Role of circulating tumor DNA in response prediction and assessment of clonal evolution Davide Rossi, M.D., Ph.D. Hematology IOSI - Oncology Institute of Southern Switzerland IOR - Institute of Oncology Research Bellinzona - Switzerland
Conflict of interest Research Support: Gilead, Abbvie, Janssen, Cellestia Employee No Consultant No Major Stockholder No Speakers Bureau No Honoraria Gilead, Abbvie, Janssen, Roche, AstraZeneca Scientific Advisory Board Gilead, Abbvie, Janssen, AstraZeneca, MSD
Agenda • General notions and practicalities • ctDNA in DLBCL • ctDNA in cHL
Agenda • General notions and practicalities • ctDNA in DLBCL • ctDNA in cHL
ctDNA is of low abundance: Optimization of sensitivity and specificity of NGS is mandatory Allele frequency in gDNA Allele frequency in cfDNA 100% 100% Allele frequency of ctDNA mutations Mutations VAF>1% Mutations VAF<1% 10% 10% 1% 1% 0% 0% ctDNA mutations True mutations Variant frequency (%) Sensitivity threshold Background noise of NGS Spina V, et al. Blood 2018 Variant position (NM_000546.5)
Pre-analytics is critical cfDNA sample of good quality: peak sized between 100 and 200 bp 500 cfDNA cfDNA of poor quality: gDNA contamination cfDNA 200 gDNA 0
cfDNA circulates in small amounts Hohaus S et al. Ann Oncol. 2009;20(8):1408-1413
Challenges in the identification of small abundant ctDNA variants by NGS • Input DNA (at least 32 ng) • Library preparation chemistry (capture based, molecular barcoding) • Coverage (>2000X >80% target region) • Bioinformatic pipeline for variant calling (catalogue of systematic errors)
The origin of cell free DNA in healthy subjects and cancer patients • In healthy individuals cfDNA derives from apoptosis of normal hematopoietic cells • In tumor patients cfDNA is released by tumor apoptotic cells • ctDNA is distinguished from other cfDNA by the presence of somatic mutations representative of tumor biology absent in normal cells Snyder, Cell 2016
Agenda • General notions and practicalities • ctDNA in DLBCL • ctDNA in cHL
Diffuse large B-cell lymphoma vs classical Hodgkin lymphoma DLBCL cHL Tumor cells are enriched in the mass Tumor cells are rare in the mass Exome sequencing data from >1000 cases Exome sequencing data from only 10 cases Reichel J, et al. Blood 2015 Pasqualucci L, et al. Semin Hematol 2015
ctDNA mirrors the genetics of DLBCL cells 30% Mutation frequency GC Non-GC N=30 20% 10% 0% Mutation identified both in gDNA and in cfDNA KMT2D STAT6 Mutation identified in gDNA only Mutation identified in cfDNA only PIM1 TNFAIP3 100% 82.8% 80% MYC CCND3 60% N=21 N=18 N=87 TP53 EZH2 40% 20% CREBBP TBL1XR1 0% Sensitivity missense mutation truncating mutation Rossi D, et al. Blood 2017
Circulating tumor DNA resolves the spatial heterogeneity of lymphomas Scherer F, et al. Sci Transl Med 2016
Longitudinal cfDNA genotyping allows Non invasive detection of ibrutinib resistance mutations Scherer F, et al. Sci Transl Med 2016
LymphoSIGHT ™ platform 4) Prepare for 3) Amplify VDJ 5) Sequence ~1M 1) Collect 10cc 2) Extract DNA sequencing with with multiplex 100bp reads peripheral common PCR PCR blood CTGGCCCCAGTAGTCATACCAACTAGCG TTGGCCCCAGAAATCAAGACCATCTAAA ACGGCCCCAGAGATCGAAGTACCAGTGT TTGGCCCCAGACGTCCATATTGTAGTAG CTGGCCCCAGAAGTCAGACCGGCTAACA Serum Genomic PCR amplicons Sequencing library Sequence data DNA
MRD at the end of treatment predicts progression 5-year TTP of 94.6% vs. 11.8% Roschewski M et al. Lancet Oncol , 2015
Limitations of Ig-HTS as tumor fingerprint in the Liquid biopsy Recovery rate of the tumor IG rearrangement from DLBCL tissue biopsies Kurtz DM et al. Blood , 2015.
