ArticlePDF AvailableLiterature Review

Genetic Alterations of Metastatic Colorectal Cancer

Authors:

Abstract and Figures

Genome sequencing studies have characterized the genetic alterations of different tumor types, highlighting the diversity of the molecular processes driving tumor development. Comprehensive sequencing studies have defined molecular subtypes of colorectal cancers (CRCs) through the identification of genetic events associated with microsatellite stability (MSS), microsatellite-instability-high (MSI-H), and hypermutation. Most of these studies characterized primary tumors. Only recent studies have addressed the characterization of the genetic and clinical heterogeneity of metastatic CRC. Metastatic CRC genomes were found to be not fundamentally different from primary CRCs in terms of the mutational landscape or of genes that drive tumorigenesis, and a genomic heterogeneity associated with tumor location of primary tumors helps to define different clinical behaviors of metastatic CRCs. Although CRC metastatic spreading was traditionally seen as a late-occurring event, growing evidence suggests that this process can begin early during tumor development and the clonal architecture of these tumors is consistently influenced by cancer treatment. Although the survival rate of patients with metastatic CRC patients improved in the last years, the response to current treatments and prognosis of many of these patients remain still poor, indicating the need to discover new improvements for therapeutic vulnerabilities and to formulate a rational prospective of personalized therapies.
Content may be subject to copyright.
biomedicines
Review
Genetic Alterations of Metastatic Colorectal Cancer
Ugo Testa * , Germana Castelli and Elvira Pelosi
Department of Oncology, Istituto Superiore di Sanità, Viale Regina Elena 299, 00161 Rome, Italy;
germana.castelli@iss.it (G.C.); elvira.pelosi@iss.it (E.P.)
*Correspondence: ugo.testa@iss.it; Tel.: +39-649902422
Received: 20 August 2020; Accepted: 9 October 2020; Published: 13 October 2020


Abstract:
Genome sequencing studies have characterized the genetic alterations of dierent
tumor types, highlighting the diversity of the molecular processes driving tumor development.
Comprehensive sequencing studies have defined molecular subtypes of colorectal cancers
(CRCs) through the identification of genetic events associated with microsatellite stability (MSS),
microsatellite-instability-high (MSI-H), and hypermutation. Most of these studies characterized
primary tumors. Only recent studies have addressed the characterization of the genetic and clinical
heterogeneity of metastatic CRC. Metastatic CRC genomes were found to be not fundamentally
dierent from primary CRCs in terms of the mutational landscape or of genes that drive tumorigenesis,
and a genomic heterogeneity associated with tumor location of primary tumors helps to define
dierent clinical behaviors of metastatic CRCs. Although CRC metastatic spreading was traditionally
seen as a late-occurring event, growing evidence suggests that this process can begin early during
tumor development and the clonal architecture of these tumors is consistently influenced by cancer
treatment. Although the survival rate of patients with metastatic CRC patients improved in the last
years, the response to current treatments and prognosis of many of these patients remain still poor,
indicating the need to discover new improvements for therapeutic vulnerabilities and to formulate a
rational prospective of personalized therapies.
Keywords:
colorectal cancer; genomic alterations; metastasis; tumor heterogeneity; tumor evolution
1. Introduction
Colorectal cancer (CRC) is one of the most frequent cancers worldwide, corresponding to the
second in males and third in females most frequent tumor. CRC is the second most common cause of
cancer death in Europe [1].
Colorectal cancer is a highly heterogeneous disease that comprises dierent tumor phenotypes,
characterized by specific molecular and morphological alterations. CRC is caused by genetic alterations
that target tumor suppressor genes, oncogenes, and genes related to DNA repair mechanisms.
Depending on the origin of these mutations, CRC can be classified as sporadic (70–75%), hereditary
(5%), and familial (20–25%). Three major pathways are involved in CRC origin and progression:
(a) chromosomal instability (CIN); (b) microsatellite instability (MSI); (c) CpG island methylation
phenotype (CIMP). Each of these three dierent groups displays peculiar pathological, genetic,
and clinical characteristics [2].
