quali sottogruppi biologici di elezione?

Roma, 10 maggio 2013 Meccanismi molecolari di resistenza alle terapie anti-egfr nella neoplasia colorettale metastatica: quali sottogruppi biologici di elezione? Andrea Sartore Bianchi Oncologia Clinica Molecolare Ospedale Niguarda Ca Granda, Milano

Different molecular alterations initiate the development of CRC, drive its progression and determine its responsiveness or resistance to antitumor agents Markowitz & Bertagnolli NEJM, 2009

Oncogenic activation of PI3K/AKT & RAS/MEK pathways in CRC Fearon ER, Annu. Rev. Pathol. Mech. Dis. 2011. 6:479 507

There is a clear molecular separation between hypermutated and non-hypermutated CRCs The Cancer Genome Atlas Network Nature 487, 330-337 (2012) doi:10.1038/nature11252

Diversity and frequency of genetic changes leading to deregulation of signalling pathways in CRC The Cancer Genome Atlas Network Nature 487, 330-337 (2012) doi:10.1038/nature11252

The Pandora s jar clinician EREG molecular biologist

The Pandora s jar pathologist EREG clinician

Established use of biomarkers in 2013 Clinical setting adjuvant first-line liver/lung resection Diagnosis of CRC mcrc MSI KRAS??? Molecular markers advanced chemorefractory disease KRAS* KRAS WT KRAS MT EGFR-targeted therapy layers of molecular selection clinical trials *NRAS, BRAF and PIK3CA (ex20) may further enhance selection for EGFR-targeted therapy

Outline 1. Primary resistance to EGFR-targeted therapy in mcrc KRASin CRC: a success of molecular selection for EGFR-targeted monoclonal antibodies Unresolved issues: increased detection sensitivity for KRAS selection; differences among KRAS alleles 2. Other candidate biomarkers of resistance: pros & cons Coexistence of different molecular alterations and multi-determinant analyses 3. Secondary resistance Liquid biopsies to monitor ongoing mutagenesis 4. Implications for the clinical decision making

Predictive markers for EGFR-targeted therapeutics in mcrc: a rationale dissection of a molecular pathway for personalized cancer therapy

100 80 60 40 20 0 0.01 0.1 1 10 100 Cetuximab Concentration (nm) SW48 DiFi DLD HT-29 SW-480 HCT-116 LoVo SW620 Benvenuti S etal. CancerRes, 2007 % Proliferation (BrdU incorporation)

Molecular selection based on KRAS: a successful story 1. It is a relevant molecular alteration in CRC biology 2. Testing provides a digital output (mutated/wt) 3. Reproducible 4. Status in metastatic deposits mirrors primary T 5. High specific negative predictive value: mutated patients display no benefit at all ( 0% RR) 6. Allows huge savings for health care systems

What is there beyond KRAS mutations? More on KRAS differences among KRASalleles : G13D mutations versus all others increased detection sensitivity for KRAS selection KRAS amplification What is beyond? solitary KRAS vs multiple signaling path mutations sensitive assessment in a phase III trial The myriad of mechanisms of EGFR resistance Her2 amplification in CRC liquid biopsies to monitor ongoing mutagenesis Integration of molecular knowledge in clinical practice

The hypothesis-generating study concerning the role of G13D Percentage alive 100 80 60 40 Any cetuximab therapy No cetuximab therapy* Overall survival p.g13d mutation Other KRAS mutation KRAS wild-type Log-rank p<0.001 Log-rank p=0.49 Log-rank p<0.001 20 0 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 Time since randomization Time since randomization Time since randomization or start of cetuximab (mo) or start of cetuximab (mo) or start of cetuximab (mo) *The no cetuximab group for all patients from the pooled data set is the best supportive care group from the C0.17 trial Test for interactionbetween p.g13d mutation vs other KRAS mutationsand overall survival benefit from cetuximab-based treatment (any vs none): P=.003 (HR, 0.30; 95% CI, 0.14-0.67) for OS and P=.05(HR, 0.47; 95% CI, 0.22-1.00) for PFS. No interaction between p.g13d mutation vs wild-type KRAS status De Roock W et al. JAMA 2010

isogenic cell lines In Vitroand In Vivo Effects of p.g13d Mutation on Cetuximab Sensitivity xenografts in immunocompromised mice De Roock W et al. JAMA 2010

