Policy ForumBiomedical Innovation

How economics can shape precision medicines

See allHide authors and affiliations

Science  17 Mar 2017:
Vol. 355, Issue 6330, pp. 1131-1133
DOI: 10.1126/science.aai8707

Many public and private efforts in coming years will focus on research in precision medicine, developing biomarkers to indicate which patients are likely to benefit from a certain treatment so that others can be spared the cost—financial and physical—of being treated with unproductive therapies and therapeutic signals can be more easily uncovered. However, such research initiatives alone will not deliver new medicines to patients in the absence of strong incentives to bring new products to market. We examine the unique economics of precision medicines and associated biomarkers, with an emphasis on the factors affecting their development, pricing, and access.

We focus on precision medicines, those focused on biomarker-defined patient subgroups. This is distinct from the broader practice of precision medicine, which includes development of new products and new ways to tailor treatments to individual patients. Successful development of precision medicines requires an evidentiary basis to support their efficacy and the establishment of analytical and clinical validity of their associated biomarkers and assays. We focus on the U.S. context, given its unique regulatory and insurance systems and the fact that as the world's largest health-care market, its innovation policies have an outsized effect on global biopharmaceutical innovation.


Incentives for precision medicine innovation differ from other drug development incentives due to precision medicines' tendency to change the perceived market size, as well as raising issues regarding companion diagnostic development. The market may contract if a biomarker identifies a patient subgroup most likely to benefit, suggesting discontinuation of the therapy for patients without the relevant biomarker. The market would expand if a therapy comes with adverse side effects whose likelihood can be reduced with knowledge about a biomarker.

To the extent that many precision medicines will shrink patient markets, the history of drugs for rare or “orphan” diseases can inform our understanding of development incentives. All else equal, smaller markets attract fewer entrants, and strong evidence of this phenomenon exists in the pharmaceutical setting (1, 2). The fact that small markets may not incentivize a socially optimal amount of pharmaceutical innovation prompted the passage of the U.S. Orphan Drug Act of 1983 (ODA). The ODA creates incentives to encourage manufacturers to develop new drugs for orphan diseases, defined (arbitrarily) as those affecting fewer than 200,000 people, corresponding to an overall prevalence of roughly 1 in 1500 (or 65 in 100,000). A manufacturer receiving U.S. Food and Drug Administration (FDA) approval for an orphan drug receives tax credits (50% of its clinical trials expenses) and extended marketing exclusivity (7 years, versus 5 years for nonorphan drugs). These incentives increase the expected profitability of new orphan drugs and are believed to have spurred commercialization of more than 516 medicines for more than 450 different orphan diseases since the passage of the ODA (3). In 2015 alone, 47% of novel drugs approved were orphan drugs (4).

The incentives provided by the ODA mean that manufacturers of precision medicines should be particularly eager to find biomarkers that allow them to bring their medicines to market as orphan drugs, including salvaging some projects by showing effectiveness in narrower populations. Indication subdividing will most likely occur when the prevalence of a disease in a subpopulation is below the 200,000-patient threshold (5). In this scenario, manufacturers may use biomarkers to identify one indication that would benefit from the ODA's provisions and then rely on off-label use or pursue subsequent trials for indications for larger populations. To the extent that innovation is cumulative—i.e., obtaining one orphan indication for a precision medicine makes it easier for the same manufacturer to develop another orphan indication and/or drug—some may call for revisions to the ODA. However, it remains to be seen how precision medicines will threaten or entrench the ODA as a policy framework.

More complicated is a situation in which a biomarker identifies a subtype with prevalence greater than that of an orphan disease and where the benefit in other populations is small, perhaps because of a unique genomic target. Here, manufacturers' incentives to develop new therapies depend on whether the biomarker allows a manufacturer to charge a high enough price to have incentives to enter a smaller patient market.

Several existing FDA regulatory designations, such as the Priority Review, Fast Track, and Breakthrough Therapy designations, as well as the Accelerated Approval Pathway, provide incentives (faster approvals and larger economic rewards) for manufacturers to commercialize medicines, regardless of the size of their target populations. If precision medicines offer higher therapeutic value relative to existing treatments, or a tighter link between surrogate endpoints and more clinically relevant outcomes, they will likely qualify for these designations. A better understanding of how precision medicines will be considered for such programs will be important for understanding which precision medicines are developed (6).


