Alkylating agents were the first drugs introduced for the systemic therapy of cancer. They are a chemically diverse group of drugs, but they all contain alkyl groups (eg, —CH2Cl). In vivo, the alkyl groups generate highly reactive, positively charged intermediates, which then combine with an electron-rich nucleophilic group, such as an amino, phosphate, sulfhydryl, or hydroxyl moieties, on intracellular macromolecules, such as DNA. Although nucleophilic groups occur on almost all biological molecules, alkylation of bases in DNA appears to be the major cause of lethal toxicity.
Alkylating agents may contain 1 or 2 reactive groups and are thus classified as monofunctional or bifunctional, respectively. Bifunctional alkylating agents have the ability to form crosslinks between DNA strands and are the most clinically useful of these agents. Interstrand crosslinking of DNA, which prevents cell replication, unless repaired, seems to be the major mechanism of cytotoxicity for bifunctional alkylating agents, whereas the toxicity of monofunctional alkylating agents is probably related to single-strand breaks in DNA or to damaged bases. Mechanisms of resistance of alkylating agents include decreased transport across the cell membrane, increased intracellular thiol concentrations (eg, glutathione); such compounds react with alkylating agents and thus reduce the likelihood of interaction with DNA (see Chap. 19, Sec. 19.2.4), increased enzymatic detoxification of reactive intermediates, and alterations in DNA repair enzymes.
Because alkylating agents bind directly to DNA, they lack cell-cycle specificity, although they may have greater toxicity for proliferating cells. Common toxicities include myelosuppression, immunosuppression, hair loss, nausea, and vomiting. Some alkylating agents have long-term effects, such as infertility and carcinogenesis caused by long-lasting DNA damage. Nitrogen mustard, melphalan, and nitrosoureas are associated with an increased incidence of acute myelogenous leukemia, whereas cyclophosphamide is associated with irritation of the bladder and rarely development of bladder cancer. Mechanisms of resistance to alkylating agents include decreased transport across the cell membrane, increased intracellular thiol concentration leading to reduced DNA interactions, increased enzymatic detoxification of reactive intermediates, and alterations in DNA repair enzymes such as guanine-O6-alkyltransferase (Pegg, 1990; see Chap. 19, Sec. 19.2.6).
18.2.1 Nitrogen Mustards, Nitrosoureas, and Others
The development of nitrogen mustard and related compounds as anticancer drugs originated from the observation of lymphoid aplasia in men exposed to the more reactive, but chemically similar, sulfur mustard gas during World War II. This family of drugs, all chemically derived from nitrogen mustard (mechlorethamine) contains several drugs in common clinical use. The structures of these drugs are shown in Figure 18–2; each is bifunctional, with 2 chloroethyl groups that form the reactive electron-deficient groups responsible for alkylation of DNA.
Structures of clinically used alkylating agents of the nitrogen mustard family.
The most common site of alkylation of DNA by nitrogen mustards is the N-7 position on the base guanine (Fig. 18–3). First, one of the chloroethyl side chains releases a chloride ion, resulting in a highly reactive, positively charged intermediate. This intermediate then binds covalently with the electronegative N-7 group on a guanine base, resulting in alkylation. Alkylation of guanine leads to mispairing with thymine or to strand breakage. The second chloroethyl side chain of nitrogen mustard may undergo a similar reaction, leading to covalent binding with another base on the opposite strand of DNA and thus to formation of an interstrand crosslink.
Reactions leading to alkylation at the N-7 position of guanine by nitrogen mustard.
Mechlorethamine was the first anticancer drug introduced to clinical use. However, it is chemically unstable, and undergoes spontaneous hydrolysis. A large number of analogs have been synthesized and 5 of them (chlorambucil, melphalan, cyclophosphamide, ifosfamide, and bendamustine) have largely replaced mechlorethamine, which is only used clinically as part of the MOPP protocol (mechlorethamine, Oncovin [vincristine], procarbazine, prednisone) for Hodgkin disease. The addition of ring structures to the nitrogen mustard molecule confers increased stability, such that these agents can be administered orally.
