Chapter 22: Pharmacology and Toxicity of Antineoplastic Drugs

The leukemias and lymphomas have been the initial testing ground for cancer chemotherapy. Because of their rapid rates of proliferation, lack of surgical treatment options, ready access to malignant cells, and availability of mouse models of leukemia, the hematologic malignancies drew the attention of early investigators interested in treating cancer with drugs. The first evidence for activity of a chemical antitumor agent came in 1942, from the experimental work and subsequent clinical trials conducted by Goodman, Gilman, and colleagues at Yale, and their observation that nitrogen mustard caused tumor regression in a patient with Hodgkin lymphoma.¹ Six years later, Sidney Farber, a pathologist at Children’s Hospital in Boston, made the even more startling discovery of remission induction by aminopterin and then methotrexate in acute lymphocytic leukemia (ALL). His work ushered in the modern era of chemotherapy.² Over the next 20 years, clinical trials in these diseases established the basic principles of cyclic combination therapy and dose intensification,³ developed effective strategies for management of infectious and hemorrhagic complications, and led to the cure of these diseases with chemotherapy. High-dose chemotherapy with marrow reconstitution has further extended the cure rate in leukemias and lymphomas. As our understanding of the biologic and molecular basis for malignancy has advanced, the concept of molecularly targeted therapy achieved its first striking success with the development of imatinib mesylate for chronic myelogenous leukemia (CML).⁴ Studies of relapsing patients on imatinib provided the first clear evidence for target mutation as a mechanism of clinical drug resistance.⁵ The first effective use of a monoclonal antibody, rituximab, has extended the cure rate for patients with large cell lymphomas, and the first clear demonstration of drug-induced differentiation by all-trans retinoic acid (ATRA) has led to a remarkable improvement in the cure rate of acute promyelocytic leukemia (APL).⁶ Other unique noncytotoxic drugs with unusual mechanisms of action, such as L-asparaginase, thalidomide, and bortezomib, have become valuable components of regimens for specific kinds of hematologic malignancies. Molecular studies of the abnormalities in pathways that control proliferation and survival lymphomas and leukemias have revealed distinct subsets of disease have identified new therapeutic targets.

The safe and effective use of chemotherapy in clinical practice requires a thorough understanding of the basic aspects of drug action as well as knowledge of the important clinical toxicities, pharmacokinetics, and drug interactions of the various agents. Antineoplastic chemotherapy is a complex undertaking, with the potential for serious or fatal side effects. Patients are best served if their treatment is based on evidence from clinical trials, which define optimal doses, schedules, and drug combinations. The specific protocol chosen for treatment should be appropriate not only for the stage and histology of the tumor but should consider individual patient comorbidities, age, and susceptibility to specific potential toxicities. Thus, bleomycin is not a safe choice for a patient with serious underlying renal or lung disease, nor is doxorubicin an appropriate drug for use in a patient with a history of congestive heart failure; even in patients with normal cardiac or pulmonary function, total dose limits should be respected for these agents. Even though clinical trials define the benefits and risks of a cohort of patients of a defined age range and physiology, these results may not be easily extrapolated to patients at the extreme ends of the spectrum.

Depending on the major route of drug clearance, doses should be modified for renal or hepatic dysfunction (Table 22–1). Changes in the dose and schedule of a drug (dose-dense chemotherapy) offer potentially greater antitumor effects, but often lead to unique toxicities. With the development of techniques for marrow or blood stem cell storage and replacement of marrow after chemotherapy, potentially lethal doses of chemotherapy can be administered in an attempt to cure malignancies refractory to standard regimens. In general, these regimens may produce organ toxicities not seen at conventional doses—including pneumonitis, cardiac failure, vascular endothelial damage, and hepatic and renal insufficiency—and are ordinarily reserved for patients of younger age and with normal baseline organ function.

The success of chemotherapy in curing hematologic malignancy is incompletely understood. Although targeted therapies exploit clear differences in biology conferred by mutations or amplification of key genes, an explanation for the differential effects of cytotoxic drugs on tumor versus normal tissues is less obvious. The greater susceptibility of malignant cells to drug toxicity, as reflected in the phenomenon of leukemia remission induction, with restoration of normal marrow function, may result from the relative resistance of normal marrow stem cells to drug injury. These stem cells exist in a nonreplicating phase of the cell cycle, where they are less susceptible to damage by DNA-directed agents, and they express genes that protect against chemical and hypoxic damage. In addition, there is growing evidence that cancer cells lack cell-cycle checkpoints that recognize DNA damage and activate repair of DNA strand breaks, base deletions, or other lesions induced by chemotherapy. This differential in repair capability may allow normal cells to repair damage and promote recovery from chemotherapy-induced injury.

COMBINATION CHEMOTHERAPY

Most leukemias and lymphomas are highly drug sensitive, but, with the exception of the curability of Burkitt lymphoma (treated with cyclophosphamide) and hairy cell leukemia (treated with cladribine), are rarely, if ever, cured with single-agent chemotherapy. Combination chemotherapy forestalls the emergence of drug-resistant cells and thus is curative in settings where individual agents are ineffective. Empiric principles have resulted from the clinical experience of the past 4 decades of combination therapy. In general, drugs selected for combination therapy should have demonstrable antineoplastic activity, or at least biologic effects, against the tumor in question. The lone exception may be targeted drugs that inhibit signal transduction or angiogenesis; antibodies such as trastuzumab or rituximab may exhibit limited antitumor activity on their own, but may significantly augment the action of cytotoxic agents.⁷ Individual agents in a combination should be chosen based on their different mechanisms of action and lack of a common mechanism of resistance such as multidrug resistance (MDR). The dose-limiting toxicities of the agents chosen should not overlap; otherwise, they could not be used together at or near full doses. The clinical use of specific combinations should be based on preclinical evidence of synergistic interaction, and single-agent activity in the disease in question. Favorable molecular or biochemical drug interactions may be dependent on specific schedules of administration. Pharmacokinetic interactions should be defined in initial trials of drug combinations so as to avoid under- or overdosing of individual agents.

Another important consideration in designing clinical protocols is dose intensity, the dose administered per unit time, which should be maintained throughout a treatment regimen. Achieving this objective may require the use of hematopoietic growth factors to hasten marrow recovery, prevent repeated episodes of febrile neutropenia, and allow ontime administration of the next treatment cycle.

Interdigitation of chemotherapy with surgery and irradiation makes it possible to take advantage of favorable cytokinetic or radiosensitizing effects of chemotherapy, but drugs may enhance radiation or surgical toxicity to normal organs, an interaction that requires careful consideration in designing a multidisciplinary regimen. Thus, 5-fluorouracil and cisplatin are potent radiosensitizers used with radiation therapy to enhance local tumor control in solid tumors. In the treatment of lymphomas, the toxicity of radiation therapy to sensitive organs such as skin, lung, heart, and brain may be significantly increased by concurrent administration of anthracyclines, a consideration that has prompted the use of radiation therapy either before or after anthracycline antibiotics, but not concurrently. Likewise, bleomycin sensitizes the lungs to damage by high concentration of inspired O₂ during surgery. Antiangiogenic therapies, rarely used in hematologic malignancies, are associated with an increased risk of bowel perforation in patients who have recently undergone intraabdominal surgery.

CELL KINETICS AND CANCER CHEMOTHERAPY

The cell-killing characteristics of cancer chemotherapeutic agents vary according to their mechanism of action. Many of the most effective agents in antileukemic therapy belong to the antimetabolite class, including cytarabine and methotrexate. These drugs kill cells most effectively during the DNA synthetic phase (S-phase) of the cell cycle. For these agents, a prolonged period of tumor exposure to drug is essential so as to maximize the number of cells exposed during the vulnerable period of the cell cycle. As would be predicted, the antimetabolite drugs are primarily active against rapidly dividing tumors such as acute leukemias and intermediate and high-grade lymphomas. Other anticancer drugs, such as the topoisomerase inhibitors and alkylating agents, do not require cells to be exposed during a specific phase of the cell cycle, although like the antimetabolites, these drugs are generally more effective against actively proliferating cells as compared to resting cells. Still others, most notably the nitrosoureas and busulfan, are equally toxic to dividing and nondividing cells, and at the same time, deplete marrow stem cells. In general, the toxicity of alkylating agents is determined by the total dose of drug, whereas for the cell-cycle-specific drugs (such as methotrexate and cytarabine), both drug concentration and duration of exposure determine cytocidal effect. However, for drugs that act through alternate mechanisms, such as the taxanes, myelosuppression correlates best with the duration of exposure above a threshold plasma concentration, which is approximately 50 to 100 nM for paclitaxel and 200 nM for docetaxel.⁸

High-dose regimens achieve a number of worthwhile objectives for these agents, including an enhancement of cross-membrane transport, saturation of anabolic pathways inside the cell, and prolongation of the period of effective drug concentration. However, achieving these objectives is realized at the cost of increased toxicity to normal proliferating marrow precursor cells and may produce significant and unexpected damage to normal organs, such as hepatic venoocclusive disease (alkylating agents), cerebellar toxicity (certain alkylating agents and cytarabine [ara-C]), or pulmonary toxicity (nitrosoureas and alkylating agents). Because hematopoietic stem cells can be harvested, stored, and reinfused, dose-limiting toxicities of high-dose chemotherapy are generally those affecting nonhematologic organs.

The choice of an appropriate dose and schedule of drug administration depends on a number of factors: (1) the drug’s cell-cycle dependence; (2) the often empirically derived relationship between antitumor effects, drug dose, and schedule; (3) pharmacokinetic behavior and the need to maintain a specific drug concentration for a given period of time; (4) potential interactions with other components of the treatment regimen; and (5) patient tolerance. Multiple clinical trials are required to establish safe and effective single-agent regimens and drug combinations. For molecularly targeted drugs, the aim is to maintain inhibition of the target for prolonged periods of time, keeping drug levels above a threshold for toxicity to tumor cells, but balancing these considerations against the potential for toxicity to normal tissues, such as skin, liver, and intestinal epithelium.

DRUG RESISTANCE

Inadequate treatment of a sensitive tumor tends to select for the outgrowth of drug-resistant clones of the original tumor. The reasons for emergence of drug resistance are manifold. Cancer cells often harbor basic defects in DNA repair as one of their hallmark mutations and spontaneously generate drug-resistant mutants, even in the absence of drug exposure. Thus it has been demonstrated in the specific example of imatinib treatment of CML that drug-resistant cells, carrying specific mutations in the BCR-ABL gene, can be identified in marrow prior to treatment and become the dominant tumor population under the selective pressure of drug treatment.⁵ A similar finding of pretreatment mutations explains drug resistance to inhibitors to the epidermal growth factor receptor in non–small-cell lung cancer.⁹ In addition, many cancer drugs, especially alkylating agents, and irradiation are mutagenic and increase the rate of generation of drug-resistant mutants, as demonstrated in the selection of mismatch repair mutants by temozolomide.¹⁰ To discourage the outgrowth of resistant cells, multiple agents with differing mechanisms of resistance should be used simultaneously, because the likelihood of there being a doubly or triply resistant cell is the product of the probabilities of the independent drug-resistant mutations occurring at the same time in a single cell. The probability of a cell division resulting in mutation at any given genetic locus is approximately 10^–6 for any given episode of cell division in somatic cells; thus the probability of two independent mutations arising in the same cell is 10^–12. Mutation rates may be distinctly higher in tumor cells and may be further increased by exposure to alkylating agents and irradiation. Some mutations, such as those affecting apoptosis, may confer resistance to multiple agents of diverse mechanisms of action. Thus, the probability of encountering MDR cells is much higher in reality.

In choosing drugs for combination therapy, one must bear in mind potential mechanisms of resistance. Classical MDR occurs as a consequence of increased expression of drug efflux pumps such as the P-glycoprotein or the MDR-associated proteins (MRPs),^11,12 and confers resistance to a broad spectrum of agents derived from natural products, including taxanes, anthracyclines, vinca alkaloids, and epipodophyllotoxins, and potentially to a number of “targeted” agents. Other mechanisms of resistance that induce amplification of a target gene, such as dihydrofolate reductase (DHFR)¹³ or BCR-ABL kinase,¹⁴ may be highly specific for a single drug. Table 22–2 lists the common mechanisms of resistance. Although the presence of these biochemical changes is not routinely determined in tumor biopsies prior to therapy, these mechanisms should be considered in developing new protocols and in choosing cytotoxic therapy. In relapse after targeted therapy of CML or certain non–small cell lung cancers, the choice of second-line therapies may rely on studies of tumor cell resistance, as reflected in repeat biopsies of solid tumors or cell sampling in CML.

In addition to drug-specific mechanisms of resistance, mutations that abolish recognition of DNA damage, such as the loss of components of the mismatch repair gene complex (MLH6 or MSH2)¹⁵ seem to block initiation of apoptosis by cisplatin, thiopurines, or alkylating agents. Other mutations that block the induction of apoptosis, such as loss of p53¹⁶ or overexpression of the antiapoptotic factors such as BCL-2,¹⁷ may render tumor cells insensitive to a broad array of drugs and modalities, including ionizing irradiation, alkylating agents, antimetabolites, and anthracyclines. Although the specific contribution of p53 mutation and altered apoptosis to clinical resistance is still uncertain, emerging evidence suggests that these factors are commonly associated with clinical resistance and aggressive tumor growth and may be more relevant causes of drug resistance in the clinic than are the classical drug-specific mechanisms found in experimental tumors.

The contribution of tumor stem cells to treatment resistance and disease recurrence is an intriguing, but as yet undefined, possibility. It is clear that many tissues, including marrow, contain stem cells capable of repopulating organs, even from single cells.¹⁸ Likewise, many tumors contain stem cells, which, on careful evaluation, preserve many of the surface antigens of their normal counterpart, and display resistance to DNA damage, reactive oxygen species generated by drugs or irradiation, and readily export toxic natural products.¹⁹ It is possible, but still to be established, that these drug-resistant stem cells represent the ultimate barrier to successful cancer treatment.

A number of anticancer drugs, particularly those developed during the era of cytotoxic chemotherapy, exert their antitumor effects on DNA synthesis. Cells are thus most vulnerable during periods of active DNA synthesis (S-phase), and least affected during quiescent (G₀) stages of their life cycle. Thus tumors that have a high proliferative rate, such as leukemias and aggressive lymphomas, are most vulnerable to these agents.

METHOTREXATE

Farber and associates showed that the folate antagonist aminopterin induced a complete remission in children with ALL, thereby launching the modern era of chemotherapy. Unfortunately, these remissions were short-lived, and the leukemia invariably became resistant to further treatment. Subsequently, methotrexate, a 4-amino, N-10 methyl analogue of folic acid, supplanted aminopterin because it had more predictable side effects. Methotrexate continues to be a key drug in the induction and maintenance therapy of ALL, in the intrathecal prophylaxis and treatment of CNS leukemia, in the primary treatment of CNS lymphomas, and in combination therapy of high-grade lymphomas.

Mechanisms of Action

Methotrexate enters cells through an active uptake process mediated in most tumor cells by the reduced folate transporter²⁰ and is actively effluxed from cells by the MRP class of exporters.²¹ A second uptake transporter, the membrane folate-binding protein (FBP), has lower affinity for methotrexate, but may contribute to uptake of other antifolates, such as pemetrexed. The FBP is found on many solid malignancies, and is an active target for folate analogue- and antibody-mediated drug development. A third, low pH transporter may also participate in methotrexate influx, particularly in the intestine, but its role in tumor uptake is uncertain.²² By virtue of its 4-amino substitution, methotrexate potently inhibits the enzyme DHFR, which recycles oxidized dihydrofolate to its active tetrahydrofolate state. Inhibition of DHFR leads to rapid depletion of the intracellular tetrahydrofolate coenzymes required for thymidylate and purine biosynthesis. As a result, DNA synthesis is blocked and cell replication stops. Methotrexate is retained intracellularly as a consequence of an enzymatic process that adds up to six glutamate moieties in an unusual peptide linkage to the γ-carboxyl group of the drug (Fig. 22–1). Polyglutamation is an important determinant of leukemic cell sensitivity to methotrexate. Methotrexate polyglutamates, in addition to their long persistence in cells and their potent inhibition of DHFR, have greatly increased inhibitory effects on other folate-dependent enzymes, including thymidylate synthase and enzymes that synthesize purines (Fig. 22–2). Cells that convert the drug to polyglutamates efficiently, such as leukemic myeloblasts and lymphoblasts, are more susceptible to the drug than are normal myeloid precursors, which have limited capability for polyglutamation.²³ Accumulation of polyglutamates correlates with increased cytotoxicity and treatment response in childhood lymphoblastic leukemia.²⁴ Hyperdiploid ALLs are particularly efficient in transporting methotrexate and in producing polyglutamated species, factors that may contribute to their favorable prognosis.²⁵ Polyglutamates are slowly degraded to their readily effluxed monoglutamate form by γ-glutamyl hydrolase, and a polymorphism (T127I) that deceases γ-glutamyl hydrolase activity is associated with enhanced polyglutamate accumulation in leukemic cells.²⁶ Acquired resistance to methotrexate in patients with leukemia is associated with several different alterations: increased levels of DHFR as a consequence of gene amplification,¹³ defective polyglutamation,²⁷ impaired drug uptake,²⁸ or increased efflux by the MRP class of transporters.²⁹

Clinical Pharmacology

Methotrexate is well absorbed when administered orally at low doses (5 to 10 mg/m²), but when doses exceed 30 mg/m², absorption is variable. Consequently, doses greater than 25 mg/m² should be administered parenterally.

The concentration of methotrexate in plasma declines in a polyexponential manner. A very rapid initial disposition phase persists for only a few minutes after intravenous administration. The intermediate disposition phase has a 2- to 4-hour half-life and continues for 12 to 24 hours after dosing. The terminal phase of drug decay is considerably slower, with an 8- to 10-hour half-life, and this phase becomes important in determining drug toxicity and the effectiveness of leucovorin rescue in patients treated with high-dose methotrexate. Methotrexate is primarily excreted unchanged by the kidney, while a minor fraction of the drug (7 to 30 percent) is inactivated by hepatic hydroxylation at the 7 position. Thus, doses should be reduced in proportion to the decrease in creatinine clearance (CrCl) in patients with renal impairment (CrCl <60 mL/min), because the prolonged exposure to high blood levels may result in life-threatening hematologic and gastrointestinal toxicity.³⁰ High-dose methotrexate (>0.5 g/m²) followed by leucovorin rescue is used to treat patients with high-grade lymphoma, osteosarcoma, and ALL.

In ALL, dose adjustment of methotrexate to maintain a specific area under the concentration × time (C × T) curve improves treatment outcome.³¹ Patients receiving high-dose methotrexate can be rescued from drug toxicity by administering small doses of N-10-formyltetrahydrofolate (leucovorin), which replenishes the intracellular pool of reduced folates. Leucovorin is administered intravenously or orally in doses of 10 to 15 mg/m² at 6-hour intervals, starting 6 to 24 hours after the infusion of methotrexate, and continuing until plasma concentrations of the drug fall below 1 μM. In patients receiving high-dose methotrexate, drug levels are routinely assayed 24 to 48 hours after dosing to determine the rate of drug elimination and the safety of discontinuing leucovorin. Both methotrexate and its hydroxylated metabolite are organic acids, which, like uric acid, are much more soluble in alkaline urine. In patients receiving such therapy, renal toxicity may result from intrarenal precipitation of the parent drug or its 7-OH metabolite, and is generally the primary cause of decreased drug clearance and overwhelming toxicity. Renal dysfunction can be prevented by alkalinizing the urine to pH 7 with intravenous sodium bicarbonate prior to and during therapy. Patients should be given intensive hydration, as well. If drug concentrations in plasma exceed 1 μM at 48 hours after high-dose therapy, leucovorin should be continued at higher doses of 50 to 100 mg/m² every 6 hours until methotrexate concentrations fall below 0.1 μM. The higher doses of leucovorin are necessary to compete with methotrexate for transport and polyglutamation. In cases of extreme renal failure, with stable drug levels in the 10 μM range, leucovorin will not be effective. In this setting, continuous flow hemodialysis may provide a sustained reduction in drug levels.³² An alternative effective measure in this circumstance is the administration of glucarpidase, a commercially available bacterial enzyme that instantly degrades antifolates³³ and prevents further toxicity.

