NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Kufe DW, Pollock RE, Weichselbaum RR, et al., editors. Holland-Frei Cancer Medicine. 6th edition. Hamilton (ON): BC Decker; 2003.

  • By agreement with the publisher, this book is accessible by the search feature, but cannot be browsed.
Cover of Holland-Frei Cancer Medicine

Holland-Frei Cancer Medicine. 6th edition.

Show details

Alkylating Agents

, MD.

The alkylating agents were the first nonhormonal drugs to be used effectively in the treatment of cancer, and the story behind the recognition of the antitumor effects of these compounds is a remarkable one. During World War I, toxic gases were used as military weapons. The most devastating of these gases was sulfur mustard (Figure 51-1). The compound was used as a weapon because of its vesicant effects, which produce skin irritation, blindness, and pulmonary damage. However, it was observed that troops and civilians who were exposed to sulfur mustard also developed bone marrow suppression and lymphoid aplasia. Because of these findings, sulfur mustard was evaluated as an antitumor agent.1 The closely related, but less toxic, nitrogen mustards of World War II vintage were selected for further study. Trials in patients with lymphoma demonstrated regression of tumors, with relief of symptoms.2–4 These results encouraged the search for nitrogen mustards that were more effective and less toxic and stimulated efforts to find other chemicals with antitumor activity.

Figure 51-1. Structure of sulfur mustard (bis[2-chloroethyl]sulfide).

Figure 51-1

Structure of sulfur mustard (bis[2-chloroethyl]sulfide).

Chemistry of the Alkylating Agents

The alkylating agents are compounds that react with electron-rich atoms in biologic molecules to form covalent bonds. Traditionally, these agents are divided into two types: those that react directly with biologic molecules and those that form a reactive intermediate, which then reacts with the biologic molecules. These types are termed SN1 and SN2, respectively, and are illustrated in Figure 51-2. The terms refer to the kinetics of the reactions; the rate of reaction of an SN1 agent is dependent only on the concentration of the reactive intermediate, whereas the rate of reaction of an SN2 agent is dependent on the concentration of the alkylating agent and of the molecule with which it is reacting. This distinction has important implications in understanding the cellular and molecular pharmacology of specific alkylating agents. The nitrogen mustards and nitrosoureas are examples of SN1 agents, whereas busulfan is an SN2 agent.

Figure 51-2. SN1 and SN2 reactions of alkylating agents.

Figure 51-2

SN1 and SN2 reactions of alkylating agents.

A large number of chemical compounds are alkylating agents under physiologic conditions, and a variety of such compounds have antitumor activity. Although it is not possible to describe all of the compounds that have been used clinically, those compounds that are currently used, look promising in clinical trials, or represent a type of alkylating agent are discussed.

Types of Alkylating Agents

Nitrogen Mustards

The most frequently used alkylating agents are the nitrogen mustards. Although thousands of nitrogen mustards have been synthesized and tested, only five are commonly used in cancer therapy today. These are mechlorethamine (the original “nitrogen mustard”), cyclophosphamide, ifosfamide, melphalan, and chlorambucil, and they are illustrated in Figure 51-3. The characteristic chemical constituent of the nitrogen mustards is the bischloroethyl group, and all of the nitrogen mustards react through an aziridinium intermediate as shown in Figure 51-4. The remainder of the molecule is important in determining the physical properties of the molecule and affects the transport, distribution, and reactivity of the specific agents. The importance of the total molecule is demonstrated by cyclophosphamide.

Figure 51-3. Structures of nitrogen mustards currently used in therapy.

Figure 51-3

Structures of nitrogen mustards currently used in therapy.

Figure 51-4. Alkylation mechanism of nitrogen mustards.

Figure 51-4

Alkylation mechanism of nitrogen mustards.

Cyclophosphamide is not a reactive compound, but undergoes activation in the body. Figure 51-5 illustrates the complex activation scheme.5 The initial activation reaction is carried out by cytochrome P450-mediated microsomal oxidation in the liver to produce 4-hydroxycyclophosphamide, which is in spontaneous equilibrium with the tautomer aldophosphamide.6 At physiologic pH, this equilibrium is predominantly in the form of 4- hydroxycyclophosphamide.7 This equilibrium mixture diffuses from the hepatocyte into the plasma and is distributed throughout the body. Because 4-hydroxycyclophosphamide is relatively nonpolar, it enters target cells readily by diffusion. Aldophosphamide spontaneously decomposes to produce phosphoramide mustard, which is the first reactive alkylating agent produced in the metabolism of cyclophosphamide. Although phosphoramide mustard is also produced extracellularly, this compound is very polar, and enters cells poorly, and phosphoramide mustard in the plasma probably plays a minor role in the therapeutic and toxic effects of cyclophosphamide. Thus, 4-hydroxycyclophosphamide/ aldophosphamide serves as an efficient mechanism to deliver the alkylating phosphoramide mustard into cells. Recent evidence suggests that after one of the chloroethyl groups of phosphoramide mustard cyclizes to form a chloroethyl aziridinium moiety, the molecule cleaves to produce chloroethylaziridine.8 Accordingly, free chloroethylaziridine may contribute to the alkylation and cross-linking of DNA by cyclophosphamide. It has been demonstrated that chloroethylaziridine can diffuse between wells in in vitro studies with 4-hydroxycyclophosphamide and produce cytotoxicity in “control wells”8a in such experiments.

Figure 51-5. Metabolism of cyclophosphamide.

Figure 51-5

Metabolism of cyclophosphamide.

The toxic compound acrolein was demonstrated to be produced by the metabolism of cyclophosphamide by Alarconand Meientrofer,9 but administration of didechlorocyclophosphamide, a compound that could produce acrolein, but not the chloroethyl alkylating species, did not demonstrate antitumor activity in an animal model.10 In 1992, Lee and colleagues11 reported that a decrease in the enzyme O6-alkylguanine-alkyltransferase in circulating lymphocytes was produced by the administration of high doses of cyclophosphamide for bone marrow transplantation. Recently, Friedman and colleagues reported that tumor cells with elevated 06-alkylguanine-alkyltransferase were sensitized to 4-hydroperoxycyclophosphamide by depletion of the enzyme.12 These and further studies have indicated that acrolein released by cyclophosphamide forms an O6-guanyl adduct that can be removed by O6-alkylguanine-alkyltransferase. Thus, acrolein contributes to the antitumor activity and probably the carcinogenic effects of cyclophosphamide, and these effects are abrogated by the action of O6-alkylguanine-alkyltransferase.

Cyclophosphamide produces less gastrointestinal and hematopoietic toxicity than other alkylating agents. The basis for this decreased toxicity is the enzyme aldehyde dehydrogenase. This enzyme oxidizes aldophosphamide to carboxyphosphamide, an inactive product, which is excreted in the urine and accounts for approximately 80% of an administered dose of cyclophosphamide in any species. This enzyme is found in high concentration in the hepatic cytosol, in primitive hematopoietic cells, and in the stem cells and mucosal absorptive cells in the intestine.13 Administration of an inhibitor of this enzyme to an animal markedly increases the hematopoietic and gastrointestinal toxicity of cyclophosphamide.13

Ifosfamide is a structural isomer of cyclophosphamide that is used particularly in the treatment of testicular tumors and sarcomas.14–16 Ifosfamide undergoes the same metabolic reactions as cyclophosphamide, but the location of the chloroethyl group on the ring nitrogen produces quantitative changes in the metabolism of the drug17,18 and subtle changes in the chemical properties of the reactive metabolite, ifosfamide mustard, so that it is less reactive than phosphoramide mustard.19 The primary metabolite, aldoifosfamide, is a substrate for aldehyde dehydrogenase, so that the bone marrow and gastrointestinal tract sparing properties are similar to those of cyclophosphamide. The oxidation of the chloroethyl side chains to produce chloroacetaldehyde is a minor metabolic pathway for cyclophosphamide (< 10% of dose) but is increased to as much as 50% for ifosfamide. The increased production of chloroacetaldehyde has been implicated in the neurotoxicity of ifosfamide,20 and may also contribute to the greater renal and bladder toxicity of ifosfamide. The greater side chain oxidation of ifosfamide and the lesser reactivity of the ifosfamide mustard are consistent with the fact that higher doses of ifosfamide than cyclophosphamide are used clinically.

Melphalan is an alkylating agent that is used in the treatment of multiple myeloma,21,22 ovarian cancer,23,24 and breast cancer.25,26 Melphalan is an amino acid analog that enters cells and crosses the blood–brain barrier through active transport systems. The natural substrates for these systems are amino acids,27,28 and the entry of melphalan into cells29 and the central nervous system (CNS)30 can be modulated by the presence of certain amino acids in the extracellular fluid.31

Chlorambucil is used for the treatment of chronic lymphocytic leukemia,32,33 ovarian carcinoma,34,35 and lymphoma,36,37 but has been used less often in high-dose combination therapies than the other nitrogen mustards that are described here. This agent is well tolerated by most patients and can be used in patients who have severe nausea and vomiting with cyclophosphamide or melphalan.

Aziridines and Epoxides

Closely related to the nitrogen mustards are the aziridines, which are represented in current therapy by thiotepa, mitomycin C, and diaziquone (AZQ), and are illustrated in Figure 51-6. These agents presumably alkylate by the same mechanism as the aziridinium intermediates produced by the nitrogen mustards, but the aziridine rings in these compounds are uncharged and less reactive than aziridinium compounds.

Figure 51-6. Structures of aziridine alkylating agents.

Figure 51-6

Structures of aziridine alkylating agents.

