Etiopatogenia y factores de riesgo para desarrollar cáncer

Etiopatogenia y factores de riesgo para desarrollar cáncer

View more presentations from CFUK 22

Genética y cáncer

Genética y cáncer

View more presentations from CFUK 22

Epidemiología del cáncer

Epidemiología del cáncer

View more presentations from CFUK 22

Cáncer de Pulmón

Cáncer de pulmón

View more presentations from CFUK 22

Principios de la Terapia del Cáncer

Chapter 192 – PRINCIPLES OF CANCER THERAPY

Michael C. Perry

Diagnosis

Approach to the Patient with Cancer

Update: New Content Added Date Added: 18  August  2009

Chapter 192 - Principles of Cancer Therapy Cecil Textbook of Medicine 23rd edition is constantly being reviewed and revised by the Editors.    •    July 1, 2009: Edited and Revised and New Grade A References added    •    March 20, 2009: Edited and Revised

Few diagnoses produce such emotional responses as “cancer” or “leukemia,” and the first moments after those words are uttered often produce a storm of feelings that limit useful discussion. When the time is right, however, the physician must discuss the diagnosis, its implications, and, so far as he or she understands them, the therapeutic alternatives. It is best if other family members or close friends are also in the consulting room, both for emotional support and for another “set of ears.” It is often useful to ask “what do you understand about your diagnosis?”

If the physician is not current with the latest treatments, advance reading, for example from the National Cancer Institute's Physician Data Query, will make the interview more meaningful. Prompt referral to a specialist, whether surgical oncologist, radiation oncologist, or medical oncologist, is imperative. The generalist should not be a therapeutic nihilist unless he or she is intimately involved in the field and knows of all current therapies and clinical trials.

The consulting medical oncologist, often advised by a local tumor board comprised of medical, surgical, and radiation oncologists, will usually outline the prognosis and alternatives: standard therapy, possible clinical trials, a second opinion, or no treatment. Many oncologists actively participate in clinical trials and may have investigational drugs available or may suggest referral to a tertiary cancer center as appropriate.

Diagnostic Procedures

In most settings, a lesion has been found on physical examination or by radiographic studies, and a subsequent biopsy has confirmed the diagnosis. It is critical that the biopsy be representative of the entire tumor and that appropriate investigations (e.g., special stains, flow cytometry, cytogenetics, hormone assays) be performed before treatment is initiated. If there is a question as to whether the lesion is benign or malignant or as to its proper classification, consideration should be given to additional biopsies, and consultation with a reference pathologist may be indicated. There is seldom a need for such rapid therapy that appropriate pretreatment evaluations cannot be performed. For many tumor sites, such as the colon (Chapter 203), there is one predominant histology; in others, for example the lung (Chapter 201), the distinction between small cell lung cancer and non–small cell lung cancers is critical for treatment. For breast cancer (Chapter 208), the treating physician is interested in a variety of factors, such as histology, tumor grade, the presence or absence of estrogen and progesterone receptors, and the presence of ERBB2 (Her-2/neu) overexpression.

Staging

After a diagnosis has been established, staging is next. The American Joint Committee on Cancer (AJCC) staging system is considered the standard in the United States and is based on the T (tumor), N (node), and M (metastases) system. The approach to staging depends on the type of cancer but commonly includes plain films, such as chest radiographs, computed tomographic (CT) scans, magnetic resonance (MR) imaging scans, radionuclide scans, and, increasingly, positron emission tomography (PET) scans. These studies are typically supplemented by routine hematologic and chemistry profiles, tumor markers, and, in some cases, bone marrow aspiration and biopsies.

