 218:705, 1993. RP, Lewis JJ, Leung D, et al: Two hundred gastrointestinal stromal tumors: Recurrence patterns and prognostic factors for survival. Ann Surg 231:51, 2000.

44. DeMatteo 45. Gaynor

JJ, Tan CC, Casper ES, et al: Refinement of clinicopathologic staging for localized soft tissue sarcoma of the extremity: A study of 423 adults. J Clin Oncol 10:1317, 1992. I, Cooper J, Henson D, et al (eds): AJCC Cancer Staging Manual, 5th ed. Philadelphia, Lippincott-Raven, 1997. MF: Staging of soft tissue sarcomas. Ann Surg Oncol 6:8, 1999.

46. Fleming 47. Brennan 48. Greene

F, Page D, Fleming I, Fritz A, et al (eds): AJCC Cancer Staging Manual, 6th ed. Heidelberg, Springer-Verlag, 2002.

49. Rosenberg

SA, Tepper J, Glatstein E, et al: The treatment of soft-tissue sarcomas of the extremities: Prospective randomized evaluations of (1) limb-sparing surgery plus radiation therapy compared with amputation and (2) the role of adjuvant chemotherapy. Ann Surg 196:305, 1982. WC, Collin C, Casper ES, et al: The changing role of amputation for soft tissue sarcoma of the extremity in adults. Surg Gynecol Obstet 175:389, 1992.

50. Williard

51. Yang JC, Chang AE, Baker AR, et al: Randomized prospective study of the benefit of adjuvant radiation therapy in the treatment of soft tissue sarcomas of the extremity. J Clin Oncol 16:197, 1998. 52. Pisters

PW, Harrison LB, Leung DH, et al: Long-term results of a prospective randomized trial of adjuvant brachytherapy in soft tissue sarcoma. J Clin Oncol 14:859, 1996.

53. Baldini EH, Goldberg J, Jenner C, et al: Long-term outcomes after function-sparing surgery without radiotherapy for soft tissue sarcoma of the extremities and trunk. J Clin Oncol 17:3252, 1999. 54. Alektiar

KM, Leung D, Zelefsky MJ, Brennan MF: Adjuvant radiation for stage II-B soft tissue sarcoma of the extremity. J Clin Oncol 20:1643, 2002.

55. Baldini EH, Demetri GD, Fletcher CD, et al: Adults with Ewings sarcoma/primitive neuroectodermal tumor: Adverse effect of older age and primary extraosseous disease on outcome. Ann Surg 230:79, 1999. 56. Esnaola

NF, Rubin BP, Baldini EH, et al: Response to chemotherapy and predictors of survival in adult rhabdomyosarcoma. Ann Surg 234:215, 2001.

814

57. Bramwell

V, Rouesse J, Steward W, et al: Adjuvant CYVADIC chemotherapy for adult soft tissue sarcomareduced local recurrence but no improvement in survival: A study of the European Organization for Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group. J Clin Oncol 12:1137, 1994.

58. Tierney JF, Stewart LA, Parmar MKB, et al: (Sarcoma Meta-analysis Collaboration): Adjuvant chemotherapy for localised resectable soft-tissue sarcoma of adults: Metaanalysis of individual data. Lancet 350:1647, 1997. 59. Singer

S, Corson JM, Demetri GD, et al: Prognostic factors predictive of survival for truncal and retroperitoneal soft-tissue sarcoma. Ann Surg 221:185, 1995. MF: Retroperitoneal sarcoma: Time for a national trial? Ann Surg Oncol 9:324, 2002. KM, Hu K, Anderson L, et al: High-dose-rate intraoperative radiation therapy (HDR-IORT) for retroperitoneal sarcomas. Int J Radiat Oncol Biol Phys 47:157, 2000. KG, Lewis JJ, Leung DH, et al: Multifactorial analysis of the survival of patients with distant metastasis arising from primary extremity sarcoma. Cancer 85:389, 1999.

60. Brennan 61. Alektiar

62. Billingsley 63. Singer

S, Antman K, Corson JM, Eberlein TJ: Long-term salvageability for patients with locally recurrent soft-tissue sarcomas. Arch Surg 127:548, 1992. MF: The enigma of local recurrence. The Society of Surgical Oncology. Ann Surg Oncol 4:1, 1997.

64. Brennan

65. Billingsley 66. Benjamin 67. Antman

KG, Burt ME, Jara E, et al: Pulmonary metastases from soft tissue sarcoma: Analysis of patterns of diseases and post-metastasis survival. Ann Surg 229:602, 1999.

RS, Rouesse J, Bourgeois H, van Hoesel QG: Should patients with advanced sarcomas be treated with chemotherapy? Eur J Cancer 34:958, 1998.

K, Crowley J, Balcerzak S: An Intergroup phase III randomized study of doxorubicin and dacarbazine with or without ifosfamide and mesna in advanced soft tissue and bone sarcomas. J Clin Oncol 11:1276, 1993. A, Ryan L, Sulkes A: Response to mesna, doxorubicin, ifosfamide, and dacarbazine in 108 patients with metastatic or unresectable sarcoma and no prior chemotherapy. J Clin Oncol 7:1208, 1989. AD: High-dose therapy for adult soft tissue sarcoma: Dose response and survival. Semin Oncol 25:19, 1998. CD, Berman JJ, Corless C, et al: Diagnosis of gastrointestinal stromal tumors: A consensus approach. Hum Pathol 33:459, 2002.

68. Elias

69. Elias

70. Fletcher 71. Hirota 72. Rubin 73. Singer 74. Druker

S, Isozaki K, Moriyama Y, et al: Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science 279:577, 1998. BP, Singer S, Tsao C, et al: KIT activation is a ubiquitous feature of gastrointestinal stromal tumors. Cancer Res 61:8118, 2001. S, Rubin BP, Lux ML, et al: Prognostic value of KIT mutation type, mitotic activity, and histologic subtype in gastrointestinal stromal tumors. J Clin Oncol 20:3898, 2002. BJ, Tamura S, Buchdunger E, et al: Effects of a selective inhibitor of the Abl tyrosine kinase on the growth of Bcr-Abl positive cells. Nat Med 2:561, 1996. DA, Willis NA, Jacks T, et al: STI571 inactivation of the gastrointestinal stromal tumor c-KIT oncoprotein: Biological and clinical implications. Oncogene 20:5054, 2001.

75. Tuveson

76. Demetri 77. Pisters 78. Lewis

GD, von Mehren M, Blanke CD, et al: Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med 347:472, 2002.

PW, Leung DH, Woodruff J, et al: Analysis of prognostic factors in 1,041 patients with localized soft tissue sarcomas of the extremities. J Clin Oncol 14:1679, 1996. JJ, Leung D, Casper ES, et al: Multifactorial analysis of long-term follow-up (more than 5 years) of primary extremity sarcoma. Arch Surg 134:190, 1999.

Copyright 2004 Elsevier Inc. All rights reserved.

www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/321.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

Bone Tumors
John H. Healey M.D. PRIMARY BONE TUMORS
General Concepts

Bone tumors have a low incidence but a high significance. All surgeons should have a basic understanding of them to avoid errors and provide optimal care. In this chapter, the most common benign and malignant bone tumors are discussed; these tumors serve as paradigms for the management of other lesions in their category of bone tumors. There is a wide spectrum of bone lesions for which benign and malignant tumors are in the differential diagnosis. Surgeons should seek appropriate radiologic and orthopedic help to diagnose and treat the lesions. The two basic questions to address are who needs a biopsy and who should perform the biopsy. Who needs a biopsy? Any person with a lesion that cannot be diagnosed by clinical and radiographic means or whose lesion requires treatment is a biopsy candidate. A tissue diagnosis is needed when the history, examination, and imaging studies cannot be reconciled. Proper preliminary investigation includes plain radiographs and not just magnetic resonance imaging (MRI). These tests are complementary. Discriminating evaluation of the plain radiographs is still the most specific way to diagnose bone tumors. Get another opinion if doubt exists. Orthopedic,
816

oncologic, and imaging expertise is available in most medical centers. Generally, if the physician is not comfortable treating the possible tumors that could be diagnosed from a biopsy, it is better to refer the patient to a specialist. The treating surgeon should have the opportunity to biopsy and analyze the tumor. This principle is crucial to diagnose bone tumors accurately and avoid inopportune biopsies that could compromise or preclude definitive surgical treatment.[1] If there is inconsistency between the biopsy results and other clinical features, further analysis is needed. This may be another test, rebiopsy, or clinical follow-up.
Clinical Features

