This story has been updated to include the company's newly announced breakthrough device designation.
NEW YORK – Extracellular vesicle diagnostics firm Mercy Bioanalytics is making progress in its plans to market and eventually garner US Food and Drug Administration approval for a novel ovarian cancer screening test to serve what oncologists have long viewed as a severely unmet need. The company announced on Wednesday that it had received breakthrough device designation from the FDA for its assay, a step that eases it's path through agency review.
Mercy CEO Dawn Mattoon said in an interview this week that Mercy has long planned to pursue IVD approval and, as such, has been careful in planning and enacting its validation efforts to make sure that it can demonstrate the value of its test, called Halo, in a field where other options have failed to meet necessary benchmarks.
"It was really important to us that we train our classifier, lock it down, and then apply that locked assay sequentially to independent verification cohorts," Mattoon said.
Mercy's assay is based on the detection of circulating extracellular vesicles (EVs), isolating cancer-associated vesicles from a background of non-cancer-associated particles using a set of three surface markers, which Mattoon said are proprietary. The enrichment of tumor-derived EVs involves the use of DNA oligonucleotide-conjugated antibodies targeting the three epitopes. Those oligos are then bound and ligated, and the cancer EV-associated signals are detected via qPCR.
Others have explored EV or exosome-based cancer testing strategies, but Mattoon said that assays that use single markers can have low specificity because each individual target suffers from overlap between cancer-derived EVs and those that are reflective of other cell mechanisms.
Mercy believes the sensitivity and specificity it has been able to achieve is a result of its unique triple discrimination method.
The company had previously shared validation results from smaller studies, including one in which the Halo assay distinguished high-grade serous ovarian cancer from benign controls with 89.9 percent sensitivity at 98 percent specificity. It also distinguished stage I and II tumors with 84.6 percent sensitivity.
This week it reached a new milestone with the release of newer data it will feature at the upcoming annual meeting of the American Society of Clinical Oncology.
These newest results hew close to what the company had seen in prior studies but reflect a much larger cohort that Mercy gleaned from a trial called UKCTOCS, the United Kingdom collaborative trial of ovarian cancer screening. Mattoon said that the trial provided a great way for Mercy to perform a new validation because the company could explore its test performance at different time points prior to a patient's cancer diagnosis.
UKCTOCS began enrollment in the year 2000 and enrolled 200,000 women who were followed for up to 20 years. The trial participants were randomized to a no-screening arm, a screening arm that used annual transvaginal ultrasound imaging, and a third arm that screened women using the blood biomarker CA-125.
Because of its size, a subset of trial participants, as expected, developed ovarian cancer.
Although CA-125 and ultrasound did demonstrate a significant stage shift in the 20-year trial, with more cases detected at earlier stages in the testing arm than the control arm, this didn't translate to a significant enough benefit to drive guideline inclusion and widespread clinical use.
"Those technologies are 40 years old now," Mattoon said. "We really haven't had an improvement on those in all that time, but we feel like we've got something that is finally a tool that can overcome those limitations and actually deliver a benefit."
In the study that the firm will present at ASCO, investigators ran the Halo test and a research-use CA-125 assay on samples from more than 1,300 control cases who did not develop cancer and 129 cancer cases from the trial who were diagnosed up to three years after their initial enrollment sample. Ten of the case samples and a little more than 100 control samples were excluded due to operator error or sample issues.
For cases sampled within 12 months of diagnosis, Halo detected 82 percent at 98 percent specificity, a significant improvement over CA-125, which showed 63 percent sensitivity and 96 percent specificity in the same sample set.
Furthermore, the improvement was not limited to late-stage tumors. Halo's sensitivity for stage I and II tumors in this group was 85 percent, compared to just 46 percent for CA-125.
"The whole objective is to detect at stage I when you really can deliver therapy with meaningful clinical benefit," Mattoon said. "It's not sufficient to have a test that does a good job of detecting stage III and stage IV cancer."
Although it only detected 15 percent of cancers in samples from two to three years before diagnosis, Halo again managed to catch more than CA-125, which had just 5 percent sensitivity.
With almost 1,200 controls, Mattoon said that the company is especially confident in the study's specificity readout, which suggests a low false positive rate.
