Friday, February 28, 2014

ASH 2013: Dr Jeff Sharman Discusses the Ideal Design of Phase 3 Trials with Novel Agent versus the Reality in the Community

As a patient, it is easy to be critical of study design.

Case in point: The phase 3 trial of rituximab (R) plus idelalisib (I) versus rituximab plus placebo presented at ASH 2013 in New Orleans.

Wait a minute, we patients scream. One arm of this trial is offering us a monotherapy option that is not recommended by most CLL experts because it produces a lousy response rate in the middle single digits. In the trial, lymph node response rate was a meager 4% for the R+ placebo versus 93% for the arm with the two active agents. And the duration of the response for the R alone is less that six months.

In comparison, in the R+I arm, median progression free survival has not been reached, and remarkably, even overall survival was improved, something unheard of in CLL just a few years ago when FCR was shown to be the first drug combination to add years to our lives.

All the data from the late breaking abstract can be accessed here. It is powerful, but one must ask, does idelalisib look so good because it was being compared to a straw man?

Is this an ethical way to conduct a trial?

Dr. Jeff Sharman, a regular on this web site, is both a community based oncologist and the medical director of hematology research for The US Oncology Network.

He points the pros and cons of the controversial trial design.

Watch the video and form your own opinion before I give you my bottom line.

First Dr. Sharman does a good job reminding us the difference between a phase 1, 2 and 3 trial.

So what do you think?

At first,  I was dismayed by the trial, from a patient's perspective, but when we have a trial design where we are pretty much guaranteed that we will get idelalisib if we need it, and in either arm we  can avoid chemotherapy, on further reflection I am more sanguine about this kind of trial.

The science bothers me more than the ethics. In the study, the results with idelalisib are very impressive and would likely have stood up to a much more active comparator arm, but in this case it was a massive mismatch as evidenced by the unheard of gargantuan P number (p = 1.3 x 10-30 ). In other words, the odds of getting this result by chance alone was not one in a million or one in a billion or one in a trillion, but about one in 10 followed by 30 zeros.

So what is my take away message?

First, I am OK with most trials that have a cross-over built in from the get-go, and especially fond of those that offer us a chance to avoid chemotherapy.

Second, idelalisib is clearly going to be a powerful addition to our armamentarium for treating CLL. 
In this study it took on a tough bunch of patients.The median number of prior therapies was three. The test subjects were all considered too unhealthy to get traditional chemotherapy. Nearly half were 17p deleted and 17 out of 20 were unmutated. We are told that all the risk groups responded better to the idelalisib arm, but the abstracts doesn't give us much detail.

So, bottom line, unless my need to get my disease under control was so acute that I could not risk the delay in getting a response if I was randomized to the R+ placebo arm, I wouldn't hesitate to enroll in a similar trial if I needed therapy.

Based on this and other positive studies, I hope and suspect idelalisib will be approved for CLL in the USA before the end of the year. And that will be a good thing for the CLL community.

Labels: , , , , , , , , , , ,


Anonymous Anonymous said...

Awesome to get more posts Brian! Miss them dearly and this was a great one.

The more I think about it the more complicated I realize these studies are...

Morally - If they pick so a low standard of care like Rituxan monotherapy... a virtual placebo and a lame duck with Gayzva taking its spot... it is good to know that the crossover was built-in from the start.

Productivity-wise - How productive is it to pick such low hanging fruit as a control arm? I was trying to think of the best therapy to compare idelalisib against when the study began... and I briefly supposed it was FCR. But so many of our relapse refractory CLL folks who can enroll in these studies are poor candidates for chemo based therapies like was the case here. Plus, if FCR were the comparison, I would guess the time to approval would be much much longer as it would take longer for idelalisib to prove its relative advantage if the cohort wasn't so filled with pretreated chemo and 17p deleted patients.

Its probably going to get more difficult for them to pick the control arm as more therapies get approved leading to more personalized cocktails depending on the CLL flavor someone has contracted. Whats the standard of care then? Going to the lowest common denominator like a monoclonal antibody like Gayzva with a crossover stipulation to the trial drug makes a lot of sense to me in that framework.

So I am especially grateful for the post Brian... because in my rash reading of these trial designs, I usually get angry... but by making me think harder about it... you have made me see that they are probably designing the studies closer to right than wrong, although they can never be perfect.

February 28, 2014 at 4:01 PM  
Blogger Terry's CLL Journey said...

As I have experienced, the lack of a crossover is one huge reason for NOT entering into a trial. Especially if you are comparing the drug to a known drug that has little lasting effect. Aren't the Trial patients in Europe a much higher number than in the US?

March 1, 2014 at 1:22 PM  

Post a Comment

Subscribe to Post Comments [Atom]

<< Home