IMPLICATIONS OF NGS AND MOLECULAR TESTING BEYOND RAS, RAF AND MSINicola Normanno

ISTITUTO NAZIONALE PER LO STUDIO E LA CURA DEI TUMORI-IRCCS-FONDAZIONE G. Pascale – NAPOLISC Biologia Cellulare e Bioterapie

DISCLOSURE SLIDE

Personal financial interests (speaker’s fee and/or advisory boards): MSD, Qiagen, Bayer, Biocartis, Incyte, Roche, BMS, MERCK, Thermofisher, Boehringer Ingelheim, Astrazeneca, Sanofi, Eli Lilly

Institutional financial interests (financial support to research projects):MERCK, Sysmex, Thermofisher, QIAGEN, Roche, Astrazeneca, Biocartis

Non-financial interests: President, International Quality Network for Pathology (IQN Path); President Elect, Italian Cancer Society (SIC)

Outline of the talk

• Different applications of NGS• Genomic, trascriptomic, liquid biopsy

• Genomic biomarkers• HER2 (CNV, variants), Fusions, POLE, TMB, etc.

• Transcriptomic subtypes of CRC• CMS classification, therapeutic implications

Current and future applications of next generation sequencing (or massive parallel sequencing) in cancer

Frese Biology 2013

Targeted sequencing of DNA and RNA is the main NGS application currently used in the clinic

NGS panels may vary from fewgenes to hundred genes

An increasing number of centers are using panelscovering large number of genes

CNB=core needle biopsy; CNV=copy number variation; FFPE=formalin-fixed paraffin-embedded; FNA=fine needle aspirate; indel=insertion and deletion; Mb=megabase; MNV=multi-nucleotide variants; mut=mutations; no=number; SNV=single-nucleotide variant; TMB=tumour mutational burden

NGS panels for comprehensive genomic profiling

5

Assay name FoundationOneCDx™ MSK-IMPACT™ Molecular

Intelligence® Tempus xT ACE ImmunoID™16-18

TruSightOncology 500(assay under development)

OncomineTM

TumourMutational Load

assay

NeoTYPE® Discovery Profile CANCERPLEX

Sequencingplatform

Illumina HiSeq4000

Illumina HiSeq2500 Illumina NextSeq Illumina HiSeq

4000Illumina

NovaSeq 600Illumina

NextSeq 550DxIon GeneStudio

S5 series Not reported Illumina NextSeq

No. of genes 324 468 592 595 >20,000 523 409 326 435

Types of alterations

Substitutions,indels, CNVs,

rearrangements

Somatic SNVs, indels, CNVs,

rearrangements

Somatic non-synonymous

missense mutations

SNVs, indels, CNVs,

rearrangements

SNVs, indels, fusions

SNVs, MNVs, indels

Nonsynonymous and synonymous

SNVs

SNPs, indels, rearrangements

SNVs, indels, CNVs,

rearrangements

Tumour slide specifications

10 FFPE slides (20% tumour

nuclei)

FFPE tissue(50–250 ng) 15 FFPE slides

FFPE slides (20% tumour

nuclei)

FFPE tissue (50 ng DNA) Not reported FFPE tissue

(20 ng DNA) FFPE tissue

FFPE slides/blocks,

CNB, FNA (>20% tumour nuclei)(>50 ng DNA)

Patient-matched normal sample No Yes No Yes No No No No No

Reported TMB value (determination for TMB high)

mut/Mb(≥20)

mut/Mb(≥13.8)

mut/Mb(≥17)

mut/Mb(not reported)

Non-synonymous mut/Mb

(not reported)

mut/Mb(≥15)

mut/Mb(not reported)

mut/Mb(not reported) Not reported

Turnaround time <2 weeks <3 weeks 10–14 days 2–3 weeks 2–4 weeks Not reported 2–3 days 14–22 days 7–10 days

