Prognostic Tests

April 2005

The valuable prognostic tests are: VH gene mutations, ZAP-70 by Flow, FISH for 17p deletions (perhaps also for del11q, del13q and trisomy 12) and CD38 by Flow.

The problem is that the test cannot be done when you are in remission. The CLL cells need to be present in order to be tested. If your hospital has some archival material saved then it might be possible to do them on this.

If the tests are not available in Manchester they can be done in Bournemouth. We have certainly done Manchester patients in the past.

April 2005

The comprehensive grid that you ask for does not exist, and no-one has really been able to construct one from the information currently available. But we can have a go.

First, there is a pattern that predicts with a fair degree of certainty that the CLL will never need treatment. I have just analyzed my own data and this is what emerges:-

If you present with a lymphocyte count less than 30, 000, a normal hemoglobin and platelet count, no enlarged lymph nodes or spleen. and if you have mutated VH genes, a CD38 less than 30%, a ZAP-70 less than 20% (or less than 10% if the Orchard method is used) and either normal chromosomes or del 13q by FISH, then you will not die of CLL nor need any treatment for it. This statement is 98% certain.

If you have the del 17p abnormality by FISH, your CLL will almost certainly be unresponsive to chlorambucil, fludarabine, FC and FCR. The only available treatments are alemtuzumab (Campath) and high dose steroids. These may produce a remission, but they almost certainly won't cure. The only option for cure is a stem cell allograft, but even this is not certain.

Rai stage III and IV need treating straight away. Stage I or II only need treatment if the lymphocyte count is doubling every 6 months or less, or there are marked symptoms such as fatigue stopping you take part in normal activities, 10% weight loss in 6 months, night sweats so bad you have to change your night clothes, a rapid increase in the size of lymph nodes or spleen.

When the prognostic factors are assessed in patients who need treatment on clinical grounds, there are three groups of patients: those who have unmutated VH genes and raised CD38 and ZAP-70; those who have mutated VH genes, and low ZAP70 and CD38; and those with discordant results.

All three groups will respond fairly well to a wide range of treatments the first time around, but increasingly less well to successive treatments. Of the three groups, the first will have the shortest remissions, the second the longest and the the third, remissions that are intermediate between the other two.

There is some dispute as to what treatment should be used. There have been few head-to-head comparisons of one treatment with another, and those that there have been have not shown a difference in overall survival, though they have shown a difference in response rates and in length of remission.

Thus, for example, in the last German trial which compared fludarabine (F) with F plus cyclophosphamide (C) the response rate for F was 84%, but for F+C was 95%, and the length of remission was 20 months for F and 48 months for F+C. But so far there is no difference in overall survival between the two, because when you relapse after F you are likely to respond a second time to F+C. Indeed there is no evidence yet that F+C is better than either F or chlorambucil in making people live longer overall.

Earlier this year there were two papers on JCO that seemed to show that FCR was better than either FC or F alone. But the comparison was not done head-to-head, but sequentially. There is no way of knowing whether the ratio of ZAP-70 negative to ZAP-70 positive patients remained the same during all three trials. So although it seems likely that FCR produced better and more long lasting responses than F+C or F alone, it has not been proved. Nor of course has it been shown that people will live longer with FCR.

One interesting fact in the German trial was that the 5% of patients who had no response to F+C died very quickly, while the 16% who had no response to F had the same overall survival as the 84% who did respond. One possible explanation for this is that when very effective treatment is used, the only cells that remain are the drug resistant ones. It is thought that there is some measure of resistance or immunity to CLL operated by the patient's T cells. Fludarabine kills most of the body's T cells, so that the drug resistant leukemia cells can grow unabated after treatments like FC or FCR. This is only a theory, so it may not be true, but until we have some hard facts we need to remain cautious about what is the best treatment for individual patients.