Skip to main content

Ovarian Cancer Study Provides Support for Myriad's HRD Test as CDx for Tesaro's Niraparib

Premium

NEW YORK (GenomeWeb) – The first data from a large, prospective study has shown that ovarian cancer patients who have lost the ability to repair double-stranded DNA breaks by molecular mechanisms other than germline BRCA mutations derive greater benefit from Tesaro's PARP inhibitor niraparib compared to placebo.

The 500-patient Phase III NOVA study used Myriad Genetics' myChoice HRD to identify best responders to niraparib. The companion diagnostic predicts if cancer patients are likely to benefit from agents that target proteins involved in DNA repair pathways, such as PARP. The tissue-based myChoice HRD detects whether patients have somatic mutations in BRCA1 and BRCA2 genes and other "genomic scars" — through heterozygosity loss, telomeric allelic imbalance, and large-scale state transitions — that make them unable to correct DNA damage and more likely to respond to DNA-damaging drugs.

However, since patients in the overall study population also responded to niraparib, though to a lesser degree than the molecularly defined patient population, market analysts questioned whether a companion diagnostic was necessary — a question that will be answered when Tesaro reports more detailed study data later this year.

In NOVA, recurrent ovarian cancer patients were stratified into either a germline BRCA mutation cohort (if Myriad's BRACAnalysis CDx blood test detected a deleterious variant) or a non-germline BRCA mutation cohort. Patients in the latter cohort were tested on myChoice HRD and included patients who were HRD positive and negative, as well as those with somatic BRCA mutations.

Tesaro, which declined to comment for this article, reported last week that ovarian cancer patients with germline BRCA mutations and those with HRD-positive tumors lived significantly longer without their disease getting worse when they received niraparib compared to patients receiving placebo. Patients with germline BRCA mutations had a median progression-free survival of 21 months on niraparib compared to 5.5 months in the control group. Patients with HRD-positive tumors experienced median progression-free survival of nearly 13 months on niraparib compared to under 4 months on placebo.

Tesaro will submit a new drug application for niraparib with the US Food and Drug Administration in the fourth quarter. During a call to discuss the results, Myriad executives told analysts that it will align the premarket approval application for myChoice HRD with the NDA so the therapy and companion test enter the market together after approval in the first half of next year.

"All of the biomarker evaluation has been done in an investigational device-exempt laboratory environment with full discussions with the FDA," Jerry Lanchbury, Myriad's chief scientific officer, told GenomeWeb last week. During the analyst call, CEO Mark Capon noted that if the timelines remain on track, myChoice HRD could start generating revenues as a companion diagnostic for niraparib during Myriad's fiscal year 2018 (which begins in the second half of 2017).

Estimating there are 22,000 ovarian cancer patients in the US and 38,000 in Canada and across five EU countries, Myriad is projecting a $200 million global market for the HRD test. Beyond ovarian cancer, Myriad believes myChoice HRD has utility as a companion diagnostic in many other malignancies and is estimating an addressable market of more than 1.4 million patients valued at $6 billion globally.

NOVA represents the first prospective data on myChoice HRD and Myriad expects the results will de-risk the more than 22 companion diagnostic collaborations it has ongoing with drugmakers. Myriad entered the companion diagnostic market with BRACAnalysis CDx, which the FDA approved in 2014 alongside AstraZeneca's ovarian cancer PARP inhibitor Lynparza. However, according to Myriad, myChoice HRD identifies twice as many patients who would benefit from PARP inhibitors such as niraparib compared to germline testing with BRACAnalysis. 

BRACAnalysis CDx identifies 15 percent of ovarian cancer patients who have germline BRCA1/2 mutations, while a Tumor BRACAnalysis CDx Myriad has launched in Europe identifies 22 percent of patients who have inherited or acquired a BRCA mutation. Comparatively, half of the ovarian cancer patients tested by myChoice HRD would get a positive score, Myriad said. 

During the call last week to discuss NOVA results, however, analysts questioned whether myChoice HRD would be required at all, since patients in the overall, molecularly unstratified population still derived a degree of benefit from niraparib compared to placebo. In the non-germline BRCA mutation cohort, which included patients who had HRD positive and negative scores by Myriad's test, patients treated with niraparib experienced median progression-free survival of 9.3 months compared to 3.9 months on placebo.

Capone noted that this question will be answered once Tesaro reports more detailed data from NOVA at the European Society for Medical Oncology congress in October, specifically how patients with myChoice HRD-negative scores fared on niraparib versus placebo. "The overall group is heavily influenced by HRD positive, so the overall group isn't useful in answering that question," he explained.