Treatment

Development of a Treatment Plan For cancers amenable to surgery, resection is usually the best alternative if the patient is a suitable candidate for anesthesia (Chapter 457) and otherwise is in acceptable condition in terms of concomitant or comorbid illnesses. A joint discussion among the internist, oncologist, surgeon, and anesthesiologist is often very useful in this regard. Determination of the patient's performance score (Table 192-1) is a simple means of assessing functional status. If life expectancy is limited or the patient is not a good candidate for surgery, radiation therapy is usually considered as the next “local therapy,” with chemotherapy reserved for patients whose disease is too extensive or metastatic. The increasing effectiveness of chemotherapy has resulted in its incorporation into therapy earlier, often as part of an “organ-sparing” approach. The ideal discussion with the patient should include a multidisciplinary approach, with clarification of diagnosis, prognosis, treatment goals, alternatives, side effects, and risks and benefits. Surgical Therapy Surgery is used to biopsy a suspected lesion, to remove the primary tumor, to bypass obstructions, and to provide palliation. A preoperative discussion may establish the need for placement of a venous access device at the time of surgery, thus eliminating a second procedure. Surgery remains the most common method to cure localized cancers, such as breast cancer (Chapter 208), colorectal cancer (Chapter 203), and lung cancer (Chapter 201), but it is limited by the location of the tumor, its extension, and distant metastases. Even if a tumor cannot be removed, a biopsy provides confirmation of the diagnosis. Occasionally, an obstructing lesion can be bypassed to provide palliation. Surgical staging also establishes the extent of the disease. For ovarian cancer (Chapter 209) surgical “debulking” aims to remove all visible disease, leaving minimal residual disease, to enhance chemotherapy. In rare circumstances when the primary tumor is controlled, removal of a single metastasis (“metastastectomy”) can result in long-term survival; an example is resection of a single liver metastasis found at the time of colectomy for colorectal cancer. A variety of surgical techniques, such as radiofrequency ablation or cryoablation, can treat hepatic metastases in carefully selected patients. “Adjuvant” chemotherapy is often given after surgery in this situation to treat microscopic metastases. Reconstruction after a disfiguring procedure is critical to long-term physical and emotional functioning. Examples include postmastectomy breast reconstruction (Chapter 208) and plastic surgery procedures to correct deformities after head and neck surgery (Chapter 200). Radiation Therapy Ionizing radiation (Chapter 18) can be delivered using high-energy rays, known as teletherapy, via a linear accelerator; by brachytherapy, through the application of radioactive implants, seeds, wires, or plaques; and intravenously by using radioisotopes. Radiation interacts with molecular oxygen, inducing the formation of superoxide, hydrogen peroxide, or hydroxyl radicals that damage DNA, leading to cell death. Like chemotherapy, radiation therapy is most effective against rapidly dividing cells. As “local therapies,” both surgery and radiation therapy are limited in their effectiveness by inapparent extension of disease, the location of tumors next to normal structures that must be preserved, and the presence of distant metastases. Normal tissue tolerance, which varies among the various organs and tissues, often prevents the use of doses that could uniformly eradicate cancers. Radiation therapy is also limited by tumor hypoxia: large, bulky tumors are frequently relatively “radioresistant,” whereas well-oxygenated tumors can be more effectively treated at lower doses. Radiation therapy can be used as the primary treatment, as part of multimodality therapy, in the adjuvant setting, and for palliation. As a single modality, radiation therapy can be curative for early-stage malignancies such as laryngeal cancer (Chapter 200), cervical cancer (Chapter 209), and prostate cancer (Chapter 211). Breast-conserving surgery (Chapter 208) requires the use of radiation to treat the remaining breast. Partial irradiation techniques using three-dimensional planning with external beam radiation or with a balloon catheter have recently been developed and used for selected patients with appropriately placed breast cancers. For localized prostate cancer (Chapter 211), implanted radioactive seeds of gold or palladium offer an alternative to surgery or external beam radiation therapy, again in carefully selected patients. It is important to note that the combination of chemotherapy and radiation therapy may result in synergistic toxicities, such as esophagitis (Chapter 140) in the treatment of lung cancer (Chapter 201) or mucositis in the treatment of head and neck cancer (Chapter 200). Newer techniques, such as intensity-modulated radiation therapy, permit more exact tailoring of the dose to the target and therefore reduce damage to surrounding normal tissues. Stereotactic radiation therapy or “gamma knife” techniques allow treatment of primary or metastatic brain tumors (Chapter 199) of up to 3 cm with pinpoint accuracy, thereby minimizing damage to normal brain. Low- to moderate-dose palliative radiation is used to ameliorate symptomatic cancer when cure is no longer the goal. For instance, radiotherapy can improve brain metastases (Chapter 199); relieve pain from bone lesions (Chapter 212); relieve obstructing lesions; and sometimes relieve hemoptysis caused by a lung cancer (Chapter 201) or bleeding from a gynecologic malignancy (Chapter 209). Bone-seeking radioisotopes such as samarium or strontium may relieve pain from bone metastases in prostate cancer (Chapter 211) or breast cancer (Chapter 208). Systemic Therapy Chemotherapy Pharmacogenomics, the study of inherited differences in interindividual drug disposition and effects, is becoming important in cancer therapy because genetic polymorphisms in drug-metabolizing enzymes are often responsible for the variations in efficacy and toxicity observed with many chemotherapeutic agents. Drugs potentially affected by polymorphisms identified to date include the thiopurines, 5-fluorouracil, irinotecan, and the platinum agents. In patients who are heterozygous or homozygous for metabolizing enzymes, toxicity can be dramatically enhanced. Currently available tests do not reliably permit assessment of the likelihood of response to therapy, so treatment is largely empirical and based on predictive factors from the tumor itself. Gene expression microarrays currently under development may predict responses reliably in the future. Assessing Treatment Assessment of the response to therapy depends largely on tumor size, determined either by direct measurement or from diagnostic imaging studies. The categories of response are complete response, with total absence of the tumor and correction of tumor-associated changes; partial response, defined as greater than 50% reduction of tumor size; stable disease, defined as greater than 25% but less than 50% reduction in tumor size; and progressive disease, characterized by either tumor growth or the development of new tumors. Leukemias can be assessed by bone marrow aspirates, and multiple myeloma is typically assessed by measurement of monoclonal proteins, peripheral blood counts, and percentages of malignant plasma cells in bone marrow samples, as well as radiographs of bone lesions. Chemotherapy is now used in a variety of settings without, before, with, and after surgery and radiation therapy (Table 192-2). Considerable experimental evidence suggests that cancers are most sensitive to chemotherapy during the early stages of growth, as a result of the high growth fraction and shorter cell cycle times, so that a given dose of drug will exert a greater therapeutic effect against a rapidly growing tumor than against a larger, quiescent tumor. Neoadjuvant Chemotherapy Neoadjuvant therapy, also called primary or induction chemotherapy, is used before surgery or radiation therapy to decrease the size of locally advanced cancers, thereby permitting better surgical resection, and to eradicate undetectable metastases. It also affords an opportunity to evaluate the effectiveness of treatment by histologic analysis of resected tissue. This approach is most often used for locally advanced breast cancer (Chapter 208), although other primary tumors can be targeted. Disadvantages result from the initially incomplete pathologic staging and the possibility that ineffective chemotherapy will permit