IgVH mutational status and molecular cytogenetics have dramatically improved the ability to predict the prognosis of CLL patients. These tests, however, are highly sophisticated, complex and costly for routine use. ZAP-70, a syk family tyrosine kinase normally expressed in T cells, is a newly described marker which correlates with clinical progression and shorter survival in CLL. A flow cytometry assay to detect ZAP-70 described by Crespo et al (1), appears to be the simplest approach for routine clinical stratification in B-CLL. It is highly informative, and has a strong correlation between the expression of ZAP-70 in CLL cells and clinical outcome. However, in this analysis there are some technical aspects that should be improved to enable it to be standardized as a routine flow cytometry assay. ZAP-70 expression in B-CLL cells is not quantitative but assessed relative to its expression in the T- and NK cells (CD3+, CD56+). This approach can be problematic at times, as ZAP-70 levels in T cells vary in CLL patients as well as in normal controls, probably due to its up regulation following activation. An additional quandary in this assay is that all results are recorded relative to the subjectively delineated T-cell gate. Accordingly, small changes in expression in the T cells can significantly alter the results obtained in some B-CLL samples. In this study we aimed to improve the resolution of the assay by performing a quantitative analysis of ZAP-70 expression within the B-CLL cell population which is uncoupled from T cells. Blood samples were stained by the method described by Crespo et al (1) and ZAP 70 levels in B cell populations in CLL patients (CD19+CD5+) and in healthy volunteers (CD19+) were determined using a standard curve generated by an absolute fluorescent standard of FITC high levels beads with a range of 50–2000 x103 molecules of equivalent soluble fluorochrome (MESF) units per microsphere. Quantitation of expression levels were generated using Quick Cal V2.2 via www.bangslab.com. (Bangs laboratories). Using this analysis system the mean expression levels of ZAP 70 were calculated in healthy B cells (n=11) to be 11,177±1812 MESF units while in CLL (n=36) the mean value was >143,000 MESF units. To determine the reliability of this new method and its clinical relevance we compared our results to data generated using the analysis method of Crespo et al (1). We found a significant correlation between the two methods (r2 = 0.7558). Using ROC curve analysis with maximum sensitivity and specificity, our minimum positive value was found to be 46,700 MESF, with >95% sensitivity at 27,000 MESF and >92% specificity at 67,000 MESF and a Pearson correlation of 0.877 (P<0.0005). We conclude that this assay can provide a more reproducible and reliable analysis of Zap-70 expression in B-CLL, which is easily standardized. This analysis is highly specific as it is quantitative, not subjective and uncoupled from T cell activation in the sample.

Crespo et al.,
N. Engl. J. Med.

Author notes

Corresponding author