Chlorambucil is a well-absorbed oral drug with a narrow spectrum of activity, and is used mainly in slowly progressive neoplasms, such as low-grade lymphomas and chronic lymphocytic leukemia. Oral melphalan is used for treatment of plasma cell myeloma and in some high-dose bone marrow transplantation protocols. Absorption of melphalan is variable and unpredictable; some patients with no effect after oral administration may respond when melphalan is given intravenously. Both chlorambucil and melphalan are almost equally toxic to cycling and noncycling cells, and may lead to delayed and/or cumulative effects on bone marrow because of their toxicity to hematopoietic stem cells.
Cyclophosphamide is the alkylating agent in widest clinical use and is part of treatment protocols for breast, lymphatic, gynecological and pediatric tumors, in high-dose chemotherapy regimens, and for a number of autoimmune diseases. Cyclophosphamide is well absorbed after oral administration; however, the parent compound is inactive, requiring metabolism by hepatic mixed-function oxidases to form the alkylating intermediate phosphoramide mustard (Fig. 18–4). Hepatic microsomal enzymes metabolize cyclophosphamide to 4-hydroxycyclophosphamide, which exists in equilibrium with its acyclic isomer aldophosphamide. 4-Hydroxycyclophosphamide enters cells and spontaneously decomposes to form phosphoramide mustard and acrolein, or it is inactivated by aldehyde dehydrogenase. Elimination of cyclophosphamide and its metabolites is mainly by renal excretion. Cyclophosphamide induces cytochrome P450 enzymes, and hence its own metabolism with repeated administration. This alters the rate but not the absolute amount of phosphoramide mustard formation, so no dose adjustment is required.
The dose of cyclophosphamide given to patients ranges from 100 to 200 mg/m2 per day given orally, to 600 to 1000 mg/m2 given intravenously every 3 to 4 weeks. Very high doses are used in preparation for bone marrow transplantation. The dose in this setting is limited by irreversible damage to the heart, which occurs with single doses greater than 60 mg/kg (approximately 2500 mg/m2). The usual dose-limiting toxicity is myelosuppression, and cyclophosphamide causes a fall in granulocyte count with rapid recovery by 3 to 4 weeks after administration (see Chap. 17, Sec. 17.5.2). There is relative sparing of stem cells and platelets, which may be a result of the higher concentrations of aldehyde dehydrogenase in early progenitor cells. Cyclophosphamide causes hemorrhagic cystitis with chronic use or at higher doses because of the direct irritative effect of acrolein on the bladder mucosa.
Ifosfamide is an analog of cyclophosphamide that differs in the presence of only 1 chloroethyl group on the oxazaphosphorine ring. It is used in the treatment of testicular cancer and sarcoma. Hemorrhagic cystitis is more common as a result of increased production of acrolein, such that all patients receiving ifosfamide should be given a sulfhydryl-containing compound, such as 2-mercaptoethane sulfonate (Mesna), which conjugates with acrolein in the urinary tract and renders it inactive. As Mesna dimerizes to an inactive metabolite in blood, and is hydrolyzed to its active form only in urine, it does not affect the cytotoxicity of cyclophosphamide or ifosfamide at other sites. Neurotoxicity, manifesting as changes in mental status including confusion, hallucination, cerebellar dysfunction, seizures, and coma, may occur with higher doses of ifosfamide, but not cyclophosphamide. Risks for ifosfamide neurotoxicity include low serum albumin, elevated creatinine and prior cisplatin treatment (Pratt et al, 1990; David and Picus, 2005).
Bendamustine is the newest bifunctional mechlorethamine derivative introduced into clinical practice. It is indicated for patients with chronic lymphocytic leukemia and indolent non-Hodgkin lymphoma.
