Skip to main content

Randomized Studies Not Needed When Relabeling To Add Genetic Data; Observational Trials OK

WASHINGTON, DC – Randomized controlled clinical trials — the US Food and Drug Administration’s gold standard for drug development — are not universally necessary when a sponsor wishes to update a label for an older drug to include genetic testing data, an FDA official said at a pharmacogenetics conference held here this week.
Separately, the official said the FDA plans to issue a final guidance on labeling that will place pharmacology and genetic data more prominently for doctors to see (see briefs, this issue).
For drugs such as warfarin, which has been on the market for more than 50 years and for which the safety and efficacy are well established, randomized controlled clinical trials perhaps set the bar prohibitively high, Lawrence Lesko, director of FDA’s Office of Clinical Pharmacology & Biopharmaceutics, told an American College of Clinical Pharmacology meeting on “Clinical Applications of Pharmacogenetics.”
In these circumstances, observational studies are sufficient to show how genetic data can be meaningful in treatment decisions, Lesko said.
“The thing I try to remind people is that old drugs that are chosen for relabeling have already shown to be safe and effective; that’s why they are in the marketplace,” Lesko said. “All we want to do with this genetic information is improve the one-size-fits all dosing in order to optimize benefit/risk.”
He acknowledged that performing prospectively designed randomized clinical trials to validate dosing based on genetic factors can be prohibitively expensive, which in turn can hinder pharmacogenomic advances. 
“I’m afraid that if we put this standard out for every genomic test that we’ll be going nowhere fast,” Lesko said.
Indeed, the FDA’s willingness to accept less-rigorous observational studies could significantly ease the burden on diagnostics companies and help encourage drug and diagnostic companies to invest in the studies necessary to “personalize” drugs already out on the market.
However, Lesko stressed that FDA’s flexibility does not mean that the agency is lowering the bar for clinical evidence. Rather, by encouraging companies to pursue relabeling, the FDA hopes to raise the level of accuracy with which older drugs are administered. 
He suggested that in order for sponsors to update older drug labels with genetic data, industry, clinicians and the FDA should together outline a hierarchy of studies that sponsors should conduct based on the clinical question being considered.
Lesko’s comments caused a stir at the AACC conference, attended by clinicians from industry and academia, since “prospectively designed randomized clinical trials” have long been considered FDA’s mantra for valid clinical trial design.
Industry observers have often criticized that FDA needs to incentivize Rx/Dx co-development in order to fuel market adoption of diagnostics. Lesko’s comments suggest that the less-rigorous requirements could make life easier for diagnostic companies by freeing resources, manpower, and time they would otherwise dedicate to conducting large-scale randomized trials.
Yet some conference speakers noted that data requirements posed by payors also have made it difficult for diagnostics firms.
Raju Kucherlapati, scientific director of the Harvard Partners Center for Genetics and Genomics, noted at the meeting that the bar for what payors consider “convincing data” is “set very high.” The HPCGG recently launched a study to measure patient outcomes and cost when genetic information is used to make clinical decisions about warfarin [see PGx Reporter 11-15-06].
Medco, the second-largest pharmacy benefit manager in the US, has said that payors need to see the “value proposition” of incorporating genomic data into coverage decisions in order to become comfortable with the concept.
According to one Medco official, the “value” of such data in treatment decisions can be established in the minds of payors by conducting large outcomes trials, by relabeling a drug to recommend a particular diagnostic, or through credible association guidelines suggesting the use of the test should be incorporated into standard medical practice [see PGx Reporter 12-06-06].
Lesko also noted that randomized clinical trials may be necessary for reimbursement purposes, but conceded that such studies are not infallible.
The Warfarin Standard
The FDA has said it plans to update warfarin’s label to include genetic testing information. At the meeting, Lesko noted that labeling language for the anticoagulant is still being negotiated, but the agency expects to complete the relabeling in June.
Although warfarin has been widely used for half a century, physicians still dose the drug by trial and error, which helps account for the roughly 2,000 adverse events associated with the drug each day.
Relabeling warfarin with genetic information “has the potential to change the standard of care,” Lesko said. “Where is the bar for evidence? It’s a moving target. The bar for this drug, I believe will have been raised a bit, when we update the label.”
The FDA requires randomized controlled trials for companies submitting applications for new drugs for which the safety and efficacy is unknown. Lesko acknowledged that for widely used drugs that have been in medical practice as long as warfarin has, observational studies are sufficient.
Lesko expressed concern that the trend in the healthcare industry to press for companies to conduct randomized clinical trials to relabel drugs “empowers people that would like to see one size fits all dosing stay where it is.”
“Obviously, there are shortcomings to [observational studies],” Lesko acknowledged. “But if they are not biased, if they are well done, if there is a hypothesis, then they could be beneficial. Especially if there are several of them, and there is a meta-analysis. …”
He also said that prospectively designed randomized clinical trials used to validate dosing based on genetic factors can be prohibitively expensive, which in turn can hinder pharmacogenomic advances. 
Proponents of randomized clinical trials for relabeling hold up warfarin to note that there are no rigorously validated data showing that incorporating genetic information in treatment improves clinical outcomes.
Warfarin dosing can be problematic, since not enough drug can lead to unwanted clotting while too much can cause patients to hemorrhage. There are sparse guidelines regarding warfarin dosing, leaving physicians to titrate doses up or down until they reach a safe and beneficial response.