CIN is the most common (85% of total CRCs) genetic mechanism occurring in CRC. CIN is
characterized by the acquisition of a consistent karyotypic variability, aneuploidy, chromosomal
and subchromosomal aberrations, gene amplifications and loss of heterozygosity. Allelic losses at
the level of chromosome arms 1p, 5q, 17p, 18p, 18q, 20p, and 22q are highly recurrent. A major
pathogenic consequence of this CIN consists in the loss of heterozygosity at tumor suppressor gene
loci. Furthermore, CIN tumors are associated with the accumulation of mutations at the level of
Biomedicines 2020,8, 414; doi:10.3390/biomedicines8100414 www.mdpi.com/journal/biomedicines
Biomedicines 2020,8, 414 2 of 29
several oncogenes, including KRAS and BRAF and of tumor suppressor genes such as APC and TP53.
The meta-analysis of the outcome of more than 10,000 CRC patients clearly indicated that CIN is
associated with a worse prognosis [3].
MSI involves several recurrent alterations in the microsatellite zone, without apparent structural
and numerical changes in the genome; approximately 15% of all CRCs have a high frequency of MSI
due to germline mutations in mismatch repair (MMR) system or somatic inactivation by promoter
hypermethylation of MLH1 gene [4].
CIMP pathway is responsible for 20–30% of total CRCs and is predominantly observed in the
proximal colon (30–40%) and more rarely in distal colon (3–12%) [4].
The Cancer Genome Atlas provided in 2002 the first genome-scale analysis of a large set (276) of
CRC samples, performing a comprehensive study involving exome sequencing, DNA copy number,
promoter methylation, messenger RNA and micro RNA expression evaluation [
4
]. This analysis
showed that CRCs can be classified according to their mutation pattern: (i) 16% of CRCs were found to
be hypermutated (75% displayed high MSI, usually associated with hypermethylation and silencing
of the MLH1 gene, whereas the remaining 25% exhibited mismatch-repair gene and polymerase
ε
(POLE) gene mutations); (ii) the non-hypermutated CRCs that formed the most consistent group of
tumors showed the recurrent mutations of APC,TP53,KRAS,PIK3CA,FBXW7,SMAD4,TCF7L2,
and NRAS genes; (iii) in hypermutated CRCs, the most frequently mutated genes were ACVR2A (63%),
APC (51%), TGFBR2 (51%), BRAF (49%), MSH3 (46%), MSH6 (40%), MYO18 (31%), TCF7L2 (31%),
and CASP8 (29%); (iv) APC (81% vs. 51%) and TP53 (60% vs. 20%) were significantly more mutated in
the non-hypermutated cancers compared to hypermutated cancers. Integrated analysis of the genetic
profiling showed that some pathways are recurrently altered in CRCs: (i) WNT pathway is altered
in 93% of all tumors (in 80% of cases due to biallelic inactivation of APC or activating mutations of
CTNNB1); (ii) PI3K signaling pathway is altered in 50% of non-hypermutated and 53% of hypermutated
CRCs; (iii) RTK-RAS signaling pathway is more frequently altered in hypermutated (80%) than in
non-hypermutated (59%); (iv) finally, TGF-
β
signaling pathway was much more frequently altered in
hypermutated (87%) than in non-hypermutated (27%) CRCs [4].
The study by TCGA showed the existence of three subtypes of CRC according to their
transcriptomic profile: microsatellite instability/CpG island methylator phenotype (MSI/CIMP);
invasive; chromosome instability (CIN). In a subsequent study, Zhang et al. carried out a proteogenomic
analysis on the CRCs previously characterized by TCGA [
5
]. This analysis showed the existence of a
limited correlation between mRNA and protein levels. Five CRC subtypes (from A to E) were identified
according to proteomic data: (i) B and C subtypes included all CRCs characterized by hypermutation,
MSI-H, POLE and BRAF mutations: B subtype was associated with the CIMP-H methylation subtype
of the TCGA study, absence of TP53 mutations and chromosome 18 loss; C subtype was associated
with a non-CIMP TCGA subtype. (ii) The A, D, and E subtypes were associated with the TCGA CIN
subtype. (iii) The E subtype displayed several remarkable features, such as the presence of TP53
mutations and chromosome 18q loss (both genomic alterations frequently associated with CIN CRCs)
and with HNF4A amplification and HNF4
α
protein abundance [
5
]. CRCs display frequent copy number
alterations (CNAs), particularly those characterized by CIN. However, only few CNAs are associated
with significant changes at protein level. Among the various CNAs, the chromosome 20 amplicon
was associated with the largest changes at both mRNA and protein level and is associated with HNF4
(hepatocyte nuclear factor 4, alpha), TOMM34 (translocase of outer mitochondrial membrane 34) and
SRC (SRC proto-oncogene, non-receptor tyrosine kinase) overexpression [5].
