Skip to main content

Geneoscopy Develops Colorectal Cancer Dx Using RNA Biomarkers, Taking Aim at Exact Sciences


NEW YORK (GenomeWeb) – For patients looking to avoid a colonoscopy by using an at-home stool-sample diagnostic test for colorectal cancer (CRC), their current options are limited to Exact Sciences' Cologuard. But if the researchers at St. Louis, Missouri-based Geneoscopy have their way, those options may soon double.

The startup — which was cofounded in 2015 by Washington University in St. Louis researchers Erica Barnell and Yiming Kang, and Wharton MBA Andrew Barnell (Erica's brother) through the WashU School of Medicine's entrepreneurial incubator SlingHealth — is developing an at-home CRC diagnostic kit that is modeled on Cologuard, but that looks for RNA biomarkers in stool rather than DNA biomarkers. That difference, Erica Barnell told GenomeWeb, means that Geneoscopy's test is much better than Cologuard at detecting high-risk adenomas (HRAs). It's a difference that the company is hoping could help it take a big bite out of Cologuard's market.

It's certainly not an exaggeration to say that Exact created the at-home CRC testing concept. Until the last five years, using stool samples to test for colorectal cancer didn't really exist — the conventional wisdom was that there was so little human DNA in stool that it would be impossible to check for the genetic mutations that are indicative of cancer. For most people, the only option was to undergo screening with colonoscopy in five or 10-year intervals. And although other options exist, such as fecal occult blood testing (FOBT) or fecal immunochemical testing (FIT), direct imaging of the colon seemed to be the most reliable way to check for early signs of cancer.

Exact's breakthrough was in finding a way to screen for cancer in stool samples, enabling patients to collect their own samples and send them to Exact's lab for processing. It obviates the need for the fasting, intestinal cleansing processes, trips to the hospital, and anesthesia that can be a part of colonoscopies. The test itself — which was developed in collaboration with the Mayo Clinic — is an automated assay for tumor-specific DNA changes, including aberrant methylated BMP3 and NDRG4, a mutant form of KRAS, beta-actin, and hemoglobin.

So far, Exact has been able to create the parameters of its own market, reaping the advantages of a first-mover — the firm reported in January that its preliminary fourth quarter revenues rose 64 percent year over year thanks to 66 percent growth in the number of completed Cologuard tests during the quarter.

Not surprisingly, however, Exact's success has attracted the attention of potential competitors. Recently, privately owned Italian company Diatech Pharmacogenetics — which bills itself as the owner of more than 70 percent of the Italian molecular diagnostic market — released its own stool-based CRC diagnostic kit. Diatech's test searches stool samples for fragments of long DNA — pieces of DNA that are 200 basepairs or longer and which characterize non-apoptotic DNA that has been shed into the fecal stream from diseased mucosa — as a biomarker for early diagnosis of colorectal cancer.

That company's test isn't something a patient could do at home and it has to be paired with FOBT in order to provide maximum benefit, so it isn't directly comparable to Cologuard. But, its stool-based, non-invasive nature positions it as a possible competitor to Exact's test if Diatech gets regulatory approval in the US. Right now, the firm is concentrating on the European market, but it has plans to approach the US market with full validation from the US Food and Drug Administration within three years.

The homegrown threat to Exact's market supremacy, however, could be more immediate.

Much like Exact developed a way to extract human DNA from stool, Geneoscopy has developed a way to extract stool-derived eukaryotic RNA (seRNA) transcripts. The company's nucleic acid extraction method amplifies the human signals in stool and degrades bacterial signals, allowing for sensitive extraction of human RNA. Barnell and her team used this technology to develop a multi-target RNA biomarker panel for the detection of precancerous lesions and CRC, and this is the basis for the firm's diagnostic test. The company plans to explore the test's utility in gastrointestinal diseases such as inflammatory bowel disease (IBD) as well as developing it for the detection and prevention of CRC.

Geneoscopy's reasoning in searching for RNA biomarkers rather than DNA is that while DNA mutations can predict a patient's likelihood of developing a specific disease, they don't provide information necessary for clinical decision-making, such as phenotypic or quantitative data related to the symptoms of disease or the body's molecular response.

