ECMC cfdna consensus meeting State of the art for cfdna technologies 24 th November 2014
Applications of ctdna analysis for drug development Potential of ctdna analysis to: Identify the right patients for targeted therapies Selection of patients for clinical trials Companion diagnostic development Monitoring of response to treatment Earlier detection of response / relapse Chetan Bettegowda et al. Sci Transl Med 6, February 19, 2014 Identification of resistance mechanisms To guide combination therapy strategies and development of next generation of therapies
The Iressa label now includes provision to use ctdna testing to identify patients where tissue is not available IRESSA label in the EU based on a positive CHMP opinion based on the IFUM data set The supporting data Douillard et. al. (2014) Special warnings and precautions for use/assessment of EGFR mutation status Tumour samples which are used for the diagnosis of advanced NSCLC are the preferred sample type for EGFR mutation testing. A tumour sample should be collected and tested where possible. If a tumour sample is not available or evaluable, then circulating tumour DNA (ctdna) obtained from a blood (plasma) sample may be used. Only robust, reliable, sensitive test(s) with demonstrated utility on ctdna should be used for the determination of EGFR mutation status of ctdna. EGFR mutations identified in ctdna are highly predictive of EGFR mutation positive tumours). However it is not always possible to detect EGFR mutations using this sample type (0.2% false positives, 34.3% false negatives)
ctdna analysis - Technology landscape QPCR Established, commonly used, well understood CE-IVDs for use on tissue; IVD status for ctdna varies according to specific kit) Sensitivity~1%, individual assays vary in sensitivity Assays designed against common mutations Digital PCR approaches (e.g. Droplet digital PCR / BEAMing) Established in research setting, in trial/translational settings. Highly sensitive <0.05% Still hotspot assays, small sets of mutations. BEAMing not yet available outside central lab setting. NGS (Illumina / Ion Torrent / other?) Tremendous potential wide breadth panel assays, many genes in parallel from single plasma DNA sample Not as sensitive as digital approaches yet. 1-2% achievable, 0.1% challenging - ultrasensitive techniques need refinement and reduction to robust standard practice Others? None optimal for variety of reasons. Mass-Spec based mutation genotyping. Mutation enrichment methods (pre-ngs). 4 Author 00 Month Year Set area descriptor Sub level 1
Mutation detection in ctdna from NSCLC patient plasma to support the clinical development of AZD9291
EGFR TKI Therapy and Resistance in NSCLC EGFR mutation positive (EGFRm+) NSCLC is effectively treated by 1 st line EGFR-TKI s with response rates of approximately 70% and a median PFS of 10 months. However, nearly all patients who initially respond to EGFR-TKI therapy eventually develop acquired resistance and over half of EGFR-TKI resistant tumors carry a T790M mutation in exon 20 of EGFR. AZD9291 is an oral, selective, irreversible EGFR-TKI that targets the T790M resistance and original sensitizing mutations AZD9291 is currently in clinical testing in patients with EGFRm+ NSCLC who have relapsed on an approved EGFR-TKI therapy
