Exact Sciences: Misunderstood And Undervalued?

May.28.14 | About: EXACT Sciences (EXAS)


EXAS is fundamentally Undervalued.

FDA approval imminent, CMS coverage should be within 90 days of FDA.

Market underestimates potential revenues. EXAS stock valuation model yields $40+.

Exact Sciences' (NASDAQ:EXAS) potential is so misunderstood and undervalued that we believe that EXAS probably has an upside of 100%-300% over the next twelve months based on the prospects of its first product, Cologuard, on a standalone basis. Keep in mind, our evaluation does not consider the Exact Sciences product development pipeline. In conjunction with the Mayo Clinic, EXAS is pursuing numerous additional markers for other gastrointestinal tract cancers that can be detected with the same or similar kit. This year's Digestive Disease Week 2014 [DDW] listed at least ten related investigations.

Exact Sciences Corporation is a molecular DNA diagnostics company whose first product, Cologuard, is a multiple-marker DNA molecular platform product. Cologuard is a non-invasive diagnostic test kit for colorectal cancer [CRC] and, importantly, CRC precursors. The test combines detection of multiple DNA markers in stool [sDNA] and a fecal blood test. According to figures produced by the CDC in 2010, CRC kills over 50,000 people a year in the United States. And the World Health Organization [WHO] reports that in 2008 over 225,000 die each year in Europe. The novel part of Cologuard is the automated, highly sensitive test for aberrant cells shed from the lining of the gastrointestinal tract. Because the progression of CRC from a small polyp to stage IV CRC takes 10-15 years, a test with the sensitivity of Cologuard should substantially reduce annual deaths from CRC by detecting pre-cancerous polyps prior to extensive harm.

Before discussing financial projections for the launch of Cologuard by EXAS, we need to put aside, in our opinion, a number of pieces of misinformation floating about the market - these generated primarily by ill-informed shorts.

FDA approval is a near certainty, and should be expected in a couple of months or even earlier. The successful conclusion of the 10,000 patient clinical trial [DeeP-C, NCT01397747] demonstrated efficacy of Cologuard comparable to colonoscopy and supremacy of Cologuard over the FIT (Fecal Immunochemical Test and its precursor the Fecal Occult Blood Test - FOBT). Cologuard was significantly better than FIT in colorectal cancer and pre-cancer detection, has comparable sensitivity to colonoscopy and is non-invasive. The subsequent publication entitled "Multitarget Stool DNA Testing for Colorectal-Cancer Screening" (April 3, 2014 available online on March 19, 2014) by Imperiale, et. al. in the New England Journal of Medicine (370: 1287-1297) - the premier impact journal for this clinical trial - shows the importance of Cologuard as a diagnostic tool.

Cologuard detects 92.3% of cancers, 66% of pre-cancers over 2cm (most likely to progress to cancer) and 42.4% of all advanced adenomas (see CC-73 and CC-79 from the EXAS presentation to the Molecular and Clinical Genetics Panel Meeting - March 27, 2014). And, Cologuard is significantly better than its nearest non-invasive competitor - FIT - at cancer and pre-cancer detection.

Some conflicted sources have tried to focus on the sensitivity for precancerous lesions and a specificity of slightly less than 90%. Cologuard has the highest sensitivity of any noninvasive test for pre-cancerous lesions. Think about it - if a patient has a false positive or misses a small precancerous dysplasia for slowly developing colon cancer, the worst case is merely additional tests and a colonoscopy, not CRC. This is a limited cost and/or inconvenience for the other 90% that are set on avoiding the costs, side effects and risks of a colonoscopy in the first place. At a suggested interval of once every three years, detection of cancer is virtually certain, and detection of precancerous lesions (and their removal with colonoscopy) should go a long way to eliminating CRC in the screening population.

Also, we believe some editorials and sponsored reports have mistakenly overstated the reasons for failure rates of participation in the DeeP-C study. Of the 12,776 participants, for the secondary endpoint - superiority to FIT, 2,787 participants were not included in the final 9,989 participants that were available for testing. The figures here and in the table below are from Figure 1 of Imperiale, et. al., op. cit. and page 18 of the NEJM presentation available on the Exact Sciences website (NEJM_Presentation_Final_031914.pdf). It appears that more participants withdrew due to an unwillingness to undergo colonoscopy, but in a study with three tests - colonoscopy, Cologuard and FIT, there can be and were multiple reasons for exclusion. This short table illustrates that point, among others, for the secondary outcomes of the DeeP-C clinical tests comparing Cologuard and FIT.