the tumor to grow beyond the point of resection. Organ-sparing therapy is the use of chemotherapy, radiation therapy, or both to permit salvage of organs that otherwise would have been surgically removed if cure were the intended result. This technique is often effective in patients with cancers of the larynx (Chapter 200), esophagus (Chapter 140), bladder (Chapter 207), and anus (Chapter 203). Adjuvant Therapy Adjuvant chemotherapy is the use of chemotherapy in patients in whom the primary tumor and all evidence of cancer (e.g., regional lymph nodes), have been surgically removed or treated definitively with radiation but the risk of recurrence is thought to be high because of the presence of involved lymph nodes or certain morphologic or biologic characteristics of the cancer. Common examples include cancers of the breast (Chapter 208) and colon (Chapter 203). The typical end points of clinical chemotherapy, such as shrinkage of measurable tumor on serial radiographic studies, are not available in this situation; instead, relapse-free survival and overall survival are the principal measures of treatment effect. For an individual patient receiving adjuvant therapy, there is no means to determine whether the toxicity and expense of the therapy were beneficial or necessary, so decisions are generally based on evidence from clinical trials. Adjuvant therapy has been used in a wide variety of tumors with variable success. In the cases of breast cancer (Chapter 208) and colon cancer (Chapter 203), the number of lives saved by the use of adjuvant therapy is significant because of the large number of affected patients, despite the modest absolute differences seen between treated and control patients with current treatment programs. Resectable lung cancer (Chapter 201) has recently been added to this list. Palliative Chemotherapy Chemotherapy rarely cures cancers that remain after surgical or radiation treatment or that recur after such therapy. Pancreatic cancer (Chapter 204) is perhaps the best example of this scenario, because few patients are deemed eligible for surgery, and most have recurrent cancer after surgery. Most adult patients with recurrent or metastatic disease are considered for palliative therapy if there is no realistic chance of cure, but the potential for prolongation of useful life and/or relief of tumor-related symptoms makes such therapy reasonable. Combination Chemotherapy Virtually all the curative chemotherapy regimens developed for hematologic malignancies or solid tumors use combinations of active agents. Combination chemotherapy is also usually superior to the use of single agents in adjuvant and neoadjuvant therapy. The improved results achieved by combination chemotherapy can be explained in several ways. Resistance to any given single agent is almost always present at diagnosis, even in clinically responsive tumors. Tumors that are initially “sensitive” to chemotherapy rapidly acquire resistance to single agents, either as a result of selection of a preexisting clone of resistant tumor cells or because of an increased rate of mutation leading to drug resistance. Combination chemotherapy theoretically addresses both phenomena by providing a broader range of coverage against initially resistant clones of cells and preventing or slowing the development of resistant clones. Combination chemotherapy follows a set of principles. All drugs must be active against the tumor. All drugs must be given at an optimal dose and schedule. The drugs should have different mechanisms of antitumor activity as well as differing toxicity profiles, and the drugs should be given at consistent intervals for the shortest possible treatment time. Hormonal Therapy Endocrine or hormonal therapy for cancer, the earliest form of systemic therapy, is almost entirely limited to breast cancer (Table 192-3) and prostate cancer (Chapter 211). Many premenopausal breast cancers are thought to be under the influence of estrogens, and hormonal deprivation (ablation) may produce long-term responses in properly selected patients (those with estrogen and/or progesterone receptor positivity who have predominantly soft tissue or bone disease). This hormonal ablation may take the form of surgical removal of the ovaries, ablative radiation therapy, or the use of luteinizing hormone–releasing hormone (LHRH) antagonists. The antiestrogen, tamoxifen, is effective against breast cancer, and it may decrease the incidence of contralateral breast cancers in both premenopausal and postmenopausal women with breast cancer. It also has an estrogen-like activity that is responsible for an increased rate of endometrial cancers. Somewhat paradoxically, postmenopausal women who are candidates for hormonal therapy may also respond to tamoxifen. Aromatase Inhibitors Patients who have experienced a prolonged objective response or stable disease with hormonal therapy may be candidates for second, third-, or fourth-line hormonal therapy. However, such responses tend to become less frequent and shorter, and many patients eventually need chemotherapy. Recently, aromatase inhibitors (e.g., anastrazole, letrozole, exemestane), which decrease the conversion of metabolites in fat and muscle into estrogen, have been found to be more effective than tamoxifen as first-line therapy in both the adjuvant and metastatic settings, although the optimal schedule for tamoxifen and the aromatase inhibitors in the adjuvant setting is still under study (Chapter 208). Prostate cancer is androgen dependent, and androgen deprivation though castration or antiandrogens can produce meaningful responses. Estrogen therapy now is used infrequently because of its cardiovascular side effects and the availability of better alternatives. Once prostate cancer becomes androgen independent, second-line hormonal therapy rarely produces useful responses. Corticosteroids The corticosteroids, typically prednisone or dexamethasone, are widely used in the treatment of hematologic and oncologic cancers. In Hodgkin's disease (Chapter 197), the non-Hodgkin's lymphomas (Chapter 196), and multiple myeloma (Chapter 198), corticosteroids have antitumor activity. In solid tumors, they are used as antiemetics, rarely for the treatment of hypercalcemia of cancer (Chapter 266), and for symptomatic relief of cerebral edema in cases of central nervous system metastases (Chapter 199) or as an adjunct to radiation therapy for spinal cord metastases. Medroxyprogesterone acetate (Megace) is often used in an attempt to relieve anorexia, which is common among cancer patients. Immunotherapy Two cancers that are characterized by often unpredictable clinical behavior, melanoma (Chapter 214) and renal cell carcinoma (Chapter 207), are treated with interferon or interleukin-2 or both (Table 192-4). Dramatic responses are uncommon, and immunotherapy is only a minor component of cancer therapy. Molecularly Targeted Agents Targeted agents (Table 192-5) are drugs directed at a specific molecular point, such as a protein tyrosine kinase, or at the presence of a specific antigen on a tumor cell. Tyrosine kinase inhibitors include imatinib and erlotinib. The current best example of the success of tyrosine kinase inhibitor therapy is the dramatic response of chronic myelogenous leukemia (CML; see Chapter 195) to imatinib (Gleevec). Imatinib also has activity against gastrointestinal stromal cell tumors (Chapter 202). Erlotinib, directed against the epidermal growth factor receptor (EGFR), has antitumor effects in patients whose non–small cell lung cancers (Chapter 201) have EGFR mutations. Current research aims to identify the specific types of mutations so that patients can be prospectively selected for therapy, analogous to the measurement of estrogen receptors to select breast cancer patients for hormonal therapy. The vascular endothelial growth factor receptor (VEGFR) inhibits the formation of new blood vessels that are critical for tumor growth. Anti-VEGFR agents, such as the monoclonal antibody bevacizumab, prevent VEGF from inducing its signal in endothelial cells, thereby preventing their division. Bevacizumab has antitumor effects in metastatic colorectal cancer, non–small cell lung cancer, and breast cancer. Thalidomide inhibits angiogenesis through an unknown mechanism and is used against multiple myeloma (Chapter 198). Bortezomib (Velcade), a unique drug, is a reversible inhibitor of the proteasome pathway that normally regulates the intracellular concentration of specific proteins, thus controlling homeostasis. It has been effective in the treatment of refractory multiple myeloma (Chapter 198) and non-Hodgkin's lymphomas (Chapter 196). The development of monoclonal antibodies directed