|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2002 American Society for Clinical Oncology Decision-Analytic Model and Cost-Effectiveness Evaluation of Postmastectomy Radiation Therapy in High-Risk Premenopausal Breast Cancer PatientsByFrom the Department of Radiation Oncology, Division of General Internal Medicine, and the Leonard Davis Institute of Health Economics, University of Pennsylvania Medical Center, Philadelphia, PA, and Department of Radiation Oncology, University of Michigan Medical Center, Ann Arbor, MI. Address reprint requests to Jason H. Lee, MD, Hudner Oncology Center, 795 Middle St, Fall River, MA 02721; email: jasonlee{at}massmed.org
PURPOSE: To present a decision model that describes the clinical and economic outcomes of node-positive breast cancer with and without postmastectomy radiation therapy (PMRT). METHODS: A Markov process was constructed to project the natural history of breast cancer following mastectomy in premenopausal node-positive women. Biannual hazards of local and distant recurrence without PMRT were derived from a large meta-analysis of adjuvant systemic therapy trials for breast cancer. The addition of PMRT reduced the risk of disease relapse by an odds ratio of 0.69. Costs of PMRT ($11,600) and recurrent breast cancer ($4,250 to 16,200/year) were estimated from available literature. The model projected number of recurrences, relapse-free and overall survival, and costs to 15 years, using a discount rate of 3%. Cost-effectiveness ratios were calculated per incremental year of life and quality-adjusted year of life gained. One- and two-way sensitivity analyses were performed to determine the sensitivity of results to clinical and economic assumptions. RESULTS: The model projected 15-year relapse-free survival of 52% and 43% with and without PMRT, respectively. Overall survival was increased from 48% to 55% with PMRT, resulting in an incremental 0.29 years of life gained per subject. PMRT increased 15-year costs from $40,800 to $48,100. Cost per year of life gained was $24,900, or $22,600 when survival was adjusted for quality of life. Results of the model were relatively sensitive to radiation therapy cost and breast cancer relapse risk. CONCLUSION: This analysis suggests that PMRT offers substantial clinical benefits achieved in a cost-effective manner, with an average cost per year of life gained of $24,900. Results of the model were robust under a wide range of clinical and economic parameters.
MORE THAN 190,000 women are diagnosed with breast cancer annually in the United States (US), and approximately 40,000 patients will die of their disease.1 Medical costs attributable to the diagnosis and treatment of breast cancer have been estimated to be greater than $2.5 billion in 1997.2 A significant proportion of women will receive radiation therapy during their course of disease, either as part of definitive management or as a palliative measure. Despite the widespread adoption of breast conservation therapy, many women with operable breast cancer undergo mastectomy as primary treatment. The use of radiation therapy following mastectomy has generally been limited to the treatment of patients with multiple positive axillary nodes, advanced primary disease, advanced nodal disease, or involved surgical margins.3-6 Three recently reported randomized trials have suggested a survival benefit with postmastectomy radiation therapy (PMRT) in high-risk breast cancer patients.7-9 These trials provide evidence that PMRT is beneficial under a broader set of indications; however, the validity of their results has been under considerable debate in the US. Differences in surgical technique, particularly axillary node dissection, and intensity of chemotherapy are two factors that may partially explain why the relatively large benefit in PMRT was observed in these non-US trials. The economic implications of PMRT compared with its incremental benefits have gained increasing attention.10-12 Economic analyses are considered appropriate when alternative management strategies for a common clinical entity may result in differing outcomes and may have substantial financial implications.13 Cost-effectiveness analysis assesses the costs of a treatment that are expended to achieve a measurable health outcome, such as number of years of life gained.14 Many consider radiation therapy to be expensive, especially the complex treatments necessary to deliver multiple fields typically administered in the postmastectomy setting.15 Robust cost-effectiveness analysis of PMRT may further justify the expense of such treatment. We have previously reported preliminary results of a decision-analytic model using clinical details from two selected trials.16 In this analysis, we used the results from recent meta-analyses to assess the incremental cost-effectiveness of PMRT over mastectomy alone.
Model Overview We constructed a Markov model to describe the clinical history of breast carcinoma following mastectomy in a hypothetical cohort of lymph node–positive premenopausal women. In this type of simulation, subjects make transitions at discrete time intervals among health states relevant to the clinical problem.17 As subjects make these transitions, they incur costs and gain years of life. These costs and life-years are aggregated for each treatment approach being evaluated and are compared with each other. This model was used to compare the costs and years of life associated with mastectomy and chemotherapy alone versus mastectomy, chemotherapy, and PMRT in node-positive 45-year-old women. The first step of this analysis involved the development of a natural history model of breast cancer following mastectomy and no radiation therapy (base case scenario). Health states used in the model described conditions of being well, having recurrent local-regional and/or distant disease, and undergoing successful salvage therapy after isolated local-regional failure (LRF), as well as death from breast cancer and death from other causes (Fig 1). States were characterized by transition probabilities into other states, costs, and utilities for quality of life. Next, the impact of PMRT (intervention scenario) was modeled by adjusting the risk of recurrence in the base case scenario. Finally, the validity and robustness of the model were tested using one- and two-way sensitivity analyses and Monte Carlo simulation. The model was developed as a Microsoft Excel (Microsoft Corp, Redmond, WA) spreadsheet and, for validation purposes, simultaneously designed on DATA 3.5 (TreeAge Software, Williamstown, MA; .tre file available from authors on request).
