Alkylating agents were the first cytotoxic drugs used in clinical oncology, derived directly from observations of lymphoid tissue destruction by sulfur mustard during World War I. Despite their age, they remain among the most widely used agents in cancer chemotherapy, appearing in regimens for lymphoma, leukemia, breast cancer, ovarian cancer, sarcoma, and brain tumors. Their mechanism, covalent modification of DNA (deoxyribonucleic acid), is shared by a structurally diverse group of compounds whose clinical toxicity profiles differ substantially and whose ADME (absorption, distribution, metabolism, and excretion) characteristics require careful attention to dosing and organ function.1
The defining biochemical event of alkylating agent action is the formation of a highly reactive electrophilic intermediate that attacks nucleophilic sites on DNA, RNA (ribonucleic acid), and protein. The primary cytotoxic lesion is alkylation at the N7 (nitrogen-7) position of guanine, the most nucleophilic site in double-stranded DNA (dsDNA). N7-guanine alkylation can result in depurination, miscoding during replication, or the more lethal cross-linking reactions that prevent strand separation. Interstrand cross-links (ICLs), which covalently bridge complementary strands of dsDNA and physically prevent replication fork progression, are the most cytotoxic lesion produced by bifunctional alkylating agents. Intrastrand cross-links, in which two adjacent guanines on the same strand are bridged, as in the 1,2-d(GpG) adduct that accounts for approximately 65% of cisplatin lesions, distort the DNA helix and block RNA polymerase and DNA polymerase progression. Because alkylation damages DNA regardless of cell cycle position, alkylating agents are classified as cycle-nonspecific, and their dose-response curves are more nearly linear than those of cycle-specific antimetabolites, making dose escalation a meaningful strategy up to the limits imposed by organ toxicity.12
Alkylating agents are classified by chemical mechanism into five major subgroups. Nitrogen mustards (cyclophosphamide, ifosfamide, mechlorethamine, melphalan, chlorambucil, bendamustine) generate aziridinium ion intermediates that alkylate DNA. Nitrosoureas (carmustine [BCNU (bis-chloroethylnitrosourea)], lomustine [CCNU (chloroethylcyclohexylnitrosourea)], streptozocin) undergo spontaneous decomposition to generate both a chloroethylating species and an isocyanate that carbamylates proteins; their high lipophilicity confers central nervous system (CNS) penetration. Platinum compounds (cisplatin, carboplatin, oxaliplatin) form platinum-DNA adducts after aquation, a hydrolytic activation step. Alkyl sulfonates (busulfan) alkylate through a different electrophilic mechanism and have particular selectivity for hematopoietic stem cells. Triazenes and hydrazines (dacarbazine, temozolomide, procarbazine) require metabolic or spontaneous activation to generate methylating or alkylating species. All bifunctional agents (those bearing two reactive groups) produce ICLs more readily and are generally more cytotoxic than monofunctional agents, which produce primarily single-strand adducts.12
Bifunctional alkylating agents carry two reactive groups and produce interstrand cross-links, which are the most lethal DNA lesions because they completely block replication fork progression. Most clinically used alkylating agents (cyclophosphamide, ifosfamide, platinum compounds, nitrosoureas, busulfan, mechlorethamine) are bifunctional. Monofunctional agents (some metabolites of procarbazine, dacarbazine) produce single-strand adducts with lower cytotoxic potency per lesion. The distinction matters clinically because bifunctional agents generally require nucleotide excision repair (NER) and Fanconi anemia pathway activity for cellular resistance, while monofunctional agents rely more on base excision repair (BER).
