|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 1999 American Society for Clinical Oncology Randomized Phase II Study of High-Dose Paclitaxel With or Without Amifostine in Patients With Metastatic Breast CancerFrom the British Columbia Cancer Agency, Vancouver; National Cancer Institute of Canada Clinical Trials Group, Kingston; Mount Sinai Hospital and Eli Lilly Company, Toronto; Ottawa Regional Cancer Centre, Ottawa; and McGill University, Montreal, Canada. Address reprint requests to Karen A. Gelmon, MD, BCCA Vancouver Cancer Centre, 600 West 10th Ave, Vancouver, British Columbia, Canada V5Z 4E6; email kgelmon{at}bccancer.bc.ca
PURPOSE: To determine whether the neurotoxicity of paclitaxel 250 mg/m2 given over 3 hours every 3 weeks could be reduced by pretreatment with amifostine 910 mg/m2. Secondary objectives included comparing myelosuppression, myalgias, and response rates of the two groups. PATIENTS AND METHODS: Forty womenwith metastatic breast cancer were randomized to receive either paclitaxel alone (arm 1) or paclitaxel preceded by amifostine (arm 2). All were assessable for toxicity, and 37 were assessable for response. At baseline and after each cycle, all patients completed questionnaires for neurologic symptoms and had standardized neurologic examinations, including objective assessments of power and vibration sense. In addition, standard follow-up assessments for other toxicities and tumor response were undertaken. Changes from baseline after courses 1, 2, and 3 were assessed. The sample size was sufficient to detect a 50% improvement in the expected determination in sensory change. RESULTS: There were no differences observed in any of the measures of neurotoxicity. Other toxicity was similar in arms 1 and 2, including hair loss (95% v 90%), neurosensory changes (100% v 100%), fatigue/lethargy (85% v 90%), myalgia (95% v 90%), and grade 4 neutropenia (47% v 60%). Nausea, vomiting, dizziness, hypotension, and sneezing were more common in the amifostine arm. Response rates (22.2% v 36.8%) and paclitaxel pharmacokinetics were not significantly different. CONCLUSION: There was no protection from paclitaxel-related neurotoxicity or hematologic toxicity in this study. These results suggest that the mechanism of action of paclitaxel-related toxic effects is not amenable to the cytoprotective action of amifostine.
SINCE THE INITIAL studies of paclitaxel, there has been interest in defining the optimal dose and schedule for delivering the drug in single-arm and some comparative clinical studies.1,2 Although a debate about schedule and dose effects on antitumor activity continues, it is clear that these parameters are important in determining toxicity. When short infusions of paclitaxel are given with cytokine support, severe neurotoxicity replaces myelosuppression as the dose-limiting toxicity.3,4 Combinations with other neurotoxic drugs, particularly cisplatin, seem to exacerbate the toxicity.5 Paclitaxel induces a variety of neurotoxic effects.6-8 The most common is a peripheral sensory neuropathy, although motor and autonomic neuropathies can occur. As well, complaints of myalgias and generalized muscle weakness can be seen and are reported more commonly with higher doses and shorter infusion schedules, and they may limit the long-term use of the drug.9 Postma et al10 reported the results of neurologic assessments in 27 patients who received 3-hour paclitaxel infusions at three different starting dose levels, 135, 175, and 250 to 300 mg/m2. They reported neuropathic symptoms in 50%, 79%, and 100%, respectively, and neuropathic signs in 83%, 86%, and 100%, respectively. Dose-limiting neurotoxicity occurred in 0%, 21%, and 71%, respectively.10 Amifostine (Ethyol; U.S. Bioscience, Inc, West Conshohocken, PA) is an organic thiophosphate compound, which in animal models selectively protects normal tissue against the toxicity of both radiation and a variety of chemotherapeutic agents.11 A prodrug, amifostine is dephosphorylated at the tissue site to the active metabolite, WR-1065. In the cell it is thought to act by binding to and detoxifying the reactive molecules of alkylating or platinum agents and acting as a scavenger for the oxygen free radicals produced by ionizing radiation and certain chemotherapeutic agents.12,13 Although the cellular target of paclitaxel is tubulin, not DNA, preclinical studies have shown that amifostine may offer selective cytoprotection of normal tissues from paclitaxel as well. In vitro studies compared the selective cytoprotection of amifostine on paclitaxel cytotoxicity in normal human lung fibroblasts (MRC-5) versus two non–small-cell lung cancer cell lines (A547 and A549).14,15 Amifostine pretreatment at two different concentrations was able to protect normal human lung fibroblasts but not non–small-cell lung carcinoma cells from the cytotoxic effects of 100 nmol/L paclitaxel. In addition, amifostine reduced the number of DNA single-strand breaks in MRC cells induced by paclitaxel but did not decrease the number of single-strand breaks in A427 cells. In an in vivo system, amifostine plus paclitaxel was shown to have greater antitumor effects than paclitaxel alone when tested against a murine