This would be perfectly valid because it provides the important efficacy information for the 75% or so of the patients who are not candidates for post-continuation therapy.
Now the above is a quick analysis on my part and further reflection might change things. However, it illustrates to me that our knowledge as investors is limited (both known unknowns and unknown unknowns) and we don't know all the options. The ONCY team on the other hand has much more complete information and the expertise to evaluate trial design options. As Brad mentions as well, the regulators are very helpful/collaborative in the trial design process. R/M SCCHN continues to require better therapies AND more cost effective therapies (than certuximab/ERBITUX).
I think they will figure it out. That actually was the strongest message I got from the conference call - in tone and content. They have strong signals of efficacy in this loco-regional group. They know where the confounding issues are. They are confident they can work out a trial protocol (with the regulators' help and blessing) that can demonstrate statistical significance. Brad's guesstimate of 250 patients in a registration trial is a pretty strong indicator of his confidence in demonstrating efficacy (especially if those numbers have been bumped up 30% to account for patients that go on to post-continuation therapy).
And we haven't even talked about the mets-only results yet.
From the conference call Brad and Matt emphasized that the way to deal with the confounding issues (the fever issue and the post-discontinuation therapy issue) was to remove the blind (so doctors wouldn't over-react to patients in the test arm with fevers) and to design the registration trial so that the post-continuation therapy patients would be segregated into a separate group (The trial has to include this segregation in its protocol - it cannot be an after-the-fact add-on).
Here are (my off the top of my head) reasons why I think it could work:
If the trial was unblinded, yes, there is the risk of some control patients being removed early and moved onto another therapy. However I don't think this number would be massive because most patients don't have other options. In the recently released loco-regional data only 30 of the 118 patients (25%) went on to post-continuation therapy. (Brad said about 1/3 of the control patients and 1/10 of the test patients went on to post-continuation therapy - I'm not sure, but I think here he is referring to the patients who progressed at the six-week scan). Anyway, the point is most patients (75%?) don't have other options because they have exhausted other options - like certuximab which many would have already received as first-line therapy (they are now resistant to it or have developed toxicities), or they are simply too sick. Another problem is that in Europe many/most/all (?) countries won't pay for certuximab in R/M SCCHN because it is too expensive given the limited time of quality life it offers. Therefore, ONCY could do something like adjust the enrolment number 25-30% higher to account for patients that would potentially move on to post-continuation therapy.
By segregating the post-discontinuation therapy group (in the trial design so it is legitimate statistically) they could then legitimately remove this group from the OS stats when doing the comparison between treatment and control. (to be continued)
d2 - We don't know the details of the fever issue or why the degree of the problem was not foreseen. I'm not prepared to jump from a position of not really knowing what went on or what constraints ONCY was under to accusing ONC/Y of poor execution of the trial. The design of the trial was I suspect the fruit of a lot of input - Matt, Brad, Principal Investigators, opinion leaders, regulators, independent statisticians, etc. These are smart and motivated people. I give them the benefit of the doubt unless it can be clearly shown they made mistakes they shouldn't have. I do take seriously that cutting edge research is a venture into the unknown - not everything can be anticipated.
One big difference between this trial and previous trials is that it was double blinded. None of the other trials with REOLYSIN were even single-blinded. ONCY may have known that there was a potential problem with fevers being interpreted as possible sepsis. They may even have informed doctors of this possibility, but they had never experienced how blinded doctors would react (or blinded patients - maybe patients were the ones who insisted on being taken off the trial). Maybe they underestimated how many blinded doctors would react to uncertainty by taking the cautious approach and removing patients from the trial. Maybe the FDA or EMA underestimated the problem and insisted the trial be double blinded. We don't know. I prefer to say "I don't know" until I do know. Until then ONCY gets the benefit of the doubt.
Thanks, casiturri, for chiming in. I agree. Brad and Matt are dealing with realities and making judgements and adjustments as needed. The market seems to prefer simplistic answers, even if they don't make any sense. So the share price has been punished because the market didn't get the simple answers it wanted. However, ONCY has learned a great deal that will make ALL of its subsequent trials better. So, it's understandable to be disappointed by delay, but progress has been made and ultimately ONCY will be stronger for it.
Nice to see the share price perk up today. More is to be learned when a complete analysis is done of the loco-regional data. The mets-only data, while almost certainly too small to give stat sig results and possibly with some of the same complications as the loco-regional data, will hopefully still give a very strong signal of REOLYSIN's effectiveness against mets. The lung results were very good. Anecdotally, we are hearing good things about REO's effectiveness in some of the phase 2 trials and hopefully soon we'll get some interim data.
D2, I believe the longer we wait the stronger the likelihood of powerful (potentially even statsig) results. Maybe that threshold has already been reached. However, we won't know (and Brad won't know) until the data is unblinded. So there is reason for hope, but nothing is certain until the data is unblinded.
