Introduction. The practice of cancer medicine has changed dramatically as curative treatments have been identified for many previously fatal malignancies such as testicular cancer, lymphomas, and leukemia. Adjuvant chemotherapy and hormonal therapy can extend life and prevent disease recurrence following surgical resection of localized breast, colorectal, and lung cancers. Chemotherapy is also employed as part of the multimodal treatment of locally advanced head and neck, breast, lung, and esophageal cancers, soft-tissue sarcomas, and pediatric solid tumors, thereby allowing for more limited surgery and even cure in these formerly incurable cases. Colony-stimulating factors restore bone marrow function and expand the utility of high-dose chemotherapy. Chemotherapeutic drugs are increasingly used in nonmalignant diseases: cytotoxic antitumor agents have become standard in treating autoimmune diseases, including rheumatoid arthritis (methotrexate and cyclophosphamide), Crohn's disease (6-mercaptopurine), organ transplantation (methotrexate and azathioprine), sickle cell anemia (hydroxyurea), and psoriasis (methotrexate). Despite these therapeutic successes, few categories of medication have a narrower therapeutic index and greater potential for causing harmful effects than the cytotoxic antineoplastic drugs. A thorough understanding of their pharmacology, including drug interactions and clinical pharmacokinetics, is essential for their safe and effective use in humans.
The compounds used in the chemotherapy of neoplastic disease are quite varied in structure and mechanism of action, including alkylating agents; antimetabolite analogs of folic acid, pyrimidine, and purine; natural products; hormones and hormone antagonists; and a variety of agents directed at specific molecular targets. Tables 60-1, 60-2, 60-3, 60-4, 60-5 summarize of the main classes and examples of these drugs. Figure 60–1 depicts the cellular targets of chemotherapeutic agents.
Table 60-1Alkylating Agents ||Download (.pdf) Table 60-1 Alkylating Agents
|TYPE OF AGENT ||NONPROPRIETARY NAMES ||DISEASE |
|Nitrogen mustards ||Mechlorethamine ||Hodgkin's disease |
| ||Cyclophosphamide Ifosfamide ||Acute and chronic lymphocytic leukemia; Hodgkin's disease; non-Hodgkin's lymphoma; multiple myeloma; neuroblastoma; breast, ovary, lung cancer; Wilms' tumor; cervix, testis cancer; soft-tissue sarcoma |
| ||Melphalan ||Multiple myeloma |
| ||Chlorambucil ||Chronic lymphocytic leukemia; macroglobulinemia |
|Methylhydrazine derivative ||Procarbazine N-methylhydrazine, MIH) ||Hodgkin's disease |
|Alkyl sulfonate ||Busulfan ||Chronic myelogenous leukemia, bone marrow transplantation |
|Nitrosoureas ||Carmustine (BCNU) ||Hodgkin's disease; non-Hodgkin's lymphoma; glioblastoma |
| ||Streptozocin (streptozotocin) ||Malignant pancreatic insulinoma; malignant carcinoid |
| ||Bendamustine ||Non-Hodgkin's lymphoma |
|Triazenes ||Dacarbazine (DTIC; dimethyltriazenoi midazole carboxamide), ||Malignant melanoma; Hodgkin's disease; soft-tissue - sarcomas; melanoma |
| ||Temozolomide ||Malignant gliomas |
|Platinum coordination complexes ||Cisplatin, carboplatin, oxaliplatin ||Testicular, ovarian, bladder, esophageal, lung, head and neck, colon, breast cancer |
Table 60-2Antimetabolites ||Download (.pdf) Table 60-2 Antimetabolites
|TYPE OF AGENT ||NONPROPRIETARY NAMES ||DISEASE |
|Folic acid analogs ||Methotrexate (amethopterin) ||Acute lymphocytic leukemia; choriocarcinoma; breast, head, neck and lung cancers; osteogenic sarcoma; bladder cancer |
| ||Pemetrexed ||Mesothelioma, lung cancer Breast, colon, esophageal, stomach, pancreas, head and neck; premalignant skin lesion (topical) |
|Pyrimidine analogs ||Fluorouracil (5-fluorouracil; 5-FU), capecitabine Cytarabine (cytosine arabinoside) ||Acute myelogenous and acute lymphocytic leukemia; non-Hodgkin's lymphoma |
| ||Gemcitabine ||Pancreatic, ovarian, lung cancer |
| ||5-aza-cytidine ||Myelodysplasia |
| ||Deoxy-5-aza-cytidine ||” |
|Purine analogs ||Mercaptopurine (6-mercaptopurine; 6-MP) ||Acute lymphocytic and myelogenous leukemia; small cell and related non-Hodgkin's lymphoma inhibitors |