Copy number alterations (CNAs) show significant changes during the progression of colorectal
carcinogenesis from benign adenoma to CRC. Thus, chromosomal aneuploidies aecting chromosomes
7, 13, and 20q (all chromosomal gains) cooperate with APC mutations in the progression from adenoma
with low-grade dysplasia to adenoma with high-grade dysplasia. Losses of chromosomes 8p, 15q, 17p,
and 18q and gain of 8q are involved in tumor progression to infiltrating adenocarcinoma [6].
Biomedicines 2020,8, 414 3 of 29
The analysis of gene expression profiles obtained through the study of thousands cases of
colorectal cancers supported a classification of colon cancer, based on four major consensus molecular
subtypes (CMS), CMS1 to CMS4 (Table 1) [
7
]. CMS1 group (MSI immune subtype, including 14%
of all CRCs) is characterized at genetic level by hypermutation, hypermethylation, enrichment for
BRAFV600E mutations (observed in 40% of these tumors) and by pronounced infiltration of the tumor
microevironment by immune cells, particularly represented by T lymphocytes (both Cytotoxic CD8
+
and CD4
+
T helper) and natural killer lymphocytes; frequent in these tumors are mutations at the level
of APC (35%), TP53 (30%) and KRAS (25%) genes. Frequent in these tumors are mutations in MSH6,
RNF43,ATM,TGFBR2,BRAF, and PTEN genes. Predominantly, these tumors originate from precursor
lesions with a serrated histology, with preferential location at the level of proximal regions of the colon;
their prognostic outcome is intermediate but poor after relapse. The CMS2 subtype corresponds to
the canonical subtype (37% of CRCs) and is characterized by CIN-high, microsatellite stability (MSS)
and low levels of gene hypermethylation; a mutational profile typically observed in CIN-high CRCs,
including recurrent APC (75%), TP53 (70%), and KRAS (30%) mutations, whereas BRAF mutations were
absent; pronounced upregulation of WNT and MYC downstream targets, elevated expression of EGFR,
HER2, IGF2, IRS2, HNF4A, and cyclin; complex tubular histological structure, predominantly located in
the distal region of the colon. The CMS3 subtype corresponds to the metabolic subtype (10% of CRCs)
that is characterized by activation of glutaminolysis and lipidogenesis and by the presence of a
distinctive genomic and epigenomic profile compared with other CIN tumors, for the presence of
a mixed CIMP-H (20% of cases), MSI-H (15% of cases), hypermutation (30% of cases), and CIN-H
(54% of cases); at mutational level, frequent KRAS and APC mutations but less frequent TP53 and BRAF
mutations are observed; these tumors predominantly display papillary morphology and are located at
the level of both proximal and distal regions of colon. CMS4 corresponds to the mesenchymal subtype
(25% of all cases) and is characterized by the presence of tumors exhibiting activation of the pathways
related to epithelial-mesenchymal transition (EMT) and stemness (TGF-
β
signaling and integrins)
and overexpression of genes involved in extracellular matrix remodeling, complement-associated
inflammation, stromal invasion and angiogenesis; marked stromal cell infiltration at the level of
peritumoral microenvironment is a typical histological feature of these tumors; these tumors are
frequently CIN-H but rarely hypermutated, CIMP-H and MSI-H; at mutational level, frequent are
the mutations of APC,TP53 and KRAS, associated with rare BRAF mutations; at histological level,
these tumors are characterized by a desmoplastic reaction with high stroma; these tumors are associated
with a poor outcome compared with the other CMS subtypes [7].
Finally, there is a residual unclassified group representing 10–15% of all tumors with mixed features,
that seemingly represents a transitional phenotype or reflects an intra-tumoral heterogeneity [7].
Importantly, the CMS classification was predictive of chemotherapy and targeted-therapy response
in CRC patients with advanced/metastatic disease [810].