Thiotepa (triethylene thiophosphoramide) has been used particularly in the treatment of carcinomas of the ovary and breast and for the intrathecal therapy of meningeal carcinomatosis.38–40 Thiotepa is oxidatively desulfurated by hepatic microsomes to produce triethylenethiophosphoramide (TEPA).41 Although TEPA is cytotoxic, it is less so than thiotepa.42 After the clinical administration of thiotepa, both thiotepa and TEPA are found in the blood,43,44 and the concentration and area under the curve (AUC) exposure to TEPA may exceed those of thiotepa.45 The AUC exposure to thiotepa correlates with the degree of myelosuppression in patients, whereas the AUC exposure to TEPA does not.45 Some studies suggest that a metabolite is produced that is more reactive than the parent compound.46,47 However, such a metabolite has not been characterized, and the activity of thiotepa may be enhanced by low pH within tumor cells. At the lower pH, the aziridine ring will be protonated and more reactive.

Mitomycin C is a natural product that is used in the treatment of breast cancer and cancers of the gastrointestinal tract.48–50 This compound contains an aziridine ring and appears to exert its cytotoxic effect through the cross-linking of DNA.51,52 Mitomycin C undergoes reduction in cells, with enhancement of the affinity of the carbon-1 atom of the aziridine ring for nucleophiles, such as the extracyclic nitrogen atom on guanylic acid in DNA. Following this alkylation, there is displacement of the activated carbamate group on the 10-carbon atom of mitomycin C by an extracyclic amino nitrogen of a guanylic acid molecule on the complementary DNA strand to produce an interstrand DNA cross-link.53–55

AZQ was designed to be sufficiently lipophilic to readily cross the blood-brain barrier for the treatment of CNS tumors.56 It has demonstrated clinical activity against brain tumors,57 other solid tumors, and leukemia.58 AZQ undergoes reduction of the quinone ring in cells. This reduction results in protonation of the aziridine rings and enhancement of reactivity of the compound.59,60

The epoxides, such as dianhydrogalactitol61,62 (Figure 51-7), are chemically related to the aziridines and alkylate through a similar mechanism of attack of a nucleophile, such as an amino nitrogen, on a carbon of a strained three-member ring. Dibromodulcitol63 is hydrolyzed to dianhydrogalactitol and thus is a pro-drug to an epoxide.64 Dianhydrogalactitol and dibromodulcitol have been mainly used in Europe, and are not commonly used in the United States.

Figure 51-7. Structures of an epoxide alkylating agent (dianhydrogalactitol) and an epoxide prodrug (dibromodulcitol).

Figure 51-7

Structures of an epoxide alkylating agent (dianhydrogalactitol) and an epoxide prodrug (dibromodulcitol).

Alkyl Sulfonates

The alkyl alkane sulfonate busulfan (Figure 51-8) was one of the earliest alkylating agents.65 This compound is one of the few currently used agents that clearly alkylate through an SN2 reaction, as shown in Figure 51-9. Hepsulfam, an alkyl sulfamate analog of busulfan with a wider range of antitumor activity in preclinical studies,66 was evaluated in clinical trials but demonstrated no superiority to busulfan. Busulfan has a most interesting, but poorly understood, selective toxicity for early myeloid precursors.67,68 This selective effect is probably responsible for its activity against chronic myelocytic leukemia (CML).69,70

Figure 51-8. Structure of alkyl sulfonate (busulfan) and alkyl sulfamate (hepsulfam) agents.

Figure 51-8

Structure of alkyl sulfonate (busulfan) and alkyl sulfamate (hepsulfam) agents.

Figure 51-9. Mechanism of alkylation by busulfan.

Figure 51-9

Mechanism of alkylation by busulfan.

The use of busulfan as first-line therapy for the treatment of CML has been succeeded by the use of the less toxic hydroxyurea,70a and more recently by Gleevec,70b a specific inhibitor of the bcr-abl oncogene found in CML. The current major use of busulfan is as a component of bone marrow ablative regimens for bone marrow and stem cell transplantation of patients with acute myeloid leukemia and other malignancies.71,72


The nitrosoureas are a class of alkylating agents that received considerable attention during the past three decades.73–75Figure 51-10 shows several nitrosoureas currently in clinical use or clinical trials. These compounds decompose to produce alkylating compounds under physiologic conditions. Although there are several mechanisms by which this may occur, the predominant mechanism is that shown in Figure 51-11, a base catalyzed decomposition to a chloroethyl diazonium moiety,76 which reacts with DNA,77,78 as discussed in the section Mechanism of Cytotoxicity, to form a unique interstrand DNA cross-link.

Figure 51-10. Structures of nitrosoureas.

Figure 51-10

Structures of nitrosoureas.

Figure 51-11. Mechanism of nitrosourea activation and alkylation of deoxyguanylic acid.

Figure 51-11

Mechanism of nitrosourea activation and alkylation of deoxyguanylic acid.

Carmustine (BCNU) was the first agent to demonstrate significant activity against a preclinical model of intracerebral tumor74 and is currently used for the treatment of primary brain tumors79 and in the treatment of multiple myeloma.80 Lomustine (CCNU) and semustine (methyl CCNU) demonstrated greater activity against solid tumors in preclinical studies.81 CCNU is used in the treatment of CNS tumors82,83 and lymphomas,84,85 and methyl CCNU has been used particularly in the treatment of gastrointestinal tumors.86,87 Nimustine (ACNU), which is more water soluble than most of the other nitrosoureas, has been employed for the intraarterial and intrathecal treatment of CNS tumors88,89 and the treatment of solid tumors.90 The clinical use of the nitrosoureas has been limited by marked and prolonged hematopoietic toxicity and by renal toxicity. The development of nitrosoureas with a higher therapeutic index, such as fotemustine91,92 and others,93,94 remains an active area of endeavor.

Triazenes, Hydrazines, and Related Compounds

These are nitrogen-containing compounds that spontaneously decompose or can be metabolized to produce alkyl diazonium intermediates that alkylate biologic molecules. Procarbazine and dacarbazine, which are illustrated in Figure 51-12, are metabolized to reactive intermediates that decompose to produce methyl diazonium, which methylates DNA.95 The metabolism of procarbazine is complex, and there are different pathways through which a reactive methyl group can be produced.95 It is most likely that the pathway responsible for the DNA methylation and cytotoxicity is the generation of methylazoxyprocarbazine.96,97Figure 51-13 illustrates activation of dacarbazine via N-methyl oxidation by a microsomal P450 enzyme.98,99 Both procarbazine and dacarbazine are used in the treatment of Hodgkin disease100,101; procarbazine is a component of combination regimens used for the treatment of primary brain tumors,102 and dacarbazine is used in the treatment of melanoma.103,104 Procarbazine was originally developed as a monoamine oxidase inhibitor, and it can produce CNS depression and acute hypertensive reactions after the ingestion of tyramine-rich foods.105

Figure 51-12. Structures of monofunctional alkylating agents.

Figure 51-12

Structures of monofunctional alkylating agents.

Figure 51-13. Interstrand cross-linking of DNA by nitrogen mustards.

Figure 51-13

Interstrand cross-linking of DNA by nitrogen mustards. (1) Site of cross-linking proposed by Brookes. (2) Site of cross-linking found by Loechler and Hopkins.

Temozolomide (see Figure 51-13) spontaneously decomposes under physiologic conditions to produce the same active metabolite produced by dacarbazine (DTIC).106 Temozolomide, which is administered orally, has demonstrated antitumor activity against gliomas and melanomas in Phase I and II trials,107–109 and is now approved for glioma treatment in the United States and Europe. Temozolomide appears at least as effective, and more predictable in its effects than DTIC, because temozolomide does not require metabolic activation by drug-metabolizing enzymes.


Hexamethylmelamine (Figure 51-14) is an active antitumor agent that is considered to be acting as an alkylating agent because the methyl groups are required for antitumor activity. The methyl groups are hydroxylated with subsequent demethylation in vivo,110,111 a reaction that can generate a reactive methyl group. Analogs in which the methyl groups are hydroxylated are also active.112,113 Few studies of the cross-resistance of this agent have been carried out, but one study114 found that O6-alkylguaninealkyltransferase was not inactivated in vivo by hexamethylmelamine, as would be expected from an O6-guanyl-methylating agent. Therefore, the mechanism of cytotoxic activity of hexamethylmelamine remains in question. The agent does have significant antitumor activity against ovarian cancer115 and is used primarily in the third-line treatment of that tumor.

Figure 51-14. Mechanisms of resistance to alkylating agents.

Figure 51-14

Mechanisms of resistance to alkylating agents.

Decomposition and Metabolism

The alkylating agents react with water and are inactivated by this hydrolysis. The alkylating agents also are inactivated by reaction with thiols, such as glutathione. The reaction of alkylating agents with glutathione can be increased by the glutathione S-transferase enzymes, as is discussed later in the sections on mechanisms of cellular resistance. The alkylating agents also undergo microsomal and other types of xenobiotic metabolism. Such metabolism may activate agents, as described above, inactivate them, or change their physical properties without inactivating them. Nitrosoureas have been reported to be denitrosated and inactivated by microsomal metabolism.116,117

Chlorambucil is metabolized to bischoroethylphenylacetic acid, which is an active alkylating agent, and probably contributes to the therapeutic and toxic effects of chlorambucil.118,119 Mitomycin C must be reductively activated intracellularly to alkylate DNA bases and cross-link the DNA, and glutathione appears to play a role in this process.120

Mechanism of Cytotoxicity

Although the alkylating agents react with a number of biologic molecules, including amino acids, thiols, ribonucleic acid (RNA), and DNA, a number of lines of evidence have led to the generally accepted conclusion that the cytotoxic effects of the agents are a result of reactions with DNA. Bifunctional agents are more effective antitumor agents than monofunctional agents, but addition of more than two alkylating groups does not further increase the cytotoxic activity. These observations121 and the early studies of Brookes and Lawley122,123 led to the suggestion that interstrand cross-linking of DNA was responsible for the cytotoxic activity of the bifunctional alkylating agents. There is a good correlation between cytotoxicity and the formation of interstrand cross-links by bifunctional alkylating agents. Recently, nitrogen mustard interstrand cross-links in oligonucleotides were chemically characterized125–127 in order to understand their mechanisms of cytotoxicity and to develop approaches to reduce resistance produced by cellular repair of these lesions.

