Friends of Cancer Research (FOCR) has issued a white paper calling on the U.S. Food and Drug Administration (FDA) to exercise greater regulatory flexibility in the validation of rare biomarker diagnostics and to streamline and expedite the review of companion diagnostics that target rare biomarkers or indications.
The white paper, which was developed without input from the FDA, includes recommendations for FDA and industry about how to safely speed the development of companion diagnostics targeted at rare biomarkers and diseases. Implementing the recommendations would not require additional statutory authority, according to the white paper authors.
“There are flexibilities that are already in place for drug developers, and these could be extended to companion diagnostic developers, as well, for rare indications. We think that the FDA could consider defining rare, in the case of a companion diagnostic, as 10,000 patients likely to have a particular diagnostic biomarker or indication,” Elizabeth Mansfield, vice president, regulatory, at Foundation Medicine and a member of the working group that authored the white paper, said during an FOCR webinar.
The FOCR working group begins its recommendations at the clinical trial enrollment stage, where multiple local assays are often used to enroll patients into the trial. The variability in the designs and methodologies of these tests can lead to discordance with the companion diagnostic test used in the trial, even when the clinical performance is on target. The group recommends that trial sponsors set minimum performance standards for accruing patients and require local labs to provide minimum performance data. That performance data would include information on accuracy, precision, and analytical sensitivity and specificity.
The white paper also outlines a framework to conserve patient samples that are positive for a rare biomarker. Currently, these samples are used to standardize test performance and support test validation but using positive samples in this way can be costly and time consuming, the FOCR working group asserted. Instead, the group called on FDA to issue guidance that would recommend the use of a combination of contrived samples, representative variant validation, variant class-based validation for certain variant types, and prior data from approved diagnostics to determine test performance. For example, to assess concordance, the working group recommended a minimum requirement of 30 biomarker negative samples and a range of up to 30 biomarker positive samples, with the option to use a contrived sample for the positive samples.
“This would relieve the need to use precious clinical samples with rare biomarkers or from rare indications to perform the analytical validation,” Mansfield said. “These are often difficult to find, and we believe it may not be an appropriate use of such precious clinical samples.”
The working group also proposed that the FDA allow substitution with similar tumor types in cases where clinical samples are difficult to obtain. Calling this a “DNA is DNA” approach, the FOCR working group explained that it would involve performing analytical validation on any sample with the biomarker of interest, regardless of the tissue origin, allowing for validation on non-small cell lung cancer samples, for instance, if small cell lung cancer samples were unavailable.
“We do not want to dilute the bar for safe companion diagnostics at all,” said Rasika Kalamegham, head of US regulatory policy at Genentech, and a member of the working group. “We do want to find a way to practically meet those bars and that’s what all of our recommendations speak to.”
The white paper also outlines the need for the FDA’s drug and diagnostic review divisions to align their processes for concurrent approval and calls on the agency to take a “risk-based approach” when determining which data elements are needed prior to approval and which could be shifted to postmarketing.
The working group also suggested that FDA create a risk-based pathway for a companion diagnostic for rare biomarkers, in situations where patients are willing to take on more risk in order to gain access to therapies, Mansfield said. “We propose that FDA could publish the risk-benefit assessment in the final summary of safety and effectiveness documentation for [the premarket approval application] PMA for a companion diagnostic,” she said. “It would explain that this precedent was only applicable in the setting of a rare biomarker or a rare indication and that it wouldn’t be generally applicable.”
Soma Ghosh, a biologist at the FDA’s Center for Devices and Radiological Health (CDRH), said FDA is aware of the roadblocks and challenges in companion diagnostics for rare biomarkers and indications and has already begun work in this area. FDA review teams are already using “flexible and least burdensome” approaches for diagnostics for rare biomarkers and they have began shifting some requirements into the postmarketing setting. The agency also finalized a guidance document to broaden the labeling of companion diagnostics, she said. (RELATED: Companion diagnostics: FDA finalizes guidance to broaden labeling for precision oncology, Regulatory Focus 13 April 2020)