Introduction: Hairy cell leukemia and some other low-grade B-LPDs are characterized by a high incidence of oncogenic mutations in the Ras-MAPK pathway, most commonly in BRAF and MAP2K1 genes. In contrast, chronic lymphocytic leukemia (CLL) has a low frequency of such mutations. To determine the patterns of emergence of Ras pathway mutations in CLL, we assessed the frequency and timing of mutations in KRAS, NRAS, BRAF and MAP2K1 and their association with other CLL-associated mutations and B-cell receptor (BCR) type. We also compared expression profiles for signaling molecules for CLL with or without these mutations.
Methods: A custom-designed 50-gene, Ampliseq/Ion-S5 next-generation sequencing (NGS) panel was performed on DNA from purified B cells for 221 CLL cases presenting for therapy at our institution. Panel included mutation hotspots for the genes above and all other common CLL-associated pathogenic mutations, with an analytic sensitivity of 1-2% VAF. For post-therapy samples demonstrating a Ras pathway mutation, prior/pre-therapy samples were tested to determine whether mutations were present at baseline or acquired with therapy (or likely so based on findings in off-therapy baseline samples). We performed transcript analysis on total RNA from highly pure CLL samples in a subset of baseline samples using the 770-gene Pan-cancer Immune Panel, which included many signaling and downstream AP1, MAPK, MAPKK and MAPKKK genes (nCounter System, Nanostring). Transcript normalization and clustering was performed in nSolver software with log-2 transcript levels compared for significance following Benjamini-Hochberg FDR adjustment. Pathway analysis was performed using R scripts with Bioconductor and Pathview visualization.
Results: Mutations in KRAS, NRAS, BRAF and MAP2K1 were identified in 32/221 CLL (14.4%), including pre-therapy/baseline in 17 cases. Ras pathway-mutated cases were IGVH-unmutated (U) in 29/32 cases (90.6%). Multiple (31.3%) and subclonal mutations (75%) were common with atypical sites of BRAF mutation predominating (Table). These BRAF mutations were mostly predicted to have minimally increased or decreased kinase activity. Particularly among cases where the mutations were likely or definitively present pretherapy, IGH variable (IGVH) gene sequencing revealed a striking predominance of cases expressing VH1-69/51p1 type (14/17, 82.4%, p<0.0001, Fisher exact). In contrast, cases likely or definitely arising post-therapy did not show any association with BCR stereotype. These most frequently arose following BTK inhibitor therapy (Table).
The case percentage of other co-occurring CLL-associated mutations included: BIRC3 6.3%, BCOR 21.9%, BCORL1 6.3%, NOTCH1 28.1%, SF3B1 25% and TP53 12.5% and 3.1% for FBXW7, MYD88, POT1, RPS15 and XPO1. There was a trend (not reaching significance) for lower TP53 and higher SF3B1 and NOTCH1 mutations than the overall data set. Cytogenetics revealed an increased proportion of +12 by karyotype and/or FISH in the Ras pathway-mutated cases (13/32, 40.6% versus 24.6% for other cases, p = 0.001). The co-association of +12 CLL requiring treatment with NOTCH1 mutations has been previously noted.
We compared RNA expression in the 770-gene panel for 11 baseline 1-69/IGVHU CLL (6 Ras pathway-mutated) and a control group of 6 baseline CLL expressing other IGVH genes. There were 50 genes that were significantly differentially expressed in the 1-69 group (adjusted p value <0.05) with 3 that distinguished Ras-mutated from unmutated 1-69/IGVHU CLL (IFNL1 and IL13 increased; TOLLIP decreased). In unsupervised clustering, Ras pathway-mutated cases clustered together except for 1 case. An analysis of differentially regulated BCR-associated genes showed coordinate shifts in expression in the components of the co-regulatory PIR-B and Leu13 complexes among Ras pathway-mutated as compared to the unmutated group
Conclusion: We identify Ras pathway mutations in CLL in two different contexts; prior to therapy in the VH1-69 subgroup of IGVHU CLL and separately emerging following therapy with BTK inhibitors. In both scenarios, multiple and subclonal mutations were common implicating Ras pathway dysregulation as a progression factor. The apparent complementation of Ras pathway mutations with the putatively antigen-driven unmutated VH1-69 BCR stereotype may indicate a unique pattern of regulated signaling in that CLL subgroup.
Lozanski:Beckman: Research Funding; BI: Research Funding; Stem Line: Research Funding; Genentech: Research Funding; Coulter: Research Funding; Novartis: Research Funding.
Asterisk with author names denotes non-ASH members.