Withdrew consent- no stated reason


Did not undergo colonoscopy


Colonoscopy excluded


No colonoscopy/colonoscopy excluded


Did not submit a stool sample


No stool sample for FIT or Cologuard


No stool sample for Cologuard


Sample that could not be evaluated with Cologuard


Sample that could not be evaluated with FIT


Sample lost


Stool sample excluded for FIT and/or Cologuard


Participants not included for the secondary outcome

Click to enlarge

There is no way that we can know why 464 withdrew consent, but one can surmise that dislike of colonoscopy was a predominant reason. Of the remaining 2,323 participants, 1,168 did not undergo colonoscopy - that's 50% of the participants that were not included in the secondary analysis for unwillingness to undergo an invasive procedure! Including the 304 participants that a had colonoscopy excluded, 63% (1,472/2,323) were excluded due to unwillingness to undergo colonoscopy or other issues with the colonoscopy procedure.

Only 128 of the 12,776 participants did not submit a stool sample - that's 1% of the total, and 5.5% (128/2,323) of those that were excluded from the secondary analysis for a stated reason. Even within the confines of a clinical trial, it's easier to get someone to use a non-invasive test with no burdensome preparation.

There were 723 (851-128) participants whose stool samples had some issue - either for Cologuard or FIT or both. FIT was affected by 51% of those stool issues (367/723) and Cologuard by 95% of the stool issues (689/723). Is this worrisome? No - this is a tempest in a teapot, in our opinion, and an aberration of the clinical trial process. We claim that the issues with stool samples are typical of common testing procedures and will be easily resolved in practice.

In a clinical trial, if the protocol is not followed exactly, a patient cannot be included in the final analysis. So, if a stool sample is not obtained, delivered too late (after colonoscopy, for instance) or damaged in any way (like a leaky delivery container), that patient would not be included in the final sample. In the real world, of course, this would not be a failure of either Cologuard or FIT. As with any noninvasive test conducted at home, at a lab or at a doctor's office, at very low cost, a new sample would be requested, and tests would be run on the new sample. And so, in the real world, there should be few, if any, mechanical testing failures that cannot be corrected with a second sample. We believe that use of the sensitive, noninvasive Cologuard diagnostic test (especially including the "compliance engine" for follow-up) will vastly increase the rates of compliance in the screening population.

Others reports have talked about the lower costs of the inferior stool blood tests - FIT and FOBT. The fact of the matter is that for $300-$500 every three years, at-risk patients will be able to avoid an expensive, intrusive, risky procedure (colonoscopy) and avoid the low sensitivity of cheaper tests to precancerous lesions. Cologuard is in the sweet spot - a noninvasive, reasonably priced, sensitive diagnostic test for CRC and its precursors. We have entered a new era for medical diagnostic tests that are enabled by multiple-marker, DNA molecular platforms, often involving multiple patents, like Cologuard.

The March 27, 2014, Medical Device Advisory Committee meeting [AdComm] was a complete success with three unanimous 10-0 votes favoring Cologuard. The favorable comments of the panel at the end of the EXAS AdComm (p 211-215 March 27, 2014 Transcript FDA AdComm meeting materials) and those reported in the final remarks of Kevin Conroy, Chairman and CEO of EXAS at the most recent conference call on May 1, 2014 provide further support for this proposition.

While nothing is certain in this world, FDA should approve Cologuard in the near future. According to comments by CEO Kevin Conroy at the May 7, 2014 Baird Growth Stock Conference, EXAS is finalizing FDA inspections for a supplier and the EXAS manufacturing site and would expect FDA approval a month and 1/2 thereafter (EXAS website, comments on the timing of inspections and subsequent approval are at 12:54 and 14:48 minutes from the start of the presentation). We believe that FDA approval is a near certainty and expect FDA approval within weeks or months and EXAS should be rolling out Cologuard kits and recognizing revenue before the end of the year.