Breast Cancer Model The simulation begins after subjects have undergone potentially curative mastectomy for operable breast carcinoma and begin to receive adjuvant therapy. The principal analysis of the model focuses on premenopausal women with node-positive disease. In the base case scenario, these patients are eligible for adjuvant polychemotherapy, and the model assumes all subjects receive systemic treatment. At 6-month intervals, subjects may undergo transitions among the health states listed in Table 1. All subjects start in a "well" state that represents no evidence of recurrent disease. During each period, subjects may remain in their current health state or make a transition to another health state. Target health states from the well state include those involving disease relapse and death from non–breast cancer mortality. After disease relapse, some may undergo successful salvage therapy and others can progress and ultimately die of disease. Subjects who die during the simulation, either from breast cancer or other causes, make no further transitions. For purposes of simplification, the model does not consider risks of contralateral breast cancer. The state transition model is depicted in Fig 1.
Risk of relapse. The probability of disease relapse without radiation therapy was derived from the relapse-free survival curve from the Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) 1998 meta-analysis of adjuvant systemic therapy.18 In this meta-analysis, the survival free of breast cancer relapse for node-positive women under 50 years of age who received polychemotherapy was 57.1% at 5 years and 47.6% at 10 years. Although this curve may include some women who did in fact receive radiation therapy, we arbitrarily defined this outcome as the relapse-free survival in the base case scenario without radiation. The annual hazard of disease relapse is estimated from this curve and listed in Table 2. The intermediate points on the curve and extrapolation beyond 10 years was estimated using a fitted exponential function. These hazards are similar to the risk of relapse observed in Eastern Cooperative Oncology Group chemotherapy trials.19
The model divides patients with disease relapse into three categories: (1) those with isolated LRF, (2) those with only distant disease, and (3) those with combined local-regional and distant disease. The distribution of failure sites was estimated from the control arm of prospective trials of PMRT and retrospective studies of patterns of failure following mastectomy alone.7,9,20-26 Isolated LRF was defined as recurrent disease involving only the chest wall and/or regional nodes. The proportion of initial relapse that is isolated LRF ranges from 19% to 49%. The proportion of initial relapse that is distant disease only ranges from 41% to 62%. The proportion of initial relapse that is combined local-regional and distant disease ranges from 8% to 23%. For the purposes of this model, the proportion of initial failures that were only local-regional, only distant, and combined was 30%, 54%, and 16%, respectively. These values were obtained from a weighted average of results from prospective clinical trials. The model assumes that the pattern of failure (ie, LRF, distant disease, or combined disease) is constant over the course of the simulation. The parameters used to describe outcome of initial disease relapse in the model are summarized in Table 3.
Salvage of LRF. Many patients who experience isolated LRF may be rendered disease-free and experience prolonged survival with surgical excision and salvage chest wall irradiation. We assumed all subjects with isolated LRF who received only adjuvant polychemotherapy were undergoing salvage treatment ("treatable"), including chest wall irradiation, which results in a complete response in 80% of cases.27-30 Subjects who do not achieve a complete response (20%) enter a state of persistent local-regional disease and may experience further progression of disease, as described below. The decision tree in Fig 2 depicts the potential outcomes after initial disease relapse following no PMRT.
Women who achieve a complete response to treatment are considered "salvaged" (Fig 1) but may subsequently develop a second relapse that is either local-regional, distant, or both. The risk of secondary failure varies depending on the initial disease-free interval, location and extent of recurrence, and type of salvage treatment. Most reports indicate that overall survival is approximately 20% to 50% at 5 years, freedom from second recurrence is 0% to 40%, and the majority of relapses are distant.28-34 Since patients may experience a second relapse of disease despite salvage local-regional treatment, we estimated an annual risk of 6% for secondary LRF, 15% for distant disease, and 6% for combined disease. These basic hazards correspond to a 5-year relapse-free survival of approximately 22% and overall survival of 44%. Subjects who relapse after salvage therapy may experience further progression and death from disease as described below. Progression of recurrent disease. Once a woman develops local-regional, distant, or combined disease, she remains at high risk for cancer progression and death from breast cancer. Subjects with isolated LRF who cannot be successfully treated with salvage therapy may develop distant disease or die of local-regional progression. Subjects with either distant disease or combined disease may also die of breast cancer. The model assumes that after disease recurs distantly, no patient can be successfully treated with salvage therapy or return to the well state. The risk of disease progression is strongly influenced by the initial disease-free interval.34-36 Because early recurrence portends worse prognosis, the relative risks of distant dissemination and death due to breast cancer are higher in the first 5 years of the simulation compared with later years. The annual probabilities of disease progression and death due to disease obtained are listed in Table 4.
Other-cause mortality. Subjects in the simulation may also die from other causes. The annual risk of death was obtained from 1996 US vital statistics data.37 For women aged 45 to 60 years, the annual mortality rate is typically less than 1.0% (range, 0.22% to 0.85%) per year. Although this mortality rate includes deaths due to breast cancer, the absolute proportion is small and should not affect the results of the model.
