Drug treatment for cancer began with the use of the mustard class of alkylating agents, initially mechlorethamine (nitrogen mustard), which entered into clinical use in the mid-1940s. Alkylating agents are used in many regimens but are rapidly being supplanted by newer classes of agents.
All alkylating agents undergo molecular rearrangements to form covalent bonds to DNA bases. Some form monoadducts, while others form cross-links either with intrastrand or interstrand bases, or with adjacent proteins. All alkylating agents are cytotoxic to tumor cells through interruption of DNA replication, induction of DNA damage repair and stress response, or through checkpoints in the cell cycle that result in apoptosis, senescence, or necrosis. All alkylating agents induce DNA strand breaks directly or through the damage response. Alkylating agents can also cause DNA mutations that can induce a cytotoxic response or can alter proteins, leading to immune responses to novel protein sequences.
Mechanisms of Resistance
Efficient removal of the DNA adducts reduces the lesion burden, while loss of DNA damage recognition can invoke damage tolerance, as in the case of loss of MMR and temozolomide tolerance. Detoxifying enzymes (glucuronidating enzymes and metabolizing enzymes residing in the liver) can serve as acceptor molecules for alkylation, alter the agent prior to DNA attack, or metabolize the parent compound. Loss of TP53 results in loss of cell cycle checkpoint induction of DNA repair signals. Increased AKT signaling promotes cell proliferation to compensate for the DNA damage response-associated toxicity.
Nitrogen Mustard
The nitrogen mustard class includes mechlorethamine, cyclophosphamide, 4-hydroperoxy cyclophosphamide, ifosfamide, chlorambucil, and melphalan. These drugs all share a common bischloroethyl group attached to nitrogen and a substituted “R” group that provides drug specificity (Fig. 1). All nitrogen mustards react with DNA in an SN2 reaction, a bimolecular nucleophilic displacement reaction also called a second-order reaction. Although numerous sites are targeted for alkylation, nucleophiles in DNA, including nitrogen (N), oxygen (O), and phosphate (P), attract the chloroethyl moiety attached to the R–N backbone, and the chlorine is displaced by the nucleophilic atom to form an aziridinium moiety. The remaining chloroethyl group is then attracted to a second nucleophilic atom, forming a second aziridinium intermediate, leading to a second alkylation, forming a cross link. Both intrastrand and interstrand cross-links are formed. The N7 guanine position is the most critical for cytotoxic cross-link formation. Clinical use of mechlorethamine is limited because the MOPP regimen (mechlorethamine, vincristine [Oncovin], procarbazine, prednisone) has been replaced by ABVD (doxorubicin [Adriamycin], bleomycin, vinblastine, dacarbazine [DTIC]) in Hodgkin lymphoma. Local use of mechlorethamine occurs in a dermatologic suspension for the treatment of cutaneous T-cell lymphomas (CTCLs).
Fig1. STRUCTURE OF COMMON ALKYLATING AGENTS.
Cyclophosphamide is metabolically activated by cytochrome P450 mixed function oxidases in the liver to 4-hydroxycyclophosphamide. 4-Hydroxycyclophosphamide is further converted to aldophosphamide and then to phosphoramide mustard, the alkylated species, and acrolein. High levels of aldehyde dehydrogenase detoxify cyclophosphamide in hematopoietic stem cells and, thus, high doses are not marrow ablative. Acrolein is a highly reactive aldehyde and the cause of hemorrhagic cystitis. Mercaptoethane sulfonate (Mesna) is used to provide prophylaxis against hemorrhagic cystitis caused by cyclophosphamide and ifosfamide, and is now standard for doses of cyclophosphamide and ifosfamide above 1000 mg /m 2. Mesna is given in divided doses every 4 hours or as a continuous infusion for 18 to 24 hours in a dose equivalent to either cyclophosphamide or ifosfamide. Other than hemorrhagic cystitis, BM suppression is dose limiting and can be rescued by reinfusion of autologous or allogeneic hematopoietic progenitor cells. Other toxicities include alopecia and cardiac toxicity, which is unusual and most often seen after high-dose therapy.