The CMS transcriptional classification was implemented through the analysis of microenvironment
signatures, showing consistent correlation between these two classification systems: CMS1 subgroup
was characterized by elevated expression of genes specific to cytotoxic T lymphocytes; CMS4 subgroup
was characterized by several microenvironmental features, including expression of monocytic markers
and a combined angiogenesis, inflammatory and immunosuppressive signature; at pathologic level,
CMS4 tumors display numerous infiltrating fibroblasts, producing cytokines and chemokines inducing
the angiogenetic and inflammatory phenotypes; CMS2 and CMS3 subgroups exhibit low inflammatory
and immune signatures [11].
Isella and coworkers have proposed a new transcriptional classification of CRC, allowing the
identification of five CRC intrinsic (CRIS) subtypes, displaying distinctive molecular, phenotypic
and functional features [
12
]. This classification was based on a methodological approach to limit
the impact of tumor stromal cells on the transcriptional classification of CRC. CRIS-A identifies a
subgroup of CRCs enriched for MSI-H, BRAF or KRAS-mutated tumors, with secretory mucinous
histology, with sustained glycolytic metabolism and inflammatory traits; CRIS-A englobes CRCs mainly
Biomedicines 2020,8, 414 4 of 29
corresponding to CMS1 and, at a minor extent, CMS4. CRIS-B identifies a subset of CRCs characterized
by an impaired dierentiation, activation of TGF-
β
signaling and epithelial to mesenchymal transition;
these tumors are mainly MSS and only in part MSI-H; these tumors are characterized by a poor
prognosis and by an elevated infiltration of fibroblasts; CRIS-B englobes both CMS4 and CMS1 tumors.
CRIS-C identifies a group of CRCs, CIN-H, and MSS, with absent KRAS mutations and exhibiting
elevated EGFR activity and MYC copy number gains; these tumors are particularly sensitive to EGFR
inhibitors; CRIS-C englobes CMS2 tumors and in part CMS4 tumors. CRIS-D tumors display a
number of typical features mainly represented by a stem-like phenotype associated with high WNT
signaling, a MSS status, strong enrichment of IGF2 overexpression/amplification and FGFR autocrine
stimulation; CRIS-D englobes both CMS2 and CMS4 tumors. CRIS-E is characterized by a Paneth
cell-like phenotype, an MSS status, numerous WNT-related features, and frequent TP53 mutations;
CRIS-E englobes both CMS2 and CMS4 CRCs [12].
Table 1. The gene expression-based consensus molecular classification of colorectal cancer.
Tumor
Subtype Frequency Gene Expression Genetic Abnormalities Tumor Location Prognosis
Signature
CMS1
Hypermutated
14%
Immune infiltration and activation
High PD-1 activation
Low stromal cell infiltration
SCNA low
Hypermutated
MSI high
CIMP high
KRAS (25%)
BRAF (40%)
APC (35%)
TP53 (30%)
Predominantly
proximal (74%)
Intermediate
Poor
prognosis
after relapse
CMS2
Canonical 40%
WNT and MYC activation
Elevated expression of EGFR, HER2, IGF2, IRS2 and HNF4A
Low immune infiltration and activation
SCNA high
No hypermutated
MSI low
CIMP negative
KRAS (30%)
BRAF (0%)
APC (80%)
TP53 (70%)
Predominantly
distal (80%) Good
CMS3
Metabolic 10%
Metabolic deregulation, with upregulation of several
metabolic signatures (glutaminolysis and lipidogenesis)
Low immune and stromal cell infiltration.
SCNA mixed
Hypermutated (30%)
MSI low; MSI high (15%)
CIMP mixed
KRAS (70%)
BRAF (10%)
APC (75%)
TP53 (30%)
Equally proximal
and distal Intermediate
CMS4
Mesenchymal
25%
Stromal infiltration
TGF-βactivation
Angiogenesis
Matrix-remodelling pathways
Complement-mediated inflammation
SCNA high
No hypermutated
MSI low
CIMP negative
KRAS (40%)
BRAF (5%)
APC (65%)
TP53 (55%)
Mainly distal (66%)
Negative
Usually
diagnosed at
advanced
stage
SCNA: somatic copy-nucleotide alteration.
