Poster Presentation 27th Lorne Cancer Conference 2015

Whole-exome sequencing of matched fresh and formalin-fixed paraffin embedded melanoma tumours: relevance for clinical decision making (#149)

Ricardo De Paoli-Iseppi 1 , Peter Johansson 2 , Marcel E Dinger 3 , Kerith-Rae Dias 3 , Gulietta M Pupo 4 , Hojabr Kakavand 1 5 , James S Wilmott 1 5 , Graham J Mann 1 4 5 , Nicholas K Hayward 2 , Alexander M Menzies 1 5 , Georgina V Long 1 5 , Richard A Scolyer 1 6 7
  1. Melanoma Institute Australia, Sydney, NSW, Australia
  2. Oncogenomics Laboratory, QIMR Berghofer Medical Research Institute, Brisbane, QLD, Australia
  3. Garvan Institute of Medical Research, Sydney, NSW, Australia
  4. Centre for Cancer Research, The University of Sydney at Westmead Millennium Institute, Sydney, NSW, Australia
  5. Discipline of Medicine, Sydney Medical School, The University of Sydney, Sydney, NSW, Australia
  6. Discipline of Pathology, Sydney Medical School, The University of Sydney, Sydney, NSW, Australia
  7. Tissue Pathology and Diagnostic Oncology Royal Prince Alfred Hospital, Sydney, NSW, Australia

The identification of recurrent driver mutations by whole-exome sequencing (WES) of fresh frozen human cancers and the subsequent development of novel targeted therapies has recently transformed the treatment of many cancers including melanoma. In routine clinical practice, fresh frozen tissue is rarely available and mutation testing needs to be performed utilising archival formalin-fixed, paraffin-embedded (FFPE) tissue in which DNA is typically fragmented, crossed-linked and of lower quality. It remains unresolved whether clinically-actionable WES data generated from genomic DNA (gDNA) extracted from FFPE tissues can be produced reliably. In this study of 10 melanoma patients, we compare WES data produced from analysis of gDNA isolated from FFPE tumour tissue with that isolated from fresh-frozen tumour tissue from the same specimen. FFPE samples were sequenced using the Illumina Nextera® and Roche NimbleGen SeqCap exome capture kits.

FFPE exomes were called utilising the matched fresh tissue sequence as a reference in order to examine variants between the two tissue sources and platforms. Of the 10 FFPE gDNA samples, 7 Nextera and 4 SeqCap samples passed library preparation quality control. On average, there were 5341 and 2246 variants lost in FFPE compared to matched fresh tissue utilising Nextera and SeqCap kits, respectively. In order to explore the feasibility of future clinical implementation of WES of archival samples, FFPE variants in 27 genes of interest for the melanoma mutation landscape were assessed. Due to sequencing errors and read failures, the average error rate was 16% in a total of 297 calls for the chosen genes using the FFPE gDNA. However, for the current clinically important melanoma mutations in BRAF and NRAS, 10/11 (91%) of FFPE calls were concordant with the fresh tissue result, which was confirmed using the Sequenom OncoCarta™ Panel.

FFPE WES therefore shows promising results in a limited number of important melanoma genes, but specialised library preparation to account for low quality gDNA and further refinements may be necessary before this approach could be used for routine clinical decision making.