Cyclophosphamide is among the most widely used cytotoxic drugs in oncology and rheumatology, appearing in regimens for breast cancer, lymphoma, leukemia, ovarian cancer, sarcoma, and autoimmune diseases. Its unique toxicity profile, determined by the metabolic pathway generating both the active alkylating species and the toxic byproduct acrolein, illustrates how understanding prodrug activation chemistry directly shapes clinical management and preventive strategy.3
Cyclophosphamide is an inactive prodrug that requires hepatic oxidation by cytochrome P450 (CYP) enzymes, principally CYP2B6 (the predominant isoform for cyclophosphamide activation) with contributions from CYP3A4 (cytochrome P450 3A4) and CYP2C9 (cytochrome P450 2C9), to generate 4-hydroxycyclophosphamide (4-OH-CP). This intermediate is in equilibrium with its tautomer aldophosphamide, which spontaneously decomposes in tissues to yield phosphoramide mustard, the active bifunctional alkylating species, and acrolein (propenal), a highly reactive alpha,beta-unsaturated aldehyde (an organic compound with a carbonyl group adjacent to a carbon-carbon double bond) that is the primary cause of cyclophosphamide-induced hemorrhagic cystitis (HC). Phosphoramide mustard is itself a bifunctional alkylating agent that produces DNA (deoxyribonucleic acid) interstrand and intrastrand cross-links. The CYP-mediated activation step occurs primarily in the liver, and the resulting active metabolites are distributed systemically; tumor cells, hepatocytes, and urothelial cells are all exposed. The aldehyde dehydrogenase (ALDH) enzyme expressed at high levels in normal hematopoietic stem cells and hepatocytes rapidly inactivates aldophosphamide to the non-toxic carboxyphosphamide, conferring relative protection of these cell populations. Tumor cells with low ALDH activity are comparatively more vulnerable to the active metabolite.34
The prevention of cyclophosphamide-induced hemorrhagic cystitis (HC) depends on understanding acrolein pharmacokinetics. Acrolein is excreted unchanged in urine and achieves high concentrations in the bladder lumen, where it causes direct urothelial cytotoxicity manifesting as dysuria, hematuria, and in severe cases, bladder fibrosis and urothelial carcinoma with long-term exposure. Two strategies prevent HC: forced hydration dilutes urinary acrolein concentrations and promotes rapid elimination, and mesna (sodium 2-mercaptoethane sulfonate) binds acrolein in the urinary tract through a thiol-disulfide exchange reaction, forming a non-toxic conjugate that is excreted renally. Mesna is administered concurrently and after ifosfamide (where HC risk is higher due to the greater acrolein load per gram of drug) and is required whenever ifosfamide is used. For standard-dose oral or intravenous cyclophosphamide, forced hydration alone is often sufficient; mesna is generally reserved for high-dose cyclophosphamide regimens. The daily first-morning urine should be tested for blood in any patient receiving these agents to detect early HC before frank hematuria develops.34
The pharmacokinetics of cyclophosphamide are favorable for clinical use. It is well absorbed orally, with bioavailability exceeding 75%, and achieves a half-life of approximately 4 to 8 hours after intravenous administration. It is renally cleared as both parent drug and active metabolites; modest dose reduction is warranted when creatinine clearance falls below 10 mL/min, but routine dose reduction is not required for moderate renal impairment. The drug crosses the blood-brain barrier (BBB) to a limited extent, which limits its efficacy in CNS (central nervous system) disease but also reduces CNS toxicity. Autoinduction of CYP2B6 and CYP3A4 occurs with repeated dosing, increasing the rate of 4-hydroxylation and potentially reducing plasma half-life with chronic administration. Significant drug interactions involve inducers and inhibitors of CYP2B6 and CYP3A4; rifampin (a potent CYP inducer) increases activation and may increase toxicity, while azole antifungals reduce activation and may reduce efficacy.34
Ifosfamide is a structural isomer of cyclophosphamide in which one chloroethyl arm has been moved from the nitrogen to the exocyclic position. It requires the same CYP-mediated hepatic activation as cyclophosphamide but has a substantially higher rate of side-chain oxidation (N-dechloroethylation), generating chloroacetaldehyde, a neurotoxic metabolite responsible for ifosfamide encephalopathy. Ifosfamide encephalopathy presents with confusion, somnolence, cerebellar ataxia, visual hallucinations, seizures, or coma, typically beginning 12 to 48 hours after the start of an ifosfamide infusion. The incidence varies from 10 to 40% depending on dose, infusion duration, and patient risk factors (hypoalbuminemia, renal impairment, prior cisplatin exposure, and high creatinine all increase risk). The treatment for established ifosfamide encephalopathy is methylene blue, given intravenously at 50 mg every 4 to 8 hours. Methylene blue acts as an electron acceptor that reverses the NADH (nicotinamide adenine dinucleotide, reduced form) accumulation caused by chloroacetaldehyde-mediated inhibition of flavin-dependent mitochondrial enzymes, restoring mitochondrial oxidative function. Ifosfamide also produces a higher acrolein load per gram than cyclophosphamide, making mesna prophylaxis mandatory at all doses. Renal tubular toxicity, manifesting as Fanconi syndrome (proximal tubular dysfunction with phosphate, bicarbonate, glucose, and amino acid wasting), is more prominent with ifosfamide than cyclophosphamide and may be dose-limiting in pediatric patients receiving repeated cycles.34