Treatment Effect on PFS & OS by KRASAllele in Three Phase 3 Studies * * * Positive interaction test Peeters M et al. ESMO 2012

Pooled Analysis of Mutant KRAS Alleles Across Panitumumab and Cetuximab Studies * 1 1 * Positive interaction test 1 Tejpar S, et al. J Clin Oncol. 2011; 29:suppl 15; abstract 3511. Peeters M et al. ESMO 2012

Molecular selection based on KRAS: How deep should we go in testing? ME-PCR Courtesy of Alberto Bardelli

The importance of minor clones carrying KRAS mutations on the response to anti-egfr therapy Marchetti A, NEJM 2009; Molinari F et al, CCR 2011

Effect of low-frequency KRASmutations on the response to anti-egfr therapy in metastatic colorectal cancer Tougeron D et al., ASCO 212 J Tabernero discussant

KRASas the major player of resistance: not only mutations but also (rare) amplifications affects response to anti-egfr therapy Valtorta E, Misale S, Sartore-Bianchi A et al, Int J Cancer 2013

EGFR amplification Balanced polisomy Scattered amplification p EGFR CEP7 q Clustered amplification Ch 7

EGFR GCN and response to EGFR-targeted moabs 2.47 2.92 2.83 3.00 Siena S et al. J Natl Cancer Inst, 2009

EGFR gene copy number: which cutoff? disomy polisomy 100% disomy 30% polisomy 50% polisomy 60% polisomy response Moroni M, Sartore-Bianchi A, Veronese S & Siena S. Lancet Oncol 2008

Standardization of EGFR FISH in colorectal cancer: Results of an international inter-laboratory reproducibility ring study Slide-exchange program to compare EGFR GCN FISH testing results among five pathology reference Centres located in Belgium (University Hospital Gasthuisberg, Leuven), Italy (Ospedale Niguarda Ca Granda, Milano and Istituto Clinico Humanitas, Rozzano), Switzerland (Laboratory of Molecular Diagnostic, Istituto Cantonale di Patologia, Locarno) and United States(University of Colorado School of Medicine, Aurora). Molecular diagnosis of EGFRGCN by FISH varied largely among pathology centres, with fluctuations covering the whole range of proposed cut-offs of predictive usefulness from literature. Definition of a detailed scoring system and implementation of a comprehensive training program for molecular diagnostic laboratories are therefore necessary before including the test into clinical practice Sartore-Bianchi et al. J Clin Pathol, 2011

Activation of both PI3K/AKT & RAS/MEK pathways is common in CRC 1 co-activation upon binding to RTK 2 multiple molecular alteration whithin the same tumor Modified from Sarat Chandarlapaty, ASCO 2011

Overlapping among molecular alterations in mcrc KRAS 21 12 xpten 23 In mcrc there is an association between PIK3CA (exon 9) and KRAS mutations (14.7% in KRAS mutants vs 6.8% in KRAS wild types; (p=0.0006) 1 1 2 6 PIK3CA 5 2 1 BRAF 3 Sartore-Bianchi et al. PLoS One 2009 De Roock et al. Lancet Oncol2010

Effect of multiple molecular alterations within PI3K/AKT & RAS/MEK pathways on response to EGFR-targeted monoclonal antibodies Because of the occurrence of multiple molecular alterationswithin the same tumor, we investigated a cohort of 132 patients treated with EGFR-targeted moabs in the chemorefractory settingby separating patients according to the actual number of molecular abnormalities within the same tumor (i.e. none vs1 vs 2 alterations) among KRAS, BRAF, PIK3CAmutations and loss of PTEN. The probability of response was: 50.0%(22/44) among patients with no alterations, 4.2%(2/47) among patients with 1 alteration and 0%(0/23) for patients with 2 alterations, objective response rate (p<0.0001 0 ( quadruple negative ) 1 2+ number of molecular alterations 50% 5% 0% Sartore-Bianchi A et al. PLoS One 2009

