Start Your Search
OA 10 - Liquid Biopsy for Genomic Alterations (ID 678)
- Event: WCLC 2017
- Type: Oral
- Track: Advanced NSCLC
- Presentations: 2
- Moderators:Adrian G. Sacher, Pasi A Jänne
- Coordinates: 10/18/2017, 11:00 - 12:30, F201 + F202 (Annex Hall)
OA 10.03 - Liquid Biopsy in the Lung Cancer Clinic: A Prospective Study of Plasma DNA next Generation Sequencing to Guide Matched Therapy (ID 8218)
11:00 - 12:30 | Author(s): D. Stephens
Liquid biopsy for plasma circulating tumor DNA (ctDNA) next generation sequencing (NGS) is now commercially available and increasingly adopted in clinical practice with a paucity of evidence based guidance. We set out to prospectively determine the utility of plasma ctDNA NGS in the lung cancer clinic.
Patients (pts) with advanced NSCLC who were driver unknown or resistance mechanism unknown were eligible. Pts were enrolled prospectively at Memorial Sloan Kettering (NY, USA) and Northern Cancer Institute (Sydney, Australia). Peripheral blood was collected in Streck tubes (10-20mL) and sent to Resolution Bioscience (Bellevue, WA) for targeted NGS of extracted DNA using a bias corrected hybrid capture 21 gene assay in a CLIA laboratory with unique reads at 3000x and sensitive detection at variant allele frequency above 0.1%. Clinical endpoints included detection of oncogenic drivers, turnaround time, comparison to tissue NGS when available, and ability to match pts to targeted therapy along with their treatment outcomes.
Seventy-six pts were prospectively accrued. Plasma NGS detected an oncogenic driver in 36% (27/76) of pts, of whom 14% (11/76) were matched to targeted therapy; including pts matched to clinical trials for HER2 exon 20 insYVMA, BRAF L597Q and MET exon14. Of the 10 evaluable pts, 10 partial responses were observed. Mean turnaround time for plasma was 6 days (3-12) vs 21 days (16-30) for tissue (P <0.0001). Plasma ctDNA was detected in 60% (46/76) of pts; detection rate was 46% (16/35) if blood was drawn on active therapy and 73% (30/41) if drawn off therapy, either at diagnosis or progression (Odds ratio 0.31, 95% CI 0.12 – 0.81; P=0.02). Of the 25 concurrent tissue NGS performed to date, there was a 96% plasma concordance with tissue and a 60% tissue concordance with plasma for driver mutations.
In pts who were driver or resistance mechanism unknown, plasma NGS identified a variety of oncogenic drivers with significantly shorter turnaround time compared to tissue NGS, and matched patients onto targeted therapy with clinical benefit. Plasma ctDNA is best detected at diagnosis of metastatic disease or at progression. A positive finding of an oncogenic driver in plasma is highly specific and can immediately guide treatment, but a negative finding may still require tissue biopsy. Our findings provide evidence to support the incorporation of plasma NGS into practice guidelines.
Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.
OA 10.05 - Non-Invasive Molecular Profiling in NSCLC by Targeted and Whole Exome Analysis of Plasma cfDNA (ID 10422)
11:00 - 12:30 | Author(s): D. Stephens
Molecular characterization of tumor can guide the choice of therapy for NSCLC patients. However, tumors are complicated by spatial heterogeneity and sometimes may not be of sufficient quality and quantity for analysis. NGS using plasma cell-free DNA (cfDNA) input may capture temporal and spatial heterogeneity, and enable genomic profiling in patients without adequate available tumor tissue. Targeted gene panels allow for robust detection of known oncogenic drivers, but may not be comprehensive enough to screen for novel biomarkers or mechanisms of acquired resistance. Whole exome sequencing (WES) allows for hypothesis-free biomarker discovery, but may be technically challenging in the setting of limited tumor-derived DNA content in plasma cfDNA. In this study, we aim to develop a workflow to guide the selection of samples for targeted and whole exome sequencing for noninvasive molecular profiling.
Plasma samples were collected from 20 NSCLC patients receiving a variety of treatment (chemotherapy, targeted therapy, or immunotherapy). Most patients (>70%) had stage III or IV disease at the time of plasma collection. CfDNA was extracted from 3 mL of plasma, and analyzed using low-pass shallow whole genome sequencing (sWGS) and MSK-IMPACT, a hybridization capture-based assay targeting over 400 cancer-related genes. Analysis of matched normal was performed for somatic variant calling.
Median cfDNA yield per plasma sample was 28ng (range 7 - 236ng). We applied z-score statistics to estimate the levels of tumor-derived mutant allele fractions in cfDNA based on sWGS data. We trained the algorithm using a separate cohort of cfDNA data from >100 patients with metastatic solid tumors to classify samples by mutant allele fraction (MAF) as either low (<5% MAF) or high (>5% MAF) tumor-derived DNA. In the subset of 10 patients with unknown drivers, two were estimated to have MAF >5% in cfDNA, and WES recapture was performed. MSK-IMPACT targeted sequencing identified actionable alterations in a subset of patients who did not have sufficient materials for tissue profiling. WES in cases with high tumor-derived DNA content by sWGS identified alterations in genes outside of the MSK-IMPACT panel.
Molecular profiling using cfDNA is feasible in lung cancer and may identify actionable alterations to inform treatment decisions in patients without sufficient tissue for molecular characterization. The application of sWGS to estimate the levels of tumor-derived mutant allele fractions in plasma cfDNA samples may help guide selection of the optimal downstream sequencing strategy.