Mechlorethamine (nitrogen mustard) is the original nitrogen mustard and is the most reactive and vesicant of the class. It is chemically unstable in aqueous solution and must be prepared and administered immediately. It appears in the MOPP (mechlorethamine, vincristine, procarbazine, prednisone) regimen for Hodgkin lymphoma and has largely been replaced in most other settings by more stable alkylating agents. Its extreme vesicant properties make extravasation a serious hazard; sodium thiosulfate is the antidote for extravasation. Melphalan (L-phenylalanine mustard) is used in multiple myeloma conditioning regimens before autologous stem cell transplantation (high-dose melphalan 140 to 200 mg/m²) and as standard-dose therapy for myeloma (oral). Its absorption orally is highly variable (15 to 90%), mandating intravenous dosing for high-dose conditioning. Dose reduction is required in renal impairment. Chlorambucil is an oral nitrogen mustard used in chronic lymphocytic leukemia (CLL) and low-grade lymphomas. It is well tolerated and produces predominantly myelosuppression as its dose-limiting toxicity. Bendamustine is a hybrid molecule combining a nitrogen mustard alkylating group with a benzimidazole ring that contributes purine analog-like properties; it is active in CLL, follicular lymphoma, and mantle cell lymphoma and produces less immunosuppression than some other alkylating agents used in these diseases.11
Ifosfamide encephalopathy is underrecognized because its early manifestations (confusion, somnolence, agitation) overlap with opioid side effects, metabolic derangements, and other chemotherapy toxicities. The key distinguishing feature is the temporal relationship to ifosfamide infusion onset (typically 12 to 48 hours). Risk factors that should prompt heightened vigilance include hypoalbuminemia below 3.5 g/dL, serum creatinine above 1.2 mg/dL, prior cisplatin nephrotoxicity, oral ifosfamide (higher chloroacetaldehyde generation), and high ifosfamide dose per cycle. Treatment is methylene blue 50 mg IV every 4 to 8 hours; do not delay treatment while waiting for confirmatory testing. Discontinue ifosfamide. Methylene blue is also used prophylactically at 50 mg IV three times daily in subsequent cycles when encephalopathy has occurred previously.
The platinum compounds are among the most clinically important cytotoxic agents, forming the backbone of regimens for testicular, ovarian, bladder, lung, head and neck, and colorectal cancers. Although they share the platinum-DNA adduct mechanism, the three compounds differ substantially in their toxicity profiles, pharmacokinetics, and clinical indications, and understanding these differences is essential for appropriate drug selection and dose adjustment.5
Cisplatin (cis-diamminedichloroplatinum[II]) undergoes a mandatory activation step called aquation in which the two chloride ligands are sequentially replaced by water molecules in the low-chloride environment of the cell interior. The resulting aquated platinum species (cis-[Pt(NH3)2(H2O)2]₂⁺) is the electrophilic species that attacks N7 (nitrogen-7) of guanine, forming predominantly 1,2-d(GpG) intrastrand adducts (approximately 65% of lesions), 1,2-d(ApG) intrastrand adducts (approximately 25%), and interstrand cross-links (approximately 5 to 8%). The extracellular environment is chloride-rich (approximately 100 mEq/L), which suppresses aquation and allows cisplatin to circulate as the neutral parent compound before entering cells. Once intracellular, where chloride concentration falls to approximately 4 mEq/L, aquation proceeds rapidly. Cisplatin enters cells primarily through the copper transporter CTR1 (copper transporter 1), and efflux through ATP7A (ATPase copper transporting alpha) and ATP7B (ATPase copper transporting beta), copper-exporting ATPases (adenosine triphosphatase enzymes), is a key resistance mechanism. Cisplatin is eliminated almost entirely by renal glomerular filtration and tubular secretion, with a triphasic plasma half-life: the initial alpha phase is approximately 20 to 30 minutes, the beta phase 60 to 90 minutes, and the terminal gamma phase extending to several days due to slow release of platinum from tissue-bound adducts.5