Although the alkylating agents can react with virtually all of the nitrogens in the DNA bases, there is selectivity, based on the electron density of the nitrogens and the local structure of the DNA. The nitrogen mustards react most readily with the N-7 position of guanylic acid.128 This nitrogen atom has a high electron density, which has been proposed to be enhanced by base stacking in the DNA helical structure.122 Brookes and Lawley suggested that the nitrogen mustard cross-link in DNA was between the N-7 guanine atoms in base-paired G-C sequences in DNA.122 However, more recent studies found the nitrogen mustard cross-link to occur between the N-7 atoms of guanylic acids in a G-X-C sequence, as illustrated in Figure 51-15.125–127 The crosslinking of mitomycin C between two extracyclic guanylic acid amino groups is described in the section Aziridines and Epoxides.53 This site of cross-linking may be determined by the orientation of mitomycin C in the minor groove of DNA.54 The reactive species of the nitrosoureas is more reactive than the aziridiniums of the nitrogen mustards and initially alkylates the O-6 position of guanylic acid.129,130 According to a mechanism proposed by Ludlum, after a series of rearrangements involving a reactive cyclic five-membered intermediate of the N-1, C-6, and O-6 atoms of guanylic acid and two carbons from the chloroethyl group of the nitrosourea (see Figure 51-11), a cross-link is formed between N-1 of guanylic acid and N-3 of a cytidylic acid on the complementary DNA strand.131,132 Such a cross-link occurs in an oligonucleotide treated with BCNU.133

Figure 51-15. Hematopoietic toxicity of alkylating agents.

Figure 51-15

Hematopoietic toxicity of alkylating agents. WBC = white blood cells.

Alkylating agents such as methylnitrosourea, procarbazine, and dacarbazine are not bifunctional and produce methylation of DNA, predominantly on the O-6 and N-7 positions of guanylic acid. These lesions can produce both spontaneous and enzyme mediated single-strand breaks,134,135 which are cytotoxic. However, it has now been demonstrated, initially by Modrich and colleagues,136,137 that active mismatch DNA repair activity is a major mediator of the cytotoxicity of monofunctional alkylating agents.

Cellular Resistance to Alkylating Agents

Cellular resistance to antitumor agents is a critical determinant of the effectiveness of therapy. Resistance mechanisms in normal tissues provide selectivity and an improved therapeutic index. Resistance of tumor cells allows these cells to escape the effects of therapy. Consideration of the pharmacology and chemistry of the alkylating agents predicts four general types of cellular resistance to alkylating agents (Figure 51-16): (1) decreased uptake of agents into or increased export out of the cell; (2) increased inactivation of agents in the cell; (3) enhanced repair of the DNA damage produced by the alkylating agents; and (4) the absence of cellular mechanisms that produce cytotoxicity in response to DNA damage. All of these mechanisms have been described. Resistance of tumor cells to mechlorethamine can occur on the basis of decreased transport into the cell,138,139 and it has also been demonstrated that certain cells resistant to melphalan have decreased active transport of the agent and of amino acid.140,141 Most alkylating agents enter cells by diffusion, however, and the alkylating agents, with the exception of mitomycin C, are not substrates for the multiple-drug-resistance export systems.

Figure 51-16. Relationship between plasma AUC of busulfan and occurrence of venoocclusive disease of the liver.

Figure 51-16

Relationship between plasma AUC of busulfan and occurrence of venoocclusive disease of the liver. Reproduced with permission from Grochow et al.

The second mechanism of cellular resistance to alkylating agents is intracellular inactivation of the agent. As discussed in the section Nitrogen Mustards, the enzyme aldehyde dehydrogenase detoxifies the primary metabolites of cyclophosphamide and ifosfamide, and the presence of this enzyme in bone marrow precursor cells and gastrointestinal epithelial cells protects these organs from toxicity of the agents. Aldehyde dehydrogenase has also been demonstrated to be a mechanism of cyclophosphamide resistance of murine,142 rat,143 and human leukemia cells,144 and of human ovarian,145 colon,146 and breast147 cancer cells.

An association between cellular resistance to alkylating agents and increased cellular levels of glutathione148–150 and the enzyme glutathione transferase has been described by a number of investigators.151–154 Glutathione (GSH) is a thiol-containing tripeptide that is present at millimolar concentrations in many cells, reacts with electrophilic (electron-deficient) molecules, and protects cells from such electrophiles.155–157 Mulcahy and colleagues158 demonstrated that increased GSH in cells resistant to melphalan can be related to increased transcription of gamma-glutamylcysteine synthetase, the enzyme that catalyzes the rate-limiting step in de novo synthesis of GSH.

Although most electrophiles of biologic significance react spontaneously with glutathione, glutathione S-transferases catalyze the reaction between glutathione and electrophiles. The glutathione conjugates of several alkylating agents have been characterized159–161 and their formation shown to be enhanced by a glutathione S-transferase.

There are three principal isozymes of glutathione S-transferase (GST), and recent studies indicate that specific isozymes may catalyze the conjugation of different alkylating agents. The alpha isozyme of GST catalyzes the glutathione conjugation of the aziridinium forms of melphalan,162 chlorambucil,163 and phosphoramide mustard.164 The GSH conjugation of 4- hydroxycyclophosphamide165 was found to be enhanced by all three classes of GST. The mu isozyme is implicated in the inactivation of BCNU.166 At this time, it is evident that glutathione alone or glutathione plus an appropriate glutathione S-transferase can render cells resistant to alkylating agents and that this mechanism is probably an important mechanism of resistance to electrophilic antitumor drugs, such as the alkylating agents and platinum compounds.

Several investigators have demonstrated that buthionine sulfoxime (BSO), an inhibitor of glutathione synthesis, can reduce cellular glutathione levels and sensitize tumors to alkylating agents in vitro and in vivo.167–169 However, normal cells can also be sensitized by BSO administration170,171 to produce significant toxicity. BSO in combination with alkylating agents has been evaluated in clinical trials. Phase I trials of the combination of BSO and melphalan have demonstrated increased myelotoxicity and depletion of tumor GSH, compared with the same dose of melphalan alone.172–174 Phase II trials are in progress to determine whether the tumor response rate is greater with the addition of BSO to melphalan.

Inhibitors of glutathione S-transferases enhance the cytotoxicity of melphalan on cells resistant to alkylating agents,175 and such inhibitors are being examined in clinical trials. A Phase I trial of the GST inhibitor sulfasalazine176 with melphalan doses of 20 mg/m2 and greater demonstrated reductions of glutathione and GST levels in the peripheral mononuclear cells of some patients, and the main toxicity of the combination was nausea and vomiting. Increased myelosuppression was not seen.

An association between increased cellular concentrations of metallothionein and resistance to platinum agents has been established,177,178 which is probably a result of binding of the platinum agents to the multiple thiol groups of this cellular protein. Lazo and colleagues179 found that transfection-induced increased cellular metallothionein also produced resistance to alkylating agents. Yu and colleagues and Wei and colleagues have demonstrated binding of melphalan180 and phosphoramide mustard181 to thiol groups in metallothionein. Thus, increased metallothionein content of cells is another mechanism of inactivation of alkylating agents.

Because the cytotoxicity of the alkylating agents appears to be mediated through the alkylation of DNA, the repair of alkylation lesions is an obvious mechanism of resistance to these agents and has been the subject of intense investigation. The best-defined DNA repair resistance to alkylating agents is resistance to the nitrosoureas and other compounds that alkylate the O-6 position of guanylic acid in DNA. The protein O6-alkylguanine-DNA-alkyltransferase (O6-AT) removes alkyl groups from the O-6 position of guanine, preventing the formation of an interstrand cross-link.129,131,182 The removed alkyl group is covalently and irreversibly bound to the alkyltransferase so that the protein can catalyze the removal of only one alkyl molecule and is then rapidly catabolized. It is now obvious that elevated O6-AT is a major mechanism of resistance to nitrosoureas in human gliomas183–185 and other human tumors.186–188

The fact that O6-AT is irreversibly inactivated by the transfer to it of an alkyl group from the O-6 position of guanine provides an approach to counteracting this mechanism of resistance. If cells are treated with a monofunctional O-6 alkylating agent, such as streptozotocin, there follows a period when the O-6 alkyltransferase activity is decreased. This decrease in activity is a result of the removal of alkyl groups from the O-6 guanine sites on the DNA and a subsequent reduction of the level of active enzyme before enzyme synthesis can restore functional levels of the enzyme. If the cells are treated with a nitrosourea (or other O-6 guanine alkylating agent) during this period of decreased O-6 alkyltransferase, the cells are more sensitive to nitrosourea.189 The enzyme will also remove O-6 benzyl groups from acid-soluble guanine analogs, and compounds such as O-6 benzylguanine, administered prior to nitrosoureas, will reverse alkyltransferase resistance in cells and animal models.190–192 Clinical trials with O-6 benzylguanine (O6-BG) are now in progress.193–195

Although O6-BG and related compounds can reverse tumor resistance to nitrosoureas and methylating agents such as temozolomide, the bone marrow toxicity of these agents is increased by O6-BG. However, human O6-AT enzymes that are resistant to the combination of an alkylating agent and O6-BG have been isolated and transfected into cell lines.196,197 In a mouse model, transfection of such an enzyme into hematopoietic cells in mice produces protection of the bone marrow from the cytotoxicity of the combination of BCNU and O6-BG196; currently, similar studies are planned in patients.

