RAS/RAF Pathway Mutations Define a Distinct Subset of Chronic Lymphocytic Leukemia

Widespread use of next-generation sequencing (NGS) based mutational analysis has identified recurrent mutations of prognostic significance in chronic lymphocytic leukemia/ small lymphocytic lymphoma (CLL/SLL). In addition to novel genes such as NOTCH1 and SF3B1, targetable mutations in genes involved in RAS-MAPK pathway such as KRAS and BRAF including V600E have been reported. However, the detailed clinicopathologic, cytogenetic and molecular genetic features of KRAS/NRAS/BRAF mutated CLL/SLL (CLL^rasmut) are not known. We describe the clinicopathologic findings in 44 cases of CLL^rasmut.

NGS analysis targeting 53 cancer-related genes was performed on Miseq Illumina on CLL/SLL cases on bone marrow (BM) and/or peripheral blood (PB) samples. ZAP70 expression was assessed by immunohistochemistry/ flow cytometry. Somatic hypermutation was performed by Sanger sequencing. Deviation from the germline above 2% was considered mutated. Interphase multi-color FISH studies were performed using probes targeting regions 11q22.3 (ATM), centromeric region of chromosome 12, 13q14.3 (D13S319 locus),13q34 (LAMP1) and 17p13.1 (TP53), Clinical data were correlated with molecular and cytogenetic findings.

We reviewed 503 CLL cases in our laboratory database that underwent next-generation sequencing based multi-gene profiling. 44 (8.8%) patients had mutations including BRAF alone (n=18), KRAS alone (n=14), NRAS alone (n=8) and dual BRAF +KRAS/NRAS (n=4), respectively. We compared these 44 cases with 75 consecutive CLL/SLL cases without mutations in KRAS/NRAS/BRAF (CLL^raswt). CLL^rasmut showed a male predilection compared to the CLL^raswt (M:F=3.9:1 versus 1.2:1, p =0.01) with a median age of 62 years (range, 42-80). There were no significant differences between CLL^rasmut and CLL^raswt in the median age of disease onset, PB counts, PB/BM tumor burden. Studies for somatic hypermutation showed unmutated IGHV in 79.4% of CLL^rasmutversus 55.4% in CLL^raswt (p =0.03). ZAP70 was positive in 73% of CLL^rasmut compared to 42.9% in CLL^raswt (0.05). TP53 was the most frequent concurrently mutated gene, noted in 22.7% (10/44) of CLL^rasmut cases. CLL^rasmut showed higher frequency of trisomy 12 (54.6% versus 30.7%, p =0.0001) and lower frequencies of D13S319 deletions (31.8% vs. 82.7%, p =0.0001) compared to CLL^raswt. In a significant proportion of cases, the mutant allelic frequencies of RAS/RAF mutations were disproportionately lower than predicted based on the tumor burden. Interestingly, 15 out of 22 (69%) BRAF mutations involved codons other than V600.

We conclude that BRAF/KRAS/NRAS mutations are seen in about 9% of CLL and could represent a distinct subset of CLL due to their frequent association with male gender, trisomy 12 and unmutated IGHV. These findings may provide additional insight into oncogenesis and guide design of potential targeted therapies in CLL.