Analytical Sensitivity of the Different Approaches Used for cfDNA Analysis

Normanno CTR 2018

Methods based on quantitative PCR have a limit of detection (LoD) up to 0.005%

The Emulsion PCR-based technologies [Droplet Digital PCR (ddPCR) and Beads, Emulsion, Amplification, and Magnetics (BEAMing)] have a LoD ranging from 0.01 to 0.001%

Technologies based on quantitative PCR and emulsion PCR can analyze up to hundreds bases and interrogate a limited number of loci, usually up to 10

Massively parallel or next-generation sequencing (NGS) technologies allow sequencing from 200 Kb to 3.2 Gb with a sensitivity up to 0.01% for targeted panels

NGS assays for cfDNA testing

bTMB=blood TMB; cfDNA=cell-free DNA; CNA=copy number alteration; CNV=copy number variation; indel=insertion and deletion; NGS=next-generation sequencing; no=number; seq=sequencing; SNV=single nucleotide variant; TMB=tumour mutational burden

Test name* FoundationOne®

Liquid/bTMBGuardant360/

GuardantOMNI™ MutatorDETECT Unnamed PredicineATLASOncomine Pan-

Cancer Cell-Free Assay

AVENIO ctDNAKits

No. of genes measured 70/394 73/500+ 64 508 600 52 17/77/197

Sequencing platform

Illumina HiSeq 4000

Guardant Health Digital Seq

PlatformIllumina NGS‡ Illumina NGS Not reported Ion GeneStudio

S5 seriesIllumina NextSeq

Types of alterations

SNVs, indels, fusions, CNAs

SNVs, indels, fusions, CNAs

SNVs, indels, fusions, CNVs

SNVs, indels, CNVs

SNV, CNV, rearrangements

SNVs, indels, fusions, CNVs

SNVs, indels, fusions, CNVs

Sample requirement (20 ng cfDNA) 1–2 mL plasma

(5–30 ng cfDNA)

Two 10 mL tubes of peripheral

whole blood or 6–10 mL plasma‡

Plasma (single blood

draw)Plasma (5 mL)§ 20 ng cfNA 10-50 ng

cfDNA

Genomic alterations detected by NGS testing of cfDNA in CRC patients

Kato JCO Precis Oncol. 2019

Outcomes of mCRC patients receiving matched vs. unmatched treatment based on NGS testing of cfDNA

Kato JCO Precis Oncol. 2019

Outline of the talk

• Different applications of NGS• Genomic, trascriptomic, liquid biopsy

• Genomic biomarkers• HER2 (CNV, variants), Fusions, POLE, TMB, etc.

• Transcriptomic subtypes of CRC• CMS classification, therapeutic implications

BRAF V600EBRAF non-V600MSIDDR

POLE mutHER2 ampl

MET ampl

Gene fusion

RAS mut +/-PIK3CA/PTEN mut

PIK3CA/PTEN mut

Wild-type

anti-EGFR therapies

BRAF inh + anti-EGFR +/- MEK inh

anti-PD1/L1

double anti-HER2

Kinase inh

45% 8%

26%

8%2%2%

1%

1%

2%2%

2%MET inh

To be defined

To be defined

To be defined

Genomic markers in CRC

Slide courtesy of R. Dienstmann

Gene mutations assessed by NGS in mCRC patients enrolled in the CAPRI-GOIM trial

Genes with >10 mutated cases

Total mutated cases, n (N=182 analyzed)

Cases with multiple mutations, n Types of concomitant mutations (n)

KRAS 45 30* TP53 (18), PIK3CA ex9 (9), PIK3CA ex20 (5), FBXW7 (5), BRAF (4), MET (1), EGFR (1), SMAD4 (1), FGFR3 (1), ERBB2 (1), PTEN (1)

NRAS 13 5 TP53 (3), PIK3CA ex9 (1), MET (1)

BRAF 15 12† TP53 (9), KRAS (4), PIK3CA ex20 (3), FBXW7 (2), PIK3CA ex9 (1), SMAD4 (1), FGFR3 (1), FGFR2 (1)