Four factors will drive up prices for precision medicines relative to conventional therapies. The first, related to the orphan drug context, affects the path of prices after product launch. In small markets, competition from other new entrants will be limited. This leads to less brand-brand competition early in a product's life cycle, followed by (statutorily) delayed and subsequently low generic entry, even after patent expiration.

Second, to the extent that precision medicines are more likely to be biologic drugs, prices will reflect their more costly and technology-intensive manufacturing. Despite the creation of an abbreviated approvals process for biosimilars (follow-on biologics), entry in the United States has been limited, with only one product launched todate. In light of anticipated FDA guidance and state-level policies that make it unlikely that biosimilars will be treated as interchangeable with their reference products by U.S. physicians and pharmacists in the near future, competition from biosimilars will be reduced relative to the role played by generic chemical drugs (7).

Third, biomarkers identify the subtype of patients in whom a treatment will be most effective, and more efficient targeting enables manufacturers to charge higher prices to reflect higher efficacy compared with unselected populations (8). Paying more for quality in the form of higher response is not necessarily bad for patient welfare. Even if prices are high for a given patient, total spending on a drug across all patients could still be small if only a few patients are candidates for the therapy. This will cease to be the case, however, as more of the population is treated with precision medicines.

Fourth, if R&D costs are higher for precision medicines than for traditional therapies, then the medicines launched will be only those with (potential) prices high enough to justify expected R&D expenditures. This applies to other medical innovations but is likely to be more acute among precision medicines, where R&D costs are expected to be higher and more salient because of the need to develop companion diagnostics.

A further dimension of pricing strategy concerns the manufacturing of cellular and gene therapies. There are two methods of production: autologous, where a patient's own cells are expanded ex vivo into treatments such as chimeric antigen receptor T cell therapies and injected back into the patient, or allogenic, where cells from one donor are used in many patients. Allogenic “off-the-shelf” therapies will resemble conventional high-cost biologics, but autologous treatments may function more like a service business, where hospitals and physicians license equipment or technology for small-batch manufacturing.

Biomarkers and Diagnostics

The promise of precision medicine relies on identifying patient or disease factors that predict the efficacy of a given therapy. Yet incentives to develop biomarkers are less well understood and their regulatory pathways less certain. One motivation is trial “enrichment,” in which a patient characteristic such as a biomarker is used to define a study subpopulation so as to maximize the likelihood of finding a drug's effect. FDA guidance on enrichment strategies for clinical trials provides illustrative examples (9): If a biomarker identifies a subpopulation that exclusively benefits from a therapy and that subpopulation represents only 25% of the overall disease population, enriching the trial for these patients provides sample size efficiencies of 16x. If the biomarker predicting exclusive benefit is present in 75% of the population, efficiencies are only 1.8x. In the former case, a positive therapeutic signal is unlikely to be discovered without enrichment, providing clear incentives to develop a biomarker. In the latter example, a therapeutic signal could be discovered in the full population without enrichment, leading to weak incentives to develop a predictive biomarker.

Embedded Image

A second motivation to find predictive biomarkers lies in the ability to segment the patient population and charge higher prices to patients who benefit most from a precision medicine. A third reason is motivated by payers and capitated providers (those in an accountable care organization or those paid by bundled payments), who have financial incentives not to overuse high-cost drugs. These entities have potential to generate additional demand for biomarkers for high-cost drugs.

Closely related to predictive biomarkers are tests for them, typically biomarker assays that are codeveloped as “companion diagnostics” with a precision medicine. Incentives for developing companion diagnostics are related to the predictive capability of the biomarker. However, because many hospitals operate their own biomarker platforms and laboratory-developed tests without purchasing third-party diagnostic devices, incentives to develop companion diagnostics will be attenuated relative to incentives to find predictive biomarkers (10).


