Monday, October 29, 2007

Late treatment in CLL

I have been looking at a group of patients with who manage to go a long time before requiring treatment, and comparing them with patients who go a long time and never require treatment. These are preliminary findings.

The database contains 424 patients in whom we have done the prognostic markers. Of these 168 have required treatment (some of them have not yet been followed for long and may yet need treatment). 19/168 survived for longer than 8 years before treatment was necessary.

The first thing to notice is that their average age at diagnosis was very low at 47 (range 36-73). There were 15 men and only 4 women. 7/19 had unmutated IgVH genes and 6 out of these 7 were also ZAP-70 positive, and 6/7 were also CD38 positive.

None of the patients had del 17p, but two had del 11q and of these one was unmutated (ZAP-70+, CD38 negative) and the other was mutated (ZAP-70 negative, CD38+).

Of the 12 with mutated IgVH genes, none were ZAP-70+ and 3 were CD38+.

Two of the mutated group were first treated because they developed autoimmune hemolytic anemia and one because he developed CLL-related nephrotic syndrome. Another patient was not treated until she developed an unrelated acute myeloblastic leukemia.

Six of the patients are still alive with post treatment survivals of 11, 20, 32, 37. 41 and 70 months. Only one patient has survived longer than 10 years post first treatment, and she was the one who was youngest when diagnosed.

There were 99 patients who remained untreated at 8 years or longer after diagnosis. This group was older with a mean age at presentation of 67 (range 38-91). There were 55 women and 44 men. 92 had mutated IgVH genes and 7 unmutated IgVH genes. Of those who were unmutated, 3 were also ZAP-70+ and 2 CD38+. Of those with mutated IgVH genes 2 were ZAP-70 positive and 12 CD38+. There were 5 cases of autoimmune hemolytic anemia that had resolved with steroids and not required any treatment of the CLL.

For completeness there are another 101 patients diagnosed long enough ago to have potentially survived more than 8 years. Their average age at presentation was 64 (24-93). There were 66 men and 35 women. 68 had unmutated IgVH genes and 33 mutated. 26/101 are still alive, 11 with mutated IgVH genes, 9 female, 9 CD38 negative and 18 ZAP-70 negative. Only one had del 17p and one del 11q. 26 (not the same 26) survived for more than 8 years after their first treatment, and 5 of these had had a transplant, and 7 had had CHOP.

Is it possible to draw any conclusions from these data? They confirm that it is better to have good prognostic markers even if you do require treatment, and it is better to be female. It seems to me that those with poor prognostic markers who live a long time without treatment are simply diagnosed earlier than those who seem to live shorter periods. It also seems that those who have the longest survival post treatment have either received a transplant or an anthracycline.

In the meantime this should be regarded as a work in progress as I think there are some missing data that I should be able to obtain.


Richard said...

"Survival since diagnosis" seems to me to tell one as much about the efficiency of the health service or the GP as it does about the progress of CLL.

For people who are diagnosed after their have counts reach 5e9/L would it not be better to least squares fit their counts to a straight line(log plot)and extrapolate back to find the time when their count would have been 5e9/L. This date could then be taken as an "effective diagnosis date". EDD for those who like acronyms.

Its not perfect, as doubling times are not always constant, but it would make a first order correction.

Or is this done already?

Vance Esler said...

I have wondered about lead-time bias, especially with the wider availability of flow cytometry which allows us to detect CLL at lower WBC levels, vis a vis the older definitions which were based upon an arbitrary absolute lymphocyte count.

Jenny Lou said...

This continues to be interesting. Is it correct that cller's with the worst prognostic marker's, (unmutated, CD38, Zap 70+) seem to make it OK for the first 6 years post dx and then the survival line falls at a rapid clip? That is where I see that newer treatment just might stop this runaway decline.

Anonymous said...

While the ALC may not be a reason to treat on an individual basis, I would think that if the average ALC at diagnosis for the group that was treated was significantly higher than for the untreated group it may suggest that they were diagnosed later in the course of their CLL.

John Liston

justme said...

I am very curious to know why some who have all good prognostic markers progress, when most others with those same markers don't.

Terry Hamblin said...

Richard - the backwards extrapolation idea does not work in practice because the blood compartment does not represent a contsnt proportion of the total available space for expansion over the course of the illness.

Vance - lead time bias is a real problem. just as it is with conditions like MDS and other diseases that are clinically silent for a long time.

Jenny Lou - its not quite as simple as that. There are some patients with good prognosis markers that behave like that too.

John, that is probably true.

Justme - So am I. There are obviously unidentified adverse factors.