Another fundamental misunderstanding concerns the potential use of Cologuard. It is not a replacement for colonoscopy as a therapeutic tool. As a diagnostic, Cologuard is a superior replacement to FIT. And, Cologuard is a non-invasive diagnostic with cancer detection capability equivalent to colonoscopy. Despite protestations by EXAS to the contrary, we also believe that Cologuard will be used to screen patients instead of the more burdensome (bowel preparation, lost day) and invasive colonoscopy test. Colonoscopy will provide more detailed diagnosis and therapy, if needed.

This will be the sensible cost-effective choice for sequencing diagnostic testing for CRC in our medical system. There is even some suggestion in the clinical trial data that Cologuard detects right-side polyps and serrated sessile adenomas more efficiently than colonoscopy. (See, for instance, the references in the journal article by Heigh, et. al, "Detection of Colorectal Serrated Polyps by Stool DNA Testing: Comparison with Fecal Immunochemical Testing for Occult Blood (FIT)", PLOS One, January 2014, Volume 9, issue 1, e85659.) If so, that is very important - as many as one-third of all CRC starts this way (see page 1, column 2 and the references to Heigh, et al 2014). Pragmatically, therapeutic colonoscopy will follow a positive Cologuard diagnostic result, and Cologuard should rapidly replace FIT as a screen and be the premier non-invasive test for CRC.

Some have claimed that a so-called "ick" factor will restrain use of Cologuard. Perhaps the proponents of this ill-conceived piece of misinformation are unaware of the fact that more than 10 million FIT/FOBT users per year face and endure the same "ick" factor with the FIT/FOBT test? The Fecal Occult Blood Test [FOBT] and the Fecal Immunochemical Test [FIT] together are currently performed at a rate of 10.2 million per year according to Exact Sciences estimates (see slide 29 of the May 2014 Corporate Presentation). Since these 10 million diagnostic tests have the same "ick" factor (if that really is an issue when testing for CRC!) why wouldn't they immediately switch to Cologuard with significantly better detection of cancer and pre-cancerous polyps? We expect a rapid transition of the FIT-using population to Cologuard and a significant displacement of routine diagnostic colonoscopies by Cologuard.

With the inevitability of FDA approval, recent misinformation has focused on the Centers for Medicare & Medicaid Service [CMS] and private coverage and reimbursement decisions. There are many moving parts to these decisions, but it is highly likely that CMS will find the use of Cologuard reasonable, necessary and cost-effective, and CMS will provide coverage. CMS has authority under 42 CFR §410.37(A)(1)(V) to include coverage of additional CRC screening tests through issuance of a National Coverage Determination [NCD]. We believe that the Cologuard diagnostic test is simple, acceptable to the general screening population (non-invasive), accurate, repeatable, sensitive, specific, and, at intervals of 2-3 years, cost-effective in providing large benefits from early detection and prevention of CRC. We believe that CMS has conclusive evidence (the DeeP-C clinical trial and other related sDNA and CRC studies) that screening with Cologuard can alter the natural history of CRC in a significant proportion of those screened.

Although the date of the CMS decision is contingent on FDA approval, EXAS has participated in a parallel process with CMS and expects to have a preliminary NCD within a week or two of the FDA approval of Cologuard and a final NCD 90 days thereafter. And, the reimbursement rate set in the final NCD will be retroactive to the preliminary NCD. For a clear explanation of Exact Sciences' progress with FDA and CMS, listen to or read the transcript of the May 1, 2014 EXAS conference call. Carefully note CEO Kevin Conroy's confidence in obtaining a preliminary coverage decision shortly after an FDA decision and a final coverage decision within 90 days thereafter.

Before discussing the reimbursement rate for Cologuard, there is another important factor in the coverage decision. When the US Preventive Services Task Force [USPSTF] evaluates Cologuard along with other CRC screening (currently scheduled for 2015), we expect that Cologuard (like colonoscopy and FOBT before it) will be graded A which means "The USPSTF recommends the service. There is a high certainty that the net benefit is substantial." The implication under the Affordable Care Act [ACA] is clear; there will be full coverage and payment as a diagnostic screen for Cologuard at regular intervals by all payers - public or private. Presently FOBT (an inferior predecessor to FIT) is rated A by the USPSTF (October 2008 USPSTF recommendations) and we see no reason why Cologuard would not also be rated A by the USPSTF.

