|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2000 American Society for Clinical Oncology Pamidronate in Prevention of Bone Complications in Metastatic Breast Cancer: A Cost-Effectiveness AnalysisFrom the Department of Internal Medicine, Virginia Commonwealth University and the Massey Cancer Center, Richmond, VA; and the Center for Outcomes and Policy Research, Dana-Farber Cancer Institute, Boston, MA. Address reprint requests to B.E. Hillner, Virginia Commonwealth University, Box 980170, Richmond, VA 23298-0170; email hillner{at}hsc vcu.edu.
PURPOSE: Pamidronate is effective in reducing bony complications in patients with metastatic breast cancer who have known osteolytic lesions. However, pamidronate does not increase survival and is associated with additional financial costs and inconvenience. We conducted a post-hoc evaluation of the cost-effectiveness of pamidronate using the results of two randomized trials that evaluated pamidronate 90 mg administered intravenously every month versus placebo. PATIENTS AND METHODS: The trials differed only in the initial systemic therapy administered (hormonal or chemotherapy). Total skeletal related events (SREs), including surgery for pathologic fracture, radiation for fracture or pain control, conservatively treated pathologic fracture, spinal cord compression, or hypercalcemia, were taken directly from the trials. Using a societal perspective, direct health care costs were assigned to each SRE. Each group’s monthly survival was equal and was projected to decline using observed median survivals. The cost of pamidronate reflected the average wholesale price of the drug plus infusion. The value or disutility of an adverse event per month was evaluated using a zero value (events avoided) or an assigned one (range, 0.2 to 0.8). RESULTS: The cost of pamidronate therapy exceeded the cost savings from prevented adverse events. The difference between the treated and placebo groups was larger with hormonal systemic therapy than with chemotherapy (additional $7,685 compared with $3,968 per woman). The projected net cost per SRE avoided was $3,940 with chemotherapy and $9,390 with hormonal therapy. The cost-effectiveness ratios were $108,200 with chemotherapy and $305,300 with hormonal therapy per quality-adjusted year. CONCLUSION: Although pamidronate is effective in preventing a feared, common adverse outcome in metastatic breast cancer, its use is associated with high incremental costs per adverse event avoided. The analysis is most sensitive to the costs of pamidronate and pathologic fractures that were asymptomatic or treated conservatively.
OVER THE LAST decade, a variety of supportive agents have been developed that benefit cancer patients. These agents, such as the selective 5-hydroxytryptamine-3 receptor antagonist antiemetics, hematopoietic growth factors, recombinant erythropoietin, and the low-molecular-weight heparins, have transformed cancer care. These agents do not directly increase survival and do not directly affect the specific cancer. However, by reducing cancer-specific symptoms or treatment-induced complications, overall quality of life is enhanced. Recent additions to this list of supportive agents are the bisphosphonates. Unfortunately, innovation is commonly associated with increased financial costs. Establishing the financial value of new medical interventions that are not associated with improved survival is a recurring challenge for cancer and noncancer therapies alike. Bisphosphonates reduce an important disease-specific complication, morbidity from bony metastases. In 1998 in the United States, approximately 44,000 women died from breast cancer. Because the median survival for women from the first detection of metastatic disease is approximately 18 to 24 months, the current prevalence of women who are alive with metastatic breast cancer exceeds 100,000 women. Bony involvement is often the only site, the initial site, and the most common symptomatic site of their metastatic disease.1 The most prominent symptom is pain, which commonly requires narcotics or radiation for control. The destructive effects of metastases on bone can lead to long bone fractures requiring surgery, vertebral fractures, and, rarely, spinal cord compression.1 Given the absolute frequency of the condition and the cost consequences of the adverse effects, the financial consequences of metastatic bony involvement are substantial. Any cost savings achieved by preventing these adverse events need to be balanced against the potential $600 to $800 million additional costs per year if all women in the United States with metastatic breast cancer received pamidronate. Pamidronate (Aredia; Novartis Pharmaceuticals Corp, East Hanover, NJ) is currently the only biphosphonate approved by the United States Food and Drug Administration for the treatment of metastatic breast cancer in conjunction with standard antineoplastic therapy in women with osteolytic bone lesions. This approval was granted predominantly on the basis of the results of two randomized controlled trials in women with lytic bone metastases. These trial results were used to conduct a post-hoc evaluation of the cost-effectiveness of pamidronate in a cohort of women as if they were entering these clinical trials. The analysis considered end points of all adverse events as well as assigned quality-of-life values for nonfatal complications. Over a 24-month time horizon, the study projected direct health care costs, cost per adverse event avoided, and cost per quality-adjusted life year (QALY).
