The recent decline in mortality observed among patients with cancer reflects not only improved methods for prevention and early detection of malignancy but also better treatments. The most striking therapeutic advances are the result of improved surgical techniques combined with cytotoxic radiation and/or drug therapy.
Historically, the notion of patients deriving potentially significant benefit from cytotoxic therapy was disparaged because of the presumption that cytotoxic therapies could not distinguish between normal and neoplastic tissues. However, the degree and duration of cytotoxic damage to normal and neoplastic tissue differ sufficiently to allow for clinically meaningful benefits in the palliation and cure of cancer. Whether the goal of therapy is palliation or cure determines the degree of risk and cost that the patient and health care system are willing to accept, since virtually all cytotoxic therapies have a narrow therapeutic index. This toxicity to normal organs not only limits the use and achievement of the full therapeutic potential of cytotoxic agents but also claims a cost in terms of patient morbidity and mortality.
Oncologists attempt to mitigate the toxic effects of chemotherapeutic agents by adjusting the dose and frequency of treatment. A major drawback to this strategy is insufficient control of the tumor. A growing area for new drug development is the evolving field of supportive care agents intended to avert or minimize treatment-limiting toxicities to normal organs.
Historically, the oldest approach of this type was the use of rescue agents, such as leucovorin rescue following high-dose methotrexate(Drug information on methotrexate). This approach has been useful for the treatment of pharmacologic sanctuaries, such as the central nervous system (CNS), as well as tumors that are relatively resistant to standard doses of methotrexate, such as osteosarcoma. Likewise, the use of marrow-stimulatory cytokines following moderately toxic doses of drugs has been useful.
An alternate approach to rescue techniques is the use of cytoprotective agents, or cytoprotectors. These drugs selectively protect normal tissues from the cytotoxic effects of drugs and/or radiation while preserving their antitumor effects. The ideal cytoprotector should be easily administered and should have a broad spectrum of activity; not only should it be able to protect multiple tissues, but also it should protect these tissues from diverse cytotoxic treatments. The agent also should have a reasonable safety profile. By improving patient tolerability to the therapy, the cytoprotector may provide multiple benefits: It may improve patient quality of life; allow for dose-inten-sification and, thus, improved response and/or cure rates; and decrease costs associated with other supportive care measures needed to treat treatment complications.
To date, three cytoprotectors have been approved by international regulatory agencies. Mesna (Mesnex) is approved for the protection of the bladder from the toxicity associated with ifosfamide (Ifex) and high-dose cyclophosphamide(Drug information on cyclophosphamide) (Cytoxan, Neosar). Dexrazoxane (Zinecard) is used to protect the heart from the cardiomyopathy associated with the cumulative cardiotoxic effects of doxorubicin(Drug information on doxorubicin). Both of these agents are drug- or drug classspecific and their protective effect is limited to specific organsthe urinary bladder for mesna(Drug information on mesna) and the heart for dexrazoxane.
Amifostine(Drug information on amifostine) (WR-2721 [Ethyol]) is the first broad-spectrum cytoprotective agent to be approved by international regulatory agencies. In this context, broad spectrum refers to cytoprotection against a broad array of cytotoxic therapies, ie, multiple drug classes as well as radiation, in multiple organ systems. Laboratory and clinical studies have shown that amifostine protects cells in virtually all organ systems except the CNS, which is a target for both acute and cumulative (or delayed) cytotoxic damage. Because of poor tissue distribution pharmacokinetics in the CNS, this organ system has not been effectively studied.
This article will review preclinical and clinical studies of amifostine, as well as trials evaluating its chemoprotective and radioprotective effects. Studies suggesting that amifostine pretreatment may also enhance the antitumor effects of chemotherapy in certain settings and prevent secondary malignancies are also detailed.
Amifostine was originally synthesized under a classified project of the US Army aimed at developing agents to protect the military from the effects of radiation from a nuclear warhead. (Its military code name is WR [Walter Reed]-2721.) Amifostine was selected for further development from over 4,000 compounds because of its superior radioprotective and safety profile. The synthesis of this series of sulfhydryl-containing compounds was based on previous observations showing that the sulfhydryl-containing amino acid, cysteine(Drug information on cysteine) and, subsequently, cysteamine could protect rodents from lethal doses of radiation.
