Friday, July 20, 2007

97, 98, 99, 100

A long gap between blogs for me. Someone once gave me the advice that if I wanted to be a writer, I should try and write something every day. Well I haven't been slacking. I have been doing some data manipulation to try and prepare a couple of abstracts for the IWCLL meeting in September in London.

The question I was addressing, was how mutated you have to be to be mutated. Another question is for which patients is it justified to conduct a randomized clinical trial of early treatment, because they have bad prognsostic markers. We now have about 400 patients for whom we have sequenced the IgVH genes plus another 600 for the CLL4 trial. 353 of our patients have also had CD38, ZAP-70 and FISH for del 11q and del 17p done. Our conslusion on the clinical trial is that almost all of the patients who have two of the following criteria: unmutated VH genes or if mutated, use of the V3-21 heavy chain, CD38 >20% expression, ZAP-70 >10% expression and del 11q or del 17p, will require treatment within three years of being diagnosed. But should unmutated mean >97% or >98% homology?

I looked first at whether there was a difference between different degrees of mutations and the acquisition of other adverse prognostic factors. There was. At 100% it was 96%, at 98 and 99% it was 82%, at 97% it was 59% and at less than 97% it was 31%. Then I looked at treatment-free survival curves for those without extra adverse factors. I was surprised to find that they all did equally well, no matter how many mutations they had, with a plateau of not needing treatment at about 60%. However, treatment-free survival curves of those who did acquire extra prognostic factors were different. For those with 97% or greater homology the average time to needing treatment was between 2 and 3 years, while for those with less than 97% homology, the average time to needing treatment was more than 10 years.

What this means is that IgVH genes have a double effect on CLL survival. First they represent a prognostic factor in their own right. If the homology is more than 97% they have a bad effect. But you need at least two things wrong to have a poor prognosis, and the VH gene mutation is only one of them. The second effect is to facilitated the acquisition of this second factor. With 100% it almost universal, whilc for those who are really mutated, only 3/10 will acquire another adverse factor (only for them it will be the first adverse factor). The 97% patients are about twice as likely as this to acquire a second factor, and those with 98 or 99% about 3 times as likely.

No wonder different labs have chosen different levels of mutation as the cut off.

20 comments:

Anonymous said...

Please explain the percentages and mutational status.

I assume homology means the 'sameness' in CLL cell markers or gene expression. (Literally, homology would be the study of sameness or likeness, wouldn't it?)

Anyway, I assume that one could have 97% of the CLL cells be identical, and yet be of a mutated status?

This almost seems like an after thought, to show 2 or 3 percent difference in CLL characteristics.

Is this right?

Terry Hamblin said...

In CLL all the cells are identical.

There are 51 VH genes and each CLL uses only one of them. As part of the normal maturation of lymphocytes these genes undergo mutations fo that the antibodies that they code for will be better suited to combine with the targets that they are designed to hit. Our discovery 9 years ago was that those CLLs that fail to undergo this mutation were more malignant than those that had. But what was the limit of mutation that we could allow?. At first we thought that small variations - up to 2% - were polymorphisma (natural variations in gene sequences that define one individual from another) but then we found that they weren't, they really were mutations. But we had already shown that small numbers of mutations didn't free you from being unmutated. In fact the Germans suggested that you could have up to 3% mutations and still be effectively unmutated. In this study I have refined this a liuttle bit.

Anonymous said...

Your entry is written at just the appropriate time for me as I have been pondering my fate since learning of my mutational status. After being diagnosed with CLL 1 1/2 years ago at age 45, the subsequent prognostic results came back with all favorable results (ZAP-70 neg, CD38 neg, 13q14 deletion, etc), the only outstanding test being the mutational status of the VH gene. I was recently surprised (or should I say shocked) to learn that I am 98.9% unmutated. My clinical course has thus far been rather benign with little progression which corresponds with the information presented in your blog. My question would be whether I am at risk for aquiring additional adverse prognostic factors (such as 11q or 17p deletion or a change from ZAP-70 neg to positive) since I am unmutated, or would it be expected that my prognostics will remain the same.

Anonymous said...

So you identify the clone (and being clones they would be identical) as being CLL cells because of their CD markers, I suppose. (CD5+, CD23+, etc)

Yet in an unmutated person, two percent of the CLL cells might be not identical to the vast majority of CLL cells, right? Does this mean there is a second population of CLL cells? Or does it mean that a small portion of CLL cells then underwent the mutation process?

I understand the basics of some of this, but the details are frustratingly slippery.

Terry Hamblin said...

No, all the cells are the same. The 2% refers to how much the gene has changed from the one that is in the germ line. What is clear is that not all 'unmutated' cases are completely unmutated. A small amount of mutation is allowed. Whether this small amount should be 2% or 3% has been disputed. This helps to settle it. The mutations have 2 effects. if they are not present, or only present at a level of less than 3% they actas a poor prognostic factor in their own right, but it is not enough to only have one poor prognostic factor; you need two. Acquiring that second prognosic factor is very likely if you are completely unmutated, less likely if you have 1 or 2% mutateions, much less likely if you have 3% mutations. Even if you have more than 3% mutations you still have a 30% chance of getting another poor prognostic factor, but, of course, you wouldn't have the first poor prognostic factor, would you?