ovarian cancer model.15,16 These findings suggested a possible selective cytoprotective utility of amifostine pretreatment in paclitaxel-containing regimens. The expectation that these laboratory findings could be extrapolated to the clinic was given credence by the experience with cisplatin where in vivo findings,17,18 were confirmed in phase II and phase III clinical trials. A lower incidence of neurotoxicity (25% with amifostine v 49% without) and a higher mean dose at which the onset of neuropathy occurred (383 mg/m2 v 635 mg/m2) were reported when cisplatin was given with amifostine as compared with results for those patients treated without the protection.19 A randomized, controlled phase III trial of cisplatin and cyclophosphamide plus or minus amifostine in 242 women with advanced ovarian cancer reported that pretreatment with amifostine protected against the acute and cumulative hematologic and nonhematologic toxicities associated with cisplatin, with no loss of antitumor efficacy and no difference in survival.20 A recent phase I trial of paclitaxel (3-hour infusion) preceded by amifostine was able to reach doses higher than the previously reported maximum-tolerated dose, which suggests that amifostine might offer protection from paclitaxel's toxic effects as well.21 To determine the possible effect of the addition of amifostine to high-dose paclitaxel, the National Cancer Institute of Canada Clinical Trials Group (NCIC-CTG) initiated a randomized, nonblinded multicenter study of high-dose paclitaxel with or without amifostine in patients with metastatic breast cancer. The primary objective of the study was to assess the efficacy of amifostine in preventing or reducing neurotoxicity by high-dose (250 mg/m2) paclitaxel delivered over 3 hours. Our secondary objectives were to document the effect of amifostine on the incidence and severity of myelosuppression, myalgias, and other toxic effects induced by the chemotherapy and to determine the response rate of paclitaxel plus amifostine in this disease.
Eligibility Criteria Patients with histologically proven metastatic breast cancer were eligible for study entry. Patients could have received up to one adjuvant chemotherapy regimen and up to one regimen for metastatic disease. None of the previous chemotherapy programs could contain paclitaxel, vincristine, cisplatin, or other neurotoxic drugs. Previous radiotherapy or hormonal therapy was permitted: radiotherapy had to have involved less than 40% of the bone marrow and had to have been terminated at least 4 weeks before randomization. Patients were nonpregnant, nonlactating women 16 years of age or older with an Eastern Cooperative Oncology Group performance status of 0 to 2 and a life expectancy of least 4 months. Adequate hematologic, renal, and hepatic functions were required and defined as a granulocyte count greater than 1.5 x 109/L, platelet count greater than 150 x 109/L, serum creatinine level within normal limits, bilirubin level less than 1.5 times upper normal limit, and AST  two times the upper normal limit in patients without liver metastses and five times the upper normal limit in patients with documented liver metastases. All patients had to be accessible for follow-up evaluation and management of complications at a participating center.
Patients were required to have clinically or radiologically bidimensional measurable disease with a minimum indicator lesion of
Patients were excluded if any of the following conditions were present: a history of malignancy other than the entry diagnosis (except for nonmelanomatous skin cancer or curatively treated cervical carcinoma in situ), pre-existing neuropathies or prior treatment with a neurotoxic chemotherapy, a history or symptoms and signs of hypotension, syncope or cardiac arrythmias, diabetes mellitus, any antihypertensive therapy (with the exception of diuretics), and newly documented brain metastases. Previously documented intracranial disease which was stable for All of the patients who entered the trial gave written informed consent. The protocol was approved by the local ethics committees. Eli Lilly Canada Inc, which licenses amifostine from U.S. Bioscience, provided the amifostine and contributed to the standard paclitaxel dose. Bristol-Myers Squibb, Canada, contributed the excess dose above the standard of 175 mg/m2.
Pretreatment Evaluation The cumulative symptom score (CSSy) is a patient self-administered neuropathy questionnaire that combines paresthesia, numbness, and pain of hand and feet together with 13 physical activities of daily life into an overall summary score with a range of 0 to 99. Each question is scored from 0 for no complaints to 10 for intolerable symptoms. All patients filled out the questionnaire at each visit as part of their neurologic assessment. This questionnaire has been used in trials to assess neurologic symptoms and in particular in a study of paclitaxel, where it was found to be a sensitive indicator of differences in neurosensory symptoms related to dose.10 This tool has recently been adopted by the European Organization for Research and Treatment of Cancer as a standard measure of toxicity to link with quality-of-life parameters.
