Clinical and Health Economic Outcomes of Alternative HER2 Test Strategies for Guiding Adjuvant Trastuzumab Therapy

James A Lee; Megan Shaheen; Thomas Walke; Matt Daly


Expert Rev Pharmacoeconomics Outcomes Res. 2011;11(3):325-341. 

In This Article


The following sections describe the methods and assumptions used to obtain the literature, calculate test concordance, perform the baseline health economic and outcomes analysis, and evaluate the economic analysis results under alternative assumptions. The analysis was performed from a US national Medicare payer perspective using 2009 reimbursements.


Two systematic reviews (Dendukuri et al.[10] and Cuadros and Villegas[22]) were used as the basis for the literature search regarding comparison of IHC and FISH. The article by Dendukuri et al. includes 17 articles comparing IHC and FISH published prior to June 2006, and Cuadros and Villegas 17 articles through September 2006. The methods of their searches are discussed in detail in their respective publications.[10,22] Search terms were aggregated and applied to Medline, Cochrane Reviews and EMBASE databases, to identify additional articles through June 2010.

For analysis of CISH and SISH, literature from the SPOT-Light® HER2 CISH Kit FDA approval documentation was acquired and literature searches were performed, again using Medline, Cochrane Reviews and EMBASE, to identify articles that compare CISH or SISH with FISH. Articles that only compared IHC to CISH and SISH were not included in this analysis as the assumed reference assay was FISH. The articles were then summarized and data abstracted, without modification, to allow analysis of concordance and performance of the economic analyses.

Search terms used included: FISH, fluorescence in situ hybridization, IHC, immunohistochemistry, CISH, chromogenic in situ hybridization, SISH, silver-enhanced in situ hybridization, trastuzumab, breast, cancer and test. Expert reviewers analyzed all found abstracts for relevance. Inclusion criteria included: the article was available in the English language; results comparing IHC, CISH or SISH to FISH were provided; and testing was performed to establish HER2 status in breast cancer.

Concordance Analysis

The ASCO/CAP guidelines recommend that an alternative test show at least 95% concordance for both positive and negative values with the validated assay to which it is compared.[12] Our analysis was based on FISH serving as the reference assay. Positive concordance is the proportion of samples that test positive using the alternative assay (IHC, CISH or SISH) within the subset of all samples that test positive with FISH. Negative concordance is the proportion of samples that test negative using the alternative assay with the subset of all samples that test negative with FISH. Example calculations are shown in the Appendix at the end of this article.

Since common practice and the ASCO/CAP-recommended guideline is to confirm IHC 2+ results with a validated assay for HER2 gene amplification, separate concordance calculations were also performed based on algorithms with confirmation of these results. This increases the negative and positive concordance with the reference assay.

Economic Analysis Methods & Assumptions

We prepared an algebraic, cost–utility model in Microsoft Excel® to calculate the incremental costs and clinical benefits of enhanced HER2 test accuracy using alternative test regimens. The model results demonstrate, in terms of both potential clinical and economic outcomes, the value of greater concordance with a reference assay. The economic analysis only applies to evaluating the cost–effectiveness of 12-month AT therapy in women with HER2-positive early breast cancer.

Table 1 presents the model baseline estimates and ranges evaluated. The additional benefit and therapy cost associated with AT considers the net impact of AT therapy resulting in reduced morbidity, mortality and healthcare costs associated with a reduced incidence of metastatic breast cancer in the treated population.[27] The cost analysis is based on 2009 Medicare reimbursement rates for tests and inflating breast cancer care payments to approximate 2009 Medicare reimbursements for care as reported by Liberato et al..[27] The benefits of AT therapy of 1.18 quality-adjusted life-years (QALYs), relative to chemotherapy alone, were obtained from Liberato et al., without modification.[27] QALY is a measure of disease burden, including both the quality and the quantity of life lived, and is typically measured as an incremental addition provided by an intervention. The range of values is 0 for death to 1 for perfect health, with values in between representing less than full health. Both costs and QALYs were discounted at a 3% annual rate using a 15-year time horizon. Incremental costs are presented in terms of cost per QALY.

For the baseline economic analysis, the prices for tests and therapy presented in Table 1 were used, but the sensitivity, specificity and HER2 positivity of the population were obtained from each study as presented. The proportion of the population with HER2 overexpression is based on the results of the FISH assay in each study.

Overall, five test strategies were considered: primary FISH; primary SISH; primary CISH; primary IHC with confirmation of IHC 2+ with FISH; and primary IHC with confirmation of IHC 2+ and IHC 3+. The latter was considered as in some countries, such as Belgium and Finland, all IHC 2+ and 3+ cases are retested by FISH or CISH and only a combination of 2+ or 3+ IHC and gene amplification by ISH is regarded as HER2 positive.[12] Specifically, for CISH and FISH, a comparison was performed assuming both a price for a single- and a dual-probe CISH relative to FISH, as reimbursement is per probe in the USA. SISH is a dual-probe assay like FISH. Example calculations of the cost–utility analysis using the baseline economic analysis assumptions are presented in the Appendix.

We chose to analyze each study separately, rather than using a pooled or meta-analysis, to enumerate studies with results that met or exceeded ASCO/CAP concordance guidelines and transparently demonstrate the impact of concordance in determining whether one test strategy is dominant or cost effective relative to other strategies. A pooled analysis or meta-analysis was not selected, as using an average sensitivity and specificity can mask the likelihood of observing the average test performance and the observed high degree of variance in test performance. Furthermore, there is substantial variation between the studies regarding method of IHC used, application of single-probe versus dual-probe CISH and cut-offs for amplification, test methods, patient population selection and whether testing was performed at a local or centralized laboratory. Instead, a sensitivity analysis using Monte Carlo simulation, described later, was performed to provide insight regarding the level of sensitivity and specificity necessary for an alternative test to be cost effective relative to a reference assay. Furthermore, we applied our methods to Dendukuri et al.[10] and Cuadros and Villegas[22] pooled analyses separately to understand the likely outcome of a pooled analysis.

Economic Analysis Results Under Alternative Assumptions

Several studies of the cost–effectiveness of AT therapy have been performed with varying estimates regarding the expected costs and benefits of intervention. While nearly all the studies conclude that AT is cost effective, the assumptions regarding the costs and benefits of AT are key.[10,23,24] Furthermore, the assumptions regarding HER2 test performance may strongly influence the conclusions, in particular in the case of CISH, SISH and FISH, where high concordance is observed.

To evaluate the impact of alternative assumptions regarding these key parameters, a generalized Monte Carlo simulation was developed to allow the values of AT therapy costs and benefits to vary between the ranges presented in Table 1. A symmetric triangular distribution around a range of published results with a mean value of that observed in Liberato et al., and used in the baseline economic analysis, was selected. For example, for the benefit of AT therapy in terms of QALYs, the midpoint is 1.18, with a low of 0.47 and a high value of 1.91.[27] The symmetric triangular distribution provides for equal and declining likelihood of observations at the upper and lower tails of the distribution, with the maximum likely observation at the mean. Only the costs and benefits of AT therapy were randomized in the model.

Test price and performance, as measured by sensitivity and specificity, were fixed in the model but assessed at different mean values. Results for the proportion of the target population with HER2 overexpression based on the reference assay of 15, 20 and 25% were also examined. Results are based upon a simulation of 10,000 model runs. The simulation generated a range of incremental costs and QALYs as output and the results are presented as the proportion of observations where the reference assay is a cost-effective alternative, using a cost per QALY threshold of US$50,000 or less.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.