Epigenomics AG Vs. Exact Sciences: Who Wins The CRC Early Detection Fight? [View article]
I am grateful to pdb100 for his answer to your previous comment. This one is for me to pick up.
In 3 words: PRESEPT IS USELESS.
And here is why. Your correctly quote the Gut article published online Feb 13th, but this is only the abstract. The abstract obviously summarises the conclusion of the trial ..... as originally designed, ie based on dupicates. The standard method for PCR though is to use a triplicate. This was ultimatelly performed, but not as part of the formal trial anymore, but as a so-called ex-post analysis. Although not official, these figures are the ones used by the company and myself in my article and previous comments. That is why PRESEPT is useless. Remember the trial was completed 3 years ago, the triplicate has been used ever since, and the improved version 2 of EpiProColon has been introduced for over 2 years. As a last reminder, the US pivotal data was run on version 2 and I expect a late-breaker presentation at this year's DDW in May, followed by a full publication.
This is the link to the full article http://tinyurl.com/9wr... "post-hoc study of a third PCR reaction" is on page 3. The discussion section on page 7 goes on to say: "improvement in the test appear to be ossbile, as shown by our post hoc analysis using the three PCR replicate emulation which mimics the second generation commercially available mSEPT9 assay and which detects nearly two thirds of the cases, but yielded false positives in 12% of non-cases. The three replicate approach needs to be tested prospectively".
for a discussion about duplicates vs triplicates, have a look at http://bit.ly/YvMEHd This time you may learn something, because listening to the Exact call the other day the only thing I learnt was that they were running one month late (and do you know why ? REPRODUCIBILITY testing. Not a good start, but is it really a surprise ?) and that Exact Sciences will be the ONLY lab of record ! No big partnership, no incentive for other labs. How do think this will translate in commercial terms ? VERY SLOW + VERY EXPENSIVE.
Epigenomics AG Vs. Exact Sciences: Who Wins The CRC Early Detection Fight? [View article]
at last some skin in the game ! yesterday CEO Taapken flagged +20000 shares and today Chairman von Prondzynski (ex CEO Roche Diagnostics) +78000 shares from the rights issue at 1,58
Epigenomics AG Vs. Exact Sciences: Who Wins The CRC Early Detection Fight? [View article]
This is only your second comment in 5 years ! I'm flattered !
1) Fair point. I've being asking myself the same question for 2 years. I just couldn't get it, and I bugged management I don't know how many times before the answer slowly emerged. It's a combination of factors. The main one is that the PRESPET study is financed by Epigenomics AG but is controlled by a high-flying and fiercely independent steering committee (easy to verify). After completion they (the steering committee) decided to look at secondary end points on sub-groups which hadn't been pre-defined. It got really messy. Not exactlyy the neat type of paper you can put in every GP's hands as a complimentary NEJM reprint to use as a surrogate marketing tool. The editor is Dr T Church. And he is a busy man. He got new projects, new studies, he got less and less keen to work on yet another draft. And don't forget that PRESEPT was a prospective trial, but not the pivotal one for FDA clearance. Don't forget either that PRESEPT was completed with the first version of the test, and that the second version, again the one relevant for the FDA, was already available in 2011. PRESEPT could, should have been peer reviewed long ago, I agree 100%, but the delay has nothing to do with bad data (they were presented in detail in several conferences). Soon the US pivotal data are going to be published, together with the FIT head to head trial (completed in Dec 2012) and PRESEPT won't be needed anymore.
2) I gave the reasons why I doubt the scientific validity of the Cologuard vs mSeptin9 trial I have nothing to add. You call me an ostrich, fair enough, but asking why I believe the authors have cheated is ostrich-like to say the least.
As to your last point,- the dog, the baby and the fetish-, it made me smile, I must say, but also made you a full member of the "head in the sand" club. Welcome. Given a choice between a blood-based and a stool based test, the vast majority will go for the blood based test. Call them whatever you want you cannot change human nature. And while the Exact Sciences crowd keeps pointing at the superiority of the data, it simply isn't the point. Once the test is cleared for screening, everybody will forget about the performance. The box won't read : WARNING ! This test has a 71% CRC sensitivity only. And even if it did, have you ever seen a smoker quit smoking after he has read SMOKING KILLS on the box ?
Epigenomics AG Vs. Exact Sciences: Who Wins The CRC Early Detection Fight? [View article]
The Septin9 vs Cologuard head to head comparison !!! I was praying for somebody to bring it up. Thank you ! You seem te have a hard time "recalling" important information, so let's refresh your memory ! This is the article you are refering to http://tinyurl.com/bzm...