Treatment Paclitaxel was administered intravenously in 500 mL of normal saline over 3 hours. To avoid hypersensitivity reactions, the following standard premedication package was administered: dexamethasone 200 mg orally 12 and 4 to 6 hours before paclitaxel infusion, and then 30 minutes before the infusion of diphenhydramine 50 mg intravenously and cimetidine 300 mg intravenously over 10 minutes. In patients randomized to the amifostine arm, the premedication included ondansetron 8 mg orally 1 hour before the amifostine and lorazepam 1 mg sublingually followed by the standard paclitaxel premedication package. Patients received 250 mL of normal saline over 30 to 60 minutes before the amifostine. The amifostine dose was given at a dose of 910 mg/m2 over 15 minutes, just before the paclitaxel infusion. During the amifostine infusion, patients were kept in a supine position and their blood pressure was monitored every 5 minutes. Vital signs, including heart rate and blood pressure, were monitored every 15 minutes for the first hour of the paclitaxel infusion and then every 30 minutes to completion. Concomitant supportive therapy, including analgesics and antiemetics, was allowed and recorded on the case report forms. Colony-stimulating growth factors were not permitted. Concomitant radiotherapy for symptom control in the absence of disease progression was allowed, but the protocol drugs were held for at least 2 weeks after irradiation.
Dose Modification Guidelines
The infusion of amifostine was interrupted if the patient developed a drop in systolic blood pressure that was significant when compared with baseline and according to specific guidelines, which were at the bedside and consistently followed by the nurses. The dose of amifostine was reduced if the patients had significant hypotension combined with other symptoms
Toxicity and Response Evaluation
Response was assessed every 3 weeks by physical examination and radiologic evaluation, except computed tomography scans and ultrasounds, which were required every 6 weeks. All patients with measurable disease who received at least one course of chemotherapy were considered assessable for response. A complete response (CR) was defined as the complete disappearance of all measurable and assessable disease and a partial response (PR) as a greater than 50% reduction in the sums of the product of the longest diameter and its perpendicular for each lesion. Stable disease (SD) was considered for all patients who had less than a PR but no evidence of progressive disease (PD). All responses had to be confirmed at least 4 weeks after the initial assessment. PD was defined as an unequivocal increase of It was planned that patients would be treated for four cycles past a CR, four cycles after a PR stabilized or until disease progression, and for a maximum of 10 cycles after SD or until serious or unmanageable toxicity supervened. Patients with PD were taken off study at the time the progression was documented clinically or radiologically. All patients were followed after discontinuation of the protocol treatment.
Pharmacokinetics Blood samples were drawn before the paclitaxel infusion, at 90 minutes into the 3-hour infusion, and at 5, 15, and 30 minutes and 1, 2, 3, 4, 6, and 24 hours after the end of the infusion. Samples were separated by centrifugation at 1,500 rpm and the plasma was stored at -70°C until analysis. For each sample (analyzed in duplicate), 100 µL of plasma was extracted using 1 mL of 0.8 µmol/L baccatin III as an internal standard in acetronitrile. Samples were vortexed for 10 seconds upon addition of organic solvent and then centrifuged at 3,000 rpm for 5 min. The organic extracts were then decanted and evaporated to dryness at room temperature under a nitrogen stream. Standard curves were generated using blank human plasma spiked with paclitaxel. High-performance liquid chromatography was carried out using an online system: Waters 600E multisolvent delivery system, 717 plus autosampler and 996 photodiode array detector; 3.0 x 150 mm Waters µBondapack Phenyl column injection volume, 120 µL; detection, 200 to 400 nm; run time, 50 min; flow rate, 1 mL/min. The mobile phase used was acetonitrile and double-distilled water mixed using the gradient profile (time, in minutes, %ddH2O, 0:90, 2:90, 30:35, 40:35, 45:90, and 50:90). Chromatograms were processed for traces obtained at 230 nm. Paclitaxel and the internal standard produced peaks with retention times, ie, 28.3 minutes and 21.8 minutes, respectively. The amount of paclitaxel in each sample was determined using formulas generated from standard curves, which were extracted separately for each day of analysis.22
Statistical Analysis Based on observations by Postma regarding the mean change in symptom score in patients receiving this dose of paclitaxel without amifostine, the sample of 20 patients per arm would have an 80% power of detecting a 75% reduction in the average change score at the first course using a two-sided 5% level test (ie, from 13.7 to 3.425). In addition, with this number of patients, there would be a 97% power of detecting a 75% reduction in the average change score of the CSSy and a 74% power of detecting a 50% reduction in the average CSSy change scores at three courses. The sample size of 40 patients would also have an 80% power of detecting a 47% difference in dose reductions between the two arms.
