FDA approval highlights potential of liquid biopsy methods
Interview by Steve Halasey
This April, FDA granted premarket approval for the Epi ProColon colorectal cancer screening test by Epigenomics AG, Berlin, Germany, making it the first blood-based circulating tumor DNA assay to receive agency approval. The agency move highlights the potential and rapid development of interest in so-called “liquid biopsies”—tests capable of detecting cancer biomarkers in body fluids, and thus reducing the need for invasive tissue biopsies.
Epi ProColon is an in vitro polymerase chain reaction (PCR) assay for the qualitative detection of Septin9 gene methylation in DNA isolated from the patient’s plasma. Cytosine residues of the Septin9 gene are methylated in colorectal cancer tissue but not in normal colon mucosa. This tumor-specific methylation pattern can be used to detect cell-free DNA shed into the bloodstream by tumor cells. Detection of colorectal cancer-derived DNA in blood plasma using the Septin9 methylation biomarker has been demonstrated in multiple clinical studies to be a reliable indicator of the presence of colorectal cancer.
Colorectal cancer remains the second-leading cause of cancer death in the United States. According to the American Cancer Society, there are projected to be more than 134,000 new diagnosed cases of colorectal cancer and almost 50,000 deaths from colorectal cancer in 2016 in the United States. Although screening and early detection of colorectal cancer can save lives, about 35% of eligible US patients are not being regularly screened.
Epi ProColon is indicated for colorectal cancer screening in average-risk patients who choose not to undergo colorectal cancer screening by guideline-recommended methods such as colonoscopy or immunochemical fecal occult blood (iFOB) tests. For patients, the test only requires a simple blood sample to be drawn as part of routine healthcare provider visits. There are no dietary restrictions or alterations in medication required for the test. Epi ProColon will be made available in the United States under a joint commercialization agreement with the company’s strategic partner, Polymedco, Cortlandt Manor, NY. Patient samples will be analyzed at a local or regional diagnostic laboratory.
FDA approval of the Epi ProColon test was based on demonstrations of safety and efficacy as established in three major clinical studies. Epigenomics also performed testing to demonstrate the potential of the test to increase participation rates in colorectal cancer screening. As typically required by FDA for new screening products, the company will initiate a postmarket study to show the long-term benefit of blood-based colorectal cancer screening using Epi ProColon.
To find out more about how the Epi ProColon test was developed, and the company’s plans for marketing the test in the United States, CLP recently spoke with Thomas Taapken, PhD, CEO and CFO of Epigenomics AG, Berlin, Germany, and Noel Doheny, CEO of Epigenomics Inc, Germantown, Md.
CLP: The Epi ProColon test is now a Class III FDA-approved diagnostic. What kind of interactions did you have with the agency about studies to support that approval?
Thomas Taapken, PhD: We had a series of interactions with the agency. We started our dialogue with a pre-investigational device exemption meeting in February 2010, and by that time we had already initiated the sample collection that was required to generate performance data for the PMA submission that we ultimately completed in January 2013.
CLP: Did the agency impose any special requirements on your studies?
Taapken: There were a total of three studies that provided data for the FDA evaluation, and we had a very intense dialogue with the agency about each of them. In fact, some of the study elements were performed specifically at the agency’s request. So FDA absolutely had input into the design of the studies.
CLP: What is the complexity categorization of the test under the Clinical Laboratory Improvement Amendments of 1988 (CLIA)?
Noel Doheny: Because it is a molecular method, in the United States Epi ProColon is restricted to use by laboratories certified under CLIA to perform high-complexity tests.
CLP: The test is a commercialized product—not a laboratory-developed test (LDT)—so you’re not intending to perform the test in your own laboratories, correct?
Taapken: That’s correct. We don’t operate our own laboratories. So we are going to sell the kits to those laboratories that have the instrumentation and qualifications required to perform the test, which in the United States means CLIA-certified high-complexity labs.
CLP: The test is intended for use with average-risk patients; what patients would be excluded by that restriction?
Taapken: The definition of ‘average risk’ is as broad as one can imagine, so the only patients excluded would be those considered to be at high risk. In this context, the definition of ‘high risk’ applies to patients who have either a history of colorectal cancer or first-degree relatives that have the disease. Those patients are advised to undergo colonoscopies starting earlier in life, and at more frequent intervals, to ensure that they don’t get the disease.
Doheny: Also, a small number of people in the US population have inherited syndromes that make them more likely to develop colorectal cancer. Those patients are also not considered to be average risk.