The medical history helps to exclude infection and congenital conditions. Focused examination may reveal the source of metastatic cancers. Rarely, the family history is diagnostic, as in multiple hereditary exostosis (osteochondromatosis). Slow-growing benign tumors may not be painful, although they often are tender. Malignant and aggressive benign bone tumors usually hurt. They grow inside a rigid bony compartment, painfully increasing intraosseous pressure, or break out through the cortex and painfully stretch the periosteum. Exceptions include slowly growing tumors such as low-grade chondrosarcomas, which may be painless in one third of patients. Tumor pain may be worse at night. Mechanical pain, exacerbated by weight bearing, reflects structural insufficiency and may herald a fracture. Coincidental findings are common. Thirty percent of patients report that they sustained trauma to the area, bringing a lesion to the patients attention. A frequent source of confusion is periarticular or intra-articular pathology. Sources of referred pain must be considered and excluded. The physician must exhibit great restraint and avoid diagnosing the obvious comorbid condition and miss the offending primary lesion or tumor. If symptoms do not resolve in a predictable fashion, patients should have a radiographic (re)examination. Physical examination findings such as a palpable mass, painful or restricted joint motion, tenderness, or an effusion are nonspecific. Vascular and neurologic findings are uncommon but may reflect compression or invasion of these structures. Lymph node involvement is rare, pointing toward a diagnosis of lymphoma, infectious, or metastatic disease. General examination of the chest, abdomen, and pelvis may reveal the site of a primary cancer in cases of metastatic disease.
Benign Tumors

Up to 43% of normal children have developmental or neoplastic bony defects during growth, the most common of which is a fibrous cortical defect.[2] This lesion usually is found incidentally and is asymptomatic. It is important to distinguish it from a true neoplasm that requires a biopsy. The most common benign bone tumor (12% of all tumors and 50% of benign bone tumors) is the solitary osteochondroma ( Fig. 308 ). Nearly one half of these develop about the knee. Symptomatic osteochondromas and those growing after skeletal maturity warrant excisional biopsy if the diagnosis is ensured. Incisional biopsy and staged resection are reserved for osteochondromas with a thick (e.g., 1.5 cm) cartilage cap on MRI. Such lesions could have converted to chondrosarcomas. Prophylactic excision of osteochondromas rarely is indicated, because the surgical risks exceed the risk of sarcomatous degeneration (estimated at 0.1%). One should not operate simply because a lesion is present. Several of the less common tumors merit comment. Giant cell tumors are the most worrisome benign aggressive tumors ( Fig. 309 ). They are lytic epiphyseal tumors that frequently (65%) occur in the distal femur or proximal tibia in young adults. Approximately 3% metastasize to lung despite their benign histologic appearance. Chondroblastomas have the same epiphyseal distribution but occur in skeletally immature patients and usually have a rim of reactive bone radiographically. The humeral head and tuberosities are frequent sites. One third of cases occur around the knee. Aneurysmal bone cysts are benign aggressive tumors that have a diaphyseal or metaphyseal distribution and may develop as secondary lesions on top of other benign or even malignant bone tumors. Treatment of these conditions is as described for giant cell tumor (see later). There are several tumors that are treated differently. Unicameral bone cysts are common lytic lesions in the metaphysis that have a predilection for the proximal humerus and femur in children and often present as a

Figure 30-8 A and B, Benign osteochondroma. Note the continuity between the cortex of the shaft and that of the lesion. Similarly, the cancellous bone spaces are continuous.

817

Figure 30-9 A and B, Giant cell tumor of bone. This is typically a lytic epiphyseal lesion in a young adult without surrounding bone reaction radiographically. MRI may show more extensive disease than is apparent on plain radiographs.

fracture. Repetitive injections of corticosteroid or bone marrow and demineralized bone matrix can cure these, with recurrence rates similar to after open curettage and grafting.[3] Osteoid osteomas are exquisitely painful lesions that provoke a large amount of reactive bone formation and are very responsive to aspirin or nonsteroidal anti-inflammatory drugs. Percutaneous radiofrequency ablation may obviate surgical excision for lesions that do not burn out during a 3- to 12-month medication trial.[4] Finally, enchondromas are benign cartilage deposits in long bones that are rarely painful and do not require surgery. One to 10 percent can progress to chondrosarcoma, so they should be monitored.[5] When they are found by a bone scan during the work-up for possible metastases of a primary breast, prostate, or lung cancer, they may necessitate a biopsy to exclude a metastasis and allow treatment of the primary cancer to proceed.
Malignant Tumors

Sarcomas, metastatic carcinomas, and hematologic malignancies affect bone. Primary sarcomas of bone are rare. Only 2100 cases per year occur in the United States. The most common tumors are mentioned here. Osteogenic sarcoma is the most common primary bone malignancy, occurring predominantly in teenagers and young adults. Approximately one half occur about the knee, usually in a metaphyseal site ( Fig. 3010 ). It provides the paradigm for treatment of spindle cell sarcomas. Chondrosarcomas, malignant fibrous histiocytomas, and fibrosarcomas are the other spindle cell mesenchymal tumors. Their incidence is approximately one half to two thirds that of osteogenic sarcomas, and they occur in an older population. Ewings family tumors classically are diaphyseal tumors in youth, yet nearly 50% of them occur in nondiaphyseal locations, and about 20% may occur in young to middle-aged adults. Ewings tumor is particularly common in the femur and pelvis and is the most common malignancy of the fibula. It serves as the prototype for treatment of round cell tumors. Metastatic deposits in bone are much more common than primary sarcomas, especially in older individuals. Hematologic malignancies affect bone. Lymphoma may

Figure 30-10 Osteogenic sarcoma. This is the most common primary bone sarcoma. It is typically a destructive bone formative lesion of the metaphyseal area of the distal femur or proximal tibia in an adolescent; however, it may occur in any bone at any age. A, Anteroposterior radiograph shows the extensive bone formation in the soft tissue and marrow space. B, Lateral radiograph. C, Axial gradient-echo MR image shows the huge soft tissue mass. The arrow shows the fluid-fluid levels typical of an aneurysmal bone cyst. D, Axial T1weighted MR image shows the marrow margin present near the popliteal vessels. E, Coronal MR image shows the extent of the disease up into the proximal femur (arrow) and disease traversing the distal femoral growth plate (arrow).

818

develop as a primary bone tumor and seems to have a predilection for the femur in children or adults. If lymphoma is a possibility, then ample biopsy tissue should be taken so that studies such as cytogenetics and lymphoma markers can be performed. Multiple myeloma and metastatic lesions are the most common neoplasms in older adults. Breast, renal, and lung cancers are the usual culprits. The bone lesions should be sampled if they are the first site of presumed metastatic disease.
Diagnostic and Imaging Studies

Conventional biplane radiography is the most specific diagnostic test to determine if the tumor is benign or malignant. Benign lesions have sharp margins, narrow transition to the adjacent normal bone, and welldeveloped surrounding reactive bone. They may expand the bone cortex and rarely extend into the soft tissue. Malignant lesions lack sharp margins and are more permeative with a wide zone of transition, and their growth outstrips the bones ability to wall off the lesion. Cortex is destroyed, and there is periosteal elevation and usually soft tissue extension. The matrix within the lesion can suggest a specific diagnosis, such as osteogenic sarcoma, which produces a white cloud of bony matrix, or chondrosarcoma, which has broken rings of calcified cartilage matrix. Serum tests should be ordered selectively. Biochemical studies should be ordered when a metabolic disease or high-grade malignancy is suspected. The erythrocyte sedimentation rate (ESR) is elevated mildly in most malignant tumors and does not distinguish them from infections. Multiple myeloma causes significant elevations of the ESR and abnormal protein electrophoresis, as well as anemia, azotemia, and even hypercalcemia in widespread disease. The alkaline phosphatase level is elevated in 80% of osteogenic sarcomas and in most patients with Pagets disease of bone. Serum lactate dehydrogenase typically is elevated in round cell tumors such as Ewings family tumors and lymphoma, with values correlating with tumor burden. Serum calcium and parathyroid hormone levels are elevated in brown tumors of hyperparathyroidism and may help differentiate them from giant cell tumor. Other tests are rarely helpful. When a lesion displays benign and latent characteristics, no further imaging is warranted. When the diagnosis is unclear, or if a malignant lesion is suspected, advanced imaging should be done to aid the diagnosis and treatment. Systemic work-up (chest computed tomography [CT] for primary sarcomas and chest and abdominal CT for suspected bone metastases) should be done selectively. A whole-body technetium-99m pyrophosphate bone scan is essential to bone tumor staging and screening for multicentric disease or metastases. Occasionally, it will identify an easier lesion to biopsy. Imaging of the lesion can be by MRI or CT. MRI provides valuable diagnostic information about both benign and malignant lesions yet remains imprecise. The technique excels at defining normal and pathologic anatomy. It is particularly accurate at detecting soft tissue, vascular, neural, and marrow space involvement.[6] CT is the most effective in assessing cortical bone and endosteal cortical erosions caused by tumors such as low-grade chondrosarcomas. CT detects cortical breakthrough and is sensitive to identify soft tissue calcification (e.g., synovial sarcoma). All bone tumors should be staged to define the extent and behavior of the lesion. Benign tumors are staged most easily by the Campanacci system, based solely on the radiograph.[7] Latent lesions (stage 1) are contained within bone and do not affect the surrounding cortex. Active lesions (stage 2) are also within bone but evoke cortical thinning, expansion, or thickening. Aggressive lesions (stage 3) breach the cortex and extend into the surrounding soft tissue. The technique chosen for tumor ablation is governed by the stage, symptoms, and potential for recurrence. Asymptomatic stage 1 lesions can be curetted and bone grafted; it may be appropriate merely to follow some regularly with examination and radiographs. Symptomatic lesions should be treated. Stage 2 lesions can be managed with similar intralesional therapy but have a higher recurrence rate; therefore, local adjuvants such as cement, phenol, or cryosurgery may be warranted to enhance local control. Stage 3 lesions usually should be excised, or, if intralesional therapy is chosen, a local adjuvant should be strongly considered. Malignant bone tumor staging is standardized and prognostically accurate. Bone cancers are graded as either low grade (grade I) or high grade (grade II) based on histologic criteria and metastatic potential. The American Joint Committee on Cancer (AJCC) recently changed the subclassification of bone cancers. Substage A denotes that the tumor is less than 8 cm, whereas substage B signifies that the tumor is greater than 8 cm. Containment within the bone compartment remains an important prognostic factor and guides treatment decisions but is no longer part of the official staging system. Grade III

identifies all metastatic tumors, whether high or low grade. Most malignant tumors are high grade and large grade IIB lesions that are also extracompartmental. The staging of Ewings family tumor also includes bone marrow assessment. Refinements that take into account biologic measures of aggressiveness may be incorporated into future staging systems.
Staging and Biopsy Techniques

