Tuesday, January 24, 2006

What is the aim of treatment? (Part 2)

What about a long remission?

There are lots of cases of acute leukemia that we can’t cure. Most people over 40 for a start. Does that mean we don’t treat them? Certainly not. We try and get them to live as long as possible in the best possible health. If you have acute leukemia then quality of life in remission is much better than quality of life in relapse, so this equates to wanting the longest possible time in remission. How about this as an ambition in CLL?

But in CLL you can have a very good quality of life in relapse. Most patients with CLL are asymptomatic. So starting treatment is delayed until symptoms start and restarting treatment is delayed until symptoms restart. In acute leukemia the first remission is the main deal. Subsequent remissions are questionable and short. In CLL it is quite usual to be able to get several remissions, many of which are long lasting. In acute leukemia the rule is to give it your best shot first because for many patients you are only going to get one shot. In CLL you may have to eke out the treatment over many years.

Most doctors who design clinical trials for CLL have trained as acute leukemia doctors. Faced with a disease that may span decades they revert to type. Pharmaceutical companies are not much interested in a clinical trial that may last 20 years – their patents will have run out. Current clinical trials, by common consent, have abandoned overall survival as an end-point; instead they have adopted complete response rate and progression free survival as surrogates.

There are problems with both of these. A German trial reported last year assessed complete response rates. They compared the complete response rate without bone marrow biopsy confirmation, complete response with bone marrow biopsy confirmation and thirdly, with CT scan confirmation. For each successive group the CR rate diminished. Their conclusion was the harder you looked for residual disease the more you found. When PCR is used to look for MRD, the ‘true’ CR rate falls even further.

Progression-free survival is even less reliable. What counts as progression? When the MRD test becomes positive? When the CLL becomes clinically detectable? When the CLL is bad enough to need treatment again? Some trials have used ‘treatment-free interval’ as the interval. But different units have different criteria for restarting treatment.

The biggest problem with these end-points is that they don’t correlate with overall survival. It is common ground that fludarabine produces more remissions than chlorambucil however the remissions are defined. It also produces longer remissions, however defined. But no trial has ever shown a longer survival for fludarabine than for chlorambucil.

The fludarabine people say that that’s unfair, because patients are allowed to crossover. Patients who fail chlorambucil are allowed to get fludarabine, and indeed, if you do fail chlorambucil you have a very good chance of responding to fludarabine; but vice versa the chances are very small. But this isn’t a game. Fairness doesn’t come into it. This is real life and in real life, patients want to live as long as possible no matter what order they take the drugs in. If someone could show that taking chlorambucil first and then fludarabine later shortens your overall survival compared to having fludarabine first then there would be no question. But the trials do not show that. Of course we will have to wait to see whether fludarabine combinations first are better than chlorambucil first, but so far in the British CLL4 trial the overall survival for chlorambucil first is no worse than for fludarabine plus cyclophosphamide first.

1 comment:

Anonymous said...

Thank you Terry for your comprehensive and astute thoughts!

Dick, in Florida