Deep-Sequencing Reveals the Molecular Landscape of Splenic Marginal Zone Lymphoma: Biological and Clinical Implications

Whilst the spectrum and clinical significance of gene mutations and immunogenetic features in common mature B-cell malignancies is well established, their incidence, biological and clinical importance in splenic marginal zone lymphoma (SMZL) remains uncertain. Accordingly we screened 175 well-characterised SMZL cases for mutations in 768 genes (Haloplex Target Enrichment System) with a known or postulated role in B-cell physiology, B-cell malignancies in general and SMZL pathophysiology in particular.

After sequencing (HiSeq 2000) we achieved a mean depth across our gene panel of 297-fold (range 128-702x), with more than 85% of all bases covered at >50-fold. After comprehensive filtering, 1,374 single nucleotide variants and insertions/deletions were identified. 221 genes were recurrently mutated at a gene frequency of 2-16% [n=2-28]. Sanger sequencing confirmed 86/86 selected variants in our recurrent genes, and showed 99% concordance between our Haloplex and Sanger sequencing of NOTCH2 exon 34, which was performed in all patients. Comprehensive validation of both germ-line Haloplex [n=18 patients] and Sanger sequencing established the sensitivity and specificity of our approach, and confirmed the biological importance and somatic origin of the genes described herein.

To identify biologically relevant genes, we employed MutSigCV analysis, an algorithm that identifies significantly mutated genes by accounting for background mutation rate, DNA replication time and the gene size. 18 mutated genes were identified with TP53 [n=28], KLF2 [n=21], MYD88 [n=12], NOTCH2 [n=17], TNFAIP3 [n=13] and CCND3 [n=15] being the most significant; genes that encode components of pathways important in the regulation and differentiation of mature B-cells were also identified, including CREBBP [n=9], MAP3K6 [n=5], KDM2B [n=7], SETD1B [n=6], TRAF3 [n=9], ARID1A [n=10], BIRC3 [n=3], BCL10 [n=5], BTG1 [n=3], ATM [n=10], NFKBIE [n=4] and DDI1 [n=4]. Then, we searched for significant pairwise gene correlations and mutually exclusive relationships between our mutated genes demonstrating the following: (1) independent events, such as MYD88, where a mutation is invariably observed as an isolated event; (2) cancer drivers that have a similar proportion of co-occurring and mutually-exclusive relationships, such as NOTCH2, TP53, TNFAIP3 and CREBBP, and (3) genes such as KLF2, CCND3 and ARID1A that have proportionally more co-occurring relationships, thus suggesting a synergistic function to promote tumorigenesis. Finally, we studied clonal evolution, by differentiating between early, clonal events, and later, subclonal mutations (ABSOLUTE algorithm), and we were able to classify clonal or subclonal mutation in 6/18 of our MutSigCV genes. Paradigmatically, we observed that all the CREBBPmutations were fully clonal.

Amongst our most novel findings was KLF2, or Krüppel-like factor 2, mutations that were distributed across the entire protein, with a cluster in the C2H2 domain and were all somatically acquired. All mutations tested were clonal, significantly associated with del(7q) (P=0.001), IGHV1-2*04 gene usage (P<0.001) and other gene mutations including NOTCH2 (all P<0.001). Together, these observations suggest that the potential cell survival advantage provided by an early KLF2 mutation allows the acquisition of additional functionally synergistic gene mutations to promote tumourigenesis.

Genes associated with reduced time to first treatment (TTFT) included KLF2 (HR 1.93, 95%CI 1.16-3.32, p=0.01), and NOTCH2 (HR 2.13, 95%CI1.26-3.58, p=0.003). TP53 mutations were associated with shorter event-free (EFS) and overall survival (OS) (HR 2.17, 95%CI 1-4.74, p=0.05 and HR 2.16, 95%CI 1.05-4.42, p=0.032, respectively). Using multivariate Cox proportional hazard analysis, which included base-line clinical varaibles, both NOTCH2 (HR 2.12, 95%CI 1.02-4.4, p=0.044) and 100% germline IGHV gene identify (HR 2.19, 95%CI 1.05-4.55, p=0.036) were independent risk factors for TTFT. Furthermore, the presence of TP53 mutation was an independent risk factor for OS (HR 2.16, 95%CI 1.05-4.43, p=0.032).

In summary, we have defined the mutational landscape in a large cohort of patients with SMZL and identified novel recurrent mutations, especially involving KLF2. Importantly we show for the first time, that gene mutations and immunogenetic features have independant prognostic significance.