against antigens found on cancer cells represents an additional treatment modality, often complementary to conventional chemotherapy. Examples include alemtuzumab, cetuximab, rituximab, and trastuzumab. Trastuzumab has recently been shown to add significantly to disease-free survival time in patients positive for ERBB2 who receive adjuvant therapy for early-stage breast cancer (Chapter 208). These monoclonal antibodies can be used alone (“naked”) or, in some cases, labeled with a radioactive molecule to enhance cell killing. This radioimmunoconjugate approach has been most effective in the treatment of non-Hodgkin's lymphomas (Chapter 196) and chronic lymphocytic leukemia (Chapter 195). The effectiveness of monoclonal antibodies is limited by changes in the antigenic composition of neoplastic cells, called “antigenic drift.” Bone Marrow/Stem Cell Transplantation Because the major dose-limiting toxicity of most chemotherapeutic agents is myelosuppression, approaches have been developed to harvest the pluripotent stem cells found in bone marrow, peripheral blood, or, less often, cord blood before marrow-damaging chemotherapy, so that the stem cells can be reinfused later (Chapter 184). This technique is most effective for acute leukemias (Chapter 194), relapsed lymphomas (Chapter 196), and germ cell tumors (Chapter 210). The effectiveness of this approach is limited more by inability to eradicate cancer cells than by the inability to achieve engraftment. Transplants may be syngeneic (from identical twin), autologous (from self), allogeneic (from a matched donor, such as a sibling or parent), or from matched unrelated donors (MUD). Nonablative hematopoietic transplants that do not completely abolish myelopoiesis reduce toxicity and allow treatment of older and medically infirm patients. Individual Agents A general list of the currently most commonly used chemotherapeutic agents (Table 192-6) can help in understanding the key issues each raises. In all cases, the most current information from the manufacturer should be sought before therapy is initiated. The number of new drugs continues to increase, and some older drugs, now less commonly used, have been omitted. The administration of chemotherapy is best done by specifically trained individuals because of the dual risks of hypersensitivity reactions and extravasation. No doses or schedules are suggested, because these agents are often used in combination, and the doses must be reduced in many cases. End-organ function also affects dosing. The administration of chemotherapy during pregnancy (Chapter 258) is an especially difficult circumstance and requires a particularly high level of expertise. Unless otherwise specified, all chemotherapeutic agents are capable of producing some degree of nausea/vomiting, myelosuppression, alopecia, mucositis, and/or diarrhea after treatment. Most agents are also teratogenic, mutagenic, and carcinogenic, so these toxicities are not repeated for each agent. Drugs used routinely to offset agent-specific toxicities are also included in Table 192-6. Management of Complications Supportive Care Nutrition is always a concern for patients newly diagnosed with cancer, even if they have not experienced weight loss. In fact, significant weight loss is an adverse prognostic factor for several cancers, especially lung cancer. Patients are often concerned with whether their diet could have contributed to the development of the cancer (Chapter 185) and whether diet can influence the results of therapy. In most settings, neither of these scenarios is the case. Malnourished patients should be evaluated by a dietitian to determine whether they are ingesting sufficient calories and to suggest dietary supplements (Chapters 235 and 237). Some patients, such as those with head and neck cancers (Chapter 200) or esophageal cancers (Chapter 140), may require parenteral nutrition through a percutaneous endoscopic gastrostomy (PEG) tube. Total parenteral nutrition (Chapter 236) is rarely indicated, is not particularly helpful, and is likely to produce an ethical dilemma when therapy has failed and a decision to discontinue it is discussed. Corticosteroids increase appetite but have many undesirable side effects. Megestrol acetate (Megace) at a dose of 800 mg daily improves appetite and allows weight gain in many patients; it is expensive, although the suspension is less expensive than the tablets. The synthetic cannabinoid dronabinol (Marinol) stimulates appetite and reduces nausea in some patients, but it can produce dysphoria, particularly in older patients. A multiple vitamin with zinc may help with abnormal taste and provide trace minerals. Larger than recommended doses of vitamins are not helpful and may be toxic. It is always useful to inquire what over-the-counter and “alternative” medications (Chapter 36) are being contemplated or used by the patient. Symptom Management Symptom management is key to successful treatment and a patient's quality of life. Pain control (Chapter 28) can be accomplished with a variety of analgesics, both non-narcotic and narcotic. Oncologists use a 10-point scale for evaluating pain control (Fig. 192-1), and start with nonsteroidal analgesics (NSAIDs) such as aspirin and acetaminophen, progressing through ibuprofen and related drugs, through combinations of NSAIDs and narcotics to stronger narcotics. Newer narcotics are available in both short-duration and long-duration forms, with patches that last 72 hours, which are ideal for patients who have severe pain and are unable to take oral medication. Oral transmucosal fentanyl is more effective than standard release morphine in this setting.[1] Oral mucositis, a common complication of intensive therapy for hematologic malignancies, may be treated with local measures or, potentially, with recombinant human keratinocyte growth factor.[2,][3] Oral anti-Candida drugs that are absorbed or partially absorbed from the gastrointestinal tract can help prevent oral candidiasis.[4] Many patients still fear chemotherapy because of the risk of nausea and vomiting. New antiemetics, used in combination, have made this side effect much less common. Chemotherapeutic drugs can be ranked according to their probability of causing nausea and vomiting, with prophylactic treatment given accordingly. The availability of the serotonin 5-hydroxytryptamine type 3 (5-HT3) receptor antagonists (dolasetron, granisetron, ondansetron) has dramatically improved the rate of complete control of nausea and vomiting. Although prochlorperazine may be adequate for mildly emetogenic chemotherapy, more emetogenic regimens require combination therapy with a corticosteroid (usually dexamethasone), a 5-HT3 antagonist, and a benzodiazepine (e.g., lorazepam). A newer antiemetic, aprepitant, is particularly useful for the treatment of delayed nausea and vomiting. Treating patients before the development of nausea and vomiting is much more effective and helps patients adhere to their treatment schedule. Growth factors, such as granulocyte-colony stimulating factor (G-CSF) and granulocyte-monocyte colony stimulating factor (GM-CSF), permit more rapid recovery of white blood cell nadirs, thus permitting chemotherapy to be given on schedule, without reducing doses in many cases. However, such therapy does not reduce hospitalizations or improve survival.[5] It is possible to determine which individuals are at greatest risk for febrile neutropenia (Chapter 173) and treat them in advance,[6,][7] based on published guidelines. Anemia induced by chemotherapy can be alleviated, and the need for transfusions and quality of life improved, by the use of either erythropoietin (Procrit) or darbepoetin (Aranesp). A meta-analysis of randomized trials showed, however, that recombinant human erythropoiesis-stimulating agents increased the risk of death in patients with cancer by 6% to 17%.[7A] The bisphosphonates (Chapter 264), pamidronate (Aredia) or zoledronate (Zometa), are very effective, not only to treat tumor-induced hypercalcemia, but also to reduce pathologic fractures in bones with metastatic lesions, particularly from breast cancer (Chapter 208) or prostate cancer (Chapter 211) and myeloma (Chapter 198). They are also used to treat osteoporosis caused by chemotherapy-induced premature menopause in young women with breast cancer. Hematologic malignancies and chemosensitive solid tumors should be treated with prophylactic allopurinol to prevent gout and renal colic from hyperuricemia. The effects of the acute tumor lysis syndrome (hyperuricemia, hyperphosphatemia, hypocalcemia, and hyperkalemia) may be minimized by the use of allopurinol, vigorous hydration, and careful assessment of serum electrolytes.

