NEW YORK (GenomeWeb) – The US Food and Drug Administration's plans to allow class labeling for companion diagnostics used to personalize cancer drugs could help ease regulatory burdens for sponsors looking to expand labels for already approved tests and encourage advancement of next-generation sequencing panels.
The agency issued draft guidelines earlier this month outlining the circumstances in which class labeling may be appropriate for companion diagnostics used to personalize cancer therapies, and the evidence test makers would need to submit to achieve such labeling. The FDA is accepting public comments on the draft guidance until Feb. 5.
According to the draft, test makers seeking class labeling for a CDx would have to show that cancer patients with the same biomarker-defined disease indication — such as non-small cell lung cancer characterized by ALK fusions — can be treated with at least two drugs in the same class. Sponsors can establish this with data from published studies or new investigations.
"It's always been easier to get supplemental premarket approval [for a test] once you have a PMA," said Kevin Hawkins, director of quality and regulatory affairs at MolecularMD, which develops molecular diagnostics to inform cancer drug development. But generally, when sponsors wanted to add another indication to an existing PMA, they had to gather clinical evidence to support that additional claim. Language in the draft guidance suggests the FDA is going to allow sponsors to submit data from the published literature, which is a pretty new move for the agency, Hawkins observed.
If the test developer already has an FDA-approved or -cleared CDx in a particular indication for a specific drug, the FDA is also willing to allow the firm to leverage that data as part of its demonstration of analytical and clinical validity for class labeling. Labs that don't already market an FDA-reviewed CDx may need to do new studies to establish the performance characteristics of their tests. However, even if a sponsor doesn't have an approved CDx to work from, the agency is willing to accept data showing concordance with another approved test as evidence of clinical validity.
Class labeling may not be appropriate for all companion tests. In the draft guidance, the agency notes the importance of pursuing class labeling after the relationship between a drug class and a biomarker of interest is well understood. "Labeling for such a broader use is not as simple as just matching diagnostic targets with therapeutic targets," the agency warns.
The FDA points out a few considerations that test developers should keep in mind when pursuing class labeling for a CDx, such as ensuring that the technology underlying a test can detect all the mutations necessary to identify the patients eligible for the drugs in a class. Additionally, the agency cautions sponsors that companion diagnostics that detect the same marker of interest and have similar analytical performance, could still identify different groups of patients as eligible for treatment because of different cutoffs for a marker-positive and marker-negative result.
"The FDA's recent draft guidance on class labeling for companion diagnostics is a good step in the right direction to bring standardization across molecular testing methods in oncology," said Lynne McBride, director of regulatory and clinical affairs at Thermo Fisher Scientific. "However, it should be noted that it does not circumvent the important analytical and clinical validation studies that are necessary to demonstrate performance and concordance with other FDA-approved testing approaches."
Despite these caveats, test developers and industry observers are intrigued by the possibility of a streamlined pathway for companion diagnostics that address a class of drugs instead of just one drug, and what this could mean for future drug/diagnostic codevelopment. Industry observers noted that for newer biomarker/drug indications the pharma and test maker would still need to collaborate to gather clinical evidence. But for indications with more commercial experience, the guidance as written would allow test makers to pursue a class label for a CDx without the involvement of every drug developer selling a drug in a particular class.
Historically, drug labels approved with a CDx don't mention that test by brand name in labeling, but the labels of companion diagnostics do note the specific drug they are indicated for. "In a way this [CDx class label] would even the playing field," said MolecularMD CEO Dan Snyder.
Improving patient access
The latest draft guidance on class labeling expands on the first set of recommendations the FDA issued seven years ago on the development of companion diagnostics, which the agency defines as providing information that is essential for the safe and effective use of a drug, Historically, most of the FDA-approved or -cleared companion diagnostics detect variants in one gene and are indicated for directing treatment with a specific drug.
In recent years, the FDA has expanded the labels of several tests to indicate their use with multiple drugs, sometimes in the same class. However, drugmakers competing for market share in a drug class will often work with different labs to advance companion diagnostics for treatments that are indicated for the same molecularly defined subpopulation of patients. These Rx/Dx alliances don't necessarily reflect how doctors practice medicine and how patients receive treatment.
When drugs in the same class, targeting the same mutations, are attached to different companion tests it can limit the drugs a patient is eligible for. "[The] FDA is concerned that the current situation is not optimal for patient care because a clinician may need to order a different companion diagnostic (i.e., one that includes other therapeutic products on the label), obtain an additional biopsy(ies) from a patient, or both, to have additional therapy treatment options," the agency said.
To illustrate the scope of the problem, the FDA noted in the draft guidance that it has approved five drugs for non-small cell lung cancer patients with EGFR exon 19 deletions or exon 21 (L858R) substitution mutations, and four companion diagnostics that can gauge these mutations, but each is approved for use with different EGFR inhibitors. Foundation Medicine's FoundationOne CDx is approved for use with Gilotrif (afatinib), Iressa (gefitinib), and Tarceva (erlotinib); Thermo Fisher Scientific's Oncomine Dx Target Test is indicated for use alongside Iressa; Roche's Cobas EGFR Mutation Test V2 is for use with Iressa, Tarceva, and Tagrisso (osimertinib); and Qiagen's Therascreen EGFR RGQ PCR Kit is for use with Gilotrif, Iressa, and Vizimpro (dacomitinib).
"In [the EGFR testing] example, the oncology community would be better served by a companion diagnostic ... indicated for 'identifying patients with NSCLC whose tumors have EGFR exon 19 deletions or exon 21 (L858R) substitution mutations and are suitable for treatment with a tyrosine kinase inhibitor approved by FDA for that indication,'" the agency said in the guidance. "This could enable greater flexibility for clinicians in choosing the most appropriate therapeutic product based on a patient's biomarker status."
