Skip to main content
Premium Trial:

Request an Annual Quote

Roche Says FDA-Approved BRAF Test Is Better Than Sanger-based LDTs, but is it Enough to Sway Labs?


By Turna Ray

Roche recently presented data showing that its PCR-based Cobas BRAF Mutation Test is more sensitive than Sanger sequencing in selecting which patients will benefit from its melanoma drug Zelboraf — a move aimed at convincing labs to replace their standard laboratory-based testing methods with the US Food and Drug Administration-approved companion diagnostic.

The new data comes just over a month after the FDA approved Roche/Plexxikon's Zelboraf (vemurafenib) and the companion BRAF diagnostic after clinical trials showed that the drug significantly extended survival in metastatic melanoma patients with BRAF V600E mutations in their tumors (PGx Reporter 8/17/2011).

The FDA-approved label for Zelboraf notes that the drug is indicated "for the treatment of patients with unresectable or metastatic melanoma with BRAFV600E mutation as detected by an FDA-approved test." Roche's Cobas test is not mentioned by name in the drug label, but it is currently the only test that the agency has approved for use as a companion diagnostic for Zelboraf.

At the European Multidisciplinary Cancer Congress this week, Roche presented data showing that its PCR-based BRAF test compares favorably to Sanger sequencing, which is currently the gold standard for cancer mutation testing in many CLIA labs.

In the study, led by Steve Anderson of LabCorp subsidiary Esoterix Clinical Trial Services, researchers analyzed patient samples with the Cobas 4800 BRAF V600 Mutation Test and 2X bi-directional Sanger sequencing. Out of 433 evaluable samples, there was 96.4 percent agreement between the two tests for cases that were mutation positive and 80 percent agreement for those that were mutation negative.

The Cobas test "had a lower failure rate than Sanger; was more sensitive in the detection of V600E mutations than Sanger, and detected a majority of V600K mutations in the cohort," Anderson et al. concluded in the study, which was funded by Roche.

Discordant results were evaluated by Roche 454 sequencing, which found that "of 42 samples that were mutation-positive by the Cobas test and negative for the V600E mutation by Sanger, 17 samples were wild-type (15) or non-V600E (2) by Sanger but V600E-positive by 454 sequencing," Anderson et al. reported in the abstract presented at EMCC. Meanwhile, 24 of those samples were V600K-positive by both Sanger and 454 sequencing, and Sanger sequencing found one sample harbored a rare GTG-to-GAC mutation.

Although Zelboraf is indicated only for patients with melanoma tumors that harbor BRAF V600E mutations, the Cobas test's ability to detect V600K mutations is noteworthy because the same test was used to measure mutation status in patients enrolled in the clinical trial that led to the drug's approval. As a result, around 10 patients with V600K mutations were included in the vemurafenib arm and these patients were shown to have some response to the drug.

"Though the Cobas BRAF Mutation Test detects V600E mutations with a high degree of sensitivity and specificity, due to the proximity of the nucleotide sequences at codon 600, like other PCR-based tests, Cobas has cross-reactivity to V600K above a certain mutation threshold," Brian Earp, international business leader in genomics and oncology at Roche Molecular Systems, told PGx Reporter.

Earp noted that only patients that were positive by the Cobas test — regardless of whether they were E or K — were enrolled in the clinical trial. As a result, "only patients positive by Cobas have been studied and shown to benefit from vemurafenib."

For the eight samples that were found to be mutation negative by Cobas but V600E-positive by Sanger, 454 analysis revealed that two cases were wild-type, two were V600K, one was V600E2, and three were V600E.

"Cobas did not misclassify any wild type tumors as V600E mutant" or give a false positive result, Earp noted. "Sanger misclassified two wild-type specimens as V600E mutants … and these patients would have inappropriately been put on vemurafenib."

Patient samples for the study were drawn from enrollees in the BRIM2 and BRIM3 studies that the FDA reviewed in approving Zelboraf. It is estimated that Roche invested more than $700 million to develop Zelboraf with the help of a companion test.

According to Earp, based on the results of this study, more than 13 percent of patients who would benefit from Zelboraf treatment would have been denied the drug if they were tested with Sanger sequencing. In comparison, the Cobas test would have denied treatment to less than 1 percent of patients who would have responded to the drug.

An Uphill Battle?

While the simultaneous approval of the drug and diagnostic appears at first glance to be a victory for Roche's Rx/Dx development strategy, it underscores the murky regulatory environment for companion diagnostics: Although Roche's test is FDA approved, it's uncertain how many labs will adopt it since they can still offer their own LDTs with the ability to gauge BRAF mutations on a variety of technology platforms.

With the results of this study, Roche is hoping to give laboratories a compelling reason — tied to patient outcomes — for using its test.

While Sanger sequencing is one of the most readily used technologies for performing DNA assessments in tumor tissue, "the high rate of invalid results and the number of false-negative results with Sanger sequencing are a concern with this application because the errors can impact treatment decisions," Roche's Earp said.

He added that the Cobas BRAF test is seeing rapid adoption due to its improved sensitivity and rapid turnaround time compared to Sanger sequencing. "Our sales organization has a significant number of requests from recognized laboratories to offer our FDA-approved Cobas BRAF test," Earp said.

A dedicated sales team at Roche is conducting educational programs and seminars to gain more customers for the Cobas BRAF test and Genentech and Roche are disseminating information about the test to doctors. Additionally, the Zelboraf website includes a direct link to labs that offer the test. So far, only three labs — Carolinas Medical Center, Clarient Diagnostic Services, and Laboratory Corporation of America — offer the Cobas test.