The most appropriate biopsy method is determined by practical considerations, such as the experience of the surgeon and pathologist and the availability of operating room space. Lesions that are virtually diagnostic radiographically can be sampled with a needle, whereas atypical radiographic presentations usually mandate an open biopsy ( Box 302 ). Ample fresh tissue increasingly is needed for genetic testing of primary bone tumors; thus an open biopsy is required to obtain adequate tissue. This may be obviated in the future with advances in fluorescent in situ hybridization that will allow a genetic profile to be obtained from scant material. Biopsy and treatment are staged unless a definitive diagnosis can be obtained by frozen section analysis. Antibiotics should not be administered
819

Box 30-2. Bone Biopsy Technique Needle biopsy if radiograph is diagnostic Open biopsy if radiograph is nondiagnostic No preoperative antibiotics Exsanguinate by gravitynot compression bandage tourniquet Small, extensile incision Radiographic localization if necessary Soft tissue biopsy if possible Round bone hole Frozen section for adequacy Fresh tissue for genetics studies Culture (fungus, tuberculosis, bacteriology) Absolute hemostasis Drain (if necessary) in line with definitive incision Protect weight bearing

before the biopsy so that microorganisms may be recovered. A tourniquet is desirable to minimize bleeding, but avoid using a compression bandage to exsanguinate the limb because this can embolize tumor. Use gravity and patience instead. The biopsy incision must be planned to be in continuity with the definitive resection incision. Generally, biopsy incisions should be extensile, avoid contaminating adjacent joints, and minimize unnecessary extracompartmental contamination. A frozen section analysis should confirm sufficiency of the biopsy. Meticulous hemostasis is very important to avoid local dissemination of tumor in hematoma. Absorbable gelatin sponge (Gelfoam) and thrombin or cementation of the cortical bone window prevents postoperative hemorrhage. If a drain is used, it should be placed in line with the incision, so that the tract of the drain can be excised easily in continuity with the biopsy incision. Fracture through the pathologic bone, which may preclude limb salvage, must be prevented. A typical biopsy weakens the bone 50%, so crutches are needed for at least 6 weeks to allow healing and prevent fracture.
Treatment of Primary Tumors

A four-part staging system specifies the nature of any tumor excision and defines the surgical margin.[8] An intralesional margin passes through the tumor; for example, curettage is an intralesional procedure suitable for benign lesions. A marginal margin courses through the layer of reactive, inflammatory tissue around a cancer. An excisional biopsy goes through the tumor pseudocapsule and almost always leaves microscopic disease behind. A wide margin removes the tumor and some surrounding normal tissue. Most limb-sparing en bloc excisions for sarcomas have a wide surgical margin. A radical margin removes the entire soft tissue and bone compartments that contain the tumor; examples are removing the entire femur for a lesion of the distal femur or the entire quadriceps for a lesion in the distal vastus medialis.
Giant Cell Tumors
Benign Tumors

Treatment of giant cell tumor embodies the methods and common technical variations that are used for other benign tumors of similar virulence, such as aneurysmal bone cyst, chondroblastomas, and chondromyxoid fibroma. Most giant cell tumors occur in the epiphyseal distal femur or proximal tibia and often extend to the subchondral surface. Curettage and bone grafting has approximately a 50% recurrence rate. At the other extreme, wide excision cures most cases, but this treatment has more morbidity and reconstructive difficulty and does not eliminate the chance of developing a metastasis. Modern treatment addresses the tumor aggressiveness and has reduced the historically high recurrence rate to less than 10% by using intralesional treatment and adjuvants.[9] This approach is oncologically sound and preserves joint function. The method consists of the following steps: (1) fully exteriorize the tumor through a cortical bone window large enough to allow direct vision into all areas of the tumor; (2) curet the lesion aggressively;

and (3) use a high-speed bur to eliminate tumor permeating cancellous bone and eradicate the subchondral tumor. Chemical and physical modalities extend the margin of tumor excision and improve tumor control. Acrylic cement is used most frequently. Cement polymerization is exothermic and thermally kills residual tumor cells, and the cement mechanically supports the defect. Despite expert technique, recurrence remains a problem. A study from the Massachusetts General Hospital reported recurrence in 7 of 25 tibial and 3 of 23 distal femoral giant cell tumors treated by these methods.[10] These disappointing results highlight the need for more aggressive treatment. Cryosurgery can dramatically extend the volume of bone sterilized and reduce the recurrence rate further. Prophylactic internal fixation prevents most fractures through the previously frozen bone. Malawer and colleagues reported a local recurrence rate of 2% for primary tumors and 8% for recurrent tumors treated with cryosurgery; the fracture rate was 6%.[9] Pathologic fracture or recurrence may disrupt the knee joint. Endoprosthetic knee replacement or allografts can salvage many such cases. Wide excision and replacement of a single condyle (or plateau) with an allograft is an excellent solution for large, eccentric tumors that destroy one tibial plateau or femoral condyle. This type of hemi-joint replacement replaces the articular cartilage, preserves joint alignment, stability, and structure and has a high success rate.
Malignant Tumors

Lesions that were once considered rapidly fatal now potentially are curable. Advances in sarcoma therapy have helped to establish and validate many of the concepts of modern surgical oncology, including principles of wide en bloc surgical excision, adjuvant multiagent
820

chemotherapy, and a multidisciplinary team approach to patient care. Surgery remains the primary curative modality for sarcomas. Wide excision is appropriate for low-grade sarcoma. When an effective adjuvant is available, a wide local excision is appropriate for high-grade cancers; radical resection or amputation may not be necessary. The effectiveness of adjuvant therapy must be considered. If the cancer is not responding to chemotherapy, for example, then it is prudent to obtain a wider surgical resection margin. Sarcomas occur throughout the body. The anatomic and functional aspects of resection and reconstruction warrant specialized surgical care. These diseases should be treated in major centers, and outcome depends on the surgeons experience. Because of the rarity of bone sarcomas, collaboration between large centers is necessary for advances to be made. Cases treated at nonparticipating centers off protocol compromise the ability to improve outcome for patients with these diseases.
Clinical Manifestations

The presenting symptoms of pain and swelling were described earlier. The patients medical and family history occasionally provides valuable information such as conditions that may predispose the patient to a secondary sarcoma, such as prior radiation, Pagets disease, bone infarct, and fibrous dysplasia. Patients with hereditary retinoblastoma or Li-Fraumeni syndrome have a genetic predisposition to developing sarcomas (particularly osteogenic sarcoma) as a result of a germline mutation in the RB tumor suppressor gene or TP53 gene, respectively. The pediatric patient with a sarcoma of bone is occasionally the index case for a family with Li-Fraumeni syndrome.
Imaging

Although the plain radiograph may seem unsophisticated, it remains the single most important radiographic test to diagnose bone sarcomas. The workup of a sarcoma of bone also should include a plain chest radiograph and chest CT, because the lung is the most common site of metastasis. A technetium bone scan should be obtained to assess both the primary tumor and to screen for distant metastases. Other nuclear medicine scans, such as thallium scans (usually for osteogenic sarcoma) have some utility in monitoring the response to chemotherapy.[11] Proton emission tomography (PET) is being used for diagnosis, staging, and assessment of response to chemotherapy.[12] Although there is occasionally merit to using this test, it is not yet established for the routine use in bone tumor imaging.
Biopsy

Biopsy is mandatory before treating bone sarcoma and should follow the guidelines described earlier. Bone-forming lesions are especially difficult to sample percutaneously. The most dependable biopsies are open biopsies.
Surgical Management