Modeling the Impact of PMRT The impact of PMRT was modeled as a risk reduction in disease relapse, based on the odds ratio reported in the meta-analysis by Whelan et al.39 Most trials in this meta-analysis included patients with node-positive disease who underwent mastectomy and received adjuvant polychemotherapy. Except for the inclusion of postmenopausal patients, Whelan et al’s study population is fairly similar to the hypothetical cohort in the model reported here. Whelan et al calculated an odds ratio of any disease relapse of 0.69 at the combined median follow-up time of the eligible trials. For purposes of this model, we adjusted the 10-year relapse-free survival without PMRT (47.6%) by an odds ratio of 0.69. Therefore, the 10-year relapse-free survival with PMRT was estimated to be 56.8%. We again fitted a parametric exponential function to obtain intermediate points on the curve and extrapolation beyond 10 years. The annual hazard of relapse with PMRT is listed in Table 2. Recurrent disease after PMRT. The patterns of failure and course of disease relapse after a patient has received PMRT were also modified compared with the base case. The proportion of subjects with isolated LRF, distant disease, and combined disease after PMRT is 10%, 82%, and 8%, respectively,7,9,20-23 obtained from a weighted average of results from clinical trials. Although there are few data on the ability to successfully salvage isolated LRFs after PMRT, we assumed the success of subsequent local therapies is reduced. We assumed that 75% of patients with isolated LRF would receive additional local therapy, but only 40% of these patients would be rendered disease-free. Patients who do not receive additional local therapy or are treated unsuccessfully remain in a state of LRF or develop distant dissemination.42 The second decision tree in Fig 2 depicts the potential outcomes after initial disease relapse following PMRT. The model assumes that the rate of disease progression after PMRT is similar to that without PMRT. Therefore, the rates of distant dissemination and death from breast cancer shown in Table 4 are also used for the intervention scenario. The principal differences between the base case scenario and the intervention scenario are the risk of disease relapse, patterns of failure, and probability of salvage of isolated LRF. Although the number of deaths due to breast cancer is reduced by PMRT, there is some evidence that the rate of other-cause mortality is increased by PMRT, thus negating its beneficial effect in early meta-analyses.43,44 Modern radiation therapy linear accelerators and careful treatment planning are thought to minimize the excess risk of PMRT on cardiac and pulmonary deaths.45 The most recent EBCTCG meta-analysis of radiation therapy suggests that the relative risk of other-cause mortality is 1.182, mostly observed 10 years after treatment.38 To allow for this potential increase in cardiopulmonary mortality, the annual risk of other-cause mortality is increased 18% beyond year 10 for women receiving upfront PMRT.
Economic and Quality-of-Life Analysis
During the simulation, costs were aggregated based on the number of subjects in each health state and the cost assigned to that health state. For the intervention scenario, all subjects incurred the cost of PMRT ($11,600). By avoiding local-regional and distant failures, PMRT has the potential of avoiding downstream costs of care. The incremental cost-effectiveness of PMRT was calculated as a ratio of the difference in aggregate costs between the intervention and base case scenarios over the difference in aggregate years of life (ie, the cost per year of life gained). The aggregate years of life for each treatment scenario were calculated as an area under the survival curve for the time horizon in the model. The incremental cost per LRF averted is also reported. Cost-effectiveness ratios are reported for 10- and 15-year time horizons. Costs and nonmonetary benefits were discounted at 3% per year. Quality of life. In the principal analysis, survival was not adjusted for quality of life, which can be affected by the morbidity of breast cancer itself or its treatment. As a secondary analysis, we report the ratio of cost per quality-adjusted life-year (QALY) gained. Utilities were estimated from published quality-of-life studies and prior decision models and are listed in Table 5.47,56-64 The methods used to obtain utility values from individual studies include time trade-off,62-64 standard gamble,60,61 and a visual analog scale.47,56,58,59 Although chest wall and nodal radiation therapy is associated with troublesome side effects, such as dermatitis, fibrosis, lymphedema, and rare instances of plexopathy and rib fracture, most symptoms are self-limited. We assumed similar utility values for the well health state after either chemotherapy or chemotherapy and PMRT, and explored the effect of differing utility values for the latter state.
Validation and Sensitivity Analysis
Survival Outcomes In the principal analysis, the model calculated a 47% relapse-free survival and 59% overall survival at 10 years without PMRT. By extrapolation to 15 years, relapse-free and overall survival rates were 43% and 49%, respectively. Of the 51 deaths occurring per 100 subjects, 47 were related to breast cancer and four were due to other causes. The 10- and 15-year clinical results of the model are summarized in Table 6.
In the intervention scenario, the 15-year relapse-free and overall survival rates were 52% and 55%, respectively. Compared with no PMRT, the addition of PMRT resulted in a 6% increase in overall survival, or a gain of 0.29 years of life over a 15 year time horizon. Undiscounted years of life were increased from 10.52 without PMRT to 10.93 with PMRT. Among 100 subjects, 45 deaths were expected at 15 years, with five due to other causes and the remainder due to breast cancer. Overall survival and relapse-free survival curves for the base case and intervention scenarios are shown in Figs 3 and 4, respectively.
At 10 years, the difference in relapse-free survival is 9% between the base case and intervention scenarios. According to the model, the odds ratio for breast cancer recurrence with the addition of PMRT is 0.70. The odds ratio for death with the addition of PMRT is 0.84 at 10 years and 0.78 at 15 years.
Patterns of Failure The addition of PMRT alters the total number and distribution of failures. For every 100 subjects, approximately 44 subjects experienced recurrent disease at 15 years with PMRT. Isolated LRF occurred in four subjects, and the total number of LRFs was eight. Distant disease was the initial site of failure in 40 subjects, and of these, four had combined failure. PMRT reduced the number of LRFs by 17 for every 100 patients treated. The cumulative risk of total LRF for the base case and intervention scenarios is shown in Fig 4.