Characterization of the resistance driver has required a fresh tumor biopsy EGFR T790M is rarely detected at initial diagnosis of NSCLC (<3%) Mok, T. S. et al N Engl J Med 2009; 361:947-957 Fukuoka, M. Et al J Clin Oncol. 2011;29(21):2866-74. A biopsy is needed post EGFR-TKI therapy to characterise the resistance mechanism. Patients whose tumors have progressed may be unable to provide an additional biopsy sample. Tumor DNA, present in the plasma of patients with NSCLC (ctdna) may offer a non-invasive opportunity to test for EGFR mutations where an evaluable tumor sample cannot be obtained. AZD9291 First Time In Patients Ascending Dose Study ( AURA )* A Phase I/II, Open-Label, Multicentre Study to Assess the Safety, Tolerability, Pharmacokinetics and Antitumour Activity of Ascending Doses of AZD9291 in Patients with Advanced Non Small Cell Lung Cancer who have Progressed Following Prior Therapy with an Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitor Agent (AURA). *NCT01802632
Plasma collections in the AZD9291 AURA clinical study to support ctdna evaluation Exploratory Biomarker objectives 1. Evaluate different technology platforms to enable sensitive and specific detection of EGFR mutations from patient plasma (T790M, L858R, Exon19del) 2. Evaluate concordance between mutation detected in plasma + tumor and relationships to clinical response to AZD9291 3. Evaluate mechanisms of resistance of AZD9291 using samples collected at progression Screening & pre-dose day 1 6 weeks 12 weeks At progression Tumor Plasma EGFRm+ patients treated with prior EGFR-TKI Relapse AZD9291 Relapse Progressive Disease
Initial evaluation of 4 different platforms for EGFR mutation detection from ctdna Gel electrophoresis (qualitative) Real-time PCR (indirect quantitation) Digital PCR (absolute quantitation) Pre-dose plasma samples collected from patients enrolled on the Phase I AZD9291 clinical study ( AURA ) Limit of detection: 5-20% 0.1-1% 0.02-0.1% Patient matched tumor re-biopsy material (FFPE), collected in 4 weeks prior to baseline plasma draw, was available for a subset of patients and genotyped by the Roche cobas platform using 3 central laboratories unless otherwise noted Allele-specific PCR Roche cobas 41 EGFR mutations FDA approved for tumor tissue (with erlotinib) Qiagen Therascreen 29 EGFR mutations FDA approved for tumor tissue (with afatinib) Digital PCR Sysmex Inostics BEAMing dpcr 22 EGFR mutations Emulsion PCR followed by flow cytometry MolecularMD droplet dpcr T790M, L858R, Exon19del Droplet digital PCR (ddpcr)
BEAMing & AS-PCR: Results from plasma samples from 72 patients in the AURA study in comparison to tissue results. High sensitivity and specificity for EGFR sensitizing mutations across both platforms Both AS-PCR and BEAMing demonstrate good sensitivity for T790M mutation detection wrt tissue result. Data from: Thress, K. et al J Clin Oncol 32:5s, 2014 (suppl; abstr 8092)
Overall Response Rates rates to AZD9291 as a function of the EGFR T790M mutation An ORR of 63% was observed in patients from AURA with a ctdna sample positive for T790M, a rate very similar to that seen from tumor genotyping Data from: Thress, K. et al J Clin Oncol 32:5s, 2014 (suppl; abstr 8092)
ctdna mutation genotyping conclusions Multiple platforms are capable of sensitive and specific detection of EGFR-TKI sensitizing mutations from NSCLC patient ctdna BEAMing digital PCR and cobas AS-PCR show highly concordant results and have a 70-80% sensitivity for T790M mutation detection with respect to tumor tissue genotyping. Genomic heterogeneity of T790M mediated resistance likely accounts for the reduced specificity observed with plasma-based detection of T790M mutations An ORR of 63% was observed in patients from AURA with a ctdna sample positive for T790M, a rate very similar to that seen from tumor genotyping. These data suggest that plasma detection of T790M mutations may be an option for NSCLC patients with acquired EGFR TKI resistance for whom provision of a rebiopsy is not possible.