Patient Characteristics Forty-two women with metastatic breast cancer were enrolled onto the study. Two were ineligible because of cardiac findings not made until after randomization and never treated. Forty patients were considered eligible, and 20 were randomized to each arm. Treatment arms were balanced with respect to most baseline characteristics, as documented in Table 2. Fewer patients in arm 1 had received prior chemotherapy for metastatic disease (three of 20 v eight of 20). More arm 1 patients had three or more sites of disease at study entry (10 of 20 v six of 20). Neither of these differences was significant.
The patients ranged in age from 32 to 75 years (median, 49 years). The majority, 35 of 40, had a performance status of 0 or 1. Twenty-six patients (65%) had been treated with adjuvant chemotherapy and 28% had received previous chemotherapy for metastatic disease. Twenty-four patients had received previous radiotherapy and 19 had received previous hormonal therapy. The majority of the patients (83%) had two or more sites of disease and six patients had four or more sites. Of the 40 patients, 17 had documented metastatic disease in their liver, 13 had disease in their lungs, 22 in their lymph nodes, 15 in their bone, and nine in their skin. Other sites of involvement included ascites (one patient), breast (six patients), pleural effusions (seven patients), soft tissue (six patients), and choroid of the eye (one patient).
Toxicity
Toxicity data categorized and graded using NCIC-CTG–modified common toxicity criteria demonstrated no differences between arm 1 and arm 2 in the common nonhematologic effects (Table 4). These effects included hair loss (95% v 90%), neurosensory changes (100% v 100%; three and two patients, respectively, had grade 3 toxicity; the remainder had grade 1 or 2), fatigue/lethargy (85% v 90%), and myalgia (95% v 90%, respectively). Nausea was seen more commonly on the amifostine arm (90%) than on the paclitaxel-alone arm (55%; P = .031, Fisher's exact text). Similarly, in keeping with the known side effects of this agent, vomiting (P = .054), hypotension (P = .003), and sneezing (P = .231) were observed more commonly on the amifostine arm, although these were not always significantly different (Table 4).
Hypersensitivity reactions were difficult to assess because of the acute toxicity of the amifostine in many patients. Symptoms were therefore carefully reviewed to differentiate those symptoms that are typical of a true paclitaxel reaction and those that are common with amifostine. In both arms, symptoms were seen and consisted most commonly of flushing, which was generally mild but was recorded as grade 3 in one patient on arm 2. One episode of grade 3 dizziness was reported on arm 1, and one patient on arm 2 experienced grade 4 shortness of breath. Hematologic toxicity was similar in both arms. Grade 4 neutropenia was seen in nine of 19 patients in arm 1 and 12 of 20 in arm 2 (P = .527). The nadir WBC count was reached at a median time of day 15 for all patients. Thrombocytopenia was similar in both arms and never more than grade 1. Mild anemia was seen equally in both arms, with no grade 3 or 4 toxicity. No significant biochemical toxicity was seen with mild elevations of liver enzyme levels, but only two measurements were elevated high enough to be considered grade 3 toxicity and only one was in a patient with a normal baseline value. It was suggested that this represented disease progression. Seven patients were taken off the study for toxicity. One patient on arm 1 was taken off for grade 3 neurotoxicity occurring after cycle 1 of paclitaxel. All the other patients were on arm 2 and seemed to have had amifostine-related toxicity. Only one patient received a single dose of amifostine before it was discontinued; all the others had between two and four courses. The toxicities included grade 4 vomiting after three cycles despite antiemetic therapy, grade 3 vomiting and grade 2 dizziness in two patients occurring after four cycles and after two cycles, respectively, grade 2 nausea and vomiting in all three cycles in another patient, and the same complaints as well as dyspnea, sneezing, tingling, and dizziness in a patient on the first cycle. One patient who received three cycles in total had upper-body flushing, dizziness, chills, and tremors on all cycles, despite a reduction in the dose of amifostine. Patients randomized to arm 2 all continued on paclitaxel after being taken off the study without reoccurrence of the symptoms described above.
