Comparing genomic landscape of early stage, treatment naïve and late stage, drug resistant EGFR-mutant lung adenocarcinomas

Uncategorized
  1. Rahul Nahar1,
  2. Yin Yeng Lee1,
  3. Alexis J. Khng1,
  4. Tong Zhang2,
  5. Angela Takano3,
  6. Xingliang Liu1,
  7. Jacob J.S. Alvarez2,
  8. Ori Zelichov4,
  9. Ezra Ella4,
  10. Zohar Barbash4,
  11. Chong Hee Lim5,
  12. Tina P.T. Koh5,
  13. Zaw Win Aung6,
  14. Tony Kiat Hon Lim3,
  15. Chee Keong Toh7,
  16. Wan-Teck Lim7,
  17. Bing Lim8,
  18. Wai Leong Tam8,
  19. Eng-Huat Tan7,
  20. Weiwei Zhai2,
  21. Daniel S.W. Tan7,8,9,
  22. Axel M. Hillmer1,10

Authors Affiliation(s)

  • 1Cancer Therapeutics and Stratified Oncology, Genome Institute of Singapore, SINGAPORE
  • 2Human Genetics, Genome Institute of Singapore, SINGAPORE
  • 3Department of Pathology, Singapore General Hospital, SINGAPORE
  • 4NovellusDx (IL) Ltd, Jerusalem, ISRAEL
  • 5Department of Cardiothoracic Surgery, National Heart Centre Singapore, SINGAPORE
  • 6Divisions of Clinical Trials and Epidemiological Sciences, National Cancer Centre Singapore, SINGAPORE
  • 7Medical Oncology, National Cancer Centre Singapore, SINGAPORE
  • 8Cancer Stem Cell Biology, Genome Institute of Singapore, SINGAPORE
  • 9Cancer Therapeutics Research Laboratory, Division of Medical Sciences, National Cancer Centre Singapore, SINGAPORE
  • 10Institute of Pathology, University Hospital Cologne, GERMANY

Can J Biotech, Volume 1, Special Issue-Supplement, Pages 246-247, DOI: https://doi.org/10.24870/cjb.2017-a231

Presenting author: rahul.nahar@gmail.com

Abstract

While, the genomic landscape of early stage, treatment naïve lung adenocarcinomas (LUADs) has been described quite elaborately in recent literature, the genomic profile of late stage, drug resistant tumors remains largely unknown. Further, most of the published studies are based on smoker dominated Caucasian cohorts and EGFR-mutant LUAD remains under-represented in them. Despite response rates of upto 70% to EGFR tyrosine kinase inhibitors (TKIs), resistance ensues in most of these EGFR-mutant patients, limiting responses to a median of 10-12 months. Thus, to better understand the evolution of these tumors in context of drug resistance, we perform a comparative analysis of the mutational and copy number landscape of early stage, treatment naïve vs late stage, resistant tumors. Whole exome sequencing was performed on: (i) 100 tumor sectors from 24 early stage, treatment naïve EGFR-mutant LUAD cases. (ii) 81 biopsies from 58 late stage, TKI and chemotherapy resistant cases. Copy number analysis using SNP arrays was performed for a subset of these patients. The significantly higher mutation burden in the late stage, drug resistant tumors elucidated a driver mutation landscape beyond just recurrent TP53 mutations, which was dominated by PIK3CA (14%), RB1 (10%), NF1 (7%) and other rare mutations in EGFR (in 5/58 cases), many of which co-occurred with the T790M mutation. Functional studies validated the oncogenicity of some of these rare mutations in the PI3K/AKT1 pathway. The copy number landscape revealed pervasive, truncal genome doubling events in both cohorts (~80% cases). While comparable fraction of genome was affected by overall copy number gains or losses (copy change >=1) across the two cohorts (49.2% vs 46.3%, P=0.51), significantly higher fraction of genome was affected by amplifications (copy change >=2; 8.7% vs 5%, P=0.02) and loss of heterozygosity (LOH; 33.1% vs 20.6%, P=0.003) in the drug resistant tumors compared to the treatment naive tumors. In summary, our study reveals (i) increased mutation and driver burden with co-occurring resistance mutations and (ii) higher fraction of genome with amplifications and LOH in the drug resistant cohort, suggesting ways in which genomic landscape of EGFR-mutant LUAD evolves, potentially making them more tolerant to drug treatment.