ZAP-70 mRNA expression provides clinically valuable information in early-stage chronic lymphocytic leukemia

Chronic lymphocytic leukemia (CLL) is a heterogeneous disease with a variable clinical course and overall survival times ranging from months to decades.1 Mutational status of immunoglobulin heavy chain variable region (IGHV) correlates with clinical behavior and is a powerful prognostic factor in CLL. Unmutated IGHV patients have a reduced survival and poor responsiveness to chemotherapy,2,3 however, IGHV sequencing is difficult to perform in a routine diagnostics laboratory. Gene profiling studies indicated that ZAP-70 was the gene that best distinguished between IGHV groups4 and that it could serve as an independent prognostic factor that is expressed in a stable manner during the course of the disease.5–7 On the other hand, assessment of ZAP-70 by flow cytometry (FC) presents some technical difficulties since T and NK cells express ZAP-70 and must be excluded from the analysis.

Our objective was to evaluate the prognostic significance of ZAP-70 determined by real-time PCR (RTqPCR) in early-stage CLL patients and compare its performance with FC analysis and IGHV mutational status to identify patients at risk of progression.

We studied 70 samples from untreated CLL patients (Binet stage A) after obtaining their informed consent. Rearranged IGHV genes were amplified by PCR using a standard protocol.8 We considered unmutated those samples with >98% homology with the closest germinal line.

FC analysis of ZAP-70 was performed on fresh samples (n= 69) according to Crespo et al.6 with some modifications. An isotype control was used as negative control. Results ≥20% were considered positive.

For RTqPCR assays, RNA was prepared from 7–10×10 CD19 selected cells and amplification was carried out using Hs00277148_m1 primers and probe sets (TaqMan Gene Expression Assays, Applied Biosystems). Amplification of GUS gene was performed in all cases to normalize gene expression.

Time to progression (TTP) was calculated from the date of diagnosis to the date of disease progression (based on NCI guidelines) or last follow-up. All statistical calculations were performed using the SPSS 13.0 software. Out of the 70 patients studied, 20 (28.6%) were unmutated. The most common families used in mutated samples were V_H3-7 (10.0%), V_H2-5 (10.0%), V_H3-30 (8.0%), and V_H1-3 (8.0%), whereas unmutated patients were V_H1-69 (30%) and V_H3-33 (15%).

The mean ± 2 SEM values of ZAP-70 measured by RTqPCR and by FC in IGHV unmutated and mutated samples were as follows: RTqPCR: 0.5115±0.284 and 0.1000±0.040; FC: 40.36+ 12.3 and 14.44±4.44 respectively (p<0.0001 in both cases, Mann Whitney test).

The area under the ROC curve for ZAP-70 expression measured by RTqPCR was higher (0.879) than for FC (0.849), indicating that RTqPCR was a better method for predicting mutational status than FC. A global concordance between RTqPCR or FC results and IGHV mutational status was almost equal (84.28% and 84.06% respectively). We obtained 11 discordant patients for both assays; however, the distributions of the discordances were different. By RTqPCR we found discordant results only in IGHV mutated group, whereas 100% of the IGHV unmutated patients show ZAP-70 positive values. By FC we had 11 discordant patients, 7 were IGHV mutated and ZAP-70 positive, and 4 were IGHV unmutated and ZAP-70 negative. In ZAP-70/IGHV discordant cases, there were 3 patients showing FC and RTqPCR concordant results. These patients were ZAP-70 positive and had other prognostic factors of poor outcome: 2 of them were CD38 positive (one had del11q), and the third was CD38 negative but was mutated in BCL-6 gene. In the whole group, FC and RTqPCR results were concordant in 82.3% of the patients. In the discordant group, RTqPCR was better in assessing IGHV unmutated cases. Follw-up for the cohort was a median of 42.07 months (range: 1.00–141.33) in order to record patient treatment requirements due to disease progression. As expected, IGHV mutations defined two subsets of early-stage CLL; IGHV mutated cases had longer TTP than those unmutated (p<0.0001) (Figure 1A). In the case of RTqPCR, the TTP curve showed that ZAP-70 negative patients had a longer TTP than the ZAP-70 positive group (p=0.001) (Figure 1B). Our data show that 100% of IGHV mutated and 100% of ZAP-70 negative by RTqPCR patients have the longest TTP. On the contrary, 7 IGHV unmutated patients and 7 ZAP-70 positive patients required treatment. In the case of ZAP-70 by FC, again patients were divided into two groups but the statistical significance was lower (p=0.041) (Figure 1C).

Our results support the proposal of Catherwood et al.9 about the usefulness of a RTqPCR assay to analyze ZAP-70 expression assigning a high percentage of CLL samples to the correct IGHV mutational status. We show that RTqPCR is a good, sensitive and accurate method to determine ZAP-70 expression having some advantages over FC, which is actually the technique most widely used. It is known that FC has some drawbacks, such as: requirement of cytoplasmic permeabilization, different performance depending on the monoclonal antibodies and fluorochromes used, and different cut-off values reported.6,10–12 RTqPCR is easier to standardize and the requirement of B cell selection could be automatized.

ZAP-70 expression as a prognostic marker was analyzed by TTP curves. In agreement with previous results, RTqPCR methodology performed well, whereas FC showed a poorer performance and less statistical significance. Rassenti et al.10 reports that measurement of ZAP-70 expression is even more significant than IGHV mutational status. In our cohort, IGHV mutational status and ZAP-70 by RTqPCR were strong predictors of TTP.

Table 1.Relation between ZAP-70 expression and IGHV mutational status.

Figure 1.TTP survival curves according to IGHV mutational status and ZAP-70 expression. (A) IGHV mutational status, (B) ZAP-70 mRNA expression by RTqPCR and (C) ZAP-70 expression by FC. Assessment of mutational status was based on a 98% cut-off value, ZAP-70 expression was determined by FC analysis of whole blood (n=69) or by RTqPCR analysis of CD-19 selected cells (n=70).

In conclusion, we have demonstrated that determination of ZAP-70 in B cells by RTqPCR has an excellent correlation with IGHV mutational status and also with TTP in early-stage CLL. These observations suggest that ZAP-70 by RTqPCR would be a useful clinical test that might be used at the time of diagnosis to identify patients who have an increased risk of early disease progression and to guide treatment decision, since this technique is more amenable to application in clinical laboratories than IGHV sequencing.

Acknowledgments

we thank to Dr. Isabel Benet for her technical assistance in the flow cytometric analysis of the samples. Key words: ZAP-70, biomarker, CLL, prognostic markers.

Footnotes

• Cristina Reinoso-Martín and Eloisa Jantus-Lewintre should be considered both as first authors.
• Funding: supported by grants from Carlos III Health Institute: G03/179, FIS PI020889, PI051001.