The Health Economics of Genomic Technologies: A Growing Evidence Base on Value

Introduction

The sequencing of the first human genome in 2003 catalysed development of the field of precision medicine over the past two decades. Next-generation sequencing approaches developed in the clinical research setting are now emerging into clinical practice. These approaches allow either the whole genome or key sections of the genome (exome sequencing or targeted panel testing) to be sequenced [1]. This genomic information can be generated at speed, and increasingly at a reasonable cost, although bioinformatics necessary to interpret sequence data and subsequent treatment costs can remain prohibitively expensive. While potentially costly, genomic information can guide diagnosis and clinical management for patients with cancer, rare diseases and chronic diseases, potentially improving health and well-being outcomes for patients and families. Large-scale genome sequencing projects, such as the 100,000 Genomes Project in England, Australian Genomics and the Canadian Precision Health Initiative continue to provide insights that not only impact current clinical management for patients, but also inform the development of new genomic and non-genomic interventions and therapies [2–4].

These scientific developments are promising, but as with all innovations in healthcare, if genome and exome sequencing are to be implemented in healthcare systems, evidence on value and value for money is required. Existing reviews suggest that the health economic evidence base for the translation of sequencing into clinical practice is under developed [5]. Although data on the cost of testing is emerging, few studies have evaluated and quantified all the elements of the value of sequencing that are important to patients and their families, and even fewer studies have brought both up-front and down-stream cost and outcomes data together in the economic evaluations required by health technology assessment agencies [6, 7].

Against this backdrop, this themed issue of Applied Health Economics and Health Policy presents a series of 12 papers on the “Health Economics of Genomic Technologies”. These papers summarise existing evidence on the value attached to the use of genomic technologies in a variety of clinical contexts, and present new health economic evidence for genomic testing in myeloid malignancies, non-small-cell lung cancer (NSCLC) and haemophilia B. Value is estimated in a variety of ways, sometimes going above and beyond standard approaches to health technology assessment, which quantify value in terms of incremental costs and quality-adjusted life-years (QALYs) gained. Study settings are international, including Australia, Canada, Scotland and Thailand, ensuring that this themed issue makes a meaningful contribution to expanding the global evidence base on value for genomic technologies. Included papers fall into three broad categories: cancer, rare diseases, and stakeholder preferences across a range of clinical contexts.

Cancer

The use of genomic information to guide clinical management for patients with cancer is inevitably a key focus of this themed issue. Oncology remains one of the most common precision medicine applications globally [8], with continuous genomic discoveries driving persistent changes in the cancer treatment paradigm [9]. Bourke et al set the scene by describing existing evidence on the cost effectiveness of genomic medicine in cancer control [10]. They report that the use of genomic testing to guide therapy is highly likely to be cost effective for breast and blood cancers and is also frequently cost effective for advanced and metastatic NSCLC. However, evidence of value for testing in other cancers is more mixed or insufficient. The authors highlight the need to expand the health economics evidence base for genomics more generally, calling for more explicit consideration of patient preferences, equity and system capacity, and increased application of value of information and budget impact analyses.

Two further papers in this themed issue present evidence on cost effectiveness for specific applications of genomic testing in oncology. Lindsay et al undertake a model-based cost-utility analysis of the use of a 37-gene panel test to guide care for patients with intermediate-risk myeloid malignancies in Australia [11]. Compared to usual care, the panel test increases costs with only a negligible improvement in QALYs. With an incremental cost-effectiveness ratio of AU$153,854 per QALY gained, panel testing has only a 1% chance of being the cost-effective alternative. Turongkaravee et al present a second model-based cost-utility analysis, this time using a genomic profiling strategy to guide the use of targeted therapy in NSCLC in Thailand [12]. They find that a sequential genomic testing approach can identify more patients eligible for targeted therapies, improving survival and QALYs gained compared to the standard genetic testing approach. However, the high cost of testing means that this sequential approach is not cost effective in this population and setting. This illustrates the challenges that might be encountered when trying to translate expensive genomic tests into clinical practice outside high-income country settings [13].