Mutational profile as tumor fingerprint PB granulocytes Plasma Ultra deep sequencing Resolution of the tumor Log fold change 1 mutation in ctDNA profile -1 -3 cHL 72 ND -5 ATM CD36 ATM PIK3CA STAT6 PIK3CA MAP3K14 CD79B IRF8 FOXO1 FOXO1 0 10 17 26 31 4
The prognostic value of molecular response is independent of interim imaging Kurtz, J Clin Oncol 2018
Agenda • General notions and practicalities • ctDNA in DLBCL • ctDNA in cHL
Understanding of the genetics of DLBCL vs cHL DLBCL cHL Tumor cells are enriched in the mass Tumor cells are rare in the mass Exome sequencing data from >1000 cases Exome sequencing data from only 10 cases Reichel J, et al. Blood 2015
ctDNA mirrors the genetics of HRS cells a b 80% (12/15) 80% % of mutated cases 53% (8/15) 60% ITPKB CATALYTIC domain 27% 1 (4/15) 20% 40% 946 13% (3/15) 7% (2/15) 20% (1/15) 0% TNFAIP3 OTU ZF ZF ZF ZF ZF ZF ZF STAT6 ITPKB TNFAIP3 B2M GNA13 HIST1H1E CIITA IRF8 ARID1A BTG1 IRF4 PCBP1 PIM1 STAT3 ATM BCL6 BTK CCND3 CD58 CXCR4 ID3 KMT2D MYC NFKBIE NOTCH1 PRDM1 SPEN TET2 TNFRSF14 TP53 TRAF3 XPO1 1 790 0 N. of mutations 5 STAT STAT6 STAT_int STAT_bind SH2 STAT6_C missense nonsense frameshift 10 alpha 1 847 3’ -UTR splicing start loss 15 missense nonsense 20 splicing frameshift c 20 N. of mutations 15 10 100% Biopsy confirmed mutations d 87.5% 5 0 80% 60% e 13 84 12 Biopsy confirmed mutations 66.6% 66.6% 40% 100% 100% 100% 100% 100% 100% 100% 100% 100% 91% 75% 50% 40% identified in ctDNA 100% 20% 50% 0% 0% Mutation identified both in gDNA and in ctDNA Mutation identified in ctDNA Patient *chemorefractory sample Mutation identified in gDNA
Mutational landscape of newly diagnosed cHL N=80 STAT6 TNFAIP3 ITPKB GNA13 B2M ATM SPEN KMT2D XPO1 TP53 ARID1A HIST1H1E BTG1 Spina V, et al. Blood 2018
Mutated pathways in newly diagnosed cHL NF- κ B Epigenetic genes 46.2% (37/80) 35% (28/80) PI3K-AKT immune surveillance genes 27.5% (22/80) 46.2% (37/80) Cytokine signaling NOTCH pathway 37.5% (30/80) 20% (16/80) Spina V, et al. Blood 2018
Interim PET/CT accuracy in cHL False negative rate = 3% False positive rate = 19% Baseline PET Interim PET 3 1 2 4 5 1. 2. 3. 4. 5. No uptake FDG < MBP FDG >MBP ≤ liver FDG > liver FDG >> liver Terasawa, et al J Clin Oncol 2009 27:1906-1914
Changes in tumor cfDNA complement iPET cHL 66 cHL 85 cHL 67 cHL 52 cHL 19 cHL 23 cHL 63 cHL 78 cHL 64 cHL 65 cHL 13 cHL 74 cHL 75 cHL 76 cHL 79 cHL 80 cHL 83 cHL 86 UPN25 UPN11 cHL 8 UPN6 UPN4 UPN3 4.5 a b d 1 Deauville Log fold change in tumor ctDNA 3 4 4 3 3 3 3 1 2 2 2 4 4 5 4 1 2 1 1 1 1 1 3 1 4.0 PD PD PD PD PD PD CR CR CR CR CR CR CR CR CR CR CR CR CR CR CR CR CR CR Outcome Standardized log-rank statistic 0 3.5 -1 3.0 -2 -3 2.5 -4 p=0.001 2.0 -4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 ND -5 Log c Score e 1 > -2 log fold reduction Log fold change in tumor ctDNA 0 -1 -2 -3 -4 ND -5 < -2 log fold reduction p<0.001 0 50 100 150 200 Days from start of therapy iPET positive – Progressive disease 18 15 2 2 0 iPET positive – Cured iPET negative – Progressive disease 6 0 0 0 0 iPET negative – Cured Spina V, et al. Blood 2018
Conclusions • Molecular markers informing treatment: BCR signaling mutations, EZH2 mutations • Clonal evolution: mutation resistance monitoring • Molecular markers informing response to therapy: minimal residual disease monitoring
Clinical trial design incorporating ctDNA assessment Intensification Salvage Biological agent +/- -/+ +/+ +/+ PET 0 Tx Cycle 1-2 PET 2 PET EOT ctDNA 0 ctDNA 2 ctDNA EOT +/- -/+ -/- -/- Tx Cycle 3-6 Follow-up
Experimental Hematology Lymphoma Unit Hematology Alessio Bruscaggin Annarosa Cuccaro Bernhard Gerber Claudia Cirillo Stefan Hohaus Alden Moccia Adalgisa Condoluci Anastasios Stathis Pathology Francesca Guidetti Georg Stüssi Maurizio Martini Gabriela Forestieri Emanuele Zucca Luigi Larocca Valeria Spina Nuclear Medicine Lodovico Terzi di Bergamo Luca Ceriani Lymphoma & Genomics Michele Ghielmini Francesco Bertoni Martina Di Trani Franco Cavalli Silvia Locatelli Carmelo Carlo-Stella Clara Deambrogi Lorenzo De Paoli Fary Diop Luca Nassi Gianluca Gaidano Unrestricted research grant from Gilead and Abbvie
Recommend
More recommend