Drug Delivery The delivered dose-intensity is summarized in Table 5. The initial four courses of paclitaxel were planned at 250 mg/m2. The median received dose was close to that planned on both arms, but only 60% of patients in arm 1 and 75% of patients in arm 2 received 90% or greater of the planned dose. The dose-intensity of subsequent treatments planned at the standard dose of 175 mg/m2 was higher and was 95% of the planned dose-intensity. The numbers of cycles with dose reductions were similar in the two arms.
Response Eighteen patients who received paclitaxel alone and 19 patients who received combination treatment were assessable for response (Table 6). In arm 1, there was one CR and three PRs, for a response rate of 22.2% (95% confidence interval, 6.4% to 47.6%), and eight SDs and six PDs. In arm 2, there was one CR and six PRs, for a response rate of 36.8% (95% confidence interval, 16.3% to 61.6%), and six SDs and six PDs. The difference in response rate between the two arms of 1.6% in favor of the amifostine arm has a 95% confidence interval of -20% to 49% and is not significant.
The sites of CR were the liver in one patient and the lung and nodes in the other. When all the CRs and PRs were considered, responses were seen in all disease sites. Only three of the 11 responding patients had previous exposure to chemotherapy for metastatic disease, and they were distributed in both arms.
Pharmacokinetics
Paclitaxel induces a wide range of neurotoxic effects that are commonly peripheral sensory in nature but can also be motor or autonomic. The sensory neuropathy typically presents as numbness and paresthesia in a glove and stocking distribution, with symptoms often appearing within 24 to 72 hours after treatment. These are troublesome to many patients and may be severe and debilitating in others. Paclitaxel is now widely used in a variety of tumors and continues to be developed for other indications. The recent reports of paclitaxel efficacy in an adjuvant breast cancer study will further promote its use in women who are otherwise asymptomatic and who may be expected to have a long survival.25 In this setting, the issue of toxicity is particularly significant, as it may affect the long-term quality of life of the patient. Furthermore, its combination with cisplatin and carboplatin lead to an excess of neurologic problems. Thus, an agent that could prevent or reduce the incidence and severity of paclitaxel neuropathy would be of benefit in protecting patients in both curative and palliative clinical situations. Our study was designed to investigate this question and randomized women to high-dose paclitaxel with or without amifostine. We failed, however, to show any protection against neurotoxicity in the schedule we used. The rationale for choosing this agent for neuroprotection stems from a number of sources. Amifostine is an organic thiophosphate compound that has been shown in animal models and some human studies to protect tissues by binding directly to the active species of the alkylating or platinum agents. The ability of amifostine to selectively protect normal tissues is based on the differential metabolism and uptake of the free thiol into normal versus malignant tissues. With platinum agents, amifostine reduces the formation of platinum-DNA adducts, which protects the tissue from acute toxicity. In in vitro studies, it seems to reduce the genotoxic and carcinogenic properties of the platinum compounds and alkylators. When given with radiation therapy and drugs such as anthracyclines, the free thiol seems to act as a potent scavenger for oxygen free radicals that are implicated in toxicity.
Despite the preclinical laboratory data suggesting that amifostine might protect against some of paclitaxel's toxic effects, the data from this trial speak against there being meaningful protection in the clinical setting. The lack of protection is likely related to the fact that paclitaxel is a tubulin-active agent whereas the purported mechanism of action of amifostine is DNA-releated. It also highlights the need to carefully consider the mechanism of action of the cytotoxics we use to better understand the damage they cause and how we may ameliorate it. It is possible, of course, that amifostine does have protective activity for paclitaxel toxicity and we simply failed to detect it. One explanation for this might be related to the schedule of amifostine administration and another explanation might be related to the study sample size. Amifostine is cleared rapidly from the plasma with an alpha half-life of less than 1 minute and a beta half-life of approximately 8 minutes.26,27 It is rapidly metabolized, especially in normal tissues, to the active free thiol metabolite WR-1065. Although concentrations of WR-1065 have been detectable in studies of bone marrow cells 5 to 8 minutes after intravenous infusion of the 910-mg/m2 dose, it may be that this clearance occurs so rapidly that it is not available to affect paclitaxel, which in this trial was administered over 3 hours. This schedule and dose of amifostine has been adopted for use with radiation and cisplatin, which are delivered in a much shorter time frame after amifostine dosing. Therefore, if the protective effect of amifostine on paclitaxel is still believed to be worthy of study on the basis of preclinical data, further studies of shorter paclitaxel infusions following amifostine might be worthwhile. Could we have missed an effect based on the number of patients we enrolled? The sample size was determined based on an 80% power of detecting a 75% reduction in the mean change in score at the first dose and a 74% power of detecting a 50% reduction in the mean change in scores at three courses. It is possible, however, that lesser degrees of protection might have been of