Adverse Effects

The dose-limiting toxicities of methotrexate are myelosuppression and gastrointestinal toxicity. Toxic doses of methotrexate can induce thrombocytopenia and/or leukopenia, although leukopenia is more common. An early indication of methotrexate toxicity to the gastrointestinal tract is oral mucositis, whereas more severe toxicity may be manifested as diarrhea and gastrointestinal bleeding. Less-common toxic effects of methotrexate are skin rash (10 percent), pneumonitis, and chemical hepatitis. Transaminase elevations are frequently seen after high-dose methotrexate but rapidly return to normal in most patients, and without sequelae, but low-dose chronic administration, as employed to treat psoriasis or rheumatoid arthritis, may lead to portal fibrosis and cirrhosis.

Methotrexate, given intrathecally in doses of 12 mg every 4 days for children older than age 3 years and for adults, is used to prevent or treat meningeal leukemia and lymphoma. Dose adjustment is required for children younger than age 3 years, and should be made according to established protocols. Because the drug distributes poorly into the ventricular system after spinal injection, patients with active meningeal leukemia are frequently treated through an indwelling ventricular reservoir. Toxicities caused by intrathecal administration of methotrexate include acute arachnoiditis with nuchal rigidity and headache, as well as more chronic CNS toxicities, such as dementia, motor deficits, seizures, and coma.³⁴ Rarely, these neurotoxicities develop hours after intrathecal drug administration, but more commonly they occur in the days or weeks after initiation of intrathecal treatment, and are most often seen in patients with active meningeal leukemia. Leucovorin is ineffective in reversing or preventing these toxicities. Patients with such signs should undergo evaluation to rule out progressive CNS tumor, and if malignancy is not found, intrathecal cytarabine should be used for further therapy.

Methotrexate and 6-mercaptopurine (6-MP) are synergistic in their inhibition of purine biosynthesis. L-Asparaginase, an inhibitor of protein synthesis, blocks cells from entering DNA synthesis and antagonizes the effects of methotrexate, when used before the antifolate. The two drugs are not used concurrently.

Nonsteroidal antiinflammatory drugs, which diminish renal blood flow, may reduce methotrexate clearance, as may nephrotoxic antibiotics and platinum derivatives, and these or other renal toxins should be avoided in patients during high-dose methotrexate.

CYTARABINE (CYTOSINE ARABINOSIDE, ARABINOSYL CYTOSINE, ARA-C)

Ara-C is an antimetabolite analogue of cytidine, differing in the configuration at the substituent on C₂′ position of the sugar, in which the C₂′-hydroxyl group is cis-oriented relative to the C₁′-N-glycosyl bond, in contrast to the trans configuration of the ribose nucleoside. Ara-C is a mainstay in the induction of remission in patients with acute myelogenous leukemia (AML).

High doses (1 to 3 g/m²) of intravenous ara-C given at 12-hour intervals for 6 to 12 doses are more effective alone or in a combination with anthracyclines than conventional doses (100 to 150 mg/m² q12h) in consolidation therapy of AML, and they confer particular benefit in patients with cytogenetic abnormalities (t[8:21], inv[16], t[9:16], and del[16]) related to the core binding factor that regulates hematopoiesis.³⁵ Other subsets of leukemia may have increased sensitivity to ara-C. ALL patients with mixed lineage leukemia (MLL) gene translocations have upregulation of the human equilibrative nucleoside transporter (hENT) and have a greater sensitivity to ara-C.³⁶ AML patients with K-RAS gene mutations seem to derive greater benefit from high-dose ara-C than do patients with wild-type K-RAS in their tumors.³⁷

Mechanism of Action

Ara-C is converted to the nucleoside triphosphate, cytarabine triphosphate (ara-CTP) intracellularly. The first step is catalyzed by deoxycytidine kinase (dCK); polymorphisms of the dCK gene may affect the rate of activation, and ultimately response.³⁸ Ara-CTP is an inhibitor of DNA polymerase and is also incorporated into DNA, where it terminates strand elongation.³⁹ If repair is unsuccessful, apoptosis is initiated. Ara-C and its mononucleotide are deaminated and inactivated by two intracellular enzymes, cytidine deaminase and deoxycytidylate deaminase, respectively.

Acquired ara-C resistance in experimental leukemias consistently results from the loss of dCK.⁴⁰ Other changes implicated in experimental tumors include decreased drug uptake because of decreased expression of the equilibrative nucleoside transporter, increased deamination, increased pool size of competitive deoxycytidine triphosphate, and inhibition of the apoptotic pathway. Some of these changes, particularly loss of dCK activity, have been reported in studies of human leukemia, but these results have not been confirmed in definitive trials.⁴¹

Clinical Pharmacology

Ara-C is administered intravenously either as a bolus injection or, more commonly, as a continuous infusion. It is not orally bioavailable because of its degradation by cytidine deaminase, which is present in the gastrointestinal epithelium and liver. Two standard schedules of administration are used: (1) rapid infusion of 100 mg/m² every 12 hours for 7 days; or (2) continuous infusion of 100 to 200 mg/m² per day for up to 7 days. Ara-C distributes rapidly throughout total-body water and is eliminated from plasma with a biologic half-life of 7 to 20 minutes. Most of the dose is excreted as arabinosyluracil (ara-U), an inactive metabolite, which is formed in plasma, the liver, granulocytes, and other tissues. Inhibition of ara-C deamination by ara-U may be responsible for the prolongation of the biologic half-life of the drug as larger doses are administered.⁴² Single-bolus injections and short infusions (30 minutes to 1 hour duration) at doses as high as 5 g/m² produce little myelotoxicity because of the drug’s rapid clearance, whereas continuous intravenous infusion of only 1 g/m² over 48 hours produces severe marrow toxicity. High-dose ara-C (3 g/m² q12h for 3 days on days 1, 3, and 5), is routinely used for consolidation therapy of AML, but lower doses of 1 gm/m² or less should be used in patients older than 60 years to avoid CNS toxicity. Unlike most drugs, a relatively high concentration of ara-C is achieved in the cerebrospinal fluid after intravenous administration, and may approach 50 percent of the corresponding plasma concentration.

Ara-C is also used intrathecally to treat meningeal leukemia. Doses of 50 to 70 mg in adults are usually employed and afford cerebrospinal fluid levels of the drug near 1 mM, which decline with a half-life of 2 hours. Ara-C (50 mg given every 2 weeks) has been impregnated into a gel matrix, in a formulation called DepoCyt, for sustained release into the cerebrospinal fluid, thus avoiding the need for repeated spinal taps. Initial clinical results in spinal lymphomatous meningitis indicate that it has efficacy equal to that of methotrexate.⁴³

Adverse Effects

The dose-limiting toxicity for conventional dosing regimens of intravenous ara-C, 100 to 150 mg/m² per day for 5 to 10 days, is myelosuppression. Nausea and vomiting also occur at these doses, the severity of which increases markedly when higher doses are employed, although repeated administration of the drug results in some tolerance. The nadir of the white count and platelet count occurs at about days 7 to 10 after the last dose of drug. Cerebellar, gastrointestinal, and liver toxicity, as well as conjunctivitis have also been observed when high-dose regimens are used. Hepatotoxicity ranges from abnormalities in serum transaminase levels to frank jaundice. The severity of these effects increases as the duration of therapy is prolonged; however, toxic effects rapidly subside upon discontinuation of treatment. Pulmonary infiltrates as a result of noncardiogenic pulmonary edema, and occasionally associated with severe pulmonary dysfunction, occur in leukemic patients receiving ara-C, as do gastrointestinal ulcerations with bleeding and infrequently perforation. Ara-C treatment is also reported to predispose to Streptococcus viridans pneumonia.⁴⁴

In patients older than 60 years of age, and in patients with renal dysfunction, intravenous high-dose ara-C (3 g/m² every 12 hours, days 1, 3, and 5, for six doses) causes a high incidence of cerebellar toxicity, manifested as ataxia and slurred speech.⁴⁵ Confusion and dementia may supervene, leading to a fatal outcome. Cerebellar toxicity is more frequent in patients with abnormal renal function because of slowed elimination of ara-U, with consequent inhibition of ara-C deamination. Intrathecal ara-C is usually well tolerated, but neurologic side effects have been reported (seizures, alterations in mental status).

GEMCITABINE

Although primarily used for solid tumors, gemcitabine, a 2′-2′-difluoro analogue of deoxycytidine, has significant activity against Hodgkin lymphoma. Its mechanism of action is similar to ara-C, in that, as a triphosphate, it competes with deoxycytidine triphosphate for incorporation into the elongating DNA strand, where it terminates DNA synthesis. It is also self-potentiating in that at a second site of action, it inhibits ribonucleotide reductase and thereby reduces competitive pools of deoxycytidine triphosphate (dCTP). It achieves higher nucleotide levels in tumor cells than does ara-CTP, and has a longer intracellular half-life. Its clinical pharmacokinetics are determined primarily by its rapid deamination by cytidine deaminase, yielding a short plasma half-life (t_1/2) of 15 to 30 minutes. Standard schedules use 1000 mg/m² infused over 30 minutes, and produced peak drug concentrations of 20–60 μM in plasma. Longer infusion times may produce higher intracellular triphosphate concentrations, but the benefit is uncertain.⁴⁶

Resistance in solid tumors arises from low expression of hENT, increased expression of ribonucleotide reductase, and low levels of the initial activating enzyme, dCK. Gemcitabine is an extremely potent radiosensitizer and should not be used concurrently with radiation therapy except in clinical trials.

Toxicities are acute myelosuppression, mild hepatic enzyme elevations, uncommonly a reversible pneumonitis, and with prolonged usage, a progressive hemolytic uremic syndrome with capillary leak, leading to pleural effusions, ascites, and renal failure.⁴⁷

5-AZACYTIDINE AND 5-AZA-2′-DEOXYCYTIDINE

Both 5-azacytidine and decitabine (5-aza-2′-deoxycytidine), its closely related deoxy analogue, exhibit cytotoxic activity and also induce differentiation of malignant cells at low doses. The latter action results from their incorporation into DNA and their covalent inactivation of DNA methyltransferase. The resulting inhibition of methylation of cytosine bases in DNA leads to enhanced transcription of otherwise silent genes.⁴⁸ The differentiating effects of 5-azacytidine are the basis for the induction of fetal hemoglobin synthesis in patients with sickle cell anemia and thalassemia⁵³ and its approved use in low-dose therapy of myelodysplastic syndromes (MDS). The usual doses of 5-azacytidine are 75 mg/m² subcutaneously or intravenously per day for 7 days, repeated every 28 days, whereas decitabine is used in doses of 20 mg intravenously every day for 5 days every 4 weeks. Responses become apparent in myelodysplasia after two to five courses.

5-Azacytidine and decitabine are rapidly deaminated to chemically unstable uridine metabolites that immediately degrade into inactive products. Pharmacologic activity results from phosphorylation of the parent compound by cytidine kinase (for 5-azacytidine) or dCK (for decitabine), with subsequent conversion to a triphosphate nucleotide that becomes incorporated into DNA. The primary clinical toxicities of both 5-azacytidine and decitabine⁴⁹ include reversible myelosuppression, nausea and vomiting with higher doses, hepatic dysfunction, myalgia, and fever and rash. Resistance likely results from defects in drug activation or alternative mechanisms for gene silencing, such as histone methylation or acetylation.

PURINE ANALOGUES

Purine analogues (Fig. 22–3) occupy an important role in maintenance for childhood ALL, and in the past decade newer analogues have shown remarkable activity in chronic leukemias and small cell lymphomas. With methotrexate, 6-MP is a critical component in the maintenance phase of curative therapy of childhood ALL. Other purine analogues include azathioprine, a prodrug of 6-MP and potent immunosuppressive agent; allopurinol, an inhibitor of xanthine oxidase, useful in the prevention of uric acid nephropathy; 2-chlorodeoxyadenosine, effective in the treatment of hairy cell leukemia and other lymphoid malignancies; 6-thioguanine (6-TG), an infrequently used antileukemic agent; and fludarabine phosphate (2-fluoroara-adenosine monophosphate), an effective agent for chronic lymphocytic leukemia (CLL) and follicular lymphomas, and for suppression of graft-versus-host disease in transplantation. A new purine analogue, nelarabine, is an ara-guanine prodrug, with strong activity against T-cell diseases, including lymphoblastic leukemias and lymphomas.⁵⁰ The basis for this T-cell sensitivity appears to be the resistance of arabinosylguanine (ara-G) to degradation by the catabolic enzyme, purine nucleoside phosphorylase. High levels of arabinosylguanine triphosphate (ara-GTP) accumulate in T-cell neoplasms, leading to Fas ligand-mediated apoptosis. The most recent addition, clofarabine, also an adenosine analogue, has notable activity against childhood ALL and adult AML. Deoxycoformycin, a potent inhibitor of adenosine deaminase, is also effective in the treatment of T-cell malignancies and hairy cell leukemia.

Mechanism of Action of 6-Thiopurines

Both 6-MP and 6-TG have a thiol group substituted for the 6-hydroxy group of hypoxanthine or guanine, respectively, and are converted to nucleotides by hypoxanthine guanine phosphoribosyltransferase. They block synthesis of purines. The nucleotides of both 6-MP and 6-TG are incorporated into DNA, where they become methylated and are recognized by the mismatch repair system. Attempts to correct miscoding lead to strand breaks and apoptosis.⁵¹ Cell death correlates with the extent of their incorporation into DNA. 6-MP has the added effect of inhibiting de novo purine synthesis through the action of its metabolite, methyl-thioinosine monophosphate.⁵²

In experimental tumor cells, resistance to 6-MP is most commonly caused by decreased activity of hypoxanthine guanine phosphoribosyltransferase (HGPRT), by increased efflux by the transporter MRP-4, and by the absence of an effective mismatch repair process. Resistance in human leukemia is poorly understood, but is linked to HGPRT deficiency. Patients differ in their rates of metabolic clearance of 6-MP and in their ability to efflux 6-thiopurines from cells. Rapid systemic clearance of the drug, as mediated by methylation of the thiol group by 5-thiopurine-methyltransferase (TPMT),⁵³ is associated with a high leukemia recurrence rate in ALL maintenance therapy. Low levels of red blood cell thiopurine nucleotides correlate with a high level of activity of TPMT (more often found in patients of African descent) and a high risk of clinical relapse in patients with ALL,⁵⁴ whereas decreased expression of TPMT, resulting from an inherited polymorphism in the number of tandem repeats in the 5′ promoter region, is associated with increased drug toxicity. A commercial test for enzyme polymorphism, based on red cell enzyme activity or thioguanine nucleotide content, is available. A second polymorphism of significance involves the cellular efflux protein, MRP-4; an inactive variant is associated with high 6-TG nucleotide concentrations in cells, and may be responsible for great sensitivity of Japanese patients to 6-thiopurines, as the variant occurs in 18 percent of the Japanese population.⁵⁵ A polymorphism affecting the inosine triphosphate pyrophosphorylase enzyme (rs41320251) responsible for degrading a thiopurine nucleotide intermediate is associated with increased methyl-mercaptopurine nucleotides and a high incidence of febrile neutropenia in children with ALL.⁵⁶

Methotrexate and 6-MP are highly synergistic, possibly because methotrexate blocks the de novo synthesis of purines, elevates phosphoribosyl pyrophosphate (PRPP), and enhances the activation of 6-MP. 6-MP blocks warfarin anticoagulation in some patients, leading to a requirement for higher doses of warfarin in patients receiving chronic 6-MP therapy for immunosuppression.

Clinical Pharmacology of 6-Thiopurines

Both 6-TG and 6-MP are given orally at doses of 50 to 100 mg/m² per day. Oral absorption of 6-MP is erratic, as only 16 to 50 percent of an oral dose is systemically available.⁵⁷ Food and antibiotics may decrease absorption. Both 6-MP and 6-TG are inactivated by metabolism, and have half-lives of approximately 1 to 1.5 hour in plasma. Peak plasma levels of 6-MP occur 2 hours after administration and reach 1 to 2 μM. During 6-TG treatment, 6-TG nucleotides accumulate to much higher levels in leukemic cells, as compared to 6-MP. Inactive 6-thiomethyl nucleotides are almost 30-fold higher after 6-MP, than after 6-TG.⁵⁸ 6-MP is inactivated by metabolism to 6-thiouric acid, a reaction catalyzed by xanthine oxidase. Allopurinol inhibits the metabolic inactivation of 6-MP, but not of 6-TG. Therefore, it is generally recommended that dosages of orally administered 6-MP be reduced by 75 percent in patients receiving allopurinol. 6-TG is inactivated primarily by S-methylation, followed by oxidation and desulfuration, but a second pathway, mediated by guanase and xanthine oxidase, contributes to clearance. Dose reduction is not necessary when 6-TG and allopurinol are administered together.

Adverse Effects of 6-Thiopurines

Both 6-TG and 6-MP are myelotoxic, producing nadirs of white blood cells and platelets at 7 to 10 days after treatment.⁵⁹ Moderate nausea and vomiting may also be observed. Patients may experience mild but rapidly reversible hepatotoxicity after treatment with either compound. Cirrhosis has occurred in some children with leukemia who are receiving long-term therapy with 6-MP. TPMT, which inactivates 6-thiopurines, occurs in several polymorphic forms that fail to metabolize the analogues. Approximately one person in 10 of the white population is heterozygous for ineffective polymorphic forms of the enzyme and will have significantly greater myelosuppression, whereas one patient in 300 is homozygous for the inactive forms, accumulates high concentrations of thioguanine nucleotides in both tumor and normal cells, and is at risk for overwhelming toxicity, even with greatly reduced doses of 6-MP.⁶⁰

Other toxicities may include hypersensitivity reactions (fever, rash), interstitial pneumonitis; pancreatitis; opportunistic infection, and an increased incidence of AML in patients receiving chronic immunosuppressive treatment with 6-MP.

FLUDARABINE PHOSPHATE

Originally synthesized as a deamination-resistant analogue of adenosine, fludarabine phosphate contains two important substitutions: a fluorine attached to the purine ring, which renders the drug resistant to deamination, and an arabinose sugar in place of deoxyribose, which leads to its pharmacologic activity as an inhibitor of DNA synthesis and ribonucleotide reductase. It has outstanding activity in CLL.⁶¹ It is strongly immunosuppressive, like the other purine analogues, and is frequently used for this purpose in nonmyeloablative allogeneic marrow transplantation⁶² and in the treatment of autoimmune diseases.

Activation of fludarabine phosphate requires removal of the phosphate group in plasma to allow cellular uptake by nucleoside transporters, and then intracellular rephosphorylation. Fludarabine is activated to the monophosphate level by dCK. The triphosphate inhibits DNA polymerase and becomes incorporated into both DNA and RNA.⁶³ Its mechanism of cytotoxicity results from DNA chain termination and induction of apoptosis, although it also inhibits ribonucleotide reductase (RNR), a self-potentiating activity that decreases intracellular deoxyadenosine triphosphate (dATP) and increases fludarabine incorporation into DNA.⁶⁴ Its triphosphate has a long intracellular half-life of 15 hours in CLL cells. Resistance has been ascribed to decreased active uptake, a deficiency of dCK, increased efflux, or increased RNR.