| ||Pentostatin (2′-deoxycoformycin) ||Hairy cell leukemia; chronic lymphocytic leukemia; small cell non-Hodgkin's lymphoma |
| ||Fludarabine ||Chronic lymphocytic leukemia |
| ||Clofarabine ||Acute myelogenous leukemia |
| ||Nelarabine ||T-cell leukemia, lymphoma |
Table 60-3Natural Products ||Download (.pdf) Table 60-3 Natural Products
|TYPE OF AGENT ||NONPROPRIETARY NAMES ||DISEASE |
|Vinca alkaloids ||Vinblastine ||Hodgkin's disease; non-Hodgkin's lymphoma; testis cancer. |
| ||Vinorelbine ||Breast and lung cancer |
| ||Vincristine ||Acute lymphocytic leukemia; neuroblastoma; Wilms' tumor; rhabdomyosarcoma; Hodgkin's disease; non-Hodgkin's lymphoma |
|Taxanes ||Paclitaxel, docetaxel ||Ovarian, breast, lung, prostate, bladder, head and neck cancer |
|Epipodophyllotoxins ||Etoposide ||Testis, small cell lung and other lung cancer; breast cancer; Hodgkin's disease; non-Hodgkin's lymphomas; acute myelogenous leukemia; Kaposi's sarcoma |
| ||Teniposide ||Acute lymphoblastic leukemia in children |
|Camptothecins ||Topotecan irinotecan ||Ovarian cancer; small cell lung cancer |
| || ||Colon cancer |
|Antibiotics ||Dactinomycin (actinomycin D) ||Choriocarcinoma; Wilms' tumor; rhabdomyosarcoma; testis; |
| || ||Kaposi's sarcoma |
| ||Daunorubicin (daunomycin, rubidomycin) ||Acute myelogenous and acute lymphocytic leukemia |
| ||Doxorubicin ||Soft-tissue, osteogenic, and other sarcoma; Hodgkin's disease; non-Hodgkin's lymphoma; acute leukemia; breast, genitourinary, thyroid, lung, and stomach cancer; neuroblastoma and other childhood and adult sarcomas |
|Echinocandins ||Yondelis ||Soft-tissue sarcomas, ovarian cancer |
|Anthracenedione ||Mitoxantrone ||Acute myelogenous leukemia; breast and prostate cancer |
| ||Bleomycin ||Testis and cervical cancer; Hodgkin's disease; non-Hodgkin's lymphoma |
| ||Mitomycin C ||Stomach, anal, and lung cancer |
|Enzymes ||l-Asparaginase ||Acute lymphocytic leukemia |
Table 60-4Hormones and Antagonists ||Download (.pdf) Table 60-4 Hormones and Antagonists
|TYPE OF AGENT ||NONPROPRIETARY NAMES ||DISEASE |
|Adrenocortical ||Mitotane (o,p′-DDD) ||Adrenal cortex cancer suppressants |
|Adrenocortico-steroids ||Prednisone (other equivalent preparations available) ||Acute and chronic lymphocytic leukemia; non-Hodgkin's lymphoma; Hodgkin's disease; breast cancer, multiple myeloma |
|Progestins ||Hydroxyprogesterone caproate, medroxyprogesterone acetate, megestrol acetate ||Endometrial, breast cancer |
|Estrogens ||Diethylstilbestrol, ethinyl estradiol (other preparations available) ||Breast, prostate cancer |
|Anti-estrogens ||Tamoxifen, toremifene ||Breast cancer |
|Aromatase inhibitors ||Anastrozole, letrozole, exemestane ||Breast cancer |
|Androgens ||Testosterone propionate, fluoxymesterone (other preparations available) ||Breast cancer |
|Anti-androgen ||Flutamide, casodex ||Prostate cancer |
|GnRH analog ||Leuprolide ||Prostate cancer |
Table 60-5Miscellaneous Agents ||Download (.pdf) Table 60-5 Miscellaneous Agents
|TYPE OF AGENT ||NONPROPRIETARY NAMES ||DISEASE |
|Substituted urea ||Hydroxyurea ||Chronic myelogenous leukemia; polycythemia vera; essential thrombocytosis |
|Differentiating agents ||Tretinoin, arsenic trioxide ||Acute promyelocytic leukemia |
| ||Histone deacetylase inhibitor (vorinostat) ||Cutaneous T-cell lymphoma |
|Protein tyrosine kinase inhibitors ||Imatinib ||Chronic myelogenous leukemia; GI stromal tumors (GIST); hypereosinophilia syndrome |
| ||Dasatinib, nilotinib ||Chronic myelogenous leukemia |
| ||Gefitinib, erlotinib ||EGFR inhibitors: Non–small cell lung cancer |
| ||Sorafenib ||Hepatocellular cancer, renal cancer |
| ||Sunitinib ||GIST, renal cancer |
| ||Lapatinib ||Breast cancer |
|Proteasome inhibitor ||Bortezomib ||Multiple myeloma |
|Biological response modifiers ||Interferon-alfa, interleukin-2 ||Hairy cell leukemia; Kaposi's sarcoma; melanoma; carcinoid; renal cell; non-Hodgkin's lymphoma; mycosis fungoides; chronic myelogenous leukemia |
|Immunomodulators ||Thalidomide ||Multiple myeloma |