It is vital to have a clear concept of whether the operation is performed for diagnostic, palliative, adjunctive, or curative intent. Both low- and highgrade cancers can be cured by local excision, and each requires at least a wide surgical resection margin to adequately ensure that all cancer was resected. Nevertheless, there is usually another opportunity to cure patients who sustain a recurrence of low-grade sarcoma, whereas most recurrences of high-grade sarcomas result in the death of the patient. Oncologic goals should take precedence over reconstructive goals, but they often are interrelated. Limb-preserving surgery is currently appropriate for over 85% of patients with a sarcoma of the bone.[13] It is important to evaluate the relationship of the tumor with the vascular and neurologic structures to establish that limb-preserving surgery can be done. The primary indication for amputation is the inability to obtain an adequate margin of normal tissue around the tumor and still retain a useful limb. Conditions that commonly lead to amputation include widespread contamination of soft tissue by previous surgery, fracture, and infection. Another broad indication for amputation is the anticipated loss of vascular or nervous supply to the extremity that cannot be adequately restored. Finally the interrelationship between surgical margin and the effectiveness of adjuvant therapies must be taken into account.[14] If the chemotherapy is ineffective or compromised, then a wider surgical margin should be considered. An amputation is both an ablative and reconstructive procedure. Advances in prosthetics such as energy-conserving feet and ischial containment sockets have increased the potential function of amputees. Tellingly, young amputees are more likely to be able to participate in sports than sarcoma patients treated by endoprosthetic or allograft reconstructions of their limb. Disadvantages of amputations include (1) the psychological burden of losing a limb, (2) the considerable expense of maintaining and replacing prosthetic limbs, (3) the increased cardiac demand and oxygen consumption needed for lower extremity amputees to ambulate, and (4) phantom pain and sensations. Limb-preserving surgery has become possible for many patients due to improved imaging, surgical technique, reconstructive materials, and adjuvant therapy. Reconstructions should address deficiencies of bone, joint, and soft tissue. Limb-sparing procedures usually require large bone grafts, endoprosthetic replacement, or a combination of biologic and artificial materials. There is no perfect reconstructive strategy, and there are advantages and disadvantages for each approach. Bone segmental replacements include autologous or allogeneic bone grafts. Single sides of joints can be replaced by osteoarticular allografts. When both sides of the joint are missing, an endoprosthesis, alloprosthetic composite, arthrodesis, or rotationplasty is needed. The majority of patients will enjoy long-term survival, and the long-term quality and
821

durability of the limb reconstructions must be assessed critically.

Autogenous Grafts

Autogenous bone grafts are inexpensive, are biologically compatible, and avoid potential transmission of infectious diseases. They are used to replace short defects, to supplement allografts, and in specialized techniques such as free fibular transfer and bone transport (Ilizarov). Large bone grafts are used to provide structural support, and these usually are harvested from the fibula, iliac crest, and rib. The amount of bone available from these sites is limited. If the structural integrity of the diseased bone is preserved, it may be possible to remove the tumor, sterilize the bone, and reinsert it into the defect. This approach has been popular in many parts of the world where large bone banks do not exist and where endoprostheses are prohibitively expensive. Vascularized autogenous grafts can be either local pedicular grafts or free grafts, the most versatile being the free fibular graft. Vascularized grafts have several theoretical advantages as the result of an intact blood supply. The limited experience with vascularized grafts after tumor excision shows that they have not yet fulfilled expectations. Complication rates are high: infection (16%), nonunion rate (6% to 19%), and fracture (8% to 43%).[15] The limb must be protected with a cast, brace, or external fixator until hypertrophy occurs, often 2 years or longer. Innovative approaches, such as combining vascularized fibular grafts with allografts, may expand indications for the procedure in the future. Bone transport uses an external fixation device and advances a cut bone segment across the defect at 1 mm daily, inducing new bone formation in its wake. Joint stiffness, pin-track infection, and nonunion are among the complications seen from this treatment.
Rotationplasty

The Van Nes rotationplasty typically is done for a distal femoral lesion in a child younger than 10 years old.[16] [17] The procedure involves intercalary amputation of the knee and either distal femur or proximal tibia. The distal segment is rotated 180 degrees so that the foot faces backward and the tibia is secured to the proximal femur. This brings the ankle joint up to the level of the knee and allows it to bend in the direction of a normal knee. It effectively converts what would have been a high above-knee amputation into a below-knee amputation. Function is usually quite good, and most patients ambulate without crutches. The procedure allows for continued growth of the limb.
Endoprosthetic Replacement

Endoprostheses usually are employed in situations involving joint reconstruction where the primary advantage is immediate skeletal stability and restoration of function, an important consideration in patients who may not have very long to live. Prosthetic infection results in removal of the implant in 2% to 9% of cases and is highest for the proximal tibia, where the soft tissue coverage is poorest. Aggressive use of muscle flaps and free flaps has significantly reduced the rate of infections. Amputation may be needed to cure refractory infections or those for which treatment would delay resumption of postoperative chemotherapy or jeopardize the patients life. Limb retention rates after initial limb preservation for sarcoma averages 90% at 10 years. [18] The major cause of late failure is aseptic loosening related to numerous factors.[19] [20] [21] [22] The rate of loosening is highest for the proximal tibia, followed by the distal femur, proximal humerus, and finally proximal femur. Younger, more active patients have significantly higher failure rates than older, sedentary patients. The amount of bone resected also affects the rate of loosening. The use of less-constrained implants has improved longevity of implants. The newer, rotating hinge prostheses for knee replacement dissipate forces by allowing multiplanar motion.[23] The implants have long stems, secured to the host bone either by traditional cemented technique or with porous coated stems that rely on bone ingrowth for fixation. At this time it is not clear which method is better. Extendible prostheses have been used primarily for tumors around the knee in adolescent children to maintain growth and prevent leg-length discrepancy.[24] The extendible portion of the device is weak in some designs and subject to fracture. The lengthening must be done in small, incremental steps to avoid nerve palsy and joint contracture, thereby necessitating multiple procedures. The rates of aseptic loosening and overall failure are higher than the corresponding rates for nonexpandable prostheses.[25]
Allografts

Bulk allografts of bone are versatile biologic implants that are available in all types and sizes. They can be ordered with soft tissue attachments to reattach ligaments, restore muscle function, and provide joint stability. Allograft tissue is dead, with the exception of limited preservation of articular cartilage. They can be used to replace a bone segment (intercalary) or one side of a joint (osteoarticular graft.) Allografts can transmit viruses, but processing seems to reduce this risk significantly. The biomechanical and biologic properties of allografts are affected by the method of processing. Freezing in liquid nitrogen reduces immunogenicity with little effect on mechanical strength. Freeze-drying dramatically reduces immunogenicity of the grafts and simplifies storage but results in a significant loss (10% to 61%) of bending and torsional strength. The clinical results of allografts vary according to the situation in which they are employed. Replacement of a part of the diaphysis with an intercalary graft has the best outcome, and long-term graft survival is between 80% and 100%.[26] [27] The worst results have occurred with allograft arthrodesis, and where only 40% of the grafts were retained at 10 years. Osteoarticular allografts have had intermediate results. These frozen allografts cryopreserve the chondrocytes with dimethyl sulfoxide or ethylene
822

glycol. Although 40% to 70% of cells survive the experimental freeze-thawing process acutely, few survive long-term. Degenerative arthritis is inevitable over time. The major complications with allografts include fracture, 16% to 19%; nonunion, 17%; and infection, 13%. Because the graft is not vascularized, infections are difficult to eradicate and, in most instances, necessitate removal of the graft. The rate of infection is especially high for large pelvic allografts. To minimize the risk of infection, coverage of the graft with vital muscle and soft tissue is important. Impregnation of antibiotic bone cement into the medullary cavity may reduce the rate of infection while at the same time retarding bone resorption and preventing fractures.
Alloprosthetic Composite Reconstruction

Composite reconstructions use a combination of biologic and man-made materials. Most involve alloprostheses, which are a combination of allografts and endoprostheses ( Fig. 3011 ). This approach is advantageous because it provides tendinous and other soft tissue attachments to the endoprosthesis, as well as restoration of bone stock.[28] In the future, custom combinations of artificially engineered biological materials and implants will be used for limb reconstructions.
Adjuvant Therapies

Multiagent chemotherapy has been successful for primary sarcomas of bone and is now the standard treatment for

Figure 30-11 Allograft-prosthetic composite reconstruction. This is an example of one of the many possible reconstructions. A, Anteroposterior radiograph shows the allograft (arrows) and the rotating hinge knee replacement. B, Lateral radiograph shows how the bone provides the patellar tendon attachment (arrows).

osteogenic sarcoma and Ewings sarcoma, and possibly for malignant fibrous histiocytoma of bone. Unfortunately, chondrosarcoma does not respond to traditional agents. New cytotoxic or biologic agents are needed to treat these sarcomas. Most sarcomas, apart from Ewings sarcoma, are resistant to radiation at tolerable doses.
Osteogenic Sarcoma