The drug is available in the United States as an intravenous preparation, and for oral use. It has 60 to 80 percent bioavailability. Because it is resistant to adenosine deaminase, fludarabine is eliminated primarily by renal excretion (60 percent), with a terminal half-life of 10 hours. For patients treated with fludarabine, the standard intravenous dose is 25 mg/m² daily for 5 days, whereas the approved oral dose is 40 mg/m² daily for 5 days. In patients with renal impairment, a 20 percent dose reduction for a CrCl of 17 to 40 mL/min/m², and a 40 percent dose reduction for a CrCl less than 17 mL/min/m² yields an area under the curve approximately equal to that seen in patients with normal renal function receiving full doses of fludarabine.^65,66

When administered at these doses, fludarabine causes only moderate myelosuppression. In CLL patients, its antileukemic effect will lead to a progressive improvement in marrow function over a period of two to three cycles of treatment, with a median time to disease progression of 31 months. However, the drug also exerts cytotoxic effects against both B and T lymphocytes, lowering CD4 T-cell counts to 150 to 200 cells/μL and predisposing patients to opportunistic infections. In patients with a large tumor burden, rapid tumor lysis may rarely lead to hyperuricemia, renal failure, and hypocalcemia (tumor lysis syndrome).⁶⁷ Thus, patients should be well hydrated and their urine alkalinized prior to beginning therapy. The primary acute toxicity is reversible myelosuppression. Peripheral sensory and motor neuropathy may occur during standard-dose therapy; autoimmune phenomena, including prolonged hypothyroidism, neutropenia and hemolytic anemia with both warm and cold antibodies, have been reported.⁶⁸ Approximately 10 percent of CLL patients receiving fludarabine may develop a hypersensitivity syndrome of pulmonary infiltrates, hypoxemia, and fever, responsive to glucocorticoids.⁶⁹ Myelodysplasia and acute leukemias, with chromosome 7p deletions, have been reported as infrequent late complications.⁷⁰

CLADRIBINE (2-CHLORODEOXYADENOSINE, 2-CDA)

The extreme sensitivity of normal and malignant lymphocytes to deamination-resistant purine analogues is further exemplified by the potent activity of cladribine in hairy cell leukemia, CLL, and low-grade lymphomas.⁷¹ A single course of cladribine, typically 0.09 mg/kg per day for 7 days by continuous intravenous infusion, induces complete response in 80 percent of patients with hairy cell leukemia. Administration by subcutaneous injection or by 2-hour intravenous infusion daily for 5 days to the same total dose achieves similar results. The drug has much the same intracellular fate as fludarabine, undergoing phosphorylation by dCK and further conversion to a triphosphate that becomes incorporated into DNA. The triphosphate of cladribine has a long intracellular half-life of 9.7 hours in CLL cells isolated from patients treated with the drug.⁷² The triphosphate has multiple metabolic effects, disrupting oxidative phosphorylation in mitochondria, inhibiting RNR and depleting nicotinamide adenine dinucleotide levels in tumor cells. All of these actions may explain the drug’s toxicity to slowly dividing lymphoid malignancies such as hairy cell leukemia and CLL. The actual mechanisms by which cladribine induces DNA strand breaks are not completely understood. However, similar to fludarabine, it inhibits DNA chain extension and daughter strand synthesis.⁷³ Furthermore, the drug’s inhibition of RNR lowers levels of the competitive dATP. The cumulative effects of cladribine induce apoptosis (programmed cell death).

Cladribine is eliminated primarily (>50 percent) by renal excretion, with a terminal plasma half-life of 7 hours. In a patient with renal failure, continuous flow hemodialysis effectively cleared the drug and prevented serious myelosuppression.⁷⁴ Cladribine retains effectiveness in at least a fraction of hairy cell leukemia patients resistant to deoxycoformycin or fludarabine, although clinical experience with sequential use of these drugs is limited. Toxicities of cladribine include transient myelosuppression, fever, tumor lysis syndrome, and occasional opportunistic infections possibly related to immunosuppression. The development of cumulative thrombocytopenia during treatment with repeated courses of the drug may limit its use. Resistance develops in experimental tumors through decreased uptake, loss of the activating enzyme dCK, increased RNR activity, increased efflux,⁷⁵ or by induction of 5′-nucleotidase activity.

CLOFARABINE (2-CHLORO-2′-FLUORO-ARABINOSYLADENINE)

This analogue has halogen substitutions on both the purine ring and arabinose sugar, resulting in a ready uptake and activation, to a highly stable intracellular triphosphate (half-life of 24 hours), which terminates DNA synthesis, inhibits RNR, and induces apoptosis. The usual adult dose is 52 mg/m² given as a 2-hour infusion daily for 5 days. Clofarabine has a plasma half-life of 6.5 hours. The primary route of clearance is through renal excretion, and dose adjustment according to CrCl is recommended for patients with abnormal renal function.

Toxicities are myelosuppression; uncommonly, fever, hypotension, and pulmonary edema, suggestive of capillary leak caused by cytokine release; hepatic transaminitis; hypokalemia; and hypophosphatemia. As a single agent, the drug is well tolerated as second-line treatment for AML patients with remission rates of 30 percent.⁷⁶

NELARABINE (6-METHOXY-ARABINOSYLGUANINE)

A guanine nucleoside analogue, nelarabine has useful activity as a secondary agent for T-cell lymphoblastic lymphoma and acute T-cell leukemias. Its mode of action is similar to the other purine analogues, in that it becomes incorporated into DNA and terminates DNA synthesis. Its selective action for T cells may relate to the ability of T cells to activate purine nucleosides and the lack of susceptibility of this drug to purine nucleoside phosphorylase, a degradative reaction.

Usual doses are an intravenous 2-hour infusion of 1500 mg/m² for adults on days 1, 3, and 5, and a lower dose of 650 mg/m² per day for 5 days for children. The drug is rapidly demethylated by adenosine deaminase after administration, yielding the ara-G, which is cleared by hydrolysis and has a longer plasma half-life of 3 hours. ara-G is converted intracellularly to its triphosphate⁷⁷ which becomes incorporated into DNA. The primary toxicities are myelosuppression and abnormal liver function tests, but the drug may cause a spectrum of neurologic abnormalities, including seizures, delirium, somnolence, and the Guillain-Barré syndrome of ascending paralysis.

PENTOSTATIN (2′-DEOXYCOFORMYCIN)

Pentostatin contains a unique seven-carbon primary ring system that closely resembles the transition-state intermediate of the adenosine deaminase reaction. As such, pentostatin is a potent inhibitor of the enzyme, leading to accumulation of intracellular adenosine and deoxyadenosine nucleotides. In addition, the triphosphate of pentostatin is incorporated into DNA. The imbalance in purine nucleotide pools produced by pentostatin probably accounts for its cytotoxicity.

Although initial trials of pentostatin demonstrated striking renal and neurologic toxicities at doses of 10 mg/m² intravenously per day or greater, lower doses (4 mg/m² biweekly) are extremely effective in inducing pathologically confirmed complete responses in hairy cell leukemia. At this lower dose, severe depletion of normal T cells occurs and may predispose to opportunistic infection.⁷⁸ The optimal dose may be lower than 4 mg/m² biweekly. The drug is eliminated entirely by renal excretion, necessitating proportional dose reduction in patients with reduced CrCl.

RIBONUCLEOTIDE REDUCTASE INHIBITOR: HYDROXYUREA

Hydroxyurea inhibits RNR, the enzyme that converts ribonucleotide diphosphates to deoxyribonucleotides. It chelates iron, an essential cofactor in the RNR reaction. In malignant disease, hydroxyurea is most commonly used for treating polycythemia vera, essential thrombocythemia, and the chronic phase of CML and to lower the myeloblast count in patients presenting with AML or blastic crisis of CML. It has also become the standard agent for preventing painful crisis and reducing hospitalization in patients with sickle cell disease and in thalassemia patients with hemoglobin (Hgb) C/SS. Its antisickling activity results from induction of Hgb F through its activation of a specific promoter for the γ-globin gene. It may also exert antisickling activity and decrease occlusion of small vessels through its generation of nitric oxide, a vasodilator, and through decreased expression of adhesion molecules such as L-selectin, on neutrophils.⁷⁹ Resistance occurs in experimental tumors as a consequence of amplification of the catalytic subunit of RNR or through mutations in RNR that lower affinity for the enzyme.

Clinical Pharmacology

Hydroxyurea is well absorbed orally, even when large doses such as 50 to 75 mg/kg orally are given for rapid lowering of the white blood cell count. In chronic therapy of myeloproliferative neoplasm, starting doses of 15 mg/kg orally are adjusted upward or downward based on neutrophil counts. In managing patients with sickle cell disease, neutrophils should be maintained above 2000 per mL.⁸⁰ Hydroxyurea may also be given intravenously to rapidly lower the white blood cell count in patients with extreme leukemic leukocytosis or thrombocytosis. Peak plasma levels following oral administration are achieved at about 1 hour and decline with a half-life of 3 to 4 hours thereafter. Renal excretion is the major route of drug elimination, and doses should be decreased in proportion to the deficit in CrCl.

Adverse Effects

The major toxicities of hydroxyurea are leukopenia and the induction of megaloblastic changes. Nausea, drug fever, pneumonitis, maculopapular skin rash, and painful leg ulcers have been observed with this drug, although it is generally well tolerated. Hydroxyurea, like ara-C, is an S-phase–specific agent. Accordingly, single large doses cause little toxicity other than myelosuppression. The nadir of the leukocyte count occurs 3 to 5 days after a single dose of drug, and the leukocyte count recovers rapidly. It is a potent teratogen and should not be used in women of childbearing age. Its potential to cause leukemic transformation is uncertain, but small cases series suggest this may occur in patients with a myeloproliferative neoplasm.⁸¹

VINCA ALKALOIDS

Vinblastine and vincristine are commonly used in the treatment of hematologic neoplasms: vinblastine because of its excellent activity in the treatment of Hodgkin lymphoma and vincristine in lymphomas and childhood leukemia. Both drugs have activity in solid-tumor therapy, particularly in treating childhood sarcomas (vincristine), and testicular cancer (vinblastine).

Mechanism of Action

The vinca alkaloids exert their cytotoxic action by their binding to tubulin, a structural protein found in the cytoplasm of cells. Microtubules, assembled through polymerization of tubulin dimers, form the spindle along which the chromosomes migrate during mitosis. Microtubules are an important structural component of neuronal axons. Binding of the vinca alkaloids to tubulin leads to inhibition of formation of the mitotic spindle,⁸² arresting cells in metaphase and inducing apoptosis. Resistance to the vinca alkaloids may be acquired through the expression of the MDR efflux pump. Alternatively, resistant cells may contain mutant tubulin with decreased avidity of vinca binding.⁸³ The clinical importance of these resistance mechanisms, however, is uncertain.

Clinical Pharmacology

Vincristine and vinblastine are both administered by the intravenous route. The average single dose of vincristine is 1.4 mg/m² and that of vinblastine 8 to 9 mg/m². Sequential doses of the drugs are usually given at 1- or 2-week intervals. These doses provide peak plasma drug concentrations of approximately 1 μM. The plasma pharmacokinetics of both vinca analogues are characterized by a very rapid initial disposition phase followed by a slow terminal phase of decay, with half-lives of 20 to 85 hours. Almost 70 percent of a dose of vincristine is metabolized by the liver and excreted in the feces. Cytochrome P450 (CYP)-mediated metabolism is also the major route of inactivation of vinblastine, producing a variety of inactive metabolic produced in the liver are excreted in the bile. Inducers of CYP3A4, such as phenylhydantoin, enhance clearance, while inhibitors delay clearance and increase toxicity. The dose of vincristine or vinblastine should be reduced in patients with hepatic impairment. Although specific guidelines for dose reduction have not been developed, a 50 percent decrease in dose is recommended for patients presenting with a bilirubin level of 1.5 to 3 mg/dL and a 75 percent reduction for levels greater than 3 mg/dL. Dose reduction is not necessary for patients with impaired renal function, as very little intact drug is excreted in urine.

Adverse Effects

The dose-limiting side effect of vincristine is neurotoxicity, which usually occurs when the total dose received exceeds 6 mg/m². The initial signs of neurotoxicity are paresthesia of the fingers and lower extremities and loss of deep tendon reflexes. Continued administration may lead to profound loss of motor strength, such as weakness of dorsiflexion of the foot and extension of the wrists. Elderly patients are particularly susceptible to such toxicities. Occasionally, cranial nerve palsies may lead to vocal cord paralysis or diplopia, and severe jaw pain may result from vincristine administration. At high doses of vincristine (>3 mg total single dose), autonomic neuropathy may cause obstipation and paralytic ileus. Sensory changes and reflex abnormalities slowly improve when the drug is discontinued; motor impairment improves less rapidly and may be irreversible. Inappropriate antidiuretic hormone release resulting in symptomatic dilutional hyponatremia has been ascribed to vincristine.

While marrow suppression is not common with vincristine administration, myelosuppression may be noted in patients with impaired marrow function as a consequence of prior treatment with other drugs. Platelet counts are relatively unaffected.

The primary toxicity of vinblastine is leukopenia. The white count reaches a nadir at day 7 and reverses rapidly thereafter. Mucositis may result from higher doses (>8 mg/m²) of vinblastine or when it is used in combination with other cytotoxic drugs. Neurotoxicity is rare, but ileus can occur at high doses.

Both drugs cause severe pain and local toxicity if extravasated. Neither drug should ever be given intrathecally. Vincristine administered inadvertently into the cerebrospinal fluid causes acute neurologic dysfunction, coma, and death. Attempts at replacement of the cerebral spinal fluid with an electrolyte solution, Ringer lactate, supplemented with 15 ml/L of fresh-frozen plasma, have been reported to avert a fatal outcome, but do not prevent severe neurologic sequelae.⁸⁴

TAXANES

The taxanes, paclitaxel, docetaxel, and Abraxane, are a second class of antimitotic compounds that differ in mechanism and toxicity profile from the vinca alkaloids that are primarily used in patients with solid tumors. Paclitaxel was purified from an extract of the bark of Taxus brevifolia, whereas docetaxel is a closely related semisynthetic derivative. Abraxane is paclitaxel embedded in an albumin microparticle. The taxanes bind to the β-tubulin subunit of microtubules and promote the polymerization of microtubules, leading to disordered mitotic spindle formation and a block in the progression through mitosis.⁸⁵ They induce apoptosis in tumor cells irrespective of the p53 status of the cells and kill cells at 10 nM concentrations or less in cell culture in a time-dependent manner.⁸⁶ In experimental settings, resistance is related to increased drug efflux, mutations in β-tubulin, or increased expression of antiapoptotic proteins such as survivin,⁸⁷ or of the mitosis-related aurora kinase.⁸⁸

The taxanes are subject to MDR mediated by the mdr and mrp genes, as well as to β-tubulin mutations. Because they are highly insoluble in aqueous solution, paclitaxel and docetaxel are formulated in lipid-­based solvents that cause occasional hypersensitivity reactions. Thus, paclitaxel is given after pretreatment with antihistamines (cimetidine, diphenhydramine), and dexamethasone. Both drugs are cleared primarily by hepatic CYP metabolism, although by different isoenzymes (paclitaxel predominantly by CYP2B6 and docetaxel by CYP3A4) with terminal plasma half-lives of 10 to 13 hours. Their metabolism is stimulated by phenytoin and other CYP-inducing drugs and inhibited by ketoconazole. Their major toxicities, aside from hypersensitivity, are a sharp but brief leukopenia, milder thrombocytopenia, and mucositis. High-dose or repeated cycles of the taxanes cause a sensory and motor peripheral neuropathy that is reversible with drug discontinuation. Occasional patients have experienced atrial conduction block or atrial or ventricular arrhythmias after paclitaxel administration, and the combination of paclitaxel with doxorubicin may produce a greater incidence of congestive heart failure than seen with doxorubicin alone.⁸⁹ A syndrome of progressive fluid retention and peripheral edema occurs in patients receiving multiple cycles of docetaxel and can be at least partially prevented by pretreatment with glucocorticoids.⁹⁰

The taxanes have not found a valuable role in the treatment of hematologic malignancy. However, a number of analogues and new formulations are under development. Abraxane, consisting of paclitaxel bound to albumen microparticles, does not require a lipid solvent, is virtually free of hypersensitivity as a side effect, and enters cells by a separate albumen-mediated transporter. It is approved for treatment of relapsed breast cancer and pancreatic cancer. Cabazitaxel, approved for prostate cancer, is a new analogue with decreased susceptibility to MDR. An entirely new class of natural products, the epothilones, have a similar mechanism of action, are less susceptible to MDR, and have activity against breast cancer.⁹¹

CAMPTOTHECINS

This group of compounds includes synthetic derivatives of 20 (S)-camptothecin, a naturally product from the Camptotheca acuminata tree. The camptothecins interact with a unique target, topoisomerase I, stabilizing the enzyme’s complex with DNA and preventing the resealing of DNA single-strand breaks induced by the enzyme. Resistance arises through mutation, deletion, or decreased expression of the topoisomerase I gene. The primary agents in clinical use are irinotecan, which is approved for treatment of colon cancer, and topotecan, approved for second-line treatment of ovarian cancer and small cell lung cancer. Irinotecan, most commonly administered intravenously at a dose of 125 mg/m² once each week for 4 weeks every 42 days, has shown promise against lymphomas in phase II trials performed in Japan.⁹² Response rates of 42 percent in previously treated patients with non-Hodgkin lymphoma, and of 38 percent in patients with refractory or relapsed adult T-cell leukemia-lymphoma, remain to be confirmed. Topotecan has remission-inducing activity in patients with myelodysplasia and chronic myelomonocytic leukemia, both as a single agent (1.5 mg/m² per day for 5 days) and in combination with ara-C.^93,94 Objective responses have also been observed in phase I clinical trials in patients with AML.⁹⁵ Irinotecan and topotecan differ substantially in their profile of toxicities and pharmacokinetic behavior. Irinotecan is a water-soluble prodrug that is converted to the active species, SN-38, by carboxyl esterase-mediated cleavage. Irinotecan and SN-38 are both eliminated by glucuronidation and biliary excretion. Therefore, irinotecan must be used with caution and at lower doses in patients with Gilbert disease (and lacking glucuronyl transferase 1A1) or in those with hepatic dysfunction.⁹⁶ In contrast to the hepatic extraction and excretion of irinotecan, approximately two-thirds of the dose of topotecan is eliminated by renal excretion, with the remainder being cleared by biliary excretion. Dose adjustment proportional to CrCl is indicated in patients with renal failure.⁹⁷ Topotecan toxicity consists mainly of myelosuppression and, to a lesser degree, mucositis, whereas irinotecan causes a profound diarrhea, which is responsive to loperamide, and a more modest myelosuppression. The maximum tolerated dose of topotecan for the 5-day schedule of 30-minute intravenous infusions/day is 4.5 mg/m² per day in patients with leukemia.⁹⁸ This is considerably greater than the approved dose for solid tumors, and gastrointestinal side effects, such as mucositis and diarrhea, become dose-limiting at these higher doses.

ANTHRACYCLINE ANTIBIOTICS

The anthracyclines are a unique class of natural products that inhibit topoisomerase II (Topo II), an enzyme important in DNA strand passage allowing the untangling of DNA prior to replication or repair. Doxorubicin, daunorubicin, idarubicin, and epirubicin are closely related in structure, each possessing a rigid planar core to which is linked a daunosamine sugar. The molecules differ in side-chain substitutions attached to the anthracycline ring system, and exhibit different spectra of antitumor activity and toxicity. Mitoxantrone, a closely related, nonglycosidic anthracenedione, has very similar pharmacologic properties to those of the anthracyclines. The anthracyclines are produced by a Streptomyces species, whereas mitoxantrone is a synthetic compound. Doxorubicin (Adriamycin) has broad activity against solid and hematologic malignancies. It is an important component of the standard multidrug regimens used to treat Hodgkin lymphoma (doxorubicin, bleomycin, vinblastine, and dacarbazine [ABVD]) and aggressive non-Hodgkin lymphoma (cyclophosphamide, doxorubicin, vincristine, and prednisone). Daunorubicin and idarubicin are used almost exclusively in combination with ara-C for the treatment of AML, whereas epirubicin is primarily effective against solid tumors. Mitoxantrone is employed for the treatment of AML and breast cancer, and as an immunosuppressive for patients with multiple sclerosis. Liposome-encapsulated doxorubicin (Doxil) and daunorubicin derivatives are approved for treatment of solid tumors; they provide a more prolonged, lower peak concentration of drug, and have decreased cardiac toxicity. The daunorubicin liposome is of interest in treating AML.⁹⁹ A novel anthracycline, pixantrone, which has lesser cardiotoxicity, has received conditional approval in Europe for refractory non-Hodgkin B-cell lymphoma.¹⁰⁰

Mechanism of Action

Anthracyclines target the replication and structural integrity of DNA. Their primary mechanism of toxicity stems from their interaction with Topo II, an enzyme that creates DNA strand breaks and promotes strand passage through those breaks. Strand passage is essential in untangling DNA in preparation for replication and repair. Once the strand passage and unwinding is complete, Topo II reseals the broken DNA strands. The anthracyclines inhibit the resealing step by forming a complex with Topo II and the broken DNA strand to which the enzyme is linked. Accumulation of strand breaks activates apoptosis. The planar molecular structure of these drugs promotes their intercalation between opposing strands of the DNA helix and may contribute to the specificity of sites of DNA breakage. In addition to their inhibition of Topo II, anthracyclines generate free radicals by virtue of the oxidation-reduction cycling of their quinone group, an action catalyzed by the binding of Fe²⁺. Free radical generation is thought to be responsible for their cardiac toxicity.

