|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2001 American Society for Clinical Oncology Phase III Comparison of High-Dose Paclitaxel + Cisplatin + Granulocyte Colony-Stimulating Factor Versus Low-Dose Paclitaxel + Cisplatin in Advanced Head and Neck Cancer: Eastern Cooperative Oncology Group Study E1393From the Johns Hopkins Oncology Center, Baltimore, MD; Dana-Farber Cancer Institute and Brigham & Women’s Hospital, Boston, MA; Vanderbilt University, Nashville, TN; H. Lee Moffitt Cancer Center, Tampa, FL; and Fairview University Medical Center, Minneapolis, MN. Address reprint requests to Arlene A. Forastiere, MD, Johns Hopkins Oncology Center, Bunting-Blaustein Cancer Research Building, 1650 Orleans St, Baltimore, MD. email: forasar{at}jhmi.edu
PURPOSE: To determine dose-response effects and the activity of paclitaxel combined with cisplatin in patients with incurable squamous cell carcinoma of the head and neck. PATIENTS AND METHODS: Two hundred ten patients with locally advanced, recurrent, or metastatic disease were randomly placed in either Arm A, paclitaxel 200 mg/m2 (24-hour infusion) + cisplatin 75mg/m2 + granulocyte colony-stimulating factor, or Arm B, paclitaxel 135 mg/m2 (24-hour infusion) + cisplatin 75 mg/m2. Cycles were repeated every 3 weeks until progression or a total of 12 cycles for complete responses. Primary outcomes were event-free and overall survival. RESULTS: No significant differences in outcomes were observed between the high- and low-dose paclitaxel regimens. The estimated median survival was 7.3 months (95% confidence interval, 6.0 to 8.6). The 1-year survival rate was 29%, and event-free survival was 4.0 months. The objective response rate (complete response plus partial response) was 35% for the high-dose patients and 36% for the low-dose patients. Myelosuppression was the most frequent toxicity: grade 3 or 4 granulocytopenia, 70% of patients in Arm A and 78% in Arm B; febrile neutropenia, 27% of patients in Arm A and 39% in Arm B. Grade 5 toxicities occurred in 22 patients (10.5%). Treatment was terminated early in 31% because of excessive toxicity or patient refusal. CONCLUSION: This phase III multicenter trial showed (1) no advantage for high-dose paclitaxel and (2) excessive hematologic toxicity associated with both regimens. Therefore, neither of the paclitaxel regimens evaluated in this trial can be recommended.
HEAD AND NECK cancer comprises 5% of invasive cancers diagnosed annually in the United States and less than one half of affected individuals can expect to be cured. Over the past two decades, trials have been conducted that integrate chemotherapy into initial curative treatment that uses various strategies. The combination of cisplatin and fluorouracil (FU) is a standard regimen integrated into initial combined modality therapy or for treatment of recurrent and metastatic disease. When used in the palliative setting, randomized trials have demonstrated significantly higher response rates to the combination therapy when compared with single agent cisplatin, FU, and methotrexate.1-3 Despite a higher response rate with combination therapy, the median survival of patients is only 6 months, with a 1-year survival rate of 20%, regardless of treatment. Induction and adjuvant chemotherapy trials that use cisplatin + FU have also failed to improve survival when compared with standard definitive local therapy.4-6 Although concomitant chemoradiation strategies can improve survival and local-regional control of patients with locally advanced disease,7,8 it is clear that new agents and more effective combination regimens need to be identified. By convention, new cytotoxic agents are tested first in the recurrent disease population and then brought forward for testing in combined modality treatments in patients with potentially curable disease. Following that course, the Eastern Cooperative Oncology Group (ECOG) conducted a phase II trial of single agent paclitaxel using 250 mg/m2 by 24-hour infusion. In 30 assessable patients, a 40% response rate was observed along with a median survival of 9.2 months and 1-year survival rate of 33%.9 A next logical step was to combine paclitaxel with cisplatin. Two questions were of interest. The first, given its unique mechanism of action to promote polymerization of microtubules, does a dose-response effect exist for paclitaxel? The second, is this combination superior to cisplatin + FU? To that end, we wished to establish the efficacy and toxicity of cisplatin and two different doses of paclitaxel in the population of patients with head and neck cancer. Based on the results of this trial, we would select the superior paclitaxel-based regimen for a subsequent direct comparison to cisplatin + FU. From July 1993 to June 1995, the ECOG conducted a trial that compared high-dose paclitaxel (200 mg/m2) + cisplatin and granulocyte colony-stimulating factor (G-CSF) support to low-dose paclitaxel (135 mg/m2) + cisplatin in patients with locally advanced incurable, recurrent, or metastatic squamous cell carcinoma of the head and neck.
