Abstract 24: Intratumoral genomic heterogeneity of primary pulmonary adenocarcinoma in never smokers

Abstract

Abstract Aim: Intratumoral genomic heterogeneity challenges personalized lung cancer care, especially where it relies upon small diagnostic samples. To explore genomic representation provided by tumor subsampling, we performed whole genome sequencing (WGS) of multiple regions of individual primary pulmonary adenocarcinomas (LUAC). Methods: An observational study was performed on three cases of never-smoking LUAC resected with curative intent. Post-diagnostic residual fresh tumor was procured with informed consent, along with constitutional samples from normal lung or blood. Selection criteria included: histologically confirmed LUAC; never-smoker [defined as fewer than 100 cigarettes consumed in a lifetime]; and no prior malignancy, cytotoxic therapy or thoracic radiotherapy. Tissue samples were procured by an anatomical pathologist and research scientist and snap frozen within 60 minutes of devascularization, then stored at -80 degrees celsius. Nine macrodissected subsamples met quality criteria of >40% tumor cellularity and <20% necrosis (Case 1, 4 regions; Case 2, 3 regions; Case 3, 2 regions) as assessed visually by 2 anatomical pathologists. High molecular weight DNA was extracted from tissue using Qiagen AllPrep DNA/RNA Mini Kit and from blood using Qiagen Blood Maxi Kit. WGS was performed on paired end libraries using Illumina's HiSeq 2000 platform to 80x (tumor), 40x (normal lung) and 30x (blood) coverage. Reads were aligned to GRCh37 with BWA-MEM. Duplicates were removed using Picard and local INDEL realignment and base quality recalibration were performed with GATK. Single nucleotide variants (SNVs) were called by MuTect, Varscan, Strelka and SomaticSniper. Variants were considered ‘high priority’ if predicted by SNPEff to have ‘moderate’ or ‘high’ functional significance. Structural variants were detected from WGS data using Breakdancer and Pindel. Sample genotyping was performed using Illumina's HumanOmni2.5-8 array and used to call copy number variations (CNVs) using the Genome Alteration Print tool. Results: All cases were Caucasian females. Case 1 consisted of a 37 year old with a well to moderately differentiated pathological stage IV (AJCC 7th Edition; T4 N1 M1a) tumor 75mm in maximal dimension for which DNA from 4 tumor regions and whole blood was available. Case 2 was an 80 year old with a 24mm, acinar predominant, moderately differentiated pathological stage 1A (T1b N0 M0) tumor for which DNA from 3 tumor regions and whole blood was available. Case 3 was an 82 year old with a 35mm, acinar predominant, pathological stage 1B (T2a N0 M0) tumor for which DNA from 2 tumor regions and non-tumor lung was available. Mean tumor cellularity (and mean sequencing coverage achieved) for regions 1, 2, 3 and 4 for case 1 were 50% (98x), 50% (100x), 73% (99x) and 58% (134x), respectively. Similarly, for regions 1, 2 and 3 of case 2, mean cellularity (and coverage) was 45% (93x), 45% (114x) and 40% (93x), respectively. Case 3 demonstrated 45% (107x) and 55% (97x) mean cellularity (and coverage) for regions 1 and 2, respectively. Less than 10% necrosis was observed in all tumor regions. Of 10275 SNVs detected in case 1, 3198 (3198/10275, 31%) were found in all 4 subsamples. 6911/15689 (44%) and 5595/9528 (59%) were shared among all subsamples in cases 2 and 3, respectively. The numbers of SNVs unique to each region relative to total SNVs observed for each region in case were: 869/5999 (14%), 1129/6437 (18%), 914/6969 (13%) and 517/5936 (9%). Similarly, the numbers of unique SNVs as a proportion of total SNVs for each region in case 2 were 1148/9835 (12%), 2556/11404 (22%) and 2632/10714 (25%); and for case 3 were 2293/7888 (29%) and 1640/7235 (23%). In case 1, 7 of 303 (2%) high priority variants were detected in all regions. Similarly, 44/303 (15%) and 29/302 (10%) high priority variants were detected in all tumor regions for case 2 and 3, respectively. Conclusion: Significant intratumoral heterogeneity was observed. These findings have significant implications not only for diagnostic testing of lung cancer but also for clinical trial design. Prospective clinical trials incorporating assessment of both geographic and temporal intratumoral heterogeneity will help explore the implications of this phenomenon on patient treatment. Acknowledgements: We acknowledge the patients, nurses, and staff of The Prince Charles Hospital for their contributions to this project. Funding: MD supported by Cancer Council Queensland and NHMRC PhD Scholarships. Supported by funding from NHMRC, Cancer Australia, TPCH Foundation, Queensland Health, Cancer Council Queensland. Citation Format: Marissa G. Daniels, Lutz Krause, Jonathan J. Ellis, Ian A. Yang, Rayleen V. Bowman, Vandana Relan, Kelly Chee, Felicia Goh, Brielle Parris, Leanne Morrison, Maria Martins, Linda Passmore, Elizabeth McCaul, Deborah Courtney, Edwina Duhig, Morgan Windsor, Rishendran Naidoo, Kwun M. Fong. Intratumoral genomic heterogeneity of primary pulmonary adenocarcinoma in never smokers. [abstract]. In: Proceedings of the AACR Precision Medicine Series: Integrating Clinical Genomics and Cancer Therapy; Jun 13-16, 2015; Salt Lake City, UT. Philadelphia (PA): AACR; Clin Cancer Res 2016;22(1_Suppl):Abstract nr 24.