The complex, variable and potentially confounding role of microenvironment in the evaluation
of the transcriptomic expression of CRC, highlights the need of performing analyses at single-cell
level as a tool to better define and understand intratumoral heterogeneity [
13
]. Only few studies
have explored single-cell transcriptomic in CRC samples. In this context, particularly relevant was
the study carried out by Li and coworkers investigating single-cell RNA sequencing on 969 tumor
cells derived from primary tumors of 11 dierent CRC patients and 622 normal mucosal intestinal
cells located near the CRC [
14
]. This analysis identified seven dierent cell clusters, corresponding to
epithelial cells, endothelial cells, fibroblasts, T lymphocytes, B lymphocytes, mast cells, and myeloid
cells [
14
]. The single-cell analysis allowed the identification of a larger set of dierentially expressed
genes compared with normal mucosa than the bulk analysis of gene expression [
14
]. Importantly,
EMT (epithelial-mesenchymal transition)-related genes resulted to be upregulated only at the level
of the cell population of cancer-associated fibroblasts but not at the level of epithelial cells [
14
].
The data obtained from single-cell transcriptomic allowed to define six dierent signatures of six
tumor cell types: epithelial dierentiated; epithelial stem, fibroblast, T cell, B cell, macrophages [
14
].
Biomedicines 2020,8, 414 5 of 29
The integration of the six cell type signatures together with the data of bulk signatures obtained through
the analysis of various cohorts of CRC patients allowed to define three tumor groups, defined as S1,
S2 and S3: S1 CRCs display a weak epithelial, an elevated myeloid and a strong fibroblast signature;
S2 CRCs exhibit intermediate level of all signatures; S3 CRCs show a strong epithelial signature,
associated with weak myeloid and fibroblast signatures [
14
]. In all the cohorts of CRC patients studied,
S3 CRCs display a better survival than the two other groups [
14
]. A more recent study based on the
analysis of >50,000 single cells from CRCs and matched normal tissues provided evidence that CRC
development is associated in all cases analyzed with changes at the level of epithelial, immune and
stromal cell compartments [
15
]. Interestingly, in the epithelium, five dierent tumor-specific stem and
progenitor-like cell populations were identified [
15
]. This single-cell analysis showed also that epithelial
tumor cells and cancer-associated fibroblasts are fundamental and essential for the assignment of each
CRC to a given CMS subtype [15].
Although single-cell transcriptomic techniques cannot be proposed for the clinical classification of
CRCs, their use may be of considerable support in the study of CRC patients undergoing immunotherapy
treatments or myeloid-targeted therapies [16,17].
Very few studies have explored the gene expression profile observed at the level of metastatic CRC
lesions. Kamal et al. reported the comparative analysis of the transcriptomic profile of primary tumors
and corresponding metastases (liver and lung metastases) in some CRC patients [
13
]. According to
the gene expression profile, two types of distant metastases were identified: M1 and M2 [
13
]. The M1
metastatic group is characterized by strong activation of inflammatory and immune response pathways
(including immune evasion pathways, such as those involving PD-1/-L1 signaling) and enrichment in
EMT activity. The M2 metastatic group exhibits MYC activation and cell proliferation [
18
]. Importantly,
treatment modifies the gene expression profile of metastatic lesions: the immune phenotype of M1
metastases is lost in post-treatment metastases; treatment induces an enrichment of EMT activity [
18
].
The analysis of CMS groups in metastases showed the absence of CMS3 and the presence of CMS1
in only few cases; the majority of metastases were classified as CMS2 (37%) or CMS4 (45%); 86% of
metastases were CMS4 in the M1 cluster, while 60% of metastases were CMS2 in the M2 cluster [
18
].
The comparison of gene expression in paired primary tumors and corresponding metastases showed
that FBN2 and MMP3 were the most dierentially expressed genes [18].
The incidence of CRC increases with the age. In a recent study, Lieu et al. on a large panel
of CRC samples reported the occurrence of CRCs in 7.8% of patients under the age of 40, 17.6% in
the age comprised between 40 and 49 years and 74.6% in patients with an age of 50 or older [
19
].
Overall genomic alterations were similar in the majority of genes currently mutated, with some
notable dierences: in MSS CRC patients, TP53 and CTNNB1 alterations were more common in
younger patients with CRC [19]; in the MSI-H cohort, most of genes displayed a similar frequency of
alterations in the two age groups, but significant dierences were observed at the level of APC and
KRAS alterations more frequent among younger than older patients and BRAF alterations markedly
more recurrent among older than younger CRC patients [19].