Cisplatin nephrotoxicity is dose-dependent, cumulative, and potentially severe, and its prevention requires meticulous pre-hydration and post-hydration protocols. The mechanism involves direct tubular cell injury from platinum accumulation in proximal tubular cells and the loop of Henle, leading to reduced glomerular filtration rate (GFR), hypomagnesemia (from tubular magnesium wasting, which is often dose-limiting before overt nephrotoxicity develops), hypokalemia, hypocalcemia, and in severe cases, acute tubular necrosis (ATN). The standard prevention protocol requires at least 1 to 2 liters of normal saline infused before cisplatin, aggressive saline diuresis during and after the infusion, and supplemental magnesium in every cisplatin cycle regardless of baseline serum magnesium. Amifostine is an organic thiophosphate prodrug that is selectively dephosphorylated by alkaline phosphatase in normal tissues to its active cytoprotective thiol metabolite WR-1065 (the dephosphorylated free thiol form of amifostine), which scavenges free radicals and binds platinum in normal cells. Amifostine is approved for cisplatin nephroprotection in non-small cell lung cancer (NSCLC) and reduces the incidence of clinically significant nephrotoxicity when given before each cisplatin dose, though its use is limited by infusion-related hypotension and nausea. Ototoxicity from cisplatin is sensorineural (affecting high-frequency hearing first), dose-dependent, cumulative, and largely irreversible; audiometric testing is recommended before and during cisplatin therapy in patients at high risk (prior ear disease, concurrent aminoglycoside use, pediatric patients).5
Carboplatin differs from cisplatin in that the two chloride leaving groups are replaced by a cyclobutanedicarboxylate (CBDCA) bidentate ligand, which is hydrolyzed more slowly than chloride and produces a less reactive platinum species. The slower aquation rate reduces reactivity with normal tissue proteins and accounts for carboplatin's substantially lower nephrotoxicity, neurotoxicity, and emetogenicity compared to cisplatin. The dose-limiting toxicity of carboplatin is myelosuppression, predominantly thrombocytopenia, which is predictably related to the area under the concentration-time curve (AUC) rather than to dose per unit body surface area. Carboplatin is eliminated almost exclusively by renal filtration as the intact complex, making GFR the primary determinant of drug exposure. The Calvert formula (dose in mg equals target AUC multiplied by [GFR plus 25]) provides AUC-based dosing that accounts for individual variation in renal function and is the standard approach for all carboplatin administration. Because carboplatin and cisplatin form the same platinum-DNA adducts, platinum resistance mechanisms confer cross-resistance between the two agents; patients who progress on cisplatin-based therapy will generally not respond to carboplatin substitution. Carboplatin cannot substitute for cisplatin in curative-intent regimens such as BEP (bleomycin, etoposide, platinum) for testicular germ cell tumors, where the superior clinical outcome data are specific to cisplatin.56
Oxaliplatin is a third-generation platinum compound in which the amine ligands of cisplatin are replaced by a 1,2-diaminocyclohexane (DACH) carrier ligand. The DACH-platinum adducts formed on DNA (deoxyribonucleic acid) are bulkier than cisplatin adducts and are not recognized by the mismatch repair (MMR) proteins hMSH2 and hMSH6 that normally initiate MMR-mediated apoptosis in response to platinum-DNA damage. This difference explains why oxaliplatin retains activity in MMR-deficient (microsatellite-unstable) tumors that are cisplatin-resistant. The combination of oxaliplatin with fluorouracil and leucovorin (FOLFOX) is standard therapy for metastatic colorectal cancer (CRC) and adjuvant stage III colon cancer. The dose-limiting toxicity of oxaliplatin is peripheral neuropathy, which has two distinct phases: an acute, cold-triggered sensory neuropathy (dysesthesias triggered by contact with cold objects or cold air, occurring within hours of infusion and reversible within days) and a cumulative sensory neuropathy (stocking-glove distribution, progressive with cumulative dose, partially reversible after treatment discontinuation). Nephrotoxicity and significant ototoxicity are not observed with oxaliplatin at standard doses, which is a clinically important advantage over cisplatin.12
Cisplatin is the preferred platinum when maximum antitumor efficacy is essential (testicular cancer BEP, cisplatin-based chemoradiation for cervical, head and neck, and bladder cancer) and when the patient's renal function allows safe administration with aggressive hydration. Carboplatin is substituted for cisplatin when renal impairment, audiometric concerns, poor performance status, or patient preference make cisplatin's toxicity profile unacceptable, and when clinical trial data support equivalence in the relevant tumor type (ovarian cancer, limited-stage NSCLC). Carboplatin should not substitute for cisplatin in settings where clinical outcomes data are cisplatin-specific. Oxaliplatin is used exclusively for colorectal cancer (and selected upper GI tumors) given its FOLFOX efficacy data; it is not interchangeable with the other platinums in most tumor types. Cross-resistance between cisplatin and carboplatin is complete; cross-resistance with oxaliplatin is partial due to the DACH carrier ligand.
