Friday, June 4, 2021

Utility of HRD Testing in Ovarian Cancer Expands, but Standardization Remains Elusive


This article first appeared in OncLive and was written by Hayley Virgil. 

Thomas C. Krivak, MD, discusses the significance of HRD in ovarian cancer, approved and prospective HRD tests, and considerations for genomic testing in this population.

Two tests for homologous recombination deficiency (HRD) have been FDA approved for patients with ovarian cancer: myChoice CDx and FoundationOne CDx. Their results provide important information that can be used to guide treatment decisions, according to Thomas C. Krivak, MD, but differences between these next-generation sequencing assays can make standardization of their use difficult.

myChoice CDx, developed by Myriad Genetics, was approved by the FDA in May 2020 as a companion diagnostic for niraparib (Zejula) monotherapy and olaparib (Lynparza) plus bevacizumab (Avastin) in patients with advanced ovarian cancer, respectively, and FoundationOne CDx, developed by Foundation Medicine, was approved in November 2017 to identify molecular abnormalities in several solid tumors, including ovarian cancer.1-3 Although both tests are useful, the clinical data to support myChoice CDx may be more compelling, according to Krivak.

Importantly, each test defines HRD in a different way. myChoice CDx identifies HRD based on deleterious or suspected deleterious mutations in BRCA1/2 and/or a positive Genomic Instability Score.4 In comparison, FoundationOne CDx defines HRD based on tumor tissue with BRCA positivity and/or a high loss of heterozygosity.5 Because approved and prospective tests have different definitions of HRD, standardization may not be possible, according to Krivak.

“I do not believe there is going to be standardization,” Krivak said. “I would have to say HRD is this concept that each company is going to define, and then it is going to be up to the individual clinician, whether that is a pathologist or someone who specializes in genetics and cancer, who will say, ‘We think this is the best test for these reasons, and this is what we are going to do in our practice.’ I do not think a perfect test is out there.”

In an interview with OncLive® during an Institutional Perspectives in Cancer webinar on ovarian cancer, Krivak discussed the significance of HRD in ovarian cancer, approved and prospective HRD tests, and considerations for genomic testing in this population. He is director of the Ovarian Cancer Center for Excellence and cochair of the Society of Gynecologic Oncology Research Institute at Allegheny Health Network.

OncLive®What are some of your key treatment considerations for patients with ovarian cancer?

Krivak: For patients with advanced ovarian cancer, you have to always think about germline testing. We do not want to have everyone focused on tumor testing. First, think about a good surgery, chemotherapy for the majority of patients, and then the germline genetic testing component. That is important [to] help us to determine who may qualify for medications after chemotherapy.

Also, an ounce of prevention is better than a pound of cure. [If] we can identify patients who have germline mutations that place their family members at a higher risk for ovarian cancer, then we can intervene in those patients with risk-reducing surgery. With technology, if you have a patient who has a daughter aged 21 [years and] tested positive, we are going to make interventions at that point for her. However, even [in terms of] reproductive capabilities, there are [institutions] in the country where you can do prenatal genetic testing before you implant the embryos.

Also, we really have to do some type of molecular interrogation to classify and select our medications. We have numerous medications that we can use to treat these patients; now we have to figure out what the optimal timing is to initiate these medications, as well as what sequence [they should be given in]. Hopefully, some of this molecular information will help.

How does HRD testing fit into the molecular testing landscape?

Someone who has HRD and is known to be BRCA [wild-type] is going to triage to some type of maintenance strategy. If patients have a germline mutation, testing them for HRD up front probably is not a great use of resources because you are going to use an oral medication. If that patient recurs, you may want to test them later on or you may want to actually use a different test; you may not want to use an HRD test. You may say, “I want to get more molecular information, such as tumor mutation burden and DNA mismatch repair [MMR] status.” This other type of information may be gained through molecular interrogation.

[We should] test up front. If someone is going to receive chemotherapy followed by maintenance in the recurrent setting, if they have not received a PARP [inhibitor and] have not been tested, especially the platinum-sensitive [patients, testing] may help triage [what] to use: oral medications or potentially intravenous medications. Or [the choice may be to not use any if the patient is] HRD proficient.

What we are really testing is the tumor. To me, that can be looked at as unfavorable if you do not have any tumor [tissue] available and maybe the patient did not undergo debulking. If I’m going to test a patient, I want to test the most recent biopsy because chemotherapy may potentially induce some changes within the tumor itself. Most of us would say, if someone has recurred [for] the third or fourth [time], and you are going to do some molecular interrogation at that time, you would want to biopsy that patient and use the most recent tumor [specimen] that you can get. If you cannot get that, then you go back to the initial specimen.

As to why [we] want to test, [it is] so we can get more information on how to best treat patients. Certain medications may be indicated for only 3% to 5% of patients. Microsatellite instability or MMR deficiency may be present in only 3% to 5% of patients with ovarian cancer, but those patients now qualify for checkpoint inhibition. That may open up that strategy for that patient who may not otherwise be [considered for it].

We do not have all the answers and we do not know how to sequence chemotherapy appropriately. This is another piece of information that will help guide us through how we are going to sequence treatments. How we [choose to] treat our patients will be based upon curability and noncurability, prior treatment-related adverse effects, and possibly cost and distance to [the institution]. If we look at the molecular profile of the tumor, that may [present] us [with] other ideas for someone who may have not experienced any response in the past.

Do you have a preference among available HRD tests?

Each company is developing a test [and is defining what] HRD is. We, as clinicians, need to be able to say, “This patient is homologous recombination deficient [according to] this test,” whether it is the Caris Life Sciences test or the Tempus test. The 2 tests that are FDA approved are by Myriad and Foundation Medicine. All those companies I mentioned are very solid companies with good research, but, no doubt, Myriad has really done a tremendous amount of research to develop their test. Foundation Medicine [has, as well,] in very different ways. However, when someone calls me and says, “Hey, listen, I have this patient who has this, this, and this, and they are HRD positive,” my immediate reflex is to answer, “Wait a second, how do you define that? Is that by Myriad or Foundation Medicine?”

Do you believe HRD testing will ever become standardized?

I wish we could say, “HRD is defined as this.” If you had spoken to me 3 years ago, I would have said that I define HRD as any molecular alteration within that homologous recombination pathway…I do not want to say that I have no data, [but we] have very limited data. It is mostly intuition. The way I defined [HRD back then is not how I define it now]. We are probably never going to have a standard definition for [HRD]; it probably just boils down to economics. All companies want to say that their test is [best].

No comments:

Post a Comment