EXOME SEQUENCING POINTS TO DIFFERENCES IN GENETIC INSTABILITY LEVEL IN MGUS COMPARED TO MM

Background
Malignant transformation of normal to tumour cells is a multistep process followed by sequential aggregation of hits at different molecular levels. Genetic events including single nucleotide variants (SNVs), insertion-deletion changes (indels) as well as copy number alterations (CNAs) affect the phenotype of the tumour population and consequently patient prognosis. Transformation from a symptomless state, monoclonal gammopathy of undetermined significance (MGUS) to multiple myeloma (MM) can be used as a unique model for cancer development studies. To date, there is very little data regarding the mechanisms leading to disease progression at molecular level.

Aims

To perform exome sequencing together with SNP array analysis in MGUS patients to describe the premalignant phenotype and compared these to advanced tumour cells at the DNA level.

Methods
Overall, 33 and 69 MGUS patients were included in a WES and SNP array study, respectively. Plasma cells (PC) were isolated from bone marrow by FACSAria (BD Biosciences) system using CD138, CD19 and CD56 markers to obtain a pure abnormal PC population with a purity >90%. For WES, NEBNext kit and SureSelect Human All Exon V5 (Agilent) were used, samples were sequenced by HiSeq2000 (Illumina). Copy number alterations (CNAs) were tested by SurePrint G3 CGH+SNP, 4x180K (Agilent). Results were compared to 463 and 91 MM patients analysed by WES and SNP arrays, respectively.

Results
CNAs and acquired somatic gene mutations (SNVs) were detected in 68% (47/69) and 100% (33/33) of MGUS patients in comparison to 100% (91/91, p<10^-4) and 100% (463/463) of MM patients, respectively. However, the overall number of both CNAs and SNVs per patient was significantly lower in MGUS (CNAs: median 2, range 0-15; SNVs: median 89, range 9-315) than in MM (CNAs: median 16, range 2-49, p<10^-18; SNVs median 123, range 1-897, p<10^-4). Non-synonymous SNVs (NS-SNVs) were present in 97% (32/33) cases with a median 19 (range 0–70) NS-SNVs per patient. Overall, 42 genes were recurrently mutated in at least 2 patients and 2 non-large protein coding genes were mutated in at least 3 cases includingKLHL6 and NPIPL2. We identified 7 genes which were significantly mutated in MM in our previous study including KRAS(n=2), HIST1H1E (n=2) and NRAS, DIS3, EGR1, LTB, PRKD2 (all n=1). IGH translocations were identified in 27% (9/33) of patients: t(11;14) in 12% (4/33), t(4;14) in 9% (3/33), t(14;16) in 3% (1/33) and t(14;20) in 3% (1/33). As previously described in MM, only one type of IGH translocation was found per patient and all 9 cases with IGH translocation did not have additional hyperdiploidy. We did not find any translocations involving MYC (8q24.21) or the light chain loci IGK (2p12) and IGL (22q11.2) and any mutations in TP53, ATM, ATR and ZNFH4 genes involved in DNA repair pathway alterations which were identified as unfavourable factors to MM patients survival.

Summary
We have performed the first comprehensive analysis of MGUS patients using exome sequencing together with SNP arrays and described the main genetic events that are already present in this premalignant state. We proved that complex genetic instability is formed before clinical manifestation first at the gene level then at the chromosome level. Then, a number of random genetic hits increases to form a landscape for significant oncogenic hits driving the transition to a malignant state. This study was supported by grants IGA MHCZ NT13492, OPVK CZ.1.07/2.3.00/20.0183.

Keyword(s): MGUS, Mutation analysis

www.eha.org