| ||Lenalidomide ||Myelodysplasia (5q− syndrome); multiple myeloma |
|mTOR Inhibitors Monoclonal antibodies ||Temsirolimus, everolimus ||Renal cancer (see Tables 62–1 and 62–2) |
Summary of the mechanisms and sites of action of some chemotherapeutic agents useful in neoplastic disease.
The strategy for the discovery of anticancer drugs has undergone a dramatic transformation in the past 15 years, based largely on advances in understanding the molecular basis of malignant transformation. In prior years, cancer drugs were discovered through the large-scale testing of synthetic chemicals and natural products against rapidly proliferating animal tumor systems, primarily murine leukemias (Chabner and Roberts, 2005). Most of the agents discovered in these screens interacted with DNA or its precursors, inhibiting the synthesis of new genetic material and causing broad-based damage to DNA in both normal and malignant cells. The rapidly expanding knowledge of cancer biology has led to the discovery of entirely new and more cancer-specific targets (e.g., growth factor receptors, intracellular signaling pathways, epigenetic processes, tumor vascularity, DNA repair defects, and cell death pathways). For example, in many tumors, proliferation and survival depends on the constitutive activity of a single growth factor pathway, or so-called oncogene addiction (i.e., the "Achilles heel"), and inhibition of that pathway leads to cell death (Weinstein and Joe, 2006). Thus, imatinib (gleevec) attacks the unique and specific bcr-abl translocation in chronic myelocytic leukemia. Imatinib also inhibits c-kit and produces extended control of gastrointestinal stromal tumors that express a mutated and constitutively activated form of c-kit. Monoclonal antibodies effectively inhibit tumor-associated antigens such as the amplified her-2/neu receptor in breast cancer cells. These examples emphasize that entirely new strategies for drug discovery and development, and advances in patient care, will result from new knowledge of cancer biology. Figure 60–1 outlines the common targets of cancer chemotherapeutic agents. New clinical trial designs, aimed at determining effects of new drugs at the molecular level, increasingly employ biomarkers derived from samples of biological fluids or tumors to assess the effects of these new agents on signaling pathways, tumor proliferation and cell death, and angiogenesis (Maheswaran et al., 2008). Imaging of molecular, metabolic, and physiological effects of drugs will become increasingly important in establishing that drugs effectively engage their targets.
Although molecularly targeted drugs have had outstanding successes in selected types of cancer, new therapies are not likely to replace cytotoxics in the foreseeable future. Rather, the targeted drugs and cytotoxics will continue to be used in combination. For instance, monoclonal antibodies or small targeted molecules, used as single agents against solid tumors, produce low response rates and modest benefits; however, in combination with cytotoxics and in early stages of disease, monoclonal antibodies such as trastuzumab (herceptin) and bevacizumab (avastin) are dramatically effective (Romond et al., 2005; Slamon et al., 2001). At the same time, the toxicities of cytotoxic drugs have become more manageable with the development of better anti-nausea medications (Chapters 13 and 46) and with granulocyte colony-stimulating factor and erythropoietin to restore bone marrow function (Chapters 37 and 62). Finally, targeted drugs are helping to overcome resistance to chemotherapeutic agents by normalizing blood flow, promoting apoptosis, and inhibiting pro-survival signals from growth factor pathways. Tumor angiogenesis leads to increased interstitial pressure and diminishes delivery of drugs to tumor cells; inhibitors of angiogenesis (e.g., bevacizumab) normalize blood flow and interstitial pressure (Batchelor et al., 2007), improve drug delivery, and are thus synergistic with cytotoxic drugs in the treatment of lung, breast, and other cancers.
