NEW YORK – More than two years after bringing on board a minimal residual disease detection platform through the acquisition of ArcherDx, Invitae is making progress in building evidence to grow its presence in the testing and monitoring of early-stage cancer patients.
Multiple clinical studies are currently ongoing, and the firm recently published its first study on the method's analytical performance. Bob Daber, Invitae's chief scientific officer, said that although the analytical validity study began prior to Invitae's acquisition of the technology, the firm has since made some notable improvements that feature in the new publication in Molecular Diagnosis & Therapy.
Like several competitors in the space, Invitae's MRD approach involves upfront sequencing of tumor tissue, which is then used to develop patient-specific cell-free DNA tests. These methodologies have been at the heart of a web of lawsuits, including one by Natera against ArcherDx/Invitae, in which a Jury awarded Natera $19.4 million in damages for royalties and lost profits this past May. Invitae shortly after announced its intent to appeal.
One distinguishing aspect of what Invitae has dubbed Personalized Cancer Monitoring is the allowance for patient-specific panels of different sizes.
"What we learned initially from some of our early studies is patients, even those with 48 or 49 variants identified, were getting excluded because we had to require 50," Daber said.
"We've run some studies where we've gone up, I think even in TRACERx we went up to 200. But there's an economic aspect here too. You can increase sensitivity and performance, but then you're increasing cost … so for us, 50 was sort of that sweet spot where the gains after that were not as significant, but the costs were still linear, and so we thought this had the right balance for commercial adoption," he added.
In the firm's analytical validation study, investigators reported that they could create personalized panels for all 63 of the patient tissue samples in which there was at least 20 percent tumor content. In another 30 samples with down to 10 percent tumor content, they could find enough variants for panels for 80 percent, or 24 individuals.
According to the authors, the PCM assays' limit of detection was 0.008 percent allele frequency when utilizing 60ng of cell-free DNA input with 18–50 variants and a baseline threshold for calling positivity. Using 10ng of cfDNA with a 18-variant panel and a more stringent monitoring threshold, they calculated a LOD of 0.05 percent allele frequency.
Daber said that the different thresholds were developed as part of the expansion of the test from single-timepoint MRD detection to longitudinal monitoring.
"The way we make a call, [ctDNA-positive or negative] essentially, is that we look at the signal to noise and we ask, 'Is the signal significantly different from what we would expect from noise alone?'" Daber said.
"Back in the day with the Archer algorithm, it had less stringent criteria because it was looking at a single time point. But when you do monitoring, because you're now increasing the number of sampling events, it was important for us to set a different threshold for making a positive call. And in the test as we run it today, we've actually abandoned those two different significance thresholds and we run everything just at the more stringent threshold."
The company's task moving forward is to follow up this analytical validation with robust clinical validity and utility data.
"We've spent a lot of time focusing in two areas over the last few years. One was our working with our pharma partners, so doing what we call fee-for-service work," Daber said. The outcome has included the CheckMate 816 Phase III clinical trial, which assessed neoadjuvant chemo-immunotherapy in resectable lung cancer, where investigators used PCM assays to detect ctDNA clearance following treatment.
Results, published last year, showed higher rates of ctDNA clearance in patients treated with immunotherapy plus chemotherapy than in those treated with chemotherapy alone, with associated longer event-free survival and higher pathological complete response rates.
In another publication earlier this year, investigators from the TRACERx trial reported on their use of the Invitae MRD tests in 197 patients with early-stage lung cancer, who were longitudinally followed up for up to 5 years.
The team was able to create personalized ctDNA assays with between 70 and 201 variants, and in initial testing performed within 120 days after surgery 25 of the patients were ctDNA-positive including 49 percent of all patients who experienced clinical relapse. In ctDNA surveillance testing every three to six months, tests also identified impending disease relapse in an additional 20 percent of previously negative patients.
In addition to lung cancer, the company has ongoing studies in breast, pancreatic cancer, and colorectal cancer, Daber said.
"What we're trying to do is send more volume through our studies as we build the clinical story [for PCM]. This paper was about the analytical performance … so we [still need to] generate the clinical evidence."
When it expands its marketing efforts for PCM for physician use, the plan is to continue a pan-cancer strategy. "The way that we're looking at this is that it is pan-cancer, but we put a heavier emphasis in the areas where there's more literature. So, commercially we're not going to restrict somebody [from ordering for] a cancer we haven't yet studied. But we're also not going to go out there and try and target those as sales points," Daber said. "The intent is to do these studies and get publications out there and then leverage those publications as we're ramping commercially with payers as well as with the physician audience."
Several other MRD tests have gained Medicare coverage in recent years under a local coverage determination issued by the MolDx program, which outlines various criteria that a test must meet. Daber said that since Invitae is under the jurisdiction of Medicare Administrative Contractor Novitas and not Palmetto, which houses MolDx, it wouldn't necessarily have to submit to that program for coverage, but in the long term the firm is "keeping an eye on all fronts," anticipating that it will likely end up offering the test in other laboratory locations, which could include those under Palmetto's purview.
In the meantime, as of March this year, PCM testing is covered by Blue Shield of California for cancer patients who have had a surgical intervention with curative intent, and for whom physicians plan to use MRD testing to inform adjuvant or targeted therapy, or monitor for relapse or progression.