4-Hydroperoxycyclophosphamide, a chemically stable form of the reactive intermediate of cyclophosphamide, 4-hydroxycyclophosphamide, is more toxic to committed hematopoietic progenitors such as colony-forming unit–granulocyte/macrophages (CFU-GM), burst forming unit–erythroids (BFU-E), and colony-forming unit–erythroids (CFU-E).
Melphalan, or phenylalanine mustard, has an amino acid side chain that alters its cellular uptake and stabilizes its structure, allowing PO administration (see Fig.1). It is available in both PO and IV forms, and has a similar effect on DNA cross-linking as cyclophosphamide and the other nitrogen mustards. Melphalan uptake by cells is by means of a neutral amino acid transporter. Its rate of cross-link formation is much slower than that of mechlorethamine, presumably because of delayed metabolism. Oral melphalan is used predominantly for the standard treatment of multiple myeloma (MM) and IV in high-dose regimens in preparation for stem cell transplantation for MM patients.
Chlorambucil has been used for more than 40 years for the treatment of CLL. Chlorambucil is the phenylbutyric acid derivative of nitrogen mustard and is very stable, entering the cell by diffusion rather than by a specific uptake mechanism. It is typically administered orally on a daily basis or intermittently. It appears to have greater bioavailability than melphalan and a more consistent half-life of approximately 2 hours.
Busulfan is an alkylsulfonate unique among alkylating agents because of two sulfur groups and lack of a chloroethyl moiety (see Fig. 1). Busulfan, similar to the nitrogen mustards, reacts predominantly at the N 7 position of guanine and produces an N 7–N 7 biguanyl DNA cross-link, although the precise nature of this cross-link appears different than that of the nitrogen mustards. The pharmacokinetics of busulfan is important for its use in high-dose therapy for ablation of the BM in patients undergoing autologous transplantation for acute leukemia or allogeneic stem cell transplantation. Because the incidence of veno-occlusive disease is lower in patients receiving high-dose busulfan with predosing pharmacokinetics performed, this is now recommended in high-dose regimens, the target being an area under the curve (AUC) of 1125 μ mol/L × min (range 900 to 1350) with every 6-hour dosing. Busulfan is a potent stem cell toxin, killing both early and late hematopoietic progenitor cells and damaging the BM stroma. Other toxicities of busulfan include nausea and vomiting, and pulmonary interstitial and intra-alveolar edema leading to fibrosis. The pulmonary fibrosis is distinct from the interstitial pneumonitis, which accompanies allogeneic stem cell transplantation and is not related to cytomegalovirus (CMV) or other viral infections.
Nitrosoureas
Four chloroethyl nitrosoureas and one methyl nitrosourea are in clinical use. These agents are different from the nitrogen mustards in that they alkylate through an SN 1 reaction, forming a highly reactive intermediate in the presence of N, O, and P nucleophiles in DNA. The commonly used clinical agent is (2-chloroethyl)- N-nitrosourea (BCNU). N-([4-amino-2-methyl-5-pyrimidinyl] methyl)- N-(2 chloroethyl)- N-nitrosourea (ACNU) is commonly used in Japan. A third agent, N-(2-chloroethyl)- N-cyclohexyl- N-nitrosourea (CCNU), is used predominantly as an PO nitrosourea in children with brain tumors. All of these compounds have high hydrophobicity, actively penetrating the blood–brain barrier.
The DNA alkylation sites include N 7 and O 6 of guanine. Chloroethylation at the O 6 position of guanine appears critical to cytotoxicity. DNA cross-linking by chloroethyl nitrosoureas include at 1-(3-cytosinyl), 2-(1-guanyl) ethane, and 1-2-bis(7-guanyl) ethane. The former is responsible for much of the cytotoxicity observed with the chloroethyl nitrosoureas. It is formed after alkylation at the O 6 position of guanine. This adduct undergoes intramolecular rearrangement to a circular intermediate, N1. O6 ethanoguanine is formed, which can then rearrange by attack at the opposite hydrogen-bonded base N 3 of cystine, forming the interstrand cross-link. This is a unique DNA cross-link and is poorly recognized by DNA repair processes, leading to marked cytotoxic potency of this cross-link.