Removal of interstrand cross-links from DNA in cells can be shown to occur in studies using alkaline elution and other techniques.198 Friedman and colleagues described a human medulloblastoma cell that is resistant to cyclophosphamide on the basis of repair of the interstrand cross-links.199 This cell does not appear to repair the BCNU or busulfan cross-link, suggesting that the recognition and repair of interstrand cross-links are quite structure specific.

Recently, evidence was presented that poly(ADP ribose) polymerase is involved in the repair of nitrogen mustard lesions.200 Also, there is good evidence that cells that react to alkylation damage by arresting in the G2 phase of the cell cycle can repair DNA during this period and are more resistant to alkylating agents than cells that proceed through mitosis despite alkylation damage. A human tumor cell line has been described that exhibits G2 arrest in response to alkylating damage and demonstrates increased resistance to nitrogen mustard.201 This cell line has increased accumulation of phosphorylated (and inactivated) cdc2 kinase associated with G2 arrest after nitrogen mustard treatment. This alteration should allow repair of DNA damage before the cell enters mitosis. This mechanism of resistance to alkylating agents is probably important for tumor cells but also may provide a degree of drug specificity for many other tumors, because normal cells may be more likely to exhibit this protective mechanism. Inhibitors of DNA repair have been shown to enhance the cytotoxicity of alkylating agents,202–204 and some of these inhibitors are being examined in clinical trials. It seems likely that increased understanding of the DNA repair process will allow more effective use of alkylating agents.

In Vivo Resistance

Murine tumors that are resistant to alkylating agents in vivo, but not in vitro, have been reported.205,206 Further studies of these tumors that are resistant to cyclophosphamide, cisplatin, and thiotepa in vivo demonstrate that the tumors are also resistant to these agents in three-dimensional in vitro culture, but not in two-dimensional in vitro culture.207 Such resistance may be acquired rapidly after drug exposure208 and may be associated with enhanced metastatic properties.209 The mechanisms responsible for this type of resistance have not yet been established. There may be differences between known cellular resistance factors or between membrane properties in the three-dimensional milieu, compared with the two-dimensional configuration, and adhesion molecules may alter drug sensitivity. Other potential mechanisms for drug resistance in vivo are poor perfusion of the tumor and changes in the intracellular pH.210

Clinical Pharmacology


After the administration of a systemic dose of 50 mg/kg, plasma levels of the parent compound of up to 400 μmol/L may be achieved and decay with a half-life of 3 to 10 h.211–213 The rate of metabolism of the parent compound varies considerably among individuals and can be modulated by the administration of compounds that affect the rate of microsomal metabolism, such as phenobarbital214 or a previous dose of cyclophosphamide.215,216 However, at conventional doses, the clearance rate of the parent compound does not appear to significantly affect the toxicity or therapeutic effect of the agent.217 This independence of effect from the rate of metabolism is probably because the parent compound is not rapidly excreted and continues to be activated, so that the AUC for systemic exposure to the active metabolites is similar after a given dose.

At the higher doses currently used in bone marrow transplantation regimens, however, the plasma concentrations of cyclophosphamide should be close to the capacity of the microsomal activating enzymes. Grochow and colleagues218 demonstrated that in patients receiving 4 g/m2 of cyclophosphamide over 90 min and achieving initial plasma concentrations of greater than 500 μmol/L, saturable pharmacokinetics are seen. These investigators concluded that when the dosing rate equals or exceeds 4 g/m2 in 90 min or the plasma concentration of cyclophosphamide exceeds 150 μmol/L (the lowest Km seen in the patients), nonlinear disposition may occur, with variable exposure to the active metabolites. This study also confirmed previous reports that cyclophosphamide can induce its own metabolism.

Studies of pharmacokinetics of the critical metabolite 4-hydroxycyclophosphamide were limited in the past by the difficulty of accurately measuring this labile molecule. However, more specific methods are now available, and the pharmacokinetics of this important metabolite are now being elucidated. Anderson and colleagues measured 4-hydroxycyclophosphamide in patient blood after cyclop hosphamide administration by using a very specific gas chromatographic-mass spectrometric technique.219 After a dose of cyclophosphamide of 110 mg/kg over 90 min, peak concentrations of 9 to 12 μmol/L and AUCs of 105 to 110 μmol/L h were measured; a cyclophosphamide dose of 170 mg/kg given as a continuous infusion over 4 days produced plasma concentrations of 1 to 5 μmol/L, with a total AUC of about 98 to 110 μmol/L h. Subsequent studies found similar results.220,221 All studies found a considerable patient variation in the exposure to 4-hydroxycyclophosphamide after the same dose of cyclophosphamide and differences in the exposure and ratios of cyclophosphamide/4-hydroxycyclophosphamide each day when short-duration infusions are given on subsequent days. These findings are most likely to be a result of differences in the cytochrome P450 complements in patients and the differing exposures to drugs that modulate the activities of these enzymes. These findings indicate that pharmacokinetically guided dose adjustment will be the best method to produce consistent patient exposures to the active metabolites of cyclophosphamide.

The majority of a dose of cyclophosphamide (< 70%) is excreted in the urine as the inactive metabolite carboxyphosphamide.222,223 Renal function does not significantly affect the toxicity of cyclophosphamide,224 most likely because spontaneous decomposition, and not renal excretion, determines the clearance of the principal active metabolites.

The clinical pharmacology of ifosfamide has been much less studied but is similar to that of cyclophosphamide, except that microsomal activation is somewhat slower, and chloroethyl side chain oxidation plays a greater role in its metabolism.225–227 Thus, for a dose of ifosfamide, lower systemic concentrations of the 4-hydroxy metabolite are achieved than for the same dose of cyclophosphamide.228 Both cyclophosphamide and ifosfamide are well absorbed after oral administration.229 Boddy and colleagues17 demonstrated that ifosfamide, like cyclophosphamide, can autoinduce its own metabolism. Because of the greater and more variable side chain oxidation of ifosfamide, differences in the P450 drug metabolizing enzymes between individuals, and the modulation of these enzymes by concomitantly administered agents, may play a greater role in altering the clinical effects of ifosfamide than cyclophosphamide.230–232


Alberts and colleagues found that peak plasma levels of 4 to 13 μmol/L were present after intravenous administration of a 0.6 mg/kg dose of melphalan, and the half-life (t½b) was 1.8 h.233,234 At this dose, the mean AUC for melphalan was 8 μmol/L h. Similar AUC per dose and pharmacokinetics have been demonstrated by other investigators after high intravenous doses of melphalan.235 After conventional oral doses of 0.25 mg/kg, peak plasma levels of up to 0.625 μmol/L were found.236 There is variable systemic availability after oral dosing,237,238 and it has been shown that oral administration of food with melphalan will inhibit absorption of the agent.239 It has been reported that myelosuppression from melphalan is increased in patients with decreased renal function.240 The half-life of melphalan is prolonged in anephric dogs,241 and significant renal clearance of the parent compound in patients has been shown by Reece and colleagues.242


After the oral administration of 0.6 mg/kg of chlorambucil, peak levels of 2 to 6 μmol/L of parent compound were found at 1 h by Alberts and colleagues.118,234 Peak plasma levels of phenylacetic acid mustard of 2 to 4 μmol/L occurred at 2 to 4 h after chlorambucil administration. The plasma half-life (t½b) of chlorambucil was 92 min and that of phenylacetic acid mustard was 145 min. At a dose of 0.6 mg/kg of chlorambucil, the plasma AUC of chlorambucil was 3 to 9 μmol/L h.118 Similar values were found by Hartvig and colleagues,243 who also found a two- to fourfold variation in systemic availability of chlorambucil and phenylacetic acid mustard after oral administration of chlorambucil.


The pharmacokinetics of thiotepa were studied by Cohen and colleagues244 after an intravenous injection of 12 mg/m2. Peak plasma levels of about 5 μmol/L were achieved and were found to decay with a t1/2 a of 7.7 min and a t1/2b of 125 min. The mean AUC was 9 μmol h. Plasma concentrations of TEPA of up to 1 μmol/L were found and remained in plasma longer than thiotepa. Henner and colleagues245 examined the plasma levels of thiotepa after 4-day continuous intravenous infusions of up to 900 mg/m2. Peak plasma levels of thiotepa of 7 μmol/L were initially attained on the first day, and then the levels gradually decreased. Plasma AUC values of up to 600 μmol/L h were achieved. When given intraperitoneally, there is rapid loss of thiotepa from the intraperitoneal cavity and a concomitant increase in plasma levels to those associated with the same dose if given intravenously.246 After intravenous injection, cerebrospinal fluid levels comparable with plasma levels are found.247 Recent studies indicate that the simultaneous administration of thiotepa and cyclophosphamide will result in lower exposure to the active metabolite of cyclophosphamide, 4-hydroxycyclophosphamide.218


Levin and colleagues studied the pharmacokinetics of BCNU.248 After intravenous infusion of 60 to 170 mg/m2, peak plasma concentrations of 5 μmol/L were reached and then decayed with an initial half-life of 6 min and a second half-life of 68 min. Henner and colleagues249 measured the pharmacokinetics of BCNU after intravenous doses of 600 mg/m2. The peak plasma level of ultrafilterable BCNU was found to be 4.7 μmol/L and the mean AUC was 5.4 μmol/L h. The ultrafilterable BCNU was 23% of the total plasma BCNU. The pharmacokinetics of CCNU after administration of 130 mg/m2 to patients have also been described.250 The parent compound could not be detected in plasma, but the monohydroxylated metabolites, trans-4-hydroxy CCNU and cis-4-hydroxy CCNU, were found in a ratio of 6:4 and at total peak concentrations of about 3 μmol/L. The plasma clearance half-lives of the hydroxy-CCNU metabolites varied from 1 to 3 h between patients.