Amifostine (Figure 1), an analog of cysteamine, is a phosphorylated aminothiol prodrug that is dephosphorylated in the tissues by membrane-bound alkaline phosphatase to its active metabolite, WR-1065. A free thiol, WR-1065 is the form of the drug that is taken up into cells and is the major cytoprotective metabolite. Oxidation of WR-1065 forms the symmetrical disulfide, WR-33278, which not only is structurally similar to spermine, a naturally occurring polyamine, but also shares certain biochemical properties with the polyamines that may contribute to some of its pharmacologic properties.
The selective cytoprotection of normal tissue by amifostine stems from its unique systemic and tissue distribution pharmacokinetics. Following intravenous administration of the drug, the half-lives of both the distribution and elimination phases in humans are extremely short (alpha-half-life, < 1 minute; beta-half-life, 8.8 minutes). Most (90%) of the drug is cleared from the plasma within 6 minutes, and only a small amount of the prodrug is bio-converted to the free thiol in the systemic circulation relative to normal tissues.[6-9]
Generation of the free thiol, WR-1065, occurs primarily at the tissue site due to local dephosporylation by membrane-bound alkaline phosphatase; this enzyme has relatively high specific activity in the endothelia of normal capillaries and membranes of normal cells but is relatively deficient in the neovascular endothelia and membranes of cancer cells.[10-12] Once inside the cells, WR-1065, the free thiol, provides protection from cytotoxic damage by acting as a scavenger of oxygen-free radicals generated by radiation and anthracyclines, and by binding to highly reactive nucleophiles that would otherwise cross-link and damage DNA.[15-17]
Recent studies suggest that amifostine or WR-1065 can up-regulate gene expression of p53; this results in cell-cycle accumulation at the G1-S interface, which, in turn, may allow for more efficient DNA repair. Extensive preclinical studies, in both cell cultures and tumor-bearing rodents, have shown that amifostine selectively protects normal tissues without protecting a wide variety of murine and human carcinomas, sarcomas, and leukemias.
Based on this preclinical profile, amifostine has been studied in a number of clinical trials. The combined laboratory and clinical experience to date has shown protection of the organs listed in Table 1 from cytotoxic damage. As a result of these studies, an expanded pharmacologic profile of amifostine is emerging (Table 2). The following summary will provide a brief overview of these areas.
Following extensive controlled laboratory investigations and a series of phase I and II clinical trials, it was evident that pretreatment with amifostine could protect normal tissues from hematologic toxicity associated with cyclophosphamide (Cytoxan, Neosar) or high-dose cisplatin(Drug information on cisplatin) (Platinol)[20,21] and from nonhematologic toxicities associated with cisplatin, and that it could achieve this cytoprotection without any negative effects on the antitumor efficacy of these drugs.[20,22-24] Amifostine-mediated cytoprotection against these hematologic and nonhematologic toxicities was evaluated in a multicenter, multinational, phase III clinical trial that enrolled women with stage III/IV ovarian cancer (Figure 2). Following surgery for tumor debulking and staging, patients were stratified based on residual tumor and cancer center and were randomly assigned to receive six cycles of cyclophosphamide plus cisplatin with or without amifostine, administered at 3-week intervals.
Prior to receiving any chemotherapy, all patients were hydrated with 5% dextrose(Drug information on dextrose) in 0.45% saline (200 mL/h) for 6 hours. Patients were to receive drug therapy only when urine output exceeded 150 mL/h for 3 hours. Patients also received mannitol(Drug information on mannitol) diuresis. Amifostine, 910 mg/m², was administered as a 15-minute intravenous infusion to patients in a supine position. Within 15 minutes of the completion of the amifostine infusion, cyclophosphamide, 1,000 mg/m², was administered over 20 minutes. Immediately following the cyclophosphamide infusion, a 30-minute infusion of cisplatin, 100 mg/m², was administered.
A total of 242 patients were enrolled in this ovarian cancer trial; 122 patients were randomized to receive amifostine and cyclophosphamide-cisplatin chemotherapy, and 120 patients were randomized to receive chemotherapy alone. The two treatment arms were well-matched with respect to age, race, International Federation of Obstetrics and Gynecology (FIGO) stage, extent of residual disease, and performance status. The clinical end points of the trial included assessment of the expected toxicitieshematologic, renal, peripheral neuropathy, and ototoxicity, as well as antitumor efficacy.
The major hematologic end point was the cumulative incidence of neutropenic events through the six cycles of therapy. Neutropenic events were defined as grade 4 neutropenia with fever and/or signs and symptoms of infection that usually require hospitalization, clinical and laboratory evaluation, and institution of empiric broad-spectrum antibiotic therapy.
