1283P - Non-invasive detection of lung cancer by identifying copy number aberrations in circulating cell-free DNA with next generation sequencing to aid ea...

Date 09 September 2017
Event ESMO 2017 Congress
Session Poster display session
Topics Cancer in Adolescents
Non-Small-Cell Lung Cancer, Early Stage
Lung and other Thoracic Tumours
Translational Research
Presenter Fiona Taylor
Citation Annals of Oncology (2017) 28 (suppl_5): v453-v456. 10.1093/annonc/mdx381
Authors F. Taylor1, J. Bradford2, P. Woll1, D. Teare3, A. Cox4
  • 1Academic Unit Of Oncology, University of Sheffield Medical School, S10 2RX - Sheffield/GB
  • 2Academic Unit Of Molecular Oncology, University of Sheffield, S10 - Sheffield/GB
  • 3School Of Health And Related Research, University of Sheffield, Sheffield/GB
  • 4Academic Unit Of Molecular Oncology, University of Sheffield, S10 - RX/GB

Abstract

Background

Population stratification with molecular biomarkers could improve the cost-benefit of lung cancer screening programmes and reduce false positives. We aim to establish that somatic copy number aberrations (SCNA) are detected in circulating cell-free DNA (cfDNA) of lung cancer cases. We hypothesise that the number and magnitude of SCNA might also serve as a discriminative test to aid early lung cancer detection.

Methods

Standard protocols were followed to process matched cfDNA and lymphocyte DNA for 51 untreated lung cancer cases, 30 high risk controls and 10 low risk controls. Low coverage DNA sequencing was carried out on the Illumina Hiseq 2500 and read copy number profiles were established with the software CNAnorm. A genomic instability score was evaluated by defining the area under the receiver operating characteristic curve (AUROC).

Results

The median coverage of the genome for cfDNA was 0.49X (range 0.2X-0.63X). There was no significant difference between the median whole genome copy number aberration (CNA) score for early stage lung cancer and high risk controls, 398 (117-15373) vs 252 (149-7122) p = 0.25. The AUROC was 0.60 (95% CI 0.47-0.78) for early stage cancer (N = 21) and high risk controls compared to an AUROC of 0.74 (95% CI 0.63-0.85) for all lung cancer cases. The CNA score was a significant prognostic factor in univariable (HR 1.22 (95% CI 1.05-1.42 p = 0.008) but not multivariable analyses (HR 0.91 (95% CI 0.73-1.15 p = 0.45).

Conclusions

Our preliminary results demonstrate non-invasive detection of tumour derived copy number alterations with low coverage whole genome sequencing. The CNA score is not recommended as a stand-alone test to aid early lung cancer detection.

Clinical trial identification

NA

Legal entity responsible for the study

Fiona Taylor

Funding

Cancer Research UK and Yorkshire Cancer Research UK and The Roy Castle Foundation UK

Disclosure

All authors have declared no conflicts of interest.