103P - Prospective comparative study of next-generation sequencing on fine needle aspirations versus core needle biopsies in cancer patients included in SHIVA02 trial

Abstract

BackgroundHigh throughput molecular screening of solid tumours using core needle biopsies (CNB) allows the identification of actionable molecular alterations. Several reports revealed that success rate of genomic analyses based on CNB is around 70% with failures being mostly related to low proportion of tumour cells. Although several studies have demonstrated the utility of small biopsy specimens for molecular testing, there remains debate as to the sensitivity of CNB to detect actionable alterations compared to the less invasive fine needle aspiration (FNA). We aim to prospectively evaluate the ability of FNA to detect molecular alterations identified on CNB in cancer patients (pts) included in SHIVA02 trial. MethodsIn-house targeted sequencing amplicon based panel (Illumina TSCA 99.3Kb, 1,504 amplicons covering 87 genes) was used to identify pathogenic variants and gene copy number alterations (CNVs) in both CNB and FNA for pts enrolled in the SHIVA02 trial (NCT03084757). ResultsCNB and FNA specimen from 39 pts enrolled in the SHIVA02 trial were assessed. Main tumour locations were breast (17pts, 44%), pancreatic (4pts, 10%), and colorectal cancers (3pts, 8%). 91 somatic variants were identified in both specimens with a good correlation of variants’ allelic ratios (r: 0.772). CNV profiles of CNB and FNA were concordant. When taking into account molecular alterations validated during the molecular biology board, 88 alterations (55 variants and 35 CNVs) were reported among which 69 alterations (78%) where concordant between CNB and FNA. Among the 50 actionable alterations, only 12 (3 variants and 9 CNVs) (24%) were discordant between FNA and CNB. Main discordances were related to homozygous deletions and amplifications but false negative results were not related to FNA samples alone (5 CNVs in favour of FNA versus 4 in favour of CNB). ConclusionsComparative analyses of molecular alterations in CNB compared to FNA showed high concordance in terms of variants as well as CNVs identified. FNA could therefore be easily used in routine diagnostics workflow and clinical trials for tumour molecular screening with the advantage of being minimally invasive. Clinical trial identificationNCT03084757 SHIVA02; Trial release date: April 1, 2017. Legal entity responsible for the studyInstitut Curie. FundingMSD Avenir Fundation. DisclosureC. Le Tourneau: Advisory / Consultancy: MSD, BMS, Merck Serono, AstraZeneca, Novartis, Roche, Nanobiotix. All other authors have declared no conflicts of interest.