Molecular classification of colorectal cancer Fred T Bosman University Institute of Pathology Lausanne
Current WHO classification of carcinoma of the colorectum Adenocarcinoma, NOS • Cribriform comedo-type adenocarcinoma Medullary carcinoma, NOS • • Micropapillary carcinoma • Mucinous adenocarcinoma • Serrated adenocarcinoma • Signet ring cell carcinoma Adenosquamous carcinoma Spindle cell carcinoma, NOS Squamous cell carcinoma, NOS Undifferentiated carcinoma
Can we use morphology for precision medicine? • Morphotypes have a poor relationship with molecular characteristics – K/NRAS mutations not reflected in morphology – MSI-H is relatively site-specific (right) and often of mucinous/medullary morphotype • Immunoscore is related to MS-status (MSI-H) • Grade not related to molecular profile • Budding? • Vascular invasion?
Pathways in the development of colorectal cancer • The chromosomal instability (CIN) pathway • The microsatellite instability (MIN) pathway • CpG Island Methylator Phenotype (CIMP) pathway • The dysplasia-carcinoma pathway in IBD
The adenoma-carcinoma sequence in the CIN pathway (After Vogelstein) + KRAS - P16 + telomerase - P53 - SMAD4 - APC Carcinoma Metastases Normal mucosa Low grade adenoma High grade adenoma Aberrant crypt focus
By array–CGH copy number variation is much higher in CIN (MSS) than in MIN (MSI-H) Xie T et al. 2012 PLoS One 2012;7:e42001
CNV heterogeneity: drivers and bystanders Xie T et al. 2012 PLoS One 2012;7:e42001
Familial adenomatous polyposis: APC/Wnt pathway
Colon ascendens cancer in the context of Lynch syndrome
Microsatellite instability A. MLH1 D. PMS2 B. MSH2 E. BAT26 (normal) C. MSH6 F. BAT26 (MSI)
MMRd CRC hypermutate TCGA Nature. 2012; 487: 330–337
Colorectal cancer 5y disease free survival of stage II + III patients by MSI Status I would not use ‘molecular grade’ (confusing terminology) but simply call them MSI/MSS
Is MSI a suppressor of lymph node and distant metastasis? Frequency Analysis Stage II Stage III Stage IV MSI-H 22% 12% 3.5% (86/395) (104/859)
The CIMP pathway
The dysplasia-carcinoma sequence in inflammatory bowel disease chronic promoterhypermethylation microsatellite instability inflammation -p53 -p27 low grade dysplasia -p16 -APC high grade dysplasia +KRAS adenocarcinoma + gain of function mutation - loss of function (through LOH, mutation or promotermethylation) 10 years ?
Pathways overlap CRC colorectal carcinoma MSS microsatellite stable FAP familial adenomatous polyposis MSI microsatellite instable MP mutator phenotype MYH MYH polyposis CI chromosomal instability (After D Snover)
Patterns of SMAD-4 expression in colon cancer: (a) complete loss of expression in tumor glands as compared with normal crypts (arrow); (b) non-homogeneous expression: loss of expression in the lower part of the field contrasts with marked expression in the upper part (R.Fiocca et al. In preparation).
SMAD4 and prognosis (stage III) (Yan P et al. Clin.Cancer Res. 2016) 1.0 / 0.9 // Proportion disease free p=0.003 0.8 / / / / / / / 0.7 / / // / /// / // / / / // / //// / / ////// ///// / / / / / // / / // / //// / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / // / / 0.6 / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / // / / / / / / / / / / / // // / /// / / / / / / // / // / // / / / // 0.5 / / / / / // // /// / / // // / // // / / / / / // / / / / / / / // / / / // // / / /// / / / / // / / / no loss expression present loss no expression 0.4 Years: 0 1 2 3 4 5 6 7 At risk: 819 709 602 535 488 417 42 2 145 114 85 74 66 61 7 0
MS- and SMAD4 status allow subtyping of CC (Roth et al. J Clin Oncol. 2012;30:1288-95) Recursive partitioning
Molecular markers differentiate T3N1 cases Roth et al. 2012 JNCI
KRAS mutations are not prognostic in stage II/III colon cancer patients but BRAF mutations are! (Roth et al. JCO 2010;28:466-74)
BRAF mutated CC express a common signature BRAFm (blue) versus BRAFwt (yellow ) Popovici et al. J Clin Oncol. 2012;30:1288-95
BRAFm signature as predictor of survival Survival after Overall relapse survival Will the BRAF signature be the companion test for BRAF inhibitor treatment? (Popovici et al. J Clin Oncol. 2012;30:1288-95)
E.Missiaglia et al. Ann Oncol. 2014;25:1995-2001
Are left and right different?