The nitrosoureas occupy a distinct pharmacological niche because their high lipophilicity enables central nervous system penetration not achievable with most other alkylating agents, making them clinically useful for primary brain tumors and CNS (central nervous system) lymphoma. Their prolonged and delayed myelosuppression distinguishes them toxicologically from all other cytotoxic drugs and demands specific scheduling awareness to avoid life-threatening cumulative bone marrow suppression.7
Carmustine (BCNU) and lomustine (CCNU) are the two principal nitrosoureas in clinical use. Both undergo spontaneous non-enzymatic decomposition in aqueous solution at physiological pH, generating two species: a chloroethyl carbonium ion (the DNA [deoxyribonucleic acid] alkylating species that produces N1-guanine (N1 denotes the nitrogen-1 position) and N3-cytosine (N3 denotes the nitrogen-3 position) adducts, which rearrange to produce interstrand cross-links more slowly than nitrogen mustards) and an isocyanate (which carbamylates lysine residues on DNA repair proteins, inhibiting their function and potentially contributing to cytotoxicity). The high lipid solubility of nitrosoureas, expressed as a log P value substantially above that of most other alkylating agents, allows passive diffusion across the blood-brain barrier (BBB), achieving therapeutically relevant CNS concentrations. Carmustine achieves CSF (cerebrospinal fluid) concentrations of approximately 15 to 30% of simultaneous plasma concentrations, adequate for treating glioblastoma multiforme (GBM), anaplastic glioma, and primary CNS lymphoma. The BCNU-impregnated biodegradable polymer wafer (Gliadel) is a local delivery formulation implanted in the surgical resection cavity after GBM debulking, providing sustained local carmustine concentrations without significant systemic exposure.7
The most clinically distinctive feature of nitrosourea toxicity is delayed and prolonged myelosuppression. Unlike other alkylating agents whose nadir occurs at 10 to 14 days post-treatment, nitrosoureas produce a nadir at 4 to 6 weeks, with recovery requiring 6 to 8 weeks from the time of administration. This delayed kinetic profile reflects the slow kinetics of the chloroethylating lesion: N1-guanine chloroethyl adducts rearrange to interstrand cross-links over a period of 12 to 24 hours, and the resulting DNA damage triggers delayed apoptosis of hematopoietic progenitors. The clinical implication is that nitrosourea cycles are administered no more frequently than every 6 weeks to allow complete marrow recovery. Patients and clinicians unaware of this delayed nadir may administer a second cycle before the first cycle's myelosuppression has reached its maximum, producing severe cumulative myelosuppression. Pulmonary fibrosis is a cumulative long-term toxicity of carmustine, typically appearing after total doses above 1,200 mg/m², and represents an absolute contraindication to further treatment. Streptozocin is a nitrosourea with unique selective tropism for pancreatic beta cells (islet cells), attributable to its glucose moiety that facilitates uptake via the GLUT2 (glucose transporter 2) transporter. It is used in metastatic pancreatic neuroendocrine tumors (PNETs) and carcinoid tumors. Its dose-limiting toxicity is nephrotoxicity (tubular injury), not myelosuppression, distinguishing it from other nitrosoureas.7
Busulfan is an alkyl sulfonate that alkylates DNA through a different mechanism from the nitrogen mustards, producing predominantly N7-guanine (nitrogen-7 of guanine) monoadducts with eventual cross-linking. Its clinical importance lies almost entirely in its use as a myeloablative conditioning agent before allogeneic and autologous hematopoietic stem cell transplantation (HSCT). High-dose busulfan (total doses of 12 to 16 mg/kg oral or equivalent IV) eradicates residual malignant cells and ablates the recipient immune system and marrow to allow donor stem cell engraftment. Oral busulfan has highly variable bioavailability (40 to 100%), and intravenous busulfan is now preferred for conditioning because it achieves more predictable plasma AUC (area under the concentration-time curve). Busulfan is metabolized by hepatic glutathione S-transferase conjugation and CYP3A4 (cytochrome P450 3A4) hydroxylation. Its most serious non-hematologic toxicity is hepatic sinusoidal obstruction syndrome (SOS), formerly called hepatic veno-occlusive disease (VOD), a life-threatening complication of high-dose busulfan conditioning characterized by tender hepatomegaly, jaundice, ascites, and thrombocytopenia due to narrowing and occlusion of hepatic sinusoids. Therapeutic drug monitoring (TDM) of busulfan plasma AUC is performed during conditioning to ensure target AUC exposure (typically 900 to 1,350 micromolar per minute per day) and reduce the risk of SOS from over-exposure or graft failure from under-exposure. Defibrotide is approved for the treatment of established severe SOS.8
The delayed nadir of nitrosourea-induced myelosuppression at 4 to 6 weeks is one of the most clinically dangerous pharmacological properties in oncology. The standard cycle interval for lomustine (the oral nitrosourea most commonly encountered in outpatient practice for glioma treatment) is every 6 weeks, not every 3 or 4 weeks as with most other regimens. Prescribers unfamiliar with nitrosoureas who apply standard 3-week or 4-week intervals will administer the second cycle before the first cycle's nadir has occurred, generating overlapping suppressions that can produce aplasia requiring hospitalization. Always verify the intended cycle interval before prescribing or dispensing any nitrosourea, and educate patients to monitor for symptoms of myelosuppression in the 4- to 6-week window after each dose.