The studies by Lindsay et al and Turongkaravee et al apply a conventional approach to health technology assessment, measuring outcomes using QALYs, often from a healthcare provider perspective. Minhinnick et al present a scoping review that confirms that cost-utility analysis is commonly undertaken when considering the value of molecular tests in cancer [14]. They consider how broadly value has been defined in economic evaluations in this context, identifying 91 relevant studies, of which 75 applied this conventional approach. Few studies considered non-clinical elements of value that are potentially important to patients and their families, such as the psychological impacts of sequencing information or access to trials. This is an important finding, because there is a growing literature demonstrating that these broader impacts of genomic testing hold real value and should be considered within health technology assessment processes [15–17].

Consequently, research studies that apply alternative approaches to estimating the value of genomic technologies are valuable additions to the literature. Pataky et al demonstrate one such approach [18]. Using administrative data, the authors explore the value of using KRAS and NRAS mutation status to guide treatment decisions for patients with metastatic colorectal cancer, applying an approach called the value of heterogeneity (VOH) framework. Their results suggest that stratification of therapy by mutation status provides value that exceeds the marginal cost of genomic testing, and that the collection of data to reduce decision uncertainty could offer additional value.

Rare Diseases

Much of the initial research on using genomic technologies in clinical practice has focused on people with rare genetic disorders, and four studies in this themed issue present new evidence on the value of genomic information in this clinical context. Abbott et al investigate willingness to pay for genomic testing in patients and families in Scotland with experience of genome sequencing for rare disease diagnosis, using a contingent valuation payment card [19]. The authors also administered two genomic-related patient-reported outcome measures to measure negative psychological outcomes: the Personal Utility Scale (PrU), and a subscale of the Feelings About Genomic Testing Results (FACTOR) questionnaire. The contingent valuation results indicate that people receiving a diagnosis were willing to pay GB£2043 for genome sequencing, versus GB£835 in those who do not receive a diagnosis. The authors also found that negative psychological outcomes impacted on the value attached to genome sequencing.

Smith et al focus on the use of exome sequencing to diagnose rare disorders in the USA, comparing test use, diagnostic outcomes, and costs for children who undergo genome sequencing and children who receive less comprehensive forms of genetic testing [20]. Using linked family claims data, the authors report that exome sequencing improves diagnostic outcomes, leading to 1.44 more new diagnostic codes after testing than current practice. However, the authors also found that out-of-pocket costs are similar in both groups, indicating that the higher costs of testing in the population undergoing exome sequencing are not being covered by families. Instead, payers who see value in using these more expensive tests to generate diagnostic information are financing testing costs.

Degeling et al also evaluate the use of exome sequencing in patients with a suspected rare genetic disorder, using a retrospective dataset for 305 Canadian patients to investigate whether the timing of testing in the diagnostic pathway impacts cost effectiveness [21]. They report that exome sequencing increases diagnostic yield from 20 to 36% depending on timing, and that implementing exome sequencing as a first-tier test achieves the highest diagnostic yield at the lowest cost. As the authors note, this finding – that early use of genomic testing improves outcomes and saves money compared to existing testing practice – is reported increasingly frequently in different settings and clinical contexts, and there may now be sufficient health economic evidence to support more widespread use of first-tier genomic testing [22, 23].

Finally, Sarker et al shift the focus from rare disease diagnosis to clinical management, evaluating the cost effectiveness of a gene therapy – etranacogene dezaparvovec (EDZ) – to treat patients with haemophilia B in the USA, compared to conventional factor IX (FIX) treatment [24]. Etranacogene dezaparvovec is expensive, costing US$3.5m, but is only administered once, whereas FIX requires frequent intravenous administration. Using a Markov model, the authors estimate that EDZ is associated with lifetime cost savings of US$11m, while also improving QALYs in this population, making it the dominant alternative.

