NEW YORK – The latest approval of niraparib (GlaxoSmithKline's Zejula) for heavily pretreated ovarian cancer patients with homologous recombination deficiency (HRD) not only increases the number of patients eligible for the drug, but also adds another biomarker that oncologists can use to help them decide whether to prescribe this PARP inhibitor or another drug.
In the precision oncology era, ovarian cancer patients may receive a variety of genomic tests, for example, germline testing to assess their inherited cancer risk or a next-generation sequencing panel that analyzes tumor tissue for alterations in hundreds of genes to identify clinical trials or potentially beneficial drugs. Doctors will now have to weigh their patients' HRD status against this increasingly complex backdrop of biomarker information. According to at least one oncologist, it will be more difficult for him and other practitioners to assess where Myriad Genetics' myChoice CDx, the HRD test approved with niraparib, fits into their diagnostic armamentarium because it uses a proprietary algorithm.
Heavily pretreated, advanced ovarian cancer patients are not a well-studied group, and in the past, doctors have had to try out different treatment strategies with little data on which patients are likely to benefit from them. For a long time, the only indicator that a patient may benefit from chemotherapy was whether she was sensitive to prior platinum-based chemo. Then, in 2014, the US Food and Drug Administration approved the use of BRCA testing (Myriad Genetics' BRACAnalysis CDx) to identify advanced ovarian cancer patients who would likely respond to the PARP inhibitor olaparib (AstraZeneca's Lynparza) after receiving three or more lines of chemotherapy. With the tissue-agnostic approval of pembrolizumab (Merck's Keytruda) in 2017, immunotherapy became a late-stage option for ovarian cancer patients whose tumors are unresectable and had microsatellite instability.
And now doctors have homologous repair deficiency, or HRD, to help predict response to niraparib. The Phase II single-arm QUADRA trial, which the FDA relied on for the approval, used multiple biomarkers to enrich the study for patients who were HRD-positive –– defined as patients with a deleterious or suspected deleterious mutation in BRCA1 or BRCA2, or patients who had responded for more than six months to platinum-based chemotherapy before progressing and had a genomic instability score of 42 or more.
In approving the new indication for niraparib, the FDA also approved Myriad's myChoice CDx, which detects large rearrangements and mutations in BRCA1/2 and uses a proprietary algorithm to analyze three processes involved in DNA repair — loss of heterozygosity, telomeric allelic imbalance, and large-scale state transitions — and calculate a genomic instability score.
"Having HRD … does allow me to identify patients who are more likely to respond and guide them to this therapy before I try something else," said Kathleen Moore, an associate professor of gynecologic oncology at the Stephenson Cancer Center at the University of Oklahoma Health Sciences Center and lead investigator of the QUADRA study. "It's not a perfect biomarker but it's better than continued empiric selection of drugs … and hoping it works. Our patients deserve better than that and this gets us closer to that goal of individualized therapy."
Moore and colleagues enrolled approximately 460 patients, out of which 189 were HRD-positive and 230 had HRD-negative or unknown status. In the HRD-positive group, 63 patients with a BRCA1/2 mutation were further stratified according to if they were platinum-sensitive, resistant, or refractory. The 126 patients who were HRD-positive but without BRCA mutations were also evaluated based on if they were platinum-sensitive or not.
In the overall HRD-positive population, the objective response rate was 24 percent and the median duration of response of 8.3 months. In BRCA1/2 mutation positive subgroup, the objective response rate was 39 percent in those with platinum sensitive disease, 29 percent in platinum-resistant disease, and 19 percent in platinum-refractory disease. In patients who were HRD positive but without BRCA mutations, the objective response rate was 20 percent.
"The overall response rate and the duration of response are the real story here, but even in terms of overall survival, from inclusion in QUADRA, it's around 17 months," Moore pointed out. "If you had asked people before this study what they though the median survival would be for women with fifth-line ovarian cancer recurrence, they probably would have told you six or eight months."
