Oral Presentation COSA 2015 ASM

Exploring the feasibility and utility of exome-scale sequencing of tumours in a clinical setting (#22)

Belinda Lee 1 2 3 , Arthur L Hsu 4 , Graham R Taylor 4 , Stephen Fox 1 , Andrew Fellowes 1 , Jennifer Mooi 1 , Jayesh Desai 1 3 , Ken Doig 1 4 , Paul Ekert 5 6 , Clara Gaff 2 , Dishan Gunawardana 3 7 , Anne Hamilton 1 8 , Paul James 1 , Andrew Roberts 2 , Ray Snyder 9 , Paul Waring 4 , Grant McArthur 1 , Ben Tran 3
  1. Peter MacCallum Cancer Centre, Parkville, VIC, Australia
  2. Walter & Eliza Hall Institute (WEHI), Melbourne, VIC, Australia
  3. The Royal Melbourne Hospital, Parkville, Melbourne, VIC, Australia
  4. Faculty of Medicine & Health Sciences, University of Melbourne, Melbourne, VIC, Australia
  5. The Royal Childrens Hospital, Melbourne, VIC, Australia
  6. Murdoch Childrens' Research Institute, Melbourne, VIC, Australia
  7. Department of Medical Oncology, Western Health, Melbourne, VIC, Australia
  8. The Royal Women's Hospital, Melbourne, VIC, Australia
  9. St Vincents Hospital, Melbourne, VIC, Australia

Background: 

The falling cost and increasing ease of access to genomic sequencing of tumours is gaining influence on clinical decision-making. Cancer genomics is approaching a tipping point where genomic sequencing enters routine clinical practice. However, evaluating the feasibility and performance of genomic technology, particularly sequencing of large numbers of genes in a clinical setting is not fully established.

Methods:

Between March 2014 and March 2015, the Victorian Comprehensive Cancer Centre (VCCC) explored the feasibility and utility of performing real-time genomic sequencing in 15 participants.  The performance of an exome-scale (ES) approach was compared against targeted sequencing (TS), alongside an assessment of fresh frozen (FF) tissue, versus formalin-fixed paraffin embedded (FFPE) tissue. A multi-disciplinary molecular tumour board (MTB) comprising scientists, pathologists, bioinformaticians and clinicians reviewed the genomic data to determine the clinical relevance of identified mutations and whether they were “actionable” and/or “druggable”.

Results:

All 15 cases were successfully analysed. Mean coverage was 100X for blood, 500X for FFPE and 1000-1500X for FF specimens. A total of 63 variants were identified by ES after filtering for genes documented to be relevant in cancer. Only 21% (13/63) were found using “hotspot” TS; 48 variants were unique, 17 were pathogenic, 9 possibly pathogenic and 32 were variants of unknown clinical significance. Mutations in TP53 (9/15 cases) were most frequent. Targeted sequencing detected a mutation in 60% (9/15 cases), whilst ES identified mutations in all (100%). Concordance between the platforms was complete. A median of 2 germline and 2 somatic variants were identified per participant. Our findings informed cancer management in 53% (8/15) of cases with results contributing towards either diagnostic, prognostic or treatment information; 47% (7/15) referred to the familial cancer clinic, and “druggable” targets identified in 53% (8/15).

Conclusion:

Genomic sequencing of tumour DNA is feasible and aided clinical decision making in 53% of cases.