As more patients are categorized into biomarker-defined subpopulations, linkages between higher efficacy and higher prices will become more conspicuous, putting pressure on public and private payers. Three specific challenges are noteworthy.

To the extent that some precision medicines reveal themselves to be curative, it is not clear how the current payment system will absorb their costs. Due to churn in insurance markets, the insurer that pays for therapies, often expensive and administered during a short window of time, and the insurer that saves from future reductions in expenditures (most likely Medicare) are unlikely to be the same, creating a disconnect between coverage and long-term value (11). An inability for insurers and patients to pay for such drugs will reduce firms' incentives to develop them. Economists have argued for new financial instruments, which would function like mortgages to spread the costs of high-value, high-price treatments over time, decreasing the upfront financial burden for patients and payers (11, 12). Such schemes increase the value of insurance, which in turn increases demand for treatments and leads to higher prices, although not automatically to higher spending. Such instruments will be more effective when combined with outcomes-based contracts and/or clinical trials establishing a treatment's effectiveness.

Closely related to the idea of spreading out the cost of precision medicines over time is the idea of spreading out their cost over more individuals. Precision medicines may require larger insurance pools to spread payments for these treatments, especially for employer-provided insurance, which covers more than half of Americans. Employers of all sizes will struggle to absorb these costs, but smaller employers would be devastated by the presence of a single employee who needed access to a million-dollar curative therapy. Publicly financed “high-risk pools” may help cover high-cost therapies. In such pools, the government uses tax revenues to cover certain precision therapies (assuming that it is possible to raise taxes or find revenues elsewhere). This would be analogous to “catastrophic” reinsurance plans for Medicare Part D (13).

Alternatively, policies that decouple insurance from specific firms—by encouraging employers to purchase insurance on exchanges where multiple employers pool patients—would help spread risk. In the absence of larger insurance pools, insurers will have an incentive to cherry-pick patients with a low propensity to need precision therapies. Although “spreading the risk” via a larger insurance pool may lower per capita premiums, this is different from reducing the price of a therapy, which would still be high.

Finally, the high effectiveness of precision medicines means that payment decisions based on comparative effectiveness are unlikely to reduce the budget pressure from these therapies; cost-effectiveness will have to be considered (14). Creating price competition can provide financial relief for patients and payers. This includes policies to expedite biosimilar review times at the FDA and to encourage physicians to actively prescribe biosimilars when there is clinical evidence that they are effective substitutes (even in the absence of automatic pharmacy substitution). It also will mean stimulating more brand-brand and biologic-biosimilar competition, which depend on payers' ability to use formularies and competitor products in those formularies. These policies require a richer understanding of the process by which companies bring drugs to market. Clear characterization of the precision medicine development pipeline—including its sensitivity to economic incentives such as exclusivity periods, effective patent length, public funding, and the roles of early stage companies and more mature players—will allow policy-makers to more accurately anticipate the likely profiles of medicines that will reach the market in years to come (15).

Looking Forward

Despite the potential link between the high price of precision medicines and lower access to them, establishment of genomic databases and validated biomarkers is expected to decrease the cost of trials and time-to-market by allowing smaller, more focused clinical studies, particularly in the more expensive, later phases of development. Platform trials that incorporate multiple therapies and biomarkers with a common control arm can provide significant efficiencies to test and prioritize promising therapeutic hypotheses (16).

Layering Bayesian adaptive techniques that use information as it accumulates to focus resources on the most promising arms and to investigate surrogate endpoint relationships in real time offers even more potential. The most notable example is the I-SPY 2 trial of neoadjuvant experimental therapies in breast cancer, which recently identified two therapies with biomarkers that have a high probability of success in phase III studies (1719).

Reductions in both the cost and length of trials mean that more drugs can clear the hurdle of commercial viability. A lower hurdle for commercial viability will lead to more innovation, which can, in turn, create more competition. But smaller and shorter trials necessitate a strong and highly engaged, appropriately resourced FDA, working at the cutting edge of regulatory science. This reinforces the importance of state-of-the-art regulatory science and policy in facilitating precision medicine development.


View Abstract

Navigate This Article