Personalized Medicine: Seven Keys to Success

Compelling science doesn't always make a good commercial case. Here's how to tell the difference.

Stephen Little
Jan 1, 2006
<figcaption> Credit: © Ole Geisselbrecht</figcaption>
Credit: © Ole Geisselbrecht

Several years ago I tried to persuade a pharmaceutical product manager that coupling his successful drug with a diagnostic assay would boost its efficacy and safety. Although he agreed with my technical arguments, he said there was no possibility of introducing such an assay because doctors would drop the drug - one of many in its class - to the bottom of their priority list if patients first required testing. My suggestion of personalized medicine wasn't flawed from a scientific perspective, but it was destined for obscurity from a business sense because it meant the company would sell fewer drugs.

The lesson for companies involved with personalized medicine is clear: You can have the most compelling arguments in the world for using a particular diagnostic. But without a good commercial case, the project is most likely doomed.

As I learned several years ago, the realities of...

Stephen Little is CEO and cofounder of DxS Ltd, a UK biotech company providing pharmacogenomic services and products to the healthcare industry.


1. B.B. Spear et al., "Clinical application of pharmacogenetics" Trends Molec Med, 7:201-4, 2001. 2. J.A. Diaz et al., "Patients' use of the Internet for medical information" J Gen Intern Med, 17:180-5, 2002.

Consider These Issues When Drawing up a Business Plan

1. Product Differentiation
Many drugs in the development pipeline are destined for treating diseases that already have successful therapies, and many of those therapies are becoming available as lower-cost generics. To compete in a crowded market, pharmaceutical companies need to ensure their new drugs really are better than those of their competitors. A second approach is to make better use of drugs through personalized medicine, which can reposition a drug from "as good as the competition" to "best in class" For example, AstraZeneca's cancer drug Tomudex was launched as a competitor to the cheaper generic 5-flourouracil. Perhaps if the company had offered an accompanying diagnostic to target the drug to those patients who would have been expected to show an improved response over the generic, the drug could have been more successful.

2. Patent Protection
By the time a therapeutic reaches the marketplace, it already has used several years of its patent protection. The discovery of a pharmacodiagnostic tends to occur later in the development cycle, typically around Phase IIa. A drug company may be able to extend its proprietary advantage for several additional years by obtaining patent protection for the combination of the drug and the diagnostic, although it's still to soon to tell whether this strategy will pay off.

3. Product rescue
New drugs don't always perform as well as expected. For example, although AstraZeneca's Iressa is very successful in Asian markets, it is not doing as well in the West owing to lack of efficacy. Some recent publications suggest the company might revive its fortunes in the United States and Europe if it introduces a genetic diagnostic to target individuals carrying mutations indicative of a positive response to the drug. AstraZeneca now must balance the cost of a diagnostic against potential sales and determine if this is a commercially sound strategy.

4. Diagnostic Threats
The diagnostic industry is eager to share in the value of pharmaceuticals, and they are developing new assays to improve drug efficacy and safety. The US Food and Drug Administration's decision to relabel Pfizer's Campto in light of recently emerged diagnostic data illustrates the impact a diagnostic can have on an established therapy. This development should encourage drug companies to be more proactive in taking control of biomarker discovery. From pharma's perspective it is bad enough that a drug label is changed, but how much worse is it if the intellectual property for the diagnostic specified on the drug label is owned by a third party?

5. Payer Pressure
It's a sobering fact that most drugs don't work most of the time,1 so it is hardly surprising that payers are interested in converting from payment for pills to payment for results. A prime example of this strategy is the risk-sharing scheme introduced by the UK's National Health Service (NHS) regarding the sale of interferons for the treatment of multiple sclerosis (MS). In the United Kingdom, the National Institute for Clinical Excellence (NICE) is responsible for health technology assessment (i.e., determining if a drug is good value for money). NICE ruled that it was not cost effective to supply interferons to all patients with MS. Faced with the prospect of selling no drugs at all, the manufacturers - Biogen, Serono, Teva, and Schering - of the drugs in question (Avonex, Rebif Copaxone, and Betaferon,) agreed to supply the therapies to patients selected by diagnostic criteria as likely to respond well. The drug makers also set agreed performance benefits for the drugs measured in quality-adjusted life-years (QALY). If the selected patients achieve the benefits predicted by the manufacturers, the NHS pays for the drugs at the agreed price. If the benefit fails, the drug price falls to maintain the cost per QALY.

6. The Educated Patient
The Internet has become a vital tool providing curious patients with medical and health data. Approximately 88.5 million adults used the Internet to seek online health information in the United States during 2005.2 The amount of time spent online is directly related to the severity of the disease, so many chronically ill patients are well informed about scientific developments, clinical trial findings, and alternative drug treatments. By way of example patients are already requesting genetic analysis to identify mutations that may predict response to EGFR antagonists such as Roche's Tarceva. At this time it is certainly not clear whether such testing has clinical validity, but the demand is evident and this will no doubt encourage clinical laboratories to provide the service. In September 2005, Genzyme announced the commercial availability of just such a test to help identify patients likely to respond to therapies targeted for the treatment of non- small-cell lung cancer.

7. Regulatory Push
The FDA encourages the use of pharmacogenomics and personalized medicine through its Critical Path Initiative, which aims to evaluate delays in the adopvtion of new technology in drug development. In the last 18 months there have been more than 15 publications, guidances, and concept papers supporting the future development of personalized medicine. For example, in the recent "pharmacogenomic data submissions" guidance for industry, the FDA writes: "As the field of pharmacogenomics advances, it is likely (and desirable) that sponsors will begin to use pharmacogenomic tests to support drug development and/or to guide therapy." Currently, the FDA does not mandate the use of pharmacogenomics in drug development, but it would be a bold pharmaceutical company that chose to ignore such clear directions.