D2, the email is still on the V2 board (message #16968). Here is Brad's email to spdstr56:
"There is a time relationship to long term survivors that supersedes the event number.
The time from un blinding to announcement is highly variable based on how complicated it is!
We don't have to consult with the FDA prior to announcement. FYI our FDA reviewers are working from home, unpaid right now. A very dedicated group of people.
I don't take Brad's statement, "The time from un blinding to announcement is highly variable based on how complicated it is!", as a hint in any direction. They are blinded so they don't know how simple or complicated it will be. Brad just wants people to be aware that the length of time can vary.
It seems to me that all of Brad's comments over the past months have been cautious and designed to give "space" regardless of whether or not statsig is achieved in part 1. As CEO of a company in a important phase 3 double blinded trial I think he has to be cautious in his pronouncements. This is the time to wait for the data to speak.
"In fact I'll just spill all the beans. There are 11 pts who are a) alive b) have been on study between 4 months and less than a year. Apparently a couple will have reached the year threshold (or died) by the time of the the conference. It is simple math that if all 11 die before reaching one year, Brads statement will be wrong. In fact it will be wrong if only 6 of the 11 die before reaching the one year threshold-- This is simple math." - rationalthought11
Hmmm. Who to trust - Brad or rationalthought11???!!! I don't think that's a hard choice.
As far as rationalthought11's simple math goes consider this:
1. On November 9th 2012 (almost a year ago) there were already 33 patients enrolled in the lung study (from the poster available in the Presentations section of the ONCY website).
2. That means that as of the Oct. 1, 2013 press release likely 31-33 patients of the 36 evaluable had enough follow up to calculate the one year survival of 53%. So with at least 31 of the 36 patients already with enough follow up that gives a pretty solid basis for the 53% number.
3. Now if it is known that the majority of the remaining 5 or less patients are doing well (remember this is an open label trial) Brad has good grounds to expect the one year survival to go up.
I think I'll stick with Brad's assertion that, "The 53% should be taken as a minimum %".
That's it for now. I will be blissfully away in the wilderness for the next two days.
It feels to me that a new randomized phase 2 trial in first-line non-small cell lung cancer is imminent, and I wouldn't be surprised if it were centred in the US and sponsored by the NCI. Also, remember, if REOLYSIN is approved in any indication (like H&N) it can be used off-label for other indications if supported by a well constructed randomized phase 2 trial. (Dr. George Gill had a lot to say about this at the 2013 AGM in Calgary). In other jurisdictions (like Europe) a well constructed randomized phase 2 alone can be sufficient for limited approval.
As we know, nsclc is the BIGGEST cancer indication worldwide. REOLYSIN used on first-line patients would be an enormous development. Big Pharma is paying attention.
I had a couple of very brief email exchanges with Brad yesterday.
1. My Question: How many of the 36 evaluable patients on enrolment were stage III and how many were stage IV? My recollection is that the vast majority were stage 4.
Brad's Answer: "Most were IVb's, and most of the rest IVa's.
And comparisons to other agents should wait until people see the correlation with specific gene markers at the end of the month!"
My comment: Notice Brad's exclamation mark at the end of the sentence. He obviously thinks the gene markers are a big deal and that important news in that regard will come out at the end of the month (most likely at the Austrailian Lung Conference).
2. My Question: Could you clarify this statement in the news release for me: "Of the 36 evaluable patients with sufficient follow up to date, progression free survival (PFS) at six months is 36% and one-year survival, 53%." At first I thought this meant that all 36 patients had sufficient follow up to demonstrate one-year survival of 53%, but now I see that this more naturally reads that not all of the 36 patients have sufficient follow up, but of those that do have sufficient follow up 53% have survived at least one year.
Is this reading correct?
Brad's Answer: "We had to stick to the content of the text in the abstract very closely, and could not reword that. There are patients still alive that may have or will have passed the one year mark by the date of the conference. The 53% should be taken as a minimum %."
My Comment: The one year OS in the lung trial is evolving and according to Brad it is evolving upward. As Nome said on the other board, such results in stage IV lung cancer patients (the majority of whom are actually stage IVb) is remarkable. Also, Brad again refers to the conference in Australia at the end of the month. This is a conference we will need to pay attention to. The new poster by Dr. Villalona-Calero could be very (positively) potent.
I misspoke. The Oct. 1 news release makes clear that the conference where the nsclc data will be presented at is the International Association for the Study of Lung Cancer (IASLC) World Conference on Lung Cancer held from October 27th to 30th, 2013 in Sydney, Australia. An updated comparison of genetic markers and survival outcomes would be most interesting.