Eligibility Criteria All patients had measurable, histologically confirmed squamous cell carcinoma from any site in the head and neck, excluding nasopharynx, and an ECOG performance status of 0 or 1. Two groups of patients were eligible: (1) those newly diagnosed with distant metastases or with local-regional disease so extensive that cure was not possible with surgery or radiotherapy; and (2) patients with recurrent or metastatic disease after initial surgery or radiotherapy. No prior chemotherapy was allowed for treatment of recurrent/metastatic disease, but patients could have received chemotherapy as part of initial curative treatment, if more than 6 months before study entry. Also required were the following: adequate bone marrow reserve, defined as an absolute neutrophil count of  1,500/µL, platelets 100,000/µL, and hemoglobin 10 gm/dL; serum creatinine  1.5 mg/dL; and total bilirubin less than 1.5 mg/100 mL. Some patients with cardiac disease were excluded, specifically those on medications known to alter cardiac conduction; also excluded were those with a history of congestive heart failure or a myocardial infarction within 6 months of study entry. All patients provided written informed consent in accordance with institutional guidelines.
Randomization and Treatment Dose modifications were specified for hematologic toxicity and neurotoxicity. Cisplatin dose was not modified for hematologic toxicity; paclitaxel dose modifications were based on the nadir absolute neutrophil count and nadir platelet count and specified for each toxicity episode. Doses were modified for grade 4 neutropenia exceeding 5 days, febrile neutropenia, or grade 4 thrombocytopenia. Patients in the high-dose arm had the following paclitaxel dose reductions: first episode, paclitaxel dose reduction to 170 mg/m2 + G-CSF; second episode, dose reduction to 135 mg/m2 without G-CSF; third episode, dose remained at 135 mg/m2 and G-CSF was added. Patients in the low-dose arm had the following dose modifications: first episode, the dose remained at 135 mg/m2 and G-CSF was added; second episode, the dose was reduced to 110 mg/m2 (without G-CSF); third episode, the dose remained at 110 mg/m2 and G-CSF was added. There was no re-escalation of dose.
The criteria for dose modification for neurotoxicity were based on patient response and toxicity grade such that responding patients who developed mild-to-moderate functional loss, grade 3 neuropathy, continued treatment with dose reductions for paclitaxel and cisplatin, whereas patients with stable disease discontinued treatment. Dose modifications for nephrotoxicity were based on the serum creatinine obtained before each course. For serum creatinine of 1.6 to 1.9 mg/dL, cisplatin dose was reduced to 50 mg/m2, and for serum creatinine Patients who received G-CSF were monitored with a complete blood count and differential count starting on day 8 until G-CSF was discontinued; otherwise, a complete blood count and differential count was obtained weekly. Liver function tests and serum creatinine were repeated before each course. Tumor measurements were documented every 3 weeks. Duration of treatment was dependent on response. Patients who achieved a complete response (CR) received a maximum of 12 courses. An interruption in treatment to receive radiotherapy was allowed at the discretion of the treating physician for those previously untreated patients who achieved a CR. All other patients continued treatment until there was objective evidence of progression of disease.
Response Criteria
Study Design and Statistical Methods The Fisher’s exact test was used to compare groups with respect to a dichotomous end point (eg, response). The logistic model was used to adjust for one or more explanatory variables simultaneously when assessing response. Overall survival was calculated from the date of registration to the date of death or date when last known to be alive. Event-free survival was calculated from the date of registration to the date of progression, relapse, death, or date when last known to have had relapse or to have been progression-free. Overall survival curves were estimated with the Kaplan-Meier method and compared with the log-rank test. The proportional hazards model was used to evaluate the effects of treatment and prognostic factors on event-free survival. All P values were two-sided. The Wilcoxon-Mann-Whiting test was use to test whether toxicity (ordered by grades 0 to 5) differed by treatment. Exact tests were considered when necessary.