Biologia del Cancer

Chapter 187 – BIOLOGY OF CANCER

Jeffrey A. Moscow,
Kenneth H. Cowan

GENETIC CANCER BIOLOGY

Cancer is an acquired genetic disease (Chapter 186). Spontaneous genetic changes overwhelm the mechanisms that maintain normal cellular homeostasis, disrupt the normal tight control of cell division and death, and result in the malignant phenotype.

The genetic damage that results in cancer can occur in several ways: translocations of genes can juxtapose two genes in ways that cause dysregulation of their function; mutations can activate cancer-causing genes or deactivate cancer-preventing genes; and epigenetic modifications of proteins that associate with DNA can alter the expression of critical genes (Chapter 186). Sometimes, the first step is the mutation of genes that normally prevent mutations in other genes—cells then quickly gain additional mutations that, through a morbid natural selection, ultimately produce the mutant clone that gives rise to a malignancy.

The very processes that generate malignant transformations also present obstacles to their treatment. The genetic plasticity that can manufacture cancer-causing mutations can also generate mutations that result in resistance to anticancer drugs.

Tumor formation after malignant transformation requires changes in cell biology that, in the processes of invasion and metastases, favor propagation of the malignant cells. Because individual cells, like organisms, are programmed to die, cancer cells must learn to evade the intricate systems of apoptosis that ensure cell death. Cancer cells must also enable the formation of new blood vessels to provide nutrition for the growing mass and develop strategies to escape the immune surveillance that suppresses tumor formation. These processes that give rise to malignant tumors can again raise barriers to successful therapies; for example, cells with impaired apoptosis may be resistant to anticancer drugs that kill cells through activation of apoptosis. However, the distinctive biology of malignancies can also provide opportunities for directed therapies.

Specific Mutations, Targeted Therapies

All cancer cells contain some genomic damage. In most cases, several abnormalities must occur, but sometimes even a single mutation appears sufficient to produce malignant transformation. The latter examples demonstrate clearly that cancer is a genetic disease of dysregulated growth, and they demonstrate that identification of a genetic cause can lead to a specific targeted therapy aimed at the product of the damaged gene.

In the case of acute promyelocytic leukemia (APL; Chapter 194), the characteristic t(15;17) translocation is the only identified genetic lesion. This genetic mishap splices the retinoic acid receptor (RAR) gene to another gene called PML, and the resulting hybrid RAR-PML protein does not function properly. In normal cells, RAR binds to its ligand, vitamin A, which causes the RAR to disassociate with a complex of proteins that repress transcription of certain genes and recruit a different complex of proteins that activate transcription of those genes. In APL, the hybrid RAR-PML protein has lost the capacity to activate gene transcription in response to normal levels of vitamin A. However, the functional defect of the RAR-PML hybrid protein product can be overcome with pharmacologic doses of a vitamin A analogue, all-trans-retinoic acid (ATRA). The addition of ATRA to chemotherapy for APL doubles disease-free survival, from approximately 40% to approximately 80%.

Cancers driven by single mutations appear to be restricted to specific tissue types. The t(15;17) creates malignancy in promyelocytes, but this translocation does not lead to cancer when it occurs in other tissues. Similarly, the accidental genetic recombination that creates the BCR/ABL hybrid protein results in malignant transformation of myeloid cells (Chapters 186 and 195), but this genetic alteration is found only in a limited number of nonhematologic tissues (Chapter 202). Thus, genetic alterations that lead to malignant transformation are also dependent on the cellular context in which they occur.

The Multistage Evolution of Malignancy

Most cancers do not have a single cause but rather are the products of a progression of genetic lesions that can be years in the making, as evidenced by their multiple and complex genetic alterations. The classical model of two stages of cancer development, initiation and promotion, has been replaced by a more dynamic and multistep model in which accumulated genetic damage leads to dysregulation of cell division and the disarming of the mechanisms of cell death.

Often the first step in the creation of a tumor is the development of genomic instability. During each cell division, some 3 billion nucleotide pairs must be faithfully copied to produce exact replicas in each daughter cell. The process of cancer can begin with alterations in any one of a number of factors that influence the accuracy of this process of genetic replication. Two major mechanisms create this acquired genetic damage: spontaneous mutations can disable the machinery that edits DNA replication and removes damaged genes, or, more frequently, cells are exposed to carcinogens that directly damage DNA, and the imprecise repair of damaged DNA results in an increase in spontaneous mutations.

The evidence for the loss of genomic integrity in cancer is exemplified by studies that examine the stability of DNA sequences scattered throughout the genome, that is, microsatellite repeats. These sequences have no apparent role in gene expression or regulation, but they can be viewed as a bellwether for the fidelity of genomic replication. In most tumors, the increase in microsatellite instability, or an increase in the variation of the length of these DNA sequences in tumor tissue versus normal tissues, demonstrates the loss of the ability to replicate the genome faithfully in malignant cells.

Heritable conditions that impair the proteins that ensure genomic fidelity cause a predisposition to a variety of types of cancer. For example, defects in genes that repair DNA damage (Table 187-1) cause disorders such as xeroderma pigmentosum (Chapter 462) and ataxia-telangiectasia (Chapter 271) and are associated with an increased risk of cancer. The colon cancer predisposition syndrome, hereditary nonpolyposis colon cancer (Chapter 203), is caused by inherited defects in a number of related genes that repair DNA mismatches (MSH2, MSH6, MLH1, and PMS2).

TABLE 187-1   -- GENOMIC INSTABILITY GENES ASSOCIATED WITH INCREASED HEREDITARY CANCER RISK

Genes	Syndrome	Hereditary Tumor Types
MUTYH	Attenuated polyposis	Colon
ATM	Ataxia-telangiectasia	Leukemias, lymphomas, brain
BLM	Bloom's	Leukemias, lymphomas, skin
BRCA1, BRCA2	Hereditary breast cancer	Breast, ovary
FANCA, FANCC, FANCD2, FANCE, FANCF, FANCG	Fanconi's anemia A, C, D2, E, F, and G	Leukemias
NBS1	Nijmegen breakage	Lymphomas, brain
MSH2, MLH1, MSH6, PMS2	Hereditary nonpolyposis colon cancer	Colon
XPA, XPC, ERCC2, ERCC3, ERCC4, ERCC5, DDB2	Xeroderma pigmentosum	Skin

Carcinogens are substances that create genomic alterations. As cancer is a genetic disease, carcinogens are genomic toxins that damage DNA and create mutations. The DNA damage from cigarette smoke (Chapter 30) results from exposure to dangerous hydrocarbon products that bind DNA and disrupt faithful replication. Ultraviolet irradiation causes characteristic DNA damage in the skin that can lead to melanoma and other skin cancers (Chapter 214). Ionizing radiation (Chapter 18) from diagnostic and therapeutic radiation can also cause cancer.