We expect CMS and private payers to all provide coverage for Cologuard as a diagnostic test based on its superior clinical trial results, superb publication record, unanimous AdComm approval and support and, soon, FDA approval. Also, we expect Cologuard to significantly reduce healthcare costs for Medicare, Medicaid and private insurers through displacement of colonoscopies and reductions of surgeries and treatments for late stage CRC.

As regards CMS reimbursement rates, we disagree with some naysayers for a number of reasons. In the broadest sense, Cologuard provides the government and healthcare professionals the first significant opportunity to materially reduce annual mortality rates and the overall healthcare costs by thousands of lives and billions of dollars, respectively. Counter to some of short-oriented opinions made with out-dated facts used out of context of historical developments, we believe that there are technical justifications for a crosswalk analysis that should result in a reimbursement rate of $300-$500.

And so, we come to the reimbursement decision, which will determine the ultimate financial success of Cologuard and EXAS. What will be the reimbursement rate? Kevin Conroy has stated, time and again, that EXAS would be pleased with a $300 per test reimbursement rate. To put this in perspective, EXAS management has highlighted two important facts:

1.) At a reasonable run rate, Cologuard costs are in the neighborhood of $100-$110 (Exact Sciences Management Discusses Q4 2012 Results - Earnings Call Transcript - February 20, 2013). In the context of a question about the gross margin on a $300 per test price, Maneesh Arora, CFO and COO of EXAS says "guidance is 65%. [ ... ] but at steady-state, this is still 65% gross margin or better." The implication is cost per test is $105 or less.

2.) A crosswalk (or sum of parts) analysis of existing reimbursement codes for molecular diagnostic tests and FIT add up to nearly $500 per test. At the February 13, 2014, Leerink Swann Global Health Conference EXAS presented the following estimate of crosswalk pricing for the components of Cologuard (* assuming Tier 2 Palmetto GBA pricing per analyte for 19 analytes for CPT 81401, see Slide 41 accompanying the remarks of Kevin Conroy).


CPT Code







$ 21.86



$128.00 *



$128.00 *


Click to enlarge

Some of the most recent misinformation from shorts claims that a gap-fill process rather than a crosswalk will be used to calculate the reimbursement rate. This could not be further for the truth in our opinion. CMS uses a crosswalk process when comparable codes exist.

§414.508 Payment for a new clinical diagnostic laboratory test states [in part]:

For a new clinical diagnostic laboratory test that is assigned a new or substantially revised code on or after January 1, 2005, CMS determines the payment amount based on either of the following:

(A) Crosswalking. Crosswalking is used if it is determined that a new test is comparable to an existing test, multiple existing test codes, or a portion of an existing code. [ ... ]

(B) Gapfilling. Gapfilling is used when no comparable test is available.

From the crosswalk estimate above, it can easily be seen that codes exist for all the primary components of Cologuard. Gap-fill analysis is only used when comparable codes do not exist.

In the May 1, 2014 conference call and his May 7, 2014 conference presentation (from about 12:20 to 18:00 minutes), Kevin Conroy clearly answered questions on this issue and indicated that in his discussions with CMS, it appears that crosswalk from existing codes would be used. And, Conroy also indicated that while he did not expect a gap-fill process to price Cologuard, that he would be pleased with the outcome of that calculation also.

Either way, gap-fill or crosswalk, it appears that Conroy's goal ($300 or more per test) should be attained. And, the shorts which claim, without analysis, to project sub-$300 reimbursement rates are, in our professional opinion, just blowing smoke with no justification for their choice of gap-fill or the resulting numbers that they propose.

A gap-fill analysis of the cost-effectiveness of Cologuard, which depends on the likely choice of a three-year interval between tests would show that Cologuard is the most effective diagnostic screen for CRC and its precursors. In other words, $100 per year ($300 every three years) for a superior test for CRC is a cost-effective way of reducing the number of colon cancers and the substantial costs of treating CRC. We expect that reimbursement in the $300 per test range and costs in the $110 per test range will provide excellent gross margins.

And now, we return to discuss the financial consequences of the above determinations why we think that Exact Sciences is currently and fundamentally undervalued.

Central to our thesis is the fact that, with the company steering conservative expectations, in our opinion, for the potential roles of Cologuard, the market has been led to believe and embrace a supplemental role for Cologuard in the screening and detection of colon cancer in the U.S. As we suggest above, Cologuard will have a primary role in screening CRC, and that is how we base our analysis.

Conservative expectations are fine for a company whose primary objective is to win FDA approval with minimal objections by the gastroenterology community and take the high road in the presentation of its clinical results; however, it has engendered projections of an extended roll out and annual revenue potential of only $1-3 billion in about five years.

Some naysayers even dwell on the quarterly financial results (i.e., revenues and EPS) of Exact Sciences which are irrelevant for a cash-rich (about $250 million: $116 million reported at the end of 2014 Q1 and $137.7 million from a stock offering closing on April 9, 2014), startup biomedical about to introduce new disruptive technology to a large marketplace.

In the May 7, 2014 Baird Growth Stock Conference, Kevin Conroy was asked by Jeff Elliot why his estimates of volume and revenue were so low. KC laughingly deferred to COO Mahesh Arora, but the point was made. The comments on the low estimate for sales - termed "prudent" - appear at 25:40-27:18 minutes into the presentation. We believe, based on our analysis that Kevin Conroy and EXAS appear to be rather conservative in their sales estimates.

We understand the genesis of the conservatism, but for valuing the prospects of EXAS in the future, we want realistic estimates of the revenue from Cologuard. We are content to model a price that at minimum would be $300.

Our analysis takes another tact which is more along the lines of a reverse engineered value analysis, not atypical for an emerging company about to release a revolutionary product for a resistant problem that results in hundreds of thousands of worldwide deaths each year.

An analyst must evaluate a range of assumptions for the potential, time, and penetration of the worldwide markets for a new, efficacious diagnostic test for colon cancer that is as accurate as a colonoscopy at a fraction of the cost and far beyond the efficacy of existing blood/fecal tests that are cheaper. There are two important sections that need to be dealt with, 1) the time frame for management to tilt-up its infrastructure and distribution and, 2) the amount of the worldwide market that is reachable and penetrable by EXAS.

The first factor, in our opinion, is rather easy to deal with as evidenced by all of the information from Exact Sciences. This company has a seasoned management team ready to immediately roll out the technology, operations, sales and administration. The first diagnostic lab capable of processing 1,000,000 tests a year is virtually ready and they already have the cash (about $254 million as of the May 1 conference call), know-how and plans to complete other labs in cookie cutter fashion in paralleled projects such that the ability to handle millions of tests a year in a matter of years is not a limiting issue.

The more debatable element is how many potential physicians (there will be more than gastroenterologists now involved) and patients will embrace Cologuard and the frequency with which they will enlist it as a detection strategy. Needless to say, we believe that Cologuard will be a worldwide revolution. If you estimate that Cologuard displaces 60-80% of colonoscopies, FOBTs, and FITs; encourages much higher compliance; is priced near $400/test; is used at a higher frequency (every 2-3 years rather than 5-10 years for colonoscopies); and achieves market share of 40-60% (there is no other viable competitor right now in our opinion), then any simulation projection model you construct points towards annual revenues of $3-$12 billion within 5-7 years.

Others may reach different numbers than ours; however, it is inescapable that Exact Sciences is looking at the potential of around $10 billion revenues by 2020. At an income margin of 9-12%, that would translate into net income in the future of nearly about $1 billion or EPS of about $10 per share. With a mature P/E of 15x, that would imply a future stock price of perhaps $150 and discount back to approximately $50+ for the stock price today with a risk-adjusted discount rate of over 20%.

Now you see why we have reached the conclusion that the stock price of Exact Sciences is highly undervalued. Sift through the cross currents of hedge funds, shorts, competitive procedures, conflicted parties and a company appearing to be trying to remain conservative in their proclamations and you will reach a similar conclusion: this stock should sharply jump 200% to 400% after they get FDA approval and CMS coverage. You can argue with our assumptions, but it appears that any set of numbers implies that EXAS's stock price should be $25-$50+ when Cologuard receives FDA approval.

Disclosure: I am long EXAS. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Additional Disclosure: We do own measureable positions in EXAS. Our article reflects our own opinions and our projections and analysis are based on public reports that we believe are reliable and our own assumptions on future events that may, or, may not prove to be accurate. We have portrayed our assumptions and our reasons and opinions for their derivation. Do your own due diligence. We are not receiving compensation for our article (other than from Seeking Alpha). We have no business relationship with Exact Sciences, Inc. other than that of an interested shareholder.