The non-germline BRCA cohort randomized more than 300 patients to receive niraparib or placebo, and around 165 patients were HRD-positive. Tesaro, in discussions with the FDA, will have to consider whether HRD-negative patients derived statistically significant and clinically meaningful benefit from the drug compared to placebo.

To address the challenge of oncology biomarkers, such as PD-L1 expression, that can't be used to neatly carve out the intent to treat population for immunotherapies, for example, the FDA has advanced a new test category, dubbed the complementary diagnostic. After studies showed that some cancer patients with low levels of PD-L1 expression respond to immunotherapies, such as Opdivo (nivolumab), the FDA approved some of these drugs with complementary tests for identifying which patients will derive a greater magnitude of benefit from treatment. 

Based on the data so far, Myriad is discussing its test as a companion diagnostic for niraparib. "The only use for [myChoice HRD] is [as] an FDA-approved companion diagnostic," Capone told analysts. Myriad views this as an advantage over competitors because of its experience bringing to market the first lab-developed companion diagnostic for Lynparza with BRACAnalysis CDx. 

MyChoice HRD is a next-generation sequencing test, while BRACAnalysis employs PCR and Sanger sequencing. Despite the differences, Capone noted that Myriad will be able to leverage its learnings from the FDA's review of BRACAnalysis CDx, during which the agency reviewed the company's variant classification processes and quality systems.  

Historically, Myriad held a monopoly over the BRCA testing market, and although the firm is still leading the space, it has faced competition from other labs who have launched NGS hereditary cancer panels. Capone maintained during the call with analysts, however, that competitors will not be able to copy the myChoice HRD test. "It's proprietary and it's based on technologies that have intellectual property and trade secrets," he said.

The three molecular mechanisms that measure genomic scars indicative of homologous recombination deficiency are unique to myChoice HRD and involve algorithms that will be challenging for competitors to replicate, Myriad executives said. "We realized a long time ago that BRCA1 and BRCA2 were only two genes that contribute to deficiency in homologous recombination," Lanchbury said. "If we were able to pick up a signature for that inability to repair double-stranded DNA breaks, we had the opportunity to expand the patient group who would qualify for treatment with a PARP inhibitor very, very dramatically."

Myriad developed a way to define loss of heterozygosity, a mechanism involved in cancer development due to the loss of the normal copy of an allele, leaving the abnormal form. Myriad licensed from Boston Children's Hospital a way to measure allelic imbalance that extends to the telomeres, and from researchers at Institute Curie in France a way to measure large-scale state transitions, which define adjacent breaks in chromosomes that are 10Mb in size.

Although multiple groups were simultaneously working on characterizing the various mechanisms involved in DNA repair defects in cancer, Lanchbury noted that Myriad is the first to develop a test that combines these three aspects under the term homologous repair deficiency, or HRD. "We've shown that the combination of those individual components is actually more powerful than any of the individual components alone in [predicting the] potential for therapeutic response," Lanchbury said.

The scientific rationale has intrigued drugmakers developing PARP inhibitors, and Myriad has successfully inked a number of companion diagnostic collaborations involving myChoice HRD, although not all of the deals are public. However, AstraZeneca, which worked with Myriad to bring Lynparza to market with BRACAnalysis CDx, has decided to take a different approach. 

The drug developer announced in June it would be working with Foundation Medicine to develop a next-generation companion diagnostic for Lynparza that will detect alterations in a panel of genes involved in homologous recombination repair pathways. An AstraZeneca executive previously explained to GenomeWeb that the firm decided to go with Foundation's test because it detects the genetic alterations that cause DNA repair deficiency, whereas Myriad's test measures the genomic scars that result from that deficiency. 

Myriad, however, has maintained that measuring the DNA scar is the most robust indicator of a tumor cell with severe defects in the homologous repair pathway, rather than detecting the "thousands of potential causes" of the deficiency. "When you see a traffic jam on the highway, you don't need to know the cause to know you need to find a detour," Capone told analysts last week.

He estimated that test panels that detect mutations in multiple cancer-linked genes would identify around 25 percent of ovarian cancer patients who would respond to PARP inhibitors, whereas its HRD test would identify 50 percent of patients as responders. This is important for advanced ovarian cancer patients who lack other options and also benefits drugmakers by expanding the intent to treat population for their PARP inhibitors, Capone said. Myriad also has a tumor panel in research phase that analyzes more than 80 clinically actionable, cancer-associated genes.