Economic and Quality-of-Life Outcomes The administration of routine PMRT increased the estimated 15-year cost per subject to $48,100, a net increase of $7,300 compared with the no PMRT scenario. The average first-year cost was $22,000 (majority due to chemotherapy and radiation therapy), and continuing costs were approximately $2,400 annually. The average cost of recurrent breast cancer was $40,400 per subject with relapsed disease. A breakdown of costs for the base case and intervention scenarios is shown in Fig 5.
Given a gain of 0.29 life-years at an incremental cost of $7,300, the cost-effectiveness of PMRT was estimated to be $24,900. As a secondary analysis, survival time was adjusted for quality of life. Quality-adjusted survival was 7.03 years without radiation and 7.35 years with radiation. This resulted in a cost per QALY gained equal to $22,600. Finally, PMRT reduced LRFs by 16%, for a cost per LRF avoided of $43,400.
Sensitivity Analysis
A plot of incremental costs versus QALYs, as obtained from the Monte Carlo simulation, is shown in Fig 7. The mean cost-effectiveness ratio was $28,500, with 95% of values ranging from $7,100 to $72,000. The dotted diagonal line in Fig 7 represents a cost-effectiveness threshold of $50,000 per QALY, and approximately 90% of replicates lie below this line.
We have developed a Markov model of node-positive breast cancer based on the annual rates of relapse, patterns of relapse, and an estimation of the natural history of recurrent breast cancer. The rate and patterns of relapse were altered in the model to reflect the impact of PMRT. Given these parameters, our model can project the clinical outcomes of node-positive breast cancer with and without PMRT in a robust manner and thereby assess the incremental value of chest wall and nodal irradiation. The actuarial curves representing LRF, relapse-free survival, and overall survival are consistent with those reported in actual randomized trials. Local-regional recurrence in the Eastern Cooperative Oncology Group trial was 25% without PMRT and 15% with PMRT (9% in those who actually received treatment).20 In the British Columbia postmastectomy trial, 10-year disease-free and overall survival rates were 56% and 64% with PMRT and 41% and 54% without it, respectively.8 Our model estimated a 4% improvement in survival at 10 years, similar to the absolute survival benefit reported by the EBCTCG meta-analysis of radiotherapy for node-positive women under 50 years of age.38 A true overall survival benefit of PMRT has been substantiated by recent randomized clinical trials7-9 and other meta-analyses.39,41 PMRT produced an average survival benefit of approximately 3.5 discounted months (0.29 years) of life over a time horizon of 15 years. Although a 3.5-month incremental survival gain may seem small, this is consistent with the benefit of other adjuvant interventions for breast cancer. For node-negative breast cancer, tamoxifen produces a benefit of 3.5 to 5.2 months,66 and chemotherapy produces a benefit of 4.0 to 5.1 months.46 Breast irradiation after breast-conserving surgery improves quality-adjusted survival by 0.35 years.67 Because the majority of the survival benefit of PMRT seems to occur beyond 5 years of follow-up, it is possible that survival projections into a longer time horizon may further improve the incremental gain in life-years. However, clinical data beyond 15 years for women receiving chest wall and nodal irradiation are lacking. Under the baseline assumptions in the model, the cost-effectiveness of PMRT was $24,900 per year of life gained. With adjustments for quality of life, the cost-effectiveness ratio improved to $22,600 per QALY as a result of decreased survival time with lower utilities from recurrent breast cancer. This ratio compares favorably to other widely adopted medical and cancer therapies. The cost-effectiveness of breast irradiation for early-stage cancer was $28,000 per QALY.67 The cost-effectiveness of chemotherapy in the node-negative setting is $15,400 for younger patients46 and $44,000 for elderly patients.68 Tamoxifen was found to be highly cost-effective, with a ratio of $4,300 to $11,400 under varying assumptions.66 Axillary lymph node dissection for clinically node-negative patients had a cost-effectiveness ratio of $36,700 per QALY.69 On the other hand, an electron-beam boost after breast tangential irradiation cost over $300,000 per QALY.70 Other studies have evaluated the economics of PMRT using approaches distinct from our decision-analytic technique. Marks et al12 related local recurrence to the number of involved axillary lymph nodes and, on the basis of a number of randomized trials, predicted a 67% reduction in recurrence with PMRT. The upfront costs of PMRT, lump sum costs of local and system recurrence, and quality of life were considered in the model. The model was computed assuming an absolute survival improvement of 1% to 12% at 10 years. The cost per QALY of PMRT was generally below $50,000 for survival gains of 6% or more, and the cost per local failure avoided was $25,000 to $130,000 depending on the number of positive nodes.12 Even for modest survival gains of 3% or less, the cost-effectiveness of PMRT ranged from $10,000 to $110,000 per QALY. One limitation in this analysis is the detachment of extent of nodal positivity, and therefore risk of relapse and net benefit of radiotherapy, from gains in survival. A "spreadsheet" analysis of the costs of PMRT was reported by Dunscombe et al.11 They identified various costs components of four-field chest wall and nodal irradiation (50 Gy) and aggregated these to be $7,000 (1997 Canadian dollars). An incremental survival gain (0.5 discounted years) was taken from two selected positive randomized trials.7,8 The cost-effectiveness was estimated to be $14,000 per year of life gained, with a range up to $28,000 in the setting of pessimistic cost assumptions. Downstream costs and effect of recurrent disease on quality of life were not considered. A second Canadian study evaluated the impact of PMRT on a national scale using a population-based simulation model.10 Disease-free and overall survival gains were derived from the Danish 82b trial, and costs were collected from a Canadian societal perspective (1995 Canadian dollars). This model calculated a survival benefit of 3.2 years in node-positive patients and a cost savings of over $2,000 per case. Regardless of type of initial surgery or number of positive nodes, radiation therapy was believed to result in cost savings due to