The progresses made in primary and adjuvant treatments of CRC patients have led to an
improvement of the survival times of these patients. The optimal treatment of CRC patients would
imply complete surgical ablation of primary tumor and metastases. However, 25–30% of CRC patients
display at diagnosis an advanced disease stage with metastatic diusion; furthermore, a remaining
20% of patients develop metachronous metastases after standard treatments. Therefore, a significant
proportion of CRC patients need an ecacious medical treatment to induce the regression of tumor
cells that cannot be removed by surgery. The current medical treatment implies first line chemotherapy
or radiotherapy that can be performed either before surgery in a neoadjuvant setting or after surgery in
an adjuvant setting. Current chemotherapy treatment implies either single-drug treatment involving
fluoropyrimidine (5-FU) and multiple-drug regimens, based on the use of irinotecan (IRI), capecitabine
(CAP) or oxaliplatin (OX), such as FOLFOX (5-FU +OX), FOXFIRI (5-FU +IRI), CAPIRI (CAP +IRI)
or CAPOX (CAP +OX) [1].
Biomedicines 2020,8, 414 6 of 29
The studies carried out in the last years have shown that CRC exhibits a clinically relevant
molecular heterogeneity related to various genetic and non-genetic mechanisms. The identification
of molecular subtypes of CRCs helped to identify new strategies of treatment for selected groups of
patients (targeted therapy): (i) the presence of KRAS or NRAS mutations allowed the identification
of a group of CRC patients refractory to EGFR inhibitors; (ii) the absence of KRAS, NRAS, BRAF,
and PIK3CA/PTEN mutations (CRC “wild-type”) identifies a group of CRC patients responsive to
EGFR inhibitors; (iii) CRCs bearing BRAF
V600E
mutations have a poor prognosis and are responsive
to targeted inhibition in combination; (iv) CRCs with HER2 amplifications display sensitivity to
dual HER2 blockade; (v) CRCs bearing rare kinase fusion events are targetable with specific kinase
inhibitors; (vi) MSI-H and POLE hypermutant CRCs are particularly sensitive to treatment with
immune checkpoint inhibitors; (vii) CRCs with a mesenchymal phenotype display immunosuppressive
mechanisms that could be removed through combined immunotherapy treatments [20].
The strategy recommended by the National Comprehensive Cancer Network (NCCN) for the
targeted therapy of metastatic CRC patients implies a dierential treatment according to the RAS
mutational status and to the colon location of the primary tumor: (i) for patients with left colon mCRC,
RAS-WT it is recommended an initial therapy based on EGFR inhibitors, and a subsequent therapy based
on mutational status for BRAF mutations (BRAF inhibitors), HER2 amplifications (HER2 inhibitors)
BRAF/HER2-WT (anti-PD-1/L1 if deficient in mismatch repair (MMR); anti-VEGF if proficient in MMR);
for patients with right colon mCRC it is recommended a therapeutic approach similar to that adopted
for RAS-mutant patients; for patients with mCRC, RAS-mutant it is recommended a dierential
therapy according to the MMR status: for patients deficient in MMR it is recommended a first-line of
therapy based on anti-PD-1/L1 and a second line based on anti-VEGF inhibitors, whereas for patients
proficient in MMR, a first line based on anti-VEGF inhibitors and a second line of therapy based on
best supportive care therapy are recommended [21].
A large body of molecular data on the genomic abnormalities observed in CRC has been generated;
the majority of these studies focused on primary tumors. However, recent studies have characterized
the molecular abnormalities observed in metastatic CRC. Some studies have molecularly characterized
metastatic lesions with their corresponding primaries. The present review paper reports a detailed
analysis of these recent studies on the characterization of metastatic CRCs, supporting the view that a
better understanding of the molecular alterations and of their heterogeneity may improve the treatment
outcome of these patients.
2. Genetic Abnormalities in Metastatic CRC
Few studies have explored the frequency of recurrent genetic alterations in metastatic CRC patients.
In 2017, Zehir and coworkers reported the mutational landscape of 10,945 metastatic tumors,
including 975 metastatic CRCs, as encountered in clinical practice [
22
]. This study showed the presence
of four recurrently mutated genes, represented by APC,TP53,KRAS, and PIK3CA. Furthermore,
according to the somatic tumor burden, metastatic CRCs can be distinguished into three groups:
normal, hypermutated, ultramutated. The metastatic CRCs with a high mutational burden displayed a
dominant MMR signature. Finally, 35% of metastatic CRCs showed actionable somatic alterations [
22
].