Proximal and distal colon carcinomas have different biology E.Missiaglia et al. Ann Oncol. 2014;25:1995-2001
Distal colon carcinomas progress faster after relapse E.Missiaglia et al. Ann Oncol. 2014;25:1995-2001
WHO classification of carcinoma of the colorectum Adenocarcinoma, NOS • Cribriform comedo-type adenocarcinoma Medullary carcinoma, NOS • Micropapillary carcinoma • Mucinous adenocarcinoma • Useful for patient stratification for • Serrated adenocarcinoma treatment decisions? • Signet ring cell carcinom a Adenosquamous carcinoma Spindle cell carcinoma, NOS Squamous cell carcinoma, NOS Undifferentiated carcinoma
Molecular subtypes of colon cancer Budinska et al. J.Pathol.2013;231:63-76
With so many different approaches what do you do? Create a consensus molecular classification! Dr.Steven Friend
Consensus molecular subtypes of CRC Guinney et al. 2015 Nat.Med. Under revision CMS4 – Mesenchymal CMS3 – Metabolic 23% 13% SNCA high Mixed MSI status , CIMP low Stromal infiltration SNCA low TGF- β β activation , Angiogenesis β β KRAS mutations Worse relapse free and overall Metabolic deregulation survival CMS2 – Canonical CMS1 – MSI Immune 37% 14% SNCA high MSI, CIMP high WNT activation Hypermutation MYC activation BRAF mutations Better survival after relapse Immune infiltration and activation Worse survival after relapse
Do molecular subtypes differ in survival? CMS1 MSI immune CMS2 Canonical CMS3 Metabolic CMS4 Mesenchymal
Consensus molecular subtypes: pathways involved Dienstman et al. Nature Reviews Cancer 2017;17:79-92
Molecular subtypes: host response pathways involved Dienstman et al. Nature Reviews Cancer 2017;17:79-92
Can a panel of immunohistochemical markers equal CMS classification? Anne Trinh et al. Clin Cancer Res 2017;23:387-398
BRAF mutated patients respond differently to targeted treatment. Are they different at molecular level? Barras et al. Clin Cancer Res. 2017;23:104-115
Which gene modules are involved? Barras et al. Clin Cancer Res. 2017;23:104-115
Prognostic significance Barras et al. Clin Cancer Res. 2017;23:104-115
Potential impact on precision treatment Barras et al. Clin Cancer Res. 2017;23:104-115
The nature of classifications
The paradigm shift From ‘what does it look like’ to ‘which key molecular mechanisms are involved’ – ER, PR, Her2/neu and breast cancer – KRAS and colon, lung, head and neck ….. cancer – ALK, EGFR, ROS1 and lung cancer – MSI and colon cancer – Immune checkpoint status (PD-1/PD-L1; CSK4) Advancing molecular technology – almost anything can be done on FFPE tissues – fast and cheap NGS – disease/target dedicated platforms (Mammaprint, Coloprint etc.)
Conclusions – For stratification of CRC TNM is not enough – A new layer of subgroups can be obtained with ‘classical’ molecular markers – New molecular technology reveals more heterogeneity (molecular and clinical) in CRC resulting in CMS classification – Understanding this heterogeneity will go along with the development of new molecular targets and accompanying predictive tests to be assessed in basket and umbrella trials – This requires large series of patients with detailed clinical data: large scale pooling of tumor genotypes based upon diagnostic testing
The PETACC3 consortium IPA Lausanne SAKK Bern Pu Yan Dirk Klingbiel Edoardo Missiaglia SIB Lausanne HUG Genève Mauro Delorenzi Arnaud Roth David Barras Vlad Popovici (now in Brno) GI oncology/Genetics Leuven Eva Budinska (now in Brno) Sabine Tejpa r Pfizer Oncology San Diego Scott Weinrich Pathology Genova Graeme Hodgson Roberto Fiocca Mao Mao Xie Tao
Recommend
More recommend