Study Schema for Pivotal Phase 3 Panitumumab Study (Study 20020408) R A N D O M I Z E Panitumumab 6.0 mg/kg Q2W + BSC Best Supportive Care (BSC) PD PD Follow- up Follow- up 1:1 Primary Endpoint: Progression-free Survival (PFS) PD, progressive disease Peeters et al Clin Cancer Res 2013

Evaluated Gene Alterations in the EGFR Signaling Pathway ligand EGFR membrane PI3K/AKT pathways PTEN PIP3 PDK1/2 AKT1/2 PI3K PIP2 P P Grb hsos RAS RAS pathway Signaling RAF MEK1 MEK2 transduction cell survival apoptosis angiogenesis ERK1/2 ß-catenin p53 transcription cell survival proliferation angiogenesis migration Peeters et al Clin Cancer Res 2013

Increased PFS Observed in Patients with KRAS Wild-type Tumors Percent Event-free 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Wild-type Median Events / n (%) (weeks) Pmab + BSC 115 / 124 (93) 12.3 BSC Alone 114 / 119 (96) 7.3 0 4 8 12 16 20 24 28 32 36 40 44 48 52 Weeks Hazard Ratio = 0.45 (95% CI: 0.34 0.59) 0.59) Stratified Log Rank Test p < 0.0001 Percent Event-free 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Mutant Median Events / n (%) (weeks) Pmab + BSC 76 / 84 (90) 7.4 BSC Alone 95 / 100 (95) 7.3 0 4 8 12 16 20 24 28 32 36 40 44 48 52 Weeks Hazard Ratio = 0.99 (95% CI: 0.73 1.36) Quantitative interaction test p < 0.0001 Adapted from Amado et al JCO 4 Peeters et al Clin Cancer Res 2013

454 Pyrosequencing Exon Target Linker A Locus-specific sequences DNA barcodes Linker B PCR amplify exons of interest, anneal to beads Emulsion PCR Clonal amplification Deposition of beads onto PTP with 1.6M wells. 4 3 2 454 1 Basepair sequences resolved as flowgrams Technology is commercially available from 454 Life Sciences (Roche)

Joint Frequency Distribution of Mutations Across All Gene Pairs Gene - n AKT1 BRAF CTNNB1 EGFR KRAS NRAS PIK3CA PTEN TP53 AKT1 BRAF 1 CTNNB1 EGFR KRAS 2 1 3 NRAS 3 PIK3CA 2 1 13 PTEN 4 1 6 1 1 TP53 8 1 1 66 9 12 9 109 tumors had mutations in more than one gene 20 tumors had more than one mutation in a single gene in this analysis Peeters et al Clin Cancer Res 2013

WT* KRAS Subset: Response Rate by Central Assessment in the Panitumumab + BSC Arm Genotype BRAF NRAS PIK3CA PTEN n Response Rate (%) WT 63 18 TP53 Genotype n Response Rate (%) WT 32 19 MT 9 0 MT 49 12 WT 76 16 EGFR WT 82 15 MT 4 0 MT 0 N/A WT 74 14 AKT1 WT 69 15 MT 5 20 MT 1 0 WT 72 15 Beta- MT 7 14 catenin WT 72 13 MT 2 50 *WT in codons 12, 13, and 61 Peeters et al Clin Cancer Res 2013