Osteogenic sarcoma is the most common primary bone sarcoma, with an incidence of 2 to 3 per million people per year, and 600 to 800 cases occur every year in the United States. The most commonly affected sites are the distal femur, proximal tibia, and proximal humerus. The peak incidence is in the adolescent years, with a second peak around the age of 60 years. Osteogenic sarcoma is composed of distinct disorders with different causes and outcomes. Primary, high-grade conventional osteogenic sarcoma is the most common. It is a highly malignant tumor, arising in the medullary canal and metastasizing in over 80% of cases treated by surgery alone. Secondary osteogenic sarcoma develops in Pagets disease, bone infarct, fibrous dysplasia, or previously radiated bone. Secondary osteogenic sarcoma is rare in young patients but accounts for over half of the cases in patients older than 60 years of age. Juxtacortical osteogenic sarcomas arise from the external surface of a bone, most commonly on the posterior aspect of the distal femur. The majority of tumors are low grade and have a favorable prognosis, but some are high grade, invade the medullary canal, and metastasize rapidly. They should be designated as high or low grade. Patients with low-grade osteogenic sarcomas of all varieties should not receive chemotherapy, whereas high-grade tumors warrant chemotherapy. The genetic basis of osteogenic sarcoma is only partially understood.[29] Complex karyotypes are present in most tumors. Ring chromosomes have been noted in a substantial number of cases of juxtacortical osteogenic sarcoma. Every case seems to have an alteration of cell cycle regulation. Abnormalities of TP53 or RB fail to be prognostic under these conditions. It has recently been established that a telomere maintenance mechanism is essential for the cancer to persist. Uniquely, a high percentage of osteogenic sarcoma cases have an alternative, nontelomerase mechanism to remain immortal.[30] Dominant oncogenes also may be involved, including RAS, MET, FOS, and MYC.[31] Recently, the erb-B oncogene (epidermal growth factor receptor) was found to be present in 42% of cases and may present a therapeutic opportunity to modulate tumor growth.
Diagnosis

The lesion usually arises in the metaphysis of a long bone and grows outward from the medullary canal. On plain radiographs, classic osteogenic sarcoma produces a characteristic osteoblastic lesion in about 60% of cases. Lytic osteogenic sarcomas are composed of predominantly fibrous or telangiectatic tissue. Typical malignant features include a permeative growth pattern, poorly defined borders, and erosion through the cortex. MRI is excellent
823

for imaging lesions in the marrow, screening for skip lesions, and identifying intramedullary extension of juxtacortical tumors and penetration of the physeal cartilage. The essential histologic feature of osteogenic sarcoma is the production of osteoid by malignant, spindle-shaped cells.
Prognosis

Factors other than stage (discussed earlier) have possible prognostic importance [32] : Site: axial and pelvic tumors fare worse than appendicular tumors. Biochemistry: elevation of alkaline phosphatase above institutional normal and lactate dehydrogenase above institutional normal are independent predictors of an unfavorable outcome. Race: African Americans have a worse outcome. Secondary osteogenic sarcomas have a worse prognosis. Skip lesions, which may represent bone-to-bone metastases, are associated with a poor patient prognosis. Histologic necrosis: the extent of necrosis in the primary tumor reflects the response to chemotherapy and predicts long-term disease-free survival.
Treatment

For high-grade osteogenic sarcoma, surgical excision of the primary, and any metastatic tumor must be combined with adjuvant chemotherapy. An oncologically sound operation must be performed to prevent local recurrence. Data support the success of limb-preserving surgery. Overall survival matches that achieved after amputation, despite an increase in local recurrence rates. In North America, Rougraff and colleagues reported that the rate of local recurrence was 0% for hip disarticulation, 7.8% for transfemoral amputation, and 11% for limb-sparing surgery.[33] Internationally, the COSS

group reported local recurrence in 2.2% of cases of amputation or rotationplasty, in contrast to 11.1% for limb-sparing procedures.[34] At the Rizzoli Institute, local recurrence was 0% for rotationplasty and radical amputation, 8% for wide amputation, and 10% for limb-sparing surgery.[35] Poor responders to chemotherapy who also had narrow surgical margins suffered local recurrence in 20% of cases. They showed the interrelationship between good surgery and successful chemotherapy and, more importantly, between inadequate surgery and ineffective chemotherapy. After local recurrence, the 5-year survival rate is only 10%. Taken together, these studies indicate that amputation is more likely to control local disease than limbsparing surgery. Nevertheless, it is not established that amputation is any better at curing patients. Limb preservation is oncologically appropriate if coupled with effective adjuvant chemotherapy. Patients who present with metastatic disease have a poor prognosis, and only 30% survive for 5 years.[36] With aggressive treatment, including resection of pulmonary metastases and intensive chemotherapy, survival can be extended for most patients, and some patients can be cured. Complete resection of recurrences seems to be the most important factor for a successful outcome.
Adjuvant Therapy

Systemic adjuvant chemotherapy is required to eradicate metastatic deposits. Several randomized studies settled the issue definitively. High-dose methotrexate (HDMTX) has been used in most protocols for osteogenic sarcoma and continues to be a major component of current multiagent regimens. Several agents in addition to HDMTX subsequently were found to have efficacy, most notably doxorubicin and cisplatin. Multiagent chemotherapy is superior to single-agent adjuvant therapy. The T4, 5, 7, 10, and 12 protocols, used previously at Memorial Sloan-Kettering, resulted collectively in a 65% disease-free survival for 279 patients with stage II (nonmetastatic) disease, now after 13 years of follow-up.[37] The search for more effective combinations of chemotherapy continues. Adding more drugs to current regimens will compromise the dose intensity of the active agents already being used. The Childrens Oncology Group/Pediatric Oncology Group (CCG/POG) study investigated two agents: ifosfamide and muramyl tripeptide phosphoethanolamine (MTP-PE). The 22 factorial design allowed for complex interactions between the experimental agents, so no definitive conclusions can be drawn. However, the use of ifosfamide without MTP-PE had the worst outcome, so this approach should probably not be taken. Chemotherapy for osteogenic sarcoma has included preoperative neoadjuvant chemotherapy. Benefits of preoperative chemotherapy include immediate treatment of micrometastatic disease, enhancement of the margin of resection, and provision of time for surgical planning; however, there has not been a clear survival benefit.[38] A major advantage of preoperative chemotherapy is the ability to assess the histologic response to chemotherapy in the surgical specimen. Patients with more than 90% tumor necrosis after chemotherapy have significantly higher disease-free survival. Response to preoperative chemotherapy is highly dependent on which agents are used and the length of neoadjuvant therapy. Conceptually, it is attractive to administer different agents postoperatively if a patient does not respond well to the initial drug selections. Unfortunately, tailoring of chemotherapy does not improve survival for patients that do not respond well to first-line therapy.[37] Nevertheless, even patients who fail to show a dramatic response to preoperative chemotherapy still have a survival advantage compared with not receiving chemotherapy. Thus, chemotherapy should not be abandoned if necrosis is less than 90%. Understanding the mechanisms of drug resistance is of paramount importance. Finally, it is hoped that elucidation of the genetic mutations underlying osteogenic sarcoma ultimately will lead to gene-based therapy.
Ewings Family Tumors

Ewings tumor is a malignant tumor of small round cells that arises in bone. Cytogenetic and molecular studies show that Ewings sarcoma is a member of a family of
824

tumors that includes primitive neuroectodermal tumors, peripheral neuroepithelioma, Askins tumor of the chest wall, and extraosseous Ewings tumor. The cause is related to a reciprocal translocation t(,22)(q24,q12) in over 90% of cases, which results in a fusion of the EWS gene to the FLI1 gene. In approximately 5% of cases there is a 21,22 translocation that fuses the EWS gene to the ERG gene, and in rare cases the EWS gene may be fused to other genes, such as the E1A gene and the ETV1 gene. The chimeric proteins affect other genes that transform the cell. Ewings tumor is one half as common as osteogenic sarcoma in the United States. Among patients under 15 years old, Ewings tumor is nearly as common as osteogenic sarcoma. The disease is distinctly uncommon in blacks and Asians. The pelvis and femur are favored locations, but any bone may be involved.
Clinical Manifestations

Most patients have pain and swelling at the affected site. Growth of the tumor is rapid, a substantial, firm mass is present, and some patients have constitutional symptoms. Pathologic fractures occasionally occur.
Diagnosis

Ewings tumor has protean radiographic manifestations. Onionskin periosteal changes are characteristic but often lacking. Ewings tumor usually produces an ill-defined, lytic defect that permeates up and down the medullary canal, giving the bone a moth-eaten appearance. A large, soft tissue mass usually is adjacent to the bone. It is important to note the lactate dehydrogenase level because it is correlated to the disease burden and has prognostic importance. Osteomyelitis, Langerhans cell granuloma, and lymphoma are in the differential diagnosis. The tumor is composed of sheets of small, round, blue cells with hyperchromatic nuclei, scant cytoplasm, and little extracellular matrix. The monoclonal antibodies HBA71 and O13 show strong membranous staining for Ewings tumor in 91% of cases. However, the antibody is not 100% specific. Reverse transcriptase-polymerase chain reaction (RT-PCR) detects the chimeric pseudogenes. Ewings tumor can disseminate widely in the marrow so the bone marrow should be sampled during staging for Ewings tumor. Lungs and other bones are the usual sites of metastases.
Treatment

