Replies to post #160044 on Biotech Values

[DewDiligence]

EXAS—I suspect that some significant fraction of undetected polyps are just "low shedding".

Polyps grow over time and hence, in a series of tests, the shedding can be expected to increase, on average, from test to test; this strengthens the argument for cumulative sensitivity.

Moreover, some portion of ColoGuard’s “false positives” are presumably true positives of precancerous lesions that were missed by colonoscopy.

Notwithstanding the above, I would agree that the lack of detailed data on these issues is bearish for the existing version of the test.

…[other precancerous lesions] just shed "benign signature" or "signature not caught by test".

Conducting a series of tests clearly won’t help in these instances; however, the next version of ColoGuard (assuming there is one) can be expected to become smarter about which signatures are recognized as worrisome.

[DewDiligence]

EXAS—As per usual in biotech, if the company hasn't provided data, it is reasonable to speculate that the data isn't particularly good.

Following up on my reply in #msg-87008129: After listening to the EXAS CC last week, I’m pretty sure EXAS does not have any hard data on cumulative sensitivity (which, I suppose, is better than having bad data and not disclosing it). This passage from the CC is revealing:

http://seekingalpha.com/article/1351641-exact-sciences-ceo-discusses-preliminary-top-line-results-of-deep-c-pivotal-clinical-trial-transcript?part=single

…the risk of developing colorectal cancer multiplies with polyp size, but fewer than 40% of polyps continue to grow past 1 centimeter. Those that do grow expand slowly, doubling in 5 to 6 years. The slow growth of these polyps provides us with an ample window to detect them over several screening intervals, which creates a cumulative sensitivity of potentially 90%.

The key words in this passage are: i) several, which is deliberately vague in a context where the devil is in the details; and ii) potentially, from which I infer that no actual data exist on series of Cologuard tests.

I continue to think the concept of cumulative sensitivity has merit for Cologuard screening of the general population, but I have to wonder why EXAS has made no attempt to quantify it.