The pharmacokinetics of the chloroethyl nitrosoureas reveal a very short half-life. BCNU is predominantly used for high-dose treatment of recurrent lymphomas. Regimens including BCNU induce sustained complete remission (CR) rates of approximately 40% to 60%. Doses of between 300 and 600 mg/m 2 have been safely administered. The chloroethyl nitrosoureas cause profound and cumulative BM suppression at conventional doses of 120 to 150 mg/m 2, limiting treatment to three to five cycles at 6-week intervals.
Complications of high-dose BCNU therapy include pulmonary toxicity and renal toxicity at doses higher than 600 mg/m 2. Pulmonary toxicity, as evidenced by a decrease in the DL CO (diffusing capacity of the lung for carbon monoxide), occurs in up to 40% of patients. It can be managed with high-dose oral steroids during the inflammatory phase. Interstitial nephritis with glomerulosclerosis, interstitial fibro sis, and dropout of tubules has been reported with BCNU or CCNU.
Methylating Agents
Four methylating alkylating agents are in clinical use. These include procarbazine, DTIC, streptozotocin, and temozolomide. Procarbazine and DTIC are triazines. Streptozotocin is a mono functional methyl nitrosourea derivative with an attached sugar moiety, and temozolomide is an imidazotetrazine. All react with DNA by undergoing SN 1 reactions forming a methyldiazonium ion, resulting in methylation of N 7 guanine (67%), O 6 guanine (9%), O 4 and O 2 thymine (2%), and N 3 adenine (3%). None form DNA cross-links. However, all induce high levels of DNA methylation, and their recognition and repair results in both single- and double-strand breaks. N 7 methylguanine is removed through the BER system. Recognition by the N-methylpurine glycosylase results in removal of the adducted base with formation of an abasic site that is recognized by the apurinic (AP) endonuclease, which then cleaves the backbone at the AP site. Subsequently, the free 5 ′ sugar is released by DNA lyase, with repair initiated by β -polymerase and DNA ligase. BER effectively removes N7 methylguanine and N3 methyladenine, and restores DNA to normal. Inhibition of BER by the investigational agent methoxyamine (TRC102) blocks this pathway and increases toxicity.
O6 methylguanine mispairs with thymine during DNA synthesis, resulting in a lesion recognized by the MMR system. Mispair recognition proteins are MSH6, MSH3, and MSH2. Recognition of the mispair recruits additional proteins to the complex, including MLH1 and PMS1/PMS2. These proteins initiate exonuclease cleavage of a long patch in the newly synthesized strand of DNA. This is then repaired by polymerases- δ and - ε . If unrepaired, a thy mine is repeatedly inserted opposite the O 6 methylguanine, resulting in multiple single-strand breaks. Cells expressing high levels of the DNA repair protein for O 6 methylguanine, O 6 methylguanine DNA methyltransferase (MGMT), are approximately 10-fold more resistant to methylating agents than MGMT-negative cells. Cells lacking MMR are very resistant to methylating agents. Acquisition of MMR defects is associated with acquired resistance to methylating agents and cisplatin, which is also recognized by this protein complex.
Procarbazine was synthesized as a monoamine oxidase inhibitor and has been used since the 1950s for the treatment of Hodgkin lymphoma and NHL, as well as a component of combination therapies for gliomas. DTIC is metabolically activated by cytochrome P450 microsomal oxidoreductases, ultimately leading to formation of the methyldiazonium ion and DNA methylation. DTIC is used in combination with ABVD for treating Hodgkin lymphoma and is also used for patients with metastatic malignant melanoma in combination with BCNU, cisplatin, and tamoxifen. Activation of DTIC requires hydroxylation of one terminal methyl group caused by demethylation forming 5-[3-methyl triazen-1-yl]-imidazole-4-carboxamide (MTIC), with spontaneous decomposition to the methyldiazonium ion, which alkylates the DNA, as noted earlier. Maximum tolerated doses of DTIC are approximately 1000 mg/m2, with myelosuppression and gastrointestinal toxicity (including severe watery diarrhea) being the most common side effects.