Aredia Breast Cancer Protocols 18 and 19 We considered a hypothetical group of women meeting the entry criteria for the Aredia Breast Cancer Study Group Protocols 18 and 19 (P18 and P19).2-4 Details of the design and patient characteristics are described in the original reports. In brief, each protocol was an international, multicenter, randomized, double-blind, parallel trial of women with metastatic breast cancer with one or more osteolytic lesions at least 1 cm in diameter. All women received systemic therapy. The specific type of systemic therapy was at the physician’s discretion and was classified as hormonal (P18) or chemotherapy (P19). For the chemotherapy patients, details of prior systemic therapy were reported: approximately 52% of patients had received two or more prior chemotherapy regimens, and approximately 50% had received two or more hormonal treatments.2 Antineoplastic therapy could be changed at any time during these trials. All women had an expected survival time of at least 9 months at entry. Patients with a skeletal event or radiation in the 2 weeks before enrollment were ineligible. The primary end point of the trials was skeletal-related events (SREs), an aggregate of all bony complications. SREs included pathologic fractures, spinal cord compression or collapse, radiation for pain relief or treatment of pathologic fracture, surgery to bone, and hypercalcemia. The median age was approximately 57 years. Approximately 60% of women had bony involvement as their sole site of metastases and approximately 65% had an Eastern Cooperative Oncology Group performance score of 0 or 1. Both trials clearly showed that pamidronate was effective in reducing SREs. Fewer total SREs, less radiation therapy, and fewer pathologic fractures occurred with pamidronate administration. The median time to the first skeletal complication was increased in both studies (13.9 v 7.0 months in P19, and 11.0 v 7.0 months in P18).
The Model and Its Assumptions
2. The number of deaths per month was projected using a constant hazard rate based on the observed median survivals. The hazard rate per month was 3% for patients in the hormonal group and 4.9% for patients in the chemotherapy group. 3. Although in the trials the probability of patient refusal to continue therapy was not substantially different with pamidronate compared with placebo, a constant hazard rate of refusal or dropout was included to avoid overestimating the treatment costs of pamidronate. The observed numbers of patients completing therapy at 1 and 2 years were used as targets to guide the refusal estimates (0.5% per month in the chemotherapy group and 1.0% per month in the hormonal cohort). 4. There was no survival benefit from pamidronate. 5. Patients experienced no more than one adverse event per month. 6. Death could not occur during the month of an adverse event. 7. The cost and quality-of-life decrement for each SRE was assumed to last only 1 month. 8. Costs due to systemic therapy were unchanged by pamidronate and, therefore, excluded. 9. The costs and quality of life associated with radiation were independent of the reason for therapy (pain control, prevention or treatment of pathologic fracture). 10. The costs of all surgery for pathologic fractures were estimated from the costs of hip procedures. 11. Pathologic fractures not requiring surgery or radiation were either totally asymptomatic or required only oral analgesics. 12. When uncertain, higher estimates for the costs of each SRE were made to bias the analysis in favor of pamidronate. 13. Although most women with metastatic breast cancer have measurable out-of-pocket and lost wages due to their disease, it seems unlikely that these differ substantially with the use of pamidronate. Therefore, these costs were excluded from the model.
