|Anastrozole is cost-effective vs tamoxifen as initial adjuvant therapy in early breast cancer: Canadian perspectives on the ATAC completed-treatment analysis
|Rocchi A, Verma S
This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn.
The study examined the use of anastrozole (Arimidex) in the adjuvant treatment of hormone receptor-positive (HR+), postmenopausal early breast cancer (BC) patients. Anastrozole was compared with the standard care for this group of women, that is, tamoxifen. Both therapies were assumed to be given for up to 5 years. The dosages were not reported.
Economic study type
Cost-effectiveness analysis and cost-utility analysis.
The study population comprised a hypothetical cohort of HR+, postmenopausal, early BC women. The typical patient was 64 years old and the distribution of early BC was 64% Stage I and 36% Stage II.
The setting was secondary care. The economic study was carried out in Canada.
Dates to which data relate
The clinical data were derived from studies published between 1986 and 2005. Most of the resource use and cost data were derived from 2004 sources. The price year was 2004.
Source of effectiveness data
The clinical data used in the decision model were:
the recurrence rates (locoregional or distant recurrence, or contralateral BC),
the mortality rates (due to BC or other causes),
the adverse event rates (minor and major adverse events),
the therapy discontinuation rates, and
treatment efficacy in reducing disease progression (in the form of relative risk for anastrozole compared with tamoxifen).
A Markov model with yearly cycles and a lifetime horizon was constructed to evaluate the clinical and economic impact of the two therapies in a hypothetical cohort of 10,000 eligible women. The health states of the model were reported, along with a schematic representation of the model. Briefly, patients could experience adverse events, BC recurrence, BC mortality, or death from other causes. BC progression was represented by standard health states (loco-regional recurrences, distal recurrences).
Sources searched to identify primary studies
Much of the clinical data relative to anastrozole over the first 5 years were derived from the "Anastrozole, tamoxifen alone, or in combination" (ATAC) trial, an RCT conducted on 9,366 postmenopausal women with early BC, with a median follow-up of 33 months. Data for the post-trial period were mainly derived from recent meta-analyses conducted by the Early Breast Cancer Trialists' Collaborative Group (EBCTCG). Most of these data were also confirmed by the literature. Rates of BC deaths were based on data from a provincial cancer database (British Columbia). All-cause mortality was derived from life table survival rates published by Statistics Canada.
Methods used to judge relevance and validity, and for extracting data
The primary studies appear to have been identified selectively. The authors did not report details of a systematic search of data, but they explicitly selected the ATAC trial for its large sample size, long follow-up and head-to-head comparison between anastrozole and tamoxifen. Also, long-term data were explicitly taken from meta-analyses, on the grounds of high internal validity. The treatment effect for anastrozole with respect to tamoxifen was calculated by means of hazard ratios (relative risk).
Measure of benefits used in the economic analysis
The summary benefit measures were the life-years (LYs) and quality-adjusted life-years (QALYs). LYs were used in the cost-effectiveness analysis and QALYs in the cost-utility analysis. Both measures were estimated using the modelling approach and an annual discount rate of 5% was applied. The utility weights used to calculate QALYs were derived from a BC utility study, the details of which were not given. However, the utility weights associated with each health state were reported.
The viewpoint of the analysis was that of a third-party payer. The analysis included the costs of medications, chemotherapy regimens, physician services, laboratory tests, procedures, hospitalisations, chemotherapy clinic visits, radiotherapy fractions, treatment of adverse events and treatment of endometrial cancer. The unit costs and the quantities of resources used were not presented separately. Patterns of resource consumption were derived from two main sources. First, a panel of seven oncologists selected from Canadian investigators of the ATAC trial, to represent the various regions of Canada in a proportional fashion, estimated the management of patients while on adjuvant hormonal therapy. Second, the Statistics Canada's Population Health Model (POHEM) for BC was used to provide resource use for the other items. The costs were estimated using data from Ontario. Specifically, the Ontario Drug Benefit Formulary and Pharmaceutical Price Index, the Cancer Care Ontario Formulary, the Ontario Health Insurance Plan Schedules of Benefits, the Ontario Case Costing Initiative, and some published studies. A 5% discount rate was used as lifetime costs were evaluated. The price year was 2004 in the base-case analysis and 2002 in the interim analysis in which 3-year clinical trial data were used.