Anonymous said...

I think I understand what you are saying.

IgVH is the grand daddy of poor prognostics because if you are unmutated it is very likely you have other poor prognostics and the more unmutated you are the more likely this is.

But if you are a mutant, but have say CD38+ and ZAP 70+ you probably will still need treatment in 2-3 years?

Is my understanding correct?

Terry Hamblin said...

Steve,
I can't say that because in our series we do not have anybody who is mutated but ZAP-70+ and CD38+ (or for that matter who has a positive FISH). We do have two patients who were mutated, CD38+ and del 11q who required treatment at 29 and 32 months respectively, and one who was CD38 positive and acquired del11q late in his disease, who required treatment at 25 years.

Anonymous said...

Dr. Hamblin: What is the homology percentage of normal B cells that have gone thru mutation? and if CLL cells aproach this percentage could it be that they are still able to produce functioning antibodies and IG levels would not become depleted?

John Liston

Terry Hamblin said...

The normal cells are not clonal. They are mutated to variable amounts - roughly the sort of range that is seen in mutated CLL. CLL cells do make small amounts of imunoglobulin, but they do not mature into plasma cells to make large quantities of it. Even if they did it would not be useful antibody - it would all be against teh same target. Vecause normal B cells are not clonal, they produce a wide variety of antibodies to target any pathogen one is likely to meet.

Anonymous said...

So although CD38+ and ZAP 70+ alone are not good surrogate markers for IgVH mutation status. When they both exist in the same person, it is almost certain that they are unmutated?

Terry Hamblin said...

According to our data. Remember that the way we do zap-70 we get a <90% correlation with VH genes. The Rassenti paper showed a 77% correlation, but that was by a different method.

Anonymous said...

Terry I am aware of the NEJM article. From Ressanti et al in 2004.

The conclusion was:

"Although the presence of an unmutated IgV(H) gene is strongly associated with the expression of ZAP-70, ZAP-70 is a stronger predictor of the need for treatment in B-cell CLL."

What I was asking is seeing your data has not shown a case where CD38+ and ZAP 70+ has occurred in a case of mutated CLL. Are these two prognostic indicators, if they occur together, conclusive in being a surrogate marker for IgVH mutation status?

Terry Hamblin said...

Obviously they are for our dataset. But it cannot be generalized because our ZAP-70 is such a close corelate with VH genes and this is not the case for all ZAP-70 assays.

Dave said...

Dear Dr. Hamblin,

Thanks for this data. It answers a lot of questions I have had since, like anonymous, I have no bad prognostic markers, except that I am 98.3% homologous. Sounds like if I don't pick up another one, the disease will be on the slower side (hopefully).
Thanks again, I look forward to the publication.

Anonymous said...

Late reading this, but Tom, my husband, is one of the changing CD38 patient's. At dx he was CD38- with only 4%. 2 1/2 years later, after FCR, he is 25% CD38, unmutated at 98.6% and Zap70+ at 40%. The interesting thing is his WBC is continuing to fall within 4 tests this year from 6.8 to 4. Sweats and serious fatigue are present. Other than that, he is rocking along.
Thanks for this explanation. It helps quite a bit.

Terry Hamblin said...

One of the reasons that Chaya is against FCR is that it might select out a more malignant, drug-resistant population of cells. Certainly our discovery that CD38 can increase was first made in patients folowing this sort of treatment. One of the first patients that I treated with FCR had an original CD38 of 11%. Following FCR he had very few residual CLL cells, but 76% were CD38 positive. Fortunately he remains in remission.

Anonymous said...

Terry,
Are you stating that the CD38 cells that come back after FCR are the ones that were resistant to FCR in the first place? At the end of FCR, Tom was a PR with MRD. No mop up was done with Campath. He is somehow managing his disease at this point. Is relapse inevitable? Dr. Keating did say that next treatment he wanted to try something that was as not as toxic to Tom's immune system. Is this because the CLL cells now are resistant to FCR?

Terry Hamblin said...

I think that relapse after FCR is always inevitable, but it may take a very long time. The CD38 cells are probably the ones that were always there.

Anonymous said...

Some patients who have had FCR are 10 years out, and still in remission. I suppose these are the exceptions, but a 10 year remission in someone who is 75 is a great achievement, isn't it?

It has been pointed out that ALL CLL treatment selects for resistant cells. That would be the case for all chemo, I would think, for all cancers, that are not cured.

Chaya can be against FCR, but what drug or drug combo gives such a high complete remission rate?

FCR really should be the new gold standard treatment.

Terry Hamblin said...

FCR is regarded as the gold standard in America, but trials that might show that it is better than FC won't report until next year. As chairman of the data monitoring committee, I shall be the first to know.