Because of its insolubility in aqueous solutions, busulfan was previously available only as an oral preparation. However, an intravenous preparation was recently made available,251 but has not been extensively used, so that most of the published pharmacokinetic information is after oral administration. For myeloablative therapy prior to bone marrow transplantation, busulfan is widely used at a dose of 1 mg/kg every 6 h for 4 days. After a 1 mg/kg dose in adults and older children, there is a considerable variation in bioavailability, with peak plasma levels of 1 to 10 μmol/L and elimination half-times between 1 and 7 h.252–254 The AUC after a single dose in adults and older children varies between 10 and 80 μmol/L h. However, in young children (ages 1 to 3 years), the peak plasma concentrations are less (1 to 5 μmol/L), the mean elimination time is approximately 40% faster, and the AUC is consistently less at 6 to 17 μmol/L h.255 Grochow and colleagues252 demonstrated that AUCs of busulfan greater than one standard deviation from the mean values for all patients are associated with a very high risk of venoocclusive disease of the liver (Figure 51-16). It has now been demonstrated that pharmacokinetic-guided adjustment of the busulfan dose can reduce the incidence and severity of this toxicity.256,257


The characteristic toxicities of the alkylating agents are hematopoietic, gastrointestinal, gonadal, and CNS toxicity. However, each of the agents has a characteristic set of toxicities, determined by the reactivity, metabolism, and distribution of the agent, and the clinician should be aware of these idiosyncrasies of the agents.

Hematopoietic Toxicity

In general, the clinical dose-limiting toxicity for alkylating agents is hematopoietic toxicity, particularly suppression of granulocytes and platelets. The nadir of granulocyte depression after alkylating agents is usually 8 to 16 days, and the granulocytes usually return to normal within 20 days after a single dose of the agent.258 Cyclophosphamide and ifosfamide are less hematopoietically toxic than other alkylating agents258,259; granulocyte levels return to normal more rapidly, platelets are affected less, and repeated doses of cyclophosphamide and ifosfamide do not produce cumulative damage and progressive deterioration of the hematopoietic elements. The reduced hematopoietic toxicity of cyclophosphamide and ifosfamide is a result of the presence of aldehyde dehydrogenase in the hematopoietic stem cells and the early megakaryocytes, as discussed earlier. In contrast, the nitrosoureas produce severe hematopoietic toxicity, with a delayed onset and nadirs of granulocytes and platelets occurring as late as 45 days.75,260 Busulfan also produces severe hematopoietic depression, with a selectivity for early myeloid precursors.67,68 The variations in the cellular patterns and time courses of hematopoietic suppression after the administration of different alkylating agents indicate that the individual agents have selectivity for different hematopoietic precursor cells (Figure 51-15).

Peptide growth factors, such as granulocyte-macrophage colony-stimulating factor (sargramostim, GM-CSF) and granulocyte colony-stimulating factor (filgrastim, G-CSF), which stimulate the differentiation and proliferation of hematopoietic precursors,261 are now used clinically. The degree and duration of granulocyte depression after antitumor drug administration can be reduced by the concomitant use of these growth factors.262–264 Currently, growth factors that may stimulate the proliferation and restoration of megakaryocytes and platelets are under investigation.265,266 The use of these factors with the alkylating agents is particularly attractive because of the steep dose-response curve of the alkylating agents and because, with several alkylating agents, a considerable increase in dose may be administered before another dose-limiting toxicity is reached. For these same reasons, combinations of alkylating agents have been used extensively in association with allogeneic and autologous bone marrow transplantation.71,267

Gastrointestinal Toxicity

Damage to the gastrointestinal tract is a toxicity that frequently occurs with high-dose regimens. Mucositis, stomatitis, esophagitis, and diarrhea occur with high doses of alkylating agents, particularly after high doses of melphalan and thiotepa or combinations of alkylating agents, including melphalan or thiotepa.268–270 Significant mucositis is unusual even after very high doses of cyclophosphamide or ifosfamide. This lack of gastrointestinal toxicity is probably caused by the presence of the enzyme aldehyde dehydrogenase in the epithelial cells of the gastrointestinal tract.13

Nausea and vomiting are frequent side effects of alkylating agents. Although these side effects are not usually life threatening, they are major discomforts to patients and may result in the delay or discontinuation of therapy. The nausea and vomiting are, at least in part, mediated through the CNS and are not caused by direct gastrointestinal toxicity.271,272 These effects are variable between patients in that some people tolerate high doses of these drugs without nausea and vomiting, whereas other patients are incapacitated by even low doses of alkylating agents. The frequency of nausea and vomiting does increase as the dose of alkylating agents is increased. Therefore, it is important, especially with the use of increasing doses of alkylating agents, to provide the patient with adequate antiemetic medication. Such medications include phenothiazines, other antiemetics, acute doses of corticosteroids, and, more recently, antiserotonin agents.273–275

Venoocclusive Disease of the Liver

This syndrome is characterized clinically by hepatomegaly, right upper quadrant pain, jaundice, ascites, and a high mortality rate from hepatic failure. Pathologically, the syndrome is associated with subendothelial thickening and narrowing of the hepatic venule lumen.276 This complication has been seen in approximately 25% of patients receiving high-dose cyclophosphamide and busulfan (see Figure 51-17) or cyclophosphamide and total-body irradiation prior to allogeneic or autologous bone marrow transplantation for leukemia or lymphoma,276 and has also been seen after other high-dose alkylating agent therapy.277,278 Liver transplantation has been used for the treatment of venoocclusive disease in patients after bone marrow transplantation.279,280

Figure 51-17. Structures of platinum antitumor agents.

Figure 51-17

Structures of platinum antitumor agents.

Figure 51-18. Aquation of platinum compounds and reaction with nucleophiles.

Figure 51-18

Aquation of platinum compounds and reaction with nucleophiles.

Gonadal Damage

A serious toxicity of the alkylating agents is gonadal damage. The characteristic lesion in men, depletion of testicular germ cells with preservation of Sertoli cells, was first described in 1948 in patients treated with mechlorethamine.281 This lesion was subsequently observed with other alkylating agents,282 and frequently results in aspermia or oligospermia in men treated with drug combinations, including alkylating agents.283 However, spermatogenesis and fertility may return after several years.284,285

Amenorrhea, associated with disappearance of mature and primordial ovarian follicles, is seen in women treated with alkylating agents.70,286,287 The frequency of amenorrhea increases with the age of the woman and is more likely to be irreversible in older women.288

Pulmonary Damage

Pulmonary damage in the form of interstitial pneumonitis and fibrosis is associated with almost all of the alkylating antitumor drugs. Although the exact mechanism of the pulmonary toxicity is unknown, it is presumably caused by direct toxicity of the alkylating agents to pulmonary epithelial cells. The typical presentation of this toxicity is the onset of a nonproductive cough and dyspnea, which may progress to tachypnea and cyanosis, and even to severe pulmonary insufficiency and death. This complication was first described in association with busulfan therapy,289 but subsequently it was described after cyclophosphamide,290,291 nitrosoureas,292,293 melphalan,294 chlorambucil,295 and mitomycin C.296 A significant incidence of pulmonary toxicity has been reported in patients receiving high doses of cyclophosphamide, cisplatin, and BCNU.297

Hemorrhagic Cystitis

The oxazaphosphorines cyclophosphamide and ifosfamide produce bladder toxicity, which is not seen with other alkylating agents. This toxicity is a hemorrhagic cystitis, which may progress to massive hemorrhage.298,299 The toxicity is caused by the metabolites of these drugs, which are excreted into the urine. The metabolite principally responsible for this toxicity is acrolein,300 although phosphoramide mustard and chloroacetaldehyde may contribute to the effect. Hemorrhagic cystitis is seen more commonly after ifosfamide therapy than cyclophosphamide, partly because higher doses of this agent are used (see Nirogen Mustards). Renal tubular damage has also been seen after ifosfamide, including a Fanconi-type syndrome with azotemia, elevated serum creatinine, and enzymuria.301

The systemic administration of thiols can prevent or ameliorate the bladder damage from cyclophosphamide and ifosfamide because the thiols conjugate the aldehyde functions of acrolein and chloroacetaldehyde. The most widely used compound to prevent oxazaphosphorine bladder toxicity is the sodium salt of 2- mercaptoethane sulfonate (MESNA).302 MESNA is usually administered to all patients receiving ifosfamide and to patients who are receiving high-dose cyclophosphamide. Subclinical renal toxicity has been observed in children receiving ifosfamide,16,303 despite MESNA administration, so that administration of MESNA does not eliminate the need for adequate hydration and careful observation of the patient.


An antidiuretic effect is commonly seen in patients receiving doses of cyclophosphamide of 50 mg/kg or greater.304,305 This syndrome is characterized by a decrease in urine output 6 to 8 h after drug administration, weight gain, a marked increase in urine osmolality, and a decrease in serum osmolality and sodium concentration. Pericardial and pleural effusions may be seen, and seizures caused by hyponatremia have occurred after cyclophosphamide therapy,306 especially if low-sodium replacement fluids were administered. This antidiuretic syndrome appears to be caused by an effect of cyclophosphamide metabolites on the distal renal tubule and is self-limited, with the excess fluid excreted over a period of about 12 h. Administration of furosemide promotes free water clearance and ameliorates the syndrome.307

Renal Toxicity

Renal toxicity is a serious toxicity of the nitrosoureas.308,309 This effect is dose-related and may produce severe renal failure and death after administration of more than 1,200 mg of BCNU. Elevation of serum creatinine and other clinical evidence of renal toxicity may not be seen until after the completion of therapy. The histology of the kidneys in patients with renal nitrosourea damage is similar to that in radiation nephritis. A case of acute renal failure after melphalan therapy has been reported.310


Although the association between an alkylating agent and alopecia was first described with busulfan therapy,311 this toxicity is predominantly associated with cyclophosphamide and ifosfamide therapy. The alopecia produced by these agents can be quite severe, especially if the agent is given in combination with vincristine or doxorubicin. Regrowth of the hair occurs after cessation of therapy and may be associated with a change in the texture and color of the hair.312 The structure–function studies of Feil and Lamoureaux313 suggest that this toxicity is a result of the entry of lipophilic metabolites into the hair follicles. This suggestion is consistent with the fact that busulfan, vincristine, and adriamycin are all lipophilic molecules.