Costs
Pamidronate. The monthly cost of pamidronate therapy was estimated to be $775; reflecting the sum of the average wholesale price of a 90-mg vial ($621, December 1998), a 2-hour intravenous outpatient infusion (Current Procedural Terminology [CPT] codes 96410 and 96412), and preprocedure magnesium and multichannel chemistry (CPT codes 83735 and 80019). No physician visit was included. SREs. Most important health care utilization details were not tracked or reported from these trials. Specifically, data regarding the frequency and duration of hospitalizations, the relationship between specific bony complications and subsequent survival, and the rate at which SREs detected by surveillance imaging subsequently became symptomatic are not available in published results. Therefore, for any specific SRE, the frequency and duration of hospitalization and specific therapies within a treatment category were estimated. Radiation. Using the Massey database, a detailed estimate of the cost of a 2-week course of local radiation of 30 Gy over 10 fractures administered on an outpatient basis was constructed. This dose and schedule is the most commonly used regimen and is widely considered to be appropriate in the United States.6 The treatment complexity was assumed to be intermediate with two treatment areas and required multiple block wedges. The costs for the radiation were estimated from CPT codes 77262, 77285, 77300, 77334, 77336, 77409, 77417, and 77425. An additional $600 was added for the physician (office, emergency room, or hospitalization costs) and diagnostic evaluation for symptoms (pain with or without fracture). Direct nonmedical costs of $700 were added for 14 half-days of lost work for a companion to accompany the patient for therapy. The $50 per half day was based on projections by Hayman et al.6 Surgery and other complications. Fractures requiring surgery, hypercalcemia, and cord compression would each require hospitalization. In the last 3 years, only 10 cancer patients were identified in the Massey database as having orthopedic procedures of joint fixation, fusion, or replacement due to International Classification of Diseases, 9th Revision (ICD-9) 733.1 "pathologic fracture," ICD-9 198 "secondary malignant neoplasm of other specified sites," or ICD-9 203 "multiple myeloma." In addition, the complications of hypercalcemia or spinal cord compression could not be separated within the broad category of metastatic breast cancer. Therefore, the estimates were made on the basis of expert opinion and data listed in Table 3. The costs for all forms of surgery were estimated from the costs for total hip replacement. It was assumed that spinal cord compression would require, at least initially, inpatient radiation therapy. Severe hypercalcemia has a poor prognosis. A recent series found a median survival of 1.4 to 2.2 months after the first episode of hypercalcemia in breast cancer patients.8 Therefore, the costs and 4-day median hospitalization for "therapeutic procedures secondary to malignancy" listed in Table 3 were used to anchor the costs of hypercalcemia. Costs for fractures not requiring radiation or surgery were projected to prompt additional outpatient visits: three outpatient visits with direct nonmedical costs, two radiographs, chemistries, and oral pain medications. Using physician relative units and local Medicare payments, the aggregate cost estimate was $600.
Quality-of-Life Assigned Values
End Points and Cost-Effectiveness
The model’s simulated baseline results for each cohort (with and without pamidronate in each systemic therapy group) are listed in Table 4. The model projected that, on the average, a patient would receive pamidronate for 13.6 months in the chemotherapy group and 15.6 months in the hormonal group. The model, compared with the observed data, slightly but minimally underestimated the number of pamidronate-treated patients: at 1 year, there were 95 versus 99 patients, respectively, in the chemotherapy group and 107 versus 113 patients, respectively, in the hormonal group.2-4 Survival at 24 months closely approximated that observed in the trials.
In both systemic therapy groups, the total costs associated with administering pamidronate exceeded the cost savings from avoided SREs. In the chemotherapy group, pamidronate reduced the costs associated with SREs by approximately 50%. However, because of the cost of pamidronate, total net costs per pamidronate patient were $17,906 compared with $13,938 for placebo, representing a net increase of $3,936. In the hormonal therapy group, the cost savings from SREs avoided were less and treatment costs were higher, because patients lived longer. Therefore, total net costs per pamidronate patient were $20,319 compared with $12,634 for patients receiving placebo in the hormonal group, representing a net increase of $7,685. The number of SREs was taken directly from the trial results. The average number of total SRE-free months in the chemotherapy group increased by 1.13 months (from 11.07 to 12.20 months) with pamidronate. In the hormonal therapy group, the increase with pamidronate in SRE-free months was smaller, at 0.82 months. When the disutility of SREs was taken into account, pamidronate was associated with a gain of 0.026 QALYs (or approximately 9 quality-adjusted days) for patients in the hormonal group and 0.037 QALYs (approximately 13 quality-adjusted days) for patients in the chemotherapy group. Varieties of ratios related to the cost-effectiveness or cost per event were determined. The cost to prevent one SRE is the cost per patient for pamidronate while ignoring the costs of treating an SRE. The cost to prevent one SRE was $9,350 in the chemotherapy group and $12,760 in the hormonal therapy cohort. Because pamidronate prevented total SREs in addition to more expensive SREs, the cost to treat an SRE, if one occurred, was lower with pamidronate. The net total costs for preventing an SRE and its subsequent treatment was $3,940 in the chemotherapy group and $9,390 in the hormonal therapy cohort. The cost-effectiveness ratios for pamidronate were estimated with a worst case, using the assigned health-state utility values, and a best case, using SRE event months avoided (ie, setting the utility of an SRE event month to zero). The cost-effectiveness ratios associated with pamidronate ranged from $108,200 per QALY in the chemotherapy group to more than $300,000 per QALY in the hormonal therapy group. If the utility or value of any month associated with any SRE was zero, then the costs per QALY were $47,280 for the chemotherapy group and $112,260 for the hormonal therapy group.
