Nick Zheng's  Instablog

Nick Zheng
Send Message
Biotech stock trading and investing for 25 years. PhD in biology, MS in computer science. Extensive experiences in biotech/pharmaceutical companies, in both USA and China, serving middle to senior executive positions. After 12 years of research at top institutions/universities and 10 years of... More
  • Why Cologuard Of Exact Sciences (Ticker: EXAS) May Be No Better Than Routine Fecal Occult Blood Test (FOBT) 3 comments
    Apr 21, 2013 4:13 AM

    After waiting for much longer time than expected, top-line data of pivotal trial of Cologuard for colorectal cancer screening (Deep-C) was finally announced early morning of April 18. The data were short of expectation and stock dropped more than 30% to under 7 at pre-hour trading.

    Technologically, Cologuard is a sDNA (stool DNA) test on top of well established routine FOBT test. EXAS uses multiple markers found in stool DNA in combination of a FOBT test to determine the likelihood of colorectal cancer and advanced adenoma (pre-cancerous tumor that may develop into cancer in later years) using a complicated algorithm. In the recent presentation by the company, slide 16 states Cologuard requires 2 multiplex DNA assays and 1 ELISA essay for fecal hemoglobin. But cross examination of top-line data seems to suggest that Cologuard is not better than routine fecal occult blood test (FOBT).

    Let's take a look at the limited data for cancer and advanced adenoma, respectively.

    1. Primary target for cancer detection is only partially achieved.

    The primary performance is targeted for >=85% cancer sensitivity and >=90% specificity, while Deep-C top-line data demonstrated 92% cancer sensitivity and 87% specificity.


    Primary Target


    Target Met?









    Specificity tells the accuracy of the positive tests. Lower specificity means that more false positive cases have to be confirmed by invasive colonoscopy.

    1. Cancer detection is not better than routine FOBT.

    In 2007, National Cancer Institute (NYSE:NCI) published a paper on improved FOBT tests that were later adopted as routine simple screening for colorectal neoplasm. NCI also accompanies that paper with a bulletin.

    Following is the table comparing the performance of cancer detection by Cologuard and FOBT methods.





    Better than FOBT?











    In the clinical trial involving 5841 subjects, NCI used two FOBT methods. One is unrehydrated guaiac fecal occult blood test, and another is called fecal immunochemical test (FIT). Both use ELISA essays to detect hemoglobin (occult blood) in stool.

    While Cologuard detects more cancer cases (higher sensitivity) than FOBT, but it is at expense of accuracy (specificity). The results may be expected considering that Cologuard is DNA essays in addition to FIT. The additional DNA assays appear to pull out more cancer cases, but also many more false positive cases who will have to go through unnecessary colonoscopy to confirm. The colonoscopy procedure is costly (about $2500) and uncomfortable. In another words, the additional stool DNA assays have performed much worse than FOBT if we could separate the additional cancer cases and analyze them independently.

    1. Detection of advanced adenoma is worse than routine FOBT.

    Management claims that the key differentiating performance of Cologuard is its significantly higher ability to detect pre-cancerous (advanced adenoma) in comparison to currently available non-invasive methods. Let's compare Cologuard and FOBT performance for advanced adnoma in the following table.

    Advanced Adenoma




    Better Than FOBT?











    EXAS did not release detailed statistical analysis of its top-line data, but it is clear from above table that Cologuard is no better than GT in term of sensitivity, and worse than either FOBT method (FIT or GT) in term of specificity.

    One secondary endpoint of Deep-C trial is to demonstrate that Cologuard is superior to FIT for advanced adenoma detection. While the press release did not provide any data, this comparison analysis shows Cologuard performs no better than FOBT overally, and worse than sensitive GT method.

    In summary, I believe Cologuard did not perform better than routine FOBT that costs only $15-35 per test. The commercial success will rely on its ability to detect pre-cancer better, but Deep-C shows otherwise. Not only is its sensitivity at 42% lower than earlier trials (>57%), but no better than simple FOBT. For cancer detection, Cologuard achieved higher sensitivity at cost of lower accuracy. The much hyper generated for Cologuard may end as a commercial flop even FDA approves it next year. I also think that the advisory committee and FDA will scrutinize Cologuard performance and question the real benefit on background of many competing screening methods. Will men over 50 love to poo over the white plastic bowl, cover it, and mail it back to EXAS as the CEO displayed the stool collecting tool at CNBC right after the conference call on the top-line data announcement? The cost of FedEx will be higher than a simple routine FOBT test.

    Disclosure: I am short EXAS. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.

Back To Nick Zheng's Instablog HomePage »

Instablogs are blogs which are instantly set up and networked within the Seeking Alpha community. Instablog posts are not selected, edited or screened by Seeking Alpha editors, in contrast to contributors' articles.

Comments (3)
Track new comments
  • stockmannick
    , contributor
    Comments (14) | Send Message


    the FOBT study u cite from 2007, do u have the sensitivity #'s for adenomas over 2cm and 3 cm by chance?


    The only saving grace I see for Cologuard is in the larger precancer adenomas. (In the topline results, 66% of 2cm or greater precancers were detected by Cologuard) and the CEO mentioned the 3cm variety (high grade dysplasia) had a slightly higher sensitivity than the 2cm group.


    I recall CEO Conroy mentioned they beat the FIT test that was used in Deep-C by a wide margin in the Secondary endpoint, but hard to speculate just how wide and at what specificity the FIT test came in at for this trial.


    As for now Im disappointed with the topline results like many were, but until we get all the stratified data, especially compared to the FIT component, I'll reserve final judgement.
    21 Apr 2013, 05:46 PM Reply Like
  • mdropp
    , contributor
    Comments (6) | Send Message
    The study you site:
    "In a large prospective study performed by investigators from three Northern California Kaiser Permanente medical centers, a type of fecal occult blood test (FOBT) called a fecal immunochemical test (FIT) showed high sensitivity and specificity for detecting left-sided colorectal cancer. "


    What about right-sided colorectal cancers? Wouldn't you want to find those also?
    30 Apr 2013, 07:30 PM Reply Like
  • Energy-Trader
    , contributor
    Comments (1051) | Send Message
    Yet another guy from the blogosphere that is learning an "expensive" lesson given the terrible knowledge base that they are operating under. EXAS traded up thru $11.00 briefly last Friday.


    The "market" seems to understand that Cologuard is vastly SUPERIOR to the traditional FIT and FOBT screening tests that you have cited. Unfortunately, you don't have a very good understanding of the data, let alone how pre-cancers develop into full blown cancer.


    Why is it that you fail to acknowledge the Sensitivity by Cologuard for High Grade Dysplasia (HGD) at several percentage points north of 66%?


    High Grade Dysplasia are pre-cancers that are right on the verge of turning into full blown cancer.


    According to the Morikawa study from the August 2005 issue of Gastroenterology and the Imperiale study found in the NEJM from December of 2004, the Sensitivity for pre-cancerous polyps by FIT and FOBT are ridiculously low.


    In fact, FIT is only 20% sensitive to detecting pre-cancers while FOBT is only 13% sensitive.


    Moreover, they also pale in comparison to Cologuard when it comes to detecting High Grade Dysplasia at 33% to 66% for HGD by Cologuard.


    It's not even close.
    22 May 2013, 11:14 AM Reply Like
Full index of posts »
Latest Followers


More »
Instablogs are Seeking Alpha's free blogging platform customized for finance, with instant set up and exposure to millions of readers interested in the financial markets. Publish your own instablog in minutes.