Allergic and Hypersensitivity Reactions

Because the alkylating agents react with many biologic molecules, it is not surprising that they would serve as haptens and produce allergic reactions.314–316 The most frequent reactions that have been reported have been cutaneous hypersensitivities. Anaphylactic reactions are rare, but they have occurred.317 Patterns of cross-reactivity have not been carefully defined, but cross-reactivity between agents of similar structure, such as the nitrogen mustards, have been described.316,318


The nonhematologic dose-limiting toxicity of cyclophosphamide is cardiac toxicity.319–321 The fulminant syndrome is seen most frequently in patients receiving a total dose of cyclophosphamide greater than 200 mg/kg preparatory to bone marrow transplantation. The clinical course of the syndrome consists of the rapid onset of severe heart failure, which is fatal within 10 to 14 days. The hearts of such patients are dilated, with patchy transmural hemorrhage and pericardial effusion. The microscopic findings consist of interstitial hemorrhage and edema, myocardial necrosis and vacuolar changes, and specific changes in the intramural small coronary vessels.320 Decreased electrocardiographic voltage and a transient increase in heart size is seen in high-dose cyclophosphamide patients without clinical symptoms, and the characteristic pathologic findings are present in such patients who die of other causes. Cardiotoxicity and cardiomegaly have been seen in patients receiving lower doses of cyclophosphamide in combination with other alkylating agents.322 Age greater than 50 years and previous Adriamycin exposure appear to increase the risk of cyclophosphamide cardiotoxicity.321


In preclinical studies of alkylating agents, convulsions were often seen.323 At the usual clinical doses of these agents, frank neurotoxicity is not usually seen but drowsiness and alterations of consciousness can be seen.324 With the increasing use of higher doses of alkylating agents and combinations of alkylating agents, more clinical neurotoxicity is being seen.325 At BCNU doses of 1,200 mg/m2, severe CNS toxicity has been seen,326 and the intracarotid administration of BCNU has produced severe eye pain and blindness.327 High-dose busulfan therapy produces seizures, and anticonvulsants are often used prophylactically in these patients.328


Studies carried out in vivo and in embryo cultures demonstrate that virtually all of the alkylating agents are teratogenic.329,330 The teratogenic effect is probably a result of cytotoxic effects on the embryo by the same mechanisms by which the compounds are toxic to tumor cells.331–334 The available clinical information indicates that there is a definite risk of a malformed infant if the mother is treated with an alkylating agent during the first trimester of pregnancy.335–337 In a review of the literature, Nicholson338 found that of 25 women who had received alkylating agents during the first trimester of pregnancy, there were four fetal malformations. However, the administration of alkylation agents during the second and third trimesters is not associated with an increased risk of fetal malformation.338–340


Since the initial reports of acute leukemia occurring in patients treated with alkylating agents,341–344 it has become increasingly obvious that this type of oncogenesis is a significant complication of alkylating agent therapy. Several studies indicate that the rate of acute leukemia after alkylating agent therapy may be 10% or higher in certain groups of patients.345–347 Procarbazine and other methylating agents appear to be the most potent oncogenic agents,348 and melphalan appears to produce a higher rate of acute leukemia than does cyclophosphamide.349 The lesser leukemogenic potential of cyclophosphamide may well be related to the hematopoietic stem cell-sparing effect of this agent.13 An increased rate of solid tumors is also seen in patients treated with alkylating agents.350,351 Although sufficient data are not yet available to be certain, it appears that high-dose alkylating agent therapy administered in intermittent pulses over a relatively short period of time is less oncogenic than prolonged alkylating agent therapy.


The immunosuppressive effect of alkylating agents was first described by Hektoen and Corper352 for sulfur mustard. Cyclophosphamide is particularly immunosuppressive353 and is used for the treatment of autoimmune diseases.354–356 Cyclophosphamide is also used in preparative regimens for allogeneic transplantation because of its immunoablative activity.357 Low doses of cyclophosphamide and melphalan can enhance the immune response by selectively inhibiting the immune suppressor cells.358–360 Because of this effect, moderate doses of cyclophosphamide have been used in conjunction with immunotherapy and biologic response modifiers such as interleukin-2.361,362

The clinical significance of the immunosuppression produced by alkylating agents in their role as antitumor agents is not certain. The two major concerns are susceptibility to infection in the immunosuppressed host and the potential interference with a host immune response to the tumor. The available evidence indicates that most intermittent antitumor regimens do not produce a profound or prolonged immunosuppression.363

Platinum Antitumor Compounds

The platinum antitumor agents are complexes of platinum with ligands that can be displaced by nucleophilic (electron-rich) atoms to form strong bonds with covalent characteristics. Thus, like the alkylating agents, the platinum agents form strong chemical bonds with thiol sulfurs and amino nitrogens in proteins and nucleic acids.

The first platinum antitumor compound was discovered by Rosenberg and colleagues364,365 while studying the effects of electric current on bacterial growth. The growth inhibition observed was found to be caused by a platinum complex of ammonia and chloride, which was produced in the medium from the platinum electrode. These investigators found several such compounds to have antitumor activity against murine tumors in vivo.365 The most active of these compounds was the one now known as cisplatin (Figure 51-17).

Cisplatin went into clinical trials in the early 1970s 366–368 and was found to have significant antitumor activity against testicular cancer, lymphoma, squamous cell carcinoma of the head and neck, ovarian cancer, and bladder cancer. Because of its significant therapeutic effect in these tumors and activity against a number of other solid tumors, it became the most frequently used antitumor agent. Because of the renal and neurotoxicities of cisplatin, there were intensive efforts to devise analogs with fewer of these toxicities. This work led to the development of carboplatin, which produces primarily hematopoietic toxicity and appears to have an antitumor effect similar to cisplatin369–373 against the tumors against which it has been used. A number of other platinum compounds are currently under investigation and are discussed in the section Molecular and Cellular Pharmacology.


The platinum compounds that are active antitumor agents can have either four or six ligands (see Figure 51-17), with a square planar or hexahedral configuration, respectively. Those with four ligands have an oxidation state of +2, and those with six ligands an oxidation state of +4. The chloride ligands of cisplatin and the other complexes with the +2 oxidation state can be exchanged for nucleophilic atoms in the biologic milieu, including the nitrogens of the DNA bases. The chloride ligands of the +4 compounds are much less reactive than those of the +2 compounds,374 and it is likely that the +4 compounds are reduced in vivo to produce the reactive +2 complexes.375–377 The ligand substitution reactions of the square planar complexes occur with retention of the configuration of the platinum complex.378 Because the trans-platinum compounds are essentially inactive as antitumor compounds, the ability of the cis compounds to form certain stereo-specific cross-links probably accounts for their antitumor activity.

In some cis-platinum compounds in clinical use, the chloride leaving ligands are replaced with carboxyl ester groups, as in carboplatin and oxaliplatin (see Figure 51-19). These ligands are less-readily displaced; thus, these compounds require higher concentrations for cytotoxicity. The decreased renal and neurologic toxicity of these compounds is also probably a result of their being less chemically reactive than cisplatin. Substitutions on the amino groups alter the lipophilicity and distribution of the agent.

Figure 51-19. Platinum-DNA adducts.

Figure 51-19

Platinum-DNA adducts.

Cellular and Molecular Pharmacology

Although the chloride and carboxyester ligands can probably be directly displaced by biologic atoms, it is likely that, in the biologic milieu, the chloride or carboxy ligands are displaced by water molecules to form the aquo ligand, which is a better leaving group than the chloride or carboxy groups.378 The high chloride content of the extracellular fluid maintains the platinum compounds in the chloride and less reactive form. However, in the lower chloride content of the cell, the more reactive aquo species is formed. The loss of a proton produces the hydroxy ligand, which is unreactive.379Figure 51-20 illustrates the proposed aquation pathway for cisplatin. The platinum compounds react with many biologic molecules, but there is considerable evidence that these compounds, like the bifunctional alkylating agents, exert their cytotoxic effect by reacting with DNA and interfering with DNA replication and cell division. Roberts and Pera380 demonstrated that the amount of platinum bound to DNA was directly related to the degree of toxicity of platinum compounds. Zwelling and colleagues381 demonstrated that the degree of DNA interstrand cross-linking in vitro and in vivo was directly related to the degree of cytotoxicity in rodent tumor cells.

Figure 51-20. Clinical pharmacokinetics of cisplatin after single injection of 100 mg/m2.

Figure 51-20

Clinical pharmacokinetics of cisplatin after single injection of 100 mg/m2. Adapted from Patton et al.

