"Personalized medicine," "precision medicine," "pharmacogenomics" or whatever you want to call it, it's pretty clear the industry is moving in the direction of having drugs to treat patients based on their genetic makeup.

According to a report from Tufts Center for the Study of Drug Development, drugs classified as personalized medicine accounted for 20 percent of FDA approvals last year, increasing the fraction of currently marketed drugs with pharmacogenomic information on their labels to 13 percent, up from 10 percent in 2010.

"We continued to see major R&D resources being pumped into personalized medicine. Coming from a baseline of pretty much zero 15 years ago, we have virtually all major companies involved in some way with personalized medicine," said Joshua Cohen, associate professor at Tufts Center for the Study of Drug Development and one of the study's authors.

Investments in developing personalized medicine nearly doubled over the last five years and companies expect to increase spending by 33 percent on average over the next five years, according to the report derived from a "survey of 15 leading manufacturers of personalized medicines and interviews with 22 leading drug and diagnostics companies involved in the development of personalized medicines." The increased spending is expected to boost the number of personalized medicines in development by 69 percent.

Oncology leads the way, with 73 percent of compounds relying on biomarker data, compared to just 42 percent of all drugs. Cohen attributed oncology's dominance to the fact that "more is known scientifically about the subpopulations." That knowledge is actually driving the development of drugs targeting those specific subpopulations.

Zelboraf (vemurafenib, Roche AG), Tafinlar (dabrafenib, Glaxosmithkline plc) and Mekinist (trametinib, Glaxosmithkline plc), for instance, all target BRAF and are designed to specifically help melanoma patients with BRAF mutations. If tumors with the mutations weren't known, it seems unlikely the drugs would ever have been developed.

For other specialties, developing personalized medicines is more challenging. It's clear that drugs often work better for some patients than others, but it isn't nearly as easy to elucidate whether there's a genetic component to the difference in efficacy, and if there is, whether a biomarker can be identified.

The Tufts' study cited the identification of biomarkers as one of the issues that companies feel is a substantial challenge in developing new personalized medications.

"In many instances, there are multiple biomarkers that each have a different weight, and how to measure that weight, in terms of the contribution of a subpopulations responsiveness to a drug, all has to be worked out. Tailoring to a subgroup is a difficult process," Cohen told BioWorld Insight.

DEVELOPING TESTS

Only eight FDA-approved tests have been co-developed with drugs, but seven of those were approved in the last five years, highlighting the gaining traction of personalized medicine.

Another 14 drugs have FDA-approved companion diagnostics that were developed after the drug was approved. The added information of how to stratify the population to find the subgroup most likely to respond often can lead to increased use.

Cohen cited the post-approval discoveries that patients with KRAS mutations didn't respond to Erbitux (cetuximab, Bristol-Myers Squibb Co. and Eli Lilly and Co.) and that Iressa (gefitinib, Astrazeneca plc) was more effective in patients with EGFR mutations as two examples of where companion diagnostics were able to help drugs become personalized medicines after they were already on the market.

The rest of the personalized medicines – recall 20 percent of drugs approved in 2014 fall into that class despite the lack of any co-developed FDA-approved test approved last year – use already-approved tests or use laboratory developed tests (LDTs) that don't require FDA approval.

LDTs often are cheaper to develop than FDA-approved tests, but their popularity also has been fueled by a lack of guidelines from the FDA on how to gain approval for co-developed tests. Fortunately, the agency has issued a number of draft guidances recently that should alleviate the regulatory uncertainty.

As the regulatory pathway becomes more certain, Cohen predicted an increase in co-developed FDA-approved diagnostics. In a separate study two years ago, Tufts identified 23 products in phase II and phase III development that have diagnostics being co-developed with the products.

"I would suspect that in the coming years we'll certainly see surges like we saw in 2012-2013 with three or four or maybe even five co-developed combinations per year," Cohen said.

POST-APPROVAL CHALLENGES

Developing a test and gaining FDA approval for a personalized medicine drug is only half the battle. Companies still need to gain reimbursement and get doctors to prescribe the drugs, which can be more difficult than for a typical medication.

Payers will only pay for tests if there's clear evidence of improved outcomes from the information gained from the test.

Cohen pointed to tests to determine patients' sensitivity to warfarin as a good example of payers' unwillingness to pay for a test even though the tests can accurately determine a patient's risk for bleeding.

A 2013 study published in The New England Journal of Medicine found that having the results from the genetic test didn't help physicians get more of their patient into the desired range of blood thinning than for patients where the genetic test wasn't performed.

"You can tell that a patient is at bleeding risk, but what is the physician supposed to do with that information?" Cohen pointed out.

Even if companies can get their drugs and corresponding tests approved, they still need to get physicians to understand how to use the two together.

In the Tufts study, 90 percent of interviewees said that familiarity with genomic medicine among physicians was inadequate. The companies are using a variety of methods to help physicians understand how to use the medicines and tests. Adding the medicine and test to clinical practice guidelines and making sure there are detailed recommendations for testing on the labels are companies' top two methods for increasing awareness.

"It's a key bottleneck that is important to overcome," Cohen said of physician familiarity of personalized medicines.