One key regulator of genomic integrity is the p53 protein, the product of a tumor suppressor gene that is commonly referred to as the “guardian of the genome.” The p53 protein can sense DNA damage and direct the cell either to cell cycle arrest by increased expression of the cyclin kinase inhibitor p21 (Fig. 187-1), which provides the cell with the opportunity to repair genetic damage prior to cell division, or down the path of programmed cell death (apoptosis) by increased expression of bax, a proapoptotic protein. Inactivating mutations in p53 are among the most common abnormalities observed in cancer, emphasizing the critical role of unrepaired genetic damage in the etiology of cancer. Individuals with inherited mutations of the p53 gene have the Li-Fraumeni syndrome (Chapter 186), with increased susceptibility to specific types of cancer. The retinoblastoma (RB) gene is another tumor suppressor gene whose normal function is to regulate cell growth (see Fig. 187-1). Mutations in a number of other tumor suppressor genes are also associated with specific cancer syndromes (Table 187-2).

FIGURE 187-1  P53 and DNA damage response. Following DNA damage, p53 expression induces the cyclin kinase inhibitor p21, resulting in cell cycle arrest, and bax, which induces apoptosis. Rb, retinoblastoma
TABLE 187-2   -- TUMOR SUPPRESSOR GENES ASSOCIATED WITH INCREASED HEREDITARY CANCER RISK

Genes	Syndrome	Hereditary Tumor Types
P53	Li-Fraumeni	Breast, sarcoma, adrenal, brain,
APC	Familial adenomatous polyposis	Colon, thyroid, stomach, intestine
CDHI	Familial gastric carcinoma	Stomach
VHL	von Hippel–Lindau	Kidney
WT1	Familial Wilms' tumor	Wilms'
PTEN	Cowden's	Hamartoma, glioma, uterus
CDKN2A	Familial malignant melanoma	Melanoma, pancreas
CDK4	Familial malignant melanoma	Melanoma
RB1	Hereditary retinoblastoma	Eye
NF1	Neurofibromatosis	Neurofibroma
MEN1	Multiple endocrine neoplasia type 1	Parathyroid, pituitary, islet cell,
NF2	Neurofibromatosis	Meningioma, acoustic neuromas

The relative impact of heritable predispositions and environmental exposures depends on the frequency of the genetic abnormality and its penetrance. Mutations in the breast cancer susceptibility genes BRCA1 and BRCA2 (Chapter 208) are rare, so relatively few women inherit abnormal copies of these gene alleles from their parents, and only approximately 10% of breast cancers can be attributed to BRCA1 and BRCA2 mutations. The mechanisms involved in carcinogenesis induced by BRCA1 and BRCA2 mutations are not clear. BRCA1 is involved in the regulation of gene expression, whereas both BRCA1 and BRCA2 associate with many intracellular proteins that are involved in DNA repair, including rad50, rad51, ATM, and p53 (Fig. 187-2). Cells defective in BRCA1 or BRCA2 display defects both in the response to DNA and in DNA repair.

FIGURE 187-2  Protein interactions with BRCA1 and BRCA2. BRCA1 and BRCA2 act as molecular scaffolds and promote assembly of protein complexes involved in cell cycle regulation in response to DNA damage as well as DNA repair

The penetrance of BRCA1 is high, so that an affected woman has a 60 to 85% lifetime risk for breast cancer. By comparison, other cancer susceptibility genes are both more common and less penetrant, and the interactions among these genes are complex. For example, studies in monozygotic twins estimate that inherited genes may account for almost a third of breast, colorectal, and prostate cancers, with the remaining attributable risk thought to be environmental.

DNA damage, whether from heritable conditions, from unfortunate somatic mutations in the DNA repair mechanisms themselves, or from carcinogen exposure, increases the rate of spontaneous mutations in cells and sets the stage for the natural selection of malignant clones. Although mutations of most of the approximately 25,000 genes in the genome do not result in malignancy, mutations or other disruptions of a few critical genes, which confer a selective proliferative or survival advantage, are frequently altered in many types of cancers.

Disease of the Messenger Proteins

In health, cells respond to external stimuli with complex and redundant protein networks that interact with external stimuli and transmit appropriate signals to the nucleus. The proteins involved in signal transduction include cell surface receptors, second messenger systems, and multiple transcription factors that directly regulate gene expression. Each of these elements of the signal transduction network is controlled by multiple proteins that tightly regulate the activation state of each element of the network. Disruption of any of the genes that encode the proteins involved in signal transduction—proteins that relay the external stimulus to the nucleus—occurs frequently in cancer.

Cell surface receptors for external stimuli involve families of receptor tyrosine kinases, proteins that become activated after binding to specific ligands, often as a result of homodimerization or heterodimerization and subsequent phosphorylation of cytoplasmic proteins. For example, the epidermal growth factor (ERBB) family receptor kinases are often amplified or found activated by mutations in breast, ovary, gastric, and lung cancers. Identification of genetic abnormalities in cancer cells has provided valuable targets for selective therapies. Thus, trastuzumab, a monoclonal antibody directed against the ERBB2/Her2neu protein, which is amplified and overexpressed in 25% of breast cancers, has proved to be very useful in improving disease-free survival and overall survival when given in combination with chemotherapy (Chapter 208). Erlotinib, an inhibitor of ERBB1 (epidermal growth factor [EGF]) receptor, may be effective in treatment of lung cancers that contain mutations in the ERBB1 (EGF) receptor (Chapter 201).

