Abstract
Abstract 1817
The two most frequent etiological translocations in multiple myeloma (MM) are t(4;14), which deregulates FGFR3 and MMSET and has a poor outcome and t(11;14) which directly deregulates cyclin D1 and has an indolent course. The t(11;14) is present at 10–15% in both MGUS and MM but the t(4;14) is in only 3–4% of MGUS compared to 11% in MM. Consequently it is thought that patients with a t(4;14) have a less stable disease which progresses more quickly to myeloma than other subtypes. In order to address the hypothesis that cases with the t(4;14) are more prone to acquire mutations and so progress, we have compared the number and mutational spectrum of cases with these two variants.
DNA was extracted from CD138-selected bone marrow cells from 10 t(4;14) and 12 t(11;14) cases of newly presenting MM. 50 ng of genomic DNA was used to capture the exome using the SureSelect Human All Exon 50Mb target enrichment set (Agilent). We have previously validated this approach and shown it to have parity with approaches using larger starting amounts of DNA. Libraries were prepared from tumor and non-tumor DNA from the same patient and sequenced using 76 bp paired end reads on a GAIIx (Illumina). Samples were sequenced to a median depth of 61x, with 99% >1x and 85% >20x exomic coverage. Following base calling and quality control metrics the raw fastq reads were aligned to the reference human genome. The Genome Analysis Tool Kit was used to call indels and single nucleotide variants (SNVs), with BreakDancer used to detect structural variants. These variant calls were recalibrated and soft filters applied to remove potential false-positives using dbSNP, HapMap and the thousand genomes project as truth sets. Variants that occurred in both the normal and tumor samples were filtered out and the tumor-specific variants were annotated using Annovar. As well as the identification of commonly affected genes, functional annotation enrichment analysis was used to identify commonly affected pathways.
The group of 22 cases sequenced at the exome level showed a mutation spectrum that comprised 32,000 SNVs and 1,800 indels per patient, with 1,600 SNVs and 500 indels in the tumor sample only. Structural and copy number variants inferred from this data were also identified and validated previous results using other technologies. We identified 250 SNVs and indels, per patient, that were not in dbSNP and constitute tumor-acquired mutations. We were able to validate some of these mutations that we had previously analyzed using other platforms (98% concordance). The Ti/Tv ratio of mutations was not consistent with any specific exposure or mechanism. The distribution of indels was biased toward insertions rather than deletions, with both predominantly being multiples of three to produce in-frame mutants. In total sequencing data from 60 exomes is available and pathway analysis of the SNVs newly identified confirmed the deregulation of pathways previously identified as being mutated in myeloma, in addition we also identify novel deregulated genes and pathways.
We note a consistent increase in the number of variants in the t(4;14). Each tumor had on average 60 non-synonymous SNVs per sample with a range of 29 to 101, some patients being clear outliers. There was a bias to an excess number of mutations within the t(4;14) group which did not reach statistical significance. Importantly, the overlap between the SNVs in individual patients was limited with few consistently mutated genes across the sample set as a whole. In contrast, pathway analysis of the genes mutated in these two different entities shows marked similarities, with more frequent involvement of genes mediating cell adhesion in the t(4;14)s. Although the t(4;14) group had a greater number of mutations, a larger number of genes were affected in the t(11;14) group with the number of mutated genes in two or more samples being 111 versus 237, respectively. This observation implies a more consistent group of genes are deregulated in the t(4;14) group, suggesting that they are under greater selective pressure than in the t(11;14) group.
In this work we show a higher mutation frequency but with more limited numbers of genes affected in the t(4;14) group compared to the t(11;14) group. Overall, the data are consistent within the two etiologically distinct groups of MM having a similar spectrum of mutations driving disease progression, with a focus on pathway deregulation rather than any single gene.
Disclosures:
No relevant conflicts of interest to declare.
Genome Expression Pathway Analysis Tool – Analysis and visualization of microarray gene expression data under genomic, proteomic and metabolic context