10de bijeenkomst Werkgroep “Moleculaire Diagnostiek in de Pathologie” 15 Januari 2016, Utrecht Disclosures Behandelkeuze voor targeted-therapie aan de hand van mutatie-analyse op cel-vrij plasma DNA Consultant/Advisory Board: Ed Schuuring Klinisch Moleculair Bioloog in de Pathologie Hoogleraar in de Moleculaire Oncologische Pathologie Hoofd Laboratorium Moleculaire Pathologie Afdeling Pathologie van UMCG AstraZeneca, Roche, Pfizer, Novartis, Amgen, BioCartis, QCMD, ESP, IQNPATH Speaker’s fee: Abbott, Novartis, Roche Stock/Royalties: None [email protected] Behandelkeuze voor targeted-therapie aan de hand van mutatieanalyse op cel-vrij plasma DNA Can peripheral blood be used to determine the presence of the tumor ? • Clinical relevance of cell free DNA testing - Cell-free DNA from plasma reflects volume/mutation-profile of tumor - Clinical applications • • • • Methods to detect circulating tumor DNA Collection of plasma (standarisation) Plasma processing and DNA extraction (standarisation) Quality control and EQA 1 Circulating tumor DNA in blood plasma Can peripheral blood be used to determine the presence of the tumor ? Plasma: * DNA (cell-free DNA, cfDNA) * RNA (exosomes) * Proteomic/immunological markers * Pharmacokinetics ([drug]) Leukocytes Platelets Schwarzenbach Nat Rev Cancer 2011 Can peripheral blood be used to determine the presence of the tumor ? Cell-free plasma: * DNA and circulating tumor DNA (ctDNA) * RNA and circulating tumor RNA (ctRNA) * Proteomic/immunological tumor markers * Pharmacokinetics ([drug]) Leukocytes and circulating tumor cells (CTC) Platelets and tumor RNA Behandelkeuze voor targeted-therapie aan de hand van mutatieanalyse op cel-vrij plasma DNA • Clinical relevance of cell free DNA testing - Cell-free DNA from plasma reflects volume/mutation-profile of tumor - Clinical applications • • • • Methods to detect circulating tumor DNA Collection of plasma (standarisation) Plasma processing and DNA extraction (standarisation) Quality control and EQA 2 Plasma-cfDNA increased in colon cancer and associated with tumor size Plasma cell-free DNA as a predictor for the presence of NSCLC Ulivi Cell Oncol 2013 Gorges Biomarkers 2012 Plasma cell-free DNA as a predictor for the presence of NSCLC Plasma cell-free DNA as a predictor for the presence of NSCLC In all these studies different assays to determine DNA concentration: • qPCR • Picogreen • Nanodrop • Qubit (approx 10x lower concentration compared to nanodrop) • In general: <2 ng/ml plasma (untreated pts !!!) Ulivi Cell Oncol 2013 Ulivi Cell Oncol 2013 3 DNA concentrations by picogreen (ng/ml plasma) classified per tumor type at baseline Liquid biopsies: the detection of cell-free circulating tumor DNA in plasma (ctDNA) Total amount of cell-free plasma DNA increases: • • • • • • Luca: 9-19 ng ~12 ng Tumor progression Inflammation Exercise Treatment Pregnancy Hemolysis GIST: 4-25 ng ~10 ng Perkins Plos One 2012 PA UMCG 2015/16 Qubit Crowly E. et al Nat. Rev. Clin. Oncol. 10, 472-484 (2013) Tumor mutation testing on liquid biopsies Tumor mutation testing on liquid biopsies sample selection: important issues sample selection: important issues Plasma collection and storage - Blood collection in EDTA-tube: processing < 4hrs (no consensus) - Streck-tube: <24hrs BUT expensive and no routine tube in hospital Blood collection, plasma processing and storage: Plasma processing/collecting: - Centrifugation: 1x slow (~800g) and 1x high (13000g) (no consensus) - Storage: - -20oC, -80oc - Storage step 1, and 2nd centrigation after thawing very time-consuming high hands-on time (many technicians) Automatisation not suitable Large