In chronic lymphocytic leukemia (CLL), biologic risk factors such as immunoglobulin variable heavy chain gene (VH) mutation status, CD38 expression level, and genomic aberrations have recently been identified, but the relative prognostic impact of the individual parameters is unknown. In the current study, we analyzed VH mutation status by polymerase chain reaction and sequencing (n = 300), genomic aberrations by fluorescence in situ hybridization (+3q, 6q−, +8q, 11q−, +12q, 13q−, t(14q), 17p−) (n = 300), and CD38 expression by triple-color FACS (CD5, CD19, CD38) (n = 157) in a unicentric CLL cohort. The prognostic influence of VH mutation rate and CD38 expression level was tested by maximally selected log-rank statistics. A corrected P value (Pcor) for a cutoff level allowing the best separation of 2 subgroups with different survival probabilities was identified at 97% VH homology (95% confidence interval [CI], 96%-98% homology,Pcor <.001) and at 7% CD38 expression (95% CI, 20%-71% expression, Pcor = .02). In univariate analyses, unmutated VH genes and high CD38 expression levels predicted for shorter survival times. The overall incidence of genomic aberrations was similar in theVH unmutated and VHmutated subgroups. High-risk genomic aberrations such as 17p− and 11q− occurred almost exclusively in the VHunmutated subgroup, whereas favorable aberrations such as 13q− and 13q− as single abnormalities were overrepresented in theVH mutated subgroup. In multivariate analysis, unmutated VH, 17p deletion, 11q deletion, age, WBC, and LDH were identified as independent prognostic factors, indicating a complementary role of VH mutation status and genomic aberrations to predict outcome in CLL.

There were characteristic modes of discordance between ZAP-70 and VH mutation status depending on the presence or absence of additional genetic high-risk features such as 11q and 17p deletion or V3-21 usage. Although the biologic background for these findings is yet to be determined, these data have biologic and clinical implications regarding ZAP-70 as a pathogenic factor and outcome predictor, respectively.

No abstract