Sensitivity Analyses
Pooling. Berenson and Lipton10 recently assessed the homogeneity of results between the two studies. They found that the odds ratio of having an SRE (except for hypercalcemia) while receiving placebo as compared with pamidronate in the hormonal therapy study did not differ significantly from the odds ratio in the chemotherapy study. Therefore, they justified the pooling of data from the two studies. The substantial difference between the two baseline cost-effectiveness estimates is driven by the differences in the number needed to treat to prevent one SRE. Therefore, the methodologic foundation for pooling the cost-effectiveness is weaker. Nevertheless, a combined analysis using all 762 patients found that cost-effectiveness of pamidronate was $175,200 per QALY and a net $5,860 per SRE avoided. Symptomatic events. In the baseline analysis, a disutility was applied to any SRE. However, many of the SREs may have been asymptomatic, because many if not most fractures were vertebral or rib fractures that were detected as part of the every-3-month radiographic survey. We therefore conducted analyses in which costs and reduced quality of life were limited to SREs requiring active therapy. In this analysis, the costs per QALY increased to greater than $134,700 per QALY and greater than $3,480 per month per SRE avoided. Hypercalcemia. Hypercalcemia may not be considered an SRE versus a systemic consequence of metastatic disease. Because pamidronate is currently the standard of care for treating this complication, it is not surprising that pamidronate was also quite effective in its prevention. Available reports do not describe the subsequent prognosis of patients with hypercalcemia. When hypercalcemia was excluded, the cost-effectiveness ratio almost doubled to $187,900 per QALY and increased to $5,220 per SRE avoided. Costs. If the costs of pamidronate were reduced 50% ($378), then the pamidronate strategy would be dominant: it would have lower costs and higher quality-adjusted survival. From a cost-minimization perspective, the threshold or break-even point for pamidronate costs, including infusion, was $484 per monthly treatment. Applying the commonly used threshold of $50,000 per QALY,11,12 pamidronate’s costs would need to decrease from $775 to $618. If the highest point estimates for costs for each individual SRE (listed in Table 2) were used, then pamidronate would again be a dominating strategy. Quality-of-life values. If the true duration of adverse quality of life associated with symptomatic SREs was greater than 1 month, then the model may have underestimated the benefit of avoiding SREs. The effect of increasing the adverse quality of life for symptomatic SREs from 1 to 2 months was evaluated. This arbitrary analysis required ignoring the possibility that patients could die during the month immediately after an SRE. This combination of assumptions lowered the cost-effectiveness ratios to approximately $77,300 per QALY, using a 2-month duration. If the utility scores for all symptomatic SREs were zero, then the cost-effectiveness ratio was $60,100 per QALY. If a modest quality-of-life penalty for receiving pamidronate were included (a 1-day reduction of 20%), then the cost-utility ratio increased to $134,100 per QALY. Treatment refusals. Now that the effectiveness of pamidronate has been established, patients may be less likely to opt to discontinue therapy then were patients in the trials. Therefore, the analysis was repeated assuming pamidronate was continued in all patients until death. Under this assumption, the incremental cost-effectiveness ratios increased to $122,200 per QALY for the chemotherapy cohort and to greater than $360,000 per QALY in the hormonal therapy cohort.
