Liquid biopsies as prognostic and predictive biomarkers; ready for - - PDF document

▶

Feb 21, 2023 230 likes •583 views

Liquid biopsies as prognostic and predictive biomarkers; ready for the clinic? AR-V7 expression in CTCs from patients with castration- resistant prostate cancer & Feasibility studies of circulation tumor DNA analysis John W. M.

SLIDE 1

1 “Liquid biopsies as prognostic and predictive biomarkers; ready for the clinic?”

AR-V7 expression in CTCs from patients with castration- resistant prostate cancer & Feasibility studies of circulation tumor DNA analysis

Donderdag 16 april 2015

John W. M. Martens Dept of Medical Oncology, Erasmus MC Cancer Institute Rotterdam

Conflict of Interest

Funding/collaboration received from: Veridex (BC/CRC CTC studies) Sanofi (CRPC CTC-ARv7 studies) Thermo-Fisher (early access to oncomine panels) Cytotrack & Olink (collaborator in EU-project CAREMORE) Philips research (Molecular diagnostics BC) Therawis (Prognostic markers) Patent applications: Our dept. owns a patent on the ARv7 in CTCs

SLIDE 2

2

Outline of the presentation

State-of-the-art of genomics research (focus BC) Introduction about liquid biopsies (CTCs, ctDNA, EVs) Application of CTC count and characteristics in metastatic cancer ARv7 positive CTCs predict progression in mCPCR Comparisons of ctDNA detection methods (dPCR & targetted NGS) Standard operation procedures/sample collection Examples of NGS in mBC and CRC ESR1 detection in CTCs and ctDNA Conclusions and outlook

Outline of the presentation

State-of-the-art of genomics research (focus BC) Introduction about liquid biopsies (CTCs, ctDNA, EVs) Application of CTC count and characteristics in metastatic cancer ARv7 positive CTCs predict progression in mCPCR Comparisons of ctDNA detection methods (dPCR & targetted NGS) Standard operation procedures/sample collection Examples of NGS in mBC and CRC ESR1 detection in CTCs and ctDNA Conclusions and outlook

SLIDE 3

3 Personalized cancer treatment

Giving:

The right drug, to the right patient, at the right dose, from the right

moment onwards, till the right moment

Personalized cancer treatment will yield:

Better outcome and less toxicity Less patients receive treatments, more will benefit More favorable cost-effectiveness of cancer treatments

To get there: early markers for response, prognostic and in particular predictive factors

Lessons learned from NGS of BC genomes

1. Mutations are recurrent in a few common and many rare driver genes (>93 breast cancer genes)

ER-positive ER-negative Stephan PJ et al Nature. 2009 Dec 24;462(7276):1005-1010

Breast tumors Mutated genes

Signatures in 560 breast cancer genomes (Nik-Zainal, nature 2016

SLIDE 4

4

Metastases can derive from subclones (un)detectable in the primary tumor

2. Cancer evolves over time (cancer evolution)

Lessons learned from NGS of BC genomes

Yates LR et al Nat Medicine 2015

Introduction

Breast tumors evolve over time during “natural” disease progression and in response to treatment Example(s) include: the appearance of activation ESR1 mutations during endocrine treatment Discrepancy between ER and/or Her-2 expression and PIK3CA mutation status between primary tumor and metastasis Evidence of clonal hierarchical relations within heterogenous tumors Drivers of diversity (APOBEC; BRCAness) are operational during tumor progression Repeated sampling during tumor progression is likely benefitial for optimal patient care

SLIDE 5

5

Outline of the presentation

State-of-the-art of genomics research (focus BC) Introduction about liquid biopsies (CTCs, ctDNA, EVs) Application of CTC count and characteristics in metastatic cancer ARv7 positive CTCs predict progression in mCPCR Comparisons of ctDNA detection methods (dPCR & targetted NGS) Standard operation procedures/sample collection Examples of NGS in mBC and CRC ESR1 detection in CTCs and ctDNA Conclusions and outlook

Liquid biopsies (CTCs/ctDNA)

In the circulation of patients:
tumor cells are present (circulating tumor cells (CTCs))
tumor DNA is detectable (circulating tumor DNA (ctDNA)

Q: Can these be “surrogate biopsies” of metastatic tissue?

