Supposing I told you that there was a drug being developed for CLL which killed CLL cells in the test tube, but had absolutely no effect on T cells. A drug that did as well for cells from Mantle cell leukemia and from Waldenstrom's Macroglobulinemia. Moreover it also killed CLL cells from patients who were refractory to fludarabine and had either 17p deletions or 11q deletions.
Supposing this drug had already been given to a couple of thousand people in an attempt to treat another (non-malignant) disease at roughly two-thirds the dose needed to kill CLL cells and it had virtually no toxicity.
Would you want to have it?
Surely there must be some drawbacks?
AS far as we know it has never been given to CLL patients so we do not know whether it will work in patients. There are really no animal models of CLL that it can be tested in. (The TCL-1 mouse described by Carlo Croce's group is sometimes thought of as a model of CLL, but it is a much more aggressive tumor).
How should the company proceed?
I tell this story, not because there is a particular drug being developed, but as a hypothetical exercise in drug development.
My ideal drug for CLL would be just like this. It would kill CLL cells but not T cells and it would kill CLL cells with deletions at 11q and 17p. I would also like it to be effective in those other two difficult lymphomas, mantle cell lymphoma and Waldenstrom's macroglobulinemia. Finally it would have no side effects.
In the old days I would just give it to a few of my patients and monitor what happened, but today that is not possible. For one thing you can no longer publish as a clinical trials something made up as you go along. It has to be registered as a clinical trial with a defined protocol. It would have to be part of a randomized controlled trial, and before we could do that we would have to do a phase 1 study which would establish the maximum tolerized dose, and a phase 2 study that establishes that it actually kills CLL cells in the patient. The phase 3 study would be a randomized comparison with the best available (or at least the best licensed)treatment.
For the phase 1 study the idea is to treat three (usually end-stage) patients at the same dose as has been given before, and to gradually increase the dose for each successive three patients until the side effects become intolerable. The increases are according to what is known as a modified Fibonacci series in which the next number is the sum of the two previous numbers (220.127.116.11.18.104.22.168 etc).
In the phase 2 study a minimum of 14 patients has to be studied. If none of the first 14 improve you can be 95% certain that the drug is a dud. If it works on some then you have to add patients up to somewhere between 20 and 45 so as to get an estimate of how well it works.
Then the drug can go forward for a comparative trial.
The problem is that this is all very costly, especially the last part. It's all very well for the big boys, Glaxo and Bristol Myers and Roche, but a small start-up doesn't want to waste money on a dud.
For our imaginary drug, we don't want to start testing it on treatment failures. They will all have damaged T cells, and the beauty of our drug is that it doesn't damage T cells. Ideally we would want to try it as a first line drug, because the real question we want to ask is can it kill CLL cells in a patient without killing T cells.
So the question I want to ask all my CLL readers is, "Would you be prepared to take part in a trial of such a drug as first line treatment?"
The question I would like to ask any ethical committee members who read this blog is "Would you be prepared to sanction it?"