Pamidronate as supportive therapy for women with osteolytic bone metastases delays the onset of skeletal complications and reduces their overall number. In addition, in the study among women with bone pain at entry using chemotherapy as systemic therapy, pamidronate reduced their bone pain. Overall, Eastern Cooperative Oncology Group performance status was better maintained, but the extent of the benefit was not described in published reports. However, overall survival and overall quality of life, as measured with the Spitzer quality-of-life index, were not changed. Therefore, the costs and modest inconvenience of pamidronate are important concerns that must be balanced against these benefits. The financial costs of a supportive agent depend on the product’s unit cost plus the costs of delivery and the frequency with which it is used. Most cytotoxic chemotherapy or radiation can be administered over a relatively short interval and is discontinued if there is evidence of cancer progression. In contrast, many supportive agents need to be given over a prolonged or indefinite period to prevent specific adverse events. In addition, continuing the supportive agent after the first adverse event occurs is often standard practice to prevent subsequent events or reduce the severity of the initial adverse event. Pamidronate was initially approved by the United States Food and Drug Administration for treatment of hypercalcemia associated with advanced stages of malignancy. For this indication, most patients require only one or two treatments, which is likely to have been an important factor in its initial pricing. Pamidronate for women with osteolytic bone metastases requires prolonged use, which increases the total costs. A complex Markov model for this analysis was unnecessary, because primary data from well-designed, appropriately sized randomized trials are available, no major difference in survival was observed, and the short life expectancies would lead to similar results of either approach. By projecting the costs, assigning disutilities to months with SREs, and conducting a sensitivity analysis, we attempted to assist oncology providers in the difficult task of prioritizing limited financial funding, especially from a disease-management perspective. In this post-hoc economic assessment of the two manufacturer-sponsored multinational trials, the costs of pamidronate were projected to greatly exceed the cost savings associated with preventing SREs. Using assigned utility values that reflect overall quality of life, the projected costs per added QALY with pamidronate were higher than those of most commonly accepted medical interventions. Because an extensive set of assumptions was made in constructing this post-hoc model, the analysis was intentionally biased toward pamidronate by including relatively high cost estimates for the SREs requiring medical interventions, assigning a $600 cost for evaluation and management of pathologic fractures that seemed to be asymptomatic, and assigning low utility scores for SREs. Even so, the cost-effectiveness ratio ranged from more than $100,000 to $300,000 per QALY. This model was inherently limited by its post-hoc assessment based on a synthesis of the published reports. However, the investigators were unsuccessful in negotiating access to the data with pamidronate’s manufacturer, Novartis. Therefore, the assumptions listed in Patients and Methods (under The Model and Its Assumptions) had to be made. The model is most sensitive to the cost of pamidronate per month, the costs and quality of life associated with fractures not requiring radiation or surgery, and the duration of the reduced quality of life if SREs occurred. Eighty percent of the projected costs of pamidronate per treatment were due to the drug costs itself. Pamidronate’s costs could be reduced if (1) the infusion time were reduced, (2) pamidronate’s dosage could be reduced, (3) the treatment intervals could be increased, or (4) the unit price was decreased. More research is needed to improve the estimates of the direct and indirect cost consequences of bony complications of breast cancer. This will require prospective data collection, as retrospective review of administrative claims is unlikely to be able to distinguish between the specific bony complications or concurrent bony and visceral disease. Using estimates of patient utilities from expert opinion is always a limitation. However, the incremental cost-effectiveness ratios are still large in using the extreme penalty of equating a month with an SRE to death. Future prospective assessments need to address the duration of symptoms associated with these bony complications and their relationship to patient mortality. Cost does not seem to be a major barrier to pamidronate’s use: sales increased 80% in 1997 and 61% in 1998, suggesting that the worldwide market is rapidly adopting this therapy.13 The economic assessments of supportive care in oncology that do not directly lead to improved survival have been inconsistent in finding that the supportive care is cost-effective.14-18 The work described here suggests that a retrospective review is warranted to assess how many patients had asymptomatic pathologic fractures versus symptoms that were controlled with analgesics or diffuse bony lesions that were not amenable to radiation. Individual oncology practitioners should not deny pamidronate to women on the basis of this analysis. How should health plans, at-risk multispecialty groups, and individual oncologists use these results in their purchasing negotiations for pamidronate? It is difficult to foresee any of these groups having sufficient leverage against the world’s third-largest pharmaceutical corporation to demand a substantial price reduction for pamidronate. In addition, patients are unlikely to accept being denied the drug for cost reasons. Pending further refinements in defining the quality of life of women with bony complications compared with other systemic manifestations of metastatic breast cancer, these results suggest that the incremental benefits associated with pamidronate come at a high financial price. Only time will tell if market forces and this type of cost-effectiveness analysis will impact pricing.