Unregulated growth can also result from mutations that affect downstream signal transduction networks (Fig. 187-3). Receptor tyrosine kinases transmit their signals through biochemical pathways and ultimately drive gene transcription. In some pathways, receptor tyrosine kinases activate Ras proteins, which act as second messengers that amplify and direct signals from receptor tyrosine kinases to other intracellular proteins, which then ultimately generate the cellular response to the stimulus. Ras proteins are guanosine triphosphate binding proteins that sit in the cell membrane, and their activation is tightly regulated in normal cells. Ras mutations, which are frequently found in cancer, typically result in the Ras protein being stuck in the “on,” or activated, position; instead of switching on and off in response to a stimulus, activated (mutated) Ras provides a constant, unregulated stimulus to downstream proteins, which in turn creates a cascading effect that stimulates cell growth. Although mutations in other genes are also commonly found in tumors that harbor Ras mutations, mutation of Ras alone has been shown to transform normal cells into malignant cells. Activating mutations of Ras family proteins can frequently be found in melanoma, myeloid leukemias, and cancers of the colon, pancreas, and lung. Genetic abnormalities are also frequently found in the other signaling pathways that transmit signals from receptor tyrosine kinases, including the PIP3/PDK1/Akt pathway, as well as in the nonreceptor tyrosine kinases signaling pathways, including the Jak/STAT and the c-Src pathways.

FIGURE 187-3  Growth factor–mediated signal transduction. Upon binding to specific receptor ligands, receptor tyrosine kinases activate downstream signal transduction cascades that result in changes in gene expression and enhanced cell growth. Alterations in various steps in the signal transduction cascade are frequently observed in cancer. MAP, mitogen-activated protein.

Disruption of the genes that encode the proteins that regulate gene transcription, which are the downstream targets of signal transduction cascades, can also create malignancy. For example, amplification of the Myc family of transcription factors is frequently observed in neuroblastoma and in cancers of the lung, bladder, breast, stomach, and colon. Certain acute leukemias (Chapter 194) arise when the normal gene recombination process that produces diverse immune responses goes awry and an immunoglobulin or T-cell receptor locus is accidentally spliced onto a gene encoding a transcription factor, resulting in loss of regulation of the transcription factor activity. Several pathognomonic oncogenic chromosomal translocations commonly found in sarcomas also involve genes for transcription factors; for example, the t(11;22) of Ewing's sarcoma (Chapter 212) creates the Fli1-EWS hybrid transcription factor, and the characteristic t(2;13) of alveolar rhabdomyosarcoma creates the hybrid Pax3/FKHR DNA binding protein.

Suppression of Tumor Suppressors

Normal cells contain proteins that prevent malignant transformation. Inactivation of the tumor suppressor genes that encode these proteins also leads to cancer. Because a loss of function is required for a malignant effect, both copies of the tumor suppressor gene must be affected. In almost all cases, patients have a physical loss of one copy of a gene and an acquired mutation of the other. The spontaneous deletion of genetic material, called loss of heterozygosity, is a frequently observed genetic abnormality in tumors.

Some familial cancer predisposition syndromes are based on the inheritance of one damaged copy of a tumor suppressor gene (Chapter 186). Hereditary retinoblastoma, which results from inheritance of a mutated RB gene, and the Li-Fraumeni syndrome (Chapter 186), which results from inheritance of a mutant p53 gene, predispose affected individuals to cancers. Both the RB and p53 proteins play critical roles in regulating the progression of proliferating cells through the cell cycle, and the loss of this checkpoint regulation can contribute to uncontrolled cell growth and cancer.

The importance of the p53 and RB signaling pathways in tumor suppression is further revealed by the mechanism by which infection with the human papillomavirus (HPV; Chapter 396) in cervical epithelial cells leads to cervical cancer (Chapter 209). HPV proteins E6 and E7 inactivate p53 and RB proteins, respectively, thereby creating a virally induced premalignant state in which the machinery that prevents damaged cells from proliferating has been turned off. For this reason, vaccination against certain HPV serotypes holds the promise of preventing most cases of cervical cancer.

DISTINCTIVE CANCER BIOLOGY

Cancer cells do not behave like normal cells. The alterations in the regulation of cell growth, differentiation, and death that give rise to cancer also produce common abnormal biologic characteristics, which are shared by tumors that arise from different cells of origin. These common features conspire to allow the transformed cell to grow into a tumor.

Normal cells are programmed to differentiate and ultimately to die, and this programming is regulated through enzymatic pathways that lead to terminal differentiation, senescence, or apoptosis. Cancer cells evade the mechanisms that are designed to steer cells toward terminal differentiation and senescence by altering the function of telomerase. As primitive cells divide and differentiate, the ends of chromosomes, called telomeres, progressively shorten and ultimately lead to a growth arrest that is termed replicative senescence. An enzyme called telomerase adds length back to the telomeres and reverses the process of replicative senescence. Telomerase is usually expressed at significant levels only in stem cells. However, telomerase is highly expressed in most malignant tissues, demonstrating a common alteration of cell biology that is necessary for creation and maintenance of the malignant phenotype.

In addition to bypassing senescence, cancer cells disable the pathways that lead to apoptosis. Because apoptosis is literally a life-and-death decision for the cell, it must be tightly controlled by intricate pathways of regulatory proteins. Apoptosis can be triggered through either external or internal pathways that converge to activate a family of enzymes, called caspases, that systematically degrade cellular proteins and DNA in a characteristic pattern resulting in cell death. Cancer cells contain many common aberrations in the machinery that regulates apoptosis, including increased activity of antiapoptotic proteins, such as Bcl-2 and Mcl-1, or increased levels of inhibitors of apoptosis, such as the protein survivin, which inhibits caspase activity.

Epigenetic changes in gene expression are also hallmarks of malignancy (Chapter 186). In normal cells, gene expression is controlled by epigenetic processes that limit the physical accessibility of genes to transcription factors. Gene expression can be silenced by processes that methylate specific DNA sequences on chromosomes, called CpG islands. In cancer, tumor suppressor genes are frequently found to be abnormally methylated, leading to a loss of their expression and function. The drug 5-azacytidine and the newer agent decitabine reverse methylation and have activity in the myelodysplastic syndrome (Chapter 193) and leukemia. However, drug resistance genes may also be methylated, and strategies to reverse global methylation may have uncertain effects on cellular sensitivity to chemotherapy. Gene expression is also controlled by histone acetylases, which influence how tightly genomic DNA is spooled around large complex nuclear proteins called histones and alter the interaction of chromatin proteins with DNA. A novel group of drugs called histone deacetylase inhibitors, such as suberoylanilide hydroxamic acid and depsipeptide, are currently under development as novel anticancer agents.