The importance of the presence of Topo II in determining response to anthracyclines is best illustrated by the greater benefit of anthracycline-based breast cancer treatment in patients with amplification of the target enzyme on chromosome 17, near the HER2 gene with which it coamplifies.¹⁰¹ Anthracycline-containing regimens are particularly effective in HER2-amplified breast cancers.¹⁰²

Anthracyclines enter cells through a passive transport process. Their lipophilic structure allows them to achieve high intracellular concentrations. Anthracyclines are pumped out of the cell by a series of ATP-dependent transporters, including the P-glycoprotein MDR transporter, the breast cancer resistance protein transporter (BCRP) and related efflux pumps.¹¹ Other mechanisms for anthracycline resistance include decreased Topo II activity or Topo II mutations in the enzyme that inhibit drug binding, as well as defects in apoptosis or impaired checkpoint recognition of DNA strand breaks.

Clinical Pharmacology

Daunorubicin and idarubicin are readily converted to active hydroxyl metabolites, whereas doxorubicin produces limited amounts of alcohol metabolite. The alcohols of daunorubicin and doxorubicin are less active as antitumor agents, but do possess cardiotoxic activity.

All anthracyclines are eliminated by the formation of inactive metabolic products (aglycons, side-chain-modified products, glucuronides, sulphates, and oxidative metabolite) in the liver. Only a minor fraction of the dose of any of the anthracyclines is excreted in the urine as the parent drug or alcohol metabolite. The pharmacokinetics of the clinically useful anthracyclines are predominantly influenced by their terminal disposition phase, which exceeds 24 hours. Although prolongation of the half-life of doxorubicin has been reported in studies of patients with compromised liver function, no clear correlations of liver function with toxicity have been established. However, in patients with elevated serum bilirubin levels, initial doses of doxorubicin and daunorubicin should be reduced by 50 percent, and adjust there after according to tolerance. Idarubicin, the only anthracycline amenable to oral administration, has a bioavailability of 20 percent for the parent drug and 40 percent for parent drug plus idarubicinol, the primary active metabolite. Idarubicinol has a very prolonged biologic half-life, ranging from 50 to 60 hours, and is likely responsible for the antitumor activity of this drug. In contrast to the metabolites of doxorubicin and daunorubicin, idarubicinol is eliminated primarily by renal excretion. No dose adjustment for hepatic dysfunction is indicated.

Mitoxantrone has a long terminal half-life of 23 to 42 hours. Only a minor fraction of unchanged drug is excreted in the urine (<10 percent) or stool (<20 percent). The majority of the drug is metabolized or bound to tissues. Patients with impaired hepatic function may have a more prolonged elimination of mitoxantrone.

The usual dose of doxorubicin when administered as a single agent by bolus intravenous injection is 45 to 75 mg/m² every 3 to 4 weeks, depending on the tumor treated and the drug combination. Less cardiac toxicity may result from schedules that avoid high peak plasma concentrations, such as weekly doses (15 to 25 mg/m²) or continuous intravenous infusion over 48 to 96 hours, as in the EPOCH (etoposide, prednisone, vincristine, cyclophosphamide, and doxorubicin) regimen.¹⁰³ When given in combination with other myelotoxic agents such as cyclophosphamide, the dose of doxorubicin is usually decreased because of overlapping marrow toxicity. Daunorubicin has been used as the anthracycline of choice in the treatment of AML. To minimize the cardiotoxic effects of daunorubicin, in standard “3+7” therapy in AML the daunorubicin dosage (45 to 60 mg/m² for adults <60 years) is given daily for 3 days to avoid high peak concentrations, although larger doses (90 mg/m²/day), may produce higher complete remission rates. In the elderly, a lower dose of 30 mg/m² daily for 3 days is typically used in combination therapy.

Adverse Effects

Myelosuppression is the primary acute toxicity of this class of drugs, with a nadir occurring 7 to 10 days after single-dose administration and recovery by 2 weeks. Mitoxantrone produces less nausea and vomiting than does either daunorubicin or doxorubicin. Doxorubicin may cause mucositis, especially when used in maximally tolerated divided doses given over 2 to 3 days or when used in combination with other drugs that cause mucositis. Anthracyclines can also cause radiation recall in previously irradiated tissues, especially when the drug is administered just prior to or in the weeks following irradiation. Alopecia often occurs. Extravasation of these drugs can result in tissue necrosis so they should be administered through an indwelling central venous catheter. Dexrazoxane injected subcutaneously, lessens tissue damage after extravasation.¹⁰⁴ Patients receiving doxorubicin should also be warned that their urine may turn red.

Cardiotoxicity is the major late toxic effect of anthracyclines.¹⁰⁵ Cardiotoxicity most likely results from free radical formation catalyzed by the anthracycline’s quinone moiety, although cardiac Topo IIb (not the major topoisomerase involved in DNA replication) may play a role in mediating this effect. Iron as a reduction-oxidation (redox) cofactor contributes to toxicity, as it accumulates in mitochondria during treatment.¹⁰⁶ Cardiac proteins, including the cardiac myosin-binding protein C, show evidence of alkylation and degradation after anthracycline treatment.¹⁰⁷

Clinically, anthracycline-induced cardiotoxicity presents after repeated cycles of treatment. Rarely, acute effects are manifest as arrhythmias, conduction abnormalities, or a “pericarditis–myocarditis syndrome.” The more common long-term consequence is congestive heart failure (CHF), which can develop during or several months after treatment. Studies in breast cancer have demonstrated a 0.5 to 1 percent risk of cardiomyopathy in patients treated with adjuvant anthracyclines.¹⁰⁸ The risk is higher in patients receiving trastuzumab or paclitaxel in combination with doxorubicin.

The risk of anthracycline-induced cardiotoxicity increases with total dose, but is difficult to estimate for any individual patient. In patients with normal cardiac function prior to treatment, the subsequent rate of doxorubicin-induced CHF reaches less than 1 percent at total doses of 400 mg/m², but climbs steeply thereafter to 7 to 20 percent at total doses of 550 mg/m².¹⁰⁹ The threshold for cardiotoxicity varies among the different anthracyclines. For example, the inflection threshold for daunorubicin (600 to 700 mg/m²) is significantly higher than for doxorubicin (400 mg/m²). However, it should be remembered that these thresholds are based on population studies, and for any individual patient the risk is impossible to predict. The clinician must pay close attention to symptoms of CHF, such as dyspnea, cough, orthopnea, and weight gain or ankle edema, throughout a course of treatment and irrespective of total dose.

Besides the cumulative dose of anthracycline, other risk factors for anthracycline-induced cardiomyopathy include mediastinal (mantle) radiation, preexisting heart disease, and patient age, the risk being highest in children younger than the age of 4 years. Children who receive greater than 300 mg/m² have a significant risk of having decreased myocardial contractility, decreased ventricular dimension, and an increased incidence of cardiac events (such as conduction defects, myocardial infarction, and CHF) in their adult years. The incidence of may be decreased by coadministering dexrazoxane, an iron chelator, during chemotherapy.¹¹⁰ It is recommended that the total doxorubicin dose be limited to 300 mg/m² in children. In addition, children treated with anthracyclines should have long-term cardiology followup.¹¹⁰

Ejection fraction measurements have been helpful in detecting a decline in myocardial function, a sign of impending myocardial failure. Ejection fraction measurements, usually by multigated acquisition scan (MUGA), should be performed to verify normal cardiac function prior to starting anthracycline-based chemotherapy, and should be repeated at the earliest clinical sign of cardiac dysfunction, and before every two cycles of treatment when the total dose of doxorubicin exceeds 300 mg/m². Anthracyclines should be discontinued if the ejection fraction falls below 40 percent, or if the ejection fraction drops a total of 20 percent from pretreatment levels.

As cardiotoxicity of anthracyclines results from the generation of free radicals by an anthracycline–iron complex, dexrazoxane, an iron chelator, decreases free radical formation in vitro and decreases the risk of cardiotoxicity in children receiving treatment for ALL, and in adults with metastatic breast cancer.¹¹⁰ Fortunately, dexrazoxane does not cause any apparent diminution of antitumor activity. Adding dexrazoxane to an anthracycline-based regimen represents an alternative to discontinuing anthracyclines in patients who are approaching total-dose thresholds of drug, but who still require treatment. Two trials have shown a higher rate of secondary leukemia and MDS in patients receiving doxorubicin and dexrazoxane.^111,112 In adult patients, dexrazoxane should be added only in patients who have received a total dose of at least 300 mg/m² doxorubicin or 540 mg/m² epirubicin.

Treatment with Topo II inhibitors, including anthracyclines, mitoxantrone, and the epipodophyllotoxins (see “Epipodophyllotoxins” below), increases the risk of AML. AML typically develops 6 months to 5 years after exposure to the Topo II inhibitor.¹⁰⁷ This heightened risk derives from the increased DNA double-strand breaks generated by the Topo II inhibitor. These double-stranded DNA breaks can give rise to balanced chromosomal translocations involving the MLL or PML genes.¹¹¹ Anthracyclines and mitoxantrone have affinity for specific DNA sequences and cause translocations at specific hot spots in the genome, including a 6-base pair breakpoint region in the PML gene causing the 15;17 translocation, the 11q23 translocation involving the MLL gene, and the 11;20 translocation involving the NUP98 gene.^113,114

EPIPODOPHYLLOTOXINS

Two semisynthetic derivatives of podophyllotoxin, VP-16 (etoposide) and VM-26 (teniposide), inhibit Topo II and have significant clinical activity in hematologic malignancies. Etoposide has been incorporated into combination therapy regimens for Hodgkin lymphoma, large cell non-Hodgkin lymphomas, leukemias, and various solid tumors, and is a frequent component of high-dose chemotherapy regimens. Teniposide has limited value in clinical oncology. Its use is generally restricted to childhood acute leukemia, where it appears to be synergistic with ara-C. These compounds induce double-stranded breaks in DNA through their sequence-specific binding to DNA in complex with Topo II.¹¹⁵ One mechanism of resistance is increased expression of the MDR drug exporter.¹¹ A second mechanism results from decreased Topo II activity or mutation of the enzyme, resulting in decreased drug binding.^116,117

Clinical Pharmacology

Etoposide is administered in doses of 100 to 120 mg/m² per day for 3 days, either consecutively or every other day. Approximately 30 to 40 percent of an intravenous dose of etoposide is excreted intact in the urine, while the remainder is cleared by hepatic glucuronidation or demethylation; thus, doses of etoposide require modification for patients with compromised renal or hepatic function.¹¹⁸ The plasma half-life of etoposide is 15 hours. The clinical activity of etoposide is highly schedule dependent. Single conventional doses are essentially without antitumor effect as compared to consecutive daily doses for 3 to 5 days. The pharmacokinetics of teniposide are very similar to those of etoposide, with a terminal plasma half-life of 20 to 48 hours. However, little parent drug appears intact in the urine, and dose modification for patients with renal dysfunction is unnecessary.

Adverse Effects

When administered intravenously, both etoposide and teniposide should be infused over a 30-minute period to avoid hypotensive episodes. The major toxicity of both drugs is leukopenia, which is rapidly reversible; thrombocytopenia is less common. Nausea and vomiting often follow etoposide administration. Alopecia may occur with both drugs. Other toxicities, such as fever, mild elevation of liver function tests, and peripheral neuropathy, are relatively uncommon. Because the major toxicity of etoposide is limited to the marrow, this drug is a valuable component of high-dose regimens used with marrow transplantation. In high-dose etoposide protocols (1.5 g/m² or greater given over 3 to 5 days) oropharyngeal mucositis becomes a prominent toxicity. Less-frequent high-dose toxicities include hepatocellular damage and, rarely, anaphylactic-like symptoms, probably related to the formulation vehicle. Secondary AML associated with translocation at 11q23 or the PML gene may follow etoposide treatment in children with ALL¹¹⁹ and in adults with solid tumors.¹²⁰

THE ALKYLATING DRUGS

These drugs are important in the treatment of hematopoietic malignancies either as single agents or as components of standard- or high-dose regimens. Their role as treatment for both acute and chronic hematologic malignancies results from their unique mechanism of cell killing and their lack of cell-cycle specificity. They may eradicate noncycling cells that escape cycle-active components of the treatment. Although these agents share the common property of forming covalent bonds with electron-rich sites on DNA (oxygen and nitrogen substituents), they exhibit important differences in their intrinsic reactivity, route of cellular uptake, favored sites of alkylation on DNA bases, and the specific mechanism of DNA repair that determines cell survival. These differences are borne out in experimental settings, where cross-resistance to alkylating agents is incomplete. Thus, protocols employing multiple alkylators, particularly in high-dose regimens, have a rational basis.¹²¹ Alkylating agents differ as well in their patterns of toxicity. The majority of these drugs cause myelosuppression and mucositis as their primary acute toxicities, as well as delayed pulmonary fibrosis and late secondary leukemias. These secondary leukemias often arise after a period of myelodysplasia, are usually highly drug resistant AML, and carry defects in chromosomes 5 or 7. Busulfan, bischloroethylnitrosourea (BCNU), or cyclophosphamide are most likely to cause vascular endothelial damage (hepatic venoocclusive disease) when used in high doses. However, these same drugs are often used in high-dose regimens, as they cause less mucositis than other alkylating agents. 4-Hydroperoxycyclophosphamide, an activated analogue of cyclophosphamide, appears to spare marrow stem cells relative to tumor cells and has been used for in vitro purging of marrow in autologous transplantation.¹²²

Although platinum analogues are not true alkylating agents in that they form metal adducts rather than carbon adducts with DNA, RNA, and protein, their range of toxicities and mechanisms of resistance have much in common with the classical alkylators. They have few indications in hematologic malignancy, aside from carboplatin and its role in high-dose chemotherapy for lymphomas. Their DNA adducts are subject to repair by nucleotide excision repair and double-strand break repair, processes dependent on functional p53 activity.¹²³ Polymorphisms of the repair pathways, especially the mismatch repair process, may be associated with drug resistance,¹²⁴ whereas errors in double-strand break repair (as found in BRCA1- and BRCA2-solid tumors), may create sensitivity to platinating drugs.

Mechanism of Action

All alkylating agents (Fig. 22–4) have in common the generation of highly reactive carbonium intermediates that attack electron-rich sites on DNA, such as the N-7, O-2, and O-6 positions of guanine and the N-1, N-3, and N-7 positions of adenine. For many of these agents, the alkylating group must undergo a preliminary activation reaction mediated either by chemical rearrangement of the molecule, as in the case of nitrogen mustard and the nitrosoureas, or by metabolic activation followed by chemical rearrangement, as for cyclophosphamide, ifosfamide, and procarbazine. Most alkylating agents have two reactive sites, usually two chloroethyl groups, enabling them to form intrastrand and, less frequently, interstrand crosslinks.

A second class of alkylating drugs, exemplified by busulfan, dimethyltriazenoimidazole carboxamide (DTIC) and the closely related temozolomide, and procarbazine, produce only single-strand alkylation but may be highly carcinogenic, as, for example, procarbazine. In general, all the commonly used alkylating drugs, including cyclophosphamide, ifosfamide, melphalan, chlorambucil, and the methylating drugs, produce the same spectrum of myelosuppressive, carcinogenic, and genotoxic actions, and depend on an intact mismatch repair system to recognize their adducts and initiate apoptosis.

Experimental systems have elucidated the mechanisms of resistance to alkylating agents.¹²⁵ Some mechanisms are specific for certain alkylating agents (e.g., impaired uptake of nitrogen mustard as a consequence of an alteration in the membrane carrier for choline, or deletion of the amino acid carrier used by melphalan), whereas others appear to be less specific (e.g., drug inactivation associated with an increase in intracellular sulfhydryl compounds, and enhanced nucleotide-excision repair of DNA adducts). The primary resistance mechanisms for various alkylating drugs, as documented in experimental tumors, include increased degradation by aldehyde dehydrogenase (specifically for cyclophosphamide)¹²⁶; increased conjugation of the reactive intermediates with glutathione or glutathione transferase (all chloroethylating agents and platinum analogues); increased repair of the O-6 guanine alkyl lesions by a specific alkyl transferase (nitrosoureas, procarbazine, temozolomide, and dacarbazine)¹²⁷; increased nucleotide excision repair (all platinum derivatives and chloroethylating agents, except possibly nitrosoureas); decreased uptake (melphalan, nitrogen mustard); decreased ability to recognize DNA damage because of defective mismatch repair, especially the loss of the MLH6 component¹²⁸ (most alkylating agents and platinum derivatives); and defective recognition of DNA alkylation and strand breaks and initiation of apoptosis (p53 loss-of- function mutants), which affects all alkylators. The basis of alkylating agent resistance in the clinic is still incompletely understood.

Clinical Pharmacology

In general, the alkylating agents and their reactive intermediates have short residence times in the systemic circulation and within cells. They are eliminated predominantly by hydrolysis of the reactive site, by chemical or biochemical conjugation to the sulfhydryl groups of glutathione or proteins, or by oxidative metabolism in the case of ifosfamide and cyclophosphamide. Therefore, dose reduction is not required in patients with diminished renal or hepatic function.

A few of the drugs require enzymatic activation. Cyclophosphamide and ifosfamide are closely related molecules that undergo hepatic CYP-mediated activation. Their active metabolites include a highly labile phosphoramide mustard and a second toxic metabolite, acrolein, which is excreted in the urine.¹²⁹ To counteract toxicity of acrolein to kidneys and bladder, mercaptoethane sulfonate (mesna) is administered simultaneously in equivalent doses to the alkylator. Procarbazine and DTIC require metabolic activation by hepatic CYP isoenzymes, whereas temozolomide, a structural congener of DTIC, spontaneously activates to a methylating intermediate, and has become the preferred drug for treating glioblastomas.

Nitrogen mustard is a highly reactive compound in its parent form, and thus can be administered topically for treatment of skin cancers and cutaneous lymphoma. It is a potent vesicant, and care must be taken in the mixing and administering the drug. It is still a preferred component of combination therapy in conjunction with vincristine (Oncovin), procarbazine and prednisone (MOPP) for childhood Hodgkin lymphoma. Extravasation may lead to severe tissue injury. The second-generation alkylating agents, which include cyclophosphamide, melphalan, busulfan, and chlorambucil, are more chemically stable and absorbed reasonably well when given orally.

The newest alkylating drug, bendamustine, is approved for both CLL and relapsed lymphomas, and consists of a purine base with a bis-chloroethyl side chain. In experimental systems it is only partially cross-resistant with other alkylators, and produces a bulky DNA adduct that is slowly removed by base excision repair. It strongly induces p53 phosphorylation and apoptosis, as well as cell necrosis, a distinct cell death response.¹³⁰ Bendamustine metabolism produces two minor toxic metabolites: hydroxylation of its 4 position and N-demethylation. The bulk of parent drug is eliminated through its reactivity with sulfhydryls and adduct formation. The drug displays much the same pattern of toxicity of other alkylating drugs, with perhaps less myelosuppression.