Patient Characteristics A total of 210 patients were entered onto the trial with 105 on each treatment arm. Four of the high-dose patients and six of the low-dose patients were ineligible, and one patient who was placed randomly to low-dose never received treatment. The remaining 199 patients (101 high-dose, 98 low-dose) comprised the analyzable population. Patient characteristics for the two treatment arms are listed in Table 1. Sex, race, and age distributions were typical of this population. Approximately one third of patients had an ECOG performance status of 0, and two thirds had a performance status of 1. Only 14 patients in the high-dose arm and 12 in the low-dose arm were previously untreated. The majority of patients (64% of high-dose, 62% of low-dose) had metastatic disease. The most common sites were lung and soft tissue. The distribution of primary sites was similar in the two treatment groups.
Toxicity Toxicity data are reported for all patients who received treatment, 105 in Arm A (high-dose paclitaxel + cisplatin + G-CSF) and 104 in Arm B (low-dose paclitaxel + cisplatin). Patients in the high-dose arm received a total of 430 courses (median, three courses per patient; range, one to 13). Those in the low-dose arm received a total of 465 courses (median, four courses per patient; range, one to 12). In Arm A, 83 patients received two or more courses of chemotherapy and 32 (38.5%) required dose modification. In Arm B, 87 patients received two or more courses of chemotherapy and 40 (46%) required dose modification. The total number of courses received at the starting dose and with dose modification for each treatment arm is listed in Table 2. In the high-dose arm, 77% of courses were at the 200-mg/m2 paclitaxel dose. The dose for 14% of courses was 170 mg/m2 of paclitaxel and 75 mg/m2 cisplatin, which reflected dose modification for hematologic toxicity; in 4% of courses, cisplatin was reduced for neurotoxicity or renal toxicity. Thus, 95% of courses were high-dose with 170 to 200 mg/m2 of paclitaxel administered. In the low-dose arm, 96% of courses were at the 135-mg/m2 paclitaxel dose.
Grades 3 to 5 toxicities are listed in Table 3. Myelosuppression and infection were the most common toxicities, with 74 patients (70%) in Arm A and 81 patients (78%) in Arm B experiencing grades 3 and 4 granulocytopenia. Serious infection occurred in 21 patients (20%) in Arm A and 24 patients (23%) in Arm B. Febrile neutropenia and hospitalization was determined in 27% of patients on the high-dose arm and 39% of patients on the low-dose arm by retrospective record review. This was defined as temperature 38.2°C and absolute neutrophil count less than 2,000 and hospitalization. The majority of these episodes occurred after the first course of chemotherapy. The incidence of febrile neutropenia was not significantly different between the two treatment arms. Hospital stay was slightly longer (median, 5 days) for patients in the low-dose arm compared with 3 days for those in the high-dose arm (P = .07).
The frequency of nonhematologic toxicities was quite similar in both treatment arms. Of note, the incidence and severity of neuromotor and neurosensory toxicities were similar in both treatment groups. Treatment complications that resulted in death were substantial, totaling 22 patients (10.5%), 13 in Arm A and nine in Arm B. As detailed in Table 3, a total of 13 patients died from infection and two, one in each arm, from renal failure. Two cardiac deaths were reported in Arm A patients, both from acute myocardial infarction. This occurred in one patient immediately after receiving a transfusion of two units of red cells on day 14 of course 3 of chemotherapy. The other patient died at home on day 7 of course 1, and the autopsy report indicated a myocardial infarction. Four patients were found dead at home of unclear cause, whereas a fifth patient was found unresponsive at home and later died in the hospital without a clear cause. The causes of these five deaths were classified as unknown. Four of the five patients were within 10 days of treatment, whereas a fifth had documented recovery from nadir counts. A weighted analysis of all grades 3 to 5 toxicities was performed. There was no difference by treatment.
