Description

Exact Sciences

I have followed Exact Sciences story for the past few years, initially because I had friends in the industry.  My recent idea was to examine EXAS as a potential short, following a steep rise in the stock price, which is up over 5x over the past year.  

Other reasons why I initially thought that a (prudently-sized) short might be interesting:

• Short-seller Citron recently published a short report on EXAS (on May 15)

• In June, the company issued 7mm shares at $35/share, raising $245mm gross, roughly doubling its cash balance.  Probably a prudent move to improve the cash position for this unprofitable company, but also perhaps an indication that management believes that today’s valuation is an attractive price to sell.

• A recent (April) licensing deal with MDX Health suggested that MDX (presumably an informed party) did not highly value Exact’s future revenue stream.  This issue was raised in a follow-on report from Citron.  

• Analyst reports seem quite optimistic – valuing the company based on revenue multiples, describing a U.S. addressable market as 80mm people (certainly not 50+ U.S. adult will be sending in stool samples to Exact), discussing the company as a “platform” for new drugs, and extrapolating the Cologuard success to new, yet-to-be developed products (a concept that I’m generally skeptical of) – especially in the competitive field of blood-based cancer tests.

HOWEVER, after digging into the business and doing some primary research, I’m more favorably disposed to EXAS as a long investment.  Certainly, the range of outcomes for an early-stage healthcare company is wide, and the investment includes relatively high risks.  However, I do think that there is substance here, and if Cologuard (EXAS’ main product) continues to be successful, there is reason to believe that revenues could surpass the current street expectations.

Background

EXAS is essentially a one-trick pony (for now), selling  Cologuard, a screening test for colorectal cancer (CRC).  CRC is the third-leading third-leading killer among adult cancer deaths in the U.S.  The gold-standard testing method is a colonoscopy – which costs roughly $1650 per treatment.  In contrast, Cologuard costs $500-600 – for a screen that is less invasive – the patient provides a stool sample from home (“poop in a box”), which is then tested in Exact’s lab.  Clinical trials have shown that Cologuard is a meaningfully better predictor than the most common alternative (much cheaper $25) FIT (fecal immunochemical test, which looks for blood in the stool.)

Summary of Clinical Study Results – from EXAS’s summary at link

Results from the clinical study demonstrated that Cologuard successfully met the primary performance measure of the study, establishing a clinically meaningful sensitivity and specificity for CRC. Sensitivity of Cologuard for CRC was 92.3% (60/65) with a one-sided 95% confidence interval lower bound of 84.5, substantially exceeding the protocol-specified threshold of 65%. In addition, Cologuard successfully demonstrated a clinically meaningful specificity according to the protocol-specified criteria. The specificity of Cologuard was 86.6%, with a one-sided 95% confidence interval lower bound of >86.0%.

Clinical study results demonstrated superiority of Cologuard to FIT for sensitivity in detecting CRC. Secondary performance measures included an analysis of performance Cologuard and FIT using colonoscopy as a reference. Cologuard correctly detected 60 of the 65 total CRC cases identified by colonoscopy (92.3%). FIT captured only 48 of the 65 CRC cases identified by colonoscopy (73.8%). FIT identified only a single cancer that was not identified by Cologuard. Cologuard, meanwhile, identified 13 cancers that were missed by FIT. Cologuard was compared to FIT using a non-inferiority test for CRC sensitivity. In addition, Cologuard demonstrated superiority over FIT with respect to sensitivity for CRC using an exact McNemar’s comparison test as the one-sided p-value (p=0.0018) was well below the p <0.025 threshold for superiority. The lower bound of the one-sided confidence interval for the Cologuard – FIT difference was 0.080, substantially exceeding the protocol-specified non-inferiority threshold of -0.05. Establishing superiority for AA sensitivity required a one-sided p-value

Establishing superiority for AA sensitivity required a one-sided p-value <0.025 (exact McNemar’s comparison test). Cologuard demonstrated superiority for AA sensitivity, with a p-value of <0.0001, substantially below the threshold for superiority of p<0.025. FIT identified only 29 AA cases that were not captured by Cologuard, while Cologuard identified 170 AA cases that were not positive on the FIT test.

Analysis was also performed to calculate the Cologuard negative predictive value (NPV) for Category 1 (CRC) and Category 2 (AA). Clinical results show that a negative patient result for Cologuard gives 99.94% assurance that the patient does not have cancer and a 94.79% chance that the patient does not have an advanced adenoma.

Summary of Results

What we have is a test that is much less invasive, time consuming, and expensive than the gold standard test (colonoscopy), for a disease that will cause roughly ~50K deaths in the U.S. in 2017, is treatable when detected early, and is expensive to treat in late stages.