[ pagebreak ]

The availability of an FDA-approved test, however, may not be enough reason for many labs to switch from a Sanger-based LDT that they've performed for some time, especially when it would require significant investment to evaluate the strengths and weaknesses of a new testing platform compared to standard methods.

For example, a pathologist at a major US hospital told PGx Reporter that the hospital's molecular diagnostics lab has been performing BRAF mutation analysis as a lab-developed test for several years in a variety of disease settings — including colon cancer, thyroid cancer, melanomas, and certain gliomas — and has no immediate plans to change its practices.

"Although there's a companion diagnostic [for Zelboraf] approved by the FDA, we didn't change our method," said the pathologist, who asked to remain anonymous because he was not approved to speak on behalf of the lab. "We may look at it in the future, but we've continued to perform the test that we've done all along, which has worked well for us."

The hospital's molecular diagnostic lab uses pyrosequencing to test for BRAF and KRAS mutations. Having evaluated Sanger sequencing, the lab found pyrosequencing to be quicker, less expensive, and better suited to gauging mutations clustered in a single codon.

According to the pathologist, there may be a number of reasons why Sanger sequencing may have been less sensitive in Roche's study.

"It's not at all surprising to me that a Sanger-based sequencing method may appear to be less desirable than the other method," he said. "Sanger sequencing has difficulty picking up mutations in highly heterogeneous tumors. So, the samples either need to be carefully manually dissected in order to enrich for the tumor population, which is typically what we do, or you can get a false-negative result."

While the risk of a false result is a "little more uncommon" in melanoma, "it would depend on the samples themselves," the pathologist explained. Depending on the ratio of malignant to benign cells within the tumor, "samples that have a high percentage of benign cells can give a false-negative result with any molecular test, and with Sanger sequencing in particular."

In order for a lab to bring in a new technology, it would have to be cheaper, quicker, or "substantively" better in terms of its impact on patient outcomes than existing methods. "We would also look into adopting new technologies if we had to, [for example] if there was a legal reason to look into it," the pathologist noted. "But we're not really looking to change our BRAF assay at this point; we've been happy with the way [the pyrosequencing-based LDT] has performed."

Dual Regulatory Pathway

Currently, molecular diagnsotics labs do not have a legal reason to switch testing platforms. While Zelboraf's label calls for an FDA-approved BRAF diagnostic, the FDA's stance on LDTs is in flux, so labs are unlikely to invest in new platforms unless they have to.

Although the FDA is working on a regulatory framework for bringing all laboratory-developed tests under its purview, for the time being, the Centers for Medicare & Medicare Services are still overseeing LDTs and the FDA is still responsible for establishing the safety and efficacy of diagnostic kits (PGx Reporter 9/7/2011).

Given this dual regulatory pathway, labs that are already performing BRAF mutation testing as LDTs have little incentive to switch to an FDA-approved test — especially when the FDA-approved test applies to only a single drug.

In its recently released draft companion diagnostics guidance, the FDA states that if a diagnostic is required to appropriately administer a drug — as is the case with Zelboraf — then that test must be submitted to the agency for regulatory review. This creates an uncertain situation for both laboratories and diagnostic firms.

Although labs cannot specifically market BRAF mutation LDTs as companion diagnostics for Zelboraf, they won't necessarily be at odds with current regulatory policy if they perform BRAF testing in this setting. On the other hand, test makers who spend the time and money to validate a companion test with a drug developer and receive FDA approval don't have a market advantage in return for their added investment.

An FDA spokesperson previously told PGx Reporter that this regulatory situation is unlikely to change until the agency clarifies its regulations around LDTs. Using the example of KRAS testing to pick out best responders to two colorectal drugs — Amgen's Vectibix and Merck's Erbitux — the FDA acknowledged that KRAS tests will likely continue to be performed as LDTs even after FDA-approved kits come on the market (PGx Reporter 7/13/2011).

Reputation and Return

During cost-conscious times, FDA approval is not high on many labs' lists when it comes to deciding which test platforms to adopt.

What matters, according to Charis Eng, chair and director of the Genomic Medicine Institute at Cleveland Clinic, is the reputation of the lab that performs the test. Since Eng directs a research lab, the genomic discoveries at her lab end up being adopted first by CLIA-certified labs providing patient care for hospitals and physicians.

"So it depends who makes the LDT," said Eng, who doesn't perform BRAF testing in her lab, but is familiar with the regulatory challenges molecular diagnostics labs are currently facing. Since lab professionals are acutely aware of the specificity and sensitivity of various LDTs out there, "the FDA specs are less important," she said.

"Times are tight … and CLIA labs are very mindful of cost-effectiveness and high analytic validity, and that's where the lab's reputation counts," Eng reflected. Although she couldn't provide a general estimate of how much it costs a lab to switch testing methods from an established platform to a new method, Eng suggested that the investment is significant enough that labs don't make the decision to change platforms lightly.

"For labs to validate and cross-validate a [new] test and make sure it's all CLIA certified, it's quite a bit of money," Eng said. "If they don't think they're going to get a reasonable return … then they'll just let someone else do it."

Although the FDA has said that its LDT regulations will be implemented in a manner that it is least burdensome for labs and will not disrupt patients' access to tests, the lab community fears that any requirement to perform only FDA-approved companion tests would be a financial burden they can't bear.

If FDA adopts such a strict requirement in its evolving LDT regulations, "I suspect it will slow down innovation and translation of discoveries … because FDA [regulation] is slow," Eng said. "Many geneticists and people who practice personalized medicine are very afraid of this."

Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at tray [at] genomeweb [.] com.