To what extent are clinical trial adverse events incorporated in the economic models of pharmaceutical products?
Biogen’s Aducanumab has just received FDA approval for the treatment of Alzheimer’s Disease. During clinical trials of the drug the most common adverse events were headaches and amyloid-related imaging abnormalities of the ARIA-E type (ARIA-Edema, defined as vasogenic edema or sulcal effusion. Dubois, 2021). One-quarter of participants in the 6 mg/kg dose group and one-third of those receiving 10 mg/kg developed ARIA-E, compared with 10 percent in the placebo group. Moreover, Seventeen percent of people on drug developed ARIA-H (brain microhemorrhages or localized superficial siderosis), compared to only 6% in the control group (Sevigny, 2016). The amyloid-related imaging abnormalities may result from increased cerebrovascular permeability as a consequence of antibody binding to deposited amyloid-beta. The expected annual price tag of $56,000 for Aducanumab raises the question of whether the adverse events associated with the treatment are incorporated into its economic model (Cubanski, 2021).
Craig (2009) investigated the current pharmaceutical practices of incorporating adverse events into economic models and found that there is a general implicit assumption that adverse events are significant and should be incorporated in modeling. However, adverse events are often not incorporated. One major obstacle to the incorporation of adverse events in modeling is the lack of adverse events data.
Better methodology and reporting of how adverse events data are incorporated, or excluded, in economic models is necessary for a model that effectively captures the significance of adverse events. Nevertheless, the process of pharmaceutical development shows large variability in the approach to integration of adverse event costs in economic models even when relevant data is available. Two key data points required as inputs are the probability, defined as the frequency of the adverse events over a set time period, and the unit cost, the cost per episode care associated with the event. By multiplying the two inputs we can reach the expected adverse event cost per patient.
The process gets tricky when identifying the cost associated with an adverse event. These can be sourced from existing literature, micro-cost approaches, clinical consensus from key opinion leaders (KOL), or real-world data such as administrative reimbursement data and electronic medical records. While the paper by Carlson (2019) identifies several methods for integrating adverse events into economic models, the author shares the same concern as Craig that there is no defined or agreed upon approach that is most appropriate. This is largely due to the possibility of different adverse events showing up in clinical trials, of varying severity.
Clinical Consensus-Based Approach utilizes expert opinion and existing clinical guidelines to estimate the cost of adverse events. This approach cannot determine which adverse events to include, but suggests as a starting point, focusing on grade 3+ or severe adverse events with a frequency above 5% for any given intervention, as these are most likely to require healthcare resources and have a meaningful impact on the quality of treatment (Wong, 2019). Additionally, the cost of healthcare resources could be derived from real-world claims data; however, it may not capture the entire economic burden associated with the adverse event and could be considered negligible when compared to the severity and cost of the disease (Roze, 2016). Each method considered in identifying the costs of adverse events has its strengths and limitations, and therefore, it may be optimal to apply a combination of methods in generating economic models. It is equally important to ensure the accuracy of the estimated adverse event cost is balanced with the effort it takes to estimate it.
References
Craig D., McDaid C., Fonseca T., Stock C., Duffy S. & Woolacott, N (2009). Are adverse effects incorporated in economic models? An initial review of current practice. Health Technol Assess 13(62), pp1-71, 97-181. https://www.journalslibrary.nihr.ac.uk/hta/hta13620/
Cubanski, J. (2021, June 10). FDA’s Approval of Biogen’s New Alzheimer’s Drug Has Huge Cost Implications for Medicare and Beneficiaries. Kaiser Family Foundation. https://www.kff.org/medicare/issue-brief/fdas-approval-of-biogens-new-alzheimers-drug-has-huge-cost- implications-for-medicare-and-beneficiaries/.
Dubois, B. (2021). Aduhelm. ALZFORUM. https://www.alzforum.org/therapeutics/aduhelm.
Roze S., Cartier-Bechu, C., Gherardi, A., Monnier, R., Pignata, M., Petijean, A. & Sivignon, M. (2016). MO2 – Considering or Not Adverse Events in Health Economic Models: A Pragmatic Approach. Value in Health 19(7), 347. https://doi.org/10.1016/j.jval.2016.09.004.
Sevigny J, Chiao P, Bussière T, Weinreb PH, Williams L, Maier M, Dunstan R, Salloway S, Chen T, Ling Y, O’Gorman J, Qian F, Arastu M, Li M, Chollate S, Brennan MS, Quintero-Monzon O, Scannevin RH, Arnold HM, Engber T, Rhodes K, Ferrero J, Hang Y, Mikulskis A, Grimm J, Hock C, Nitsch RM & Sandrock, A. (2016). The antibody aducanumab reduces Aβ plaques in Alzheimer’s disease. Nature 537:7618, 50-56. PubMed.
Wong, W., Carlson, J. & Cloutier, M. (2019). Estimating the Costs of Adverse Events in Economic Models: Is There a “Right” Approach? Value & Outcomes Spotlight. https://www.ispor.org/docs/default-source/publications/value-outcomes-spotlight/may-june-2019/heor -article—william-wong.pdf?sfvrsn=f0c0af42_0
20+ years of healthcare and data sciences experience including BI, ML, Policy Analysis, HTA, Value Dossiers, Predictive modeling, meta-analysis, and Budget Impact; experience with many claims and EHR databases; strong background in health policy and economics.