Drug resistance remains a major obstacle to successful cancer treatment. Resistance results from a variety of pharmacokinetic and molecular changes that can defeat the best designed treatments, including poor drug absorption and delivery; genetically determined variability in drug transport, activation, and clearance; and mutations, amplifications, or deletions in drug targets. The resistance process is best understood for targeted agents. Tumors developing resistance to bcr-abl inhibitors and to inhibitors of the epidermal growth factor receptor (EGFR) express mutations in the target enzyme. Cells exhibiting drug-resistant mutations exist in the patient prior to drug treatment and are selected by drug exposure. Resistance to inhibitors of the EGFR may develop through expression of an alternative receptor, c-met, which bypasses EGFR blockade and stimulates proliferation (Engelman et al., 2008). Defects in recognition of DNA breaks and overexpression of specific repair enzymes may also contribute to resistance to cytotoxic drugs (Holleman et al., 2006), and a loss of apoptotic pathways can lead to resistance to both cytotoxic and targeted agents.
In designing specific clinical regimens, a number of factors must be considered. Drugs in combination can negate the effects of a resistance mechanism specific for a single agent, and they may be synergistic because of their biochemical interactions. Ideally, drug combinations should not overlap in their major toxicities. In general, cytotoxic drugs are used as close as possible to their maximally tolerated individual doses and should be given as frequently as tolerated to discourage tumor regrowth. Because the tumor cell population in patients with clinically detectable disease exceeds 1 g, or 109 cells, and each cycle of therapy kills <99% of the cells, it is necessary to repeat treatments in multiple, carefully timed cycles to achieve cure.
The Cell Cycle. An understanding of the life cycle of tumors is essential for the rational use of antineoplastic agents (Figure 60–2). Many cytotoxic agents act by damaging DNA. Their toxicity is greatest during the S, or DNA synthetic, phase of the cell cycle. Others, such as the vinca alkaloids and taxanes, block the formation of a functional mitotic spindle in the M phase. These agents are most effective on cells entering mitosis, the most vulnerable phase of the cell cycle. Accordingly, human neoplasms most susceptible to chemotherapeutic measures, including leukemias and lymphomas, are those having a high percentage of proliferating cells. Normal tissues that proliferate rapidly (bone marrow, hair follicles, and intestinal epithelium) are thus highly susceptible to damage from cytotoxic drugs.
Cell cycle specificity of antineoplastic agents.
Slowly growing tumors with a small growth fraction (e.g., carcinomas of the colon or non–small cell lung cancer) are less responsive to cycle-specific drugs. More effective are agents that inflict high levels of DNA damage (e.g., alkylating agents) or those that remain at high concentrations inside the cell for extended periods of time (e.g., fluoropyrimidines). The results of comparative clinical trials provide evidence for the most effective regimens for specific tumors. The clinical benefit of cytotoxic drugs has primarily been measured by radiological assessment of drug effects on tumor size; newer "targeted" agents, however, may simply slow or halt tumor growth so their effects may best be measured in the assessment of time to disease progression. More recently, there is growing interest in designing drugs that selectively kill the stem cell component of tumors because these cells are believed to be responsible for the continuous proliferation and repopulation of tumors after a toxic exposure to chemotherapy or targeted therapy. For example, bone marrow and epithelial tissues contain a compartment of normal nondividing stem cells that display resistance to cytotoxic drugs and retain the capacity to regenerate the normal cell population. Tumor stem cells exhibit the same resistance to chemotherapy, radiotherapy, and oxidative insults, and thus they may represent a significant barrier to the cure of neoplasms (Diehn et al., 2009).