CTCs

ER, and HER2 status can differ between the primary tumor and metastases (or CTCs) Specific mutations are acquired in metastatic disease (EGFR; KRAS; ESR1)

(Tewes et al. BCRT 2009; 115: 581-590)

SLIDE 6

6

CTCs vs cell-free DNA (cf-DNA) CTCs cf-DNA Intact, living tumor cells cell-free DNA, RNA and protein DNA only (histones?) Present in 65% of MBC Present in 82% of MBC More complex processing (EpCAM-based enrichment) Relatively easy processing (plasma isolation) Both: Present in extreme low frequency Calls fo sensitive and specific assays Circulation tumor cells (CTCs)

SLIDE 7

7

CTC detection

Circulating tumor cells (CTCs) From peripheral blood CellSearch System Median 1 to 9 cells / 7,5 mL In 30 - 80% 1 CTC

CTC isolation and detection: CellSearch™

SLIDE 8

8 CTC isolation and detection: CellSearch™

CTC definition

EpCam positive
CD45 negative
CK 8/18/19 positive
DAPI positive

CD45- CK+

DAPI+

CTC count and prognosis

Prognostic factor for OS:

De Bono, J.S., et al., Circulating tumor cells predict survival benefit from treatment in metastatic castration-resistant prostate

cancer. Clin Cancer Res, 2008. 14(19): p. 6302-9.

SLIDE 9

9

CTC count and treatment response

Early response marker:

Already 2 – 5 weeks after start treatment

De Bono, J.S., et al., Circulating tumor cells predict survival benefit from treatment in metastatic castration-resistant prostate

cancer. Clin Cancer Res, 2008. 14(19): p. 6302-9.

CTC characterization

CTC characterization Use as liquid biopsy?

Adapted from Beije N., et al. Circulating tumor cell enumeration: the clinician's guide to breast cancer treatment? Cancer Treat Rev, 2014.

SLIDE 10

10

One human cell contains 5 5 5 5-

10 pg

10 pg 10 pg 10 pg total RNA. Needed for traditional qRT-PCR: 5 5 5 5-

20 ng

20 ng 20 ng 20 ng per assay.

Characterization of CTCs

Analyse multiple transcripts in limited material

5 10 15 20 25 30 35

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Ct-values after pre-amplification

MAGEA3 TERT KRT20 ESR2 CGB TFF1 TACSTD2 (EpCAM check) BST1, blood_check TWIST GALGT ALDH1A1 ESR1 MET MDM2 SERPINB5 SCGB2A2 HMBS EGFR CXCL1 HPRT PTPRC (CD45),blood_che GUSB TACSTD1 (EpCAM check) TFF3 SPDEF RPL13A MUC1 ERBB2 KRT19 ACTIN TMSB10

input RNA equivalent to approximately one cell

With pre-amplification With pre-amplification

B. Nucleic acid ampliciation required

Sieuwerts et al, Breast Cancer Res Treat 2009

Input for screening: Reference genes Epithelial specific control genes Leukocyte specific control genes Subtypes specific genes Prognosis/predictive markers Targets for therapy

Characterization of CTCs

Generating profiles 250 candidates mRNAs & 436 miRNAs for qRT-PCR

CTC fractions of 50 metastatic breast cancer patients

(collected before starting first line systemic therapy),

primary tumors of 8 of the patients

Final panel:

3 reference genes

(HMBS, HPRT1, GUSB)

2 leukocyte markers

(PTPRC, BST1)

55 CTC-specific mRNAs
10 CTC-specific miRNAs
can reliably be measured in 7.5 ml

with 5 CTCs vs. 0 CTCs or HBD

Sieuwerts, submitted, 2011

SLIDE 11

11

ER and Her-2 mRNA discrepancy

.