Statistical analysis of costs
The costs were treated deterministically in the base-case analysis.
Productivity costs were not considered.
A univariate sensitivity analysis was carried out to evaluate the robustness of the base-case results to variations in key parameters of the model. The parameters investigated included difference in recurrence rate between therapies, rate of distant recurrence subsequent to local recurrence, adverse event rates, BC treatment cost, discount rate and time horizon. Changes in parameters were either based on published estimates (often confidence intervals) or were set by the authors. A probabilistic sensitivity analysis was also performed by assigning stochastic distributions to all inputs of the analysis. This generated cost-effectiveness acceptability curves showing the probability of anastrozole being cost-effective given a threshold for a QALY gained (usually CAD 50,000).
Estimated benefits used in the economic analysis
Under base-case assumptions (5-year complete analysis), the expected LYs were 11.3454 with anastrozole and 11.1531 with tamoxifen (difference 0.1923).
In the 3-year interim analysis, the expected LYs were 11.0973 with anastrozole and 10.9034 with tamoxifen (difference 0.1938).
Under base-case assumptions (5-year complete analysis), the expected QALYs were 10.9301 with anastrozole and 10.7222 with tamoxifen (difference 0.2079).
In the 3-year interim analysis, the expected QALYs were 10.6794 with anastrozole and 10.4614 with tamoxifen (difference 0.2180).
Under base-case assumptions (5-year complete analysis), the expected costs were CAD 16,041 with anastrozole and CAD 10,244 with tamoxifen (difference CAD 5,796).
The higher acquisition costs for anastrozole (CAD 6,974) were only partially offset by a reduction in the costs associated with morbidity (-CAD 1,143).
In the 3-year interim analysis, the expected costs were CAD 16,147 with anastrozole and CAD 10,518 with tamoxifen (difference CAD 5,629).
Synthesis of costs and benefits
Incremental cost-effectiveness ratios and cost-utility ratios were calculated in order to combine the costs and benefits of the alternative strategies.
Under base-case assumptions, the incremental cost per LY gained with analysis over tamoxifen was CAD 30,137 (CAD 29,043 in the interim analysis), while the incremental cost per QALY gained was CAD 27,877 (CAD 25,818 in the interim analysis).
The results of the deterministic sensitivity analysis showed that the cost-utility ratio was particularly sensitive to the extent and duration of anastrozole benefit. The incremental cost per QALY gained remained below the threshold of CAD 50,000 per QALY if the benefits of anastrozole were to endure for 6 years or longer, but was higher than that threshold if the benefit lasted only 5 years (CAD 51,706). The only other case where the cost per QALY was over CAD 50,000 was when the time horizon of the analysis was reduced to 10 years (CAD 83,217). The model was slightly sensitive to the discount rate as the additional costs for anastrozole were incurred early while the benefits were accrued over a longer horizon.
The probabilistic sensitivity analysis showed that anastrozole remained cost-effective in almost 95% of simulations under base-case assumptions (in 80% of simulations in the interim analysis). In the base-case analysis, the 95% confidence interval for the Monte Carlo simulations was CAD 17,428 to CAD 54,605.
Anastrozole was an effective and cost-effective alternative to tamoxifen for the adjuvant therapy of hormone receptor-positive (HR+), early stage breast cancer (BC) in postmenopausal women.
CRD COMMENTARY - Selection of comparators
The authors justified the choice of the comparators. Tamoxifen was considered to be the standard treatment for the patient population considered in the study, while anastrozole represents one of the available third-generation aromatase inhibitors (anastrozole, exemestane and letrozole). The choice of anastrozole was based on several factors, principally the availability of robust data on the effectiveness and safety profile and the fact that anastrozole had been approved as an alternative for the treatment of HR+, early BC in postmenopausal women. You should decide whether they are valid comparators in your own setting.