Carboplatin, often used in high-dose therapy of lymphomas, is primarily excreted by the kidneys. Its dosing is based on renal function, aiming at a specific area under the curve of 5 to 7, according to the formula:

Adverse Effects of Alkylating Agents

Marrow toxicity, which is cumulative and a function of total dose, is the most important acute toxic effect of alkylators. Nitrosoureas produce a characteristic delayed myelosuppression that reaches a nadir 4 to 6 weeks after administration. Busulfan, like the nitrosoureas, depletes stem cells and can cause profound marrow hypoplasia or permanent aplasia. The dose-limiting toxicity of DTIC is nausea and vomiting rather than marrow suppression. Carboplatin causes an acute thrombocytopenia, as well as a more chronic sensory neuropathy.

Other common toxicities include denudation of the gastrointestinal epithelium, pneumonitis, cardiac, and endothelial damage, which become evident during high-dose therapy. Virtually every organ system may be damaged by alkylating agents. Because alkylating agents react with DNA, mutations and secondary leukemias are major long-term effects of these agents. This hazard appears to be related to the total dose administered. The monofunctional methylating agents (e.g., procarbazine) are especially potent in this regard. All alkylating agents, but particularly busulfan and the nitrosoureas, may produce pulmonary fibrosis. The nitrosoureas also cause nephrotoxicity, particularly after total doses of 1200 mg/m² BCNU, whereas cyclophosphamide and ifosfamide cause chronic bladder toxicity, hemorrhage, and, in rare cases, bladder carcinomas. Urinary toxicity of the latter two agents is prevented by coadministration of mesna, a sulfhydryl that detoxifies acrolein at acid pH.

High-Dose Alkylating Agent Therapy

The development of hematopoietic cell transplantation has made it possible to administer doses of chemotherapy that would otherwise produce life-threatening aplasia. To be of benefit, however, high-­dose therapy must employ agents that have a relatively steep dose–response relationship. The drugs used must not have lethal extramedullary toxicity at high doses. Among the classes of cytotoxics, alkylators have a particularly favorable linear relationship between dose and cytotoxicity in experimental tumor systems. Extramedullary organ toxicities are infrequent until doses are increased manyfold, making them ideal candidates for high-dose regimens. Depending on the agent and the toxicity profile, doses may only be escalated by as little as twofold as seen with cisplatin because of renal toxicity, or to as high as 18-fold in the case of thiotepa (Table 22–3).^{131,132,133,134,135,136,137} However, when agents are combined into a high-dose regimen, overlapping extramedullary ­toxicities of the agents must be considered so as to avoid serious organ compromise (Table 22–4).^{138,139,140,141,142} Overlapping extramedullary toxicities ­(particularly the risk of pulmonary or hepatic dysfunction or secondary leukemia) cannot be completely avoided, but rational drug selection can minimize the dose reductions of the individual agents, compared to their single-agent maximum tolerated dose (MTD), while at the same time ensuring safety of the combination regimen. This is illustrated in Table 22–4, which shows the fraction of the single-agent MTD that can be administered in combination with other drugs. As might be expected, this fraction is quite variable depending on the drug combinations, with the average fractional MTD used in combination ranging from 0.5 to 1. Depending on the regimen, significant gastrointestinal, pulmonary, hepatic, and/or renal toxicities are encountered and become dose ­limiting. For these reasons, high-dose regimens are safest in patients who are younger (<70 years) and who have had minimal prior chemotherapy and radiation therapy.

BLEOMYCIN

Bleomycin is a mixture of cytotoxic peptides produced by the fungus Streptomyces verticillis.¹⁴³ Because it has antitumor activity with minimal marrow toxicity, it is commonly used as part of combination regimens (such as ABVD) to treat Hodgkin lymphoma and with cisplatin and vinblastine to treat germ cell tumors. Bleomycin acts by causing both single- and double-strand breaks in DNA. These breaks form as a consequence of a bleomycin–Fe (II) complex that binds to DNA and undergoes redox cycling with molecular oxygen. The drug’s reactive complex abstracts a proton from deoxyribose, leading to cleavage of the sugar at the 3′-carbon.¹⁴⁴ In experimental tumors, resistance to bleomycin has been attributed to increased tumor cell concentrations of an aminohydrolase that cleaves and inactivates the drug.¹⁴⁵ Some resistant cell lines exhibit enhanced capacity to repair strand breaks, and in others, resistance results from decreased drug accumulation. Additional factors, such as increased free radical detoxification, may also influence toxicity. The tumor specificity of bleomycin, its severe cutaneous and pulmonary toxicity, and its lack of toxicity to marrow and the gastrointestinal tract may be a result of widely differing levels of metal ions and bleomycin hydrolase, the detoxifying enzyme, in these tissues. A polymorphism in the hydrolase gene, identified by SNP A1450G, may confer resistance to the drug as the result of its enhanced hydrolase activity.¹⁴⁶ Cell killing occurs throughout the cell cycle.

Clinical Pharmacology

Bleomycin may be administered intravenously or intramuscularly in doses of 10 to 20 U/m² per week to cumulative doses of 250 U for systemic therapy, as well as intrapleurally or intraperitoneally for control of malignant effusions. The half-life of drug elimination from plasma is estimated to be 2 to 3 hours. After a single intravenous injection, more than half the dose is excreted, unchanged, in the urine within 24 hours.¹⁴⁷ Bleomycin elimination may be markedly impaired in patients with poor renal function; such patients are at risk of overwhelming skin and lung toxicity. Dose reduction by 50 percent should be considered in patients with a CrCl in the range of 30 to 80 mL/min, and drug should be withheld in the present of CrCl less than 30 mL/min.

Adverse Effects

Bleomycin has minimal effects on normal marrow; however, in patients given other myelosuppressive drugs or who are recovering from marrow toxicity from these agents, additional mild myelosuppression may be observed. The primary toxicities that result from bleomycin are pulmonary fibrosis and skin changes. In experimental settings, the drug activates the Hedgehog pathway and induces the secretion of numerous cytokines, including interleukin (IL)-6, tumor necrosis factor-α and transforming growth factor-β, by alveolar macrophages and inflammatory cells, leading to collagen deposition.¹⁴⁸ The risk of pulmonary toxicity is related to the cumulative dose administered, increasing to 10 percent in patients given more than 450 mg. Risk is also greater in patients older than age 60 years,¹⁴⁹ in patients with underlying lung disease, in patients receiving bleomycin who are given high oxygen concentrations, and in patients who have had previous radiotherapy to the lungs. Single intravenous doses of 25 mg/m² or more predispose to this toxic effect. Symptoms of pulmonary toxicity include cough and dyspnea. Chest radiographs show nonspecific infiltrates, especially in the lower lobes. Chest computed tomography changes may show extensive infiltrates, fibrosis in later stages of evolution, atelectasis, or cavitation. Positron emission tomography scans are strongly positive. Open-lung biopsy may be required to distinguish bleomycin pulmonary toxicity from infection or malignant disease. Pathologic findings of bleomycin toxicity include an inflammatory alveolar infiltrate with edema, pulmonary hyaline formation, and squamous metaplasia of the alveolar lining cells. These changes progress to intraalveolar and interstitial fibrosis over a period of months. Patients with bleomycin lung toxicity have a measurable decrease in carbon monoxide diffusing capacity, a test of possible value in predicting potential pulmonary toxicity.¹⁵⁰ Because there is no specific therapy for patients with bleomycin lung toxicity, close attention should be paid to early pulmonary symptoms and radiographic changes. In patients with bleomycin pulmonary toxicity, some improvement may be seen on discontinuation of the drug, but the pulmonary fibrosis is usually not reversible. Glucocorticoids may decrease inflammation, but are of no proven benefit once fibrosis has occurred. O₂ supplementation must be avoided, as it promotes the oxidative injury to pulmonary tissue.

The dermatologic toxicity of bleomycin is also dose related. Erythema, hyperpigmentation, hyperkeratosis, and even ulceration may occur when the drug is given in conventional daily doses for longer than 2 to 3 weeks. Areas of skin pressure, especially of the hands, fingers, and joints, are initially affected, and Raynaud phenomenon may become apparent in the distal digits. Nail changes and alopecia may also occur with continued use of the drug. In combination regimens (e.g., ABVD) where bleomycin is used intermittently, skin toxicity is rarely a dose-limiting problem.

Fever and malaise after injection are common symptoms and may be alleviated by acetaminophen. Hypersensitivity reactions have also been observed. Idiosyncratic cardiovascular collapse has been rarely noted. A 1- or 2-mg test dose administered to such susceptible patients may result in hypotension, tachycardia, pulmonary insufficiency, or anaphylactoid reactions within 30 to 60 minutes. Their occurrence precludes further treatment with bleomycin.

L-ASPARAGINASE

The enzyme L-asparaginase is used clinically in the treatment of lymphoid malignancies, particularly in poor-risk B-cell ALL, T-cell ALL, natural killer (NK)-cell leukemia, and in high-grade lymphomas.

Mechanism of Action

The cells causing these lymphoid malignancies require exogenous L-asparagine for growth; they obtain this amino acid from the systemic pool of amino acids generated by the liver. The enzyme L-asparaginase, which catalyzes the hydrolysis of asparagine to aspartic acid and ammonia, rapidly depletes L-asparagine from plasma and induces an asparagine deficiency in lymphoid malignant cells. Resistant tumors are able to respond by induction of asparagine synthetase,¹⁵¹ thereby restoring intracellular pools of asparagine. For reasons not well understood, hyperdiploid ALL cells are particularly sensitive to L-asparaginase, whereas cells containing the BCR-ABL translocation are more resistant.¹⁵²

Three L-asparaginase preparations are available in the United States.¹⁵³ The product purified from Escherichia coli is employed as a first-line agent, while a second preparation (pegaspargase), derived by attachment of polyethylene glycol to the E. coli enzyme, is used for first-time therapy and for patients hypersensitive to the unmodified enzyme. A third preparation, purified from Erwinia chrysanthemi, can be obtained from the National Cancer Institute of the United States for patients hypersensitive to the E. coli enzyme or to pegaspargase. The various preparations differ in their pharmacokinetics, immunogenicity, and recommended doses. The E. coli enzyme is usually given in doses of 6000 to 10,000 IU intramuscularly every third day for 3 to 4 weeks, although much higher doses (25,000 IU once weekly) may be more effective in ALL treatment. Levels are maintained continuously above 0.2 IU/mL plasma, leading to total abolition of asparagine in the systemic circulation. The E. coli enzyme has an elimination half-life of 14 to 24 hours. Monomethoxypolyethylene glycol (PEG) conjugated to the E. coli enzyme reduces its immunogenicity and extends its half-life to 6 days. Pegaspargase is used in patients hypersensitive to the unmodified enzyme, in doses of 2500 IU/m² intramuscularly every 2 weeks. Single doses deplete L-asparagine from plasma for 2 to 3 weeks. Some patients develop hypersensitive to both preparations of E. coli enzyme, particularly if first exposed to the unmodified enzyme; they may be treated with enzyme from Erwinia, which has a low incidence of hypersensitivity and approximately equal catalytic activity to the E. coli preparation, but a more rapid clearance.¹⁵⁴ Consequently, the Erwinia enzyme must be used in higher doses.

Adverse Effects

Reactions to the first dose are uncommon, but after two or more doses of the unmodified enzyme, hypersensitivity may develop in up to 20 percent of patients, varying from urticarial reactions to hypotension, laryngospasm, and cardiac arrest. Skin testing to predict allergic reactions is helpful in some, but not all, cases, and should be performed to confirm a clinical suspicion of hypersensitivity. Hypersensitive patients may have antibodies to L-asparaginase in their plasma. More than half the patients with such circulating antibodies will not display an overt allergic reaction to the drug, but these patients may have more rapid disappearance of drug from plasma and an inadequate clearance of asparagine from plasma and cells, leading to therapeutic failure. Patients who are treated with L-asparaginase should be observed carefully for several hours after dosing, and epinephrine should be available in case anaphylactic reactions occur. Anaphylaxis is less likely when E. coli L-asparaginase is given intramuscularly than when it is administered intravenously. Pegaspargase has much reduced immunogenicity and hypersensitivity reactions are uncommon. However, up to 20 percent of patients previously exposed to unmodified L-asparaginase will develop allergy to subsequent pegaspargase, with undetectable enzyme levels in plasma, and an additional 8 percent will have silent inactivation of the enzyme. The other major toxic effects of L-asparaginase are a consequence of the ability of this drug to inhibit protein synthesis in normal tissues. Inhibition of protein synthesis in the liver will result in hypoalbuminemia, a decrease in clotting factors, a decrease in serum lipoproteins, and a marked increase in plasma triglycerides. Inhibition of insulin production may lead to hyperglycemia. The clotting abnormalities that are regularly observed as a consequence of L-asparaginase treatment include initial decreases in the anticoagulant factors antithrombin III, protein C, and protein S, leading to either arterial or venous thrombosis in occasional patients, and a predilection to thrombosis of cortical sinus vessels.¹⁵⁵ With more prolonged therapy, bleeding sequelae may result from inhibition of the synthesis of procoagulant proteins such as fibrinogen and factors II, VII, IX, and X. Consequently, monitoring of coagulation factors is recommended. High doses of L-asparaginase may cause cerebral dysfunction that manifests as confusion, stupor, seizures, or coma, and cortical sinus thrombosis has been documented by magnetic resonance imaging scan in such patients.¹⁵⁶ Clinical thromboembolic episodes may occur in up to 35 percent of children with ALL.¹⁵⁷ These events are mostly asymptomatic thrombi associated with central venous catheters; less frequently, cortical sinus and atrial thrombi may occur. Altered mental status may also result from hyperammonemia and diabetic ketoacidosis.¹⁵⁸ Preexisting clotting abnormalities, such as antiphospholipid antibodies or factor V Leiden deficiency, may predispose to thromboembolic complications.¹⁵⁹

Acute nonhemorrhagic pancreatitis occurs as a complication of L-asparaginase treatment, especially in patients who have extreme elevations of plasma triglycerides (>2 g/dL).¹⁶⁰ Because L-asparaginase manifests little toxicity in marrow or gastrointestinal mucosa, it has been used in combination with other drugs that do have such toxicities.

THALIDOMIDE, LENALIDOMIDE, AND POMALIDOMIDE

Thalidomide (α-phthalimidoglutarimide; Fig. 22–5), approved in 1953 as a sedative, was withdrawn shortly thereafter because of its teratogenicity. It causes dysmelia (i.e., stunted limb growth) when used during early pregnancy. However, it has since reemerged as an important antibacterial and antitumor agent, with clear effectiveness against leprosy and myeloma, especially when combined with other agents.¹⁶¹ Its analogues, lenalidomide and pomalidomide (see Fig. 22–5), have proven to be less toxic, and more effective for treating relapsed and refractory patients with myeloma. Lenalidomide is highly active in first-line combination therapy with dexamethasone, and also with bortezomib¹⁶² for myeloma, as well as being approved for the treatment of myelodysplasia in patients with the 5q– variant of this syndrome. The newest immunomodulatory drug (IMiD), pomalidomide, is approved for patients refractory to lenalidomide and bortezomib.¹⁶³

The mechanism of action of thalidomide and analogues is incompletely understood, as the compounds have a number of different actions, including a prominent antiangiogenic effect against tumors,¹⁶⁴ immune modulation, and inhibition of cytokine secretion. They inhibit neovascularization in the mouse cornea, block proliferation of endothelial cells in culture,¹⁶⁵ and inhibit secretions of vascular endothelial growth factor and other angiogenic cytokines.¹⁶⁶ Thalidomide potently stimulates phosphorylation of the CD28 costimulatory molecule.¹⁶⁷ This effect can lead to enhancement of T-cell function and activation of signaling pathways. Thalidomide has inhibitory effects on cytokine secretion, lowering levels of tumor necrosis factor-α and γ-interferon in leprosy patients. In addition it enhances NK-cell numbers and function, suppresses T-regulatory cells, and stimulates cytolytic T-cell function. It downregulates interferon regulatory factor 4 (IRF4), a myeloma survival factor. Although lenalidomide and pomalidomide have not been studied as extensively as thalidomide preclinically, they have the same spectrum of biologic actions but with greater potency.

An additional IMiD site of action related to degradation of key proteins has been identified. The IMiDs interact with cereblon, a protein that forms a complex with ubiquitin E3, and thereby promote the degradation of myc protein and other transcription factors. This effect on E3 and cereblon binding partners has been implicated in the teratogenic effects and antitumor activity of the IMiD compounds.¹⁶⁸

Clinical Pharmacology of Thalidomide and Its Congeners

Thalidomide consists of two enantiomers that rapidly interconvert in solution and biologic fluids. Its two imide bonds are unstable and undergo hydrolysis in solution. The poorly soluble drug undergoes slow and somewhat variable oral absorption with peak levels achieved in 2.9 to 4.3 hours^169,170 after oral doses ranging from 50 to 400 mg. There is no evidence for induction of metabolism on a daily dosing regimen. Drug concentrations in plasma decay with a half-life of 5 to 7 hours, the major pathways for elimination including spontaneous hydrolysis of the imide esters, and further CYP-mediated metabolism by the liver. At high doses of 1200 mg, the rate of clearance of drug from plasma decreases. Less than 1 percent of the drug is excreted unchanged in the urine. No dose adjustment is required for renal dysfunction, although its propensity for causing neuropathy may aggravate any underlying neuropathy secondary to prior exposures, renal failure or amyloidosis, making dose reduction prudent, especially when used in combination with other agents.

Lenalidomide is well absorbed orally in doses up to 400 mg, and usually given in doses up to 25 mg daily; it exhibits a plasma half-life of 3 hours. Approximately 70 percent of administered drug is excreted unchanged by the renal route, the remainder appearing in the feces unchanged. Dose adjustments are therefore recommended for patients in moderate (10 mg/day for CrCl of 30 to 50 mL/min) or severe (10 mg every other day for CrCl <30 mL/min) renal failure. For those on dialysis, the recommended dose is 5 mg once daily, with the same day dose given following dialysis.

Pomalidomide is given orally in doses of up to 4 mg per day. It has a long plasma half-life of 7.5 hours in myeloma patients, and is eliminated by CYP1A2 and CYP3A4 hydroxylation in the liver with minimal renal clearance. Inducers or inhibitors of CYP3A4 metabolism (prominently antibiotics and HIV drugs) and inhibitors of the MDR1 efflux transporter (natural products and targeted cancer therapies) should be used in combination with pomalidomide with caution.¹⁷¹

Clinical Use

Thalidomide has been evaluated against a number of human malignancies, with occasional responses in brain tumors, renal cell cancer, hepatoma, and Kaposi sarcoma. It has established value in treating myeloma refractory to first-line chemotherapy, as well as in newly diagnosed patients.^172,173 In responding patients, all aspects of the disease, including marrow infiltration with tumor cells, anemia, and performance status, improved with therapy. Thalidomide has synergistic myeloma-inhibiting activity with glucocorticoids, interferon-α, bortezomib, and cytotoxic agents. However, lenalidomide, which has less-prominent side effects and probably greater efficacy, has replaced thalidomide in first-line regimens for myeloma. Pomalidomide is reserved for patients refractory to lenalidomide and bortezomib, and retains impressive activity in this setting.

Thalidomide is generally well tolerated in oral doses of 50 to 1200 mg daily, although higher doses are more challenging. In treating myeloma, a 1-month trial is usually sufficient to observe a decline in paraprotein and an improvement in symptoms, with doses typically used ranging from 50 to 200 mg/daily. Doses can be escalated up to 200 mg every 2 weeks until dose-limiting toxicity is reached at 600 to 800 mg/day, but these higher doses are rarely used. Patients older than age 65 years are less tolerant of side effects, particularly sedation, constipation, fatigue, and peripheral sensory neuropathy, and receive a median dose of at most 400 mg/day, with lower doses (e.g., 100 mg/day) being preferable,¹⁷⁴ whereas younger patients may tolerate up to a median of at most 800 mg/day, although again lower doses (e.g., 200 mg/day) are preferable. With extended treatment at doses below 400 mg/day, the peripheral sensory neuropathy may become bothersome, but usually improves with dose reduction or drug discontinuation. To avoid undue sedation, the drug is given either in divided doses, morning and evening, or as a single evening dose. Other side effects include rash, dizziness and orthostatic hypotension, neutropenia, mood changes or depression, and nausea. Hypersensitivity and bradycardia also have been reported. Rarely patients may develop an interstitial pneumonitis or fulminant hepatic failure.