Brad was quite excited by slide 13 at this year's AGM. It was titled: REO 016: REOLYSIN Efficacy and Genotype. It contains the same information as the poster by Dr. Villalona-Calero at the EORTC-NCI-AACR Symposium last November (referred to in my previous post). The slide compares the mutation of each of the 33 patients with their BEST Response (by RECIST). That is a measure of tumour shrinkage (an indication of efficacy), but the gold standard as we all know is OVERALL SURVIVAL. Now Dr. Villalona-Calero is in a position to compare a patient's genetic marker with their response to REOLYSIN in terms of actual SURVIVAL. This could be very powerful and it may give ONCY the ability to target patients who have the best opportunity for prolonged survival when treated with REOLYSIN. I wonder if something like that is going to be presented at this year's EORTC-NCI-AACR Symposium from October 19th to 23rd?
If one goes to the For Investors == Presentations section of the ONCY website one will find the following poster presentation: 2012-11-09 Dr. Villalona-Calero EORTC-NCI-AACR Symposium on Molecular Targets and Cancer Therapeutics Nov 2012 (Lung Cancer)
If you look at the poster you will see that at the time (November 2012) the patient number is given as 33 of which 13 are stage IVa and 20 are stage IVb (not a stage 3 to be seen). These are not loosey goosey numbers.
From the 2013 AGM transcript:
The other one that’s interesting, on those Reo studies, I mean the really key thing I think for this first one is looking at the stable disease or better in this relatively diverse patient population. The entrance criteria they were all supposed to be stage 3B’s or 4’s, which is very late stage, it turned out all the patients, except for one I think, were stage 4 and in this patient population you don’t expect to see a 90% or better stable disease rate by RECIST which is what we got out of the study It's probably, arguably about twice, what you would expect to see with this particular patient population, these are very late stage patients and we saw a very, very aggressive if you want to think of clinical benefit as being stable disease or better.
The 53% one year survival in these stage IV patients is remarkable. It's great to get confirmation that stable disease seems to lead to increased survival. Here is the pertinent section from the news release: "Response evaluation for 36 evaluable patients showed 11 partial responses (PR) (30%) (EGFR amplified, five; BRAF two; Kras, three; EGFR mutated, one), 21 stable disease (SD), and four progressive disease (PD). Of the 36 evaluable patients with sufficient follow up to date, profession free survival (PFS) at six months is 36% and one-year survival, 53%."
Notice also that the 53% survival at one year includes only "evaluable patients with sufficient follow up to date". That means this remarkable one year survival number may yet get bigger.
Did you notice that 20 of the 36 evaluable patients were Kras? Tarceva is ineffectual against Kras patients. Here is a quotation from Wikipedia (you can find it under KRAS, subheading Lung Cancer): "Lung cancer patients who are positive for KRAS mutation (and the EGFR status would be wild type) have a low response rate to Tarceva or Iressa estimated at 5% or less." Therefore, the suggestion that Tarceva is comparable or better is plain wrong. ONCY specifically designed this trial to include KRAS patients because besides understanding that KRAS is one of REO's preferred targets they also knew that other therapies (like Tarceva) are ineffective against Kras mutated cancers.
ONCY helpfully provided references to these claims in this News Release which we are free to check out:
1 As reported in The New England Journal of Medicine (N Engl J Med, Vol. 346, No. 2, January 10, 2002) by Schiller, et al, a total of 1,207 patients with advanced non-small-cell lung cancer were randomly assigned to a reference regimen of cisplatin and paclitaxel or to one of three experimental regimens: cisplatin and gemcitabine, cisplatin and docetaxel, or carboplatin and paclitaxel. The response rate for all 1,155 eligible patients was 19 percent, with a median survival of 7.9 months (95 percent confidence interval, 7.3 to 8.5), a 1-year survival rate of 33 percent (95 percent confidence interval, 30 to 36 percent).
2 As reported in Clinical Epidemiology (2011:3 139-148), Cetin, et al, using data from the Surveillance, Epidemiology and End Results (SEER) Program, stratified 51,749 incident stage IV NSCLC patients (1988-2003 with follow-up through 2006) by major histologic subtype. Overall one-year survival (95% confidence interval) in stage IV non-small cell lung cancer patients who survived at least 31 days (n=44,172) was 15.9%.
"Dr. Thompson is also presenting at the WDM 1st Global Life Sciences Conference
on Monday, September 16, 2013 at 10:30 a.m. CEST. The conference is taking place
on September 16th and 17th in Warsaw, Poland."
10:30 a.m. CEST (Central European Summer Time) is equal to 4:30 a.m. Eastern
Daylight Time or 1:30 a.m. Pacific Daylight Time. So when you get up tomorrow
morning you might be able to listen to the recording of the webcast. If you go to the Presentations page on the Oncolytics Biotech website it will give you a link to the webcast.
I believe you can listen to the webcast live. I'm not sure when the archived
webcast will be available - possibly soon after the live webcast, but possibly
not until the conference is over.