Treatment related MDS/AML in CLL

One of the problems facing doctors treating patients with CLL and, indeed, their patients, is the risk of over-treatment. We have all heard the the surgeon's excuse, "The operation was successful but, unfortunately, the patient died." There is no point in achieving a complete remission is the treatment is so toxic that it kills the patient.

We have all heard about the notorious CALGB study where FCR was followed by Alemtuzumab that had 5 deaths due to infection (cunningly disguised as Grade 5 toxicities). I have heard that there was a 6th death caused by transfusion associated graft versus host disease from receiving a non-irradiated blood transfusion. It is axiomatic that if anyone who has previously received either fludarabine of Alemtuzumab requires a blood or platelet transfusion, they must receive irradiated blood. This was a story of over-treatment.

But it is not what I want to draw attention to today. I want to highlight the risk of the combination of alkylating agent and purine analog use together. The first bit of evidence comes from an old trial - the first CALGB fludarabine trial in fact, (Fludarabine Compared with Chlorambucil as Primary Therapy for Chronic Lymphocytic Leukemia Kanti Rai, BL Peterson, Fred Appelbaum et al N Engl J Med 2000; 343:1750-1757). This compared fludarabine with chlorambucil with the combination of fludarabine and chlorambucil. This third arm was abandoned half way through the trial because it was too toxic but we have the results of long term follow-up. Among 188 patients treated with fludarabine alone there was one case of treatment-related MDS or AML; among 140 treated with chlorambucil alone, there were none; but among 141 patients treated with the combination of these drugs there were 5. The numbers are small and perhaps were due to bad luck, but there is more evidence.

At ASH, Kanti Rai presented the long term follow-up of the CALGB comparison of F with FC (E2997) (Smith MR, Neuberg D, Flinn IW, et al. Increased incidence of therapy related myeloid neoplasia (t-MN) after initial therapy for CLL with fludarabine-cyclophosphamide (FC) vs fludarabine (F): long-term follow-up of US Intergroup Study E2997. Program and abstracts of the 52nd American Society of Hematology Annual Meeting and Exposition; December 4-7, 2010; Orlando, Florida. Abstract 924.) Among 137 patients who had fludarabine alone, there were 4 cases of treatment related MDS or AML but among 141 who received FC there were 9 cases. Interestingly, among those who had had the IGVH mutational status done, all 7 were mutated (ie low risk).

We have two papers reporting on the long term follow up of FCR. Tam et al (Tam CS, O’Brien S, Wierda W, et al. Long-term results of the fludarabine, cyclophosphamide, and rituximab regimen as initial therapy of chronic lymphocytic leukemia. Blood. 2008;112:975-980.) found 8 cases out of 300 developing MDS or AML, while Carney and Mulligan (Carney DA, Mulligan SP. Chronic lymphocytic leukaemia: current first-line therapy. Intern Med J. 2009;39:44-48) found 3 cases out of 61.

What this says is that there is a small but definite risk of treatment related MDS/AML from a purine analog/alkylating agent combination. This does not impact on overall survival, but it is probably an unnecessary risk in some patients. Since there seems to be a greater danger in the mutated subset - possibly because they live longer and have more time for the MDS to mature. It pushes me into saying that those with mutated IGHV genes might be better served by using chlorambucil and rituximab as first line treatment.

My theory of why the combination is so much more risky is that alkylating agents do cause genetic damage which is normally dealt with by immune surveillance but since immune surveillance is severely curtailed by fludarabine, the aberrant clones are permitted to expand and become lethal.