Abstract P6-07-09: What should be done after PHARE and HERA? Using value of information analysis as a guide

Abstract

Abstract Background: Adjuvant trastuzumab therapy totaling 12 months (AT-12) is the current standard of care in the treatment of early-stage, HER-2/neu amplified breast cancer. Two studies, PHARE and HERA, examined whether the duration of adjuvant trastuzumab should be decreased safely to 6 months (AT-6) or increased to 2 years (AT-24), respectively. Could the implementation of value of information (VOI) analysis inform future clinical trial priority setting in this sphere? Methods: A Markov health-state transition model was employed to compare 3 adjuvant therapy options: AT-12, AT-6, and AT-24. Patients transitioned through 3 health states, disease free, metastatic recurrence, and death, over 3-month time intervals. The simulation was run using a lifetime time horizon. Transition probabilities of the health states in the AT-12 cohort were estimated from the National Surgical Adjuvant Breast and Bowel Project B-31 and North Central Cancer Treatment Group N9831 trials. Based on the efficacy assumptions that provided the foundation for HERA, the AT-24 probabilities were imputed. AT-6 transition probabilities were assumed to equal AT-12 probabilities. Health outcomes were measured in quality adjusted life years (QALY) and costs in 2012 United States dollars. Health outcomes and costs were discounted to the present value by 5% annually, compounded quarterly. Result: In the reference case, AT-24 had an incremental cost-effectiveness ratio of $57,784/QALY. Probabilistic sensitivity analysis revealed that AT-24 was likely to be 0.3%, 11%, 58%, and 93% cost effective at a threshold willingness-to-pay (WTP) of $20,000, $40,000, $60,000, and $80,000/QALY, respectively. At a WTP threshold of 65,000, the expected VOI payoff from conducting the HERA trial was computed as $204,692 per patient treated per year. Conversely, AT-6 would be expected to cost $100,592 less than AT-12 over a patient's life, corresponding to an expected $249,564 per patient per year payoff for the PHARE trial. Since neither PHARE nor HERA were able to achieve their primary objectives, a new trial testing the duration of trastuzumab could be considered. Employing the same model, AT for 9 months would be expected to save $50,596 per patient compared to AT-12. The VOI of such a trial is estimated to be $224,504 per patient treated per year. Conclusion: Despite the results of PHARE and HERA, VOI analysis suggests that a non-inferiority trial examining 9 months of AT compared to AT-12 should be considered. The actual value of such a trial may be even greater if new HER-2 targeted therapies may be added to AT in the future. However, determining the total net benefit to society of performing such research requires information on its associated budgetary outlays. Cooperative groups should make the their individual trial expenses public to facilitate improved collective decision-making on future trials. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P6-07-09.