averted costs of recurrence and terminal care. Discounting of downstream costs or survival time was not performed in obtaining these initial results. Furthermore, the survival gain of 3.2 years seems to represent the difference in median survival with and without PMRT, which may overestimate survival benefits compared with an "area-under-the-curve" approximation. It is hoped that the final report of this analysis will address some of these issues. Quality of life in addition to survival is becoming an increasingly important end point in clinical trials and decision analyses. The judicious administration of PMRT in order to minimize the chances of debilitating morbidity due to progressive local-regional disease71 is a legitimate use of this intervention, regardless of effect on survival. The decrement in quality of life as a result of LRF or potential long-term toxicity from chest well and nodal irradiation has not been well quantified. In our model, we assumed equivalent utilities (0.85) for subjects who were free of disease after chemotherapy alone or chemotherapy and PMRT. Aside from the acute morbidity and inconvenience of 5 to 6 weeks of radiotherapy, PMRT may result in long-term morbidity, such as lymphedema and tissue fibrosis, particularly after reconstructive surgery. Severe chronic morbidity from PMRT was not modeled as a separate health state because the frequency and severity of morbidity vary, and a utility value for this group of conditions is difficult to estimate. At the same time, widespread progressive local-regional disease, such as in the case of "carcinoma en cuirasse," was not assigned a separate health state from small-volume local-regional disease. In the unlikely circumstance that women perceive a substantial decrease in utility of 0.82 or less from potential effects of PMRT (compared with 0.85 without PMRT), the cost-effectiveness ratio of adjuvant radiotherapy rapidly worsens. On the other hand, there is evidence that patients fear situations that place them at higher risk of breast cancer recurrence and accept the risks and inconveniences of treatments, such as radiotherapy, that enhance local control.55 Clearly, the issue of quality of life requires further study. The calculated cost-effectiveness ratio of $24,900 remains a point estimate and is subject to biases from uncertainties in model parameters. We attempted to account for these uncertainties with rigorous sensitivity analyses and the Monte Carlo simulation. The cost-effectiveness ratios are relatively stable under many circumstances, as shown in Table 7. Our baseline estimate of the cost of radiation therapy ($11,600) may be considered conservative. Previous economic studies have estimated breast radiotherapy charges between $12,00012 and $18,74272; actual costs are expected to be lower. Sensitivity analysis allows us to examine the effect of higher or lower costs of PMRT. As expected, there is significant variation in the cost-effectiveness of PMRT when its cost is increased and/or when the risk of breast cancer relapse is reduced. If a ratio of $50,000 per year of life gained is treated as a threshold level, PMRT could be considered cost-effective when the 10-year risk of relapse is 60% or greater, regardless of the cost of radiation therapy. When the risk of relapse is between 30% and 50%, the cost-effectiveness of PMRT is highly dependent on the cost of the treatment. The survival benefits and resultant cost-effectiveness ratio produced by this model apply to premenopausal node-positive patients in general, as rates of relapse were derived from combined data reported by large meta-analyses. The meta-analyses by the EBCCTG and Whelan et al39 do not allow direct cost-effectiveness evaluation of PMRT in higher-risk subgroups, such as patients with four or more positive nodes or advanced primary tumors. A detailed analysis of various patient subgroups is beyond the scope of this report. However, it is conceivable that PMRT may be more cost-effective in patients at a greater risk of local-regional recurrence, as suggested by sensitivity analysis. There are several limitations to this analysis. The clinical impact of PMRT was based on data derived from mostly non-US trials. Some of these trials have been criticized for deviation in surgical technique and incorporation of older chemotherapy regimens, compared with practice in the US. The few randomized trials performed in the US were smaller and inconclusive with regard to an improvement in survival.20-22 The time horizon in the model was truncated at 15 years, thus restricting our ability to assess lifetime benefits of PMRT. Quality-adjusted survival is limited by the lack of well-described patient utilities for conditions following LRF or chest wall irradiation. Second, costs of breast cancer recurrences were derived from retrospective examination of claims data from Medicare or health maintenance organizations. Resource utilization can vary widely depending local practice patterns, age of the patient, and extent of cancer recurrence. Not all studies separate net or attributable costs of breast cancer care from total health care costs, which may include costs not pertinent to the model. Finally, simplification of health states and their transitions was necessary in order to create a useable Markov model. Construction of special health states describing severe complications of treatment or breast cancer recurrence would permit analysis of these important end points but would also require additional assumptions on probability, costs, and utilities. In conclusion, the results of our model suggest that PMRT is a cost-effective intervention for premenopausal node-positive breast cancer patients. Radiation therapy is a moderate expense in the overall care of high-risk breast cancer patients, but this initial investment substantially decreases downstream costs of recurrent disease. The most important factors influencing the cost-effectiveness ratio are the cost of PMRT and the risk of breast cancer relapse. When considering quality of life, the cost per QALY was estimated to be $22,600. This ratio of cost per QALY was highly sensitive to the relative utility of being well after PMRT compared with being well after chemotherapy alone. Future research is necessary to assess the cost-effectiveness of PMRT for subgroups of node-positive breast cancer patients, such as those patients with four or more involved nodes and/or large primary tumors and those with one to three involved nodes where PMRT is currently under clinical investigation.