CTCs versus the primary tumor: ER Her-2

primary tumor primary tumor CTCs CTCs Characterization of CTCs

Generating profiles

Sieuwerts, submitted, 2011

.

unsupervised hierarchical cluster Epithelial CTC-specific gene expression patterns in CTC-enriched samples from metastatic BC patients ER-related Epithelial Proliferation GFR enriched

Sieuwerts, submitted, 2011

SLIDE 12

12

Molecular profiling of multiple genes is feasible if: linear pre-amplification is applied genes transcripts at least 10/100-fold higher expressed in CTCs compared to leukocytes can be considered Difference in Her-2/ER between primary tumor and CTC CTCs not always cluster along with their primary tumor Profiling CTC identifies three CTC positive subclusters driven by growth factor receptor, ER & proliferation

CTC characterization

Outline of the presentation

State-of-the-art of genomics research (focus BC) Introduction about liquid biopsies (CTCs, ctDNA, EVs) Application of CTC count and characteritics in metastatic cancer ARv7 positive CTCs predict progression in mCPCR Comparisons of ctDNA detection methods (dPCR & targetted NGS) Standard operation procedures/sample collection Examples of NGS in mBC and CRC ESR1 detection in CTCs and ctDNA Conclusions and outlook

SLIDE 13

13 CRPC treatment dilemma

Which agent to which patient at which moment? Cross-resistance!

Docetaxel Cabazitaxel Local disease Castration- sensitive mPC Castration- resistant mPC

Prostatectomy
Androgen deprivation

therapy (ADT)

ADT
Docetaxel
Abiratone?

Enzalutamide? Sipuleucel-T?

Docetaxel
Cabazitaxel
Abiratone
Enzalutamide
...

AR-V7 in CTCs

Predictive value of AR-V7 in CTCs 62 mCRPC patients: Abiraterone: N = 31 Enzalutamide: N = 31 AR-V7 in CTCs before treatment AdnaTest Enza group: 39% positive (65% docetaxel/abi) Abi group: 19% positive (16% docetaxel/enza)

SLIDE 14

14

AR-V7 in CTCs

AR-V7 50% PSA response Absent Present P-value Enzalutamide 10/19 (53%) 0/12 (0%) 0.004 Abiraterone 17/25 (68%) 0/6 (0%) 0.004

Study design

Research question: Should AR-V7-positive patients receive cabazitaxel? Hypothesis: AR-V7-positive patients still benefit from cabazitaxel given its different, AR-independent mechanisms of action CABARESC trial

Phase II trial mCRPC patients progressing after docetaxel and starting cabazitaxel Objective: to investigate whether budesonide (corticosteroid) diminishes the occurence of severe diarrhea as side-effect of cabazitaxel Blood for CTC enumeration and characterization before c1 and c3

SLIDE 15

15 Sample processing

Blood draws:

enumeration and characterization

CellSearch system RT-qPCR after preamplification for:

AR-WT
KLK3 (PSA)
AR-V1
AR-V3
AR-V7
AR-V9
TMPRSS2:ERG
EPCAM
KRT19
PTPRC (CD45)
HMBS, HPRT,

GUSB

AR-V7 and response to cabazitaxel

CTC response Yes No N AR-V7 Yes 3 12 15 No 2 8 10 N 5 20 25

20% response rate

SLIDE 16

16 Conclusies

CTCs are a strong prognostic factor for patients with mCRCP More reliable/sensitive than conventional monitoring tools (serum PSA, radiology, clinical symptoms) We developed a robust method to determine AR-V7 status of CTCs Presence of AR-V7 in CTCs predictive for resistance to abi and enza In 29 docetaxel pretreated patients 55% AR-V7 positive No difference in CTC and PSA response rates and median OS Indeed Cabazitaxel seems a valid treatment option for patients with AR- V7-positive CTCs

Circulation tumor DNA (ctDNA)

SLIDE 17

17

Outline of the presentation

State-of-the-art of genomics research (focus BC) Introduction about liquid biopsies (CTCs, ctDNA, EVs) Application of CTC count and characteritics in metastatic cancer ARv7 positive CTCs predict progression in mCPCR Comparisons of ctDNA detection methods (dPCR & targetted NGS) Standard operation procedures/sample collection Examples of NGS in mBC and CRC ESR1 detection in CTCs and ctDNA Conclusions and outlook

Circulating cell-free DNA (cfDNA) refers to small DNA fragments (~150 bp) shed by all types of cells into circulation Circulating Tumor DNA (ctDNA) is released by Tumor cells cfDNA can be isolated from plasma or serum

Crowley et al., Nature Review Clinical, 2013

Cell free tumor DNA (ctDNA)

Diaz & Bardelli , JCO 32 (2014) (FEB 20)

SLIDE 18

18

ctDNA detection by digital PCR and Ampliseq-NGS

Detectie 0,1% variant: 20ng input ~ 6000 haploïde genomen ~ 6000 templates 25000x coverage 6000 unieke moleculen 0,1% = 6 moleculen variant

Relationship DNA-input, limit of detection (LOD), and sequence coverage

SLIDE 19

19

Digital PCR

1-6 mutations Coverage <20,000 reactions Sensitivity (>0.01%) Costs €5-€10 per sample Costs system <100k€

NGS: WES/targeted

> 100 mutations Coverage > 100 to >50,000 reads Sensitivity (>0.01%-5%) Costs €250-€3000 per sample Costs system >200k€

Digital PCR versus NGS

Amplicon based Capture based

Targeted NGS & Molecule tagging

SLIDE 20

20

Targeted NGS-panels Amplicon based Thermo Ampliseq*

(45 genes; 3100 amplicons)

Thermo Oncomine

(~10 genes; ~40 amplicons)

Qiagen BC panel

(93 genes; 4800 ampicons)

* No molecule tagging Capture based NEBNext Direct HS panel

(50 genes; 200 baits)

NEBNext Direct prostate panel

(39 genes; 1000 baits)

Panel performance on references

SLIDE 21

21

Outline of the presentation

State-of-the-art of genomics research (focus BC) Introduction about liquid biopsies (CTCs, ctDNA, EVs) Application of CTC count and characteritics in metastatic cancer ARv7 positive CTCs predict progression in mCPCR Comparisons of ctDNA detection methods (dPCR & targetted NGS) Standard operation procedures/sample collection Examples of NGS in mBC and CRC ESR1 detection in CTCs and ctDNA Conclusions and outlook

Pulmonary Medicine Case: cfDNA mutations at baseline and at progressive disease

SLIDE 22

22

ctDNA detection by dPCR and Oncomine-NGS ctDNA detection by digital PCR and WES

20 40 60 80 100 120 140 coverage depth

Number of reads

SLIDE 23

23

Summary ctDNA methodology

Amount Input important: Define limit of detection [0.25% (5ng), 0.1% (20ng)] Feasibility shown for: digital PCR, OnTarget enrichment, Targetted NGS (using UMIs for single molecule analyses) WES is feasible (5% allele frequency) Negative controls essential Single molecule analyses increases sensitivity up to 0.1% allele frequency. Feasibility shown voor breast, colon and lung ctDNA analysis for lung cancer has entered the clinic

Acknowledgements

Medical Oncology Martijn Lolkema Ronald de Wit Stefan Sleijfer Agnes Jager All the patients for participating in

ur studies

Breast Cancer Genomics and Proteomics Maurice Jansen Nick Beije Jean Helmijr Silvia Vitale Mai Van Joan Bolt-de Vries Zahra Alawi Lindsay Angus Lisanne van Dessel Molecular Pathology Winand Dinjens Erik-Jan Dubbink Ronald van Marion Peggy Atmodimedjo Oncomine - Thermo Fisher Scientific

SLIDE 24

24

De laatste studie niet besproken; maar ter info Outline of the presentation

State-of-the-art of genomics research (focus BC) Introduction about liquid biopsies (CTCs, ctDNA, EVs) Application of CTC count and characteritics in metastatic cancer ARv7 positive CTCs predict progression in mCPCR Comparisons of ctDNA detection methods (dPCR & targetted NGS) Standard operation procedures/sample collection Examples of NGS in mBC and CRC ESR1 detection in CTCs and ctDNA Conclusions and outlook