clinical interest. Some nonsignificant trends favoring the amifostine arm are noted when looking at toxicity, as shown in Table 4, for example, 11 of 20 patients had grade 2 or 3 neurosensory toxicity in the control arm compared with only seven of 20 in the amifostine arm. Similarly, seven of 20 had grade 3 myalgia in the control arm versus four of 20 in the amifostine arm. However, interpretation of these trends, which seemingly favor amifostine, must be balanced by similar, nonsignificant observations in the opposite direction: nine of 20 patients with grade 4 neutropenia in the control arm versus 12 of 20 in the amifostine arm and one of 20 patients with thrombocytopenia in the control arm compared with eight of 20 in the amifostine arm. Thus, while the trial may have been underpowered to find significance in lesser differences, it is important to note that there is no consistent trend favoring the amifostine arm when one examines all the major toxic effects of paclitaxel (neurosensory, myalgia, and myelosuppression). In fact the only significant differences in toxicity between the two arms were the increased frequency of amifostine-related toxicities on that arm, ie, nausea, vomiting, and hypotension. Could the unblinded nature of the study obscure results? With the side effects of amifostine, it was difficult to do a blinded study. As patients knew which arm they were on and hoped that they would have protection, we would have expected that the CSSy would reflect a minimizing of neurotoxicity, if it were biased in reporting toxicity, and this was not seen. The toxicity of amifostine was significant in some patients. Six of the women enrolled on arm 2 required discontinuation of amifostine for symptomatic effects, whereas only one of the women on arm 1 stopped protocol therapy with paclitaxel for toxicity. The phase I study of paclitaxel plus amifostine escalated the dose of paclitaxel to 310 mg/m2 and reported little paclitaxel or amifostine toxicity. The investigators hypothesized that the premedication package normally used with paclitaxel protected their patients against the side effects of the thiol. Our patients did not enjoy the same protection. This may have been related to the use of oral ondansetron, the dose and timing of the antiemetic, the timing and compliance with lorazepam, or simply the further experience with these drugs and larger patient numbers. The widespread use of amifostine, however, will be limited by these side effects, which, although of short duration, were extremely uncomfortable for some patients. They may be less significant with the current recommended dose of 740 mg/m2, which is lower than the 910-mg/m2 dose used in our study. The response rate for the amifostine group was slightly higher, although not statistically different, confirming that it does not protect the tumor. It may, however, cause a pharmacokinetic interaction, which could be obscured if there is not a dose-response curve. Schüller et al28 reported that the AUCs of patients pretreated with amifostine be fore paclitaxel treatment were 29% lower than the AUCs obtained without pretreatment. Following his report, we initiated a pharmacokinetic study as described in Patients and Methods. Our pharmacokinetic assessments resulted in identical curves and did not show any clear interaction of amifostine and paclitaxel. As well, when one considers the metabolism of both of these agents, it is difficult to hypothesize what the mechanism of an interaction may be. The dose of paclitaxel in this study was 250 mg/m2 and not the standard 175-mg/m2 dose. This dose was chosen to explore the toxicity question, as neurotoxicity and myalgias are more common and severe at increasing doses; thus, this dose was chosen so that any differences produced by the addition of amifostine would be readily discernible. At the time the study was initiated, there were questions about the optimal dose of short-infusion paclitaxel. A recent report of the Cancer and Leukemia Group B failed to show a dose-reposne curve in metastatic breast cancer and suggests that there is no role for high-dose paclitaxel in this setting.29 Our response rate was consistent with this study and other studies using standard doses and does not give any support to higher doses, particularly as the toxicity of the high-dose paclitaxel was considerable. If there is a pharmacokinetic interaction, it may be obscured by the lack of a dose-response curve, but this simply lends further support to using a lower, less toxic, and equally efficacious dose. The study could be criticized for not showing more neurotoxicity in terms of NCIC grade 3 and above and for reducing the dose of paclitaxel for grade 2 neurotoxicity. We did, however, show significant toxicity in terms of both the NCIC grading and the CSSy beginning at cycle 1 in both arms. As well, we did not believe it was ethically justified to continue treating patients in the metastatic setting at high dose levels if neurotoxicity was reported. Although our study did not show that amifostine prevented or reduced the incidence of paclitaxel toxicity when given in this schedule, it does not provide information about the use of amifostine with other chemotherapeutic agents, with radiation, or with paclitaxel in combination with neurotoxic drugs such as cisplatin. Several studies demonstrating the feasibility of using amifostine in a number of other regimens, including high-dose chemotherapy for breast cancer and anthracycline-based therapy, have been reported.30-33 In all cases, these are single-arm trials, and although the use of the thiol compound may be feasible, it is premature to draw conclusions about efficacy from such data. Randomized trials are necessary before a role for amifostine is clearly established in these clinical settings.
