By Turna Ray
The US Food and Drug Administration's regulatory plans for personalized medicine are advancing rapidly in the wake of the recent approval of two genomically targeted cancer treatments, an FDA official said at a recent industry conference.
While the agency's evolving regulatory strategy for personalized medicine products has fomented frustration over the past several years among industry groups who have argued that the agency is out of step with advancing technologies, FDA's Elizabeth Mansfield sought to assure attendees of last week's Personalized Medicine World Conference that the agency is on a fast learning curve.
Following the approval last year of the pharmacogenomic therapies Zelboraf and Xalkori, FDA is learning to coordinate internal processes to facilitate simultaneous review of drug/test combination products and is refining the types of clinical trial data it wants to see from sponsors, Mansfield said at the conference.
Furthermore, she intimated that advancing regulation supporting the development of personalized treatments is a high priority at the agency, noting that personalized medicine is part of FDA Commissioner Margaret Hamburg's performance plan for the agency in 2012.
"You don't put that in your performance plan if you're not planning on making it a reality," said Mansfield, who is in charge of personalized medicine efforts at the Office of In Vitro Diagnostic Device Evaluation and Safety in FDA's Center for Devices and Radiological Health.
In the past year, the FDA has made some headway toward crafting a more consistent regulatory strategy when it comes to personalized medicine products. For example, the agency released two draft guidelines, one on the development of companion diagnostics (PGx Reporter 7/13/2011) and another on the development and marketing of research-use-only and investigational-use-only in vitro diagnostics (PGx Reporter 6/8/2011).
In addition, as part of an ongoing effort to keep the agency in step with science and technology advances, Hamburg made several management changes. One of her key appointments was Steven Spielberg, former dean of Dartmouth Medical School and director of the Center for Personalized Medicine and Therapeutic Innovation at Children’s Mercy Hospital in Kansas City. Spielberg is now serving in the newly created position of Deputy Commissioner for Medical Products and Tobacco (PGx Reporter 7/20/2011), and in that role he'll be spearheading inter-agency coordination.
Despite forward momentum in these areas, some industry stakeholders have recently expressed dissatisfaction about the companion diagnostic and RUO/IUO draft guidances (PGx Reporter 1/4/2012). Particularly in areas where the FDA has yet to issue guidance, such as for laboratory-developed tests, lab industry groups are working against the agency's plans through legislative channels. For example, lab proponents support a bill introduced late last year by Texas Congressman Michael Burgess that would ensure that LDTs remain under the oversight of the Centers for Medicare & Medicaid Services (PGx Reporter 10/19/2011).
Mansfield didn't shy away from acknowledging that that the FDA has much left to accomplish in the personalized medicine arena and will need to work with stakeholders to craft regulation that satisfies the agency's charge to ensure the safety and efficacy of marketed medical products while promoting innovation within industry. "An increasing number of stakeholder groups, such as the Personalized Medicine Coalition, the Pharmaceutical Research and Manufacturers of America, the Biotechnology Industry Organization, and AdvaMed are all driving us toward getting something done that's really going to help them," she said.
"The progress is pretty rapid, and it's been pushed by [our] learning by doing," Mansfield continued. "Everyone is playing pretty well together, and we're learning from each other a lot. We have a sense that the system is working and will work."
One of the main boasting points for FDA last year was that the agency approved two personalized oncology drugs — Pfizer's Xalkori and Roche/Plexxikon's Zelboraf — with their companion tests well head of the review time frames designated under the Prescription Drug User Fee Act.
"We learned [through these experiences] that codevelopment works," Mansfield said. "If people are dedicated to this and they plan it out right, it works, and we didn't really have a lot of major issues."
Other industry players interested in advancing molecularly targeted treatments have no doubt closely tracked the regulatory processes for Xalkori and Zelboraf. The volume of companion diagnostics development that the CDRH is encountering is "growing at an alarming rate" and occurring mostly in the oncology space, according to Mansfield.
At the same meeting, Larry Lesko, former director of the Office of Clinical Pharmacology at FDA's Center for Drug Evaluation and Research, predicted that "we're going to see an increasing number of submissions of personalized medicine drugs."
Learning from Success Stories
Although the agency has yet to finalize its companion diagnostic guidance and hasn't yet issued a drug/diagnostic codevelopment guidance, Xalkori and Zelboraf's development and regulatory histories contain lessons for other sponsors developing personalized medicine products.
For both of these personalized medicine success stories, the genomic markers of interest were well associated in published studies with disease pathology before drug and diagnostic development progressed. As such, when it came time to garner regulatory approval for Xalkori and Zelboraf, two drugs that were intended for patient subpopulations with limited treatment options, the FDA worked with the sponsors to figure out a way to quickly bring these treatments to market.
