NEW YORK – Shares of Exact Sciences rose more than 27 percent Thursday after the company presented new data at an investor conference on a pan-cancer liquid biopsy test that it's developing in partnership with the Mayo Clinic.
The test — which identifies esophageal, liver, lung, ovarian, pancreatic and stomach cancers — showed overall sensitivity of 86 percent at a specificity of 95 percent in initial results. Its sensitivity was 89 percent for esophageal cancer, 83 percent for liver cancer, 78 percent for lung cancer, 90 percent for ovarian cancer, 90 percent for pancreatic cancer, and 87 percent for stomach cancer.
The data was presented today during the Cowen Liquid Biopsy Summit by Exact Chairman and CEO Kevin Conroy.
He also said that Exact had collected more than 3,500 samples from cases across all cancer stages, more than 1,400 of which were from stages 0 to 2.
The company has been focusing most of its recent discussions on its upcoming liver cancer test and the development of a new version of its colorectal cancer test Cologuard. But the possibility of a pan-cancer test using methylation biomarkers was brought up by Mayo Clinic Associate Professor John Kisiel — who collaborated with Exact on the development of the liver test — in November 2019.
During Exact's second quarter earnings conference call in July, Conroy said that the company had the key pieces in place to introduce new tests for colorectal, liver, pancreatic, esophageal, and prostate cancers, and that it had presented data at a scientific conference across six cancers in different sample types — including stool, tissue, and blood — spanning indications from screening to recurrence monitoring.
According to a note today from William Blair analyst Brian Weinstein, investors had assumed that the tests as discussed by Conroy were being put together in an early form of a pan-cancer screening test, but he hadn't heard much commentary from the company on this topic until today and hadn't seen any data.
"At the conference, management was forceful in its commentary about why it thinks it is the most primed company in cancer diagnostics to bring a pan-cancer test to market," Weinstein wrote. "The team and laboratory infrastructure built over the last 11 years has the required experiences to not only discover and develop markers for a test, but also be able to cross regulatory, reimbursement, test processing, IT, and sales-and-marketing hurdles upon launch. These are no small tasks, and Exact today laid out investments totaling well over $3 billion across these channels."
The early data the company presented is encouraging. While they're not as far along as other datasets in the pan-cancer space, they do give Exact "the mandate to move into the pan-cancer space," he added. The company didn't discuss any timelines for product launches, but will likely update investors over the next several quarters on progress around increasing specificity and additional marker discovery and development.
In their own note to investors on Thursday, Evercore ISI analysts Luke Sergott and Vijay Kumar asked how Exact's tests would compare to Grail's multi-cancer blood-based screening test, called Galleri, which relies on targeted methylation sequencing. Grail, which is in the process of being acquired by Illumina for $8 billion, plans to launch the test next year.
Given that Grail's CCGA 2 study showed a sensitivity of about 60 percent for 12 pre-specified cancers for Galleri, Exact's headline number of more than 80 percent sensitivity looks good in comparison, Sergott and Kumar said. That means Exact's test could hypothetically detect 24 percent more cancers than Grail's. Exact's test showed a specificity of around 95 percent compared to specificity of 99.3 percent for Grail's test.
"While sample sizes are different, generalizing the headline sensitivity/specificity results to all cancers, we think [positive predictive value] for EXAS is about 14 percent versus 40 percent to 50 percent for Grail (ILMN)," Sergott and Kumar wrote. "We think clinicians care about the sensitivity, especially if it helps catch cancers at an earlier stage and can bend the mortality curve. Given this, we think the [approximate] 14 percent implied [positive predictive value] for EXAS (based on prelim results) is good enough."
The analysts noted that this will bring up interesting questions for payors when they consider the higher sensitivity compared to the lower positive predictive value as they look at Exact versus Grail in the future, when pivotal prospective data is available for both tests.
Exact's shares were up 27 percent in mid-day trade on the Nasdaq at $94.13.