Because heavily pretreated advanced ovarian cancer patients aren't well studied, Moore felt that there aren't standard benchmarks that would have allowed for a randomized trial. Still, she described the single-arm QUADRA trial as an important study that provides the field with "a modern view" of ovarian cancer patients who have undergone multiple lines of treatment.
In Moore's view, the latest niraparib approval is particularly important since ovarian cancer patients are living longer with their disease due to effective therapies in the recurrent setting.
"There are hundreds of thousands of women out there who have ovarian cancer but it hasn't recurred yet. … Or they have experienced recurrence and are on some line of therapy and they missed out getting a PARP inhibitor as a front-line option," she said. "They may not have been eligible for maintenance treatment if they aren't platinum sensitive and didn't have a BRCA mutation until now. This opens it up for them."
An estimated 20 percent of ovarian cancer patients have a germline or somatic BRCA1/2 mutation, making them potentially eligible for a PARP inhibitor. An additional 30 percent of patients who are negative for BRCA1/2 mutation will likely be homologous recombination deficient by Myriad's test.
However, as the QUADRA trial shows, the niraparib-eligible population in the late-line setting is small once platinum-response status and prior lines or therapy are factored in with biomarker information. In the study, 13 patients saw their tumors shrink on niraparib out of 47 patients who had received three or four prior lines of treatment, had HRD-positive tumors, were sensitive to the most recent platinum therapy, and hadn't previously received a PARP inhibitor.
"It would be a pretty small subgroup of people who would be eligible for the [treatment] indication," said David O'Malley, director of the division of gynecologic oncology at Ohio State University Comprehensive Cancer Center (OSUCCC) and the James Cancer Hospital. He felt that the rarity of the patient population may limit use of the test in ovarian cancer.
Though the numbers may be small, alternatively, the availability of a new test could also provide additional impetus to improve genetic testing rates in women who need it. While practice guidelines recommend all ovarian cancer patients receive BRCA1/2 testing, a study published in the Journal of Clinical Oncology this year using registries in California and Georgia found low testing rates in this setting, suggesting patients who may qualify for PARP inhibitor treatment may not be getting it.
"For years, we've had well-established, published guidelines that say that all women with ovarian cancer should have genetic testing, at least for BRCA1/2," said Leigha Senter, associate director of the division of human genetics and the genetic counseling graduate program at OSUCCC. "This [approval] is another motivating factor to actually achieve that."
The QUADRA study was conducted specifically to explore and establish HRD as a predictive biomarker, according to Moore. Myriad's BRACAnalysis CDx has been FDA approved as a predictive test alongside other PARP inhibitors, olaparib and rucaparib (Clovis Oncology's Rubraca). However, until now, HRD hadn't been solidified from a regulatory standpoint as a predictive biomarker.
In the earlier NOVA trial, BRCA mutation status and HRD were both used to stratify recurrent ovarian cancer patients who were platinum-responsive and showed that biomarker-positive patients had longer progression-free survival than biomarker-negative patients. However, the efficacy in the entire niraparib-treated population was impressive enough that the agency approved the drug in 2017 regardless of biomarker status and approved BRACAnalysis CDx as a complementary diagnostic.
The data presented at the recent European Society for Medical Oncology conference showed that PARP inhibitors olaparib, niraparib, and AbbVie's veliparib were broadly efficacious as front-line treatment options in advanced ovarian cancer, regardless of biomarker status. This data taken together has led some to conclude that these drugs have no role in later lines of treatment. "The data from QUADRA really shows you that's not true," Moore said.
In the QUADRA study, the aim was to investigate whether HRD status could be predictive of niraparib-response in heavily pretreated patients. "That's the signal we were looking for," Moore said.
The latest approval of the myChoice test with niraparib "is hopefully the start of the application of this companion diagnostic within this field," said Myriad CSO Jerry Lanchbury. He highlighted that Myriad's myChoice test was used to stratify patients in the three PARP inhibitor studies presented at ESMO, and the data showed these drugs have a particularly pronounced response in the HRD-positive subpopulation of patients. Data from these studies will also add to the utility of myChoice as a test oncologists can use to inform treatment strategies in ovarian cancer, said Lanchbury.