Although cells from different tumors display differences in the duration of their transit through the cell cycle, and in the fraction of cells in active proliferation, all cells display a similar pattern of cell cycle progression (Figure 60-2):
a phase that precedes DNA synthesis (G1)
a DNA synthetic phase (S)
an interval following the termination of DNA synthesis (G2)
the mitotic phase (M) in which the cell, containing a double complement of DNA, divides into two daughter G1 cells
a probability of moving into a quiescent state (G0) and failing to move forward for long periods of time
At each transition point in the cell cycle, specific proteins such as p53 and chk-1 and 2, monitor the integrity of DNA and, upon detection of DNA damage, may initiate DNA repair processes or, in the presence of massive damage, direct cells down a cell death (apoptosis) pathway. Some cancer chemotherapeutic drugs act at specific phases, in the cell cycle, mainly at the S phase and M phase; other agents are cytoxic at any point in the cell cycle and are termed cell cycle phase nonspecific agents.
Each transition point in the cell cycle requires the activation of specific cyclin-dependent kinases (CDKs), which, in their active forms, couple with corresponding regulatory proteins called cyclins. The proliferative impact of CDKs is in turn dampened by inhibitory proteins such as p16. Tumor cells often exhibit changes in cell-cycle regulation that lead to relentless proliferation (e.g., mutations or loss of p16 or other inhibitory components of the so-called retinoblastoma pathway, enhanced cyclin or CDK activity). Consequently, CDKs and their effector proteins have become attractive targets for discovery of anti-neoplastic agents.
Because of the central importance of DNA to the identity and functionality of a cell, elaborate mechanisms ("checkpoints") have evolved to monitor DNA integrity. If a cell possesses normal checkpoint function, drug-induced DNA damage will activate apoptosis when the cell reaches the G1/S or G2/M boundary. If the p53 gene product or other checkpoint proteins are mutated or absent and the checkpoint function fails, damaged cells will not divert to the apoptotic pathway but will proceed through the S phase and mitosis. The cell progeny will emerge as a mutated and potentially drug-resistant population. Thus, alterations in the regulation of cell-cycle kinetics and checkpoint controls are critical factors in determining sensitivity to cytotoxic drugs and understanding the success or failure of new agents.
Achieving Therapeutic Integration and Efficacy. The treatment of cancer patients requires a skillful interdigitation of pharmacotherapy with other modalities of treatment (e.g., surgery and irradiation). Each treatment modality carries its own risks and benefits, with the potential for both antagonistic and synergistic interactions between modalities, particularly between drugs and irradiation. Furthermore, individual patient characteristics determine the choice of modalities. Not all patients can tolerate drugs, and not all drug regimens are appropriate for a given patient. Renal and hepatic function, bone marrow reserve, general performance status, and concurrent medical problems all come into consideration in making a therapeutic plan. Other less quantifiable considerations, such as the natural history of the tumor, the patient's willingness to undergo difficult and potentially dangerous treatments, and the patient's physical and emotional tolerance for side effects, enter the equation, with the goal of balancing the likely long-term gains and risks in the individual patient.
One of the great challenges of therapeutics is to adjust drug regimens to achieve a therapeutic but nontoxic outcome. Although it is customary to base dose on body surface area for individual patients, this practice is not necessarily supported by evidence in the literature (Sawyer and Ratain, 2001). Orally administered drugs are now frequently prescribed using uniform dosing for all adult patients. Dose adjustment based on renal function, on hepatic function, or on pharmacokinetic monitoring does facilitate meeting specific targets such as desired drug concentration in plasma or area under the concentration-time curve (AUC), a measure of the pattern of systemic exposure to the agent in question. Unfortunately, there are few good guidelines for adjusting dose on the basis of obesity or age. Elderly patients, particularly those >70 years of age, exhibit less tolerance for chemotherapy because of decreased renal and hepatic drug clearance, lower protein binding, and less bone marrow reserve, but individuals vary widely. Drug clearance decreases in morbidly obese patients, and dosage should probably be capped for these patients at no more than 150% of the dosage for patients of average body surface area (1.73 m2) (Rodvold et al., 1988), with adjustment upward for tolerance after each subsequent dose.
Even patients with normal renal and hepatic function exhibit significant variability in pharmacokinetics of anticancer drugs that can reduce efficacy or cause excess toxicity. The following examples illustrate the potential of using pharmacokinetic targeting to improve therapy:
The thrombocytopenia caused by carboplatin is a direct function of AUC, which in turn is determined by renal clearance of the parent drug. Calvert and Egorin (2002) have devised a formula for targeting a desired AUC based on creatinine clearance.
