NEW YORK (GenomeWeb) – The importance of diagnostics in supporting precision medicine became more apparent in 2016, as pharmaceutical companies increasingly incorporated biomarker strategies into their therapeutic development programs and as the US Food and Drug Administration demonstrated flexibility and speed in its approvals of codeveloped products (see chart below).
Following the US elections, the FDA also announced that it would hold off on finalizing regulatory plans for laboratory-developed tests, which means that for the foreseeable future, FDA-approved companion diagnostic kits will co-exist with LDTs that operate in a CLIA environment. The dual regulatory pathway for diagnostics will continue, which many test manufacturers investing to take their kits through the FDA have complained creates an uneven playing field. However, as the precision medicine field matures, the dual pathway may not be such a bad thing, according to some industry observers, since labs may end up being better partners to pharma in the precision medicine space.
Meanwhile, pharmas continued to invest in diagnostics to inform drug development. A survey of the activities of 23 top pharma companies by consulting firm Diaceutics suggests 2016 was a record year for Rx/Dx partnerships, with 60 agreements inked already in the first nine months, a 60 percent increase from the year-ago period. Based on data from the first half of last year, Diaceutics also projected that 300 drugs that are in some way enabled by diagnostics were on track to see revenue increases of between 5 percent to 10 percent in 2016 compared to 2015.
Biomarker programs are also expanding in scope. Cancer Genetics, for example, a molecular diagnostics shop that also operates a reference lab, is looking at three times the number of biomarkers that it was two years ago in its pharma collaborations, according to CEO Panna Sharma.
"It's not uncommon today for a [pharma] company to say, 'We want to look at PD-L1 and PD-L2, but we also want to look at the gene expression patterns in RNA and we want to look at the upstream DNA, and try to find a durable signature that we can use,'" Sharma said, noting that pharma seems committed to using a range of diagnostics to better understand the drugs under development. "What we've seen, in 2016 especially, is the rise of the diagnostic or biomarker liaison," he said.
Although the FDA requires premarket review of companion tests and consider them necessary for the safe and effective use of a drug, over the past year, the agency has also shown less rigidity on this front by approving more complementary diagnostics, which are tests that can guide therapeutic strategy but aren't required like companion tests. The agency even allowed a drugmaker to launch a therapy in a new molecularly-defined indication with the condition that it would develop an FDA-approved companion test in the post market setting.
In that instance, involving the approval of Pfizer's Xalkori in non-small cell lung cancer patients with ROS1 alterations, the FDA is "speaking to the policy of having an approved companion test, but they are showing flexibility in order to get drugs out there," said Mya Thomae, VP of regulatory, clinical, and medical affairs at Illumina.
While such actions signal a more nuanced understanding on FDA's part of how diagnostics fit into the treatment continuum, there is also reason to worry whether this flexibility, coupled with the agency's decision to back off from finalizing the LDT guidance, will allow labs to launch unproven tests, confuse physicians about which test to order, and harm patients.
Delaying the LDT guidance will allow "companion-like" testing with unapproved LDTs, Thomae said. While this might facilitate greater access to testing for patients, "the other side of that is that [testing] will continue to be variable," because there's no requirement that the LDT be compared to the approved CDx that was used in the drug trial.
Diverse tools
Although last year, the FDA approved several drugs with companion or complementary diagnostics, Diaceutics' survey of the personalized medicine space in 2016 revealed that these categories weren't capturing all the ways in which testing was impacting patient care and drug revenues. When the firm factored in so-called conduit diagnostics ― tests that channel the right drugs or appropriate dosing to patients regardless of the platform used ― and found that 42 percent of more than 700 treatments marketed by 23 drug firms were biomarker-enabled in some way. In 2015, 21 percent of 600 therapies were similarly Dx-enabled.
Last year, Diaceutics also noted increased testing in the context of treating diabetes, infectious diseases, and inflammatory conditions, though diagnostics are still most commonly used in relation to cancer drugs. One of the most diverse testing environments that emerged last year was in non-small cell lung cancer, where patients can be tested for mutations in EGFR, ALK, and ROS1, and if they have alterations in these genes, can receive targeted treatments. They can also receive PD-L1 expression testing to assess their likelihood of benefitting from new immunotherapies.
According to Sharma, one of the surprise trends of 2016 was that there are now four immunotherapies launched on the market with immunhistochemistry PD-L1 antibodies as complementary and companion diagnostic tests. "Two years ago, people probably wouldn't have predicted that immunohistochemistry would be responsible for such major usage of companion diagnostics," he said.