Figure 3 shows the cumulative incidence of neutropenic events in both treatment arms. Pretreatment with amifostine resulted in a significant reduction in this potentially life-threatening event. As shown in Table 3, amifostine pretreatment significantly decreased all parameters of hematologic toxicity. Most notably, it reduced both the number of days the patient spent in the hospital and the cumulative number of days of treatment with broad-spectrum antibiotics as a consequence of neutropenic events.
The protocol-defined renal end point to assess the ability of amifostine to protect against cisplatin-induced nephrotoxicity was the need to delay or discontinue cisplatin therapy because of elevations in serum creatinine > 1.5 mg/dL. If serum creatinine level was > 1.5 mg/dL at day 22, cisplatin was to be delayed for a maximum of 2 weeks; if it remained at a level exceeding 1.5 mg/dL at day 35 (defined as a protracted elevation of serum creatinine), cisplatin was to be discontinued.
Figure 4 displays the proportion of patients who met these criteria for a delay in or discontinuation of cisplatin therapy. Consistent with the cumulative nature of cisplatin-induced nephrotoxicity, by cycles 5 and 6 a significantly greater proportion of patients in the control arm compared to the amifostine arm could not receive cisplatin as scheduled because of elevated serum creatinine levels. These differences were statistically significant despite the fact that patients in both groups received comparable doses of cisplatin (median cumulative doses, 555 and 500 mg/m² for the amifostine and control groups, respectively).
Since, in most hospitals, the upper limit of normal for serum creatinine in women is 1.0 mg/dL, a rise in serum creatinine to 1.5 mg/dL would represent a substantial deterioration in renal function. Estimations of creatinine clearance utilizing serum creatinine, age, weight, and gender were calculated utilizing the formula of Cockcroft and Gault. As shown in Table 4, by the end of therapy, 33% of patients treated with chemotherapy alone had a ³ 40% reduction in creatinine clearance. This was reduced to 10% in the amifostine-treated patients (P = .001).
The 33% incidence of a ³ 40% reduction in creatinine clearance in the control arm of the ovarian cancer trial is consistent with the incidence observed in women treated with up to six cycles of the same regimen in a trial conducted by the Southwest Oncology Group (SWOG). In this trial, after the last cycle of therapy, 40% of patients (51/127) sustained a ³ 40% reduction in creatinine clearance from baseline. Similarly, in two other trials that used single-agent cisplatin (100 mg/m2 for 5 monthly cycles) in previously untreated patients with advanced ovarian cancer, a > 40% decrease in glomerular filtration rate, as measured by endogenous creatinine clearance or ethylene-dinitrilo tetraacetic acid (EDTA) clearance, was noted in 35% and 45% of patients, respectively.[28,29]
Other reports have shown comparable cumulative nephrotoxic effects of cisplatin, as measured by progressive increases in serum creatinine values and progressive decreases in creatinine clearance measurements through five or six cycles of cisplatin (80 to 100 mg/m²).[27,30-40] In contrast, in phase II clinical trials using monthly courses of amifostine (740 or 910 mg/m²) prior to cisplatin (120 mg/m²) in patients with non-small-cell lung cancer (NSCLC), melanoma, breast, or head and neck cancer, the frequency of a ³ 40% reduction in creatinine clearance from baseline was 8%, which is consistent with the 10% incidence observed in the amifostine arm of the ovarian cancer trial (Table 5).[20,22,23,27]
The literature shows that the cumulative nature of cisplatin nephrotoxicity is generally permanent. The data from the ovarian cancer trial support this conclusion. The reduction in glomerular filtration rate at the end of cisplatin therapy observed in the control arm of that trial persisted through a 30-month follow-up period (Figure 5). This loss in renal reserve will have an impact on further antineoplastic and other therapies that require renal elimination or that have intrinsic renal toxicitya clinical concern that also pertains to the use of certain radiographic contrast agents.
In the ovarian cancer trial, clinical neurotoxicity was scored in accordance with the National Cancer Institute (NCI) grading system. The higher grades include troublesome paresthesias and inability to perform fine finger motions. Treatment with amifostine resulted in a significant reduction in the severity of clinical neurotoxicity (Table 6).
This is consistent with amifostine neuroprotection reported by Mollman et al. They found a 40% reduction in ototoxicity (defined as troublesome tinnitus and clinical hearing loss requiring a dose reduction or discontinuation of cisplatin) in the amifostine-treated patients compared to the control patients, although the difference was not statistically significant, as only 7% of control patients experienced this toxicity.