Trials of thalidomide in combination with cytotoxic drugs or biologics originally disclosed some unexpected toxicities.¹⁷⁵ Thalidomide in combination with doxorubicin or with prednisone is associated with an increased incidence of thromboembolism, a complication that can be prevented by concurrent treatment with low-molecular-weight heparin or aspirin.¹⁷⁶ Because of its teratogenicity, patients of childbearing age should take precautions to prevent pregnancy while on therapy. In trials of thalidomide and interferon against renal cell carcinoma, in which high doses of interferon (9 million IU subcutaneously three times per week) were used, four of 13 patients developed complex partial seizures and visual disturbances.¹⁷⁷ Two of 19 patients on thalidomide with low-dose interferon (1.5 to 3 million IU three times weekly) developed complex partial seizures in a trial against melanoma.

In the United States, thalidomide and its analogues are approved for use under a special risk evaluation and mitigation strategy (REMS), with restricted pharmacy access and a special consent form, to assure that pregnant women are not given these agents.

The analogue, lenalidomide, with significant activity against myeloma, causes much less sedation, constipation, and neurotoxicity, but prominent myelosuppression in 20 percent of patients. It is proving to be highly effective in remission induction with bortezomib and prednisone or with prednisone alone. Used in oral doses of up to 25 mg/day for 21 of 28 days, it is dramatically effective in normalizing hematologic parameters in the subset of patients with myelodysplasia who have a 5q– deletion on cytogenetics. A gene expression profile characteristic of lenalidomide responders has been reported.¹⁷⁸ Lenalidomide produces dramatic tumor swelling (tumor flare reaction) and tumor lysis in patients with CLL, a potentially fatal complication, even in patients with disease refractory to conventional agents. In CLL, it is equally effective in patients with poor prognostic cytogenetics (chromosomes 11 and 17 deletions). It must be started in low doses (beginning at 2.5 to 5 mg/day and escalating thereafter) to avoid tumor flare reaction and renal failure.¹⁷⁹ It has rarely been associated with severe hepatic and renal toxicity.

Like thalidomide, lenalidomide in combination with anthracyclines or glucocorticoids causes a 15 percent incidence of thrombotic events, and in these combinations should be administered with low-molecular-weight heparin, although prospective trials of anticoagulation are lacking.¹⁸⁰

Pomalidomide’s prominent toxicity is neutropenia in 50 to 60 percent of patients, and thrombocytopenia in 25 percent. It has little sedating effects and causes neuropathy in 10 percent or fewer subjects. It is highly active in relapsed, refractory myeloma and particularly in combination with various agents, including dexamethasone and proteasome inhibitors, such as bortezomib and carfilzomib. Rare toxicities include thromboembolism (3 percent) and isolated cases of hepatic failure.¹⁶³

Certain chemical agents have the ability to cause terminal differentiation (maturation) of malignant cells.^181,182 The most prominent among these are members of the vitamin A family (carotenes and retinoids), vitamin D and its analogues, phenylacetic acid, various cytotoxic agents used in low concentrations (such as hydroxyurea), inhibitors of DNA methylation such as 5-azacytidine and 5-aza-2′-deoxycytidine or decitabine, and inhibitors of histone deacetylase, exemplified by vorinostat, depsipeptide, and various benzamides.¹⁸³ In addition, biologic agents such as the interferons and interleukins induce terminal differentiation of both malignant and normal cells, but the role of terminal differentiation in the anticancer action of these drugs in humans is uncertain, as they have multiple biologic effects.

RETINOIDS

As the first effective terminal differentiating agent in cancer therapy, ATRA induces complete responses in a high percentage of patients with APL, and has become a standard member of the combination regimen for treatment and cure of this disease.¹⁸⁴ ATRA acts through binding to a nuclear receptor formed by the heterodimerization of the retinoic acid receptor-α (RARα) and its partner, the retinoid X receptor. In APL, an abnormal fusion protein, composed of portions of the RARα and a unique transcription factor (the PML gene product), results from the characteristic 15;17 chromosomal translocation found in this disease.¹⁸⁵ The fusion protein has a lower affinity for retinoids than does the wild-type molecule. High concentrations of retinoids are required to displace a corepressor bound to the protein, and activate key differentiation factors such as CCAAT/enhancer binding protein (C/EBP) and PU.1.¹⁸⁶ The fusion protein forms a variety of homo- and heterodimers that regulate genes and increase leukemic stem cell renewal, and suppress apoptosis and DNA repair, further contributing to progression of leukemia. In experimental settings, resistance to ATRA differentiating activity results from mutation or loss of retinoid binding in the PML-RARα fusion gene, indicating that the fusion gene product plays a role in retinoid responsiveness, and sensitivity can be restored by transfection of a functional RARα gene.¹⁸⁷

ATRA is administered to APL patients in oral doses of 25 to 45 mg/m² per day until complete remission is achieved and reaches peak serum levels of 300 ng/mL 1 to 2 hours after administration.¹⁸⁸ It is also used in remission maintenance, with 6-MP, methotrexate, or ara-C. The parent drug disappears from serum with a half-life of less than 1 hour during the initial course of treatment, but its rate of clearance greatly accelerates with continued treatment, a factor that may contribute to resistance to ATRA therapy. Induction of CYP26A1-mediated metabolism is suspected to underlie this accelerated clearance, and may account for the high rate of disease recurrence if ATRA is used as a single agent.¹⁸⁹ The primary toxicities of ATRA resemble those of other retinoids and vitamin A, specifically dry skin, cheilitis, mild and reversible hepatic dysfunction, bone tenderness and hyperostosis on radiography, hypercalcemia, hyperlipidemia, and occasional cases of pseudotumor cerebri. Imidazole antifungals block the degradation of ATRA and may lead to hypercalcemia and renal failure. In addition, approximately 15 percent of patients with APL, particularly those with an initial leukemic cell count greater than 5000/μL, develop a syndrome of hyperleukocytosis, fever, altered mental status, pleural and pericardial effusions, and respiratory failure (the “retinoic acid syndrome”).¹⁹⁰ Hyperleukocytosis and leukocyte adherence to small vessels results from a rapid increase in the number of mature leukemic cells in the blood and from the increased expression of integrins on the leukemic cell surface and secretion of cytokines in response to ATRA. In patients with white blood cell counts above 20 × 10³ cells/μL, pleural and pericardial effusions and peripheral edema develop rapidly, and respiratory distress, cardiac failure, and renal insufficiency may lead to death. Anecdotal reports indicate that high-dose glucocorticoids reverse this syndrome, which is mediated by leukocyte adhesion and clogging of small vessels and/or by cytokine release.¹⁹¹ The early introduction of cytotoxic chemotherapy during remission induction, and the use of dexamethasone sodium phosphate (10 mg twice daily for 3 or more days in patients with initial leukemic counts of greater than 5000 cells/μL), drastically lower the incidence of the syndrome and improve the safety of ATRA therapy.

ARSENIC TRIOXIDE

In the 1930s, arsenic was used to treat CML and other malignancies with little effect. Based on further clinical trials of arsenic trioxide (As₂O₃) in Harbin, China, in 1992, it resurfaced as an impressively effective treatment for relapsed APL, and appears to also be active against myeloma and myelodysplasia.¹⁹² Its mechanism of action probably stems from its ability to promote free radical production.¹⁹³ It inhibits the detoxification of free radicals and inactivates glutathione, an important radical scavenger.¹⁹⁴ It promotes degradation of the PML-RARα fusion protein,¹⁹⁵ and upregulates p53 and proapoptotic proteins. The cumulative effect is to induce maturation and promote apoptosis in APL cells. In addition, it has antiangiogenic effects. The sum of these actions is potent antitumor activity in some but not all tumor cells. In APL patients refractory to ATRA and conventional chemotherapy, it produces strikingly durable complete responses, and is therefore under study as a part of primary treatment regimens for this disease.

Patients are treated with a 2-hour intravenous infusion of 0.15 mg/kg day for up to 60 days, or until marrow remission is achieved, with further consolidation therapy beginning 3 weeks after remission. Remissions appear in 2 to 3 months with evidence of leukemic cell differentiation and a progressive blood leukocytosis after 2 weeks of therapy.¹⁹⁶ Side effects of arsenic trioxide in APL may include hyperglycemia, elevated liver enzymes, and hypokalemia, none of which require discontinuation of therapy. Occasional patients complain of fatigue, dysesthesias, and lightheadedness. A pulmonary distress syndrome, similar to that encountered with APL cell maturation after ATRA therapy, occurs in approximately 10 percent of patients, and is managed with glucocorticoids, oxygen, and temporary withholding of arsenic trioxide. Arsenic trioxide prolongs the cardiac QT interval, and uncommonly produces atrial or ventricular arrhythmias; it is important to maintain serum potassium at normal concentrations during arsenic trioxide therapy, and to avoid use of other drugs that prolong the QT interval, such as macrolide antibiotics, methadone, or quinidine.

Torsade de pointes occurs infrequently during arsenic trioxide treatment, but requires immediate treatment with intravenous magnesium sulfate, potassium repletion, and defibrillation if the arrhythmia and hemodynamic instability persist.¹⁹⁷

A maximum plasma concentration of 5.5 to 7.3 μM was achieved in the initial studies from China, and small amounts of drug and the methylated metabolite are eliminated in the urine, the rest remaining in tissues.¹⁹⁸

DEMETHYLATING AGENTS

DNA methyltransferases (DNMTs) regulate transcription by methylating CpG promoter regions of DNA and thereby silencing gene expression. Mutations associated with AML, including isocitrate dehydrogenase 1 and 2 (IDH1 and IDH2) mutations, TET2, and mutations that activate DNMT, result in increased DNA methylation. Hypermethylation blocks differentiation and drives cellular proliferation. Two DNA demethylating agents, azacitidine and decitabine, have been approved for treatment of MDS. Both drugs become incorporated into DNA, substituting for cytosine bases, and form a suicide covalent bond with DNMT. The two drugs differ in their activation pathways and in their effects on nucleic acid methylation. Decitabine is phosphorylated by dCK whereas azacitidine is activated by cytidine kinase. Decitabine nucleotide is incorporated only into DNA, while azacitidine is found in both RNA and DNA. The clinical response to these drugs has not been correlated with either global changes in DNA methylation or with methylation of specific genes. Indeed, while traditional thinking has taught that DNA methylation invariably leads to gene silencing, newer studies have demonstrated that many actively transcribed genes have high levels of DNA methylation and that the tissue context and specific patterns of methylated DNA may play an important role in determining transcriptional activity.¹⁹⁹

Azacitidine received FDA approval for therapy of MDS on the basis of reported improvement in Hgb, white cells, or platelets, delayed progression to AML, improved quality of life, and improved overall survival.²⁰⁰ In a more recent four-arm phase III trial in higher-risk MDS patients, azacitidine was compared with best supportive care, low-dose ara-C, or induction chemotherapy with anthracycline and ara-C. The median overall survival was 24.5 months in the azacitidine arm compared with 15 months for the other arms.²⁰¹

Decitabine was approved on the basis of a 30 percent overall hematologic response rate in MDS patients.²⁰² For those with intermediate- or high-risk disease, time to AML or death was significantly delayed, but there is no improvement as yet in overall survival with this drug.²⁰³ While the two agents have not been compared head to head in MDS, a meta-analysis showed a significant overall survival benefit versus supportive care only for azacitidine.²⁰⁴ A retrospective review found no significant difference in efficacy between the two compounds except in patients older than 65 years of age, for whom azacitidine resulted in improved survival and a more favorable toxicity profile.

Both agents increase HgbF levels in sickle cell anemia and thalassemia, and reduce symptomatic episodes, but have been superseded by hydroxyurea for this indication.

Azacitidine is approved for parenteral or subcutaneous administration at 75 mg/m²/day for 7 days every 28 days, a regimen that has been shown to result in maximal DNA hypomethylation. A daily-times-five regimen appears to have similar efficacy.²⁰⁵ The median number of cycles needed for response is three, but 80 percent of responses occurred before the sixth cycle.²⁰⁶ An oral formulation is currently being developed and has demonstrated activity in MDS and chronic myelomonocytic leukemia. The decitabine dose resulting in optimal hypomethylation is 20 mg/m² intravenously daily for 5 days every 4 weeks,²⁰⁵ although alternative schedules employing lower doses have been used. Dose adjustment and delays in repeat cycles may become necessary because of myelosuppression.

The primary clinical toxicities of both drugs include reversible myelosuppression, nausea and vomiting with higher doses, hepatic dysfunction, myalgias, fever, and rash. Both compounds are rapidly deaminated and converted to a chemically unstable azauridine or aza-deoxyuridine metabolite that immediately degrades into inactive products.²⁰²

A significant portion of patients with MDS does not respond to demethylating agents, and all will ultimately relapse. Azacitidine is initially phosphorylated by a different kinase (uridine-cytidine kinase) and so may benefit patients unresponsive to azacytidine.²⁰⁵

Clinical trials of azacytosine nucleosides with histone deacetylase (HDAC) inhibitors have demonstrated promising results in phase I trials in MDS and AML.^205,207 Second-generation hypomethylating agents are currently in clinical trials and hold the promise of more convenient dosing schedules and improved toxicity profiles.²⁰⁵

HISTONE DEACETYLASE INHIBITORS

Histone acetylation is an important determinant of gene expression and is regulated by the addition and removal of acetyl groups from lysine amino acid residues on histones. The removal of acetyl groups facilitates chromatin compaction thereby decreasing gene expression and is mediated by HDACs. The four classes of HDACs are differentially expressed. HDACs 1 to 3 are overexpressed in many cancer types. Overexpression of these HDACs is associated with repression of tumor suppressor and DNA repair genes, and confers a poor prognosis.^208,209 HDAC inhibitors reverse these changes. They maintain chromatin acetylation and decompaction and promote expression of tumor-suppressor genes, thereby inducing terminal differentiation and apoptosis of tumor cells. Interestingly, HDACs may paradoxically play a role as tumor suppressors as well, as knockout models in mice exhibit spontaneous tumorigenesis through a p53-mediated mechanism.^208,210,211 HDACs deacetylate multiple nonhistone proteins, including p53 and members of the DNA repair complex, but these actions have uncertain significance.

Three HDAC inhibitors are approved for clinical use: vorinostat and romidepsin. Vorinostat is a hydroxamic acid derivative, whereas romidepsin is a natural product produced by Chromobacterium violaceum. Both HDAC inhibitors block the zinc-dependent enzymatic activity of these HDACs and work predominantly against class 1 HDACs (HDACs 1 to 3).²¹² In lymphoma cells, romidepsin was able to overcome the prosurvival effects of BCL-2 whereas vorinostat was not.

HDAC inhibitors are active in cutaneous T-cell lymphoma. Pabinostat was recently approved, in combination with bortezomib and dexamethasone, for relapsed multiple myeloma. Romidepsin is useful against peripheral T-cell lymphoma. FDA approval of vorinostat was based upon an overall partial response rate of 30 percent and a median response duration of 168 days in patients with cutaneous T-cell lymphoma who had progressed on at least two prior regimens.²¹³ Romidepsin was approved for use in cutaneous T-cell lymphoma as well after clinical trials demonstrated an overall response rate of 34 percent, including 7 percent complete responses in previously treated patients.²¹⁴

Romidepsin is approved for use in patients with who have received at least one prior therapy.²¹² In these patients it achieved objective response rates of up to 38 percent. The median progression free survival for the 15 percent of patients in complete remission was 29 months.^215,216

Vorinostat is administered orally at 400 mg/day. It is predominantly cleared by glucuronidation and side change oxidation by cytochrome metabolism, and should be dose reduced to 300 mg/day for mild and moderate hepatic impairment. It is contraindicated in patients with severe hepatic impairment.²¹⁷ It has a plasma half-life of 2 hours, although histones remain hyperacetylated for many hours. Renal elimination does not play a major role in the drug’s clearance.²¹⁸ Romidepsin is administered as a 4-hour intravenous infusion of 14 mg/m² on days 1, 8, and 15 of a 28-day treatment cycle. A dose reduction to 10 mg/m² is possible in patients who experience high-grade toxicities.²¹⁹ Romidepsin is metabolized primarily by CYP3A4 and by glucuronidation, and is rapidly cleared with a short half-life of approximately 3.5 hours. Asian patients with the 2B57 genotype of uridine diphosphate (UDP)-glucuronyl transferase may experience delayed clearance and increased toxicity.

Both HDACs are generally well tolerated. The most common adverse events for vorinostat are diarrhea, fatigue, nausea, and anorexia, and laboratory abnormalities including, hyperglycemia, thrombocytopenia, and proteinuria.²¹³ For romidepsin, nausea, vomiting, infection, fatigue, and myelosuppression were the primary toxicities.²¹² Significant QTc prolongation and T-wave flattening but the electrocardiogram (ECG) changes were not associated with clinical cardiotoxicity. Still, it is important to ensure electrolyte levels are normalized prior to and throughout therapy.

Multiple newer HDACs have been extensively tested in clinical trials, although none has proven superior to vorinostat or romidepsin. One new agent, panobinostat, has demonstrated promising results in phase II trials against cutaneous T-cell lymphoma, Hodgkin lymphoma, and Waldenström macroglobulinemia.²⁰⁸

FUTURE EPIGENETIC TARGETS

Isocitrate Dehydrogenase 2 Inhibitors

Isocitrate dehydrogenase (IDH) is a metabolic enzyme that converts isocitrate to α-ketoglutarate (αKG) (Fig. 22–6), a cofactor in dioxygenase reaction that precedes demethylation of histones and the DNA.

Certain cancers, including subsets of AML, gliomas, intrahepatic cholangiosarcomas, breast and lung cancers, and central chondrosarcomas harbor IDH1/2 mutations that confer a gain of function. These mutated enzymes (commonly mutated at arginine 132 of IDH1 or arginine 170 of IDH2) convert αKG to the oncometabolite (R)-2-­hydroxyglutarate (2HG). High concentrations of 2HG inhibit histone and DNA demethylation by competing with the dioxygenase function of the TET family of enzymes and Jumonji-C domain (JMJC) family of histone lysine demethylases. The result of IDH mutations is hypermethylation of both DNA and histones and a block in differentiation.²²⁰ IDH1 or IDH2 mutations are present in 25 percent of AML where they confer a favorable prognosis. At the same time, IDH mutations in MDS and other myeloproliferative neoplasms may place patients at an increased risk for transformation to AML.²²¹ High levels of 2HG are measurable in plasma and urine in patients with IDH mutations, and may serve as a sensitive measure of treatment response and may reflect the presence of minimal residual disease.²²² 2HG in IDH mutant solid tumors may also be imaged by magnetic resonance imaging (MRI) spectroscopy.

Small molecule inhibitors of IDH mutant enzymes have been ­identified. One such inhibitor binds at the dimer interface in an allosteric fashion and exhibits uncompetitive inhibition of mutant IDH2, with great specificity for mutant IDH2 as compared to the wild-type enzyme. Mutant AML cells cultured ex vivo in the presence of these inhibitors undergo differentiation.²²⁰

An oral formulation of the mutant IDH2 inhibitor has shown promising remission-inducing activity in its initial clinical trial.^220A

DOT1L Inhibitors

Histone methylation also plays a role in transcriptional regulation. Histone methylation leads to exposure of DNA promoter sites, with effects on gene expression. DOT1L, a histone methyltransferase, catalyzes the hypermethylation of specific lysine residues on histone H3 (H3K79), and thereby regulates RNA polymerase II–mediated transcriptional elongation. In leukemias characterized by translocations involving the MLL gene, fusion proteins are created that recruit DOT1L to transcription factor (HOXA9 and MEIS1) promoter sites, resulting in leukemogenesis. MLL translocations are seen in 5 to 10 percent of acute leukemias of lymphoid, myeloid, or mixed/indeterminate lineages and are especially common in infant acute leukemias and secondary AML induced by Topo II inhibitors.²²¹ An aminonucleoside DOT1L inhibitor has entered clinical trials. It occupies the S-adenosyl-L-methionine (SAMe) binding pocket of DOT1L and induces conformational changes, thereby contributing to high-affinity binding and specificity of the inhibitor.