Supported by grants from the National Cancer Institute of Canada, Eli Lilly Co Canada, and Bristol-Myers, Squibb, Canada We thank the following investigators who, in addition to the authors, enrolled patients onto this study: Dr I. Hings, McGill University, Montreal; Drs S. Gertler, R. Goel, and S. Verma, Ottawa Regional Cancer Centre, Ottawa; and Drs J. Ragaz, S. O'Reilly, B. Melosky, and T. Shenkier, British Columbia Cancer Agency, Vancouver, Canada. In addition our thanks go to the study nurses who administered the standardized neurologic tests and questionnaires.
1. Eisenhauer EA, ten Bokkel Huinink WW, Swenerton KD, et al: European Canadian randomized trial of paclitaxel in relapsed ovarian cancer: High dose versus low dose and long versus short infusion. J Clin Oncol12:2654-2666, 1994
2.
Nabholtz JM, Gelmon KA, Bontenbal M, et al: Multicenter, randomized comparative study of two doses of paclitaxel in patients with metastatic breast cancer. J Clin Oncol14:1858-1867, 1996 3. Sarosy G, Kohn E, Stone DA, et al: Phase I study of Taxol and granulocyte colony stimulating factor in patients with refractory ovarian cancer. J Clin Oncol10:1165-1170, 1992[Abstract] 4. Seidman AD, Reichman BS, Crown JP, et al: Paclitaxel as second and subsequent therapy for metastatic breast cancer: Activity independent of prior anthracycline response. J Clin Oncol13:1152-1159, 1995[Abstract] 5. Chaudhry V, Rowinsky EK, Sartorius SE, et al: Peripheral neuropathy from Taxol and cisplatin combination chemotherapy: Clinical and electrophysiological studies. Ann Neurol35:304-311, 1994[Medline] 6. Masurovsky EB, Peterson ER, Crain SM, et al: Microtubule arrays in Taxol treated mouse dorsal root ganglion spinal cord cultures. Brain Res217:392-398, 1981[Medline] 7. Roytta M, Raine CS: Taxol induced neuropathy: Chronic effects of local injection. J Neurocytol15:483-496, 1986[Medline] 8. Rowinsky EK, Eisenhauer EA, Chaudhry V, et al: Clinical toxicities encountered with paclitaxel (Taxol). Semin Oncol20:1-15, 1993 (supple 3) [Medline] 9. Warner E: Neurotoxicity of cisplatin and Taxol. Int J Gynecol Cancer5:161-169, 1995[Medline]
10.
Postma TJ, Vermorken JB, Liefting AJM, et al: Paclitaxel induced neuropathy. Ann Oncol6:489-494, 1995 11. Yuhas JM, Spellman JM, Culo F: The role of WR-2721 in radiotherapy and/or chemotherapy. Cancer Clin Trials3:211-216, 1980[Medline] 12. Calabro Jones PM, Fahey RC, Smoluk GD, et al: Alkaline phosphatase promotes radioprotection and accumulation of WR 1065 in V79-171 cells incubated in medium containing WR 2721. Int J Radiat Biol47:23-27, 1985
13.
Yuhas JM: Active versus passive absorption kinetics as the basis for selective protection of normal tissues by S-2(3-aminopropylamino)-ethylphosphorothioic acid. Cancer Res40:1519-1524, 1980 14. Wang LM, Wang QW, Fernandes DJ, et al: Amifostine protects MRC-5 human lung fibroblasts from Taxol toxicity without reducing its cytotoxic effect against human non-small cell lung cancer cells. Proc Am Assoc Cancer Res 36:288a, 1995 (abstr 1712) 15. Taylor CW, Wang LM, List AF, et al: Amifostine protects normal tissues from paclitaxel toxicity while cytotoxicity against tumour cells is maintained. Eur J Cancer33:1693-1698, 1997 16. Paine GD, Taylor CW, Lopez MH, et al: Effects of amifostine and paclitaxel on growth of human ovarian carcinoma xenografts in the severe combined immunodeficient mouse: Preliminary results. Semin Oncol23:35-39, 1996 (suppl 8)
17.