Treatment consists of multiagent chemotherapy and definitive local therapy. The mainstays of chemotherapy are doxorubicin, cyclophosphamide, vincristine, and dactinomycin. Ifosfamide and etoposide also have been shown to be effective in phase II and preliminary phase III trials. There has been an effort to stratify patients into high-risk versus standard-risk categories based on tumor size, location, and metastases.[39] [40] [41] Definitive local therapy of the primary tumor consists of surgery, radiation therapy, or a combination of both modalities. Approximately 50% of patients with Ewings tumor in the United States undergo surgery. The response to chemotherapy also affects the likelihood of local recurrence. Patients who responded well to chemotherapy had significantly better local control than patients who respond poorly.[42] Patients who had an objective response to chemotherapy and small tumors (<8 cm) achieved 90% local control, despite being given only a low radiation dose (35 Gy). Patients with large tumors

and an objective response to chemotherapy were given a high radiation dose (50 to 60 Gy) but only achieved 52% local control. Finally, patients who had no objective response to chemotherapy had only 17% local control, despite being given high-dose radiation (50 to 60 Gy). Sometimes after induction chemotherapy and the administration of radiation the lesion may regress and become resectable. Surgery controls local disease better than radiation. Despite improvements in imaging and radiation technology, radiotherapy shows disappointing results. Bacci reported that local recurrence was only 5% after primary surgical treatment, with or without radiation, and 41% after radiation without surgery.[43] The CESS study reported only 77% local control at 5 years. Ozaki and associates reported 15% local recurrence with radiation therapy alone, compared with 4% local recurrence with surgery alone and 4% with surgery and radiation therapy.[44] Surgery is often a superior alternative to remove a focus of radioresistant and chemoresistant cells that resides within a large tumor and to avoid the risk of radiation sarcoma. The cumulative risk of secondary sarcoma increases with time and has been estimated to be 8.6% at 20 years, with a mean latency of 7.6 years.[45] Surgery should no longer be reserved for expendable bones, such as the ribs, clavicle, and fibula. With modern surgical techniques, all bones are potentially resectable and can be reconstructed. The appropriate margin of resection is debated between (1) outside of the original tumor, (2) outside of the tumor after induction chemotherapy, or (3) outside of the tumor after induction chemotherapy and radiation. It is likely that no single therapeutic strategy will be ideal for all patients. The early and late morbidity of surgery varies with each tumor. Treatment efficacy and morbidity must be balanced when deciding how to manage the primary Ewings tumor site.
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/322.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

METASTATIC BONE DISEASE More than one third of the 965,000 new cancer patients annually in the United States will develop skeletal metastases. Bone metastases outnumber primary bone cancers 100 to 1. Bone disease and fractures cause most of the pain and disability endured by cancer victims. Patients deserve the highest quality diagnostic and therapeutic care to ameliorate symptoms and preserve their independence in the terminal phase of their illnesses. Metastatic bone cancer should be cared for in as knowledgeable and sensitive a fashion as primary sarcomas, even though most intervention is palliative. Patients have complex end-of-life issues to deal with, and the surgeon has an important role to play in the diagnosis and management of this most-feared sequela of cancer.

825

Treatment should be multidisciplinary and address the systemic disease as well as the local manifestations. Metabolic complications of hypercalcemia, bone pain, and fractures can be prevented or ameliorated with bisphosphonates, particularly in multiple myeloma, breast cancer, and other forms of lytic metastatic disease. Fracture prevention is a major goal. Appropriate use of rest, crutches, and braces provides the protection necessary while medications and radiation control the disease. Radiation therapy is usually appropriate for patients with symptomatic metastatic bone lesions. Even the so-called radioresistant tumors respond in at least 30% of cases. Surgery benefits those who fail radiation and medical management and those who sustain fractures. Bones at risk for fracture should be identified and treated early for the best results.
Epidemiology

Every cancer can metastasize to bone, but breast, prostate, lung, and kidney cancer do so most frequently. Bone metastases can occur in any bone, but clinically significant metastases are most common in the femur, humerus, and spine. Certain bones are more likely to develop metastases from particular primary tumors, that is, scapula for kidney cancer and hand metastases from lung and stomach primary tumors. Metastatic deposits are found most often in the metaphyses of long bones; they usually are multiple. The so-called solitary metastasis is usually the first of many lesions to be identified. Thyroid cancer, renal cancer, and myeloma (plasmacytoma) are the most likely to present as isolated metastases. Even these favorable cases typically develop widespread disease, suggesting that there is unrecognized dissemination of cancer at the time the first bone metastasis is identified. Metastases to the cortex alone are unusual, most commonly occur in lung cancer patients, and rapidly cause fracture. Surgical treatment of pathologic fractures is needed for approximately 9% of patients with clinically identifiable metastatic bone disease. The proximal femur and humerus are the most frequent sites operated on.
Pathogenesis

Metastatic cancer cells localize in bone because of bones unique microvasculature and microenvironment. Interactions between the tumor cell and the host bone are critical. Bone lacks lymphatic channels, and metastases are hematogenous in origin. A complex interconnecting valveless network of vertebral, epidural, and perivertebral veins, also known as Batsons plexus, influences the distribution of blood flow and metastases. Tumor cells extravasate easily from the marrow vasculature, and small implants develop. Local angiogenesis promotes implant growth. Anti-angiogenesis may block this critical stage. Bone turnover may release growth factors that promote tumor growth. Osteopontin and bone sialoprotein help to fix tumor cells within the matrix and recruit osteoclasts that release interleukin, tumor growth factor-, and other proteins. Cytokine feedback between the cancer and stromal cells conspires to destroy local bone and stimulate growth of the tumor deposit. Bone lysis is mediated by osteoclasts.[46] Potentially, the eccrine and autocrine loops could be interrupted, preventing further tumor growth and bone destruction. Microarray techniques show promise in identifying a metastatic genetic profile of cells that spread to bone.[47]
General Considerations

Pain is the principal symptom of bone metastases. It is composed of a biologic and a mechanical aspect. Although sometimes difficult to distinguish clinically, both must be treated. The mechanical properties of bone correlate best with the plain radiograph. Radiographs should be obtained of every symptomatic area. Lytic metastases are seen with lung and renal cancer. Mixed lytic and blastic metastases occur most often in breast cancer. Breast lesions may vary throughout the skeleton and over the course of the disease. Osteoblastic metastases are less common and typically occur with prostate and breast cancer. Bone scans are the most efficient way to screen the entire skeleton for metastatic disease. Cancers that are progressing very rapidly and overwhelm the bone reparative response may not be identified by the bone scan. MRI may be helpful diagnostically in selected cases, especially in the spine and to distinguish pathologic fracture due to osteoporosis from that caused by tumor. There is no biochemical test that is specific for bony metastases. Breakdown of bone collagen can be evaluated by measuring pyridinium or pyridinoline crosslinks. Biopsy is needed to diagnose the first bone metastasis. Special histochemical stains are helpful to define the site of origin of the unknown primary metastasis. In addition to the specific markers typically used, such as prostate-specific antigen and thyroglobulin, rare markers can be crucial in guiding treatment. For example, lymphoma can masquerade as metastatic carcinoma and can be identified only by Ki-1 stains. Biopsy or tumor defects larger than inch create stress risers that decrease torsional strength by 50% to 70%. Defects larger than the diameter of the bone are termedopen segment defects and may reduce strength by as much as 90%. The nature of a metastatic lesion affects the overall bone strength.
Treatment Goals

The goals of patient comfort and independence usually are met nonoperatively by radiation therapy and chemotherapy. Most patients without a fracture do not require surgery for the bone metastasis. Fractures are best treated by operative procedures to allow immediate weight bearing. If this cannot be achieved, then surgery should be avoided. Pathologic fractures through weight-bearing bones (e.g., femur) should be treated if the patient has more than 1 month to live and through nonweight-bearing bones if life expectancy is more than 3 months. Joint replacement and stabilization with polymethyl methacrylate (PMMA) is usually the most effective. Imaging identifies areas of impending fracture that may require prophylactic surgery to fix the bone defect,
826

prevent fracture, and maximize function. CT very effectively distinguishes between the presence of disease and structurally significant disease. Mirels and associates have proposed a graduated scoring system.[48] Four clinical and radiographic factorsanatomic site, pain pattern, radiographic nature, and

lesion sizeare each graded 0 to 3. The sum creates a composite score (0 to 12) that correlates with fracture risks ( Table 305 ). The rapid growth of metastatic cancers may overwhelm the healing response. Gainor and Buchert evaluated 129 long-bone fractures.[49] Healing occurred in 45 cases (36%). Among patients who lived 6 months or more, 50% of fractures healed. The best healing rates were in multiple myeloma (67%), kidney (44%), and breast cancer (37%). No patients with lung cancer healed their fractures. Chemotherapy effects in animal models are difficult to extrapolate to humans because of variation in dose intensity and treatment scheduling. Most studies suggest that healing is reduced 50% by common agents such as methotrexate or doxorubicin. Radiation suppresses osteoblast function and adversely affects fracture healing. Doses as low as 2000 cGy begin to interfere with normal fracture healing, and healing is very difficult to achieve after doses exceeding 5000 cGy.
Treatment: Surgery, Radiation, Systemic Therapy
Surgery