The study carried out by Zehir et al. was based on targeted gene analysis [
22
]. A more recent study
by Priestley et al. involved deep whole-genome sequencing of 2399 metastatic solid tumors, including
372 CRCs [
23
]. Metastatic CRCs are among the tumors displaying the highest levels of single-nucleotide
variants (SNVs), with only urinary tract, esophagus, lung cancers and melanoma exhibiting higher
levels among 20 dierent types of metastatic cancers [
23
]. Only 4% of metastatic CRCs displayed an
MSI genotype/phenotype, a frequency that is lower than that reported for primary CRC, a finding that
can be explained by the lower tendency of these tumors to metastasize [
13
]. Copy number alterations
are frequent in metastatic CRC; an extreme form of CNA can be caused by whole genome duplication
(WGD), an event frequent (>60% of cases) in metastatic CRCs, among the metastatic tumors most
frequently showing WGD [
23
]. Metastatic CRCs displayed a mean number of total candidate driver
Biomedicines 2020,8, 414 7 of 29
events (6.5 per patient) only slightly higher than the mean number (5.7 per patient) observed in 20
dierent metastatic cancers [
23
]. The whole-genome sequencing (WGS) approach allowed to accurately
define the frequency of genetic alterations occurring in mCRC at the level of genes possessing oncogenic
activity when mutated or of tumor suppressor genes (Figure 1). The analysis of the co-mutation pattern
of driver genes showed negative associations within the same transduction pathway for KRAS-BRAF
and KRAS-NRAS, for APC-CTNNB1, for APC with BRAF and RNF43 [
23
]. Interestingly, this study
showed in 9 CRC patients with absent APC driver mutations, the occurrence of in-frame deletion of the
complete exon 3, leading to activation of the WNT and
β
-catenin pathway [
23
]. Furthermore, 5.4% of
mCRC samples displayed an amplification of CDX2, acting as a survival oncogene for these tumor
cells [
23
]. The exploration of the mutational spectrum of metastatic CRC indicates that only 30% of
these tumors possess biomarkers with either an approved therapy or with strong biological evidence
or clinical trials that are actionable [23].
Figure 1.
Frequency of the most recurrent gene alterations observed in metastatic CRC patients.
The data on the frequency of the major genetic alterations were reported by Priestly et al. [
23
] and were
based on the wide-genome sequencing analysis of 372 metastatic CRC patients.
Particularly relevant was the study carried out by Yaeger et al. [
24
] who reported the sequencing
analysis of most 1134 CRCs, including 979 patients with metastatic disease. These tumors corresponded
Biomedicines 2020,8, 414 8 of 29
to three dierent molecular groups: POLE mutant (0.7%), MSI-H/hypermutated (8.7%) and MSS (90.5%),
with predominant left colon localization of MSS tumors and predominant right colon localization of
POLE and MSI-H tumors [
24
]. The WNT pathway resulted to be altered in 85% of MSS tumors and in
93% of MSI-H tumors: APC gene alterations were more frequent in MSS CRCs than in MSI-H CRCs
(81% vs. 61%), while CTNNB1 and RNF43 gene alterations were less frequent in MSS CRCs than in
MSI-H CRCs (6% vs. 25% and 4% vs. 53%, respectively) [
24
] (Figure 2). Other remarkable dierences
in the rates of several genetic alterations between these two types of metastatic CRCs are represented
by the more frequent alterations of ERBB3,PIK3CA,PIK3R1,PTEN,NF1,BRAF,BRCA1, and BRCA2
gene alterations in MSI-H CRCs than in MSS CRCs [
24
]. (Figure 2) The analysis of mostly recurrently
mutated genes in MSS CRCs showed a mutational frequency of 79% for APC, 78% for TP53, 44% for
KRAS, 18% for PIK3CA, 16% for SMAD4, 10% for TCF7L2 and 10% for FBXW7 [
24
]. The analysis of the
frequencies of some gene mutations in early-stage tumors, primary metastatic CRC and metastases
from metastatic CRCs showed that most of these mutations do not display significant dierences, but a
minority of them are stage-related: the frequency of TP53 mutations progressively increases from
early-stage to primary mCRC and to metastases of mCRC; FBXW7 mutations are more frequent in
early-stage and primary mCRCs than in metastases of mCRC; ERBB2 mutations are more frequent in
early-stage than in metastatic CRCs [
24
]. BRAF mutations display a tendency to be more frequent in
metastatic CRC than in early-stage CRC [
24
]. This study also showed some remarkable dierences
between primary tumor sites, i.e., right colon or left colon. Right-sided primary mCRC displayed
fewer DNA copy-number alterations than left-sided mCRC; furthermore, an enrichment of genetic
alterations in KRAS,BRAF,PIK3CA,PTEN,AKT1,RNF43,SMAD2, and SMAD4 was observed in
right-sided primary mCRC and in APC and TP53 in left-sided primary mCRC [
24
]. Left-located mCRC
had a significantly better overall survival than right-located mCRC [
24
]. The analysis of the overall
survival in various molecular subgroups of mCRCs showed a poor survival for patients bearing KRAS
mutations alone or in combination with PI3K pathway mutations. These CRCs showed also a greater
tendency to have multiple first sites of metastases [24].