Both drug and test combination products were approved based on data from two single-arm studies and sponsors weren't required to enroll biomarker-negative patients. The agency made these concessions in these cases, despite the FDA's historical preference for at least one well-controlled, randomized study comparing an investigational oncologic to the standard of care.
"These approvals represent a real proof of concept, if you will, in the regulatory decisions to move a concept of a targeted therapy to practice via the approval process," reflected Lesko, who currently leads a new pharmacometrics and systems pharmacology initiative at the University of Florida's interdisciplinary Institute of Therapeutic Innovation.
"Years ago when we had meetings like this, I remember we used to say, 'Well how many biomarker-negative people have to be in a clinical trial for regulatory approval?' Because of the strong biological hypothesis, neither of these programs, as far as I know, had a biomarker-negative population in them," Lesko said. "So, does this set a precedent? Possibly."
Although the FDA has said that it prefers sponsors to test their drugs on marker-positive and -negative patients in order to ensure the ability of the marker to separate responders from non-responders, the Xalkori and Zelboraf examples illustrate that the agency will make exceptions to this requirement in certain cases.
"There has been a lot of discussion in the past [with sponsors] about what's going to happen if we don't have marker-negatives [in the study]," Mansfield said. "We have now witnessed, with two different drugs, that what happens is the [companion] diagnostic gets its selection claim, by selecting people who were in the pivotal trials that established that the drug was safe and effective."
Such a study design doesn't confirm that the marker can predict who will benefit and who won't benefit from the drug, Mansfield conceded, though she added that because of the strength of the data for these two drugs the agency was pretty confident that the markers of interest would pick out responders from non-responders. "Marker-positive trials are okay," she said, warning that with this design, if the drug company picks the wrong marker then "you may be taking a risk that … your drug isn't going to be considered safe and effective."
Learning to Align Submissions
One of the major lessons that emerged from Xalkori's approval process was the advantage of the modular premarket approval submission for companion diagnostics. Because of the divergent development time frames for drugs and diagnostics, an oft-cited challenge for sponsors commercializing personalized medicine products is aligning the validation of the test with the pivotal trial for the drug.
[ pagebreak ]
To account for this challenge, the FDA has been allowing developers of companion tests to submit parts of the submission package for the diagnostic while the development program for the drug is still ongoing. Noting that most companion diagnostics will require premarket approval, Mansfield said that this modular submission mechanism is the agency's "preferred process" when it comes to reviewing submission packages for assays used to discern which patients should receive a drug.
"What it does is it allows you to submit different modules of the PMA over time," she explained. "The last one submitted is the clinical module that establishes the safety and effectiveness of the test, so that you get all the mess out of the way early. Then, when the [new drug application] comes in, we're ready to look at the clinical module [for the test]."
Pfizer and Abbott utilized this rolling submission process to harmonize the development and review of Xalkori and ALK mutation test. Pfizer had initially used laboratory-developed tests to gauge ALK mutations in patients, but then decided to work with Abbott Molecular to develop a diagnostic kit that was submitted to the FDA. Normally, in such a case the agency would require a bridging study to ensure that the diagnostic kit is not inferior to the earlier LDTs, but Pfizer and Abbott could forgo this requirement since Pfizer had enough patients for a new Phase II study to facilitate approval for the kit.
To accomplish the simultaneous approval and market launch of Xalkori and Abbott's ALK mutation test, FDA's drug and diagnostics division had to work "very closely together" and hold "many, many meetings," Mansfield said. "The diagnostic typically has a shorter review time than the drug, but sometimes the drug review is accelerated … so we had to come up with a process to allow that to happen."
Because Xalkori was available immediately, but the companion test kit had to be placed in various labs, the FDA allowed Abbott's test to be offered in labs that participated in the clinical trial because those labs already had the equipment and had validated the test on site. "So, we allowed that under an extension of the investigational device exemption," Mansfield said.
Learning to Work Together
Through the Xalkori and Zelboraf approvals, the agency's drug and device divisions also had to learn how to work together to coordinate reviews of the drug and the test. "Our inter-center communications developed probably orders of magnitude reviewing these two over five months or so," Mansfield said. "We learned how to attend meetings with each other, and ask the right questions. We learned whose questions were whose, and to submit them in a timely manner. We had to coordinate our approvals and our press [announcements]."
The agency is planning to take the lessons it learned from these two submissions and develop internal standard operating procedures that can guide future Rx/Dx combination product reviews.