The chloroethylnitrosoureas, BCNU (carmustine) and CCNU (lomustine), are lipid-soluble drugs that can penetrate into the central nervous system (CNS). These drugs have limited clinical application because they cause prolonged myelosuppression and are leukemogenic, likely because of direct effects on bone marrow stem cells. Streptozotocin, a methylnitrosourea, is used mainly in the treatment of pancreatic islet-cell tumors; it has less hematological toxicity than other nitrosoureas.
Busulfan, an alkyl alkane sulfonate, has a different mechanism of alkylation from the nitrogen mustards. It reacts more extensively with thiol groups of amino acids and proteins, but its ability to crosslink DNA is uncertain. Busulfan has selective effects on hematopoietic stem cells, and is now used mainly in high-dose bone marrow transplantation regimens. Busulfan is eliminated via hepatic metabolism, and the higher doses of busulfan used in marrow transplantation may cause hepatic venoocclusive disease in patients who metabolize the drug slowly.
A group of other compounds, with diverse chemical structures, that are also capable of forming covalent crosslink with intracellular macromolecules includes procarbazine, dacarbazine (DTIC), and temozolomide. Procarbazine is a synthetic derivative of hydrazine that was used in combination to treat lymphoma, including Hodgkin disease. It undergoes extensive metabolism to produce alkylating species, although details of its metabolism and mechanism of action remain unclear. It has largely been replaced by other alkylating agents. DTIC was synthesized originally as an antimetabolite to inhibit purine synthesis, but is believed to function through formation of methylcarbonium ions with alkylating properties. DTIC is metabolized by CYP450 enzymes to MTIC ([methyl-triazene-1-yl]-imidazole-4-carboxamide) which alkylates DNA at the O6 and N7 guanine positions (Marchesi et al, 2007). DTIC is sensitive to light, but is stable in neutral solutions away from light. Temozolomide is a prodrug that is stable under acidic conditions but undergoes rapid, spontaneous, nonenzymatic conversion to MTIC at pH levels greater than 7 (Marchesi et al, 2007); it is rapidly and completely absorbed after oral administration with a t1/2 of 1 to 2 hours. Temozolomide is indicated for patients with newly diagnosed or recurrent malignant glioma, and has also shown activity in melanoma.
18.2.2 Platinating Agents
The prototype agent is cisplatin (cis-diamminedichloroplatinum II; Fig. 18–5), a drug whose discovery followed an observation that an electric current delivered to bacterial culture via platinum electrodes led to inhibition of bacterial growth. The active compound was identified as cisplatin, and it was shown subsequently to exert broad cytotoxic activity. Cisplatin is one of the most useful anticancer agents and is part of first-line therapy for testicular, urothelial, lung, gynecological, and other cancers. It is also associated with substantial toxicity, which limits both the number of patients who are able to receive the drug, as well as the cumulative dose that can be given. There has been a major effort to identify other platinum analogs, either to reduce the toxicity while maintaining efficacy, or to expand the use of these compounds to tumors resistant to cisplatin. The 2 analogs in routine clinical usage are carboplatin and oxaliplatin (Fig. 18–5).
Structure of platinum agents.
Platinum drugs can exist in a 2+ (II) or 4+ (IV) oxidation state, with 4 or 6 bonds linking the platinum atom, respectively. All currently used platinum drugs are platinum II compounds that exhibit a planar structure and have 4 attached chemical groups. The nature of these groups dictates the efficacy and pharmacokinetic properties of these compounds. Two of the groups are considered carrier groups, and are chemically inert, whereas the 2 leaving groups are available for substitution and reaction with molecules such as DNA.
Cisplatin acts by a mechanism that is similar to that of classical alkylating agents. The chlorine atoms are leaving groups that may be compared to those of nitrogen mustards; these atoms may be displaced directly by nucleophilic groups in DNA or indirectly after chloride ions are replaced by hydroxyl groups through reaction of the drug with water. These reactions occur more readily in environments where the chloride concentration is low, such as within the cell. The preferred sites for binding of cisplatin to DNA are the 7 positions of the guanine and adenine bases. The fact that structurally similar analogs, such as transplatin, will produce DNA binding but are devoid of cytotoxicity suggests that the stereochemistry of the compound is critical. Cisplatin binds to 2 sites on DNA and 95% of the binding produces intrastrand crosslinkages, usually between 2 adjacent guanine bases or adjacent guanine and adenine sites, with the remainder being interstrand guanine crosslinkages. The binding of platinum compounds to DNA is responsible for their cytotoxicity, although the mechanism whereby this leads to cell death is not clear.
