By Tony Fong
NEW YORK (GenomeWeb News) – Shares of Exact Sciences tumbled 20 percent on Friday, a day after the firm reported validation study results on its stool-based DNA colorectal cancer screening test.
The firm reported that the test had a sensitivity of 64 percent for pre-cancerous lesions and 85 percent for cancerous lesions and a specificity of 88 percent.
Though the results are still early-phase and further studies will be needed before the test, called Cologuard, can reach the market, Kevin Conroy, president and CEO of Exact Sciences, heralded the results and said that they suggest the test could offer a new screening strategy for colorectal cancer, which kills 50,000 people in the US each year and 150,000 patients in Europe each year.
"We think that … Cologuard has real potential as a colorectal cancer screening tool," Conroy said during a presentation on Thursday at a conference held by the American Association for Cancer Research in Philadelphia. "We also think that if you look at other products, there are no non-invasive products that can detect anywhere close to 60 percent of pre-cancer," the main indication that the study targeted.
Cologuard is not meant to replace colonoscopies, he stressed, but is being developed for use as an adjunct to colonoscopies.
Conroy also noted that the study was to validate Exact Sciences' underlying Quantitative Allele-specific Real-time Target and Signal, or Quarts, amplification assay technology which was used to develop Cologuard, as well as the markers contained in the test. It was the first major validation of the test, Conroy said.
Despite the results, which were generally regarded as positive, Exact Sciences' stock took a beating on the Nasdaq, trading down $1.71 to close at $6.86 in Friday trade — though that may have been due to a misunderstanding that the study was a clinical trial, an analyst said.
Since the start of September, Exact Sciences' share price has jumped about 80 percent leading up to yesterday as anticipation about the study results increased. While the stock has slid since the results were unveiled, Quintin Lai, an analyst at investment firm R.W. Baird, said that it may have been more the result of misguided expectations and confusion that it was a validation study and not a clinical trial.
"The Street was hoping to see perfection," he said.
In a research note published today, Lai increased the price target on Exact Sciences' shares from $9 to $13.
In the study, the company and its collaborator, the Mayo Clinic, took three distinct pathways in their approach, using four DNA methylation markers, two DNA mutation markers, and a human blood protein marker.
"If any of these are positive, then the test result is positive, and the patient should go to colonoscopy," Conroy said.
The study consisted of a specificity study, a training set, and a test set. In total, 1,187 stool samples were processed and analyzed.
The specificity study was to determine whether any biases were introduced by age, gender, ethnicity, and geography. "And what we learned [is] that it was only age-related methylation bias that occurred," Conroy said. An algorithm was developed to "mute" the bias.
In the training set, the test achieved 89 percent sensitivity and 64 percent pre-cancer sensitivity and 89 percent specificity. In the test set, it achieved 78 percent cancer sensitivity and 64 percent pre-cancer sensitivity and 85 percent specificity.
In his presentation, Conroy said the test set was the first performed outside of Exact Sciences, "so clearly we have some improvement to do with the protocols and training of outside labs."
Also, in a pilot study carried out with three markers, Cologuard achieved 85 percent cancer sensitivity, 55 percent pre-cancer sensitivity, and 90 percent specificity.
A difference between the pilot study and the training and test sets, Conroy said, was that Exact Sciences used multiplex assays in the pilot study that had not been optimized yet for the training and test sets. As the company continues optimizing those assays, it will lead to improvements in sensitivity, he added.
Taken all together, Cologuard achieved the results presented at the conference yesterday: the ability to detect 64 percent of colorectal pre-cancers, 85 percent of cancers, and a false positive rate of 12 percent.
One potential shortcoming of the test, so far at least, is that it was not especially sensitive at detecting late-stage cancer — 69 percent sensitivity for Stage 4 colorectal cancer. Conroy said that Exact Sciences hopes to improve that, but added that the goal as it pursues US Food and Drug Administration approval, with a 2012 target for a submission, would not be to target Stage 4 cancer, since at that point the cancer is already likely to be showing symptoms.
Plans for Improvement
Moving ahead, he identified several areas that the company will home in on as it further develops Cologuard, including the improvement of quality control processes and assay reproducibility.
Also, it needs to improve DNA capture efficiency and its mutation assays. During his presentation, Conroy said that the KRAS mutation assay worked, but the story for the BRAF assay was different. "It didn't contribute at all," Conroy said.
Graham Lidgard, Chief Scientific Officer for Exact Sciences, said during the presentation that a BRAF assay was developed for a specific mutation as part of the study. "As we were running the assays, the signal was higher than expected, and we're still investigating the cause," he said.
Another area of improvement will be hemoglobin detection. Exact Sciences is developing a new fecal immunochemical test "and we are going to work very hard to make sure that that increases cancer detection sensitivity," Conroy said. Hemoglobin tests detect cancers, but not pre-cancers, he said.
As it moves Cologuard toward possible FDA clearance, Exact Sciences is working with the agency as well as the Centers for Medicare and Medicaid Services to ensure proper study design. The goal is to run one study that would satisfy both FDA and CMS, Conroy said.
"Presently the agencies have a lot of questions about this product, and we're working very hard to answer those questions," he said.
The firm is now preparing for a clinical trial and is reviewing potential sites to determine which ones will get the highest enrollment and the best results, he said.
If Cologuard does eventually receive FDA clearance, getting the test to be run in large numbers by laboratories would mostly involve building out manufacturing capability, Lidgard said.
"This is the first platform that has all the components that can be commercialized for this type of test," he said, adding that although stool-based colorectal cancer tests "have been shown to be relevant over the past 10 years," they could not be practically implemented in clinical labs. "This is the first set of technologies [and] assays that can be run in the clinical lab," he said.
Commercialization of the test is still a few years away, but R.W. Baird's Lai said that the study generated "a lot of buzz" at the AACR conference and for good reason. While the study results were not perfect, he said the results were positive and at the very least, it showed that "you can sort markers that are cancer and non-cancer" and that Exact Sciences has a system that can potentially be used by different laboratories, he told GenomeWeb Daily News.
In his research note, Lai wrote that Exact Sciences "has set a new bar for non-invasive colorectal cancer testing … We think it takes a 'good' first step to get to a 'great' final product, and we think [the company's] preclinical results and further refinements bode well for the 2011 clinical trial."
He noted that the Mayo Clinic's David Ahlquist, lead investigator on the validation study, feels that Cologuard could "provide excellent programmatic sensitivity over a three-year screening interval."
Lai anticipates the test to be launched in 2013 and achieve 18 percent market penetration by 2019, generating about $830 million in revenue for Exact Sciences. By then, he estimates the colorectal cancer test market will reach $4 billion to $5 billion in the US.
Cologuard, he said, has the potential "to meaningfully improve compliance, as the test is more convenient than colonoscopies and fecal-occult blood tests."