Taapken: FDA’s approval of the Epi ProColon test applies to average-risk patients within the screening age. According to US guidelines, colorectal cancer screening should start at age 50 and continue, depending on which guideline you read, up to age 79 or thereabouts. Those patients are the target for the Epi ProColon test, and they are also the patients with the worst compliance history for other forms of colorectal testing—and especially for colonoscopies or annual fecal immunochemical testing (FIT).
CLP: What about people who have previously had a colonoscopy, but had noncancerous polyps removed; are they still eligible for this test?
Taapken: Those people will get a recommendation from their gastroenterologist to follow up with a colonoscopy at shorter intervals. The recommendation currently in the guidelines for such patients is that they should undergo a colonoscopy every 5 years.
CLP: So they are considered to be in a high-risk group rather than an average-risk group?
Taapken: Correct. They would be classified as high risk—or at least as belonging to a category that requires testing at more frequent intervals. And also, their recommended testing would be done by colonoscopy, not by use of an in vitro diagnostic (IVD).
Doheny: Those recommendations are, of course, dictated by the size and stage of the individual patient’s polyps. So they need to be understood in that context.
CLP: The Epi ProColon test is intended for the qualitative detection of Septin9 gene methylation in cell-free DNA isolated from a patient’s plasma. How rare are the entities you’re trying to detect?
Taapken: The amounts of cell-free DNA present in a patient’s blood sample may vary, but we have developed the test in such a way that its analytical performance is extremely sensitive. We’re able to pick up one to two copies of methylated Septin9 gene in one mL of a patient’s plasma. This is a really sensitive test, picking up even the slightest quantities of methylated Septin9.
Doheny: The analytical sensitivity of the test has been shown to be approximately 6 pg/mL, or slightly less than that. If we had the most sensitive balance in the world, one copy of our genome would weigh about 6 pg. So the researchers concluded that we are able to detect at a level of one to two copies per mL.
CLP: When do tumors begin to shed fragments that make them detectable?
Taapken: Fragments of cancer cells can be detected whenever tumors grow. The cancer cells undergo either apoptosis or necrosis, and shed fragments of their DNA into the environment. That environment may include the inner lining of the colon, or it may also include the bloodstream by which the tumor is vascularized.
In our work, we have found that the limiting factor on detection is not the ability of the cancer cells to shed Septin9, but the level of a tumor’s vascularization, which in turn determines the quantity of methylated DNA available in the bloodstream of the patient to be detected.
Doheny: The 2014 study led by David A. Johnson, MD, at the Eastern Virginia Medical School demonstrated that some of those fragments are shed into the bloodstream even at the pathological stage 0 or stage I of cancer development.1
CLP: How was Septin9 identified as a colorectal cancer biomarker?
Taapken: Epigenomics discovered Septin9 in 2006 as part of a research effort in which we compared healthy tissue samples to diseased tissue samples, and specifically looked for differential methylation from the various genes in those tissues. We identified Septin9 as a gene that gets highly methylated in colorectal neoplasia, whereas in healthy tissue there is almost no methylation of the Septin9 gene. That’s how we started to develop this test to detect the methylated Septin9 gene in the bloodstream of patients.
CLP: In December 2011, you completed a prospective study on the sensitivity of the test in screening applications, which was just a little over 68%. Have studies conducted since then confirmed or changed that number?
Taapken: The 2014 study by Johnson concluded that the sensitivity of the test was in a comparable range, at 73%. In some previous investigative studies that were not prospective, the sensitivity of the test was found to be even better. Those studies have also been published.
We feel that sensitivity in the 70% range is good enough for a screening test. By way of comparison, for instance, fecal immunochemical testing (FIT) has comparable sensitivity. When performed annually, FIT has been proven to be a very effective test for reducing mortality from colorectal cancer, because a regular screening program has a cumulative effect.
CLP: In practice, what does the test’s specificity of 80% tell clinicians?
Taapken: Specificity indicates the percentage of patients that the test is correctly able to identify as healthy. A specificity of 80% means that 20% of the patients that our test picks up as positive do not have colorectal cancer.
In other instances, a specificity of 80% might not seem to be high enough. But in this case, patients with a positive result will be referred to undergo a colonoscopy, which is the first-line screening method recommended in current practice guidelines. So, in effect, the specificity level of the Epi ProColon test means that more people will be referred into the recommended standard of care.
Such referrals also help with the preventive effect of screening. In our studies, we have observed that a significant number of patients who received a positive result and were referred to undergo a colonoscopy wind up having adenomas or polyps—and sometimes tumors—resected at the same time.