Going beyond hotspot assays Applying Next-Gen Sequencing to profiling patient plasma ctdna AIMS Assess mutation pickup versus other techniques Further characterise AURA plasmas Identify mechanisms of resistance Monitor mutation burden over time Illumina sequencing platforms (MiSeq / HiSeq) Variety of approaches being taken Targeted deep NGS (Multiplex PCR-based panel, Qiagen). Boreal OnTarget mutation enrichment + seq. Guardant Health digital sequencing. CRUK collaboration (Baird/Mayes/Rosenfeld) TAm-Seq (Not applied to 9291 currently) AKT1 EGFR MET PTEN ALK ERBB2 MTOR PTGS2 BRAF HRAS NRAS RB1 CDKN2A KIT PDGFRA STK11 CTNNB1 KRAS PIK3CA TP53 Qiagen 20 gene GeneRead Lung panel OnTarget 46 mutation panel Whole genome on some samples to look at tumour genome state (Copy Number / Structural Variants). 54 gene panel
Proof of concept - Targeted NGS assays for EGFR mutations deliver excellent performance compared to BEAMing data (n=54) Exon 19 deletion assays Targeted NGS* Sensitivity Specificity L858R assay Sensitivity Specificity T790M assay Sensitivity 84% (21/25) 100% (29/29) 71% (15/21) 97% (32/33) 88% (22/25) Summary -High concordance between BEAMing and targeted NGS -The majority of missed calls were due to the sensitivity of NGS or PCR-based artifacts of NGS Specificity 90% (26/29) *wrt to BEAMing plasma data
Summing Up / Discussion Topics Can we stratify patients using ctdna analysis? Demonstrated feasibility of QPCR and NGS of ctdna for EGFR mutation detection to support to support the AZD9291 clinical programme. Further work ongoing to demonstrate the utility of ctdna as a method of selecting patients for treatment Can we monitor response and detect relapse earlier? Work in progress, but external evidence suggests yes. Can we identify resistance mechanisms? Work in progress, preliminary data suggests yes in at least some cases. Discussion Topics: What is the current state of the art for ctdna technologies? Which technologies have been successfully deployed for analysis of patient samples? What level of specificity / sensitivity has been achieved? What concordance rates have been observed between plasma and tumour tissue? What is the current success rate for whole genome / whole exome sequencing for ctdna? What are the current limitations of ctdna technology? Which areas for should be prioritized for further evaluation of new technologies? What is our current knowledge of ctdna yield across indications / lines of therapy? 15
Acknowledgements Translational Science Hedley Carr Dan Stetson Ambar Ahmed Brian Dougherty Justin Johnson Ken Thress Roz Brant Carl Barrett Personalised Healthcare & Biomarkers Simon Dearden Suzanne Jenkins Helen Brown Marianne Ratcliffe Benoit Destenaves Oncology informatics Zhongwu Lai Aleksandra Markovets Early Clinical Development Tracey Hammett Global Medicines Development Mireille Cantarini Serban Ghiorghiu Rose McCormack 16
Recent precedents Douillard, J.-Y. et al J Thorac Oncol. 2014; 9: 1345-1353. Analysis of EGFR mut in IFUM* (Iressa Follow Up Measure) study population. Allele-specific QPCR (Qiagen Therascreen EGFR RGQ PCR kit). 94.3% concordance across 652 matched tumour and plasma samples Sensitivity 65.7%; Specificity 99.8%. 96.9% concordance between 224 paired plasmas. ORR very similar between patients with EGFRm+ ctdna vs EGFRm+ tumour (76.9% vs 69.9%) ddpcr assays demonstrated high sensitivity and specificity Assay of serial plasma samples from EGFRm+ NSCLC patients on erlotinib demonstrated: Pre-treatment detection of EGFR mutations Complete plasma response in most cases Increasing levels of EGFR T790M emerging before objective progression. Oxnard, G. R. et al Clin Cancer Res. 2014;20:1698-1705. *NCT01203917 Phase IV single arm study
Confidentiality Notice This file is private and may contain confidential and proprietary information. If you have received this file in error, please notify us and remove it from your system and note that you must not copy, distribute or take any action in reliance on it. Any unauthorized use or disclosure of the contents of this file is not permitted and may be unlawful. AstraZeneca PLC, 2 Kingdom Street, London, W2 6BD, UK, T: +44(0)20 7604 8000, F: +44 (0)20 7604 8151, www.astrazeneca.com 21 Author 00 Month Year Set area descriptor Sub level 1