SLIDE 25

25

~ 70% Breast Cancer (BC) are positive for the estrogen receptor (ER) BC-ER+ are eligible to Endocrine therapy (ET), such as aromatase inhibitors (AIs) ~ 40%

BC-ER+ acquire resistance during treatment and relapse in metastatic disease (MBC) ER acquired ESR1 mutations (mESR1) are a major mechanism of resistance to AIs Patients with mESR1 can benefit from treatment for example with Fulvestrant

Robinson et al., Nature Genetics, 2013 Fribbens et al., JCO, 2016

Metastatic Breast Cancer (MBC) Objectives of this study

Primary objective: ESR1 mutations more frequently present in CTCs at PD than at baseline in CTCs? Secondary objectives: ESR1 mutation detection rate in CTCs vs cf-DNA What clinical factors and therapies associate with ESR1 mutations

SLIDE 26

26

Methods [clinical characteristics]

Retrospective cohorts of MBC patients in CTC studies with 5 CTCs Baseline cohort: patients before the start of first-line endocrine therapy for MBC PD cohort: patients progressing under any line of endocrine therapy for MBC Baseline cohort n=43 Adjuvant endocrine therapy for 41% of the patients 81% of patients experienced progression in follow-up PD cohort n=40 Mostly progressing on first-line endocrine therapy (55%)

r second-line (30%)

93% received prior AI treatment Baseline and PD are separate cohorts, with only 6 matched samples available

Methods [clinical characteristics]

SLIDE 27

27

EDTA tubes

LIMITATIONS

Low Volume of cfDNA (<20 uL)
Low Concentration of cfDNA (<1 ng/uL)

LIMITATION

Only 200 uL of plasma available

QIAamp Kit

cfDNA

Workflow

Without With cfDNA input: 7.8 uL (maximum) cfDNA input: 2 uL (maximum)

SLIDE 28

28

Detection of ESR1 mutations in CTCs vs cf-DNA cut-offs for positivity in CTCs

D538G; 0.6% Y537S; 0.3% Y537N; 0.3% Y537C; 0.5%

Patient HBD

Cut-offs cfDNA

Patients HBD

SLIDE 29

29

Detection of ESR1 mutations in CTCs vs cf-DNA

sensitivity and reproduciblity

Detection of ESR1 mutations in CTCs vs cf-DNA

similarities Both available for 44 samples (baseline n=18, PD n=26) 3 out of 4 ESR1 mutations in CTCs confirmed in cf-DNA

CTC code baseline CTCs baseline cf-DNA Adjuvant therapy PD CTCs PD cf-DNA Progression

n therapy

Prior therapies for MBC CTC1581 Y537N (0.42%) Y537N (0.05%) none not available not available CTC1571 Y537N (3.77%) not available tamoxifen not available not available CTC1567 not available not available none Y537S (1.98%) Y537S (1.21%) tamoxifen CTC1587 not available not available tamoxifen D538G (0.84%) D538G (15.98%) fulvestrant AI CTC1406 not available not available tamoxifen D538G (1.13%) D538G (10.18%) AI

SLIDE 30

30

11 more ESR1 mutations identified in cf-DNA but not in CTCs Plasma baseline 2/18 (11%) vs PD 11/26 (42%); p=0.04

CTC code baseline CTCs baseline cf-DNA Adjuvant therapy PD CTCs PD cf-DNA Progression

