PHARMA: ‘Our drug is better than theirs.’
PAYOR: ‘Prove it.’
As the cost of health care continues to climb, molecular diagnostic tools will play a greater role in helping insurers and other payors decide which drugs they will reimburse, according to an industry expert.
If a pharmaceutical company employs a biomarker test to demonstrate that its drug is more efficacious than a competitor’s — especially in a me-too category — insurers may be more apt to reimburse for the drug backed by the test. In addition, it may in some cases become more difficult for insurers to justify paying for drugs that aren’t backed by such molecular-diagnostic validation, and patients and their physicians may quickly abandon these drugs for generics or me-too competitors.
This phenomenon, which is being driven by the increased bills that insurers are paying for prescription medications, is slowly beginning to take root in the United States, and will likely drive the development of pharmacogenomics technologies and applications such as molecular diagnostics.
Between 1990 and 2001, the cost of pharmaceutical products incurred by patients, health-care systems, and payors in the United States roughly doubled, to $1.4 trillion, according to the Baltimore, Md.-based Centers for Medicare and Medicaid Services. During those 11 years, the portion of that bill paid by private insurance companies more than doubled, to more than 50 percent. By comparison, patient out-of-pocket costs for pharmaceuticals fell to 30 percent in 2001 from 60 percent in 1990; government programs such as Medicare experienced a slight, gradual increase to around 23 percent in 2001 from roughly 16 percent in 1990.
It’s clear, according to experts, that payors can’t accommodate much more of this growth without significant increases in premiums and cuts in reimbursement.
“What we have been hearing from various managed-care organizations … is that their major driver in costs … is pharmacy costs,” said Glenn Miller, scientific director of Genzyme Genetics.
Since most insurers have already begun increasing their premiums, the belief is that pharmacogenomics will help drive down the cost of prescription drugs by injecting into the industry a massive bolus of competition. Insurers will be happy to use gene-based data, derived from molecular diagnostics, to help them decide where to spend their money, the theory goes.
“There’s a tremendous push towards generics, if possible,” Miller told SNPtech Reporter. “However, they are willing to reimburse [for branded drugs] if … there are tests available that would demonstrate why an individual patient would benefit substantially from a particular drug as opposed to the generic equivalent.”
To be sure, there have been few “really good examples” of this phenomenon, he said. This is likely the result of a dearth of molecular diagnostics that test for efficacy, and a paucity of large-scale, head-to-head comparisons between me-too drugs. One reason for this is that most drug makers are reluctant to collaborate with competitors in a study of this nature. “They see it as a fragmentation of their market,” said Miller.
Indeed, gene-based studies designed to compare efficacy — such as Genaissance Pharmaceuticals’ STRENGTH trial, which sought to link haplotypes with drug response — have been known to irk the companies whose drugs are being reviewed.
Genaissance performed its STRENGTH trial largely on its own, and as a result, the company “has taken some harsh criticism from various quarters in the pharmaceutical industry for doing this,” Miller said. “It was a tough road for them to travel.”
Kevin Rakin, CEO of Genaissance, said: “I think we did take some flak … but I think the flak we took was more because we were a little too overly confident about ‘We will do this’ and ‘We will do that,’ as opposed to saying ‘This is a logical way to evaluate drugs and get a more rational prescription out there.’”
Rakin also said the concept of the STRENGTH study won praise — and a lucrative partnership — from Bayer Diagnostics when the German company asked Genaissance to develop SNP-based assays to help physicians determine which patients were more prone to adverse events, or did not respond. This trial was penned in January and will initially study statins [see 1/31/03 SNPtech Reporter]. Though the means in this case were not meant to influence economic ends — Bayer no longer sells a statin; rather, it hopes to sell the assays to the much smaller clinical-lab market — trials like this will have de facto influence over which drugs are prescribed, and which ones are reimbursed by payors. Makers of statins, and eventually other me-too categories, ought to take note.
“I think most drugs are doing pretty well if they’re efficacious,” said Miller, who explained that a 30-percent response rate for a chemotherapy “is a good number.” The FDA, he said, “has traditionally been a ‘safety-first’ organization, and not an ‘efficacy-first” organization. And that’s as it should be. But what that does is take everything to a safety level and not necessarily to an efficacy level.
“I think for pharmaceutical companies that know they have a share based on marketing … as opposed to a share of a genetically determined market, they’re very comfortable in the marketing end of it, and they don’t know the absolute endpoint of a genetically determined market,” he added.
Miller said this is the case because for many drugs genetic studies have not been done. “It’s a bit of the share of the unknown; it’s a bit of how good those studies actually are. And for many of the drugs, it’s relatively unclear what the genetic background does, and what the real mechanism of it is, and if it’s more than just a phenomenological event.”
This isn’t to say that the ideas of gene-based medicine and molecular diagnostics are completely lost on pharma. In fact, “a great number of drug makers are busy creating pharmacogenomics operations; very few companies don’t have some sort of pharmacogenomics capability right now,” said Miller. However, he said, “the general push is to create drugs that don’t depend on genetics, and it is always going to be cheaper to screen compounds than it is to screen people.”
Consequently, if pharmas can develop a compound that works in the majority of patients for whom it will be prescribed, “that’s a far better proposition, and that’s why they’re using genetics. They’re not using genetics to come up with a diagnostic-therapeutic pairing; they’re using genetics to avoid having to use them,” he said. “Then you truly have your blockbuster.”
Miller declined to say which payors Genzyme Genetics had spoken with in identifying this trend, but said they were “the very large national payor organizations, certain regional organizations thinking about how they can control their costs; and organizations that manage self-funded health plans. It’s really across the board,” he said. “By far their major concern is the pharmacy costs now, because the drugs are so expensive.”
Bill Crown, vice president of outcomes research and econometrics for the Medstat Group, said that molecular diagnostics may indeed become factors in payors’ reimbursement decisions, and that payors will try to use these technologies as leverage for keeping at bay their exposure to risk — but within limit.
“Some of the best examples of targeted therapies right now are the biologics,” he said, referring to drugs like Remicade, made by Centocor, a unit of Johnson & Johnson. “Those treatments tend to be significantly more expensive than traditional pharmaceuticals. And the issue of biomarkers, in some sense, is at what point do they come in? Do you give everybody the tests and identify the people that would respond to a particular therapy, but who also might respond to a conventional therapy? Or do you wait until there is evidence that patients aren’t responding to any of the traditional therapies, and then … give them the diagnostic test and treat those that the test suggests would be? This is the way I think it will likely play out right now.
“This is still early days,” added Crown. “For most of these biological treatments, I don’t think there are specific biomarkers. But that doesn’t mean they couldn’t be developed; the ability to combine a diagnostic test with a targeted therapy is something I think exists only in very narrow market segments,” such as Gleevec and Herceptin.
Asked if the early success of therapeutics such as Herceptin, Gleevec, and Xeloda might embolden drug makers to pursue companion diagnostics for their therapeutics, Crown said “there are definitely two camps: there are the companies that are thinking about targeted therapies and personalized medicine. And those companies are very definitely thinking about this issue of developing diagnostic tests and pairing them with treatment,” he said. “But the traditional pharmaceutical model — to treat a broad a patient population as possible — is still the more pervasive. We haven’t won that battle yet.”