If you actually read the authors list you'll realise that it is Mayo and Exact Sciences. Compare that to the actual work done ? I mean the work done by the ARUP people on the septin9 arm ? Gone. Why ? They dropped out because of some serious methodological differences. That;s the reason why. This is Epigenomics open letter to the editor of the Journal http://tinyurl.com/bah...
Exact/Mayo (difficult to make the difference these days) have 1) performed the septin9 test on Cologuard positives only 2) the control arm had 8% other cancer prevalence and 3) probes came from uncontrolled and uncontrollable origin.
Thanks for pointing out the quality of this "leading journal", or is it just the fact that the Mayo clinic is an Exact Science shareholder and ARUP is not an Epigenomics shareholder ?
This is a serious issue though. The Mayo Clinic is maybe the most respected medical instituion in the world. (I have used a stamp depicting the Mayo brothers on my company's homepage, they are my heroes !). But the quasi symbiotic relationship between Dr Ahlquist and Exact Sciences, as well as this direct shareholding is a threat to its reputation.
Let me finish on a lighter note to comment on your comment on Epi's 5 years chart. Have you ever heard of the story of Bertrand Russel's chickens ? It is called the "benevolent farmer". It very simply explains how sometimes things leap from the impossible to the inevitable without transition (Exact's market cap ?), and I may add from the unescapable to the impossible (Epi's market cap ?).
Epigenomics AG Vs. Exact Sciences: Who Wins The CRC Early Detection Fight? [View article]
My point was to give as comprehensive a picture as possible and not to limit the discussion to the performance and forget about the other criteria critical to success in the marketplace. Having said that, let's take a closer look at some of your points.
EXAS "cut-off" study is exactly that: a cut-off study, not a prospective study, representative of an asymptomatic screening population. So don't jump to conclusion and let's wait for DEEP-C for a true assesment of Cologuard's performance.
I must contradict two of your statements: yes detailed data is available for EPC and yes Epigenomics has already completed a US prospective trial in an asymptomatic screening population. I mentioned early stage (phase 1+2) performance data for EPC, in both the case control and the prospective studies, let us split it further up (sensitivity) US pivotal Stg1 41% (7/17), Stg2 83% (10/12), pT1 33%, pT2 67% CE Mark Stg1 89% (24/27), Stg2 93% (27/29), pT1 na, pT2 89% H2H FIT Stg1 61% (14/23), Stg2 76% (12/16), pT1 58%, pT2 73%
It is very important to realise the difference between clinical staging (stage1, 2,3,4) and histological staging (pT1,...).. Usually pT1 cancers don't show up in case controls studies as the endoscopist has already cut them out during colonoscopy. As a rule of thumb pT1 represent 1/4 of all CRC cases and 2/3 of stage1. It is actually cancer but dealt with as any polyp or adenoma. No follow up treatment (chemo, adjuvant, radiotherapy etc) is necessary. So let me restate my point: as far as sensitivity is concerned EPC is good enough: as good as FIT and more importantly as good as required by the FDA. Let's move on to specificity. I have explained in principle why you should expect great variability of specificity according to the design of the study. (enrichment of control arm with "stuff"). But no study is as good as real life, right ? What about the 80 000 tests completed by Quests since early 2011 ? larger than 12 700, right ? Only 6.7% of all 80 000 turned out positive ! So mathematically specificity must be at least 93,3% ! Check it for yourself page 43 http://tinyurl.com/ay3....
So let's skip the compounding false positive nightmare scenario sketched out and move on to your last point.
Why is the current EUR 5m financing round done by the existing investors and not by Quest for instance ? Because they have something called a preferential right and because they can ! Because they'd rather be taken out at 800m rather than diluted a 20m.
Exact Sciences Could Be Exactly What Biotech Investors Need [View article]
if Exact really is going for a $550-$600 price tag (something GeneNews started with back in 2008, with a poor blood-based test), I believe they've come to realize that they have to maximize the value of a niche and leave the mass market to Epigenomics (blood-based) or FIT/FOBT (stool-based).
Because the real question is: HOW TO IMPROVE SCREENING COMPLIANCE ? By ticking a box on a form next time you visit your GP, or by having yet another version of a stool-based test ?
I can understand the emphasis on "pre-cancerous" detection, but polyps (hyperplastics) are not polyps (inflammatory), which are not polyps (adenomatous). Only adenomatous polyps (also called adenomas) 1cm or larger are considered pre-malignant ... and 90% of adenomas are smaller than 1cm. Finally, knowing that 50% of the people aged 60 carry at least 1 adenoma 1cm or larger, you must by now have figured out what I am hinting at: do we really want, can we really afford to send all these people to colonoscopy ? It becomes very clear that the debate has shifted from science to public health policy.