The cis-platinum compounds, like the alkylating agents,382–384 react with nitrogen atoms of DNA and preferentially react with the N-7 atom of deoxyguanylic acid. Specific adducts of Pt compounds with DNA have now been characterized and studied.380 The consensus of the studies is that the most frequent adducts are dGpdG and dApdG, which result from the cis-platinum complex binding to adjacent deoxyguanylates or an adjacent deoxyadenylate and deoxyguanylate in a strand of DNA to produce an intrastrand cross-link in both situations. A less-common lesion is one that results from binding of the platinum atom to the N-7 of a deoxyguanylate in one strand of DNA and to the N-7 atom of a deoxyguanylate in the complementary strand of DNA, thereby producing an interstrand cross-link. Repair of these lesions does occur, and the cytotoxicity to the cell is probably determined by the resultant formation and repair of the lesions.385 As mentioned above, a close correlation between interstrand DNA cross-linking has been demonstrated, but equally precise methods for quantifying intrastrand cross-links in whole cells after drug exposure are not available. Thus, intrastrand DNA cross-links might correlate equally well or better with cytotoxicity. The DNA adducts formed by Pt compounds other than cisplatin have been less well studied but appear to be similar to those formed by cisplatin.386–388

Although there is considerable evidence that the formation of DNA adducts is responsible for the cytotoxicity of the platinum antitumor agents, the mechanisms through which the cytotoxic effects are mediated are not well understood. Evidence has been presented that the platinum adducts inhibit replication.389,390 Heiger-Bernays and colleagues391 have demonstrated that as few as two platinum adducts per genome were sufficient for inhibition of DNA replication by cisplatin. Sorenson and Eastman392 found that cytotoxicity with cisplatin was correlated with the duration of arrest in the G2 phase of the cell cycle and postulated that the G2 arrest was caused by the inability of the cells to transcribe the Ptdamaged DNA and produce the messenger ribonucleic acid (mRNA) essential for mitosis.

In 1996, Drummond and colleagues393 demonstrated that ovarian tumor cells resistant to cisplatin were deficient in the MutL alpha MLHI subunit and suggested that the mismatch repair system recognized the cisplatin cross-link and played a role in the cytotoxicity of cisplatin. Similar findings were reported by others,394 and Mello and colleagues395 postulated that the mismatch repair protein hMSH2 played an active role in mediating cisplatin cytotoxicity. Vaisman and colleagues suggested that mismatch repair defects result in increased replicative bypass of cisplatin adducts.396 It has also been established that transplatin lesions, and those produced by oxaliplatin, tetraplatin, and Bis-aceto-ammine-dichloro-cyclohexamine-platinum IV (JM-216), are not recognized by the mismatch repair system,397 and that these agents produce cytotoxicity in mismatch repair deficient cells. The latter three agents produce a cross-link containing the bulky cyclohexylamine group (see Figure 51-17). Takahara and colleagues and Gelesco and Lippard have now reported both the crystal structure398 and the nuclear magnetic resonance (NMR) solution structure399 of the cisplatinum d(GpG) cross-link. This type of structural understanding should increase the interpretation of the functional effects of the platinum cross-links.

Mechanisms of Cellular Resistance to Platinum Agents

A number of mechanisms of cellular resistance to platinum compounds have been described. These mechanisms include decreased uptake of the platinum compound into resistant cells, inactivation of the drug by cellular thiol compounds, enhanced repair of the platinum-related DNA damage, and the absence of mismatch repair, as described above.

Decreased cellular uptake of cisplatin by cells resistant to the compound has been described by a number of investigators.400–404 The uptake of cisplatin into cells is linear for more than an hour and does not appear to be an active transport process, although it is partially inhibited by metabolic inhibitors.400 There has also been a report of increased efflux of cisplatin in a resistant cell line.405 Mann and colleagues406 could not demonstrate changes in the physical properties of the cell membrane of cells resistant to cisplatin. Thus, although decreased cellular accumulation of the platinum compounds appears to be one type of cellular resistance, the mechanism of this type of resistance remains undefined and may be related to altered binding of the agents to cellular proteins, rather than alteration of passage through the cell membrane.407

A number of investigators have demonstrated that both rodent and human tumor cells that are selected in vitro or in vivo by exposure to the platinum antitumor compounds frequently demonstrate elevated glutathione levels in association with resistance to these drugs.408–413 Tumor cell lines derived from patients resistant to therapy with cisplatin have also been found to have elevated glutathione levels.414,415

Further evidence that glutathione is involved in resistance to platinum compounds can be inferred from the fact that several investigators have shown that tumor cells can be sensitized to the platinum agents by depletion of cellular glutathione by treatment with buthionine sulfoximine, an inhibitor of glutathione synthesis.413–418

The mechanism(s) through which glutathione-associated resistance is mediated have not been definitively elucidated. Andrews and colleagues416 demonstrated that cisplatin binds to glutathione, and Dedan and Borch419 have studied the reaction rates of cisplatin with various thiols, including glutathione, and characterized a reaction product in which two glutathiones appeared to be bound to each platinum through the cysteine residues of the glutathiones. The thiol platinum ligand is very stable and thus will not react further. Eastman410 has presented evidence that glutathione may react with monofunctional adducts on DNA to quench the second reactive ligand and prevent cross-link formation. Resistance to cisplatin has also been associated with elevation of glutathione transferase enzyme activity, increased levels of the pi (acidic) isozyme of the protein, and increased levels of the mRNA for the pi isozyme.420–422 However, the catalysis of the conjugation of glutathione with platinum agents by this enzyme has not been characterized.

Cellular resistance to platinum agents is also associated with another sulfhydryl-containing protein, metallothionein. Several investigators have found that tumor cells exposed to heavy metals, such as cadmium, develop resistance to cisplatin, which is associated with increased cellular levels of metallothionein.423–425 In one report, transfection of cells with the metallothionein gene resulted in increased metallothionein levels and resistance of the cells to cisplatin, melphalan, and chlorambucil.425 Naganuma and colleagues426 reported that administration of bismuth subnitrate to mice produced increased levels of metallothionein in the kidneys and resulted in protection of the mice from the renal and gastrointestinal toxicity of cisplatin but did not affect the response of transplanted tumors to cisplatin in the mice. Cisplatin binds to metallothionein in Ehrlich ascites tumor cells427 and in the liver and kidney of rats,407,428 and the systemic administration of cisplatin or its hydrolyzed product can induce metallothionein in the liver and kidney.429 These findings indicate that metallothionein can protect both tumor and normal cells from cisplatin, although the binding of the drug to this protein has not been characterized.

As with the alkylating agents, there is extensive evidence that enhanced DNA repair can be responsible for resistance to the platinum compounds. Van Den Berg and Roberts430 first reported that caffeine, a known inhibitor of DNA repair, potentiated cytotoxicity and chromosomal damage in mammalian cells, and shortly thereafter Fravel and Roberts demonstrated that excision repair of cisplatin-damaged DNA does occur in treated cells.431 Many subsequent studies have demonstrated that cells deficient in DNA repair, such as those from patients with xeroderma pigmentosum or Fanconi anemia, are very sensitive to cisplatin.432–436

Agents that are known to inhibit the activity of enzymes involved in the repair of DNA, such as aphidicolin and novobiocin, sensitize cells to cisplatin and to reverse the resistance of repair-resistant cell lines.437–440 The antitumor agents hydroxyurea and cytosine arabinoside, which inhibit DNA repair synthesis, both produce a synergistic cytotoxic effect with cisplatin.441,442

Studies by Beck and colleagues and Husain and colleagues support the above indications that platinum adducts in DNA are repaired by an excision repair mechanism,443,444 and these investigators445 now report that the nucleotide excision repair system can remove DNA adducts produced from cisplatin, oxaliplatin, and JM-216. These findings are consistent with studies by Dabholkar and colleagues and Li and colleagues demonstrating that ERCC-1 is involved in platinum repair.446,447

It has also been shown that platinum interstrand DNA cross-links are removed more rapidly in cisplatin-resistant cells,448 and that very sensitive tumor cells may have a decreased ability to remove DNA interstrand cross-links.449 A protein, XPE-BF (xeroderma pigmentosum complementation group E binding factor), which binds to Pt-damaged DNA450–452 and may mark it for repair, has been identified. A series of proteins, the HMG domain proteins, which bind to Pt intrastrand cross-links, produce bending of the DNA, and may inhibit the repair of these lesions, has also been described.453,454 It has also been found that cisplatin-resistant cells can have elevated thymidylate synthase activity and be cross-resistant to 5-fluorourocil (5-FU)455 and that c-fos may play a role in the cellular response to Pt agent damage by mediating DNA repair pathways.456–458

Although it is clear that each of these mechanisms can be associated with the resistance of tumor and normal cells to the platinum agents, the relative roles of these mechanisms in the resistance of tumors to treatment in patients have not been established. Such studies and attempts to overcome resistance with BSO and inhibitors of DNA repair are currently in progress.

Platinum Analogs in Clinical Use

Cisplatin and carboplatin are licensed in the United States and internationally, and are used extensively. Because the primary toxicity of carboplatin is hematopoietic, it has replaced cisplatin for use in many patients and is being used particularly in situations where nonhematopoietic toxicity should be avoided, such as high-dose treatment with bone marrow support459,460 or with hematopoietic stimulatory factors. There is no evidence for cross-resistance between these two agents. Iproplatin has been evaluated in Phase II trials but was found to be no more or less effective than carboplatin and produced more hematopoietic and gastrointestinal toxicity.461–465 Tetraplatin (ormaplatin) produced severe neurotoxicity in initial clinical trials,466 but is still being evaluated in Phase I trials.467 Oxaliplatin is similar to tetraplatin in its preclinical toxicity.468 However, this compound has shown promising activity in gastrointestinal tumors, especially in combination with 5-FU and leucovorin.469–471 Oxaliplatin has also demonstrated significant activity in patients with ovarian cancer who previously received cisplatin.472,473 Oxaliplatin demonstrated a modest effect (15% partial response) in advanced, cisplatinresistant non–small-cell lung patients.474

A lipid-soluble platinum compound, JM-216,475 which can be administered orally, is now being evaluated clinically.476–478 The discovery that oxaliplatin and JM-216 are active against tumors lacking mismatch repair (see Mechanisms of Cellular Resistance to Platinum Agents) has stimulated interest in these and related compounds, and some of the new platinum analogs may not be totally cross-resistant with cisplatin because of differences in either cellular uptake402 or cellular detoxification.479

Although oxaliplatin has shown interesting activity against colon cancer, and the oral availability of JM-216 has been established, cisplatin, and especially carboplatin, continue to be the predominant platinum antitumor agents in clinical use.