Metastatic deposits should be removed whenever possible in the course of fixing a pathologic fracture. This achieves an immediate partial response that could take weeks to achieve by other methods. Intralesional curettage is the typical method. Bleeding may be profuse but generally stops when the lesional tissue is removed back to normal bone so the tumor vessels can contract and stop bleeding. Preoperative angiography and tumor embolization greatly reduces blood flow and intraoperative hemorrhage. Adjuvants such as cryosurgery with liquid nitrogen can be helpful in treating metastatic disease. Cryosurgery extends the surgical margin and is indicated for tumors that are no longer responsive to other treatments. Radiation therapy is the principal surgical adjuvant. It should be delivered to the entire surgical field and extend the length of any internal fixation device. Surgery has an important role to play in the terminal care of cancer patients. Amputation continues to have a role in the management of metastatic cancer for (1) an unreconstructable extremity lesion, (2) treatment failures or complications such as a fungating infected lesion, and (3) intractable pain. Distal sites are most suitable for amputation to relieve symptoms and resume function. Occasionally, rhizotomy and cordotomy provide good pain relief of unilateral disease. Stabilization of a fracture requires control of the proximal and distal fracture fragments. The considerations are different for epiphyseal, metaphyseal, diaphyseal, and apophyseal locations. The different regions are discussed in sequence. Epiphyseal fractures present the easiest problem. Fracture healing is not a consideration. Resection of the epiphyseal fractures is appropriate, and arthroplasty should be performed. Metaphyseal fractures have more complex geometry and require more complex surgical decision making. Defects on the tension side of the bone may be secured by plating techniques, particularly for pathologic fractures that have a decent chance to heal or those in very sick patients. Plating techniques are more prone to failure, however, than rodding methods. Eleven percent of plated pathologic fractures fail within 7 weeks, and a cumulative 40% failed after 5 years in a series of 167 fractures. Diaphyseal lesions are best treated by intramedullary fixation, combining an intramedullary rod TABLE 30-5 -- Scoring System to Predict Rate of Pathologic Fracture Points Variable Site Pain Radiograph Size (% of shaft) Score 06 7 8 9 1012 Upper extremity Mild Blastic 033 1 Lower extremity Moderate Mixed 3467 Patients (n) 11 19 12 7 18 2 Pertrochanteric Mechanical Lytic 68100 Fracture Rate (%) 0 5 33 57 100 3

827

with interlocking screws and cement. Intramedullary rods restore flexion and bending strength beautifully and torsional strength and stiffness poorly. Closed intramedullary rodding should be reserved for fractures that will heal when stabilized and supplemented by radiation or for patients with rapidly advancing preterminal disease in whom the proximal and distal fixation will outlast the patients projected survival.
Radiation

Radiation effectively treats the pain of metastatic cancer affecting the bone. Radiation also arrests local tumor growth and permits functional improvement in most patients.[50] Seventy percent of patients have pain relief within 2 weeks, and 90% experience relief after 3 weeks. Fifty-five to 70 percent of patients who initially respond do not develop recurrent pain in the treatment field, according to the Radiation Therapy Oncology Group study. [51] [52] There is no convincing evidence that different histologic cancers respond differently to radiation. Despite its efficacy, radiation should not be used indiscriminately because it can be expensive, does not address mechanical disease, and can have complications such as marrow fibrosis and neutropenia or thrombocytopenia that could preclude chemotherapy. Radiation is best thought of as an effective trump card to be played at an appropriate occasion in the course of the cancer patients treatment for metastatic disease.
Systemic Therapy

Patients with widespread skeletal involvement require systemic therapy. Chemotherapy or endocrine therapy is the most appropriate. This is tailored to the nature of the metastasis. Monthly bisphosphonate treatment does not interfere with systemic chemotherapy and effectively prevents fractures, progression of existing bone lesions, and development of new symptomatic lesions; it is a well-established treatment of breast cancer and multiple myeloma. Although controversial, there is a rationale for extending the use of these drugs to patients with other lytic bone diseases. Despite enthusiasm for this approach among medical oncologists, there is sparse evidence that established lesions will heal during bisphosphonate therapy. It is preferable to treat most symptomatic lesions surgically, rather than have the patient endure a protracted course of bisphosphonates and crutches while waiting for undependable bone healing.[53]
Copyright 2004 Elsevier Inc. All rights reserved. www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/323.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

Selected References
Clohisy DR, Ogilvie CM, Carpenter RJ, Ramnaraine ML: Localized, tumor-associated osteolysis involves the recruitment and activation of osteoclasts. J Orthop Res 14:26, 1996. This careful study conclusively shows that all tumor-related bone loss occurs via osteoclastic bone resorption. Treatment of bone loss, be it from metastatic disease or around prostheses, is geared to block the osteoclasts. Bisphosphonate therapy is based on these observations. Malawer MM, Bickels J, Meller I, et al: Cryosurgery in the treatment of giant cell tumor: A long-term follow-up study. Clin Orthop 359:176188, 1999. Local adjuvants are very helpful when intralesional therapy is chosen to reduce the amount of local resection required to control the bone tumor. This is particularly important when it means that a joint can be spared, thereby avoiding resection and reconstruction. Epiphyseal tumors such as giant cell tumor, some low-grade chondrosarcomas, and metastases are good candidates for cryosurgery treatment. This large study shows that stable internal fixation can reduce the rate of fracture complications. Mankin HJ, Mankin CJ, Simon MA: The hazards of the biopsy, revisited. Members of the Musculoskeletal Tumor Society (see comments). J Bone Joint Surg Am 78:656663, 1996. Inappropriate or careless biopsies complicate cancer care greatly. Biopsies performed in community centers had much higher complication rates than those performed in tumor centers. The conclusion is that bone biopsies should be done by a surgeon who will provide the definitive surgical treatment. Mirels H: Metastatic disease in long bones: A proposed scoring system. Clin Orthop 249:256265, 1989. Prediction of fracture is important to optimize the care of patients with metastatic bone disease. This popular system is simple, inexpensive, and reasonably accurate. It should be used by all oncologists and surgeons who deal with metastatic bone disease until a better system is devised. Picci P, Sangiorgi L, Rougraff BT, et al: Relationship of chemotherapy-induced necrosis and surgical margins to local recurrence in osteosarcoma. J Clin Oncol 12:26992705, 1994. Effective chemotherapy and good surgery reduce the chance of local recurrence. This study documents the complex relationship between these variables. It is provocative in implying that cases that have a poor response to chemotherapy need more effective surgical treatment to avoid local recurrence. Rougraff BT, Simon MA, Kneisl JS, et al: Limb salvage compared with amputation for osteosarcoma of the distal end of the femur: A long-term oncological, functional, and qualityof-life study. J Bone Joint Surg Am 76:649656, 1994. This retrospective review is hailed as the evidence that limb preservation achieves equivalent results as amputation in the treatment of osteogenic sarcoma. Although it remains the best there is, the study is flawed by selection bias and low power, among other things. Unwin PS, Walker PS: Extendible endoprostheses for the skeletally immature. Clin Orthop 322:179193, 1996. Limb preservation in young children must take into consideration the anticipated growth of the extremities. Extendible implants have been a major advance that has allowed limb preservation to become a reality for children 6 to 10 years old. The implants suffer from high failure and revision rates, as documented in this, the largest, series.

Copyright 2004 Elsevier Inc. All rights reserved.

www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/324.html/top

Use of this content is subject to the Terms and Conditions of the MD Consult web site.

Townsend: Sabiston Textbook of Surgery, 17th ed., Copyright 2004 Elsevier

828

References
1. Mankin 2. Caffey

HJ, Lange TA, Spanier SS: The hazards of biopsy in patients with malignant primary bone and soft-tissue tumors. J Bone Joint Surg Am 64A:11211127, 1982.

J: On fibrous defects in cortical walls of growing tubular bones. Adv Pediatr 7:13, 1955. RM: Unicameral bone cysts. J Am Acad Orthop Surg 8:217224, 2000. DI, Hornicek FJ, Wolfe MW, et al: Percutaneous radiofrequency coagulation of osteoid osteoma compared with operative treatment. J Bone Joint Surg Am 80A:815821,

3. Wilkins

4. Rosenthal

1998.
5. Marco

RA, Gitelis S, Brebach GT, et al: Cartilage tumors: Evaluation and treatment. J Am Acad Orthop Surg 8:292304, 2000. A: The accuracy of imaging in the local staging of appendicular osteosarcoma. Skeletal Radiol 31:191201, 2002. M, Capanna R, Picci P: Unicameral and aneurysmal bone cysts. Clin Orthop 2536, 1986.

6. Saifuddin

7. Campanacci 8. Enneking 9. Malawer

WF: A system of staging musculoskeletal neoplasms. Clin Orthop Rel Res 204:924, 1986.

MM, Bickels J, Meller I, et al: Cryosurgery in the treatment of giant cell tumor: A long-term follow-up study. Clin Orthop Rel Res 359:176188, 1999.