PIK3CA ex9 16 14‡ KRAS (9), TP53 (8), PIK3CA ex 20 (2), NRAS (1), BRAF (1), MET (1), EGFR (1), ERBB2 (1)

PIK3CA ex20 10 7‡ KRAS (5), BRAF (3), TP53 (3), PIK3CA ex9 (2), FBXW7 (2), ERBB2 (1)

TP53 72 36KRAS (18), BRAF (9), PIK3CA ex9 (8), FBXW7 (5), NRAS (3), PIK3CA ex20 (3), MET (1), EGFR (1), SMAD4 (1), CTNNB1 (1), FGFR3 (1), ERBB2 (1)

*11 cases with KRAS mutated tumors had >2 concomitant mutations (maximum 5 mutations); †5 cases with BRAF mutated tumors had >2 concomitant mutations (maximum 4 mutations); ‡9 cases with PIK3CA mutated tumors had >2 concomitant mutations (maximum 4 mutations)

Ciardiello, Normanno et al Ann Oncol 2014

Heterogeneity Score (HS) in mCRC patients enrolled in the CAPRI trial

Normanno et al Ann Oncol 2015

The heterogeneity score (HS) was obtained by normalizing the frequency of mutant alleles for the fraction of neoplastic cells

The HS virtually corresponds to the fraction of neoplastic cells that carry a specific mutation

Trastuzumab plus lapatinib inHER2-positive mCRC (HERACLES) 914 patients with KRAS exon 2 (codons 12 and 13) wild-type mCRC were screened 48 (5%) had HER2-positive tumors

Sartore-Bianchi Lancet Oncol 2016

RR: 30%, SD 44% Median PFS 21 weeks

Plasma copy number of HER2 correlates with activity of trastuzumab + lapatinib in mCRC pts (Heracles A)

Siravegna CCR 2019

Pertuzumab plus trastuzumab for HER2-amplified mCRC (MyPathway)

Meric-Bernstam Lancet Oncol 2019

Raghav JCO Precis Oncol 2019

HER2 amplification in mCRC

Found in about 2% of CRC More frequent in patients with KRAS/NRAS/BRAF wild type tumors (5%) Associated with resistance to anti-EGFR moAbs in retrospective analyses (2nd/3rd line

treatment)

HER2 amplification as predictive biomarker of resistance to anti-EGFR MoAbs in mCRC

Sartore Bianchi Oncologist 2019

HER2 mutations predictive value in mCRC

Kavuri et al, Cancer Discov 2015

Neratinib in HER2/HER3 mutant cancer

Hyman Nature 2018

Gene fusions predictive value in metastatic CRC

Alteration Prevalence Targetabilityevidence Enrichment

NTRK1-3 fusion < 1% Case reports (> if right colon, RAS/BRAF wt, MSI)1

ALK fusion < 1% Case reports (> if right colon, RAS/BRAF wt, MSIcolitis-associated)2

ROS1 fusion < 1% Other tumors (> if right colon, RAS/BRAF wt)3

RET fusion < 1% Other tumors (> if right colon, RAS/BRAF wt)3

FGFR2-3 fusion <1% Other tumors (> if RAS mut)4

1Russo et al, Cancer Discov 2016; 2Yaeger et al, Gastroenterology 2016; 3Kloosterman et al, Cancer Res 2017; 4Clifton et al, ASCO 2018

Slide courtesy of R. Dienstmann

Best change in tumor size in adult patients with NTRK fusion cancer treated with larotrectinib

Hong ASCO 2019

Fusions

Presented By Scott Kopetz at 2019 Gastrointestinal Cancer Symposium

Presented By Katherine Clifton at 2018 ASCO Annual Meeting

Presented By Katherine Clifton at 2018 ASCO Annual Meeting

Tumor Mutation Burden (TMB) is a prognostic factor in mCRC

Innocenti J Clin Oncol 2019

Distribution of TMB in advanced colorectal cancer

Fabrizio J Gastrointest Oncol 2018

MSI and TMB in mCRC patients treated with ICIs

Schrock Ann Oncol 2019

MSI CRC heterogeneity

Marisa JNCI 2018

Mutation frequencies in human CRC

TCGA Nature 2012

The somatic mutation rates varied considerably among the samples

84% of the cases with a mutation rate of 8.24 per 106 (median number of non-silent mutations, 58) and 16% with mutation rates of >12 per 106 (median number of total mutations, 728), which we designated as hypermutated