Sponsors have often cited the lack of coordination between the FDA's device and drug centers as a major barrier to getting Rx/Dx codeveloped products approved. In a recent interview with PGx Reporter, industry consultant Eric Lawson discussed the need for a formalized way to ensure that a combination, personalized medicine product is reviewed by all four parties – CDER, CDRH, the drug maker, and the device firm (PGx Reporter 1/4/2012).
"There is a disconnect between how meetings between sponsors and the agency are handled on the therapeutic side, with CDER and [the Center for Biologics Evaluation and Research] with the drug sponsor, versus the way meetings are handled with the diagnostics sponsor at the CDRH's OIVD," Lawson said. "There doesn't seem to be a way to link them up together."
"Certainly, the different centers … are operating under different laws and regulations, and have very, very different cultures historically," Mansfield said during her presentation at the PMWC. "We weren't co-located ever, so we couldn't recognize other people who worked at FDA. Now, we're almost all located in the same campus and we talk to each other in the hallway."
Closer proximity has helped spur more inter-center communication and collaboration. When the drug center holds a meeting, officials from the diagnostics center are invited, and vice versa. This way, "the drug people understand diagnostics more, and the diagnostics people understand drugs more," Mansfield said.
One of the major challenges in the past was figuring out how to detect early in the development of a drug that a companion test was necessary and CDRH expertise was required. Now, for investigational new drug applications, if a test is added to a study, it triggers that CDRH involvement is necessary, and the center may require more information from the sponsor. "This had been totally missed over the past couple of years," Mansfield acknowledged.
Drug and device center officials also decide together salient regulatory issues for treatment/test combination products that need to be addressed and are working together to come up with policy to address problems. The topics discussed in these internal meetings will likely be published as a codevelopment guidance, slated for release sometime this year or next. Center officials are also close to finishing an internal standard operating procedure document that outlines the work each center is responsible for when the agency changes a drug label to include a diagnostic.
"We certainly have regular interactions [between the drug and device center] on a much broader scale than we ever did before," Mansfield noted.
Despite being located on the same campus, the drug and device centers still use separate information technology systems to process submissions and communicate with staff and sponsors. This IT barrier also makes it difficult to streamline review of a drug/test combination product requiring the involvement of multiple centers.
"We all have different IT systems. We all have different ways of tracking submissions. We all have different ways of archiving submissions. And in the past we couldn't see across these center boundaries," Mansfield said. "We're working now to resolve that so we can communicate with each other's systems."
Future Developments
During his talk at the conference, Lesko suggested that the changes afoot at the FDA mean that the agency's "regulatory cloud of certainty" is now being lifted.
Lesko highlighted four FDA guidances issued last year – on clinical pharmacogenomics, on making benefit/risk determinations for device premarket reviews, on in vitro companion diagnostics devices, and on pivotal clinical trial design considerations for medical devices – as assurance that the agency is working through the regulatory and scientific challenges of developing personalized treatments.
Furthermore, CDER plans later this year to release a draft guidance on enrichment of clinical trials. This "will be very fundamental for the generation of evidence for the pivotal clinical trials prior to registration," Lesko said. "It will provide options …. [to sponsors] about the design appropriate for clinical trials for [the] drug and diagnostic."
Additionally, the FDA and drugmakers recently agreed to the PDUFA V goals, which will guide regulatory interactions between the agency and industry players between 2013 and 2017. Among the PDUFA goals are measures to increase FDA's expertise and staff to facilitate greater use of biomarkers and pharmacogenomics in regulatory submissions
"It's a commitment supported by user fees," Lesko said. "It's going to mean more people for FDA in the area of personalized medicine and genomics."
Finally, CDRH has launched an Innovation Pathway, an effort through which the agency aims to shorten the development and review time frame for innovative medical devices. Under this effort, regulatory experts are developing tools and procedures that will hopefully make regulatory processes more predictable and streamlined for advanced technologies.
The team has so far developed a decision tool to help reviewers and sponsors quickly determine the classification and regulatory direction for a particular product; a tool to help determine requirements for first-in-human clinical trials; and a forum for FDA, industry, and other stakeholders to discuss the intended use of an innovative device and the unmet medical need addressed by the technology.
The Innovation Pathway "hasn't explicitly been linked to personalized medicine, but if you read the contents it could be linked to personalized medicine as much as other devices," according to Lesko.
Given these ongoing efforts, "the next five years is going to be interesting," Lesko reflected. "The challenges that are out there are being addressed in one way or another. I think we're moving in the right direction."
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at tray [at] genomeweb [.] com.