Developing Next Generation Sequencing (NGS) Panels for Uveal Melanoma to Detect Copy Number and Single Nucleotide Variants; comparing PCR and Hybrid Capture Target Enrichment.

Abstract

Purpose : Recent data demonstrate the clinical importance of integrating somatic copy number variations (CNV) and mutational status when determining prognosis for uveal melanoma (UM) patients. To address this need we developed two custom-designed NGS panels (Illumina and Agilent) and tested them in a preliminary sample cohort. We assessed their efficacy to detect CNV in chrom. 3, 6 & 8 and recurrent mutations in GNAQ, GNA11, BAP1, SF3B1, EIF1AX, which are associated with UM development and metastasis.

Methods : All UM samples were from consenting patients (8 males/6 females; median age at diagnosis, 60yrs) and the study was ethically approved. Primary treatment was enucleation in 12 and local resection in 2. DNA was extracted from 14 snap frozen UM samples; all 14 had previous CNV data and 8 had mutation data from TCGA analyses. Target enrichment was performed for all tumours and 2 reference standards using Agilent SureSelect (SS) and Illumina TruSeq Custom Amplicon (TSCA), according to manufacturer’s instructions. Variants were identified and annotated with SnpEff version 3.2a and Integrative Genomics Viewer.

Results : The SS and TSCA panels successfully detected CNV in chrom. 3, 6 and 8 in 13/14 (93%) and 11/14 (79%) cases, respectively. The SS panel provided good coverage and only one sample failed to produce results for both CNV and mutation analyses. Mutation data from this panel was concordant with TCGA data except that an additional BAP1 mutation was detected by both SS and TSCA in one case. Two samples had differing chrom. 3 classification, with one tumour possibly being explained by isodisomy not detected by MLPA. The TSCA demonstrated 100% concordance for all CNVs. However, 3 samples failed to produce discernible copy number results, and mutation data were concordant with the SS panel with the exception of a BAP1 mutation.

Conclusions : Our data show that it is possible to develop a custom-designed NGS panel to detect CNV and mutations in UM. Differences were noted in the performance of the individual panels, and thus an additional cohort of UM samples are required for further testing.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.