WT* KRAS Subset: Relative Risk of Progression or Death of Mutant Genotypes Gene KRAS NRAS BRAF Genotype WT MT KRAS Wild-Type Efficacy Analysis Set PIK3CA PTEN TP53 WT MT WT MT WT MT WT MT WT MT *WT in codons 12, 13, and 61 0.001 0.01 0.1 1.0 10.0 100.0 Favors pmab plus BSC arm No difference in hazard Favors BSC alone arm Peeters et al Clin Cancer Res 2013

Mutations in PTEN and/or PI3K do not predict lack of progression on Panitumumab therapy Genotype Hazard Combination Groups N Ratio KRAS WT 153 0.392 KRAS, NRAS, BRAF all WT KRAS WT + (PTEN MT or PI3K MT) KRAS WT, PTEN MT, PI3K WT KRAS WT, NRAS WT + (PTEN MT or PI3K MT) KRAS, NRAS, BRAF all WT + (PTEN MT or PI3K MT) 103 18 8 16 10 0.352 0.185 0.109 0.193 0.282 0.01 0.1 1.0 10.0 100.0 Favors Pmab No difference Favors BSC plus BSC arm in hazard alone arm Peeters et al Clin Cancer Res 2013

Prognostic Value of the BRAF Mutant Genotype in the KRAS WT* Subset Gene Treatment WT N MT N Hazard Ratio BRAF BSC 52 6 2.551 Pmab + BSC 63 9 3.265 Total 115 15 2.389 0.01 0.1 1.0 10.0 100.0 Favors MT group No difference in hazard Favors WT group *WT in codons 12, 13, and 61 Peeters et al Clin Cancer Res 2013

Other players of primary resistance to EGFR-targeted therapy in mcrc: Her2 Bertotti et al. Cancer Discovery 2011

Correlation between HER2 amplification and therapeutic resistance to cetuximab in xenopatients In mcrc patients there is a progressive enrichment of HER2 amplification along with refinement of genetic selection Bertotti et al. Cancer Discovery 2011

Effect of anti-egfr and anti-her2 therapies in cetuximab-resistant, HER2-amplified metastatic colorectal cancer xenopatients Bertotti et al. Cancer Discovery 2011

HERACLES TRIAL HER2 Amplification for CoLo-rectal Cancer Enhanced Stratification An Open-Label Phase II Study of Lapatinib in Combination With Trastuzumab In Subjects With HER2-amplified Metastatic Colorectal Cancer (mcrc) KRAS Wild Type mcrc oxaliplatin-, irinotecan-, 5-FU, Cmab/Pmab resistant, HER2 amplified Cohort A Cohort B Treatment Cohort A: Lapatinib 1000 mg daily per os + Trastuzumab 4 mg/kg iv load, followed by 2 mg/kg iv weekly Cohort B: Pertuzumab 840 mg/kg iv load, followed by 420 mg/kg iv Q3weeks + Trastuzumab 8 mg/kg iv load, followed by 6 mg/kg on day 1 of each subsequent 3 week cycle Statistic Patients in this settings which would respond to a chemotherapy (H0) = 10%; Patients expected to respond to experimental treatment (H1) = 30%; Given α=0.05 and power=0.85, 27 patient (N) are required in each trial cohort (Cohort A 27 + Cohort B 27 = 54 patients overall). Each combination will be considered positive if at least 6 responses/27 patients are observed. Biomarkers Companion Study Cohort A: Patients will be sampled at baseline, every two weeks of each Q3wks cycle (i.e wks 2, wks 4, wks 6, etc), and at disease progression Cohort B: Patients will be sampled at baseline, on Day 1 of each Q3wks cycle, and at disease progression

Secondary resistance to targeted therapy in mcrc: an evolving landscape

Preclinical models of cetuximab-resistant CRC DiFi Lim1215 Cetuximab-resistantvariants of two colorectal CRC cellular models (DiFiand Lim1215 cells) that are highly sensitive to EGFR inhibition were generated Courtesy of Alberto Bardelli

CRC cells become cetuximab-resistant by two distinct mechanisms: 1. KRAS amplification 2. KRAS mutation DiFi DiFi R Lim1215 Lim1215 R Misale et al., Nature 2012