I stick to my "first things first" approach: get the screening compliance rate up in a cost-effective way and refer the positives to colonoscopy.
I wouldn't even exclude a scenario on a 5 years horizon by which colonoscopy loses its first line screening tag and be confined to 2nd line. And you don't even have to wait for guidelines changes to do the job. Convenience and market forces will tip the balance. GEs certainly will fight it. They should join it.
Exact Sciences Could Be Exactly What Biotech Investors Need [View article]
I have nothing to add to the Exact Sciences picture, but the competition one is superficial at best. To start with, OncoMethylome (Belgium) has been renamed MdxHealth and has dropped its own CRC screening test effort 18 months ago. Exact has licensed-in part of the IP, but that's it. But the elephant in the room is Epigenomics. First of all they don't have a real collaboration with Qiagen just yet, only an option to develop a commercial CRC screening test based on Epigenomics IP. And very conveniently that option expires Feb 28th 2013, around an expected PMA authorization date ! Filing is modular as in Exact's case, and has started in December 2011. Epigenomics has already completed 2 prospective studies (2x 7940), with the second one being the pivotal one. The performance data for this last trial (68% sensitivity, 80% specificity) has been a disappointment considering what had been established in the first pivotal trial (67% sens, 88% spec) and more recently in the various case control studies(sens ranging from 91% to 96% and spec from 85% to 87%). Even if one can say that the mix of cancer cases has been somewhat unfortunate, the FDA has requested a small head to head non-inferiority study against FIT. This study is about to enroll and should be completed sometime in summer. There is hardly any doubt that this is an academic exercise considering the emphasis put on sensitivity by the FDA. I find the Epigenomics vs Exact Sciences head to head confrontation more thrilling: blood based vs stool based, PMA expected 1Q13 vs 1Q14, commercial partners Abbott (global), Quest (US only but already marketing a LDT version since early 2011), Qiagen (maybe global if option exercised) vs ... blank, selling price between $100 CMS and $200 private vs $300 to $600 (by the way there is a CPT code for the Epi test for a couple of weeks now: 81401), market cap EUR 23 million (cash run rate to 2Q13) vs USD 525 million. Which one is the most likely 50 bagger do you think ?
Epigenomics AG Vs. Exact Sciences: Who Wins The CRC Early Detection Fight? [View article]
As for the rest, it is not worth commenting.
Epigenomics AG Vs. Exact Sciences: Who Wins The CRC Early Detection Fight? [View article]
In 3 words: PRESEPT IS USELESS.
And here is why.
Your correctly quote the Gut article published online Feb 13th, but this is only the abstract. The abstract obviously summarises the conclusion of the trial ..... as originally designed, ie based on dupicates. The standard method for PCR though is to use a triplicate. This was ultimatelly performed, but not as part of the formal trial anymore, but as a so-called ex-post analysis. Although not official, these figures are the ones used by the company and myself in my article and previous comments.
That is why PRESEPT is useless. Remember the trial was completed 3 years ago, the triplicate has been used ever since, and the improved version 2 of EpiProColon has been introduced for over 2 years. As a last reminder, the US pivotal data was run on version 2 and I expect a late-breaker presentation at this year's DDW in May, followed by a full publication.
This is the link to the full article http://tinyurl.com/9wr...
"post-hoc study of a third PCR reaction" is on page 3.
The discussion section on page 7 goes on to say: "improvement in the test appear to be ossbile, as shown by our post hoc analysis using the three PCR replicate emulation which mimics the second generation commercially available mSEPT9 assay and which detects nearly two thirds of the cases, but yielded false positives in 12% of non-cases. The three replicate approach needs to be tested prospectively".
for a discussion about duplicates vs triplicates, have a look at
http://bit.ly/YvMEHd
This time you may learn something, because listening to the Exact call the other day the only thing I learnt was that they were running one month late (and do you know why ? REPRODUCIBILITY testing. Not a good start, but is it really a surprise ?) and that Exact Sciences will be the ONLY lab of record ! No big partnership, no incentive for other labs. How do think this will translate in commercial terms ? VERY SLOW + VERY EXPENSIVE.
Game over ? Really ? for Epi or Exact ?
