Poster Title Researchers’, Presenters’ Names Institution/Organization/Company
Combined circulating tumour cell (CTC) and circulating tumor DNA (ctDNA) analysis of blood from patients with pancreatic cancer Ged Brady1, Dominic G. Rothwell1, Sumitra Mohan1, Mahmood Ayub1, Jakub Chudziak1, Kyaw Aung1, Richard Hubner4, Crispin Miller2,3, Alison Backen4, Hui Sun Leong2, Sakshi Gulati1, Angela Lamarca4, Mairéad McNamara4, Juan W Valle4 and Caroline Dive1
1Nucleic Acid Biomarker Laboratory, Clinical Experimental Pharmacology Group, 2Computational Biology Support and 3RNA Biology Group, Cancer Research UK Manchester Institute, The University of Manchester, Wilmslow Road, Manchester, United Kingdom, 4The Christie NHS Foundation Trust, Manchester, United Kingdom
The challenge to improve outcomes for patients diagnosed with advanced pancreatic cancer remains very real with only small improvements in median survival gained by the use of systemic chemotherapy and little improvement in 5-year survival over the past decades. The advent of next generation sequencing (NGS) of tumour nucleic acids has opened up the possibility of improving outcomes through personalised therapies selected on the basis of tumour genetics. Whilst NGS biomarkers can be measured in tumour biopsies sampled shortly before and after treatment this is often not practical for ethical or logistical reasons, or simple lack of availability. Increasingly liquid biopsies, including circulating tumour DNA (ctDNA) and circulating tumour cells (CTCs) provide an alternative to determine the genetic profile of cancer patients. We set out to: 1) develop methods which will allow combined analysis of both CTCs and ctDNA from a single blood collection tube (BCT); 2) establish a NGS panel for genes frequently mutated in pancreatic cancer; 3) evaluate developed methods by applying them to blood from patients with pancreatic cancer; 4) for ctDNA, compare tumour detection sensitivity using KRAS ddPCR, KRAS NGS and NGS of a panel of 654 genes; 5) compare epitope-dependent (CellSearch) and independent (Parsortix) CTC detection. Figure 1. Workflows showing fractionation and analysis of the cellular and plasma components of clinical whole blood samples. Blue text indicates studies in progress.
Background
Aims
7.5mlCellSearch
10mlCellSaveBlood
2.5mlStore/Process
PlasmaCTCCount
DEPArrayIsolaAonofSingleCells
BankCellSuspension
DNAWGA
NGSCopyNumberandMutaAonAnalysis
ctDNAAnalysis
EnrichCTCsParsor.x
10mlStreckCFBCT
Blood
Store/ProcessPlasma
SeparatePlasmaand
Cells
EluteEnrichedCTCs
BulkRTqPCR
RNA-Seq/FluidigmqPCR
AnalyseRNAandDNAComponents
PrepareTotalNucleicAcids
StoreCellSuspension
BulkDNAWGA
NGSCopyNumberandMuta.on
SingleCellWGANGS
CopyNumber
SingleCellRNA-SeqandRNA-ISH
Direct Single Cell Analysis and
Confirmation of Enriched Sample Data
ctDNAAnalysis
Approach
Figure 2. Sensitive ctDNA and CTC NGS analysis methods established. A) shows NGS generated copy number analysis (CNA) of matched single CTCs, EDTA cfDNA, CellSave cfDNA, and CDX tumours from SCLC as recently published1. Matching patterns of gain (red) and loss (blue) were seen across all related samples. B) External quality assessment samples were used to evaluate an Agilent SureSelect panel of 654 gene chosen primarily on the basis of frequent alterations detected in pancreatic cancer2. Combined CTC enrichment and cfDNA isolation is readily achievable using a single Streck ctDNA BCT. Results indicate that combined CTC and cfDNA analysis is more sensitive than either approach alone. Initial data indicates presence of RNA in enriched CTCs with more analysis required to establish utility. Developed ctDNA methods also suitable for monitoring response to therapy and identifying mechanisms of resistance4. 1. Rothwell et al. (2016). Molecular oncology 10, 566-574. 2. Bailey et al. (2016). Nature 531, 47-52. 3. Chudziak et al. (2016). The Analyst 141, 669-678. 4. Girotti et al. (2016). Cancer discovery 6, 286-299.
Results
Summary and Conclusions
References
CellSearch Parsortix
Gene Mutation Predict.Freq QC1 QC2 QC3 QC4 QC5 Sample.
Name Gene Mutation Predict.Freq
Expt.Freq
KRAS G12C 0.15 0.09 ND ND ND ND QC1 KRAS G12C 0.15 0.09EGFR G719S 0.33 0.23 ND ND ND ND QC1 EGFR G719S 0.33 0.23TP53 R273H 0.15 0.28 ND ND ND ND QC1 TP53 R273H 0.15 0.28EGFR G719S 0.33 ND 0.36 ND ND ND QC2 EGFR G719S 0.33 0.36NRAS Q61K 0.5 ND 0.43 ND ND ND QC2 NRAS Q61K 0.50 0.43KRAS G13D 0.2 ND ND 0.22 ND ND QC3 KRAS G13D 0.20 0.22PIK3CA H1047R 0.2 ND ND 0.20 ND ND QC3 PIK3CA H1047R 0.20 0.20KIT D816V 0.2 ND ND 0.12 ND ND QC3 KIT D816V 0.20 0.12KRAS G13D 0.5 ND ND ND 0.43 ND QC4 KRAS G13D 0.50 0.43BRAF V600K 0.5 ND ND ND 0.53 ND QC4 BRAF V600K 0.50 0.53PIK3CA H1047R 0.5 ND ND ND 0.46 ND QC4 PIK3CA H1047R 0.50 0.46KRAS G13D 0.15 ND ND ND ND 0.15 QC5 KRAS G13D 0.15 0.15BRAF V600E 0.15 ND ND ND ND 0.18 QC5 BRAF V600E 0.15 0.18PIK3CA H1047R 0.15 ND ND ND ND 0.20 QC5 PIK3CA H1047R 0.15 0.20
0.91Correlation.Coefficient.Predict.v.Expt
Mutation(Detection(Across(all(Samples Summary(of(Mutation(Frequencies
Samples
KRASMUTddPCR(Low)
KRASMUTddPCR(High)
KRASMUTNGS
AdditionalGenesNGS
ddPCRonly NGSOnlyNGSorKRASddPCR
Sum 21 14 6 20 2 2 23% 70% 47% 21% 69% 7% 7% 77%
NGSddPCR 182 2
(1samplepending)
A B
CellSearchParsor-x 44 3
(Numbersshownformatching19pa9entsonly)
CellSearchCTCs>1in7.5ml
ParsortixCTCs>1in
1ml
CombinedCTC
approach
Sum 7 8 11% 37% 42% 58%
Figure 3. Detection of mRNA in enriched CTCs. RT qPCR of B2M was applied to RNA extracted from Parsortix bulk CTC enrichment of samples from patients with colorectal cancer. Blood collection tubes were maintained at room-temperature for up to 4 days prior to Parsortix enrichment and analysis
Figure 4. Summary of Pancreatic Cancer Patient Pilot Study. A) a comparison of mutant KRAS ddPCR and NGS applied to the same extracted ctDNA samples. B) a comparison of CTCs identified by CellSearch and Parsortix (enriched as previously described3 and CTCs detected in bulk enriched populations by mutant KRAS ddPCR).
0
2
4
6
8
10
12
14
Parsor&xEnrichCTCs B2MRTqPCR
TotalNucleicAcids
Prepara&on
40-C
t CellSave
Streck cfDNA
A B