Platinum antitumor compounds have been measured in human plasma and other human tissues as total platinum, as ultrafilterable platinum, and as the specific parent compounds. Total platinum can be measured by administering analogs containing the radioactive 193Pt or 195Pt isotopes,480,481 by trapping the platinum with an ultraviolet absorbing ligand, such as diethyldithiocarbamate,482 or by flameless atomic absorption spectroscopy.482,483 Ultrafiltration of plasma and other biologic fluids separates the free platinum compounds from those bound to protein. The protein-bound species are biologically inactive and essentially irreversibly bound to the protein.484 Both cisplatin and carboplatin have been measured specifically by separation from other species on high-performance liquid chromatography (HPLC) columns and detection by electrochemical detection or by collecting fractions and quantifying the total platinum in each fraction.485,486 The cisplatinum concentration is consistently between 60% and 80% of the ultrafilterable platinum and follows the same kinetics as the ultrafilterable platinum.485,487 Carboplatin represents a higher percentage of the ultrafilterable platinum and follows kinetics similar to the ultrafilterable platinum. Because of the sensitivity, accuracy, and convenience of the method, flameless atomic absorption spectroscopy is the most common technique used to measure the platinum agents. Furthermore, because measuring filterable species appears to measure the reactive compounds and to approximate closely the measurement of the parent compounds, measurement of ultrafilterable platinum is most commonly used in pharmacokinetic studies.

In pharmacokinetic studies after cisplatin administration, total platinum in the plasma follows a triphasic pattern, with the first phase t½ approximately 30 min, the second phase t½ approximately 60 min, and the third phase t½ more than 24 h.487,488 Measurements of the ultrafilterable platinum indicate that the initial, more rapid clearance phases are a result of the renal clearance of filterable platinum, the majority of which is the parent compound.488 Carboplatin exhibits similar pharmacokinetics, except that the initial half-lives are somewhat longer, less of the total platinum is protein bound, and a greater percentage of the agent is excreted by the kidneys.486,489 The pharmacokinetics of total and filterable platinum after iproplatin administration appears to be similar to those of carboplatin.490 Decreased creatinine clearance results in higher plasma levels of both cisplatin and carboplatin and potentially greater toxicity.

After bolus administration of 100 mg/m2 of cisplatin, initial peak plasma concentrations of 3 to > 5 μg/mL are achieved,485 with this value decreasing to less than 0.2 μg/mL at 2 h. After the usual clinical dose of about 300 mg/m2 of carboplatin, peak plasma levels of about 30 μg/mL are reached, declining to about 5 μg/mL at 2 h.486,489

In typical clinical use, usually in combination with other agents, the platinum antitumor agents are given intravenously, either as a single dose or daily for several days, with repeat courses at 3 to 4 weeks. The agents are given as an infusion over several hours rather than as a bolus dose and, especially with very high doses, may be given as 24 h or longer infusions. Because of the close relationships between plasma AUC of carboplatin and renal function and between AUC of carboplatin and toxicity, dosing algorithms based on renal function have been established and are now widely used in the dosing of carboplatin.491–493

Cisplatin and carboplatin have also been administered regionally. There has been considerable experience with the intraperitoneal route, particularly in the treatment of ovarian cancer.494–496 Very high intraperitoneal concentrations can be obtained, and systemic toxicities can be reduced by the concomitant systemic administration of thiosulfate.497,498 Cisplatin has also been administered intraarterially for the treatment of tumors in the extremities,499–502 brain tumors,503–505 carcinoma of the head and neck,506,507 carcinoma of the liver,508 and carcinoma of the bladder.509,510 Intravesicular instillation of cisplatin has been used for the treatment of superficial cancers of the bladder.511–513 Cisplatin has also been instilled into the pericardial sac for the treatment of malignant pericardial effusions.514,515



The most serious, and usually dose-limiting, toxicity of cisplatin is renal.516,517 This toxicity is manifested clinically by elevated blood urea nitrogen (BUN) and creatinine, is cumulative with continued cisplatin exposure, and is potentiated by other nephrotoxins.518 Decreases in serum electrolytes have been associated with platinum renal toxicity, including symptomatic hypomagnesemia.519 Although the toxicity may remain subclinical, or the renal function return to normal, significant pathologic damage appears to persist.520 The pathology of the renal damage is characterized by focal acute tubular necrosis, dilatation of convoluted tubules, thickened tubular basement membranes, formation of casts, and epithelial atypia of the collecting ducts.520,521 High fluid intake with forced diuresis522,523 can reduce the incidence and severity of the renal toxicity. Systemic administration of thiols can reduce renal toxicity of cisplatin in animal models, and in a clinical trial, systemic diethyldithiocarbamate appeared to reduce nephrotoxicity without affecting ototoxicity or myelosuppression.524 The nephrotoxicity of the second-generation platinum complexes, such as carboplatin and iproplatin, is markedly less than that of cisplatin.


Ototoxicity has been a significant problem with cisplatin. This toxicity is characterized by tinnitus and hearing loss.366,367,525 The hearing loss is usually in the high-frequency range—4,000 to 8,000 Hz—but may occur in the lower ranges, which include the speech frequencies.525,526 Because the higher frequencies are usually involved, the hearing loss may not be symptomatic. Vestibular toxicity does not usually occur but can be seen.527,528 The ototoxicity of cisplatin is dose related and is usually cumulative with subsequent courses of the agent.529,530 Radiation prior to or simultaneous with the cisplatin administration enhances the toxicity,531,532 but this additive effect may be less if the cisplatin precedes the radiation.526

The pathologic findings associated with ototoxicity, in both experimental animals and patients, are selective damage to the outer hair cells of the cochlea and lesions in the organ of Corti, the spiral ganglion and cochlear nerve, and the stria vascularis.533–536 In studies of organ cultures of the cochlear structures, the hair cells are very sensitive to very low concentrations of cisplatin.537 Vestibular toxicity is associated with degeneration of the maculae and cristae.528


The neurotoxicity seen with the administration of cisplatin consists principally of peripheral neuropathy involving both the upper and lower extremities, with paresthesias, weakness, tremors, and loss of taste.538 Seizures and leukoencephalopathy have also been described.539–542 The neurotoxicity may be persistent543 and may progress after cessation of cisplatin therapy.542 The quantitative determination of vibratory perception threshold has been reported to correlate with cisplatin neurotoxicity.544

Particularly severe neurotoxicity has been reported after intraarterial infusions of cisplatin, with cranial nerve paralysis occurring after intraarterial infusions for head and neck cancer,541,542 and severe peripheral neuropathy occurring after lower limb perfusion.545 In experimental animals, severe CNS toxicity was seen when compounds that open the blood-brain barrier were administered prior to systemic cisplatin treatment, and intracarotid cisplatin produced damage to the blood-brain barrier and severe neurotoxicity.546 However, severe neurotoxicity was not seen in patients treated with intracarotid cisplatin for primary brain tumors.547 The neurotoxicity of ifosfamide has been reported to be enhanced by prior treatment with cisplatin.548

Because various pharmacologic maneuvers have been able to control or reduce the nephrotoxicity and severe nausea and vomiting produced by cisplatin, neurotoxicity has become the dose-limiting toxicity of cisplatin.549 An interesting observation is that treatment of animals with an adrenocorticotropin analog will prevent neurotoxicity from cisplatin and will facilitate the recovery of established neurotoxicity,550,551 but will not interfere with the antitumor effect of the agent. In a randomized, placebo-controlled clinical trial, this compound appeared to prevent or ameliorate the neurotoxicity of cisplatin.551 Neither carboplatin or iproplatin appear to produce significant neurotoxicity with the doses used thus far with autologous bone marrow transfusion.552–554

Gastrointestinal Toxicity

Severe nausea and vomiting have been a significant problem with cisplatin, occurring in almost all patients receiving the drug.517,555 The cause of this toxicity is not firmly established. Work in animal models indicates that abdominal visceral innervation and 5-hydroxytryptamine receptors on visceral afferent nerves play a role in mediating this toxicity,556 but there is also evidence that the chemoreceptor trigger zone in the medulla plays a role.557,558 The use of metoclopramide, a dopamine antagonist, prior to and during cisplatin administration is effective in controlling this toxicity559,560; the steroids dexamethasone and methylprednisolone, alone or in combination with metoclopramide, are also useful.561–563 More recently, antiserotonin analogs such as ondansetron and granisetron have proven highly effective in controlling nausea and vomiting after platinum administration. The gastrointestinal toxicities of carboplatin and iproplatin are much less than those of cisplatin.564–566 With the increasing use of carboplatin, and the more effective antiemetic regimens, the gastrointestinal toxicity has not limited the increasing use of platinum chemotherapy.

Immune Effects

In contrast to the alkylating agents, many of which are significantly immunosuppressive, cisplatin appears to have no immunosuppressive effect at the usual clinical doses and may even augment immune function at these doses.567 Monocyte-mediated cytotoxicity was found to be increased in ovarian cancer patients after cisplatin treatment,568 and OKT81 cytotoxic cells were increased in patients after cisplatin therapy.569

By agreement with the publisher, this book is accessible by the search feature, but cannot be browsed.

Copyright © 2003, BC Decker Inc.
Bookshelf ID: NBK12772


  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...