Response
Approximately one third of patients were classified as having stable disease. This included 14 patients who had more than 50% reduction in tumor but did not have measurements repeated 4 or more weeks later. Only 20% of patients in the high-dose arm and 18% in the low-dose arm had disease that progressed immediately. Nineteen cases were deemed not assessable; 10 were randomly placed to the high-dose arm. The reasons for not being assessable were as follows: died before evaluation (13 patients) and inadequate baseline tumor measurements (six patients). These cases are included in the response analysis as nonresponders.
Objective response (CR + PR) was also analyzed by a number of patient characteristics: sex, race, age ( < 55, 55 to 64,
Duration of Treatment
Survival Estimated median survival was 7.3 months with a 95% CI of 6.0 to 8.6 months. There was no difference in survival by treatment (P = .759). The median survival in the high-dose arm was 7.6 months and in the low-dose arm was 6.8 months. One-year survival rates were 28.6% and 29.4% for the high- and low-dose arms, respectively. Overall survival for all eligible patients is illustrated in Fig 1.
At the time of this report, 191 patients had either relapsed, progressed, or died. The overall Kaplan-Meier estimates of the probability of remaining event-free at 6, 12, and 24 months respectively are 0.33, 0.09, and 0.03 ( Fig 2) for all eligible patients. The median event-free survival (EFS) for all patients is 4.0 months. By treatment, the EFS medians are 4.1 and 4.0 months for high- and low-dose, respectively. When all patients are considered in an intent to treat analysis, there is still no EFS difference between the arms.
An analysis of EFS by treatment and by other patient characteristics showed that weight loss, cell differentiation, disease status, and metastases significantly predicted EFS. Performance status was of marginal significance, whereas treatment, sex, race, age, and presence or absence of metastases were not significant. In a proportional hazards regression model of prognostic factors simultaneously adjusted for other patient characteristics, disease status (previously untreated or recurrent), weight loss (dichotomized into < 5% or 5% categories), and metastases were statistically significant (Table 6). An analysis of response and survival based on prior chemotherapy showed no significant differences in outcome. Twenty of 22 patients with prior exposure to chemotherapy had received cisplatin. The median survivals of chemotherapy-naive and prior cisplatin treated patients was 7.7 months and 5.3 months (P = .286). Although this difference corresponds to a 0.45% decrease, there is low power to detect such a difference with the distribution of patients (90% v 10%). There is no difference in objective response or CR: 30% of prior cisplatin-exposed and 36% of cisplatin-naive patients.
The objectives of this trial, designed in 1992, were to estimate whether a dose-response effect was evident for 24-hour infusion paclitaxel and to determine the activity of the combination of paclitaxel and cisplatin in this head and neck cancer patient population. EFS and overall survival were the primary end points. On comparing the two treatments, no survival advantage was observed for high-dose paclitaxel, and thus, one can conclude that there was no evidence for a dose-response relationship. This same conclusion has been reached in comparative trials for three other solid tumors: lung,10 ovary,11 and breast cancer.12,13 An ECOG randomized trial in non–small-cell lung cancer (E5592) compared paclitaxel (24-hour infusion schedule) in doses of 250 mg/m2 plus G-CSF and 175 mg/m2 combined with the same dose of cisplatin as used in our trial.10 Response rate, failure-free survival, and overall survival were virtually identical for the two arms. The Gynecologic Oncology Group studied this question in a randomized trial of single-agent 24-hour infusion paclitaxel 175 mg/m2 versus paclitaxel 250 mg/m2 plus G-CSF and reported no differences in progression-free or overall survival.11 Two trials involving patients with metastatic breast cancer have addressed the dosing question for 3-hour infusion paclitaxel.12,13 One trial compared doses of 135 mg/m2 and 175 mg/m2 (Bristol Myers Squibb 048),12 and the other compared doses of 175 mg/m2, 210 mg/m2, and 250 mg/m2 (Cancer and Leukemia Group B 9,342).13 In both trials, no differences were observed in response rates or overall survival. These results may be explained by the plateauing of cytotoxicity observed in vitro as paclitaxel concentration increases. This is probably a result of saturation of paclitaxel binding sites on beta-tubulin at the paclitaxel plasma steady-state concentrations achieved with doses of 135 mg/m2 or greater (24-hour infusion).14,15 Doses that produce plasma concentrations in excess of 10 µm seem to increase toxicity but not efficacy. Plasma concentrations of 10 µm are readily achievable with lower doses of paclitaxel.15 The most frequent toxicities were myelosuppression