Stakeholder Preferences

The final group of three papers in this themed issue consider how information on preferences could be used to value genomic technologies across a range of clinical contexts. First, Salisbury et al present a systematic review of discrete choice experiments (DCEs) investigating public preferences for genetic/genomic risk-tailored screening for chronic diseases [25]. They identify 12 studies, most of which focus on cancer screening in targeted populations. The largest number of programme attributes were identified in the process category, but the attributes with the greatest impact on preferences and uptake were survival, test accuracy and screening impact. In alignment with past research, results indicate that both clinical management changes and test information drive value for genomic technologies [17].

A second study by Salisbury et al presents a DCE evaluating Australian preferences for non-invasive prenatal screening using cell-free DNA, with a particular focus on urban-rural differences [26]. Again, process-related attributes (e.g., wait times) impact preferences, in particular for participants from rural settings. The authors also report that participants were willing to pay AU$323 for a non-invasive prenatal test that screens for a wide range of conditions, over and above the standard prenatal screening offered in Australia. Differentiation between preferences of patients in urban and rural areas is a novel addition to the stakeholder preferences literature and highlights the importance of considering outcomes variation for equity-centred precision medicine implementation [27].

The third paper in this group takes a critical look at the published literature to examine how preferences have been incorporated into model-based economic evaluations of genomic medicine interventions, and to develop a conceptual framework of ways in which preferences influence economic value [28]. The authors (Smith et al) derive insights from 14 articles and form two important conclusions. First, accounting for preferences in the form of model inputs and valuation of outcomes in economic evaluations of genomic technologies helps to avoid biased implementation decisions, and second, incorporating preference data into these models could improve alignment between predicted and real-world uptake.

Concluding Comments

The 12 papers in this themed issue provide a comprehensive overview of the current evidence base and research focus of the field of health economics and genomics, which has evolved considerably over the past two decades [5, 29–32]. Today, a growing number of health economists apply cutting-edge methods to value the costs and benefits of innovative genomic interventions with the potential to transform population health and well-being [33–35].

However, there are challenges on the horizon. Many of the publications in this themed issue have clear policy relevance, but the extent to which economic evidence actually informs healthcare decision making in this clinical context is unclear, particularly non-traditional health technology assessment evidence on patient preferences and evidence derived from real-world data. While some jurisdictions are signalling increasing acceptability of patient preference information and real-world evidence for health technology regulation and reimbursement, apprehension remains [36–39]. Formal, harmonised, stakeholder-driven guidance is needed to ensure uptake of this health economic evidence into precision medicine decision making [40, 41].

A second challenge is the pace at which new applications of genomic technologies are emerging. The papers in this special issue broadly consider the value of genomic information in specific cancer, rare disease or chronic disease contexts. These cases likely represent just a subset of all future applications. There is growing interest in understanding the costs and benefits of newborn genome sequencing, polygenic risk scores, and multi-cancer early detection tests [42–44]. In health economics more generally there is a move towards equity informative economic evaluation, and genomics is one context in which issues around equity and fairness may fundamentally impact on estimates of value [45]. Finally, there is a growing need to understand the value of genomics at the population (macro) level [46], as well as in specific geographic contexts such as lower- and middle-income country settings. To date, health economists have undertaken little to no research on these topics for genomics. This should change in the coming years, and we are excited to see how this field of research continues to develop in the next decade.

Data availability

Not applicable.

Declarations

Funding

No funding was received to support the writing of this editorial.

Conflict of Interest

James Buchanan, Ilias Goranitis and Deirdre Weymann are the Guest Editors of this themed issue of Applied Health Economics and Health Policy. For articles in the themed issue with any of the Guest Editors as co-authors, that/those Guest Editor(s) were not involved in the selection of peer reviewers nor any of the subsequent editorial decisions. James Buchanan reports travel funding from Illumina, Inc. and consulting income from Genomics England. Deirdre Weymann co-directs IMPRINT Research Consulting and has consulted for AstraZeneca, Birota Economics Group, and LHS Labs.