Presented in part at the Thirty-Fourth Annual Meeting of the American Society of Clinical Oncology, Los Angeles, CA, May 16-19, 1998.
1. Greenlee RT, Hill-Harmon BM, Murray T, et al: Cancer statistics, 2001. Ca Cancer J Clin 51: 15-36, 2001 2. Hoerger TJ, Downs KE, Lakshmanan MC, et al: Healthcare use among U.S. women aged 45 and older: Total costs and costs for selected postmenopausal health risks. J Womens Health Gend Based Med 8: 1077-1089, 1999[Medline] 3. Fowble B: Postmastectomy radiation: Then and now. Oncology 11: 213-240, 1997[Medline]
4.
Fowble B, Gray R, Gilchrist K, et al: Identification of a subgroup of patients with breast cancer and histologically positive axillary nodes receiving adjuvant chemotherapy who may benefit from postoperative radiotherapy. J Clin Oncol 6: 1107-1117, 1988
5.
Katz A, Strom EA, Buchholz TA, et al: Locoregional recurrence patterns after mastectomy and doxorubicin-based chemotherapy: Implications for postoperative irradiation. J Clin Oncol 18: 2817-2827, 2000
6.
Recht A, Edge SB, Solin LJ, et al: Postmastectomy radiotherapy: Clinical practice guidelines of the American Society of Clinical Oncology. J Clin Oncol 19: 1539-1569, 2001
7.
Overgaard M, Nahsen PS, Overgaard J, et al: Postoperative radiotherapy in high-risk premenopausal women with breast cancer who receive adjuvant chemotherapy. N Engl J Med 337: 949-955, 1997
8.
Ragaz J, Jackson SM, Le N, et al: Adjuvant radiotherapy and chemotherapy in node-positive premenopausal women with breast cancer. N Engl J Med 337: 956-962, 1997 9. Overgaard M, Jensen M-B, Overgaard J, et al: Postoperative radiotherapy in high-risk postmenopausal breast-cancer patients given tamoxifen: Danish Breast Cancer Cooperative Group DBCG 82c randomised trial. Lancet 353: 1641-1648, 1999[CrossRef][Medline] 10. Verma S, Will BP, Berthelot J-M, et al: Economic impact of post-mastectomy locoregional radiotherapy (LRRT) for node positive breast cancer patients. Proc Am Soc Clin Oncol 17: 121a, 1998 (abstr 464) 11. Dunscombe P, Samant R, Roberts G: A cost-outcome analysis of adjuvant postmastectomy locoregional radiotherapy in premenopausal node-positive breast cancer patients. Int J Radiat Oncol Biol Phys 48: 977-982, 2000[CrossRef][Medline] 12. Marks LB, Hardenbergh PH, Winer ET, et al: Assessing the cost-effectiveness of postmastectomy radiation therapy. Int J Radiat Oncol Biol Phys 44: 91-98, 1999[CrossRef][Medline] 13. Weeks JC: Economic considerations in comparing whole abdominal radiotherapy with combination doxorubicin-cisplatin chemotherapy in advanced endometrial carcinoma: How much economic data should be collected? J Natl Cancer Inst Monogr 19: 17-19, 1995
14.
Eisenberg JM: Clinical economics: A guide to the economic analysis of clinical practices. JAMA 262: 2879-2886, 1989 15. Cady B: Cost-effective preoperative evaluation, operative treatment, and postoperative follow-up in the breast cancer patients. Surg Clin North Am 76: 25-34, 1996[CrossRef][Medline] 16. Lee JH, Solin LJ, Glick HA: A decision-analytic model and cost-effectiveness evaluation of postmastectomy radiation therapy for high-risk breast cancer. Proc Am Soc Clin Oncol 17: 420a, 1998 (abstr 1621)
17.
Sonnenberg FA, Beck JR: Markov models in medical decision making: A practical guide. Med Dec Making 13: 322-338, 1993 18. Early Breast Cancer Trialists’ Collaborative Group: Polychemotherapy for early breast cancer: An overview of randomised trials. Lancet 352: 930-942, 1998[CrossRef][Medline]
19.
Saphner T, Tormey DC, Gray R: Annual hazard rates of recurrence for breast cancer after primary therapy. J Clin Oncol 14: 2738-2746, 1996 20. Olsen JE, Neuberg D, Pandya KJ, et al: The role of radiotherapy in the management of operable locally advanced breast carcinoma: Results of a randomized trial by the Eastern Cooperative Oncology Group. Cancer 79: 1138-1149, 1997[CrossRef][Medline]
21.
Griem KL, Henderson IC, Gelman R, et al: The 5-year results of a randomized trial of adjuvant radiation therapy after chemotherapy in breast cancer patients treated with mastectomy. J Clin Oncol 5: 1546-1555, 1987 22. Muss HB, Cooper MR, Brockschmidt JK, et al: A randomized trial of chemotherapy (L-PAM vs CMF) and irradiation for node positive breast cancer: Eleven year follow-up of a Piedmont Oncology Association trial. Breast Cancer Res Treat 19: 77-84, 1991[CrossRef][Medline] 23. Velez-Garcia E, Carpenter JT Jr, Moore M, et al: Postsurgical adjuvant chemotherapy with or without radiotherapy in women with breast cancer and positive axillary nodes: A South-Eastern Cancer Study Group (SEG) trial. Eur J Cancer 28A: 1833-1837, 1992[CrossRef][Medline]
24.