Preclinical data demonstrate nanomolar antiproliferative activity specific for MLL-rearranged cell lines, and in a rat xenograft model of MLL-rearranged leukemia, addition of the inhibitor led to complete tumor regression. Based upon these preclinical data, the DOT1L inhibitor has recently entered clinical trials in relapsed/refractory patients with leukemia involving the translocation of the MLL gene at 11q23 (NCT01684150).²²¹

Enhancer of Zest Homologue 2 Inhibitors

Enhancer of zest homologue 2 (EZH2) is the catalytic subunit of the polycomb repressive complex 2 (PRC2), and methylates lysine K-27 on histone H3 as well as other nonhistone targets. Monoallelic point mutations in EZH2 lead to increased catalytic activity and hyperdimethylation or hypertrimethylation of H3K27 with concomitant transcriptional repression of DNA-damage response pathways in germinal center diffuse large B-cell lymphoma and follicular lymphoma.²²³ The Y641 residue is most commonly mutated with up to 22 percent of germinal center diffuse large B-cell lymphoma and 10 percent of follicular lymphoma harboring the mutation.²²³ Interestingly, loss-of-function mutations of EZH2 in myeloid malignancies confer a worse prognosis, suggesting that EZH2 may act either as an oncogene or as a tumor suppressor depending on context and site of mutation.

EZH2 inhibitors have been identified through high-throughput biochemical assays using mutant EZH2 with the PRC2 complex and histone substrates. Preclinical data demonstrate increased EZH2 inhibitor selectivity for cell lines carrying activating EZH2 mutations. In mouse models with subcutaneous xenografts of diffuse large B-cell lymphoma cell lines harboring EZH2 activating mutations, marked tumor regression was observed. Based upon these results, EZH2 inhibitors are now in early phase 1 and phase 2 clinical trials in patients with relapsed and refractory diffuse large B-cell lymphoma with EZH2 mutations.²²⁴

Bromodomain and Extraterminal Inhibitors

The class of bromodomain and extraterminal (BET) family of proteins couple histone acetylation with transcriptional activation by recognizing acetylated lysine residues in histones and recruiting members of the pTEF-b complex to promoters.²²⁵ BET proteins also interact with nonhistone acetylated proteins including p53. BET inhibitors selectively bind to the high conserved bromodomains of the BET proteins thereby inhibiting their ability to bind acetylated lysine residues within histones.

Preclinical data²²⁶ with BET inhibitors demonstrate activity in both myeloma and AML, particularly in cell lines containing MLL fusion proteins or mutant NPM1c+. BET inhibitors induce early cell-cycle arrest and apoptosis via inhibition of c-MYC and other downstream targets.²²⁶ Survival benefits have also been seen in MLL mouse models. HDAC and BET inhibitors act synergistically by maintaining both histone lysine residues and p53 in an acetylated state thereby inducing greater reliance on BET mediated transcription.^227,228 Based upon these results the first BET inhibitors have entered early phase clinical trials.²²⁹

BCR-ABL TYROSINE KINASE INHIBITORS

The era of targeted cancer therapy was pioneered in the treatment of CML with imatinib mesylate. This carefully designed compound is an inhibitor of ABL tyrosine kinase activity and particularly efficacious against the mutant ABL characteristic of the BCR-ABL fusion protein. This protein is a result of the translocation t(9;22)(q24,q11.2), also known as Philadelphia chromosome, a transforming genetic event, which produces growth factor independent proliferation and sensitizes affected leukemic cells to inhibition with imatinib and other tyrosine kinase inhibitors (TKIs).²³⁰ Imatinib was selected for clinical study by scientists at Ciba-Geigy (later Novartis) based on a high-throughput screen for kinase inhibition and was the first drug of this class approved in 2001 for the treatment of CML based on the large phase III International Randomized Study of Interferon and STI571 (IRIS) trial, which showed induction of durable remissions in a large proportion of patients.^231,232 Since then the three additional second-generation agents dasatinib, nilotinib and bosutinib, as well as the third-generation TKI ponatinib, have been approved for imatinib-refractory or intolerant patients (Table 22–5). Another non-TKI agent, omacetaxine is approved for patients with resistance or intolerance to two or more TKIs.²³³

Mechanism of Action

Imatinib, nilotinib, dasatinib, bosutinib, and ponatinib (Fig. 22–7) are all inhibitors of the BCR-ABL kinase, as well as the c-KIT kinase²³⁴ and the platelet-derived growth factor receptor (PDGFR) kinase. The c-KIT kinase is the target for imatinib in the treatment of gastrointestinal stromal tumors,²³⁵ while activating mutations of PDGFR are the target of inhibition in the treatment of hypereosinophilia syndrome,²³⁶ chronic myelomonocytic leukemia,²³⁷ MDS with PDGFR rearrangement^237A and dermatofibrosarcoma protuberans.²³⁸ Dasatinib and ponatinib also inhibit the Src family kinases, an important secondary target in CML.²³⁹ Dasatinib (inhibiting growth by concentration 50 percent [IC₅₀] = <1 nM), nilotinib (IC₅₀ = <20 nM),²⁴⁰ bosutinib (IC₅₀ = <1 nM), and ponatinib (IC₅₀ = <1 nM and 10 nM for T315I) are more potent inhibitors of BCR-ABL compared to imatinib (IC₅₀ = 100 nM). Crystallographic and mutagenesis studies indicate that imatinib and nilotinib bind to a segment of the BCR-ABL tyrosine kinase domain that fixes the enzyme in a closed or inactive state, in which the protein is unable to bind its substrate, ATP.^240,241,242 The contact points between imatinib and the enzyme become sites of mutations in drug-resistant leukemic cells, preventing tight binding of the drug and locking the enzyme in its active configuration, in which it has access to substrate. Nilotinib has been modified to overcome a number of resistance mutations to imatinib.^{239,241,243,244,245} Dasatinib is unique in that it is able to bind BCR-ABL in both the active and inactive configuration, which may be one of the mechanisms that allows it to overcome resistance.²³⁹ Importantly, ponatinib is the only agent with significant activity against the most common resistance mutation T315I, a gatekeeper mutation that prevents other TKIs from binding to the ATP binding site of BCR-ABL.²⁴⁶

Clinical Pharmacology

The BCR-ABL kinase inhibitors are all well absorbed by the oral route and subject to clearance by hepatic CYP3A4 metabolism. The absorption of dasatinib and ponatinib is pH-dependent and may be affected by the use of H₂ blockers and proton pump inhibitors. The absorption of bosutinib may be impaired by concomitant magnesium intake. The bioavailability of nilotinib is increased if taken with meals and therefore should be taken on an empty stomach. Clearance of imatinib is delayed in patients with renal dysfunction, apparently as a result of decreased P450 activity in the presence of renal failure. Limited data indicate that imatinib penetrates poorly into the cerebrospinal fluid, achieving concentrations of 1 percent of simultaneous drug levels in the systemic circulation.²⁴⁷ There are no data about the penetration of other BCR-ABL inhibitors in the cerebral spinal fluid.

All FDA-approved BCR-ABL inhibitors are more than 94 percent protein bound, largely by α₁-acid glycoprotein, a binding protein present in higher concentrations in humans than in mice.²⁴⁸ Thus, therapeutic studies in mice may overpredict drug activity. α₁-Acid glycoprotein concentrations vary over a fourfold range in human subjects, and total drug concentrations in plasma appear to be a function of α₁-acid glycoprotein levels. Drugs that compete for binding sites with α₁-acid glycoprotein, such as clindamycin, displace imatinib mesylate from binding to α₁-acid glycoprotein and, in mice, increase the concentration of drug found in cells.

There is significant variability among different BCR-ABL inhibitors with regard to their bioavailability. Imatinib has the highest bioavailability (98 percent), followed by bosutinib (23 to 64 percent), nilotinib (50 percent), and dasatinib (14 to 34 percent), while the bioavailability is unknown for ponatinib.

Despite their unparalleled benefit in the treatment of patients with CML and other malignancies, resistance to TKIs represent a major limitation. Although there are multiple mechanisms of resistance, one can divide resistance to TKIs into two general categories: primary resistance, which refers to a de novo lack of response to a drug, and secondary (acquired) resistance, in which resistance to a drug emerges after a period of drug response. The most important and most frequently found resistance mechanisms to TKIs are point mutations in three different segments of the kinase domain leading to inability of the drug to effectively inhibit the BCR-ABL kinase activity.²⁴⁹ The most common mutations associated with clinical resistance affect amino acids 255 and 315, both of which serve as contact points; these mutations confer high-level resistance to imatinib and nilotinib. There is variable efficacy of different BCR-ABL inhibitors to particular resistance mutations. For example, dasatinib can bind to both the active and inactive conformation and can overcome resistance to substitution at 255 but not 315.^241,250 Nilotinib has good activity against most resistance mutations, but lacks activity against substitutions at 255 and 315. Of all approved TKIs, ponatinib is the only agent with activity against the gatekeeper mutation T315I. In a phase II trial of heavily pretreated patients with Philadelphia chromosome positive leukemias, including those who failed dasatinib or nilotinib and with T315I mutations, ponatinib produced major cytogenetic responses in more than half of the patients.²⁴⁶

Mutations also affect the phosphate-binding region and the “activation loop” of the domain with varying degrees of associated resistance. Some mutations, such as at amino acids 351 and 355, confer low levels of resistance to imatinib, while these tumor cells remaining sensitive to higher imatinib doses and also sensitive to both nilotinib and dasatinib.^241,251 This may explain the clinical response of some resistant patients to dose escalation of imatinib.

Some kinase mutations known to cause drug resistance may be present at low allelic frequencies prior to initiation of therapy. Malignant cells carrying these mutations may grow out under the pressure of drug and are therefore detected at higher rates after drug exposure. This observation was made particularly in patients with Philadelphia chromosome–positive ALL or with CML progressing to blastic crisis ^252,253 and strongly supports the hypothesis that drug-resistant cells arise through spontaneous mutation, and are further selected by drug exposure.

The site of the resistance mutation has predictive and prognostic implications. Among CML patients, mutations are detectable in some patients receiving imatinib, including one-third of those undergoing treatment in the accelerated phase and in late (longer than 4 years from diagnosis) chronic phase CML.²⁵² Most patients with mutations demonstrate clinical resistance at the time a mutation is detected, or shortly thereafter. Prior to the development of second generation TKIs, mutations involving the phosphate binding loop were associated with rapid disease progression and death within a median of 4.5 months. The availability of highly active second- and third-generation TKIs that are effective against all known kinase mutations resulted in improved disease control and overall survival in these patients.^{246,254,255,256} As first-line agents, nilotinib and dasatinib are similarly effective, demonstrating high rates of progression free survival and overall survival. The choice of a second-line TKI depends on disease and patient characteristics as well as the side-effect profile of the different agents. For example, currently, there is no alternative to ponatinib for CML with T315I mutation. In patients with targeted CHF or pleural effusions, dasatinib should be avoided given high rate of pleural effusions in up to 35 percent of patients. These patients should be given nilotinib or bosutinib instead. In patients with severe diabetes or a history of pancreatitis, one should avoid nilotinib and choose another second-generation TKI when possible. Patients who fail two or more TKIs, including those with T315I mutation, may also respond to omacetaxine mepesuccinate, a cytotoxic natural product that was FDA approved.²³³

In addition to kinase mutation, amplification of the wild-type kinase gene, leading to overexpression of the enzyme, has been identified in tumor samples from a few patients with resistance to treatment.²⁵⁷ The MDR gene, which codes for a drug efflux protein, confers resistance to imatinib experimentally²⁵⁸; thus far this mechanism has not been implicated in clinical resistance. In addition to efflux mechanisms, influx mechanisms may also play an important role. Recent studies indicate that imatinib but not nilotinib or dasatinib is taken into cells via the organic cation transporter-1 (OCT1) and that downregulation of this pathway may confer resistance.²⁵⁹

Not all resistance is explained by kinase amplification or mutation, or by pharmacokinetic factors. There is a growing awareness of the appearance of mutant Philadelphia chromosome–negative clones carrying the karyotype of myelodysplastic cells in patients receiving imatinib for CML, and a few cases of progression to MDS and AML have been reported.^260,261 Ongoing research into novel agents for resistant CML includes evaluation of HDAC inhibitors, heat shock protein inhibitors, and targeted therapies of alternative pathways.

Adverse Effects

Imatinib, dasatinib, nilotinib, and bosutinib have modest toxicity. All cause low levels of gastrointestinal distress, including nausea and vomiting. Significant diarrhea occurs more frequently with use of imatinib and bosutinib. All can promote fluid retention resulting in peripheral edema and pleural effusions, with dasatinib causing significantly more edema than the other drugs of this class.^262,263 Although all agents cause rashes, imatinib and bosutinib tend to promote more-severe (grades III/IV) rashes in up to one-third of patients. Mild elevation of transaminases is seen with all BCR-ABL inhibitors, but more-severe elevation with bosutinib. Bilirubinemia is a rather uncommon adverse effect but it is frequently observed with use of nilotinib. Dasatinib and nilotinib cause a prolongation of the QT interval that is not seen in patients receiving other TKIs. Other nonhematologic adverse effects include hypophosphatemia (primarily dasatinib and nilotinib), muscle pain, pancreatitis, and weight gain (particularly imatinib). All agents can induce neutropenia, anemia and thrombocytopenia that can require transfusion support, dose reduction, or discontinuation. Most nonhematologic adverse reactions are self-limited and respond to dose adjustments. After the adverse events have resolved, many times the drug may be retitrated back to initial dosing. Ponatinib poses a clearly increased risk of arterial thrombosis, and should be used with caution in patients with a history of myocardial infarction, angina, stroke, or peripheral arterial disease.

JANUS KINASE INHIBITORS

Myeloproliferative neoplasms (MPNs) are a group of heterogeneous clonal hematopoietic stem cell disorders that include CML and “BCR-ABL–negative” MPNs polycythemia vera (PV), essential thrombocythemia (ET), and primary myelofibrosis (PMF).²⁶⁴ A phenotypic characteristic of these disease is the accumulation of mature-appearing myeloid cells. The majority of patients with BCR-ABL–negative MPNs carry a mutation in the Janus-type tyrosine kinase (JAK) gene, the most common being JAK^V617F.^265,266,267 Family members of the JAKs include JAK1 to JAK3 and TYK. Physiologically, JAK are necessary for intracellular signal transduction of receptors that have no intrinsic tyrosine kinase activity, such as receptors for erythropoietin (EPO), thrombopoietin (TPO), and granulocyte-macrophage colony-stimulating factor (GM-CSF).²⁶⁸ Upon ligand binding to the receptor, JAK autophosphorylation leads to binding of signal transducer and activator of transcription (STAT), which dimerizes with another STAT protein, translocates into the nucleus, and promotes transcription of STAT-responsive genes, which are involved in the control of cell proliferation, apoptosis, and cell differentiation.²⁶⁵ The substitution mutation JAK2^V617F is the most common gain-of-function alteration and occurs in 65 to 97 percent of cases in PV, 23 to 57 percent of cases in ET, and 34 to 57 percent of cases in PMF.^261,262,263 Expression of this mutation results in ligand independent growth or increased sensitivity to the cytokine/growth factor.

In 2011 the FDA approved the first specific JAK inhibitor ruxolitinib (Jakafi) for the treatment of PMF, PV, and ET. This approval was the consequence of two phase III trials: Controlled Myelofibrosis Study with Oral JAK Inhibitor Treatment (COMFORT)-I trial that investigated the activity of ruxolitinib (15 mg or 20 mg orally twice daily) versus placebo in 309 patients with PMF, PV, or ET, and the COMFORT-II trial that assessed ruxolitinib versus best available therapy in 219 patients with PMF, PV, or ET.^266,267 In the COMFORT-I trial, ruxolitinib produced greater than 35 percent spleen volume reduction (primary outcome) in 42 percent of patients at 24 weeks, improved symptom control, and overall survival compared to the placebo group. In the COMFORT-II trial 28 percent versus 0 percent of patients had greater than 35 percent spleen volume reduction, as well as superior reduction of disease-related symptoms, functionality and quality of life with modest toxicities compared to best available treatment (mostly hydroxyurea and glucocorticoids).

Mechanism of Action

Ruxolitinib inhibits all JAK kinases independent of their mutational status and to similar degree independent from the disease subtype. There is variable activity against JAK1 (IC₅₀ = 1 nM), JAK2 (IC₅₀ = 7.2 nM), TYK2 (IC₅₀ = 9.3 nM), and JAK3 (IC₅₀ = 98 nM), respectively. Molecular dynamics simulations suggest that ruxolitinib targets the ATP-binding site of the kinase in its active conformation.²⁶⁸ Administration of ruxolitinib results in decreased expression of STAT responsive genes.

Clinical Pharmacology

Ruxolitinib is orally administered, rapidly absorbed, and with a high bioavailability of at least 95 percent. It undergoes CYP3A4-dependent metabolism and is primarily eliminated in urine. The elimination half-life is approximately 3 hours. High-throughput in vitro screens identified mutations that confer primary resistance to ruxolitinib, including the gatekeeper mutation M929I.²⁶⁹

Adverse Effects

Most nonhematologic adverse effects are relatively infrequent and mild in nature. These include, but are not limited to, diarrhea, nausea, peripheral edema, nasopharyngitis, pyrexia, arthralgia, cough, and dyspnea. Ruxolitinib can cause anemia and thrombocytopenia that may require transfusions. However, toxicity can be ameliorated with dose reduction or treatment interruptions and only a very small portion of patients had to discontinue the drug. The rate of higher-grade neutropenia was low (~7 percent).^266,267

IBRUTINIB

CLL is a slowly progressive, indolent hematologic malignancy of mature B lymphocytes.²⁷⁰ For decades, alkylating agents were the mainstay for treatment of CLL, with modest benefit for survival. The addition of ­rituximab, an antibody directed against CD20 on B lymphocytes, improves survival in CLL patients.²⁷¹ However, subsets of patients with particular deletion, such as 17p13.1, have poor response to this treatment. Furthermore, none of the available therapies are curative.²⁷¹ Although no common driver mutation is found in CLL, Bruton tyrosine kinase (BTK), a downstream mediator of the B-cell receptor, appears to be critical for B-cell activation, proliferation, and survival. A selective BTK inhibitor, ibrutinib, was developed and proved highly active in high-risk relapsed CLL. It achieved a high rate of overall (88 percent) and progression-free survival (75 percent) at 26 months of followup.²⁷² In a subsequent phase III trial including 391 patients with previously treated CLL, ibrutinib was compared to ofatumumab (an anti-CD20 antibody approved for relapsed CLL). Ibrutinib achieved a superior overall response rate, progression-free and overall survival.²⁷³ Based on these findings, ibrutinib was granted accelerated approval by the FDA for this indication. It has also been approved for the treatment of patients with mantle cell lymphoma.

Mechanism of Action

Ibrutinib irreversibly inhibits the BTK active site by forming a covalent bond with a cysteine residue. Inhibition of the BTK results in disruption of activation of multiple downstream pathways important for B-cell activation, proliferation, and adhesion.

Clinical Pharmacology

Ibrutinib is orally administered once daily. The bioavailability has not been established. The half-life is 4 to 6 hours. Ibrutinib is metabolized in the liver primarily by CYP3A4 and only minimally cleared intact in urine (1 percent). No dose adjustments are required for liver or renal dysfunction. Dose adjustment may be necessary for patients who have more than grade 3 hematologic of nonhematologic adverse effect. Using whole exome sequencing, distinct resistance mutations were identified in six patients.²⁷⁴ In five patients, a cysteine-to-serine mutation (C481S) at the binding site of BTK was found. Interestingly, the resulting protein remains sensitive to ibrutinib; however, inhibition in this case is reversible. Three other mutations (L845F, R665W, and S707Y) found in two patients involving a downstream target of BTK, phospholipase γ2 (PLCγ2). Currently, it is unclear whether dose escalation may overcome these resistance mechanisms.