Epigenomics AG Vs. Exact Sciences: Who Wins The CRC Early Detection Fight? [View article]
yesterday CEO Taapken flagged +20000 shares and today Chairman von Prondzynski (ex CEO Roche Diagnostics) +78000 shares from the rights issue at 1,58
http://bit.ly/XsflpY
That's another box ticked off on my wish list (and any potential new investor I imagine)
Epigenomics AG Vs. Exact Sciences: Who Wins The CRC Early Detection Fight? [View article]
1) Fair point. I've being asking myself the same question for 2 years. I just couldn't get it, and I bugged management I don't know how many times before the answer slowly emerged. It's a combination of factors. The main one is that the PRESPET study is financed by Epigenomics AG but is controlled by a high-flying and fiercely independent steering committee (easy to verify). After completion they (the steering committee) decided to look at secondary end points on sub-groups which hadn't been pre-defined. It got really messy. Not exactlyy the neat type of paper you can put in every GP's hands as a complimentary NEJM reprint to use as a surrogate marketing tool. The editor is Dr T Church. And he is a busy man. He got new projects, new studies, he got less and less keen to work on yet another draft. And don't forget that PRESEPT was a prospective trial, but not the pivotal one for FDA clearance. Don't forget either that PRESEPT was completed with the first version of the test, and that the second version, again the one relevant for the FDA, was already available in 2011. PRESEPT could, should have been peer reviewed long ago, I agree 100%, but the delay has nothing to do with bad data (they were presented in detail in several conferences). Soon the US pivotal data are going to be published, together with the FIT head to head trial (completed in Dec 2012) and PRESEPT won't be needed anymore.
2) I gave the reasons why I doubt the scientific validity of the Cologuard vs mSeptin9 trial I have nothing to add. You call me an ostrich, fair enough, but asking why I believe the authors have cheated is ostrich-like to say the least.
As to your last point,- the dog, the baby and the fetish-, it made me smile, I must say, but also made you a full member of the "head in the sand" club. Welcome. Given a choice between a blood-based and a stool based test, the vast majority will go for the blood based test. Call them whatever you want you cannot change human nature. And while the Exact Sciences crowd keeps pointing at the superiority of the data, it simply isn't the point. Once the test is cleared for screening, everybody will forget about the performance. The box won't read : WARNING ! This test has a 71% CRC sensitivity only. And even if it did, have you ever seen a smoker quit smoking after he has read SMOKING KILLS on the box ?
Epigenomics AG Vs. Exact Sciences: Who Wins The CRC Early Detection Fight? [View article]
I cant get no satisfaction out of it, could you please elaborate ?
Epigenomics AG Vs. Exact Sciences: Who Wins The CRC Early Detection Fight? [View article]
I was praying for somebody to bring it up. Thank you !
You seem te have a hard time "recalling" important information, so let's refresh your memory !
This is the article you are refering to http://tinyurl.com/bzm...
If you actually read the authors list you'll realise that it is Mayo and Exact Sciences. Compare that to the actual work done ? I mean the work done by the ARUP people on the septin9 arm ? Gone. Why ? They dropped out because of some serious methodological differences. That;s the reason why.
This is Epigenomics open letter to the editor of the Journal
http://tinyurl.com/bah...
Exact/Mayo (difficult to make the difference these days) have 1) performed the septin9 test on Cologuard positives only 2) the control arm had 8% other cancer prevalence and 3) probes came from uncontrolled and uncontrollable origin.
Thanks for pointing out the quality of this "leading journal", or is it just the fact that the Mayo clinic is an Exact Science shareholder and ARUP is not an Epigenomics shareholder ?
This is a serious issue though. The Mayo Clinic is maybe the most respected medical instituion in the world. (I have used a stamp depicting the Mayo brothers on my company's homepage, they are my heroes !). But the quasi symbiotic relationship between Dr Ahlquist and Exact Sciences, as well as this direct shareholding is a threat to its reputation.
Let me finish on a lighter note to comment on your comment on Epi's 5 years chart. Have you ever heard of the story of Bertrand Russel's chickens ? It is called the "benevolent farmer". It very simply explains how sometimes things leap from the impossible to the inevitable without transition (Exact's market cap ?), and I may add from the unescapable to the impossible (Epi's market cap ?).
Epigenomics AG Vs. Exact Sciences: Who Wins The CRC Early Detection Fight? [View article]
EXAS "cut-off" study is exactly that: a cut-off study, not a prospective study, representative of an asymptomatic screening population. So don't jump to conclusion and let's wait for DEEP-C for a true assesment of Cologuard's performance.
I must contradict two of your statements: yes detailed data is available for EPC and yes Epigenomics has already completed a US prospective trial in an asymptomatic screening population.