and infection, as expected with the 24-hour infusion schedule of paclitaxel.9 There was a higher incidence of grade 4 neutropenia in the low-dose arm, 71.1% of patients compared with 60.9% of patients in the high-dose arm. There was also a higher incidence of hospitalization for febrile neutropenia, 39% of patients on the low-dose arm compared with 27% on the high-dose arm, which suggests a benefit from the addition of G-CSF. There was, however, no difference in the percentage of full doses of chemotherapy received according to whether G-CSF was given. There was also no difference in the overall incidence of grades 4 and 5 toxicities in the two treatment groups. The doses of paclitaxel and cisplatin selected for combination were based on phase I trials that determined the maximum tolerated doses that could be administered with and without growth factor support.16 Thus, it is not surprising that differences in hematologic toxicity were not observed between the two dosing regimens. The 24-hour infusion schedule of paclitaxel is associated with the high rate of life-threatening neutropenia occurring 1 week after treatment.9,17 These two paclitaxel and cisplatin regimens resulted in an unacceptable toxic death rate of 10%. Whether this can be attributed to the 24-hour infusion schedule cannot be discerned from this study design. This level of grade 5 toxicity, however, exceeds that reported in previous multicenter trials of cisplatin and FU.1-3 The relative efficacy and toxicity of these regimens will need to be determined in appropriately designed randomized comparative trials. Patients with head and neck cancer are a particularly fragile population because of the frequency of complications from comorbid disease. This was clearly apparent in our trial with approximately 40% of patients refusing treatment or having treatment terminated because of toxicity. Objective response was evaluated, although this was not the primary end point of interest. All patients were required to have measurable disease by physical examination or imaging. The CR plus PR rates were similar for the two treatments, 35% for the high-dose paclitaxel regimen and 36% for the low-dose paclitaxel regimen. A significantly greater number of CRs was observed in the low-dose arm (P = .038). An explanation for this difference is not readily apparent. However, it is recognized that obtaining accurate serial tumor measurements for response determination is difficult in patients with head and neck cancer, particularly when the indicator lesion is a local-regional recurrence. The frequent use of computed tomography or magnetic resonance imaging to obtain objective measurements results in fewer CRs because of the sensitivity of these scans. Residual abnormalities of unclear significance are often reported despite a clinical evaluation that indicates CR. These limitations have led the ECOG to choose the clear end point of survival as the primary outcome to determine efficacy in phase III trials for this patient population. The logistic regression model for response and survival showed disease status (previously untreated or recurrent) as the most significant prognostic factor. This is consistent with other reports. The higher response rate in the previously untreated patients, 58%, represented only a small percentage of the accrual. The response rate of the 173 patients who had failed prior curative therapies was 32%. In summary, the two paclitaxel and cisplatin regimens compared in this trial were similar in toxicity profile. Severe myelosuppression that led to life-threatening infection was substantial in both arms. There was no evidence of a dose-response effect and thus, no benefit to the added expense of the high-dose regimen and growth factor support. Based on these multicenter results, neither of these paclitaxel + cisplatin regimens is recommended for treatment of patients with recurrent/metastatic head and neck cancer.
This study was conducted by the Eastern Cooperative Oncology Group (Robert L. Comis, MD, Chair) and supported in part by Public Health Service grant nos. CA16116, CA23318, CA49957, CA20365, CA66636, CA21115 from the National Cancer Institute, National Institutes of Health, the Department of Health and Human Services, Bethesda, MD. Its contents are solely the responsibility of authors and do not necessarily represent the official views of the National Cancer Institute.
Preliminary results were presented at the American Society of Clinical Oncology Meeting in Denver, Colorado, May 1997.
1. Forastiere AA, Metch B, Schuller DE, et al: Randomized comparison of cisplatin plus fluorouracil and carboplatin plus fluorouracil versus methotrexate in advanced squamous-cell carcinoma of the head and neck: A Southwest Oncology Group Study. J Clin Oncol 10: 1245-1251, 1992 2. Jacobs C, Lyman G, Velez-Garcia E, et al: A phase III randomized study comparing cisplatin and fluorouracil as single agents and in combination for advanced squamous cell carcinoma of the head and neck. J Clin Oncol 10: 257-263, 1992[Abstract]
3.