Crowe JP Jr, Gordon NH, Antunez AR, et al: Local-regional breast cancer recurrence following mastectomy. Arch Surg 126: 429-432, 1991
25.
Recht A, Gray R, Davidson NE, et al: Locoregional failure 10 years after mastectomy and adjuvant chemotherapy with or without tamoxifen without irradiation: Experience of the Eastern Cooperative Oncology Group. J Clin Oncol 17: 1689-1700, 1999 26. Lee YT: Breast carcinoma: Pattern of recurrence and metastasis after mastectomy. Am J Clin Oncol 7: 443-449, 1984[Medline] 27. Magno L, Bignardi M, Micheletti E, et al: Analysis of prognostic factors in patients with isolated chest wall recurrence of breast cancer. Cancer 60: 240-244, 1987[CrossRef][Medline] 28. Deutsch M, Parsons JA, Mittal BB: Radiation therapy for local-regional recurrent breast carcinoma. Int J Radiat Oncol Biol Phys 12: 2061-2065, 1986[Medline] 29. Andry G, Suciu S, Vico P, et al: Locoregional recurrences after 649 modified radical mastectomies: Incidence and significance. Eur J Surg Cancer 15: 476-485, 1989 30. Aberizk WJ, Silver B, Henderson IC, et al: The use of radiotherapy in patients with isolated locoregional recurrence of breast carcinoma after mastectomy. Cancer 58: 1214-1218, 1986[CrossRef][Medline] 31. Halverson KJ, Perez CA, Kuske RR, et al: Survival following locoregional recurrence of breast cancer: Univariate and multivariate analysis. Int J Radiat Oncol Biol Phys 23: 285-291, 1992[Medline] 32. Schwaibold F, Fowble BL, Solin LJ, et al: The results of radiation therapy for isolated local regional recurrence after mastectomy. Int J Radiat Oncol Biol Phys 21: 299-310, 1991[Medline] 33. Janjan NA, McNeese MD, Buzdar AU, et al: Management of locoregional recurrent breast cancer. Cancer 58: 1552-1556, 1986[CrossRef][Medline]
34.
Schmoor C, Sauerbrei W, Bastert G, et al: Role of isolated locoregional recurrence of breast cancer: Results of four prospective studies. J Clin Oncol 18: 1696-1708, 2000
35.
Fortin A, Larochelle M, Laverdiere J, et al: Local failure is responsible for the decrease in survival for patients with breast cancer treated with conservative surgery and postoperative radiotherapy. J Clin Oncol 17: 101-109, 1999 36. Di Pietro S, Bertario L, Cantu G, et al: An analysis of 800 breast cancer patients relapsed after radical mastectomy. Tumori 62: 99-112, 1976[Medline] 37. National Center for Health Statistics, CDC: United States abridged life tables, 1996. Natl Vital Stat Rep 47: 5-16, 1998 38. Early Breast Cancer Trialists’ Collaborative Group: Favourable and unfavourable effects on long-term survival of radiotherapy for early breast cancer: An overview of the randomised trials. Lancet 355: 1757-1770, 2000[CrossRef][Medline]
39.
Whelan TJ, Julian J, Wright J, et al: Does locoregional radiation therapy improve survival in breast cancer? A meta-analysis. J Clin Oncol 18: 1220-1229, 2000 40. Arriagada R, Rutqvist LE, Mattson A, et al: Adequate locoregional treatment for early breast cancer may prevent secondary dissemination. J Clin Oncol 13: 2869-2878, 1995[Abstract] 41. van de Steene J, Soete G, Storme G: Adjuvant radiotherapy for breast cancer significantly improves overall survival: The missing link. Radiother Oncol 55: 263-272, 2000[CrossRef][Medline] 42. Metz JM, Schultz DJ, Fox K, et al: Long-term outcome after postmastectomy radiation therapy for breast cancer patients at high risk for local-regional recurrence. Cancer J Sci Am 5: 77-83, 1999[Medline] 43. Cuzick J, Stewart H, Peto R, et al: Overview of randomized trials of postoperative adjuvant radiotherapy in breast cancer. Cancer Treat Rep 71: 15-29, 1987[Medline] 44. Cuzick J, Stewart H, Rutqvist L, et al: Cause-specific mortality in long-term survivors of breast cancer who participated in trials of radiotherapy. J Clin Oncol 12: 447-454, 1994[Abstract] 45. Kuske RR: Adjuvant chest wall and nodal irradiation: Maximize cure, minimize late cardiac toxicity. J Clin Oncol 16: 2579-2582, 1998 (letter)[Medline] 46. Hillner BE, Smith TJ: Efficacy and cost effectiveness of adjuvant chemotherapy in women with node-negative breast cancer: A decision-analysis model. N Engl J Med 324: 160-168, 1991[Abstract]
47.
Barlow WE, Taplin SH, Yoshida CK, et al: Cost comparison of mastectomy versus breast-conserving therapy for early-stage breast cancer. J Natl Cancer Inst 93: 447-455, 2001
48.