Temozolomide represents an imidazotetrazinone. It differs from DTIC in that it is chemically degraded to the monomethyl tri azine, MTIC, at neutral pH and does not require P450 enzymatic demethylation. Compared with DTIC, temozolomide has much more consistent pharmacokinetic parameters, including peak serum concentrations, volume of distribution and clearance, and conversion to MTIC. Clinical studies documented considerable activity in acute leukemias. The dose-limiting toxicity was thrombocytopenia and, less frequently, neutropenia, with maximum tolerated doses of 1000 mg/ m 2 given over 5 days on a daily or twice-daily regimen. Nausea and vomiting were the other common side effects, easily controlled with antiemetics.
Bendamustine
Bendamustine is comprised of a 2-chloroethylamine nitrogen mustard alkylating group, a benzimidazole ring, and a butyric acid side chain. Its mechanism of action is unknown but appears to be different than other alkylating agents, causing DNA damage that is repaired pre dominantly by the base-excision repair system and has activity against lymphoid cell lines resistant to alkylating agents. Myelosuppression with leukopenia and thrombocytopenia is the most frequent side effect, with mild nonhematologic side effects including nausea, fatigue, constipation, and diarrhea. Phase III studies comparing standard treatment with bendamustine and rituximab in front-line treatment of CLL and indolent lymphomas have established this combination as a first-line treatment. Bendamustine is also active against MM, and combinations with steroids and bort ezomib or the immunomodulatory agents thalidomide and lenalidomide have been reported to result in high response rates in patients with relapsed or refractory disease.
Alkylating Agent–Induced Leukemias
Alkylating agents induce dose-limiting myelosuppression and cause sublethal DNA damage to hematopoietic progenitors, causing mutational events that lead to malignant transformation to preleukemic and leukemic states. A concern exists about the use of hematopoietic growth factors after exposure to alkylating agents. There is evidence of increased cytotoxicity to hematopoietic pro genitors during simultaneous exposure to these agents and growth factors. Treatment-related AML (t-AML) accounts for approximately 15% of all adult AML. Approximately 50% of t-AML patients have a preleukemic phase compared with only 10% of patients with de novo AML. CRs are achieved in 15% to 30% of patients with t-AML and a mean remission duration of 2 months. Chromosomal abnormalities and gene mutations characteristic of t-AML establish this as a distinct disease requiring novel therapeutic approaches. Loss or deletion of all or part of the long arm [q] of chromosomes 5 or 7 is common, as are trisomy 8 and deletions of the short arm of chromosomes 12, 17, and 21.
Historically, patients with Hodgkin lymphoma treated with mechlorethamine and procarbazine in the MOPP regimen or with CCNU were at the highest risk if exposed to radiation as well as an alkylating agent combination. Patients with polycythemia vera treated with chlorambucil were at much higher risk than patients treated with phlebotomy alone, which can contribute to a shift in treatment strategy. Patients with myeloma and ovarian cancer have developed t-AML, especially after prolonged exposure to alkylating agents. Patients treated with alkylating agents for benign diseases such as nephritis, lupus, psoriasis, rheumatoid arthritis, and Wegener granulomatosis also have an increased risk of t-AML. The mean latency between exposure and t-AML from alkylating agents is 4 to 5 years, in contrast to t-AML from etoposide, which has a latency period as short as 1 year. The cumulative risk of developing t-AML is between 10% and 17% at 4 to 6 years for myeloma patients treated with melphalan and between 2% and 10% at 7 to 10 years in patients with Hodgkin lymphoma. Alkylating agent-associated t-AML has also been recognized in patients with breast and colon cancer.
