Competing with precision: incentives for developing predictive biomarker tests

• Published date: 01 January 2024
• Author: Kurt R. Brekke,Dag Morten Dalen,Odd Rune Straume
• Date: 01 January 2024
• DOI: http://doi.org/10.1111/sjoe.12543

Scand. J. of Economics 126(1), 60–97, 2024

DOI: 10.1111/sjoe.12543

Competing with precision: incentives for

developing predictive biomarker tests

Kurt R. Brekke†

Norwegian School of Economics (NHH), NO-5045 Bergen, Norway

kurt.brekke@nhh.no

Dag Morten Dalen

BI Norwegian Business School, NO-0442 Oslo, Norway

dag.m.dalen@bi.no

Odd Rune Straume‡

University of Minho, 4710-057 Braga, Portugal

o.r.straume@eeg.uminho.pt

Abstract

We study the incentives of drug producers to develop predictive biomarkers, taking into account

strategic interaction between drug producers and health plans. For this purpose, we develop a

two-dimensional spatial framework that allows us to capture the informational role of biomarkers

and their effects on price competition and treatment choices. Although biomarkers increase the

information available to prescribers, we identify an anticompetitive effect on the prices set

by producers of therapeutically substitutable drugs. We also find that better information about

each patient’s most therapeutically appropriate drug does not necessarily lead to more efficient

treatment outcomes.

Keywords: Pharmaceutical markets; precision medicine; therapeutic competition; predictive

biomarkers

JEL classification:I11; I18; L13; L65

1. Introduction

Although the advancement of medicine offers new treatment opportunities

for patients with severe diseases, individual treatment responses often vary

substantially. If the average treatment effect of a drug (i.e., measured by

gained quality-adjusted life years) is sufficiently high relative to incremental

treatment costs, the traditional approach by many health plans has been to

allow physicians to prescribe the drug based on a “trial and error” basis.

†Also affiliated with the Centre for Applied Research at NHH.

‡Also affiliated with the University of Bergen.

c

2023 The Authors. The Scandinavian Journal of Economics published by John Wiley & Sons Ltd on behalf of F¨

oreningen

f¨

or utgivande av the SJE.

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution

and reproduction in any medium, provided the original work is properly cited.

K. R. Brekke, D. M. Dalen, and O. R. Straume 61

Consequently, many patients who receive expensive treatment will not see the

health improvements they could hope for, or even experience more serious

side effects than others. According to Anto˜

nanzas et al. (2018), over 90 percent

of drugs work for fewer than half of the patients they are prescribed for.

Improvements in the technology for sequencing the human genome have

enabled more precise tailoring of treatments within groups of patients sharing

the same diagnosis. Increased precision of interventions is achieved by

exploring predictive biomarkers, which “identify individuals who are more

likely than similar individuals without the biomarker to experience favourable

or unfavourable effects from exposure to a medical product”.1Instead

of treating many patients and accepting lower response rates, biomarkers

associated with molecular and genetic characteristics are used to narrow

down the number of patients that are given a specific treatment. Patients

without these biomarkers can instead be offered other treatment alternatives

or no treatment at all, thus avoiding the burden of receiving ineffective

treatment.

The potentially large benefit to patients and society of improved precision

of medical treatment has been recognized for several decades already, since

the start of research efforts to determine the DNA sequence of the entire human

genome (Langreth and Waldholz, 1999). Parallel to the race between the two

major sequencing projects, The Human Genome project and Celera Genomics,

the pharmaceutical industry started to invest in mapping variations in the

human genome, referred to as the Single Nucleotide Polymorphisms (SNP)

Consortium, already then aiming for precision, or personalized medicine

(International SNP Map Working Group, 2001). So far, predictive biomarkers

have made most progress in oncology, but other therapeutic areas are also

experiencing progress in detecting biomarkers that can provide prescribing

physicians with better information about which individuals are likely to

respond to a given therapy (for a recent review, see Jørgensen, 2021). Although

initial excessive optimism was replaced with a period of dissatisfaction about

the progress of personalized medicine (Towse and Garrison, 2013), it is

expected that we will continue to see research effort into precision medicine,

with development of specific biomarkers to inform prescription choices

(Stern et al., 2017).

Predictive biomarkers challenge economic regulation and coverage

decisions of regulators and health plans. By detecting biomarkers for new

and existing therapies, drug producers run the risk of reducing the size of the

market as non-responding patients no longer are going to be treated. Unless

drug prices are sensitive to improved precision, the incentives to develop

1See the definition offered by the FDA-NIH Biomarker working group (https://www.ncbi.nlm.

nih.gov/books/NBK338449/).

c

2023 The Authors. The Scandinavian Journal of Economics published by John Wiley & Sons Ltd on behalf of F¨

oreningen

f¨

or utgivande av the SJE.

62 Incentives for developing predictive biomarker tests

biomarkers are weak (see, e.g., Scott Morton and Seabright, 2013; Towse and

Garrison, 2013; Stern et al., 2018).

Despite regulatory challenges being identified and discussed in the

literature, the effect of biomarkers and precision medicine on competition

in pharmaceutical markets remains underexplored. On the one hand,

patent-holding drug producers enjoy market power, giving rise to

price-setting flexibility. Health plans, on the other hand, enjoy countervailing

monopsony power, first and foremost by controlling access to their plans

(Lakdawalla, 2018). The decision to develop a biomarker is clearly a

strategic choice by drug producers that is likely to affect the dynamics of

competition. An illustrating example is the introduction of a biomarker for the

immuno-oncology drug Keytruda sold by Merck.2This drug faced competition

from Opdivo by Bristol Myers Squibb for treating several types of cancer.

While the biomarker reduced the sales of Keytruda due to fewer patients, the

efficacy of the drug improved relative to Opdivo not using a biomarker and

being tested on a broader population. Merck’s launch of a biomarker turned

out to be a crucial and profitable strategy for the success of Keytruda.

Our paper aims at developing new knowledge about how predictive

biomarkers affect the strategic interaction between drug producers and health

plans, and how this feeds back to the incentives to develop biomarkers

in the first place. By exploring the equilibrium impact of biomarkers on

prescription choices, drug prices, and health benefits, the analysis improves

our understanding of the economic regulatory challenges of precision medicine

by identifying potential sources of inefficiency.

We consider a market for prescription drugs that is served either by

a monopolist or by two producers supplying different, but therapeutically

substitutable drugs. A drug producer can only gain access to the market if

the health plan is willing to sign a contract with the producer, and these

contractual decisions determine which of the drugs can be prescribed by

physicians affiliated with the health plan. Both producers can develop a

predictive biomarker that, if included in the plan, will inform prescribing

physicians about the therapeutic match between the specific drug and the

patient. A drug without a biomarker can still be included in the health plan,

but physicians then need to rely on the average performance of the drug, as

learned from clinical trials and use, when making treatment choices.

We develop a spatial competition framework with up to two drugs available

and a distribution of patients who differ with respect to their therapeutic match

with each drug. The two drugs represent alternative treatment options, with

2See, for instance, the article “A pharmaceutical firm bets big on a cancer drug” in The Economist,

24 February 2018 (https://www.economist.com/business/2018/02/22/a-pharmaceutical-firm-

bets-big-on- a-cancer-drug).

c

2023 The Authors. The Scandinavian Journal of Economics published by John Wiley & Sons Ltd on behalf of F¨

oreningen

f¨

or utgivande av the SJE.