A few of the criticisms of Cologuard (from Citron and others):

• It’s a lot more expensive than FIT – about 20x the cost ($500 vs. $25).  Fair, but it also identified 60 out of 65 cancer cases (92%) vs. 48 out of 65 (73%).  Given the costs of late stage treatment and the number of CRC deaths, it seems that patients (and perhaps that payers) would prefer the higher quality test.

• It’s not as good a CRC indicator as a colonoscopy.  Absolutely correct.  This seems to be a key misunderstanding.  Cologuard is not being pushed as substitute for your colonoscopy.  However, many patients avoid colonoscopies, and the CDC estimates that ~60% of the eligible population is up-to-date on colonoscopies, which are recommended at least every 10 years, or more frequently, depending on age and other indications.  So, for those patients who are keeping up with their recommended colonoscopy protocol, they should NOT attempt to use Cologuard as a substitute.  However, for the remaining population that isn’t being screened, Cologuard is a useful first-line test that will detect cancers 92% of the time (and the later stage cancers at a higher rate.)  This is one reason that Cologuard is often prescribed by primary care physicians—who may see patients that have not been keeping up with their protocols.

• There are too many false positives.  It is true that the test gives false positives at a rate of ~13%, more than twice the rate of FIT.  However, if we step back and consider this—aside from some patient anxiety and additional cost—the result of this is more colonoscopies, which GI doctors generally believe is profitable for them (more procedures) and good for patients in general (a higher testing rate).  False positives are not ideal, but they’re also not a reason to throw out this test.

Adoption & Reimbursment

The major payers (CMS and Aetna, Cigna, HCSC) have approved reimbursement for Cologuard, and UnitedHealth was the most recent major payer to come on board (with coverage starting July 1.)   The remaining payers to convert are some of the regional Blue Cross plans, but in general payer adoption has been strong.

Physician surveys by sell-side analysts have been positive, with primary care physicians (who often prescribe the screen) indicating that doctors expected strong adoption in the next 1-2 years.

Volumes & Profitability

The company is manifestly unprofitable and is spending heavily on sales and marketing (including B2C marketing.)  Sales and marketing spend was $113mm in 2016 and an estimated $150mm in 2017E, vs. revenues of $99mm and an estimated $206mm, respectively, and gross profits of $54mm and an estimated $133mm for those years.  

Gross profits are high, with estimates for 2017E revenue per tests of $455 vs. a cost of $167 (a cost which is declining), thus the key to profitability will be gaining scale, public awareness, and widespread adoption, so that Exact can moderate their sales and marketing blitz and leverage their fixed costs.

In terms of volumes, EXAS completed roughly 104K tests in 2015, 244K tests in 2016, and analysts estimate are 487K tests in the 2017.  The company is currently expanding capacity at its lab to allow for 2mm or more test per year.  Considering that there are roughly 15-20mm colonoscopies performed per year (15mm in 2012), and considering that EXAS’ test would be, in theory, administered at greater frequency compared to a colonoscopy, it’s not hard to imagine volumes ultimately to 4-5MM/year, which would suggest revenues of $1.5-2B and gross profits in the $1B range.  If EXAS can then curtail some of its operating expense growth (current OpEx expected to be $260M in 2017), we could imagine $600M of EBIT and $400mm of earnings, or nearly $3.40/share.  (The math is illustrative, so I’ll dispense with valuing the current cash position [less cash required to reach profitability] and with discounting future earnings back to present.)  The broader point is that—although EXAS is burning cash for now—it could ultimately be quite profitable volumes of if its high-margin test continue to grow.

Outlook

Given that EXAS has cleared a number of hurdles: payer approval, positive feedback from providers, and a clinical efficacy advantage over the low-cost alternative (FIT) – ongoing adoption is a reasonable expectation.  EXAS is clearly planning on this -- ramping production capacity and expanding their salesforce. And given the number of deaths from CRC and the large number of colonoscopies currently performed, it’s reasonable to expect that the volumes for this test could be quite high.  The most addressable population, perhaps the ~40% of individuals who are currently (per the CDC) non-compliant with their colonoscopy protocol – which might be around 30mm people – would be prime candidates for this test.  Again, if EXAS were to reach 5mm in annual tests (which is 20x growth over current volumes), the profitability issues would be a distant memory (although I suspect that EXAS would be “bought out” before then.)

In conclusion, there is undeniably a wide range of outcomes, and there are numerous risks -- for example, new testing or clinical evidence could dramatically impair this investment.  But after investigating the “short” case, I think that the current market consensus is actually under-estimating the potential for this screen—and although competitors will surely emerge, the development time is slow (Cologuard has been in development since 2010).  

Therefore, if appropriately sized, EXAS could be a worthy long investment—whether at the current price, or ideally—if the recent volatility persists—after a pricing decline.   

Catalyst

Ongoing approval by payers, uptick in screening volumes, additional favorable clinical data, M&A speculation, or success with pipeline products.