Monitoring of 5-fluorouracil levels in plasma allows dose adjustment to improve response rates in patients with rapid drug clearance, and to avoid toxicity in those with slow drug clearance (Gamelin et al., 2008).
High-dose methotrexate therapy requires drug-level monitoring to detect patients at high risk for renal failure and severe myelosuppression. Patients with inappropriately high concentrations of methotrexate at specific time points can be rescued from toxicity by the administration of leucovorin, and in extreme cases, by dialysis or administration of a methotrexate-cleaving enzyme and orphan drug, glucarpidase (voraxaze; recombinant carboxypeptidase G2).
Molecular Testing to Select Patients for Chemotherapy. Molecular tests are increasingly employed to identify patients likely to benefit from treatment and those at highest risk of toxicity (Roberts and Chabner, 2004). Pretreatment testing has become standard practice to select patients for hormonal therapy of breast cancer and for treatment with antibodies such as trastuzumab (her-2/neu receptor) and rituximab (CD20). The presence of a mutated k-ras gene indicates that a colorectal cancer patient's tumor will not respond to anti-EGFR antibodies (Tol et al., 2009); mutations of EGFR signal a 70% likelihood of response to erlotinib (tarceva) and gefitinib (iressa), both inhibitors of this receptor (Sequist and Lynch, 2008). Although not yet routinely employed in traditional cytotoxic therapy, molecular testing of tumors could improve outcomes by pairing patients with drugs likely to be effective against mutations that drive tumor proliferation or survival. Mutations of the b-Raf, HER 2/neu, and Alk, which are found in subsets of solid tumors in human subjects, represent examples of promising targets for solid tumor chemotherapy.
Inherited differences in protein sequence polymorphisms or levels of RNA expression influence toxicity and anti-tumor response. For example, tandem repeats in the promoter region of the gene encoding thymidylate synthase, the target of 5-fluorouracil, determine the level of expression of the enzyme. Increased numbers of repeats are associated with increased gene expression, a lower incidence of toxicity, and a decreased rate of response in patients with colorectal cancer (Pullarkat et al., 2001). Polymorphisms of the dihydropyrimidine dehydrogenase gene, the product of which is responsible for degradation of 5-fluorouracil, are associated with decreased enzyme activity and a significant risk of overwhelming drug toxicity, particularly in the rare individual homozygous for the polymorphic genes (Van Kuilenburg et al., 2002). Other polymorphisms appear to affect the clearance and therapeutic activity of cancer drugs, including tamoxifen (Schroth et al., 2007), methotrexate, irinotecan, and 6-mercaptopurine (Cheok and Evans, 2006).
Other aspects of molecular biology are entering into clinical decision making in oncology. Gene expression profiling, in which the levels of messenger RNA from thousands of genes are randomly surveyed using gene arrays, has revealed tumor profiles that are highly associated with metastasis (Ramaswamy et al., 2003). The expression of the transcription factor HOX B13 correlates with disease recurrence in patients receiving adjuvant hormonal therapy in breast cancer (Ma et al., 2004). Gene expression profiles also predict the benefit of adjuvant chemotherapy for breast cancer patients (Sotiriou and Pusztai, 2009) and the response of ovarian cancer patients to platinum-based therapy (Dressman et al., 2007).
New molecular tests and their more widespread use likely will shorten the time for drug development and approval, realize savings by avoiding the cost and toxicity of ineffective drugs, and ultimately improve patient outcome (Chabner and Roberts, 2005; Roberts and Chabner, 2004). Undoubtedly, molecular testing to select patients for specific treatments will be a cornerstone of cancer chemotherapy for years to come.
A Cautionary Note. Although advances in drug discovery and molecular profiling of tumors offer great promise for improving the outcomes of cancer treatment, a final word of caution regarding all treatment regimen deserves emphasis. The pharmacokinetics and toxicities of cancer drugs vary among individual patients. It is imperative to recognize toxicities early, to alter doses or discontinue offending medication to relieve symptoms and reduce risk, and to provide vigorous supportive care (platelet transfusions, antibiotics, and hematopoietic growth factors). Toxicities affecting the heart, lungs, or kidneys may be irreversible if recognized late in their course, leading to permanent organ damage or death. Fortunately, such toxicities can be minimized by early recognition and by adherence to standardized protocols and to the guidelines for drug use.