Novel Gene Mutations In Chronic Lymphocytic Leukemia: Prevalence and Clinical Implications In A Series Of 3185 Cases - Initial Results From The European Research Initiative On CLL

In chronic lymphocytic leukemia (CLL), genomic aberrations identify subgroups of patients with distinct treatment outcomes. In particular, patients with deletion of chromosome 17p and/or TP53 gene mutations have inferior prognosis with standard therapy. Hence, screening for these abnormalities is mandatory prior to enrolling patients in clinical trials (CT) and can be also useful in general practice (GP). Recent studies suggest that aberrant p53 function (p53ab), due to TP53 gene mutations and/or del(17p), is only one of several mechanisms underlying clinical aggressiveness. Indeed, mutations in the NOTCH1, SF3B1 and BIRC3 genes have been reported as recurrent in CLL and correlated to poor outcome which is in contrast to mutations in MYD88. An important caveat stemming from published studies is the relative small and hetereogeneous sample size thus making it difficult to draw conclusions regarding the actual incidence and prognostic relevance of these mutations in CLL. The European Research Initiative on CLL (ERIC) is pursuing several studies to shed light on these issues in large and well-annotated CLL series, comprising both GP CLL and CT CLL cases, and here we report the initial results of this project. To date, a total of 3185 patients have been included in the analysis; 596 (18.7%) are CT cases and 2589 (81.3%) GP cases. Mutational screening provided data on NOTCH1 (entire exon 34 or targeted analysis for del7544-45), TP53 (exons 4-9), SF3B1 (exons 14-16), BIRC3 (exons 6-9) and MYD88 (exon 5). For each gene, almost two-thirds of the cases were analyzed within a year of diagnosis. Overall, BIRC3 and MYD88 mutations were relatively rare [25/923 (2.7%) and 24/1085 (2.2%) respectively], however: (i) BIRC3 mutations were significantly (p<0.05) more frequent in cases with (a) unmutated IGHV genes (U-CLL), (b) del(11q) and (c) trisomy 12; (ii) all MYD88 mutations were found in cases with mutated IGHV genes. Regarding the other genes analyzed: (i) NOTCH1 mutations: 246/3038 (8%) cases, with a clear bias (p<0.01) to CT cases; (ii) TP53 mutations: 217/2154 (10.1%) cases, with no significant differences between GP vs. CT cases, likely reflecting the policy of some institutions to routinely screen only U-CLL; (iii) SF3B1 mutations: 225/2028 (11%) cases (K700E: 39%), enriched among CT cases (p<0.00001). Co-existing mutations were very infrequent: TP53+SF3B1: 28/1756, 1.6%; NOTCH1+TP53: 24/1978, 1.2%; NOTCH1+SF3B1: 15/1983, 0.7%. Overall, the prevalence of mutations differed significantly between institutions, probably due to differences in the composition of the respective cohorts and/or the detection method: (i) TP53: 4.5-39%; (ii) NOTCH1: 4-13.4%; SF3B1: 3.6-20.6%. Taking this into account, no clear geographical differences in the prevalence of different mutations were observed. For all three genes, mutations were significantly (p<0.05) associated with Binet stages B/C, shorter lymphocyte doubling time and U-CLL. For each gene, mutant cases displayed significantly (p<0.05) different profiles of genomic aberrations: (i) NOTCH1: high frequency of +12, low frequency of del(13q); (ii) TP53: high frequency of del(17p), low frequency of +12 and del(11q); (iii) SF3B1: high frequency of del(11q), low frequency of +12. The clinical impact of NOTCH1 and SF3B1 mutations as well as p53ab on time-to-first treatment (TTFT) was assessed in 647 treatment-naïve Binet A cases. In univariate analysis, mutations in NOTCH1 or SF3B1 and p53ab were significantly associated (p<0.001) with shorter TTFT. In multivariate analysis (tested variables: IGHV mutational status, p53ab, +12, del(11q), NOTCH1 mutations, SF3B1 mutations), significant independent factors for shorter TTFT were SF3B1 mutations (p=0.01), p53ab (p<0.0001), del(11q) (p=0.048) and U-CLL (p<0.0001). In conclusion, despite some caveats (e.g. retrospective nature of the study, patient selection biases, and differences in laboratory methods), this study offers information on the prevalence and prognostic impact of recurrent gene mutations in CLL based on the largest series of patients examined thus far. This study also indicates that novel mutations are likely to be integrated into a prognostic biomarker index in CLL. Nevertheless, to achieve this goal, important pre-requisites must be met including the standardization/harmonization of laboratory methods and the prospective evaluation of these markers in clinical trials.