Exome-Wide Sequencing Reveals Oncogenic Mutations in Both Progressive and Non-Progressive MBL

While virtually all CLL cases are preceded by an asymptomatic phase of monoclonal B cell lymphocytosis (MBL), only 1-2% of MBL cases develop to CLL requiring chemotherapy annually. Specific determinants that lead to progression are unknown thus far, but are highly relevant, since this might contribute to a better understanding of CLL biology. Recently, the application of whole genome and exome sequencing using next generation sequencing (NGS) in CLL resulted in the discovery of novel genes including mutations in SF3B1 and NOTCH1, that could potentially play important roles in CLL leukemogenesis. Additionally NGS enabled in-depth examination of genetic alterations, and therefore has shed new light on clonal evolution of driver mutations which is considered a hallmark for progression. However, the role of genetic mutations and clonal evolution in the development of MBL and progression of MBL to CLL is still largely unknown and NGS studies in MBL samples and paired longitudinal samples is currently scarce. In this study we investigated the role of genetic mutations in the development of MBL and progression of MBL to CLL.

Thirty-six clinical MBL (CD19⁺ cell count: 0.5-5.0x10⁹/L) cases were ascertained through flow cytometric screening of a population based cohort of non-hematology patients identified via an absolute lymphocytosis on a routine complete blood count. From thirteen cases a follow-up sample was available obtained at an average time of 4.8 years after initial study enrollment. All cases had a detectable CLL-phenotype clone but did not meet iwCLL criteria for a diagnosis of CLL. Samples with clinical follow-up were characterized as progressive or non-progressive based on whether the CLL phenotype clone expanded at the time of follow-up. CLL-phenotype cells and non-malignant monocytes were purified using flowcytometry and DNA was extracted. We performed whole-exome sequencing using the Agilent solution-based system of exon capture, which uses RNA baits to target all protein coding genes (CCDS database), as well as ~700 human miRNAs from miRBase (v13).

A total of 76 exomes from 33 unique individuals were sequenced for this study. Six MBL cases were progressive and 27 MBL cases were non-progressive. Paired normal DNA was sequenced for all of these cases. On average, we achieved 30-40 fold coverage of the targeted regions. Over 90% of the targeted exons were covered at an average depth exceeding 3-fold. Sequencing reads were mapped to the reference human genome, and high-quality mismatches were classified as synonymous and nonsynonymous variants. We identified somatic mutations affecting 328 genes in at least one pair among these cases. Transitions comprised the majority of the somatic variants (P<10^-3, chi-squared test). On average, we observed 10 somatic alterations per sample (range 3-30). These variants were not present in publicly available data from normal controls including dbSNP135, the 1000 Genomes Project, and available exome sequencing data from healthy individuals without lymphoma. We further required each of the identified genes to have three somatically acquired and/or non synonymous rare events predicted to be functionally significant. A striking heterogeneity among these cases were observed with no gene mutation occurring at a frequency exceeding 15%. The most frequently mutated genes were LAMA3 and MLL3, each mutated in 13% of the cases. A number of known cancer-related genes were also mutated in MBLs at a frequency of roughly 5-8% including MLL, NOTCH2 and TP53. Acquired mutations in MADCAM1 were identified only at progression and not at diagnosis in selected progressive MBL samples.

This study represents an in-depth investigation of MBL exomes. Our data show a striking degree of heterogeneity in the genetic mutations that occur in MBLs. Mutations in known oncogenes and tumor suppressor genes occur commonly in MBL. Their presence does not indicate poor prognosis or a clear risk for progression to CLL. Our data indicate that individual genetic events are likely to be much less important than the sequential oncogenic mutations and combinations of specific genetic events for the progression of MBL to CLL. Thus, our work sheds new light on the genetic underpinnings and clonal diversity of MBL and provides a starting point for a better understanding of the genetic drivers and sequence of progression to CLL.