Doheny: In a recent study, we found that more than 58% of patients who received apparent false-positive test results, and subsequently underwent reflex colonoscopy, were found to have an adenoma, polyp, or some other form of actionable medical condition. So medically, this testing protocol enables clinicians to identify potential lesions during a colonoscopy, and to remove them before they become malignant.
CLP: What is the negative predictive value of the test?
Taapken: The negative predictive value is 99.7%
CLP: Is the breakthrough methodology of the Epi ProColon test also suited for other types of cancer, using different markers? Is the company involved in developing markers and tests for other cancers?
Taapken: Yes, absolutely. DNA methylation is a naturally occurring phenomenon that is observed in a variety of cancers. Our R&D efforts have been devoted to identifying DNA markers that are differently methylated in healthy and diseased tissue, and we have so far discovered more than a dozen such biomarkers for other types of cancer. Our most advanced product in active development is a test for the diagnosis of lung cancer, where we are using two DNA methylation gene marker signatures to diagnose lung cancer in the blood sample of a patient.
CLP: At what stage of development is that test?
Taapken: We have completed technical development of the product and we are currently entering into clinical trials. Those will begin first in Europe, with the idea of moving over to US trials in the course of the next couple of quarters.
CLP: So that test is still several years away from commercialization, correct?
Taapken: For the United States, certainly. In Europe, we will be a bit earlier, based on the fact that the EU regulatory path is a little bit less cumbersome. We might have a product ready to be launched into markets that accept the CE mark sometime next year.
CLP: Does approval of the Epi ProColon test mean that the liquid biopsy technique is fully ready for clinical application?
Taapken: It is front and center, and coming to patients in the United States now.
Doheny: This is the first FDA-approved circulating tumor DNA assay, so it really is a breakthrough. This is the first time we have been able to go all the way to blood with a circulating tumor marker, and it represents a dramatic move toward liquid biopsy. This is really a pioneer product.
CLP: What is the reimbursement status of the test in the United States?
Taapken: We have historically permitted two US laboratories, Quest and ARUP, to offer an LDT version of the Epi ProColon assay under license. Originally, beginning 5 years ago, they were able to get paid by using existing CMS stacking codes.
In 2013, with the implementation of the new American Medical Association (AMA) common procedural terminology (CPT) codes for molecular pathology, Epi ProColon was assigned a Tier 2 CPT code. But earlier this year we submitted the test to go through the entire AMA gauntlet of molecular review, pathology review, and general review, and it was awarded a Tier 1 CPT code that will become fully effective in January 2017.
The laboratories that have been offering the test have educated payors about its availability. Meanwhile, we have developed both a comprehensive dossier available to all payors, and a full-blown health economic model to support payors in their adoption decisions.
Doheny: The economic model is being prepared for publication as we speak.
CLP: What do the health economics of the test look like?
Doheny: The first publication on the health economics of the Epi ProColon test was completed a couple of years ago by Uri Ladabaum, MD, a professor of medicine at the Stanford University School of Medicine.2 The health economics model used for the study was based on data from the surveillance, epidemiology, and end-results (SEER) database of the National Cancer Institute (NCI), which is the recognized modeling database for colorectal cancer population studies. The study showed that it is cost-effective for any unscreened patient to be screened with a blood test, because doing so has the potential to reduce downstream costs related to the disease.
The model revealed that the Epi ProColon test has the potential to accomplish two important objectives that will help to reduce the number of people who have not been screened for colorectal cancer. First, the high compliance rate of the test enables clinicians to increase the number of previously unscreened people who undergo screening. There is a tremendous concentration of cancer among the unscreened, and we’ve shown that we can detect more cases of cancer by adding blood testing methods to the repertoire of screening organizations.
Second, early testing can reduce overall healthcare costs. In the United States, 50,000 patients die of colorectal cancer each year, with an average healthcare cost of $100,000 during their final year of life. For a full-blown case of colorectal cancer, healthcare costs are between $350,000 and $400,000. But the health economics model demonstrates that early screening brings about long-term economic benefits to the overall cost of healthcare. There is an immediate, short-term cost for screening more patients; but the results ultimately bring about huge downstream savings.
CLP: Do you know what the average cost savings per patient are?
Doheny: The model that we’ve used looks at costs over a 3-year period, and calculates the cost to implement screening on a per member per month basis. That approach is preferable to trying to create a 7-, 10-, or 12-year model, because payors don’t really subscribe to those models, for a couple of reasons. They don’t trust the long-term data needed to inform such models. And moreover, there is so much change going on with therapeutic agents—especially with new agents that may cost more but also help people survive longer—that outcomes data are not yet fully available.