Lesko expressed concern that the trend in the healthcare industry to press for companies to conduct randomized clinical trials to relabel drugs “empowers people that would like to see one size fits all dosing stay where it is.”

The genetic basis for pharmacokinetic and pharmacodynamic variability can help clinicians dose warfarin more precisely. For gene-based warfarin dosing studies, monitoring relies on patient’s prothrombin time – a blood test that measures how long it takes blood to clot – expressed as an international normalized ratio, or INR.
“I emphasize clinical outcomes because this has been a conundrum for us,” Lesko told the meeting. “Clinical outcomes means what’s the percent of days you bleed. That’s the outcome practitioners are interested in. This will impact INR. But some people feel that INR is not good enough, and you need a randomized clinical trial, even though INR is an accepted surrogate for monitoring patients.”
Lesko said that the FDA has “taken the position that INR is a valid surrogate on which to base our decision” for relabeling warfarin.
Additionally, Lesko noted that the test doesn’t necessarily need to be cleared by the FDA to be mentioned in a drug’s label.
In the case of warfarin, there is one FDA-approved test, Roche’s Amplichip, and several homebrew tests, including recently launched assays from Genelex and Clinical Data, as well as Kimball Genetics’ warfarin sensitivity DNA test for research/investigational purposes [see PGx Reporter 10-18-06].
“The point of the matter is there is no regulation that requires the test to be approved by FDA before it goes into the label,” Lesko said. “It’s desirable, but there is no regulation that says you can’t update a label with a test that’s not approved.”
Randomized Trials Are Fallible
Lesko suggested that clinicians, the FDA, and industry work together to formulate a “hierarchy of studies” to outline the types of tests companies should conduct to justify relabeling a drug with genetic data, depending on the clinical questions being investigated.
While under certain circumstances randomized controlled trials may be appropriate, the agency and associations should be cautious not to set the bar unnecessarily high for relabeling purposes.
“I’ve been advocating that sometimes we will need [randomized] trials, but it depends on the question we have to ask,” Lesko said. But “if all we’re talking about is drug dose, then it’s hard for me to understand why we’re talking about randomized prospective controlled trials for a drug in which I know CYP2C9 and VKORC1 are controlling outcomes, and I can measure INR.”
“On the other hand if we’re talking about reimbursement, maybe these are studies that we need. If we’re talking about selecting the treatment for a patient with serious disease we may need a randomized controlled trial,” he said.
Besides, “randomized controlled trials are not infallible,” Lesko pointed out, noting that 53 percent of Phase III studies fail. 
Moreover, the widespread use of prostate-specific antigen testing and mammograms suggest that physicians don’t necessarily look for randomized controlled trials to justify incorporating a diagnostic into their treatment decisions. Widespread physician adoption of these tests aren’t fueled by outcomes data, Lesko said, noting that the literature on PSAs is largely inconclusive about whether screening reduces mortality.
PSA and mammogram adoption by physicians is a sign that they will use tests that aren’t necessarily fueled by outcomes but that fill a gap in knowledge, Lesko said.

Filed under

The Scan

Possibly as Transmissible

Officials in the UK say the B.1.617.2 variant of SARS-CoV-2 may be as transmitted as easily as the B.1.1.7 variant that was identified in the UK, New Scientist reports.

Gene Therapy for SCID 'Encouraging'

The Associated Press reports that a gene therapy appears to be effective in treating severe combined immunodeficiency syndrome.

To Watch the Variants

Scientists told US lawmakers that SARS-CoV-2 variants need to be better monitored, the New York Times reports.

Nature Papers Present Nautilus Genome, Tool to Analyze Single-Cell Data, More

In Nature this week: nautilus genome gives peek into its evolution, computational tool to analyze single-cell ATAC-seq data, and more.