POLE mutations in CRC

Domingo Lancet Gastro Hepat 2016

Negative Hyperselection of RAS and BRAF Wild-Type mCRC Patients Who Received Panitumumab-BasedMaintenance Therapy

The PRESSING panel: HER2 amplification;MET amplification;ALK/ROS1/NTRKs/RET fusions;SNVs in HER2/PIK3CAex.20/PTEN/AKT1;SNVs RAS (low mutant allele fraction);MSI

Morano JCO 2019

Genomic profile of quadruple-wt mCRC tumors

Rachiglio Cancers 2019

Genomic Profiles of Cases with Prolonged CetuximabBenefit and Primary Progressors

Woolston Cancer Cell 2019

What are the actionable alterationsobserved in a cancer type ?

(breast, lung, CRC, pancreas, prostate, stomach, liver)

ESCAT ranking of each alteration

% of patients presenting a Level I-II alteration

(minus the most frequent hotspot)

Value (nb to test to get a benefit)

ASSESSMENT OF THE VALUE OF MULTIGENE SEQUENCING: METHOD

Which alterations must betested in each disease ?

Recommendation on the use of multigene sequencing

Open questions

• Which patients should receive a comprehensive genomic profiling (CGP)?• Advanced or metastatic• Failure of previous lines of standard therapy• Progression after target therapy• Young patients, rare diseases, unusual clinical presentation, CUP…..• Fit for enrollment in a clinical trial or to receive additional treatments

• How should the result of a CGP test be interpreted?• Tumor heterogeneity: multiple actionable mutations in the same tumor• Clonal vs sub-clonal driver alterations• Tissue and cfDNA testing to assess clonal architecture of the disease• Monitoring clonal evolution of the disease• Results of CGP should be discussed in the Molecular Tumor Board

Outline of the talk

• Different applications of NGS• Genomic, trascriptomic, liquid biopsy

• Genomic biomarkers• HER2 (CNV, variants), Fusions, POLE, TMB, etc.

• Transcriptomic subtypes of CRC• CMS classification, therapeutic implications

Gene expression–based subtyping of CRC

CIMP, CpG island methylator phenotype; MSI, microsatellite instability; SCNA, somatic copy number alterations

Guinney Nat Med 2015

Loree CCR 2018

Relative prevalence of CMS at specific primary tumor locations

Intratumor heterogeneity of CMS

Alderdice J Pathol 2018

Central tumor Invasive front

Prognostic value of the CMS classification

Guinney Nat Med 2015

Overall Survival Relapse-free Survival Survival after relapse

CMS1; CMS2; CMS3; CMS4

Predictive value of the CMS classification

Aderka Lancet Oncol 2019

Transcriptomic Subtypes of Cases with Prolonged Cetuximab Benefit and Primary Progressors

Woolston Cancer Cell 2019

Immune phenotypes associated with CMS of CRC

Roelands Int J Mol Sci 2017

Schematic representation of CRC subtypes

Dienstmann Nat Rev Cancer 2017

Conclusions• Comprehensive genomic profiling may identify relevant prognostic and predictive

biomarkers in patients with mCRC• Comprehensive genomic profiling should be offered to patients who failed standard

lines of treatment for enrollment in clinical trials• In the near future, the availability of more targeted therapies for CRC patients will

lead to a greater use of comprehensive genomic profiling in the initial phases of the disease for a better molecular classification

• Integration of trascriptomics with genetic and clinical data might further improveprecision/personalized treatment of mCRC patients

• Results of molecular profiling of tumors sould be discussed within the MolecularTumor Board