I mentioned early stage (phase 1+2) performance data for EPC, in both the case control and the prospective studies, let us split it further up (sensitivity)
US pivotal Stg1 41% (7/17), Stg2 83% (10/12), pT1 33%, pT2 67%
CE Mark Stg1 89% (24/27), Stg2 93% (27/29), pT1 na, pT2 89%
H2H FIT Stg1 61% (14/23), Stg2 76% (12/16), pT1 58%, pT2 73%
It is very important to realise the difference between clinical staging (stage1, 2,3,4) and histological staging (pT1,...).. Usually pT1 cancers don't show up in case controls studies as the endoscopist has already cut them out during colonoscopy. As a rule of thumb pT1 represent 1/4 of all CRC cases and 2/3 of stage1. It is actually cancer but dealt with as any polyp or adenoma. No follow up treatment (chemo, adjuvant, radiotherapy etc) is necessary.
So let me restate my point: as far as sensitivity is concerned EPC is good enough: as good as FIT and more importantly as good as required by the FDA. Let's move on to specificity.
I have explained in principle why you should expect great variability of specificity according to the design of the study. (enrichment of control arm with "stuff"). But no study is as good as real life, right ? What about the 80 000 tests completed by Quests since early 2011 ? larger than 12 700, right ? Only 6.7% of all 80 000 turned out positive ! So mathematically specificity must be at least 93,3% ! Check it for yourself page 43 http://tinyurl.com/ay3....
So let's skip the compounding false positive nightmare scenario sketched out and move on to your last point.
Why is the current EUR 5m financing round done by the existing investors and not by Quest for instance ? Because they have something called a preferential right and because they can ! Because they'd rather be taken out at 800m rather than diluted a 20m.
Exact Sciences Could Be Exactly What Biotech Investors Need [View article]
Because the real question is: HOW TO IMPROVE SCREENING COMPLIANCE ? By ticking a box on a form next time you visit your GP, or by having yet another version of a stool-based test ?
I can understand the emphasis on "pre-cancerous" detection, but polyps (hyperplastics) are not polyps (inflammatory), which are not polyps (adenomatous). Only adenomatous polyps (also called adenomas) 1cm or larger are considered pre-malignant ... and 90% of adenomas are smaller than 1cm. Finally, knowing that 50% of the people aged 60 carry at least 1 adenoma 1cm or larger, you must by now have figured out what I am hinting at: do we really want, can we really afford to send all these people to colonoscopy ? It becomes very clear that the debate has shifted from science to public health policy.
I stick to my "first things first" approach: get the screening compliance rate up in a cost-effective way and refer the positives to colonoscopy.
I wouldn't even exclude a scenario on a 5 years horizon by which colonoscopy loses its first line screening tag and be confined to 2nd line. And you don't even have to wait for guidelines changes to do the job. Convenience and market forces will tip the balance. GEs certainly will fight it. They should join it.
Exact Sciences Could Be Exactly What Biotech Investors Need [View article]
To start with, OncoMethylome (Belgium) has been renamed MdxHealth and has dropped its own CRC screening test effort 18 months ago. Exact has licensed-in part of the IP, but that's it.
But the elephant in the room is Epigenomics. First of all they don't have a real collaboration with Qiagen just yet, only an option to develop a commercial CRC screening test based on Epigenomics IP. And very conveniently that option expires Feb 28th 2013, around an expected PMA authorization date ! Filing is modular as in Exact's case, and has started in December 2011. Epigenomics has already completed 2 prospective studies (2x 7940), with the second one being the pivotal one. The performance data for this last trial (68% sensitivity, 80% specificity) has been a disappointment considering what had been established in the first pivotal trial (67% sens, 88% spec) and more recently in the various case control studies(sens ranging from 91% to 96% and spec from 85% to 87%). Even if one can say that the mix of cancer cases has been somewhat unfortunate, the FDA has requested a small head to head non-inferiority study against FIT. This study is about to enroll and should be completed sometime in summer. There is hardly any doubt that this is an academic exercise considering the emphasis put on sensitivity by the FDA.
I find the Epigenomics vs Exact Sciences head to head confrontation more thrilling: blood based vs stool based, PMA expected 1Q13 vs 1Q14, commercial partners Abbott (global), Quest (US only but already marketing a LDT version since early 2011), Qiagen (maybe global if option exercised) vs ... blank, selling price between $100 CMS and $200 private vs $300 to $600 (by the way there is a CPT code for the Epi test for a couple of weeks now: 81401), market cap EUR 23 million (cash run rate to 2Q13) vs USD 525 million. Which one is the most likely 50 bagger do you think ?