Clavel M, Vermorken JB, Cognetti F, et al: Randomized comparison of cisplatin, methotrexate, bleomycin and vincristine (CABO) versus cisplatin and 5-fluorouracil (CF) versus cisplatin (C) in recurrent or metastatic squamous cell carcinoma of the head and neck. A phase III study of the EORTC Head and Neck Center Cooperative Group. Ann Oncol 5: 521-526, 1994 4. Laramore G, Scott C, Al-Sarraf M, et al: Adjuvant chemotherapy for resectable squamous cell carcinoma of the head and neck: Report on Intergroup Study 0034. Int J Radiat Oncol Biol Phys 23: 705-713, 1992[Medline] 5. The Department of Veterans Affairs Laryngeal Cancer Study Group: Induction chemotherapy plus radiation compared with surgery plus radiation in patients with advanced laryngeal cancer. N Engl J Med 324: 1685-1690, 1991[Abstract] 6. Sessions RB, Harrison LB, Forastiere AA: Tumors of the larynx and hypopharynx, in DeVita VT Jr, Hellman S, Rosenberg SA (eds): Principles and Practice of Oncology (5th ed). Philadelphia, PA, JB Lippincott Co, 1997, 830-847 7. Adelstein DJ: Recent randomized trials of chemoradiation in the management of locally advanced head and neck cancer. Curr Opin Oncol 10: 213-218, 1998[Medline] 8. Forastiere AA, Trotti A: Radiotherapy and concurrent chemotherapy: A strategy that improves locoregional control and survival in oropharyngeal cancer. J Natl Cancer Inst 24: 2065-2069, 1999 9. Forastiere AA, Shank D, Neuberg D, et al: Final report of a phase II evaluation of paclitaxel in patients with advanced squamous cell carcinoma of the head and neck: An Eastern Cooperative Oncology Group Trial (PA390). Cancer 82: 2270-2274, 1998[Medline] 10. Bonomi P, Klim K, Chang A, et al: Phase III trial comparing etoposide-cisplatin versus Taxol with cisplatin-granulocyte-colony-stimulating factor versus Taxol-cisplatin in advanced non-small cell lung cancer. An Eastern Cooperative Group trial. Proc Am Soc Clin Oncol 15: 382a, 1996 (abstr 1145) 11. Omura GA, Brady MF, Delmore JE, et al: A randomized trial of paclitaxel at 2 dose levels and Filgastrim (G; G-CSF) at 2 dose levels in platinum pretreated epithelial ovarian cancer (OVCA): A Gynecologic Oncology Group, SWOG, NCTTG and ECOG study. Proc Am Soc Clin Oncol 15: 280a, 1996 (abstr 755)
12.
Nabholtz JM, Gelmon K, Bontenbal M, et al: Multicenter, randomized comparative study of two doses of paclitaxel in patients with metastatic breast cancer. J Clin Oncol 14: 1858-1867, 1996 13. 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, 1997 (abstr 388)
14.
Rowinsky EK: On pushing the outer edge of the outer edge of paclitaxel’s dosing envelope. Clin Cancer Res 5: 481-486, 1999
15.
Rowinsky EK, Jiroutek M, Bonomi P, et al: Paclitaxel steady-state plasma concentration as a determinant of disease outcome and toxicity in lung cancer patients treated with paclitaxel and cisplatin. Clin Cancer Res 5: 767-774, 1999
16.
Rowinsky EK, Chaudhry V, Forastiere AA, et al: A phase I and pharmacology study of paclitaxel and cisplatin with granulocyte colony-stimulating factor: Neuromuscular toxicity is dose-limiting. J Clin Oncol 11: 2010-2020, 1993 17. Vermorken JB, Catimel G, Hoekman K, et al: Randomized phase II trial of methotrexate (MTX) versus two schedules of paclitaxel (TAXOL) in patients with recurrent or metastatic squamous cell carcinoma of the head and neck (SCCHN). Proc Am Soc Clin Oncol 17: 391a, 1998 (abstr 1508) Submitted September 20, 1999; accepted October 18, 2000.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2001 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