storage capacity (-80oC) - No standards/consensus for processing - Existing plasma-biobanks NOT suitable (hemolysis) - No references/EQA available 4 The clinical utility of ctDNA CT-scanning to identify lung disease Diagnosis and prediction Gouden standaard: > verkrijgen tumorweefsel via biopt om diagnose en behandelkeuze te kunnen stellen in routine moleculaire pathologie De detectie van EGFR-mutaties in circulerend tumor DNA in plasma komt heel goed overeen met DNA uit het pretreatment tumor biopt > Biopteren niet altijd mogelijk (erg belastend, risicovol neveneffecten, diffuse nodules) Meta-analysis of 20 studies comparing tumor with baseline ctDNA for EGFR mutations > In 15-25% van pretreatment biopten: geen representatief weefsel (te weinig weefsel of te weinig neoplastische cellen) Specifiteit: 99.8% Pooled specificity: 92.2% Pooled sensitivity: 69.1% D Vals-positiviteit: 1/546 Sensitiviteit: 34% niet gedetecteerd in plasma A small biopsy; B core-needle biopsy; C surgical specimen; D cytology E biopsy with very few tumor cells E Luo, Sc Reports 2014 Douillard JTO 2014 5 Molecular diagnostics of lung cancer for treatment planning using gene-targeted therapy in NL dutch guidelines In 2007: starting with EGFR-mutation screening In 2013: Dutch guideline in NSCLC Only EGFR-mutation analysis Validation of EGFR mutation testing (G719S, exon19del, L858R, T790M) in ctDNA plasma (ongoing) comparing ddPCR, cobas-EGFR, NGS and Idylla • ddPCR (BioRad vs in house primers probes) Relative cheap (~50-100 ?) TAT: resp 8-12 hrs Hands-on time: 2-3 hrs In 2013: ALK-translocation (Registration crizotinib July 2012) • Validation on tumor tissue (MD-sneltest) In July 2015: revisited Dutch guideline for NSCLC: • Biobanking plasma/platelets of all lung cancer patients at t=0, 2, 6, 12 weeks (starting July 2015) (1) EGFR, ALK; (2) HER2, BRAF, RET and ROS1 > testing ctDNA from plasma if no appropriate biopsy is available • Plasma vs pretreatment tumor (preliminary): - EGFR-L858R: 3/5 positive - EGFR-ex19del: 2/3 positive Bosman, Tibbesma, Elst, Groen, Schuuring Validation of EGFR mutation testing (G719S, exon19del, L858R, T790M) in ctDNA plasma (ongoing) comparing ddPCR, cobas-EGFR, NGS and Idylla Validation of EGFR mutation testing (G719S, exon19del, L858R, T790M) in ctDNA plasma (ongoing) comparing ddPCR, cobas-EGFR, NGS and Idylla Commercial systems: NGS iontorrent of tumor: • Cobas EGFR plasma (dec 2015) Relative expensive (~200-300 ?) TAT: 4 hrs Hands-on time: 1-2 hrs • Idylla EGFR plasma (2016) Relative expensive (~200-300 ?) TAT: 90 min Hands-on time: 5-10 min L858R No mutation G719S Ex19del + T790M G719S + S768I Ex19del + T790M L858R Ex19del Bosman, Tibbesma, Elst, Groen, Schuuring 6 Cobas EGFR mutation blood test (in development) sensitivity/specificity The clinical utility of ctDNA Monitoring treatment response Meldgaard BMC Cancer 2014 Specificity: 94% Sensitivity: 71% (39% not detected in plasma) Agreement: 91% False-positive rate: 6/168 Procedure of plasma cfDNA mutation testing Mutation testing using plasma cell free DNA applications in lung cancer (treatment response monitoring) tumor Tumor volume plasma Resistant mutations Druggable mutation Druggable mutation DNA Tumor volume DNA sequencing Define tumor specific mutation Aim: •Success of treatment response upon decreased plasma mutation levels of druggable mutation cfDNA quantification ctDNA ddPCR •Early detection of therapy resistance (prior to clinical manification) by increased plasma mutation levels of druggable mutation •Early detection of therapy-resistent plasma mutation levels (prior to clinical manification) enabling new