We assessed the relative value of CD38 in predicting the need for early treatment in 307 patients with chronic lymphocytic leukemia (CLL) previously characterized for ZAP-70 and immunoglobulin heavy-chain-variable region gene (IgVH) mutational status (

NEJM 2004, 351;9;893–901
). Recursive partitioning by maximally selected chi-square analyses of flow cytometry and clinical data identified the optimal cut-off for designating a CLL sample as CD38+ was at 34%, which was highly similar to the conventional cut-off of 30%. Since these cut-offs identified highly similar cohorts of CD38+ CLL patients (e.g. 32.2% (99/307) vs. 35.5% (109/307)) we chose to use 30% as the cut-off for this analysis. We used the logrank test to compare CD38 expression-status to the time from diagnosis to therapy (TTT), initiated per established NCI-Working Group guidelines. Patients with CLL cells that were CD38-negative by this criterion had a median TTT of 7.8 years; whereas, patients with CD38+ CLL cells had a median TTT of 3.7 years (p<0.0001). There was a significant association between expression of CD38 and ZAP-70 or use of unmutated IgVH genes (p < 0.0001). The median level of CD38 among cases that used mutated IgVH was 4.0%, compared to 27.4% among cases that used unmutated IgVH (p < 0.0001). The median level of CD38 among ZAP-70-negative cases was 5.3% (range from 0.1% to 99.5%), compared to 27.2% among ZAP-70-positive cases (range from 0.1% to 99.4%) (p < 0.0001).

We explored the TTT based only on flow cytometric parameters (CD38 and ZAP-70) and investigated whether addition of mutation status significantly altered our predictions. For most patients, the knowledge of CD38 and ZAP-70 provided a reliable means for predicting the TTT and the knowledge of the IgVH mutation status did not substantially alter the prediction. Only in 41 cases (14% of the total), where the CLL cells were CD38+ but negative for ZAP-70, was the prediction significantly improved by inclusion of IgVH mutation status. The median TTT of this group of 41 patients was 7.8 years. Addition of mutation status identified a subgroup of 20 patients with unmutated IgVH genes for whom median TTT was estimated at 4.6 years; 21 patients with mutated IgVH genes had a median TTT of 8.4 years. For patients with CLL cells that were CD38-negative and either negative or positive ZAP-70, stratification via IgVH mutation status did not identify subgroups within each of these categories that had significantly different median TTT. We conclude that knowledge of CD38 and ZAP-70, as assessed by the CRC, can reliably predict TTT in most patients with CLL. For the 14% of cases that are CD38+ but negative for ZAP-70, determining the IgVH mutation status may provide added value in predicting the time from diagnosis to initial therapy.

Topics:
cd38, chronic b-cell leukemias, chronic lymphocytic leukemia, disease progression, immunoglobulins, mutation, zap-70 kinase, flow cytometry, data mining

Disclosure: No relevant conflicts of interest to declare.

Author notes

*

Corresponding author

2006, The American Society of Hematology
2006
Sign in via your Institution