A lot of drug companies Cancer Genetics is working with are now looking at different IHC tools as possible CDx options. "You're going to see a rise of immuno-phenotyping and kind of proteomic-type testing," Sharma said, "which is in contrast to the genomic-driven mindset we've had in the last three to five years."
In one example of an IHC-like Rx/Dx deal, last year, Caris Life Sciences said it would use its ADAPT Biotargeting System to develop a companion test that could identify which pancreatic cancer patients might benefit from Threshold Pharmaceuticals' evofosfamide. The drug had previously failed to meet its primary endpoint in a late-stage trial, but a companion test might rescue the drug by identifying best responders. To develop the test, Caris is using fundamental components of IHC, but instead of using a single antibody, the company is applying an approach that will enable the measurement of "millions of different biological features simultaneously," Caris Life Sciences CEO David Spetzler said. "It's a systems biology way of assessing the difference between responders and non-responders."
The molecular diagnostics industry is also moving toward NGS testing and liquid biopsies, particularly in oncology, where tumor tissue samples can be challenging to procure and testing one analyte at a time isn't realistic. Toward this end, the FDA last year approved Roche's Cobas EGFR Mutation Test v2 for gauging mutations in blood, and approved the first NGS companion test, Foundation Medicine's FoundationFocus CDxBRCA, alongside Clovis Oncology's PARP inhibitor Rubraca (rucaparib).
In the future, some industry players are betting, patients will be directed to the right drugs based on so-called universal companion diagnostics. Foundation Medicine, Illumina, and Thermo Fisher Scientific all have deals with multiple drug firms to advance such tests, and in November Thermo filed a premarket approval application for its Oncomine Universal Dx test for non-small cell lung cancer that it developed in partnership with Novartis and Pfizer.
But Sharma doubts the efficiency of this model, given the pace of scientific advances. "To go down that FDA path is going to be challenging because it's going to take you a minimum of a year or maybe two years, and in that time period you will have new markers that will be relevant," he said.
Cancer Genetics is working with pharma on a dozen trials in B-cell cancer and in lung cancer, and while the companies may be considering a similar set of genes, they are investigating different codons and exons, and exploring a range of mutation levels. Some pharmas are studying monotherapies, others are looking at combination agents, and all these programs are seeking to treat patients in different stages of illness. "The indications are all so different that every time I hear the words 'universal' and 'diagnostic', I cringe a little bit." Sharma said.
Incentivizing CDx development
While pharma is increasingly using biomarkers and diagnostic tools in the development of drugs, Diaceutics' market research suggests they often don't have an organized diagnostics strategy to support drug commercialization. However, drugmakers are starting to see the financial impact of a disorganized diagnostics strategy in the commercial setting, particularly in crowded drug markets, like lung cancer, where first and second generation molecularly-targeted treatments and immunotherapies are clamoring for market share.
For example, PD-L1 is a biomarker being investigated in more than 400 trials, Keeling estimated. But according to Diaceutics' survey, the first anti-PD-L1 drugs launched with inadequate investments in testing infrastructure, resulting in lost revenues of $770 million for a single indication in the first 18 months. Diagnostics warrant better and earlier communication between regulatory and commercial teams, Keeling said, but this isn't happening.
Moreover, Rx/Dx deals are structured in a way where diagnostics firms make money out of the development contracts but not out of the diffusion of the test into the market. Keeling estimates that a test left to its own devices will achieve 4 percent market adoption a year, but pharma needs at least 50 percent CDx adoption in the first year for a successful drug launch.
"Someone has to intervene at the market stage," he said. "Diagnostic companies don't have enough skin in the game to do that."
Labs might step in, and they may be a more natural partner to pharma in the precision medicine space, in Keeling's view, since they serve a common customer: the physician. For example, this past year, Intermountain Healthcare spun out Navican, which will bring precision oncology genomic profiling services to community care providers, such as a molecular tumor board; provide assistance with procuring drugs they recommend; identify clinical trials that might be right for the patient; and help make the case to payors that they should reimburse for testing. Other labs, such as Foundation Medicine, Caris Life Sciences, and Cancer Genetics offer similar services to physicians.
Perhaps readying itself for closer ties with pharma, Foundation Medicine decided to pursue Parallel Review last year for its NGS-based cancer profiling test FoundationOne, aiming to achieve simultaneous FDA approval and a national coverage determination from the Centers for Medicare & Medicaid Services. "If you get reimbursement right it will drive utilization of testing," and then perhaps labs can come to the table with pharma as equal partners, Keeling said.
Historically, reimbursement has been a challenge for molecular diagnostics firms, but things are looking slightly more promising. At the end of 2016, Medicare contractor Palmetto GBA proposed local coverage of comprehensive genomic profiling tests for a number of cancer indications, which will benefit companies like Foundation Medicine and Caris.
