Thursday, May 17, 2007

To treat or not to treat: that is the question.

This is a point that confuses a lot of people so I need to spell it out clearly. It has become an accepted fact that the new treatments for CLL do not extend life. This is not what the trials show.

The first thing that they show is that delaying treatment until symptoms appear (or the other features indicated by the NCI guidelines) does not put a patient at risk compared to starting treatment at diagnosis. The caveat for this is that when the trial was done we had no way of distinguishing the patients who would never need treatment from those who would need it eventually and the best available treatment was chlorambucil. The trial that demands to be done now is to compare outcomes in patients with poor-risk prognostic markers looking at a treatment better able to produce complete remissions than chlorambucil compared with watch and wait. These trials are either under way or in late stages of planning, using FCR, FR or Campath as the more effective treatment.

The second thing that they show is that overall survival is not affected whether the first treatment is chlorambucil, fludarabine or FC. Certainly FC is more effective in terms of response rate, CR rate, molecular remission rate, and length of remission, but despite there having been three trials comparing F v FC and 7 trials comparing fludarabine and alkylating agents and 1 trial comparing chlorambucil and FC, none has shown any advantage for starting treatment with anything other than chlorambucil. This is because those who fail chlorambucil or relapse after chlorambucil can have a second bite of the cherry with fludarabine or FC.

Unfortunately, there are no randomized trials comparing FCR or FR with chlorambucil, though there is a trial comparing FC with FCR, but it is not yet mature enough to be analysed; we will have to wait until 2008 or 2009. It may be, as MDACC think, that everybody should be given FCR as first line, but honestly there are no data to justify this.

What the trials do not show is that patients who are left untreated do as well as those who are treated. Quite the contrary. They show that patients who achieve a remission, especially a CR, live longer than those who do not. Of course getting drugs and not responding probably shortens your life, so this is not a fair comparison. But we do know that most patients who receive chemotherapy get a response. Patients who meet the NCI guidelines and are not treated mostly run into severe problems that are frequently fatal.

There are suggestions that patients with CLL are living longer than they used to, but this observation is unreliable because we are diagnosing CLL at an earlier stage than we used to.

There are suggestions that some treatments make the serious complications of CLL like transformation, immunodeficiency and MDS more likely. This may be true, but a properly designed clinical trial will show this, and the bottom line is, "Do the patients live longer?"

At the moment I am sure that, overall and when necessary, treatment is better than no treatment, but how and when the treatment is given is still a matter for individual decision. The NCI guidelines are just that: guidelines, not regulations.

No comments: