Tuesday, August 17, 2010

ZAP-70 changes

In order to make this paper available to all.

ZAP-70 levels do change in CLL but not very often.Terry J Hamblin

Evidence is accumulating that in some patients with CLL the expression of ZAP-70 changes over time. This diminishes its role as a prognostic marker but increases our understanding of the pathophysiology of the disease.

Although the mutational status of the immunoglobulin variable region (IgVH) genes remains the gold standard for prognostic markers in chronic lymphocytic leukaemia (CLL), it is a complicated assay unsuited to routine laboratories. Unfortunately, attempts to find a surrogate have run into several obstacles. CD38 expression measured by flow cytometry was an early contender and initially seemed to have as much predictive value IgVH mutations. Unfortunately, it was subsequently found to give disparate results in 28% and even worse was described as changing over time in 24% (1).

More recently, ZAP-70 expression has attracted great interest. Although transiently expressed during certain stages of B-cell development, most mature peripheral blood B cells do not express ZAP-70, a cytoplasmic tyrosine kinase which is the key signalling molecule for T-lymphocytes and NK-cells. In CLL it acts as an auxiliary molecule for B-cell signalling in the more aggressive cases, probably acting to block inhibitors of signalling through syk. It’s presence in aggressive CLL appears to be due the chaperoning function of heat shock protein 90 (Hsp90). After it was originally identified in gene expression microarrays as a gene differentially expressed in CLL cells with unmutated IgVH genes, several laboratories developed flow cytometric assays that potentially could be used by routine laboratories. Initially, there was 95% concordance with IgVH mutational status, but a later assay using the much more convenient direct immunofluorescence, showed concordance of only 77%, hardly better than CD38, although with this assay ZAP-70 appeared to perform better than IgVH mutations as a prognostic factor (reviewed in 2).

The disparity betweens these assays has prompted a search by several laboratories for an assay that would be reliable, reproducible and of prognostic value. This quest is currently unresolved. The type of anticoagulant to use, allowable storage and transit time, which gating strategy, which monoclonal antibody and which fluorescent dye all remain in dispute and how to express the results is still a contentious issue (3). Nevertheless, several publications have attested to the fact that like IgVH mutations and unlike CD38, ZAP-70 is stable over time.

In today’s issue of Leukemia and Lymphoma, Poulain et al (4) dispute this. Four out of 33 ZAP-70 negative cases subsequently became ZAP-70 positive and this was associated with clinical progression. Two out of 27 ZAP-70 positive patients became negative. In one case this was associated with corticosteroid treatment, a phenomenon that has been observed previously. Although previous publications have reported on the stability of ZAP-70 over time, close scrutiny of these publications confirms that in a small number of patients (around 10%) ZAP-70 expression does change as time passes, although until Poulain et al there has been no indication as to why.

Recently Boelens et al (5) reported that ZAP-70 expression in CLL is higher in lymph node cells than in peripheral blood cells and Cutrona et al (6) demonstrated that within normal germinal centres there is a population of ZAP-70 positive, CD38 positive B cells representing in vivo activated cells.

These observations add to our understanding of ZAP-70 in CLL. It seems likely that both ZAP-70 and CD38 are normal features of activated B cells, rather than markers of neoplastic aberration. In CLL activation events and cell division both take place in proliferation centres and here upregulation of both markers occurs. There seems to be a free exchange of cells between tissue and blood making the presence of both CD38 and ZAP-70 surrogates for the rapidity of cell turnover.

The mutational status of the IgVH genes is fixed in CLL making it unique as a prognostic marker; it assigns cases into one immutable subgroup or the other. Other markers, whether CD38 or ZAP-70 expression or chromosomal abnormalities, can be acquired during the course of the disease. Early stage patients can gain little comfort from knowing that they are not present at diagnosis.


1. Hamblin TJ, Orchard JA, Ibbotson RE, Davies Z, Thomas PW, Stevenson FK, Oscier DG. CD38 expression and immunoglobulin variable region mutations are independent prognostic variables in chronic lymphocytic leukemia, but CD38 expression may vary during the course of the disease. Blood 2002, 99:1023-1029.
2. Hamblin AD, Hamblin TJ. Functional and prognostic role of ZAP-70 in CLL. Expert Opinion on Therapeutic Targets 2005; 9:1165-1178.
3. Marti G, Orfao A, Goolsby C. ZAP-70 in CLL: Towards standardization of a biomarker for patient management: History of Clinical Cytometry Special Issue. Cytometry Part B Cin Cytometry 2006; 70B:197-200.
4. Poulain S, Benard C, Daudignon A, Le Baron F, Morel P, Duthilleul P. Is ZAP-70 expression stable over time in B chronic lymphocytic leukaemia? Leukemia and Lymphoma, 2007; 48: 1219-21.
5. Boelens J, Philippe J, Offner F. B-CLL cells from lymph nodes express higher ZAP-70 levels than B cells from peripheral blood. Leuk Res 2007; 31:719-20.
6. Cutrona G, Colombo M, Matis S, Reverberi D, Dono M, Tarantino V, Chiorazzi N, Ferrarini M. B lymphocytes in humans express ZAP-70 when activated in vivo.
Eur J Immunol. 2006; 36:558-69.


George said...

excellent new insight. Terry can you name 2-3 labs in the UK that perform the mutational analysis? (In Cambridge we didn't do it)

Terry Hamblin said...

It is done at Bournemouth, Kings, Southampton, Liverpool and Dublin (although I think the last has left the UK).

Anonymous said...

I went to Kings (Dr Devereux) as I couldn`t get prognostic tests done in Scotland. Be prepared to spend a little over £1000.00.Came back that I have the unmutated gene, 11q del, CD38 20% and CD49 0%. Dr Hamblin can you explain what CD49 0% means. They don`t do the ZAP-70 test,I was told it is too unreliable.

Terry Hamblin said...

I think they charge £150 at Bournemouth, but if you want the other markers and FISH done the bill is about £350. CD49c is associated with a good prognosis and CD49d with a poor prognosis.

Anonymous said...

I had testing for Zap-70 and mutational status done at UCSD. Tests showed positive for Zap-70 and unmutated. At the time, I was under the impression that UCSD was among the most reliable in their testing protocol for both these markers. Is that still the case? Thanks.

Terry Hamblin said...

Yes and no. The method there finds more positives than the Crespo and Orchard methods and less consistancy with the IgVH mutations result, but figures on over 1000 patients suggest that it is is a better predictor of poor outcome than IgVH mutations. Warning! when the same test is done by Quest and other labs it does not show the same benefit.

Grateful said...

An earlier article in Blood said that CD38 and ZAP70 are functionally linked. If so, it makes sense that if CD38 can change, then so can ZAP70.

However, I could not follow the biochemistry in this article. The statistics are a bit shaky in support of this conclusion.

CD38 and ZAP-70 are functionally linked and mark CLL cells with high migratory potential

2007 110: 4012-4021

Cammie Novara said...

"This diminishes its role as a prognostic marker but increases our understanding of the pathophysiology of the disease." I agree fully. There's a really fascinating debate that I thought would be of interest on evolution vs. intelligent design going on at http://www.intelligentdesignfacts.com

Anonymous said...

I am one of the 27% percent that at diagnosis am IgVH mutated but have high (82%)CD 38 Levels with absolute counts of around 5000 and only other symptom some internal abdominal nodes...live and watch is current plan. Your thoughts for those of us with "mixed cll messages"? Zap 70 if one wants to "know more"?

Terry Hamblin said...

Our results showed a median survival of 15 years, but at this level predictions are impossible. Intermediate is all I can say. ZAP-70 may add something to the information.

Anonymous said...


Am I to take your " at this level predictions are impossible" to be in part a reference to the never ending variations in how this disease acts from person to person or the 82% is such a high number all prognostic factors fall by the wayside and the IgVH mutation will not likely slow things down?

So I guess I just stay fit, eat well and with a good FISH (only 13 q) hope treatment moves me closer to the mean of 15 years if the 82% moves me away from it!

I just found your blog this week and am enjoying the variety of posts and topics! Keep it up and hope your feeling well.

gerrysarson said...

Terry, only just discovered your blog, so apologies for this late posting.
Recent CLL diagnosis - who please should I contact at Bournemouth re testing?

Terry Hamblin said...

Dr Rachel Ibbotson.

gerrysarson said...

Thanks Terry - and forgot to ask - who please at Southampton?
best regards
Gerry Sarson

Terry Hamblin said...

Dr Kathy Potter in Tenovus

James said...

Dr. Hamblin - I have been recently diagnosed (Trisomy 12, CD38 of 48%, LDH high at 255IU/L, and serum IgM levels low at 15mg/dL). I am trying to construct an argument for my oncologist (in the US) to convince him to perform either the ZAP-70 or preferably the IgVH mutation test. His argument against this (and he is rather adamant) is that it he has does not trust teh accuracy of these tests and that in any event the results will not affect his initial treatment decision of w&w (which by the way I am completely comfortable/happy with). However, I am 48 years old, so a prognosis of 6-8 years on average versus a prognosis of 25 years on average makes a big difference to me in terms of what planning and living I do with my wife and children as well as my career, even if it has no affect on treatment today... Any words of advice?

Terry Hamblin said...

Your oncology is correct in the sense that the prognostic test are geared to populations rather than individuals. But then so are things like white count and doubling time which no doubt he relies on. People with mutated IGVH can die next week and those with unmutated IGVH can live for 15 years, we are talking about averages. The prognostic marker will not affect whether you are treated or not nor which treatment you receive (at least not yet) but they will settle your own mind as to which bucket you fall in. This can be important for some patients (though other patients prefer not to know). Your doctor should offer what you want not what he wants. It's your body, your disease, not his.

I agree over the accuracy of ZAP-70, but there is no doubt about IGVH mutations. He did CD38 so what is his problem?

James said...

Dr. Hamblin,

First of all, thank you very much for your prompt response - it says a tremendous amount about your character that you take the time to respond to so many strangers around the world.

Of course, I cannot necessarily state with a high degree of confidence why my oncologist is responding the way he is (I often regret skipping the mind reading class while at University!), but I can infer his reasoning based on his response to me, which was entirely centered on what would affect his decision of when and how to treat.

In saying no to my request, my oncologist has been very diplomatic and logical. From what I understand my oncologist firmly believes in an evidence based approach to treatment, and stresses that nothing is likely to change in my condition over the next couple of years, by which time he would be able to draw more conclusions as to what potential course of my disease and the treatment options to pursue. He stresses that the IgVH test is difficult and expensive, and the ZAP-70 tests have too many errors to trust. My position is that there is the whole patient to take into account. Considering my relatively young age, the mean survival prognosis would play a significant role in how I live my life (career and financial decisions, even family decisions). Finally, while some patients may not want to know their prognosis, my personality is certain to obsess over this more without the information, negatively affecting my quality of life.

I have an appointment with a second oncologist today and will make this argument to him. Thank you both listening as well as for your words support.

Terry Hamblin said...

Not as expensive as one course of fludarabine and not that difficult - it is offered at many labs in America and it can be done at my old lab for $250 if a sample is sent by FedEx.

James said...

Thanks Dr Hamblin! The oncologist I met today completely understood my perspective and was much more open to completeely exploring the prognostic indicators, including re-running the previous tests and adding IgVH mutational status, ZAP-70, B2M, etc... In fact the nurse that was drawing the blood was a little surprised that the doctor's tests called for 10 vials of blood... Of course, the doctor reminded me that all of this will only help to compare me to a body of statistics, but that I am decidedly not a statistic...

The downside is that they will use Quest for the ZAP-70 test, so I suppose we will have to take those results with a boulder of salt...!

James said...

Dr. Hamblin,

I have what may be a silly question... In a paper you were an author of, titled "Differential expression of CD180 and IgM by B-cell chronic lymphocytic leukaemia cells using mutated and unmutated immunoglobulin VH genes", there is a sentence that reads:

"This was in contrast to the higher levels of expression of surface immunoglobulin M by B-CLL cells using unmutated, rather than mutated IGVH genes."

Does this mean that serum IgM levels would be lower in the blood of those with mutated IGVH? If yes, what would "lower" actually mean (are they both relatively low, or just that IgM levels in mutated cases are much lower, or what)?

Obviously, I am engaged in that odious past time, the "waiting game", until the results of my test for mutation status and ZAP-70return. Of course, I am going to fill my hours with speculation :-), and trying to figure out how to extrapolate from what I already know (Trisomy 12, CD38 48%, WBC 21,000, Lymphocytes 17,000, serum IgM of 15 mg/dL, LDH 255), so was looking at the IgM level as "low" and wondered if that meant more than the CD38 @ 48% (which at 72% correlation to mutation status is only slightly better than chance - I know, I know, a statistician would take that apart...).

Alexander Pope once observed that a little knowledge is a dangerous thing, and if it is so then Google has made us all extremely dangerous :-)

Terry Hamblin said...

Surface IgM has no relationship to serum IgM.

James said...

Thank you Dr. Hamblin!

I have test results trickling in from the tests run by my second oncologist. This doctor uses Quest rather than Labcorp, and we are seeing very different results in some areas... CLL is still confirmed, but CD38 is now only 8% (versus 48%), the lymphocyte count dropped almost in half and the neutrophil count about tripled (went from low to high!). I had two Labcorp CBC's that were very similar over a month apart, and this one from Quest a month after the last Labcorp test is very different from either of the labcorp results...

Still no results on my IGVH mutation status and ZAP-70, which I think will be much more important than what the CD38 reading is anyway, but still the big swing is very interesting...

Keeping the theme of the original post (sort of!) I understand that there is some thinking that the expression of ZAP-70 changes over time, and similarly CD38 may also, but does it make sense that the CD38 would change by 40 percentage points over a 5 week period? Yes it is two different labs, so comparing the results is not as easy, but going from well above the presumed magic 30% level to well below that level in just over a month seems odd...

Terry Hamblin said...

Thanks for the information. I do not trust the results from commercial labs and your results seem to justify my distrust.