Adverse Effects

The vast majority of adverse effects to ibrutinib are mild in nature. Almost half of the patients experience diarrhea, and approximately one-third of patients develop upper respiratory infections, cough, or fatigue. Other adverse effects include nausea, vomiting, constipation, pyrexia, rashes, edema, hypertension, and headaches. Approximately 15 percent of patients developed grades 3 to 4 neutropenia, dose reduction allowed continuation of the drug.^272,273

PROTEAOSOME INHIBITORS

Bortezomib and Carfilzomib

This class of agents targets the ubiquitin-proteasome pathway, the complex regulatory system whereby normal and malignant cells eliminate potentially toxic misfolded proteins and control the intracellular levels of important regulatory proteins.²⁷⁵ The importance of this pathway and its substrates in diverse aspects of the pathophysiology of human tumors creates opportunities for therapeutic interventions. The multimeric 20S core particle of the proteasome exhibits three distinct proteolytic activities, namely chymotryptic, tryptic, and post–glutamyl peptide hydrolytic-like activities. Both bortezomib (previously known as PS-341) and carfilzomib (Fig. 22–8) inhibit the chymotryptic-like activity by binding to the β5 subunit of the 20S core. Bortezomib, a boronic dipeptide, is a reversible inhibitor of the chymotryptic-like activity, while carfilzomib, an epoxyketone, forms with the β5 subunit an irreversible adduct through two covalent bonds.²⁷⁶ Bortezomib has potent clinical activity against myeloma^277,278,279 and other plasma cell dyscrasias, including amyloidosis²⁸⁰ and Waldenström macroglobulinemia,²⁸¹ and is also active in mantle cell lymphoma.^282,283 Bortezomib induces complex molecular sequelae, such as decreased levels of antiapoptotic molecules (including nuclear factor [NF]-κB, caspase inhibitors, and Bcl-2 family members).^{284,285,286,287} The composite outcome of these events is an irreversible commitment of myeloma cells to apoptosis,²⁸⁵ as well as the sensitization of myeloma cells to diverse established agents (e.g., alkylator,^287,288 anthracyclines,^287,289 and thalidomide derivatives) or investigational agents such as HDAC inhibitors²⁸⁷. As a result, bortezomib has been a key component of diverse antimyeloma combination regimens.²⁹⁰ However, patients eventually develop resistance to bortezomib or experience sensory peripheral neuropathy,²⁹¹ the main dose-limiting toxicity for this drug. Carfilzomib and other second-generation proteasome inhibitors were thus developed to circumvent these limitations. In 2012, carfilzomib received FDA accelerated approval for the treatment of myeloma patients relapsed and refractory to bortezomib and at least one thalidomide derivative.²⁹² The observation that carfilzomib can be active in some bortezomib-refractory cases has been attributed to the fact that new proteasomes must be synthesized to fully restore proteasome capacity in cells treated with this irreversible inhibitor. However, the reversible-versus-irreversible nature of a proteasome inhibitor may not be the only determinant of its clinical activity; for example, another second-generation proteasome inhibitor in clinical trials, MLN2238, and its clinically administered prodrug, ixazomib (MLN9708), exhibit preclinical²⁹³ and clinical activity in bortezomib-resistant cells, despite reversible binding to β5.²⁹⁴

Despite the important role of the proteasome for normal cells, the administration of bortezomib and carfilzomib is associated with a clinically meaningful therapeutic window, likely because the clinically achievable levels of these agents do not completely shut down the chymotrypsin-like activity²⁹⁵ of the proteasome and also spare its other proteolytic (trypsin-like and caspase-like)²⁹⁶ activities. Consequently, there is only modest (<40 percent) decrease in the overall rate of protein degradation in either normal or tumor cells. Although normal cells can recover from this degree of perturbation, malignant plasma cells may not, because their available active proteasome particles (“proteasome capacity”) are apparently close to their functional saturation by the increased quantities of unassembled or misfolded proteins (“proteasome load”), such as immunoglobulins.²⁹⁷ It is not yet determined, if this concept also applies to mantle cell lymphoma.

The safety profile of both bortezomib and carfilzomib includes thrombocytopenia, possibly reflecting a requirement for constitutive proteasome activity in platelets to degrade the apoptosis regulator, Bax, and preserve their normal life span.²⁹⁸ Unlike bortezomib, carfilzomib does not cause peripheral neuropathy, but cardiopulmonary side effects (e.g., dyspnea, hypoxemia, pulmonary hypertension) and serum creatinine elevations have been reported.²⁹² These differences may be explained by reports that bortezomib, but not carfilzomib, also inhibits the neuroprotective molecule, Htra2/Omi,²⁹⁹ and blocks a number of serum proteases, including cathepsin G and cathepsin A, which may contribute to renal injury.^300,301,302; inhibition of proteases may contribute to renal injury.³⁰³ The mechanistic basis for cardiopulmonary adverse events with proteasome inhibition remains under investigation.

The clinical activity of bortezomib and carfilzomib, as well as the promising early studies with additional second-generation proteasome inhibitors, has validated the notion that pathways for intracellular regulation of protein homeostasis represent an important therapeutic target for plasma cell dyscrasias, mantle cell lymphoma, and potentially other neoplasias.

Monoclonal antibodies are an important class of agents for the treatment of hematologic malignancies. As a group, lymphoid cells express a variety of antigens that are attractive targets for monoclonal-based therapy, as shown in Table 22–6. Development of monoclonal antibodies against specific targets has been largely accomplished by the empiric method of immunizing mice against human tumor cells and screening the hybridomas for antibodies of interest. Because murine antibodies have a short half-life and induce a human antimouse antibody immune response, they are partially or fully humanized when used as therapeutic reagents. Presently, several monoclonal antibodies have received FDA approval for non-Hodgkin lymphoma and CLL, including rituximab and alemtuzumab. Although several mechanism(s) of action have been described for monoclonal antibodies, including direct induction of apoptosis, antibody-dependent cellular cytotoxicity (ADCC), and complement-dependent cytotoxicity, the clinically important mechanisms for most antibodies remain uncertain.³⁰⁴

Monoclonal antibodies may also be engineered to combine the antibody with a toxin (immunotoxin) or a radioactive isotope (radioimmunoconjugates), or to contain a second specificity (bispecific antibodies).^305,306,307 For example, it is possible to conjugate an antibody with specificity to B-cell lymphomas with an antibody against CD3, which binds to and activates normal T cells, so as to enhance T-cell–mediated lysis of the lymphoma cell. One such example of a bispecific antibody blinatumomab contains anti-CD3 and anti-CD19 specificity. Monoclonal antibodies raised against the immunoglobulin idiotype on a B-cell lymphoma represent another therapeutic strategy, which was first reported in 1982 by Miller and associates.³⁰⁸

NAKED MONOCLONAL ANTIBODIES

Rituximab

Rituximab, the first monoclonal to receive FDA approval, is a chimeric antibody containing the human immunoglobulin G₁ and κ constant regions with murine variable regions. Rituximab targets the B-cell antigen CD20 expressed on the surface of normal B cells and on more than 90 percent of B-cell neoplasms, and is present from the pre–B-cell stage through terminal differentiation to plasma cells.³⁰⁹ To date, the biologic functions of CD20 remain uncertain, although incubation of B cells with anti-CD20 antibody has variable effects on cell-cycle progression, depending on the monoclonal antibody type.^310,311 Monoclonal antibody binding to CD20 generates transmembrane signals that produce a number of events including autophosphorylation and activation of serine/tyrosine protein kinases, and induction of c-myc oncogene expression and major histocompatibility complex class II molecules.³¹² CD20 promotes transmembrane Ca²⁺ conductance through its possible function as a Ca²⁺ channel.³¹² These studies demonstrate the importance of CD20 in B-cell regulation, but do not in themselves indicate how ligation of the receptor produces cell death independent of ADCC or complement-mediated pathways.

Rituximab was initially approved as a single agent for relapsed indolent lymphomas, but it has shown activity in a wide variety of clinical settings. It is approved in combination with chemotherapy for the initial treatment of follicular and diffuse large B cell lymphoma.^313,314 Rituximab has also been shown to be effective in combination with chemotherapy for CLL and indolent B-cell non-Hodgkin lymphomas including mantle cell lymphoma, Waldenström macroglobulinemia, and marginal lymphomas.³¹³ It is used in combination with salvage chemotherapy for many indolent and aggressive B-cell non-Hodgkin lymphomas even after prior rituximab treatment.³¹⁵ The use of maintenance rituximab has gained increased acceptance based the demonstration of delayed time to progression and improved overall survival in the relapse setting.^316,317,318

Rituximab is given by intravenous infusion both as a single agent and in combination with chemotherapy at a dose of 375 mg/m². As a single agent it is given weekly for 4 weeks with maintenance dosing every 3 to 6 months. It has a half-life of approximately 22 days.³¹⁹ Given the risk of infusional reactions, pretreatment with antihistamines, acetaminophen, and glucocorticoids have become standard. During the first administration, the rate must be increased slowly to prevent infusional reactions. Infusions are started at 50 mg/h and in the absence of infusion reactions the rate can be increased in 50 mg/h increments every 30 minutes to a maximum rate of 400 mg/h. On subsequent infusion in the absence of reactions, infusions may start at 100 mg/h and increased in 100 mg/h increments every 30 minutes to a maximum rate of 400 mg/h. Patients with a high degree of circulating tumor cells are at increased risk for tumor lysis syndrome and are given a reduced dose of 50 mg/m² on day 1 of treatment in addition to standard tumor lysis prophylaxis. The remainder of the dose can then be given on day 3.

Resistance to rituximab may occur by down regulation of CD20, impaired ADCC, decreased complement activation, limited effects on signaling and induction of apoptosis, and inadequate blood levels.^304,320 Studies in relapsed indolent B-cell non-Hodgkin lymphoma have shown a correlation between higher mean serum rituximab levels and clinical responses,³²¹ suggesting that dose escalation may be a way to overcome rituximab resistance. Also polymorphisms in two of the receptors for the antibody Fc region responsible for complement activation, FcγRIIIa, and FcγRIIa may predict the clinical response to rituximab monotherapy in patients with follicular lymphoma but not in CLL.^322,323

Rituximab has a several known toxicities including infusional reactions which can be life-threatening. Most are usually mild and include fever, chills, throat itching, urticaria, and mild hypotension, all of which can respond to decreased infusion rates and antihistamines. There are reports of severe mucocutaneous skin reactions including Stevens-Johnson syndrome.³²⁴ Reactivation of hepatitis B virus also has been reported and it is recommended that patients be screened for hepatitis B prior to initiation of therapy. There also are case reports of progressive multifocal leukoencephalopathy caused by the John Cunningham virus and resulting in death.³²⁵ Hypogammaglobulinemia and delayed neutropenia may occur from 1 to 5 months after administration with resultant serious infections.^326,327

Ofatumumab

Ofatumumab is a fully human immunoglobulin G₁ kappa (IgG₁κ), monoclonal antibody which targets a unique epitope on the CD20 antigen found on the surface of normal and malignant pre-B and mature B cells. Ofatumumab offers many potential advantages over rituximab. It has an increased affinity for CD20 compared to rituximab.³²⁸ In vitro experiments have shown it is superior to rituximab in ability to induce cell lysis.³²⁸ Ofatumumab and rituximab both show similar antibody-dependent cellular cytotoxicity, but ofatumumab displays increased complement dependent cytotoxicity.³²⁹ In addition, ofatumumab is a fully human monoclonal antibody with a low incidence of human–antihuman antibody formation.³³⁰

Ofatumumab has been approved for use in patients with CLL refractory to fludarabine and alemtuzumab. In a phase II trial, 138 patients received eight weekly infusions of ofatumumab followed by four monthly infusions during a 24-week period (dose 1 = 300 mg; doses 2 to 12 = 2000 mg). The overall response rate was 58 percent with median progression-free survival of approximately 14 months.³³¹ No dose-limiting effects were observed in the phases I/II studies of ofatumumab. Most common adverse reactions (>10 percent) were infusion reactions, neutropenia, pneumonia, pyrexia, cough, diarrhea, anemia, fatigue, dyspnea, rash, nausea, bronchitis, and upper respiratory tract infections. Similar to rituximab above, ofatumumab is being investigated in combination with chemotherapy for both indolent and aggressive B cell non-Hodgkin lymphoma.

Obinutuzumab

Obinutuzumab is a third naked monoclonal antibody that targets the CD20 antigen. It differs from rituximab in that it is a glycoengineered type 2 antibody that, in preclinical studies, showed improved efficacy by inducing direct cell death and by enhanced ADCC.³³² Obinutuzumab was recently approved in combination with chlorambucil of the treatment of patients with previously untreated CLL. This approval was based on a phase III, 781-person trial with three arms comparing single-agent chlorambucil versus chlorambucil plus rituximab versus chlorambucil plus obinutuzumab.³³³ A main feature of this trial was that it enrolled mainly older patients (median age: 73 years) with multiple medical comorbidities representative of the majority of patients diagnosed with CLL. The median progression-free survival was greatly improved in obinutuzumab plus chlorambucil arm at 26.7 months versus 11.1 months with chlorambucil alone and 16.3 months with rituximab plus chlorambucil. Obinutuzumab is similar to other monoclonal antibodies associated with infusion reactions. The first dose was split over 2 days with the patient receiving 100 mg intravenously on cycle 1 day 1 and 900 mg on day 2. Patients then received 1000 mg on days 8 and 15 of cycle 1 and subsequently 1000 mg on days 1 of cycles 2 through 6. Patients concurrently received chlorambucil orally at a dose of 0.5 mg/kg on days 1 and 15 of each cycle. This combination was, in general, well tolerated. There was a slightly increased risk of grade 3 neutropenia (33 vs. 28 percent) compared to the rituximab control group, but no increased risk of infection (12 vs. 14 percent). Other common toxicities included infusion reactions, anemia, thrombocytopenia, and leukopenia.

Alemtuzumab

Alemtuzumab is a humanized monoclonal antibody targeted against the CD52 antigen present on the surface of normal neutrophils and lymphocytes as well as most B- and T-cell lymphomas.³³⁴ CD52 is expressed at reasonable levels and does not modulate with antibody binding, making it a good target for unconjugated monoclonal antibodies. Mechanistically, alemtuzumab can induce tumor cell death through ADCC and complement dependent cytotoxicity.³³⁵ Clinical activity has been demonstrated in CLL, including in patients with purine analogue refractory disease.^336,337 In refractory CLL, overall response rates are approximately 38 percent with complete responses of 6 percent in multiple series. Response rates in patients with untreated CLL are higher (overall response rates of 83 percent and complete responses of 24 percent).³³⁶ The most concerning side effects are acute infusion reactions and depletion of normal neutrophils and T cells. Opportunistic infections are a serious consequence, particularly in patients who have received purine analogues.^336,337 Patients should receive antibiotic prophylaxis against Pneumocystis carinii and herpes virus during treatment. Serious infections with cytomegalovirus are seen and monitoring is recommended. Alemtuzumab in combination with chemotherapy has been explored in the treatment of T-cell lymphomas but trials were limited by significant infectious complications.³³⁸

Elotuzumab, Daratumumab, and Others

A number of naked antibodies have been developed for the therapeutic targeting of myeloma. Mechanistically, this approach depends on the recruitment of ADCC, complement-dependent cytotoxicity, and apoptosis, as well as growth arrest via the selected targeting of signaling pathways (Fig. 22–9).³³⁹ Targets include surface molecules as well as signaling molecules (Table 22–7).³³⁹ Of these antibodies, one of the most promising is elotuzumab, which specifically targets CS1 (also known as SLAMF7) through which its mechanism of action includes the enhancement of NK-cell activation directly via SLAMF7 and indirectly by CD16, as well as the targeted killing of SLAMF7-expressing myeloma cells by ADCC.^340,341 Combinatorial strategies for this particular approach have been especially promising, including regimens with lenalidomide and dexamethasone where high qualities of response and impressive progression-free survival have been seen.³⁴¹

Separately, daratumumab as a first-in-class monoclonal antibody that effectively targets CD38 has shown particularly encouraging activity as both a monotherapy and in combination.³⁴² Similarly, the CD38-targeting monoclonal antibody SAR650984 has shown impressive activity, validating this as an appropriate target as well as further supporting the therapeutic potential of this class of antibodies.^339,340 Combination studies are ongoing and the outlook for the rational development of several monoclonal antibody combinations in myeloma now looks very promising.^341,342

IMMUNOTOXINS

Immunotoxins combine immune proteins such as antibodies, antibody Fab fragments, or interleukins and toxins such as ricin A chain or Pseudomonas exotoxin. These molecules have the advantage of the high specificity of the protein for its receptor or antigen, and its ability to internalize once bound to its receptor, together with the potency of the toxin molecule. Two agents, denileukin diftitox (which combines IL-2 and the catalytically active fragment of diphtheria toxin) with activity in non-Hodgkin lymphoma and gemtuzumab ozogamicin (composed of an antibody that recognizes CD33 linked to a potent chemical toxin, calicheamicin) with activity in AML, show a proof of principle but have been removed from the market.^343,344 Brentuximab vedotin is currently the only approved antibody drug conjugate for hematologic malignancies but several others are in development.

Brentuximab Vedotin

Brentuximab vedotin is an antibody drug conjugate comprising an anti-CD30 monoclonal antibody conjugated by a protease-cleavable linker to the potent antimicrotubule agent monomethylauristatin E. When brentuximab vedotin binds to CD30-expressing cells, it is internalized with proteolytic cleavage of the linker resulting in release of monomethylauristatin E, disrupting the microtubule network within the cell, and subsequently inducing cell-cycle arrest and apoptotic death of the cells.³⁴⁵ It is approved for the treatment of Hodgkin lymphoma after the failure of an autologous stem cell transplant or after the failure of at least two prior multiagent chemotherapy regimens in patients who were not candidates for transplantation. It is also approved for the treatment of patients with systemic anaplastic large cell lymphoma after the failure of at least one prior multiagent chemotherapy regimen. In a study of refractory Hodgkin lymphoma, brentuximab vedotin had an impressive overall response rate of 75 percent and a complete remission rate of 34 percent with 96 percent of patients having some improvement in their disease.³⁴⁶ The median progression-free survival for all patients was 5.6 months but notably the median duration of response for those achieving a complete remission was 20.5 months. In refractory anaplastic large cell lymphoma, a similarly impressive overall response rate of 86 percent was seen with 57 percent of patients achieving a complete remission and 97 percent of patients having a reduction in tumor volume.³⁰⁷ The overall duration of response was 12.6 months and those in a complete remission experiencing a duration of response of 13.2 months. Dosing is at 1.8 mg/kg as an intravenous infusion every 3 weeks. Dose reductions to 1.2 mg/kg are recommended for patients with hepatic or severe renal impairment. The main toxicity is cumulative peripheral neuropathy reported in up to 54 percent of the patients in the above trials. Patients experiencing new or worsening peripheral neuropathy may require a delay, dose reduction, or discontinuation of brentuximab vedotin. Other common toxicities include neutropenia, thrombocytopenia, fatigue, and nausea.

RADIOIMMUNOCONJUGATES

Radioimmunoconjugates provide monoclonal antibody targeted delivery of radioactive particles to tumor cells.^306,347 Iodine-131 is a commonly used radioisotope because it is readily available, relatively inexpensive, and easily conjugated to a monoclonal antibody. The β-emitter ⁹⁰Y has emerged as an attractive alternative to ¹³¹I, based on its higher energy and longer path length, which may be more effective in tumors with larger diameters. It also has a short half-life and remains conjugated, even after endocytosis, providing a safer profile for outpatient use. Clinically, radioimmunoconjugates were developed with murine monoclonal antibodies against CD20 conjugated with ¹³¹I (tositumomab) or ⁹⁰Y (ibritumomab tiuxetan). Response rates in relapsed lymphoma of 65 to 80 percent were achieved.^306,347,348 Both drugs were well tolerated with most toxicity attributable to marrow suppression; however, there have been worrisome reports of secondary leukemias. Collaboration between treating physicians and nuclear medicine departments is required for administration. ¹³¹I tositumomab is no longer available, but ⁹⁰Y ibritumomab tiuxetan remains in clinical use and is being investigated as a consolidation therapy and as conditioning prior to hematopoietic cell transplant.