Kattlove H, Liberati A, Keeler E, et al: Benefits and costs of screening and treatment for early breast cancer: Development of a basic benefit package. JAMA 273: 142-148, 1995 49. Fireman BH, Quesenberry CP, Somkin CP, et al: Cost of care for cancer in a health maintenance organization. Health Care Fin Rev 18: 51-76, 1997[Medline]
50.
Legorreta AP, Brooks RJ, Leibowitz AN, et al: Cost of breast cancer treatment: A 4-year longitudinal study. Arch Intern Med 156: 2197-2201, 1996 51. McDonough KL: Breast cancer stage cost analysis in a managed care population. Am J Manag Care 5: S377-S382, 1999 (suppl)[Medline] 52. Polednak AP, Shevchenko IP, Flannery JT: Estimating breast cancer treatment charges in Connecticut. Conn Med 60: 263-267, 1996[Medline] 53. Riley GF, Potosky AL, Lubitz LD, et al: Medicare payments from diagnosis to death for elderly cancer patients by stage at diagnosis. Med Care 33: 828-841, 1995[CrossRef][Medline]
54.
Taplin SH, Barlow W, Urban N, et al: Stage, age, comorbidity, and direct costs of colon, prostate, and breast cancer care. J Natl Cancer Inst 87: 417-426, 1995
55.
Hayman JA, Fairlough DL, Harris JR, et al: Patient preferences concerning the trade-off between the risks and benefits of routine radiation therapy after conservative surgery for early-stage breast cancer. J Clin Oncol 15: 1252-1260, 1997 56. de Koning HJ, van Ineveld BM, van Oortmarssen GJ, et al: Breast cancer screening and cost-effectiveness, policy alternatives, quality of life considerations and the possible impact of uncertain factors. Int J Cancer 49: 531-537, 1991[Medline]
57.
Earle CC, Chapman RH, Baker CS, et al: Systematic overview of cost-utility assessments in oncology. J Clin Oncol 18: 3302-3317, 2000
58.
Hillner BE, Smith TJ, Desch CE: Efficacy and cost-effectiveness of autologous bone marrow transplantation in metastatic breast cancer: Estimates using decision analysis while awaiting clinical trial results. JAMA 267: 2055-2061, 1992 59. Hristova L, Hakama M: Effect of screening for cancer in the Nordic countries on deaths, cost and quality of life up to the year 2017. Acta Oncol 36: 1-60, 1997 (suppl 9)[Medline] 60. Hutton J, Brown R, Borowitz M, et al: A new decision model for cost-utility comparisons of chemotherapy in recurrent metastatic breast cancer. Pharmacoeconomics 9: 8-22, 1996 (suppl 2)[Medline] 61. Launois R, Reboul-Marty J, Henry B, et al: A cost-utility analysis of second-line chemotherapy in metastatic breast cancer: Docetaxel versus paclitaxel versus vinorelbine. Pharmacoeconomics 10: 504-521, 1996[Medline] 62. Grann VR, Jacobson JS, Sundararajan V, et al: The quality of life associated with prophylactic treatments for women with BRCA1/2 mutations. Cancer J Sci Am 5: 283-292, 1999[Medline] 63. Grann VR, Panageas KS, Whang W, et al: Decision analysis of prophylactic mastectomy and oophorectomy in BRCA1-positive or BRCA2-positive patients. J Clin Oncol 16: 979-985, 1998[Abstract] 64. Gerard K, Dobson M, Hall J: Framing and labelling effects in health descriptions: Quality adjusted life years for treatment of breast cancer. J Clin Epidemiol 46: 77-84, 1993[CrossRef][Medline]
65.
Doubilet P, Begg CB, Weinstein MC, et al: Probabilistic sensitivity analysis using Monte Carlo simulation: A practical approach. Med Decis Making 5: 157-177, 1985 66. Smith TJ, Hillner BE: The efficacy and cost-effectiveness of adjuvant therapy of early breast cancer in premenopausal women. J Clin Oncol 11: 771-776, 1993[Abstract] 67. Hayman JA, Hillner BE, Harris JR, et al: The cost-effectiveness of routine radiation therapy following conservative surgery for early-stage breast cancer. J Clin Oncol 16: 1022-1029, 1998[Abstract] 68. Desch CE, Hillner BE, Smith TJ, et al: Should the elderly receive chemotherapy for node-negative breast cancer? A cost-effectiveness analysis examining total and active life-expectancy outcomes. J Clin Oncol 11: 777-782, 1993[Abstract] 69. Orr RK, Col NF, Kuntz KM: A cost-effectiveness analysis of axillary node dissection in postmenopausal women with estrogen receptor-positive breast cancer and clinically negative axillary nodes. Surgery 126: 568-576, 1999[Medline]
70.
Hayman JA, Hillner BE, Harris JR, et al: Cost-effectiveness of adding an electron-beam boost to tangential radiation therapy in patients with negative margins after conservative surgery for early-stage breast cancer. J Clin Oncol 18: 287-295, 2000 71. Bedwinek J: Natural history and management of isolated local-regional recurrence following mastectomy. Semin Radiat Oncol 4: 260-269, 1994[CrossRef][Medline] 72. Palit TK, Mittenburg DM, Brunicardi C: Cost analysis of breast conservation surgery compared with modified radical mastectomy with and without reconstruction. Am J Surg 179: 441-445, 2000[CrossRef][Medline] Submitted July 3, 2001; accepted March 6, 2002.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2002 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
, without radiation) and intervention scenarios (|b, with radiation). Ten- and 15-year results are shown next to curves.