We are continuing to extend our model over a longer period. But in the meantime, for a 3-year window, we have been able to demonstrate the effectiveness of increasing cancer detection. And, if you’re willing to accept the presumption that every prevented cancer might otherwise have resulted in a $400,000 cost, those numbers become favorable very rapidly.
Taapken: The full details of this analysis haven’t been published yet. But when the health economics study has been published, those details should also be available.
CLP: Can you provide any details about the postmarket studies you are required to conduct for the Epi ProColon test?
Taapken: FDA usually requires companies to conduct postmarket studies as a condition of receiving premarket approval (PMA) for a screening test. Since screening tests are typically approved on the basis of one-time testing data, the idea behind the requirement to conduct postmarket studies is to establish the long-term value of the screening test.
Our study will perform repeat testing of patients with Epi ProColon after a year, to see if the test is still useful in picking up additional cancers that might have been missed in the first round of screening. The agency did not require a full-blown clinical trial protocol as a condition of premarket approval.
We have 6 months to negotiate the final details of the study design with FDA. We will also use that time to select the centers that will be conducting the study, and to settle the parameters. The study itself will probably start toward the end of this year. It’s clear that this is a multiyear effort, and we do not expect to have results from the postmarket study until 4 or 5 years from now, at the earliest.
CLP: What roles will Polymedco play as your distribution partner in the United States?
Doheny: We have been working with Polymedco for several years. Recently, the entire team of 20-plus sales reps received training, and they are now fully prepared to represent the Epi ProColon product. Our product will be inventoried in Polymedco’s warehouse, and we will use their order entry, customer service, technical service, credit, and collection services in support of the product.
Polymedco already has more than 15,000 customers who perform colorectal screening, making it the largest supplier of colorectal screening tests in the United States. Our agreement puts us in a position to leverage Polymedco’s history and standing in colorectal cancer testing, not only to reach the company’s installed base, but also to penetrate other areas where the company isn’t already present.
The plan is that our area managers will support Polymedco’s local sales representation. We will work together to identify key sales targets and programs, to execute contracts with those accounts, and to coordinate customer support; and Polymedco will then maintain responsibility for customer interactions after the start-up of testing in each laboratory.
Epigenomics will remain customers’ primary contact for technical service. Epigenomics will also retain responsibility for marketing, but we will work with Polymedco to handle key targets.
CLP: You’ve conducted a study comparing Epi ProColon to the OC FIT-Chek test, also distributed by Polymedco. Have you also compared Epi ProColon versus the Cologuard test by Exact Sciences?
Taapken: We have not compared Epi ProColon to the Cologuard test because Cologuard had not yet received FDA approval at the time we were performing our studies for regulatory approval of Epi ProColon. However, in all of our comparative studies we have worked with the current guideline-recommended product, such as FIT, where the OC FIT-Chek test distributed by Polymedco is the US market leader. So far, the US Preventive Services Task Force (USPSTF) has not recommended the use of Cologuard, so we haven’t studied it as a comparator to Epi ProColon.
It may eventually be desirable to compare the performance of Epi ProColon and Cologuard. But in the meantime, it should also be noted that the use of stool testing in the United States is still at a really low level, and it has actually been declining over the past decade. Presently, only about 11% of the US population is being screened for colon cancer on a regular basis by way of stool tests. So far, the introduction of Cologuard has not changed that number in any significant way.
The availability of a blood test opens a new category apart from invasive methods such as colonoscopies and stool testing methods. A blood-based test poses a much lower barrier for patients who have so far not been complying with existing methods to participate in colorectal cancer screening.
CLP: You have also conducted a study in which Epi ProColon achieved a remarkable 99.5% rate of patient compliance. Did FDA ask you to perform that study?
Taapken: FDA requested that we undertake that study in order to support our labeling for the test to be used by patients who have a history of noncompliance with recommended testing methods such as colonoscopies and stool testing.3 The agency wanted to see data to demonstrate that offering a blood-based test would actually increase the number of patients willing to undergo screening.
We discussed this request with the agency and incorporated its input to design the Adherence to Minimally Invasive Testing (ADMIT) study. In a patient population that at least twice within 15 months had declined screening by colonoscopy or stool testing, the study ultimately resulted in 99.5% of the patients agreeing to undergo a blood-based test. The study demonstrated that most patients—even those unwilling to undergo other types of tests—are very much at ease with the option of a blood test.