The triazene and hydrazine alkylating agents occupy specific clinical niches defined by oral bioavailability, CNS (central nervous system) penetration, and tumor-specific sensitivity markers. Temozolomide transformed the treatment of glioblastoma and is the model drug for understanding how DNA (deoxyribonucleic acid) repair enzyme expression predicts alkylating agent response, a paradigm that has influenced the entire field of precision oncology.9
Temozolomide (TMZ) is an orally bioavailable imidazotetrazine prodrug that undergoes spontaneous non-enzymatic hydrolysis at physiological pH to the active methylating species methyltriazenoimidazole carboxamide (MTIC). MTIC further decomposes to the ultimate reactive intermediate, the methyl diazonium ion, which methylates DNA predominantly at O6-guanine (O6-MeG, accounting for approximately 5% of lesions but responsible for the majority of cytotoxic effect), N7-guanine (approximately 70% of lesions, less cytotoxic), and N3-adenine (approximately 9% of lesions). O6-MeG lesions cause cell death by generating persistent G:T mismatches that trigger futile cycles of mismatch repair (MMR) without successful repair, ultimately causing double-strand breaks and apoptosis. TMZ is 100% orally bioavailable, crosses the BBB (blood-brain barrier) effectively due to its small size and lipophilicity, and achieves CSF (cerebrospinal fluid) concentrations of approximately 30 to 40% of simultaneous plasma concentrations. It is used as the standard of care for newly diagnosed glioblastoma in the Stupp protocol (concurrent daily TMZ with radiation, followed by adjuvant TMZ for 6 cycles), for anaplastic oligodendroglioma, and for recurrent high-grade glioma. Renal clearance of TMZ and its metabolites is predominant; the drug is also excreted in bile, but hepatic metabolism is minimal.9
The critical predictive biomarker for temozolomide response is MGMT (O6-methylguanine-DNA methyltransferase) promoter methylation status. MGMT is a DNA repair enzyme that directly reverses O6-guanine alkylation by transferring the methyl group from the damaged guanine to a cysteine residue in its own active site in a stoichiometric, suicide reaction (the enzyme is irreversibly inactivated after each repair event). Tumors that silence MGMT expression through promoter methylation cannot repair O6-MeG lesions and are more than twice as likely to respond to TMZ as MGMT-unmethylated tumors. In the pivotal Stupp trial, patients with MGMT-methylated GBM (glioblastoma multiforme) had a median overall survival of approximately 21.7 months compared with 12.7 months in unmethylated patients treated identically, establishing MGMT promoter methylation as the strongest predictor of TMZ benefit. MGMT methylation testing by polymerase chain reaction (PCR)-based methylation-specific assay or pyrosequencing is now standard in the initial workup of any newly diagnosed glioblastoma. Myelosuppression (predominantly lymphopenia and thrombocytopenia) is the dose-limiting toxicity of TMZ; Pneumocystis jirovecii pneumonia (PJP) prophylaxis with trimethoprim-sulfamethoxazole is mandatory during concurrent chemoradiation and recommended during adjuvant TMZ due to the sustained lymphodepletion.910
Procarbazine is a methylhydrazine compound used in the PCV (procarbazine, lomustine [CCNU], vincristine) regimen for anaplastic oligodendroglioma and anaplastic oligoastrocytoma with 1p/19q co-deletion, and historically in MOPP (mechlorethamine, vincristine, procarbazine, prednisone) for Hodgkin lymphoma. It requires oxidative metabolism by hepatic CYP (cytochrome P450) enzymes and monoamine oxidase (MAO) to generate an azoprocarbazine intermediate and ultimately a methyl radical or methyl carbonium ion that alkylates DNA predominantly at N7-guanine. Because procarbazine undergoes MAO-mediated metabolism, it is a weak MAO inhibitor (MAOI) itself, creating clinically important drug and food interactions. Concomitant use of sympathomimetics, tricyclic antidepressants, selective serotonin reuptake inhibitors (SSRIs), or tyramine-rich foods (aged cheeses, red wine, fermented meats) risks hypertensive crisis or serotonin syndrome. Alcohol consumption with procarbazine causes a disulfiram-like reaction (flushing, headache, nausea, hypotension) due to inhibition of aldehyde dehydrogenase. Procarbazine is teratogenic and has significant secondary leukemia risk from its alkylating action on hematopoietic progenitors; the cumulative leukemia risk in MOPP-treated Hodgkin lymphoma patients was a primary driver of the shift to ABVD (doxorubicin, bleomycin, vinblastine, dacarbazine).2
Dacarbazine (DTIC) is a triazene requiring hepatic CYP1A1 (cytochrome P450 1A1) and CYP1A2 (cytochrome P450 1A2)-mediated hydroxylation to generate the active methylating species MTIC, the same intermediate produced by spontaneous TMZ hydrolysis. Because dacarbazine requires enzymatic activation rather than spontaneous hydrolysis, it is available only as an intravenous formulation and does not achieve meaningful CNS concentrations. It is used in ABVD for Hodgkin lymphoma and in combination regimens for malignant melanoma, soft tissue sarcoma (MAID regimen), and pheochromocytoma. Severe nausea and vomiting are prominent acute toxicities; it is classified as highly emetogenic when given at high doses. Myelosuppression is moderate. Photosensitivity reactions occur during and after dacarbazine infusion and patients should avoid direct sunlight for several days after treatment. Hepatotoxicity, including hepatic vein thrombosis, is a rare but serious adverse effect.2
MGMT promoter methylation is the only validated predictive biomarker for alkylating agent benefit in glioblastoma and should be tested in all newly diagnosed GBM before initiating Stupp protocol. A methylated MGMT promoter means the tumor cannot repair O6-guanine alkylation effectively and is sensitized to TMZ. An unmethylated MGMT promoter means robust O6-MeG repair is intact and TMZ benefit is substantially reduced. In the elderly (age above 65 or poor performance status), MGMT methylation status guides treatment selection: TMZ alone is preferred over chemoradiation in methylated patients (equivalent outcomes, less toxicity), while hypofractionated radiation alone is preferred in unmethylated patients. MGMT methylation status does not predict response to the nitrosoureas (lomustine, carmustine) in the same binary fashion, because nitrosourea DNA lesions are processed differently from TMZ lesions.