Does the DiFi patient exist? PR PR PD PD KRAS amplification in a patient with secondary resistance to panitumumab FISH by Emanuele Valtorta Ospedale Niguarda Ca Granda

A platform in place for studying resistance to EGFR-targeted moabs in mcrc at ONCG Start of cmab/pmab (chemorefractory patients) Objective tumor evaluation by RECIST (CT scan, MRI) as per clinical practice Magnetic resonance imaging as early indicator of clinical outcome (experimental) Treatment with anti-egfr moab Patient signs informed consent approved by EC Baseline blood sample for: -ctdna -mirna Analysis on FFPE tissue of: -KRAS/NRAS/BRAF/PIK3CA/PTEN -Her2, KRAS IHC & FISH Blood samples for ctdna and mirna at each time-point associated with response re-evaluation by RECIST (8-12 wks) until PD

Mutational status of the KRASgene before and after development of resistance to anti EGFR therapy Patient ID Anti EGFR sensitive tumor KRAS Mutational status (454 a or BEAMing b on tumor) EGFR-targeted monoclonal antibody Anti EGFR resistant tumor KRAS Mutational status (454 a or BEAMing b on tumor) Patient #1 wt pmab + irinotecan wt Patient #2 wt a pmab + irinotecan G13D a Patient #4 wt pmab + irinotecan G13D a Patient #5 wt cmab + irinotecan G13D a Patient #6 wt cmab + FOLFIRI G13D a Patient #7 wt a cmab + irinotecan wt a Patient #8 wt b cmab + irinotecan Q61H b Patient #9 wt b cmab + irinotecan G12D b G13D b Patient #10 wt b pmab wt b Patient #11 wt b pmab wt b KRAS amplified

Onset of KRASmutations giving rise to resistance is detectable in peripheral blood (liquid biopsies) Patient ID Anti EGFR sensitive KRAS mutational status in plasma Anti EGFR resistant KRAS mutational status in plasma Mutation Percentage Events Mutation Percentage Events Patient #8 wt 0.1% 4/46300 Q61H 1.12% 731/65400 Patient #9 wt 0.03% 0% 3/11600 0/16800 G12D G13D 0.48% 3.3% 89/18400 427/12500 Patient #10 wt 0% 0/85500 wt 0% 0/14200 Misale et al., Nature 2012

Patient #8 received a third-line treatment with weekly cetuximab and irinotecan, achieving partial response of liver metastasis maintained for 15 months He was enrolled in our study of secondary resistance to EGFR-targeted therapies and underwent liver biopsy at disease progression August 2009 liver biopsy of the same metastasis that initially responded and then progressed during cetuximab therapy Treatment with anti-egfr moab January 2011

Clinical-molecular synopsis RECIST evaluation CEA ctdna (liquid biopsies) Tumor biopsies

Montagut et al. Nature Medicine 2012 Bardelli & Janne Nature Medicine 2012

Conclusions 1. Molecular selection according to KRASstatus and other downstream effectors (BRAF, PIK3CA) provided a step forward to personalized cancer therapy in mcrc 2. Multiple molecular alterations do occur within the same tumor and multi-determinant evaluation may improve selection 3. Her2is an effector of secondary resistance and a target for therapy 4. We are gaining insights into mechanisms of acquired resistance and KRAS is still a key player (secondary mutations, amplification) 5. Tumor heterogeneity might hamper clinical selection and increased detection sensitivity improve detection of minor clones 6. Liquid biopsies are a tool for monitoring ongoing mutagenesis

Integration of new knowledges in clinical practice at Oncologia Niguarda Ca Granda mcrc chemorefractory WT EGFR-targeted therapy KRAS BRAF Her2 EGFR S492R BRAF MGMT met MT regorafenib clinical trials Anti-HER Tx panitumumab LGX818 + MEK162 temozolomide ALK/TRKA/ROS2 unselected NMS anti-alk TAS-102