n therapy

Prior therapies for MBC CTC798 D538G (0.14%) D538G (1.93%) none not available not available CTC1581 Y537S (0.39%) * Y537N (0.42%) Y537S (0.47%), Y537N (0.05%) none not available not available CTC1571 Y537N (3.77%) not available tamoxifen not available not available CTC1007 not available not available none Y537S (0.01%) Y537S (9.26%) fulvestrant AI CTC1364 not available not available none D538G (0.25%) D538G (40.05%) tamoxifen AI CTC1565 not available not available tamoxifen + AI D538G (0.14%) D538G (5.14%) fulvestrant AI CTC1569 not available not available none Y537N (0.25%) Y537N (1.96%) AI CTC1352 not available not available none D538G (0.47%) D538G (20.93%) AI tamoxifen CTC1567 not available not available none Y537S (1.98%) Y537S (1.21%) tamoxifen CTC1360 not available not available none D538G (0.52%) D538G (2.86%) AI CTC1587 not available not available tamoxifen D538G (0.84%) D538G (15.98%) fulvestrant AI CTC1406 not available not available tamoxifen D538G (1.13%) D538G (10.18%) AI CTC1393 not available not available none D538G (0.18%) & Y537C (0.23%) D538G (27.1%) & Y537C (12.96%) AI CTC1410 not available not available tamoxifen D538G (0.37%) D538G (23.84%) AI

Detection of ESR1 mutations in CTCs vs cf-DNA dissimilarities Conclusions and Discussion

ESR1 mutations more frequently observed in cf-DNA than in CTCs Much higher variant allele frequencies in cf-DNA Problem of leukocyte background in CellSearch-enriched CTCs ESR1 mutations rarely present in patients starting first-line endocrine therapy, but enriched after endocrine therapies Despite previous literature, not only after AI exposure ESR1 mutations are not mutually exclusive

SLIDE 31

31

Outline of the presentation

State-of-the-art of genomics research (focus BC) Introduction about liquid biopsies (CTCs, ctDNA, EVs) Application of CTC count and characteritics in metastatic cancer ARv7 positive CTCs predict progression in mCPCR Comparisons of ctDNA detection methods (dPCR & targetted NGS) Standard operation procedures/sample collection Examples of NGS in mBC and CRC ESR1 detection in CTCs and ctDNA Conclusions and outlook

Overview NGS platforms & panels

IonTorrent (Thermo Scientific):

Ampliseq-panels (Customized): 45 genes panel (<3000 amplicons) 21-genes CRC-panel (~1100 amplicons) Oncomine cfDNA-panels (<40 amplicons, ~ 160 hotspot mutations):

Lung (ALK, BRAF, EGFR, KRAS, MAP2K1, MET, NRAS, PIK3CA, ROS1, and TP53) Breast (AKT1, EGFR, ERBB2, ERBB3, ESR1, FBXW7, KRAS, PIK3CA, SF3B1, and TP53) Colon (APC, AKT1, BRAF, CTNNB1, FBXW7, GNAS, KRAS, MAP2K1, NRAS, PIK3CA,

SMAD4, and TP53)

MiSeq/Hiseq (Illumina):

WES (GATC; 120x coverage, paired end; 125bp) NebNext Direct Cancer HotSpot Panel (Bioke; ~190 hotspot mutations in 50 genes; ~150bp) OnTarget (Boreal; 96 hotspot mutations in 9 genes: BRAF, CTNNB1, EGFR, KRAS, FOXL2, GNAS, NRAS, PIK3CA, TP53) TrueSight170 (Illumina; 170 genes for DNA and RNA: SNV, CNV, fusion)

SLIDE 32

32

panels

Testing

Oncomine cfDNA (updated) panel Qiagen BC panel WES NEBNext panels

Clinical studies Oncomine Lung cfDNA Diagnostics (N>100) Oncomine Colon cfDNA CRC-IMPACT (N=64)

Oncomine Breast cfDNA Everolimus Biomarker (N=170)

Application of NGS-panels

“Buizenproef” study

Patient selection

Metastatic disease No current systemic treatment Known somatic variant

Experiments

cfDNA concentration cfDNA fragmentation Somatic variant detection

Van Dessel, L. et al. Mol Oncol. 2017 11:295-304

SLIDE 33

33 cfDNA concentration increased in EDTA 96 hours samples

P<0.001

Variant allele frequency (VAF) decreased in EDTA @ 96 hours

Somatic variant detected in ctDNA

11/16 patients (69%) 13/19 somatic variants (68%)