Validity of estimate of measure of effectiveness
The clinical data used to populate the decision model were mainly derived from the ATAC trial, which was selectively identified by the authors in order to use a valid source of data. Both the interim and final results of the ATAC trial were reported and used in the analysis. Details such as length of follow-up, sample size and key clinical results were reported. An important feature of this trial was the head-to-head comparison between the two drugs under study. Long-term data were obtained from meta-analyses and this should have ensured high internal validity. The use of hazard ratios to estimate the relative impact of anastrozole compared with tamoxifen appears appropriate. Other data were derived from published studies, which were described in detail. Provincial statistics were used to assess mortality data. The authors investigated the impact of variations of key clinical parameters in the sensitivity analysis. The authors stated that both the ATAC trial and the EBCTCG overviews represent two valid sources of BC data.
Validity of estimate of measure of benefit
The benefit measures were modelled using the Markov model. Both benefits were appropriate as they capture the impact of the therapies on survival and quality of life, which are relevant for cancer patients. Discounting was performed and the use of alternative discount rates was investigated in the sensitivity analysis. The use of quality of life data obtained from BC patients was adequate, but there was little information on the methodology used to elicit utility weights.
Validity of estimate of costs
The analysis of the costs was consistent with the authors' stated perspective. The costs were presented mainly as macro-categories, and a detailed breakdown of the cost items was not given. This might limit the possibility of replicating the analysis in other settings. However, the use of macro-categories is common in economic evaluations of cancer based on Markov models. In addition, extensive information on the sources of the costs and resource use was given. Statistical analyses of the costs and quantities were performed only in the sensitivity analysis, in which the impact of variations in cost estimates was also considered. The price year was reported, which will enhance the generalisability of the study results to other time periods.
The authors reported the results from a similar cost-effectiveness analysis, the results of which were comparable to those from the current economic evaluation. The issue of the generalisability of the study results to other settings was not explicitly stated, but the use of both deterministic and probabilistic sensitivity analysis enhances the external validity of the analysis. The authors presented their results in full and provided a cost-effectiveness acceptability curve.
Implications of the study
The study results support the use of anastrozole in the treatment of HR+, early stage BC in postmenopausal women. The authors pointed out that the analysis should be replicated as soon as further results from the ATAC trial (longer follow-up and sub-group analysis) are available.
Source of funding
Supported by a grant from AstraZeneca Canada.
Rocchi A, Verma S. Anastrozole is cost-effective vs tamoxifen as initial adjuvant therapy in early breast cancer: Canadian perspectives on the ATAC completed-treatment analysis. Supportive Care in Cancer 2006; 14(9): 917-927
Other publications of related interest
Because readers are likely to encounter and assess individual publications, NHS EED abstracts reflect the original publication as it is written, as a stand-alone paper. Where NHS EED abstractors are able to identify positively that a publication is significantly linked to or informed by other publications, these will be referenced in the text of the abstract and their bibliographic details recorded here for information.
Hillner B. Benefit and projected cost-effectiveness of anastrozole versus tamoxifen initial adjuvant therapy for patients with early-stage estrogen receptor-positive breast cancer. Cancer 2004;101:1311-2.
ATAC Trialists' Group. Results of the ATAC (Arimidex, tamoxifen alone or in combination) trial after completion of 5 years' adjuvant treatment for breast cancer. Lancet 2005;365:60-2.
Early Breast Cancer Trialists' Collaboration Group. Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet 2005;365:1687-717.
Subject indexing assigned by NLM
Analysis of Variance; Antineoplastic Agents, Hormonal /administration & Antineoplastic Combined Chemotherapy Protocols /economics /therapeutic use; Aromatase Inhibitors /administration & Breast Neoplasms /drug therapy /economics /mortality; Canada; Chemotherapy, Adjuvant; Cost-Benefit Analysis; Disease-Free Survival; Female; Humans; Neoplasm Recurrence, Local /prevention & Nitriles /administration & Quality-Adjusted Life Years; Randomized Controlled Trials as Topic; Risk Reduction Behavior; Sensitivity and Specificity; Survival Rate; Tamoxifen /administration & Time Factors; Treatment Outcome; Triazoles /administration & control; dosage /economics; dosage /economics; dosage /economics; dosage /economics; dosage /economics
Date bibliographic record published
Date abstract record published