Cancer cells also demonstrate characteristic alterations in glucose metabolism, known as the Warburg effect. Malignant cells tend preferentially to shunt glucose into the glycolytic pathway, taking up excessive amounts of glucose and metabolizing glucose into lactate instead of channeling glucose into the typical aerobic pathway that more efficiently captures energy and creates the end products of carbon dioxide and water. This disruption of normal cellular energy metabolism is thought to be due to dysregulation of a gene that is also involved in the regulation of apoptosis, Akt, and its downstream effectors. This abnormal metabolism of glucose in cancer cells is also the principle behind the use of positron emission tomographic scans to image tumors. Inherited inactivating mutations of the phosphatase and tensin homologue (PTEN), which inactivates Akt, are also responsible for Cowden's syndrome (Chapter 186), in which the susceptibility to breast and thyroid cancers is increased.

Normal cells have tightly regulated mechanisms to degrade and recycle proteins by attaching one or more ubiquitin molecules to a protein. Ubiquitinylation can serve as a signal that a protein should be trafficked to and degraded by lysosomes or the proteasome. Ubiquitinylation plays an important role in the regulation of receptor tyrosine kinases, in cell cycle progression, and in repair of DNA damage. This regulatory mechanism is altered in many types of cancer. The first anticancer drug targeted at this abnormality in cancer cells is bortezomib, which inhibits proteasome activity and is effective in the treatment of multiple myeloma (Chapter 198).

Because tumors must find mechanisms to feed themselves as they grow, the malignant transformation must include the ability to stimulate new blood vessel formation, or angiogenesis. Malignant cells have in common the ability to stimulate the formation of endothelial cells and the breakdown of extracellular membranes, often by secreting vasculature endothelial growth factor (VEGF). The resulting tumor vasculature, although functional, does not have the vessel architecture or endothelial wall characteristics of a normal vascular bed. The realization that tumor cells have unique angiogenesis has led to the novel therapeutic approach of targeting tumor vessel formation, and not the malignant cell itself, such as with the monoclonal anti-VEGF antibody bevacizumab in metastatic colorectal cancer (Chapter 203).

Tumors also develop strategies to evade immune surveillance. The T cells and natural killer cells of the immune system play a role in protecting the host against malignancy, just as they also protect against infectious agents. Tumors can grow unimpeded after malignant cells have been selected with properties that disarm the host's immune response to tumors; these mechanisms include down-regulation of costimulatory and major histocompatibility complex molecules, as well as secretion of cytokines that inhibit the immune system, such as transforming growth factor β, interleukin-10, and signal transducer and activator of transcription 3 (STAT3). The development of therapies designed to harness the immune system for cancer treatment, including tumor antigen vaccines, has been hindered by properties that allow tumors to escape immune destruction.

Drug Resistance

Genomic instability sets the stage for the natural selection of cells with acquired genetic alterations that permit dysregulated growth. The same genetic plasticity also allows cancer cells access to the repertoire of the human genome and its capability to express and mutate any of its genes—sometimes resulting in resistance to anticancer therapy.

Cancer cells can specifically alter the target to become resistant. In the case of imatinib mesylate, the targeted therapy for chronic myeloid leukemia (Chapter 195), cells can become resistant by mutating the binding site of the drug to the already mutant BCR/ABL protein. Cancer cellular resistance to older targeted drugs such as methotrexate, which targets the enzyme dihydrofolate reductase, can be mediated by multiple steps in the folate metabolic pathway.

Cancer cells can also reach into their genomes and call upon more general mechanisms of protection from stress. Normal cells can upregulate genes to protect against environmental toxins. In cancer, malignant cells use the same proteins to evade chemotherapy. For example, when exposed to lipid-soluble anticancer drugs that diffuse through the cell membrane, cancer cells respond by increasing the expression of cell membrane proteins that can pump a wide variety of anticancer drugs out of the cell. These ATP-dependent efflux drug pumps, which create resistance to multiple drugs, include MDR1 (for multiple drug resistance) and the MRP (for multidrug resistance–related protein) family of transmembrane proteins. Cancer cells also use other detoxification pathways, such as those involving glutathione, for protection against chemotherapy.

For every anticancer drug developed, a cancer cell has found a way to circumvent it. Also, because cancer cells recruit many different physiologic mechanisms that are also used by different tissues within the host, clinical approaches to overcoming drug resistance have largely been disappointing, as drug resistance reversal strategies also reverse the protective mechanisms in normal tissues and lead to increased toxicity. In this way, the genomic plasticity of cancer cells and the rich human genomic repertoire, the very properties essential for malignant transformation, are also properties that make cancer a difficult therapeutic challenge.

Cancer Stem Cells

Tissues are composed of a vast majority of cells that are irreversibly committed toward terminal differentiation. Hidden within tissues are also a very small minority of primitive, seemingly nondescript cells that are capable of repopulating the tissue with new cells and have the potential for self-renewal, that is, the ability to divide without differentiating. Cells that possess both the ability to produce the different, specialized cells of a tissue and the additional capacity for self-renewal are called stem cells (Chapter 160). For example, a very small population of hematopoietic stem cells in the bone marrow gives rise to more committed erythroid, myeloid, and lymphocytic lineage progenitor cells that ultimately undergo terminal differentiation to produce the formed elements in blood and have the ability to repopulate the marrow, such as in a stem cell transplantation (Chapter 184).

The view of cancer as a homogeneous mass of clonally derived malignant cells has been replaced by a view that cancer cell division is organized more like a tissue in that most of the cells of a tumor do not have the capacity for self-renewal but rather are the progeny of a minute population of cancer stem cells. In hematopoietic malignancies and other tumors, flow cytometric techniques have demonstrated small, unique populations of cancer cells within tumors that have the ability to re-form the tumor, whereas the vast majority of cells in a malignancy do not have this potential.

This concept of cancer, in which tumors contain small numbers of malignant stem cells and a vast majority of malignant but biologically differentiated cells, has enormous consequences for the study and treatment of cancer. Most important, all previous studies that have examined the overall expression of genes in a malignancy may not reveal critical and unique characteristics of the cancer stem cells from which the tumor has arisen. The pattern of treatment response followed by treatment relapse may be due not only to the acquisition of drug resistance but also to a failure of the therapy to treat the distinctive biology of the cancer stem cell. Isolation and characterization of cancer stem cells derived from different malignancies may reveal new cancer stem cell–specific targets that distinguish them from both their malignant progeny and normal stem cells. Specific therapeutics aimed at cancer stem cells hold the potential for markedly improving cancer therapy.