10. ODonnell RJ, Springfield DS, Motwani HK, et al: Recurrence of giant-cell tumors of the long bones after curettage and packing with cement. J Bone Joint Surg Am 76:1827 1833, 1994. 11. Murata H, Kusuzaki K, Takeshita H, et al: Assessment of chemosensitivity in patients with malignant bone and soft tissue tumors using thallium-201 scintigraphy and doxorubicin binding assay. Anticancer Res 20:39673970, 2000. 12. Franzius C, Bielack S, Flege S, et al: Prognostic significance of (18)F-FDG and (99m)Tc-methylene diphosphonate uptake in primary osteosarcoma. J Nucl Med 43:10121017, 2002. 13. Bacci

G, Ferrari S, Lari S, et al: Osteosarcoma of the limb: Amputation or limb salvage in patients treated by neoadjuvant chemotherapy. J Bone Joint Surg Br 84:8892, 2002.

14. Picci P, Sangiorgi L, Rougraff BT, et al: Relationship of chemotherapy-induced necrosis and surgical margins to local recurrence in osteosarcoma. J Clin Oncol 12:26992705, 1994. 15. Minami

A, Kasashima T, Iwasaki N, et al: Vascularised fibular grafts: An experience of 102 patients. J Bone Joint Surg Br 82:10221025, 2000.

16. Gottsauner-Wolf

F, Kotz R, Knahr K, et al: Rotationplasty for limb salvage in the treatment of malignant tumors at the knee: A follow-up study of seventy patients. J Bone Joint Surg Am 73:13651375, 1991. A, Hoffmann C, Gosheger G, et al: Malignant tumor of the distal part of the femur or the proximal part of the tibia: Endoprosthetic replacement or rotationplasty: Functional outcome and quality-of-life measurements. J Bone Joint Surg Am 81:462468, 1999.

17. Hillmann

SM, Glasser DB, Lane JM, et al: Prosthetic and extremity survivorship after limb salvage for sarcoma. How long do the reconstructions last? Clin Orthop, Aug (293):280286, 1993.
19. Kawai A, Backus SI, Otis JC, et al: Interrelationships of clinical outcome, length of resection, and energy cost of walking after prosthetic knee replacement following resection of a malignant tumor of the distal aspect of the femur. J Bone Joint Surg Am 80:822831, 1998. 20. Kawai

18. Horowitz

A, Healey JH, Boland PJ, et al: A rotating-hinge knee replacement for malignant tumors of the femur and tibia. J Arthroplasty 14:187196, 1999. F, Krepler P, Dominkus M, et al: Long-term followup of uncemented tumor endoprostheses for the lower extremity. Clin Orthop 388:167177, 2001.

21. Mittermayer

22. Unwin PS, Cannon SR, Grimer RJ, et al: Aseptic loosening in cemented custom-made prosthetic replacements for bone tumours of the lower limb. J Bone Joint Surg Br 78:513, 1996. 23. Wunder JS, Leitch K, Griffin AM, et al: Comparison of two methods of reconstruction for primary malignant tumors at the knee: A sequential cohort study. J Surg Oncol 77:8999, 2001. 24. Wilkins 25. Unwin 26. Donati

RM, Soubeiran A: The Phenix expandable prosthesis: Early American experience. Clin Orthop, Jan (382):5158, 2001.

PS, Walker PS: Extendible endoprostheses for the skeletally immature. Clin Orthop, Jan (322):179193, 1996. D, Di Liddo M, Zavatta M, et al: Massive bone allograft reconstruction in high-grade osteosarcoma. Clin Orthop, Aug (377):186194, 2000. HJ, Gebhardt MC, Jennings LC, et al: Long-term results of allograft replacement in the management of bone tumors. Clin Orthop, Mar (324):8697, 1996.

27. Mankin 28. Gitelis

S, Piasecki P: Allograft prosthetic composite arthroplasty for osteosarcoma and other aggressive bone tumors. Clin Orthop Rel Res 270:197201, 1991. N, Wunder JS, Mousses S, et al: Comparison of p53 mutations in patients with localized osteosarcoma and metastatic osteosarcoma. Cancer 92:21812189, 2001.

29. Gokgoz 30. Ulaner

GA, Huang HY, Otero J, et al: Absence of a telomere maintenance mechanism as a favorable prognostic factor in patients with osteosarcoma. Cancer Res 63:17591763,

2003.

31. Sandberg AA, Bridge JA: Updates on the cytogenetics and molecular genetics of bone and soft tissue tumors: Osteosarcoma and related tumors. Cancer Genet Cytogenet 145: 1 30, 2003. 32. Bacci

G, Ferrari S, Bertoni F, et al: Histologic response of high-grade nonmetastatic osteosarcoma of the extremity to chemotherapy. Clin Orthop, May (386):186196, 2001.

33. Rougraff BT, Simon MA, Kneisl JS, et al: Limb salvage compared with amputation for osteosarcoma of the distal end of the femur: A long-term oncological, functional, and quality-of-life study. J Bone Joint Surg Am 76:649656, 1994. 34. Fuchs N, Bielack SS, Epler D, et al: Long-term results of the co-operative German-Austrian-Swiss osteosarcoma study groups protocol COSS-86 of intensive multidrug chemotherapy and surgery for osteosarcoma of the limbs. Ann Oncol 9:893899, 1998. 35. Picci

P, Sangiorgi L, Bahamonde L, et al: Risk factors for local recurrences after limb-salvage surgery for high-grade osteosarcoma of the extremities. Ann Oncol 8:899903, 1997.

36. Bielack

SS, Kempf-Bielack B, Delling G, et al: Prognostic factors in high-grade osteosarcoma of the extremities or trunk: An analysis of 1,702 patients treated on neoadjuvant cooperative osteosarcoma study group protocols. J Clin Oncol 20:776790, 2002. PA, Gorlick R, Heller G, et al: Intensification of preoperative chemotherapy for osteogenic sarcoma: Results of the Memorial Sloan-Kettering (T12) protocol. J Clin Oncol 16:24522458, 1998. AM, Schwartzentruber DJ, Devidas M, et al: Presurgical chemotherapy compared with immediate surgery and adjuvant chemotherapy for nonmetastatic osteosarcoma: Pediatric Oncology Group Study POG-8651. J Clin Oncol 21:15741580, 2003.

37. Meyers

38. Goorin

829

39. Bacci

G, Mercuri M, Longhi A, et al: Neoadjuvant chemotherapy for Ewings tumour of bone: Recent experience at the Rizzoli Orthopaedic Institute. Eur J Cancer 38:22432251,

2002.
40. Cotterill

SJ, Ahrens S, Paulussen M, et al: Prognostic factors in Ewings tumor of bone: Analysis of 975 patients from the European Intergroup Cooperative Ewings Sarcoma Study Group. J Clin Oncol 18:31083114, 2000. M, Ahrens S, Dunst J, et al: Localized Ewing tumor of bone: Final results of the cooperative Ewings Sarcoma Study CESS 86. J Clin Oncol 19:18181829, 2001.

41. Paulussen 42. Wunder

JS, Paulian G, Huvos AG, et al: The histological response to chemotherapy as a predictor of the oncological outcome of operative treatment of Ewing sarcoma. J Bone Joint Surg Am 80:10201033, 1998.

43. Bacci G, Ferrari S, Longhi A, et al: Local and systemic control in Ewings sarcoma of the femur treated with chemotherapy, and locally by radiotherapy and/or surgery. J Bone Joint Surg Br 85:107114, 2003.

T, Hillmann A, Hoffmann C, et al: Significance of surgical margin on the prognosis of patients with Ewings sarcoma. A report from the Cooperative Ewings Sarcoma Study. Cancer 78:892900, 1996.
45. Kuttesch 46. Clohisy 47. Kang

44. Ozaki

JF Jr, Wexler LH, Marcus RB, et al: Second malignancies after Ewings sarcoma: radiation dose-dependency of secondary sarcomas. J Clin Oncol 14:28182825, 1996.

DR, Ogilvie CM, Carpenter RJ, et al: Localized, tumor-associated osteolysis involves the recruitment and activation of osteoclasts. J Orthop Res 14:26, 1996.

Y, Siegel PM, Shu W, et al: A multigenic program mediating breast cancer metastasis to bone. Cancer Cell 3:537549, 2003. H: Metastatic disease in long bones: A proposed scoring system for diagnosing impending pathologic fractures. Clin Orthop, Dec (249):256264, 1989. BJ, Buchert P: Fracture healing in metastatic bone disease. Clin Orthop 178:297302, 1983.

48. Mirels 49. Gainor 50. Rose

CM, Kagan AR: The final report of the Expert Panel for the Radiation Oncology Bone Metastasis Work Group of the American College of Radiology. Int J Radiat Oncol Biol Phys 40:11171124, 1998. PH: Reanalysis of the RTOG study of the palliation of symptomatic osseous metastasis. Cancer 55:14681472, 1985. V, Powers WE, Moss WT, et al: Bone metastasis: Review and critical analysis of random allocation trials of local field treatment. Int J Radiat Oncol Biol Phys

51. Blitzer

52. Ratanatharathorn

44:118, 1999.
53. Diel IJ, Mundy GR: Bisphosphonates in the adjuvant treatment of cancer: Experimental evidence and first clinical results. International Bone and Cancer Study Group (IBCG). Br J Cancer 82:13811386, 2000.

Copyright 2004 Elsevier Inc. All rights reserved.

www.mdconsult.com

Bookmark URL: /das/book/view/41779730-2/1235/325.html/top