Investigators also performed a relatively small sub-study, she added, using 100 samples from women who were deemed positive for cancer by transvaginal ultrasound imaging, but had benign findings when they received surgery. Testing samples from that subset with both Halo and CA-125, investigators saw four times fewer false positives for the EV-based assay — 2 percent versus 10 percent.
"The theme here has been consistent across every study that we've conducted: that we are significantly superior to these other methods, both in sensitivity, which is very important, obviously, for early disease detection, but also specificity, which is vital in a disease like ovarian cancer," said Mattoon.
"We don't want to be referring women unnecessarily for downstream diagnostic work-up so it's really quite important to have a low false positive rate," she added. "This has been one of the historical objections of the clinical community in adopting ovarian cancer screening, and now we've demonstrated empirically that we've been able to really improve upon that false positive rate."
Elizabeth Suh-Burgmann, a gynecologic oncologist with the Permanente Medical Group, said that some results from Mercy's study raise questions. For one, although Halo performed better than CA-125 overall, this "was not consistently the case," she wrote in an email. For example, the test performed "no better than CA-125 for patients whose blood was sampled 13-24 months prior, and worse among patients in this subset with early-stage disease."
Another question is whether the Halo test may be confounded by non-serous ovarian tumors that present a low risk to patients. "Since high-grade serous ovarian cancer is responsible for over 70 percent of ovarian cancer deaths, for a test to provide real clinical value as a screening test, it must specifically detect high-grade serous cancers," Suh-Bergmann wrote.
Although the assay outperformed CA-125 in terms of false positives in ultrasound-detected masses, the current study didn't include a specific control group of individuals with other types of ovarian cancer to explore its potential positivity rate in this population.
"More detailed information is needed regarding the sensitivity and specificity of the test by histologic type to be able to assess the potential clinical utility," she added.
UKCTOCS recruited postmenopausal women, requiring enrollees to have at least one ovary and to not currently have or have previously had an ovarian malignancy. Mattoon said that this matches precisely what Mercy envisions for its intended use population, although one major caveat to the cohort is that it included overwhelmingly white individuals, with very little representation of other racial groups.
Next steps for the firm include additional studies to support an application to the FDA for premarket approval, which it aims to file by early 2026, as well as a push to engage with various guidelines groups with the hope that, eventually, the test will be endorsed by the United States Preventive Services Task Force and reimbursed by Medicare.
"What we think will support our case is that there is an enormous clinical unmet need here," Mattoon said, citing the statistic that 80 percent of ovarian cancer is diagnosed at an advanced stage, and that only 30 percent of individuals diagnosed at stage IV will survive five years. Meanwhile, in the rare cases diagnosed at stage I, five-year survival rises to 90 percent.
"There is, as you might imagine, a very active patient advocacy community that we've begun to engage with … and there is some legislative activity already ongoing to try to help mandate coverage for ovarian cancer screening," she added.
Mattoon said that Mercy doesn't believe a randomized controlled trial of Halo is either likely or necessary.
For one, there have already been RCTs conducted more recently than the UK trial, which Mattoon said demonstrated an even greater stage shift, including a study led by MD Anderson that enrolled almost 8,000 women, again following them for 20 years. That trial came very close to demonstrating not just improved early detection, but also a mortality benefit.
Meanwhile, academic groups have begun to develop models that can translate test performance into its inferred stage-shift and impact on outcomes, which could be applied to an assay like Mercy's based on reported sensitivity and specificity compared to previously studied markers like CA-125, Mattoon added.
Finally, she said, it's important to recognize that when the UKCTOCS trial was initiated, oncology was a relative therapeutic desert compared to the current landscape of targeted treatments and immunotherapies, which could make the prospect of translating earlier detection into improved outcomes even easier.
"There are all these layers of evidence, including clinical real-world evidence that we can start to generate subsequent to commercialization, modeling studies, and then some of this other trial data that we think will be sufficient," she said.
In the background, Mercy has also been investigating its platform for other tumor types, with lung cancer being the most advanced at about 18 months behind the Halo test, and other data in breast, prostate and colorectal cancer.
"We're actively exploring which of those we might elevate into our next clinical program," Mattoon said.