Resistance to alkylating agents arises from multiple simultaneous mechanisms operating at the levels of drug uptake, DNA (deoxyribonucleic acid) repair, glutathione-mediated drug inactivation, and apoptosis signaling. Understanding which mechanisms are relevant to each drug class allows prediction of cross-resistance patterns and guides the rational selection of combination partners.11
MGMT (O6 [oxygen-6 position of guanine]-methylguanine-DNA methyltransferase) overexpression is the best-characterized resistance mechanism specific to methylating and chloroethylating alkylating agents. As described in Section 5, MGMT directly reverses O6-guanine (O6-methylguanine) alkylation in a stoichiometric suicide reaction. Tumors with high constitutive MGMT expression (unmethylated MGMT promoter) efficiently repair O6-guanine lesions before they can generate lethal mismatches or interstrand cross-links. MGMT activity is expressed in virtually all normal tissues; its absence in MGMT-promoter-methylated tumors is a tumor-specific vulnerability that is therapeutically exploited by temozolomide, carmustine, and lomustine. MGMT inhibitors such as O6-benzylguanine (a pseudosubstrate that irreversibly inactivates MGMT) have been investigated to sensitize MGMT-expressing tumors to alkylating agents, but this strategy has not translated to improved clinical outcomes in trials, partly because MGMT inhibition in normal tissues (including bone marrow) amplifies myelosuppression to prohibitive levels.11
Nucleotide excision repair (NER) is the primary pathway for removing bulky platinum-DNA adducts and nitrosourea interstrand cross-links. The NER machinery recognizes helix-distorting lesions through the XPC-RAD23B (xeroderma pigmentosum C and RAD23 homolog B) damage recognition complex and recruits the TFIIH (transcription factor IIH) complex, XPA (xeroderma pigmentosum A), RPA (replication protein A), and the endonucleases XPF-ERCC1 (xeroderma pigmentosum F-excision repair cross-complementation group 1) and XPG (xeroderma pigmentosum G) to excise a 24 to 32 nucleotide oligonucleotide containing the damaged base. ERCC1 (excision repair cross-complementation group 1) expression level is inversely correlated with cisplatin sensitivity in non-small cell lung cancer, ovarian cancer, and gastric cancer: tumors with high ERCC1 expression repair platinum adducts more efficiently and are more resistant. The ERCC1-XPF (excision repair cross-complementation group 1 paired with xeroderma pigmentosum F) complex, the primary incision endonuclease of NER, is also essential for resolving interstrand cross-links in conjunction with the Fanconi anemia pathway. Loss-of-function mutations in the Fanconi anemia pathway, including FANCA (Fanconi anemia complementation group A), FANCC (group C), FANCD2 (group D2), and other FA (Fanconi anemia) complementation groups, sensitize cells to ICL (interstrand cross-link)-forming agents (platinum, nitrogen mustards, mitomycin C) and are associated with marked hypersensitivity in patients with Fanconi anemia exposed to these drugs.1112
Glutathione (GSH) and glutathione S-transferase (GST) conjugation is a broad-spectrum detoxification mechanism relevant to all electrophilic alkylating agents. Glutathione is a tripeptide (gamma-glutamylcysteinylglycine) present at millimolar concentrations in most cells. The thiol (-SH) group of the cysteine residue acts as a nucleophile that reacts non-enzymatically and enzymatically (via GST isoenzymes) with electrophilic alkylating agent intermediates to form water-soluble glutathione conjugates that are exported from the cell via the multidrug resistance protein MRP1 (multidrug resistance-associated protein 1, gene designation ABCC1 [ATP (adenosine triphosphate)-binding cassette subfamily C member 1]). Tumors with elevated GSH levels or GST-pi (glutathione S-transferase pi, the predominant isoform in many cancers) overexpression are resistant to nitrogen mustards, platinum compounds, and anthracyclines. Buthionine sulfoximine (BSO), an inhibitor of gamma-glutamylcysteine synthetase that depletes intracellular GSH, has been investigated as a resistance modifier but has not reached clinical application due to normal-tissue toxicity amplification.1112