According to Jonathan Arnold, Qiagen's VP and head of partnering for precision diagnostics and oncology, the company is considering pursuing class labeling for its Therascreen EGFR companion test, and anticipates the regulatory process would be relatively straightforward since the test has been previously approved as a CDx for three EGFR inhibitors. Arnold added that NGS "lends itself to this type of drug class-specific labeling and this is something we're considering across our existing CDx portfolio."
NGS-based companion diagnostic panels, recently approved by the FDA, have the potential to bring more flexibility to physicians because they allow patients to be tested once for a panel of biomarkers associated with responses to multiple drugs. But even then, the CDx indications for NGS panels may not cover all the drugs in a class.
For example, doctors looking at the companion diagnostic indications on Foundation Medicine's FoundationOne CDx label will see it can be used to gauge HER2 overexpression and identify patients who may benefit from one of three HER2-targeted drugs, Herceptin (trastuzumab), Perjeta (pertuzumab), and Kadcyla (ado-trastuzumab emtansine), all marketed by Genentech. (Foundation, like Genentech, is a subsidiary of Roche.) But the test isn't approved for CDx use with other HER2-targeted drugs on the market, such as Puma Biotechnology' Nerlynx (neratinib).
CDx test labels can get even more confusing. Physicians considering the label for Thermo Fisher's Oncomine Dx Target Test will see it is FDA approved to assess ROS1 rearrangements in NSCLC patients to determine their chance of benefiting from Xalkori (crizotinib).
However, Pfizer's Xalkori is approved for NSCLC patients who have ROS1 rearrangements or ALK fusions. The information for physicians included with the Oncomine label states that the test cannot gauge ALK fusions, and highlights other tests that can. In contrast, the FDA-approved companion diagnostic indications on the FoundationOne CDx label feature its ability to detect ALK rearrangements for patients considering Xalkori, Alecensa (alectinib), or Zykadia (ceritinib), but not ROS1 rearrangements.
When to seek a class label
These examples illustrate the fact that drug/diagnostic codevelopment deals aren't solely inked on the strength of the association between a drug and a predictive biomarker. Beyond scientific considerations and patient need, competition and economics are factors that impact which drugmaker partners with which test developer.
While the agency's aim is to ease patient access to personalized cancer drugs via companion tests that can be used to direct treatment to a class of drugs instead of just one drug, the agency's guidance doesn't address the market forces that might still impact a test maker's ability to seek a broader label for a CDx.
However, this is something the agency wants stakeholder feedback on. In a federal register notice announcing the draft guidance, the FDA asked stakeholders to describe specific difficulties sponsor may face in developing the evidence needed to secure class labeling for a CDx, "for example, ... any challenges resulting from industry or business practices, including business agreements."
For example, when a biomarker doesn't neatly separate responders and non-responders to a drug, as is the case with PD-L1 expression and checkpoint inhibitors, the FDA sometimes approves a complementary diagnostic. "That's literally a diagnostic that failed to have enough correlation to be called a companion diagnostic," Hawkins said.
Drugmakers tend to like complementary diagnostics because they don't restrict the treatment-eligible population in the way required companion tests to. But for indications where some drugs in a class have been approved with a companion test and others have been approved with a complementary diagnostic, achieving a CDx with class labeling may be more challenging and may require new studies, according to industry observers.
Qiagen's Arnold pointed out that it may also be difficult to seek class labeling for a CDx when data to support a class-specific claim isn't available because this information in confidential or restricted by intellectual property. "Pharmaceutical companies often consider these mechanisms of action and the associated biomarkers as competitive differentiators," he noted.
This gives test developers even more reason to seek class labeling for a CDx in the post-market setting, after several drugs in a class have been commercialized and there is more experience with biomarker testing. "Then, when a new drug is being developed in the same drug class, the corresponding CDx with the class-specific labelling claims will already be approved," Arnold said.
Opening up the CDx market
While the specific language in the guidance has to be worked through, FDA's aim in releasing these recommendations is to spur greater competition and patient choice when it comes to companion diagnostics, observed Charles Mathews, principal at consulting firm ClearView Healthcare Partners. "What's happening with the FDA guidance is it's loosening up the potential criteria for who could get approval [by taking] more of a class approach as opposed to every single test needing to partner with individual [drug companies] to get approval for a specific indication," Mathews said. "The thinking is that that would create more opportunities in the market."
FDA guidance may also prod payors to adjust their coverage policies to be more market friendly. Mathews noted that payors have never been very keen to restrict coverage and payment to FDA-approved companion diagnostics only, recognizing that commercial lab-developed tests without the agency's blessing may be just as good as the FDA-approved version and have cost advantages.
"[Payors] like companion diagnostics … but they also like competition and they don't like to get in the business of picking winners and losers from an industry perspective," he said. "They like to say, 'Well they are equal options here and we'll let the marketplace decide and we're just going to cover testing across the board." This is reflected in the Centers for Medicare & Medicaid's national coverage determination for NGS tests to personalize cancer treatments, in which the government payor stipulated coverage criteria for FDA-approved or -cleared companion diagnostics, but also included a pathway for tests without the agency's approval to gain coverage.
Given this, Mathews believes payors would be receptive to FDA's efforts to open up the CDx market. "Part of the story arc, particularly around NGS, is payors know they're going to be covering this eventually, and not just FoundationOne," he said. "There may be 20 labs offering something that's very Foundation-like, and by that point there is probably going to be some commoditization of pricing and it's going to be easier for them to bear paying for this."
The FDA loosening the reins on who can pursue CDx claims will increase competition, Mathews observed, and "this could be better for payors, which which may mean more coverage" for these tests.