treatment opportunities Adapted from Diaz, J Clin Oncol 2014 Adapted from Diehl Nat Med 2008 Mutant DNA concentration 7 Disappearance of EGFR mutations during TKI treatment using cobas-ctDNA test in NSCLC KRAS-mutations are associated with acquired resistance upon anti-EGFR-treatment using sensitive digital beaming assay CRC patients with KRAS-wt in pretreatment biopsy ctDNA testing in plasma before start treatment and after resistence Bettegowda, Sc Transl Med 2014 Marchetti JTO 2015 13-04-2015 30-03-2015 25-03-2015 16-03-2015 15-12-2014 12-12-2014 13-10-2014 09-2013 22-09-2014 25-05-2013 10-09-2014 KWF-Alpe d’Huez: Pieter Boonstra, Marco Tibbesma, Arja ter Elst, Ed Schuuring, An Reyners (GIST-consortium UMCG, Radboud, LUMC, ErasmasMC and NKI/AvL) 01-09-2014 Example of workflow of GIST-patient during KIT-mutation targeted therapy 25-08-2014 GALLOP-study: Assessment of mutations in tumors and circulating tumor DNA and measurement of TKI plasma exposure to optimize treatment (KIT) Tumor volume Resistant mutations Operatie Druggable mutation Sunitinib 37.5mg Imatinib 400mg Druggable mutation Tumor volume Adapted from Diaz, J Clin Oncol 2014 Recidief Partiële respons Progressie Boonstra Tibbesma Elst Schuuring Reyners Gallop 8 Location of mutations/deletions in exon 11 of KIT gene in GIST in two hotspot regions Design of multiplex ddPCR drop-off assay to detect mutations in hotspot 1 or 2 of exon 11 KIT gene Boonstra Tibbesma Elst Schuuring Reyners Gallop KIT exon 11 drop-off ddPCR assay to detect hotspot 1 or 2 mutations in pretreatment biopsies of GIST Boonstra Tibbesma Elst Schuuring Reyners Gallop KIT exon 11 drop-off ddPCR assay to detect hotspot 1 or 2 mutations in pretreatment biopsies of GIST Patient NGS+/ddPCR+: 16 NGS+/ddPCR-: 2 (outside hotspot) NGS+/ddPCR-: 1 (false-negative) NGS-/ddPCR-: 9 Agreement: 27/28 (96%) Boonstra Tibbesma Elst Schuuring Reyners Gallop Mutatie Type mutatie Baseline Twee weken Zes weken 2 KIT 11 c.1667_1669del; p.Q556_W557>R - - - 3 KIT 11 c.1669T>C; p.(W557R) - - - 4 KIT 11 c.1727_1729del; p.(L576del) 8% - - 5 KIT 11 c.1671_1676delGAAGGT; p.W557_V559>C 0% - - 6 KIT 11 c.1655_1660delTGTATG; p.(M552_E554delinsK) - - 7 KIT 11 c.1726_1728del; p.L576del 0% - - 9 KIT 11 c.1676T>A; p.V559D 0% 0% 0% 10 KIT 11 c.1669_1674del; p.(W557_K558del) - - - 13 KIT 11 c.1671_1672delinsTG 0% 0% 0% 14 KIT 11 c.1669_1674del; p.(W557_K558del) 0% - 16 KIT 11 insertie 36bp - - - 17 KIT 11 c.1679_1680delinsAG; p.V560E 0% 0% 0% 18 KIT 11 c.1673_1717del 0% - - 19 KIT 11 c.1662_1674delinsGGAAGAA; 2,5% 0,0% 0,7% 22 KIT 11 c.1669_1674del; p.(W557_K558del) - - - 26 KIT 11 c.1649_1663del; p.K550_V555>I - - - 28 KIT 11 c.1649_1663del.; p.(K550_V555delinsI) - - - 29 KIT 11 c.1669_1674del; p.(Trp557_Lys558del) 16,5% 61,0% 3,0% 30 KIT 11 c.1676T>A; p.V559D 0,9% 5,5% 0% 31 KIT 11 c.1676T>A; p.(V559D) 0,1% - - 32 KIT 11 c.1669_1674del; p.(W557_K558del) - 7,2% - Only 7/12 baseline plasma samples positive - - Boonstra Tibbesma Elst Schuuring Reyners Gallop 9 mutation detection using ddPCR during imatinib-treatment of GIST patient mutation detection using ddPCR during imatinib-treatment of GIST patient Tumor FA: 35% Baseline FA: 0.7% Pretreatment plasma ctDNA: FA = 16% Pretreatment tumor: FA = 88% Tumor : 20x14cm 2 week FA: 5.5% 6 week FA: 0% 4 wk after TKI-treatment Plasma ctDNA: FA = 3% 2 wk after TKI-treatment Plasma ctDNA: FA = 61% KWF-Alpe d’Huez: Boonstra, Tibbesma, ter Elst, Schuuring, Reyners (GIST-consortium UMCG, Radboud, LUMC, ErasmasMC and NKI/AvL) KWF-Alpe d’Huez: Boonstra, Tibbesma, ter Elst, Schuuring, Reyners (GIST-consortium UMCG, Radboud, LUMC, ErasmasMC and NKI/AvL) GALLOP-study: Assessment of mutations in tumors and circulating tumor DNA and measurement of TKI plasma exposure to optimize treatment Tumor : 7.5x12cm Clinical utility of ctDNA testing (KIT) Tumor volume Druggable mutation Resistant mutations Druggable mutation Tumor volume Adapted from Diaz, J Clin Oncol 2014 Detection of resistant mutations (switching treatment) • • • • • • • Detection of early disease Assessment of molecular heterozygeneity of overall disease Identification of molecular markers for targeted therapy Evaluation of early treatment response Monitoring treatment response Diagnosis of minimal residual disease Detection of resistant mutations NGS of pretreatment GIST: activating KIT exon 11 del (Q556_V560delinsH) 86% NGS of TKI resistant GIST: resistant KIT mutation (Y823D) 92% NGS of plasma at progression: both mutations (resp 18% and 17%) 10 Technologies for detection and characterization of ctDNA Assays to detect low copy number mutations in plasma cfDNA Haber and Velculescu, Cancer Disc 2014 Power In Partitioning Digital droplet PCR technologie ddPCR Nanodroplet PCR reactions are independent, single amplification events One measurement Many thousands of discrete measurements ddPCR-assay: primer-set with HEX-probe for mutation and FAM-probe for wt 11 Rare Mutation Detection: Example 2 (KRAS G12V) mutant only mut 1% mutant 0.1% mutant 0.001% mutant 0% mutant High analytical sensitivity of mutation detection using ddPCR 20 ng input A549 (KRAS-G12S) 20ng input H1650 (EGFR-E746_A750del) mut + wt wt 0.01% mutant wt Biorad DNA concentrations by picogreen (ng/ml plasma) classified per tumor type at baseline 0.1% 20 ng input A549 (KRAS-G12S) 2 ng input A549 (KRAS-G12S) Luca: 9-19 ng ~12 ng GIST: 4-25 ng ~10 ng Perkins Plos One 2012 0.01% 20 ng input H1650 (EGFR-E746_A750del) 2 ng input H1650 (EGFR-E746_A750del) Using 2 ng (from ~1 ml plasma) the sensitivity decreases 10x For sensitivity of 0.1/0.01% at least 10ml plasma is needed per test !!! PA UMCG 2015/16 Qubit 12 Other methods to detect low copy mutations in plasma ctDNA at UMCG • Digital droplet PCR (BioRad) • Cobas EGFR mutation plasma test (launched nov 2015) (reflab) • Biocartis (in development; reflab) for EGFR, BRAF, KRAS mutation • NGS (IonTorrent-in house-panel) and with Maier (Leipzig), Bubnoff (Freiburg), Dinjens (ErasmusMC) and Ligtenberg (RadboudMC) Quality and assessment of proficiency • Proper control and reference samples (artificial) • Spiking plasma with plasmid or degraded cellular DNA Assessment: • Processing plasma and ctDNA extraction (no plasma for so many labs ???) • Detection of mutations in plasma (IQNPath-pilot-program) (in development) • European consensus meeting in summer 2016 Disadvantages of ctDNA testing • Yield of total cell-free DNA might be low • To detect low copy-mutations 2-4 ml plasma is needed (equivalent of 10 ml EDTA tube) • Because of low copies quantitative assay is needed • Processing blood/plasma is laboreous • No standarisation of collection/processing blood plasma Cell-free, circulating tumor DNA testing in plasma • • • • Non-invasive Routine blood tubes (EDTA buis) Cheap Multiple testing UMCG is reference-lab for EQA-IQNPath, cobas and Biocartis 13 Illumina Launches GRAIL, Focused on Blood-Based Cancer Screening MONDAY, JANUARY 11, 2016 Dank voor uw aandacht [email protected] Illumina has joined investors that include Bill Gates and Jeff Bezos in committing more than $100 million toward launching a new company with the ambitious goal of screening for the most frequent types of early-stage cancers through a simple test that measures circulating tumor DNA (ctDNA) in the blood. 14