"Geneoscopy has really found a niche in a trending field. What we've observed within the liquid biopsy space is that 96 percent of diagnostics evaluate blood or tissue and about 92 percent look at either circulating tumor cells or cell-free DNA," said Barnell, who serves as the company's CSO and who is completing her MD/PhD at WashU School of Medicine. "We look at RNA, which is a real-time snapshot of what the cells are doing and how they're responding to their environment. So, we can look at things beyond just colorectal cancer. We can evaluate inflammatory diseases such as ulcerative colitis, Crohn's, or IBS. And then we can also look at the change over time."

Further, she noted, most individuals have a confluence of genomic variants that can cause cancer. But a diagnostic has to go beyond the variants if it wants to detect precancerous adenomas, which is the key to preventing the disease in the first place. "We observed that this RNA signature is associated with a lot of precancerous adenomas that all have varied genomic alterations," Barnell added.

Indeed, a study of 275 people conducted by researchers at Geneoscopy and WashU found that the company's test attained a 91 percent sensitivity for detecting CRC, a 73 percent sensitivity for detecting HRAs, and an 89 percent specificity for all other findings, such as medium-risk adenomas, low-risk adenomas, and benign polyps.

Cologuard has a 92 percent sensitivity for CRC but only a 42 percent sensitivity for HRAs. "Cologuard has received strong adoption, with 147,000 prescribing physicians and 934,000 tests completed in 2018. This test volume represents 4.1 percent of the total market for CRC screening, demonstrating the significant opportunity that exists for an improved noninvasive screening tool," the authors wrote in the study, which is currently on the preprint server BioRxiv.

"Despite Cologuard's strong commercial adoption, the test's accuracy profile does not significantly improve upon existing, cheaper alternatives," they added. "One fecal immunochemical test (Polymedco, OC-Light S FIT) has a documented accuracy of 93 percent for CRC at a 91 percent specificity (estimated HRA accuracy is 25 to 35 percent at unknown specificity). Since Cologuard is a FIT-DNA test, it is clear that the addition of the nine stool-derived DNA biomarkers do not improve the CRC accuracy and offer insignificant improvement in HRA accuracy."

That detection of precancerous HRAs is the value that Geneoscopy's test gives patients, Barnell said, particularly in that 73 percent rate compared to Cologuard.

"Cologuard has a FIT test in their assay and then nine DNA biomarkers. We have a FIT test and 11 RNA transcript markers. So, when we parse out the fecal immunochemical test, 50 percent of our positive predictive value is derived from RNA biomarkers for precancerous adenomas and the remaining 50 percent comes from the FIT test," she added.

In the study, the team noted that the accuracy profile of the Geneoscopy test represented a "significant improvement relative to all existing noninvasive alternatives." Specifically, the reported 73 percent sensitivity for HRAs represented a 74 percent improvement over Cologuard's detection of HRAs, and the 74 percent blended sensitivity for CRC and HRAs for Geneoscopy's test represented a 61 percent improvement over Cologuard.

"Improved sensitivity is attributable to the use of seRNA biomarkers, which offer several advantages relative to other stool- or blood-based biomarkers. First, seRNA biomarkers are derived from epithelial cells shed within the gastrointestinal tract. Therefore, the signal of the seRNA represents a homogenized sampling of the perilesional tissue, which is preferentially shed into the lumen and excreted in stool," the authors wrote. "Second, seRNA provides a concentrated and amplified signal that can be observed across multiple transcripts in a single pathway. This is a direct advantage over DNA-based biomarkers, which are limited to two copies per cell."

The researchers also found that the transcriptome provided them with a look at the downstream effects of many precancerous variants, regardless of the tumorigenesis pathway. In other words, a relatively small panel of seRNA biomarkers can encapsulate the larger genomic landscape of mutations that cause adenoma development, they added, and using these seRNA biomarkers to attain high sensitivity for the detections of HRAs will enable the development of a diagnostic that could facilitate early intervention and prevention of CRC.