Myriad launched an early-access program for myChoice a few years ago, allowing oncologists to use the test to inform treatment strategies with platinum-based chemotherapies in triple-negative breast cancer. Additionally, academic researchers who have had the opportunity to use the test to enroll patients in various clinical trials for PARP inhibitors are also familiar with the test. However, outside of the academic setting, "there's a lot of education to still do in the field" about myChoice, said Lanchbury.
"The field is getting confused about what is an HRD test," he said, noting there are other tests claiming to gauge HRD but that only analyze a few DNA repair genes. In conversations with physicians, Myriad will highlight that its test provides the most comprehensive assessment of genes and biological processes involved in DNA repair that when disrupted make tumors more sensitive to niraparib.
"These biomarkers are capturing the hallmark of the inability to repair double-stranded DNA breaks," Lanchbury said. "It's a consolidator of all those processes that end up [hindering] double-stranded DNA repair and cause sensitivity to a PARP inhibitor like niraparib." Moreover, he emphasized that its test is the only FDA-approved companion diagnostic that can assess HRD status and predict response to niraparib.
A companion diagnostic by FDA's definition is required for the safe and effective use of a drug. A complementary diagnostic, by comparison, is not required for the safe and effective use of a drug but may be used to inform treatment strategy.
The FDA last year approved an HRD indication for Foundation Medicine's FoundationFocus CDx as a complementary test for rucaparib. A patient is HRD-positive by this test if she harbors a BRCA1/2 mutation (the test doesn't gauge large rearrangements) and has a percent of loss of heterozygosity (LOH) score of 16 or greater.
The myChoice test, by comparison, using its own proprietary algorithm, scores HRD positivity differently. The FDA-approved cutoff for a positive genomic instability score is at 42, and at this threshold, the test captures 95 percent of patients with BRCA1/2 mutations.
In another analysis with AstraZeneca and olaparib, researchers looked at using a HRD-positive cut off of 33, which Lanchbury said captures 99 percent of BRCA1/2 mutation carriers. "What you do by moving the threshold downward is you improve the sensitivity of the test," he said. "You bring more positives into the treatment group, but it comes at the cost of a little bit of decreased sensitivity."
The current approval sets the test's cutoff at 42, and "we would wait for data to clarify the situation with respect to other cutoffs," Lanchbury said.
OSUCCC's O'Malley certainly would like to learn more about the genomic instability score and the information used to calculate it. "It's not currently clear what the score really means," he said.
Although Myriad claims that its test provides the most comprehensive analysis of HRD, O'Malley is concerned that the company has provided limited data to support that claim and its proprietary algorithm further limits his ability to compare myChoice to other tests. "We as oncologists obtain next-generation sequencing test results on most of our patients looking for abnormalities in the tumor," he said, noting that these tests can identify mutations in genes known to be implicated HRD, such as RAD51, BRIP1, and BARD.
However, because Myriad doesn't report out the individual components that comprise its genomic instability score, it's not possible to know whether HRD-positive patients are responding to a PARP inhibitor primarily because of these HRD-implicated genes or because of some other mechanism the myChoice algorithm is measuring. "If Myriad were to provide those mutational analyses, the clinically meaningful impact [of the test] could be increased," O'Malley said.
While the latest approval enables more ovarian cancer patients to be eligible for PARP inhibitors, BRCA testing and HRD assessments aren't the only considerations in this setting. Patients get germline testing to gauge their inherited cancer risk. At top cancer centers, patients are increasingly offered testing on large NGS somatic panels to help guide them to clinical trials or precision oncology drugs.
At OSUCCC, Senter oversees genetic testing for ovarian cancer patients with a goal of ensuring that all patients have germline and tumor genomic assessments. If a patient's genetic test results reveal that she has microsatellite instability, for example, then she may be a candidate for immunotherapy with pembrolizumab. Therefore, doctors will have to weigh patients' HRD status in the context of other biomarker information and clinical factors when choosing the right treatment strategy for ovarian cancer patients, Senter said.