Carboplatin is an analog of cisplatin with substitution of cyclobutanedicarboxylate for the chloride leaving groups. This leads to a less-reactive compound that also has less toxicity but either comparable or slightly reduced efficacy to cisplatin. Oxaliplatin is one of a series of analogs with a substitution of a diaminocyclohexane (DACH) for the amine carrier groups (see Fig. 18–5). The DACH analogs have a different efficacy profile from cisplatin and have shown activity in tumors resistant to cisplatin, such as colorectal cancer. Carboplatin and oxaliplatin produce the same types of DNA adducts as cisplatin, although a higher concentration of carboplatin is required to produce a comparable number of adducts to cisplatin. Adducts formed by oxaliplatin are more likely to cause cell death, probably because of the different 3-dimensional structure that results from the DACH groups.
The pharmacokinetic differences between cisplatin and its analogs are a result of the differences in the leaving groups. Cisplatin is the most reactive and, following administration, it is rapidly and irreversibly bound to plasma proteins, with greater than 90% of free cisplatin lost in the first 2 hours. Total cisplatin (free and bound) disappears more slowly from plasma, with a prolonged t1/2 of 2 to 3 days. Cisplatin is excreted mainly via the urine, and 15% to 30% of the administered dose is excreted during the first 24 hours. Carboplatin is more stable in plasma and is excreted primarily unchanged by the kidney, with 90% of administered dose excreted within 24 hours. The clearance of carboplatin is predicated by creatinine clearance; therefore, it is possible to determine the dose of carboplatin based on the desired carboplatin AUC and creatinine clearance (Calvert et al, 1989):
Similar to cisplatin, oxaliplatin also binds to plasma proteins, although at a somewhat slower rate; it is also excreted primarily by the kidney.
Cisplatin causes little toxicity to bone marrow as a single agent, but can add to the toxic effects of other drugs, and may lead to anemia. Its major dose-limiting toxicities are nausea and vomiting, and damage to the kidney, to nerves, and to the ear with resulting loss of hearing. Vigorous intravenous hydration and maintaining a rapid urine output during and after drug administration minimize the effects on the kidneys; there is no known method for reducing the auditory or neurotoxicity.
Carboplatin has comparable activity to cisplatin against ovarian and lung tumors but is less active against urothelial and testicular cancers. Carboplatin has a better overall toxicity profile, which may make it preferable in palliative treatment regimens. There is minimal nephrotoxicity, and the drug causes less nausea and vomiting than cisplatin, but bone marrow suppression, particularly thrombocytopenia, is the dose-limiting toxicity. Carboplatin is used in some high-dose regimens prior to stem cell transplantation because its toxicities other than myelosuppression are relatively mild.
Oxaliplatin is used mainly in the treatment of colorectal cancer; it has minimal renal toxicity, and causes less vomiting than cisplatin and no ototoxicity. Oxaliplatin causes a unique spectrum of sensory neurotoxicity, ranging from acute neuropathy (paresthesia, muscle spasm and fasciculations or muscle twitching) immediately after infusion with marked sensitivity to the cold, particularly in the oropharynx, to a late cumulative dose-limiting sensory neuropathy resulting in sensory ataxia and functional impairment. Coadministration of calcium and magnesium may delay the onset of neuropathy and reduce its severity (Gamelin et al, 2008).
Resistance to platinating agents can be a result of reduced platinum-DNA adduct formation, or of increased repair or tolerance of the platinum-DNA adduct. Decreased uptake or increased binding to intracellular scavengers can result in reduced platinum-DNA adduct formation (Koberle et al, 2010).