Barnell noted that Cologuard's lower rate of HRA detection might also be affecting its reorder rate and market penetration. "What we've observed with Cologuard is that even though they've prescribed over 2 million tests and they've demonstrated that they can attain market penetration, the reorder rate for physicians is relatively low. Relative to colonoscopies, which [physicians] are prescribing on a day-to-day basis, it seems like they're only prescribing Cologuard in very niche circumstances when the patient absolutely refuses to get a colonoscopy," she said. "And the reason is because Cologuard cannot actually detect with high efficacy the precancerous adenomas, and that makes physicians concerned that individuals are going to be walking around with a high-risk adenoma that has the potential to transform to colorectal cancer before they get a subsequent screening. So, our value is stating that … we can actually prevent colorectal cancer development using our diagnostic rather than just merely detecting the presence of colorectal cancer."

Exact Sciences declined to comment for this article.

But before Geneoscopy can become the at-home stool-sample CRC diagnostic kit provider of choice, it still has a lot of work to do. Barnell and her colleagues are focusing on getting their study published. It's currently in peer review, and once they get feedback from the reviewers, she feels confident they'll be able to respond in 30 to 60 days.

"We did a vigorous testing of the machine-learning model that we've developed, and we already have validation of that in the wings," she noted.

The firm also needs to raise money. After it was founded, Geneoscopy received non-dilutive funding from BioGenerator, Arch Grants, Accelerate St. Louis, Washington University, and Wharton, as well as an investment from First Round Capital's Dorm Room Fund. In April 2018, the firm also received $1 million in seed funding from Billiken Angel Network and Missouri Technology Corporation, among other investors. The firm is also anticipating future fundraising efforts, but couldn't give specific details.

Also, because it was founded through SlingHealth, Geneoscopy owns its own patents. It filed three provisional patents and one utility patent to protect its technology.

Finally, and most importantly, the company needs a product to sell. And although Geneoscopy believes it has a better diagnostic than Exact, it does plan to follow Exact's path to the marketplace.

Its kit will be modeled on Cologuard. "Individuals who are recommended for screening would be prescribed a Geneoscopy test. We get shipped to their house, and they send it back to a centralized lab, and then we provide information to the physician," Barnell said.

Geneoscopy even plans to take advantage of the path to regulatory approval that Exact has already laid out. "We're finalizing the assay development and doing a lot of manufacturing in terms of creating the kit that would be given to the patient. After that we think we have about 18 months before we're able to obtain FDA approval," she noted. "We have Cologuard that we can look at and see what they did and follow their same regulatory pathway, which we don't have to carve out ourselves."

Further, although the company has already completed two clinical trials testing more than 340 patients, the company is currently enrolling patients in a 1,000-patient trial to complete the development of the diagnostic. This would serve as the final development study prior to Geneoscopy's pivotal study to generate data for a PMA submission to the FDA.

"Since our indication is for prevention of colorectal cancer, the incidence of high-risk adenomas is about 8 percent," Barnell said. She noted that Exact had to conduct a 10,000-patient prospective PMA to obtain regulatory approval because colorectal cancer incidence is only 0.5 percent, "but when you increase that to 8 percent, you can have a much smaller study size and still get the same indication. So, it's potentially possible that we can reduce our PMA study size to expedite time to market."

Geneoscopy is also looking to the future: though this specific diagnostic is for colorectal cancer screening, the firm is conducting three other clinical studies with researchers from WashU to develop subsequent diagnostics for other indications like IBD and IBS, to look at recurrence and predicting response for patients with rectal cancer, and looking at response to therapeutics for patients with Crohn's disease, all with the same underlying technology.

"We're still in preliminary development, but we hope to have comparable papers that we have for the colorectal screening hopefully by the end of the year or early next year," Barnell said. "We loved being in stool samples in RNA because there's really no one else in this space, so we have a huge area of whitespace for development. I think we can use this specific signature for colorectal cancer recurrence, but we'll be developing new signatures for other disease indications and applications."