Meanwhile, on the regulatory front, because of FDA's decision to delay the final LDT guidance, "labs will still be reticent to go through the premarket approval process, unless pharma business is a significant part of their revenues," Illumina's Thomae said.
Drugmakers might also push back against having to take a CDx through FDA approval if there are unapproved LDTs on the market. "If we're really back in a place where we're not sure that [lab testing] needs to be regulated by the FDA … it leaves [the agency] open to a potential challenge on their position on companion diagnostics," she observed.
Impact of no FDA regulation of LDTs
Although pharmaceutical companies have a clear responsibility from a regulatory standpoint to bring an FDA-approved companion test to market if the drug's safety and efficacy is dependent on a biomarker, there is no obligation for labs to adopt that test. Often, as soon as the FDA approves a companion test, the developer faces competition from labs with in house tests that gauge the same analyte.
For example, Myriad's BRACAnalysis CDx was the first LDT to garner FDA approval to identify best responders to AstraZeneca's Lynparza (olaparib) but that test has had to contend with competition from LDTs that analyze the same genes. This "creates disincentives for diagnostics companies to innovate when 'generic' competitors can immediately launch tests without having undergone substantial regulatory review that focuses on safety and efficacy," Myriad spokesperson Ron Rogers said over email, adding that Myriad will work with industry trade associations and policymakers to support regulatory proposals that favor innovation and broad access to tests.
Meanwhile, LDT developers have argued that unencumbered by FDA's review requirements, they can launch tests faster in a public health crisis, and in many cases, their tests can perform better than FDA-approved assays. "The original kits for HER2 were not perfect and it was only through LDTs that a migration to a quality algorithm was arrived at," Keeling said. "I think PD-L1 will be the same." Keeling, like many other industry observers, is betting that neither the FDA's LDT guidance nor a legislative alternative will emerge during the new administration. For many, this is a not necessarily a bad thing.
While drug companies have pursued development of personalized treatments with FDA-approved tests, in the past, most have not kicked up much fuss that LDTs that gauge EGFR mutations or HER2 expression, for example, are also being used to determine which patients should or should not get their drug. "I think pharma will now go into a period of reevaluation," Keeling said. "They will look at whether a singular [FDA-approved kit] approach will serve their long-term interests."
Novartis, for example, has supported standards-based BCR-ABL lab testing and Pfizer and AstraZeneca have worked with the UK government to establish next-generation sequencing in cancer. Similarly, he predicted, pharma may be more inclined now in the US to support improvements to laboratory testing infrastructure that support commercialization of their drugs. Companion diagnostic kits will be a part of the supply chain, but also LDTs.
The FDA seems cognizant of these realities on the ground, industry players said, and this was reflected in two draft guidances last year on the regulation of next-generation sequencing tests. In one guidance, the agency outlined how sponsors might demonstrate analytical validity of hereditary disease tests and provided a path to exempt such tests from premarket review through the de novo process. In another, the FDA laid out how publicly accessible genetic variant databases could be recognized by the agency and provide clinical validity support for genetic tests.
"FDA is being mindful of trying to stimulate innovation," Caris' Spetzler said. "They are trying to balance advocating for quality testing and safeguarding patients, while at the same time trying to create an environment that allows rapid advancement of the science."
For example, the agency approved Foundation Medicine's FoundationFocus CDxBRCA, an LDT, without approving the NGS platform on which the test runs. "In the old days, you'd have to have a master device [application] filed, and so the instrument manufacturer would play an integral role in any FDA approval," Spetzler observed, noting that the agency's handling of Foundation's CDx suggests a more "progressive approach."
Caris molecularly profiles cancer patients' tumors in its lab, which can help doctors prescribe personalized treatments and recommend clinical trials. Spetzler said his company intends to seek FDA approval for this profiling test, even in the absence of a final LDT guidance. "We're actually in favor of FDA regulation of the LDT environment," he said.
Existing lab quality standards under CLIA are not sufficient, in Spetzler's view. "There is a lot of disservice done to patients by having a loose regulatory environment," he said. "There is an important balance between innovation and protecting patients. We would like to see a little more drift in the direction of regulation."
Similarly, Myriad's Rogers said that without FDA regulation of LDTs, "other laboratories will continue to market tests with claims that have not been subjected to consistent regulatory scrutiny, a fact which is not readily apparent to providers or patients."
The trends discussed in this article will be further explored in a panel discussion at the Personalized Medicine World Conference on Jan. 25. GenomeWeb Editor-in-Chief Ed Winnick will moderate the session, entitled "Companion Diagnostics: Where is the Incentive?"