Supported in part by a Faculty Research Award from the American Cancer Society.
1. Harris JR, Morrow M, Norton L: Malignant tumors of the breast, in Devita VT, Hellman S, Rosenberg SA. (eds): Cancer: Principles and Practice of Oncology (ed 5PA,Lippincott, 1997, pp 1557-1616
2.
Hortobagyi GN, Theriault RL, Lipton A, et al: Efficacy of pamidronate in reducing skeletal complications in patients with breast cancer and lytic bone metastases: Protocol 19 Aredia Breast Cancer Study Group [see comments]. Med 335:1785-1791, 1996 3. Hortobagyi GN, Theriault RL, Lipton A, et al: Long-term prevention of skeletal complications of metastatic breast cancer with pamidronate: Protocol 19 Aredia Breast Cancer Study Group. J Clin Oncol 16:2038-2044, 1998[Abstract]
4.
Theriault RL, Lipton A, Hortobagyi GN, et al: Pamidronate reduces skeletal morbidity in women with advanced breast cancer and lytic bone lesions: A randomized, placebo-controlled trial—Protocol 18 Aredia Breast Cancer Study Group. J Clin Oncol 17:846-854, 1999 5. United States Agency for Health Care Policy and Research: Healthcare Cost and Utilization Project. Http:www.ahcpr.gov/data/hcup/hcupnet.htm 6. Rose CM, Kagan AR: The final report of the expert panel for the radiation oncology bone metastasis work group of the American College of Radiology. Oncol Biol Phys 40:1117-1124, 1998 7. Hayman JA, Hillner BE, Harris JR, et al: Cost-effectiveness of routine radiation therapy following conservative surgery for early-stage breast cancer. J Clin Oncol 16:1022-1029, 1998[Abstract] 8. Kristensen B, Ejlertsen B, Mouridsen HT, et al: Survival in breast cancer patients after the first episode of hypercalcaemia. Med 244:189-198, 1998 9. Weeks J: Taking quality of life into account in health economic analyses. J Natl Cancer Inst Monogr 23-27, 1996 10. Berenson JR, Lipton A: Use of bisphosphonates in patients with metastatic bone disease. (Huntingt) 12:1573-1579, 1998
11.
Edelson JT, Weinstein MC, Williams L, et al: Long-term cost-effectiveness of various initial monotherapies for mild to moderate hypertension. JAMA 263:407-413, 1990
12.
Smith TJ, Hillner BE, Desch CE: Efficacy and cost-effectiveness of cancer treatment: Rational allocation of resources based on decision analysis. J Natl Cancer Inst 85:1460-1474, 1993 13. Novartis Pharmaceutical Corp: 1998 Annual Report. Http:www.novartis.com 14. Weeks JC, Tierney MR, Weinstein MC: Cost effectiveness of prophylactic intravenous immune globulin in chronic lymphocytic leukemia. N Engl J Med 325:81-86, 1991[Abstract] 15. Roila F, Ballatori E: The efficacy and cost-effectiveness of various antiemetic regimens. Curr Opin Oncol 10:310-315, 1998[Medline]
16.
Smith TJ, Hillner BE, Schmitz N, et al: Economic analysis of a randomized clinical trial to compare filgrastim-mobilized peripheral-blood progenitor-cell transplantation and autologous bone marrow transplantation in patients with Hodgkin’s and non-Hodgkin’s lymphoma. J Clin Oncol 15:5-10, 1997 17. Silber JH, Fridman M, Shpilsky A, et al: Modeling the cost-effectiveness of granulocyte colony-stimulating factor use in early-stage breast cancer. J Clin Oncol 16:2435-2444, 1998[Abstract] 18. Barosi G, Marchetti M, Liberato NL: Cost-effectiveness of recombinant human erythropoietin in the prevention of chemotherapy-induced anaemia. Br J Cancer 78:781-787, 1998[Medline] Submitted December 21, 1998; accepted August 1, 1999.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2000 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