Treskes M, Boven E, Holwerda U, et al: Time dependence of the selective modulation of cisplatin-induced nephrotoxicity by WR 2721 in the mouse. Cancer Res52:2257-2260, 1992 18. Treskes M, Boven E, Loosdrecht AA, et al: Effects of the modulating agent WR2721 on myelotoxicity and antitumour activity in carboplatin treated mice. Eur J Cancer30:183-187, 1994 19. Mollman JE: Cisplatin neuropathy: Risk factors, prognosis and protection by WR 2721. Cancer61:2192-2195, 1988[Medline]
20.
Kemp G, Rose P, Lurain J, et al: Amifostine pretreatment for protection against cyclophosphamide induced and cisplatin-induced toxicities: Results of a randomized control trial in patients with advanced ovarian cancer. J Clin Oncol14:2101-2112, 1996 21. DiPaola R, Rodriguez R, Recio A, et al: A phase I study of paclitaxel and amifostine in patients with advanced malignancies. Proc Am Soc Clin Oncol15:488, 1996 (abstr)
22.
Harris JW, Rahman A, Bok-Ryang K, et al: Metabolism of Taxol by human hepatic microsomes and liver slices: Participation of cytochrome p450 3A4 and an unknown p450 enzyme. Cancer Res54:4026-4035, 1994 23. Yates F: Continuity tables involving small numbers and the z2 test. J R Stat Soc Suppl1:217-235, 1934 24. Fleiss J: Statistical Methods for Rates and Proportions (ed 2). New York, NY, Wiley, 1985 25. Henderson IC, Berry D, Demetri G, et al: Improved disease-free (DFS) and overall survival (OS) from the addition of sequential paclitaxel (T) but not from the escalation of doxorubicin (A) dose level in the adjuvant chemotherapy of patients (PTS) with node-positive primary breast cancer (BC). Proc Am Soc Clin Oncol 17:101a, 1998 (abstr 390A) 26. Shaw LM, Turrisi AT, Glover DJ, et al: Human pharmacokinetics of WR-2721. Int J Radiat Oncol Biol Phys12:1501-1504, 1986[Medline] 27. US Bioscience Pharmacokinetics of double dose amifostine (WR 2721) Ethyol) with corresponding measurements of WR 1065 in plasma and bone marrow cells. ETH PK 4 1994. Palo Alto, CA, U.S. Bioscience, 1994 28. Schüller J, Czejka M, Pietrzak C, et al: Influence of the cytoprotective agent amifostine (AMI) on pharmacokinetics (PK) of paclitaxel (PAC) and Taxotere® (TXT). Proc Am Soc Clin Oncol Abst 16:224a, 1997 (abstr 785) 29. Winer E, Berry D, Duggan D, et al: Failure of higher dose paclitaxel to improve outcome in patients with metastatic breast cancer: Results from CALGB 9342. Proc Am Soc Clin Oncol 17:101a, 1998 (abstr 388) 30. Mitchell RB: Amifostine to ameliorate the neurotoxicity of carboplatin/paclitaxel for advanced non-small-cell lung cancer. Proc Am Soc Clin Oncol 17:495a, 1998 (abstr 1906) 31. Carlo-Stella C, Dodero A, Regazzi E, et al: Amifostine enhances chemotherapy-elicited peripheral blood progenitor cell mobilization. Proc Am Soc Clin Oncol 17:78a, 1998 (abstr 301) 32. Reece D, Glenn C, Hale G, et al: Acute effects of multiple doses of Ethyol® (amifostine [AF]) plus high dose paclitaxel (P), cyclophosphamide (CY), cisplatin (cDDP) and autologous stem cell transplantation (ASCT) in patients (PTS) with metastatic breast cancer. Proc Am Soc Clin Oncol 17:94a, 1998 (abstr 364) 33. De Lena M, Catino A, Lorusso V, et al: Amifostine (AMF) protects against doxorubicin (D) toxicity with preservation of therapeutic activity in advanced breast cancer (ABC) patients. Proc Am Soc Clin Oncol 17:137a, 1998 (abstr 521) Submitted December 17, 1998; accepted June 11, 1999.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 1999 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 cm x 2 cm on computed tomography scan or ultrasound and 
2 test with Yates' continuity correction or Fisher's exact test was used to compare the dose reduction rates. The cumulative percentage of patients with dose reductions at cycle 3 was also used as an end point in the study. Other toxicities, such as myalgia and neutropenia, were compared using a Pearson 
, arm 1; 
, arm 2).