CLP: What does your US marketing and promotional plan for Epi ProColon look like, beyond the group of labs with which Polymedco already has contact?
Taapken: Our primary customers, of course, will be the laboratories that conduct outpatient testing and cancer screening. We pretty much have the address of every CLIA high-complexity laboratory in the United States, and we know which ones are engaged in our areas of interest. Polymedco will be targeting these laboratories, with our support, and that targeting effort is now fully under way.
Our second step will be to make the test known in the medical community. We will be working with key medical and public health opinion leaders to explain the value of the Epi ProColon test, so that we will have their support for including the test in relevant practice guidelines for colorectal screening. Selling to these opinion leaders will be accomplished primarily by our laboratory customers, whose sales forces already address this audience of physician customers. Laboratory corporations such as Quest and LabCorp already have large sales forces devoted to the medical community, and they have indicated that they will be committing sales resources toward selling to physicians.
Finally, we will also need to address the general public—including especially patients who have been noncompliant in the past. We will need to tell them about the availability of blood testing as an additional option for colorectal cancer screening. However, this step will need to follow a bit later.
Doheny: It’s important to note that we have adopted a marketing strategy that is totally different from the LDT approach being used for Cologuard. Our goal is to make US diagnostic laboratories a vital element in colorectal screening. Using their horsepower, we hope to help the National Colorectal Cancer Roundtable and the American Cancer Society achieve their goal of increasing the colorectal cancer screening rate to 80% by 2018.4 In the United States, that means we need to increase the proportion of people who are screened by 20% compared to the number being screened today.
Having a blood-based test available will enable major laboratories to participate in colorectal cancer screening in a way that previous methods have prevented them from doing. And it will permit those labs to offer an alternative to proprietary LDTs such as Cologuard. We think that’s a very important factor for improving patient access as well as the ease of getting tested.
CLP: Will Epigenomics and its key opinion leaders be conducting any webcasts or other activities to influence the guidelines and clinical adoption?
Taapken: We are in the beginning phases of setting up our plans to address key opinion leaders, but we will certainly be using a variety of different tools to reach that audience. It makes sense for us to engage in giving talks and conference presentations, as well as webcasts. I think that we will not avoid any of these media.
CLP: Have you been working with Quest and ARUP to incorporate their study findings into the development and promotion of the test?
Doheny: Both labs have published articles and presented poster sessions centered on their historical and utilization data, so they have been talking about the test for years. However, the version of the test they have been running was an LDT that was only intended for the detection of Septin9; it was never labeled to be used for cancer screening.
The new kit bears a label for screening among average-risk patients, and will permit laboratories to address this as a screening opportunity among the more than 25 million US patients who have yet to be screened. From a marketing perspective, the new label will permit Quest and ARUP to use the word ‘screening,’ which they were not previously able to claim.
CLP: How would you sum up the potential impact of this new testing methodology?
Taapken: For colon cancer, the demographics and magnitude of the problem are significant. And in our view, the costs associated with cancer deaths in the United States are fully preventable. But the key to prevention is better compliance with screening, and we feel that existing methods will not cut it. We need to get blood-based, convenient methods out to patients and physicians, to make sure everyone gets tested for this disease.
Steve Halasey is chief editor of CLP. He can be reached via [email protected].
- Johnson DA, Barclay RL, Mergener K, et al. Plasma Septin9 versus fecal immunochemical testing for colorectal cancer screening: a prospective multicenter study. PLoS One. 2014;9(6):e98238; doi: 10.1371/journal.pone.0098238.
- Ladabaum U, Allen J, Wandell M, Ramsey S. Colorectal cancer screening with blood-based biomarkers: cost-effectiveness of methylated septin 9 DNA versus current strategies. Cancer Epidemiol Biomarkers Prev. 2013;22(9):1567–1576; doi: 10.1158/1055-9965.epi-13-0204.
- Epigenomics reaches 99.5% adherence for Epi ProColon in ADMIT study [press release]. Berlin, Germany: Epigenomics, 2015. Available at: http://www.epigenomics.com/news-investors/news-media/pressemitteilungen/article/epigenomics-epi-procolonR-erreicht-in-admit-studie-eine-teilnehmerrate-von-995.html. Accessed May 19, 2016.
- Meester RGS, Doubeni CA, Zauber AG, et al. Public health impact of achieving 80% colorectal cancer screening rates in the United States by 2018. Cancer. 2015;121(13):2281–2285; doi: 10.1002/cncr.29336. Accessed May 19, 2016.