Clinically significant pharmacokinetic drug interactions with alkylating agents are primarily mediated through CYP (cytochrome P450) enzyme induction or inhibition affecting prodrug activation (cyclophosphamide, ifosfamide, dacarbazine) or through effects on renal drug clearance (cisplatin, carboplatin, methotrexate used with platinum). Rifampin, a potent inducer of CYP2B6 (cytochrome P450 2B6), CYP3A4 (cytochrome P450 3A4), and other CYP isoforms, accelerates cyclophosphamide and ifosfamide activation, increasing formation of active phosphoramide mustard and acrolein and potentially worsening toxicity; conversely, the same induction may deplete plasma prodrug more rapidly, reducing AUC (area under the concentration-time curve) exposure and potentially compromising efficacy. Azole antifungals (fluconazole, voriconazole, itraconazole), which are potent CYP3A4 inhibitors, reduce cyclophosphamide activation, potentially compromising antitumor effect; this interaction is clinically relevant in hematology patients who frequently receive prophylactic azoles during conditioning regimens. Nephrotoxic agents (aminoglycosides, amphotericin B, contrast agents, NSAIDs) administered concurrently with cisplatin compound renal toxicity and reduce GFR (glomerular filtration rate), potentially prolonging carboplatin and methotrexate clearance and increasing toxicity of subsequent drug cycles. Warfarin anticoagulation is destabilized by cyclophosphamide and most other alkylating agents through mechanisms including displacement from protein binding, hepatic CYP competition, and direct effects on clotting factor synthesis from myelosuppression; INR (international normalized ratio) monitoring frequency must be increased during chemotherapy in patients on warfarin.3
Cisplatin and carboplatin produce identical platinum-DNA adducts and share essentially complete cross-resistance: a patient whose ovarian cancer has progressed on carboplatin-based therapy will not respond to cisplatin substitution unless platinum-free interval has allowed partial restoration of platinum sensitivity. Oxaliplatin cross-resistance with cisplatin/carboplatin is partial rather than complete, because the bulky DACH-platinum adducts evade MMR-mediated recognition (relevant in MMR-deficient tumors) and because the structural bulk of DACH adducts may partially evade ERCC1-XPF repair. Clinically, oxaliplatin may retain activity in some cisplatin-resistant settings, particularly colorectal cancer, but it should not be substituted for cisplatin or carboplatin outside its validated indications on the assumption of retained efficacy.
Colvin OM. An overview of cyclophosphamide development and clinical applications. Curr Pharm Des. 1999;5(8):555–560.
Fu D, Calvo JA, Samson LD. Balancing repair and tolerance of DNA damage caused by alkylating agents. Nat Rev Cancer. 2012;12(2):104–120.
doi:10.1038/nrc3185Emadi A, Jones RJ, Brodsky RA. Cyclophosphamide and cancer: golden anniversary. Nat Rev Clin Oncol. 2009;6(11):638–647.
doi:10.1038/nrclinonc.2009.146Sweiss KI, Beri R, Shord SS. Encephalopathy after high-dose ifosfamide: a retrospective cohort study and review of the literature. Drug Saf. 2008;31(11):989–996.
doi:10.2165/00002018-200831110-00003Dasari S, Tchounwou PB. Cisplatin in cancer therapy: molecular mechanisms of action. Eur J Pharmacol. 2014;740:364–378.
doi:10.1016/j.ejphar.2014.07.025Calvert AH, Newell DR, Gumbrell LA, et al. Carboplatin dosage: prospective evaluation of a simple formula based on renal function. J Clin Oncol. 1989;7(11):1748–1756.
doi:10.1200/JCO.1989.7.11.1748Tong WP, Kirk MC, Ludlum DB. Formation of the cross-link 1-[N3-deoxycytidyl],2-[N1-deoxyguanosinyl]ethane in DNA treated with N,N′-bis(2-chloroethyl)-N-nitrosourea. Cancer Res. 1982;42(8):3102–3105.
Mohty M, Malard F, Abecassis M, et al. Sinusoidal obstruction syndrome/veno-occlusive disease: current situation and perspectives — a position statement from the European Society for Blood and Marrow Transplantation (EBMT). Bone Marrow Transplant. 2015;50(6):781–789.
doi:10.1038/bmt.2015.52Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med. 2005;352(10):987–996.
doi:10.1056/NEJMoa043330Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med. 2005;352(10):997–1003.
doi:10.1056/NEJMoa043331Galluzzi L, Senovilla L, Vitale I, et al. Molecular mechanisms of cisplatin resistance. Oncogene. 2012;31(15):1869–1883.
doi:10.1038/onc.2011.384Kelland L. The resurgence of platinum-based cancer chemotherapy. Nat Rev Cancer. 2007;7(8):573–584.
doi:10.1038/nrc2167