Using a multigene panel sequencing, Belardinilli and coworkers have explored the co-mutational
profile of metastatic CRC; this study involved the analysis of 779 metastatic CRC primary tumors [
25
].
The results of this analysis showed the existence of positive associations between EGFR and KRAS,
EGFR and SMAD4,BRAF and PTEN, and NRAS and TP53 mutations, whose biological and clinical
significance is at the moment unknown [
25
]. Importantly, according to the presence of TP53 and KRAS
mutations, metastatic CRCs can be subdivided into four dierent groups: MAP1, characterized by the
co-mutation of TP53 and KRAS and subdivided into a less frequent MAP1.1 subgroup, in which TP53
and KRAS mutations are associated with other recurrent mutations, such as PIK3CA,FBXW7,SMAD4
and PTEN mutations and a more frequent MAP 1.2 subgroup in which TP53 and KRAS mutations are
not associated with other recurrent mutations; MAP 2, characterized by the mutation of the KRAS
gene and subdivided into a MAP 2.1 subgroup in which KRAS mutation is associated with highly
recurrent PIK3CA mutations and a MAP 2.2 subgroup in which KRAS mutations are not associated
with other recurrent mutations; MAP 3, characterized by TP53 mutations, subdivided into a MAP 3.1
subgroup in which TP53 mutations are associated with recurrent PIK3CA,BRAF,NRAS, and SMAD4
recurrent mutations and a MAP 3.2 subgroup in which TP53 mutations are not associated with other
recurrent mutations; MAP 4, characterized by the absence of TP53 and KRAS mutations, subdivided
into a less frequent 4.1 subgroup, characterized by highly recurrent BRAF mutations and recurrent
PIK3CA,NRAS and FBXW7 mutations and a more frequent 4.2 subgroup, characterized by absence of
recurrent mutations [25].
Biomedicines 2020,8, 414 9 of 29
Figure 2.
Frequency of the most recurrent genetic alterations observed in metastatic CRC patients
(data reported by Yaeger et al., 2018) [
24
]. Top Panel: most recurrent genetic alterations observed in
the whole population of metastatic CRC patients; Middle Panel: most recurrent genetic alterations
observed in the population of metastatic CRC patients subdivided into MSI-H and MSS; Bottom Panel:
tumor location in metastatic CRC patients exhibiting either MSI-H or MSS.
BRAF-mutant CRC represent a peculiar subgroup of mCRCs. In the metastatic setting,
600E
BRAF mutation occurs in 10% of cases and is associated with a poor prognosis [
4
].
Among
V600E
BRAF-mutated CRCs, two subgroups have been distinguished according either to
the activation of KRAS/mTOR/AKT/4EBP1 pathway (BM1 subtype) or to the deregulation in the cell
cycle (BM2 subtype) [
26
]. In addition to
V600E
BRAF-mutated CRCs, there is a rarer (occurring in 2%
of metastatic CRC patients) subgroup of
nonV600E
BRAF-mutated CRCs; these
nonV600E
BRAF-mutated
CRCs involve mutation at the level of 19 dierent codons [
27
,
28
]. Patients bearing mutations at
the level of codons 594 and 596 seem to form a distinct subgroup with longer overall survival
Biomedicines 2020,8, 414 10 of 29
compared with
V600E
BRAF-mutated patients [
27
,
29
]. A recent study reported the classification
of BRAF-mutated CRCs into three sub groups: BRAF mutations activating RAS-independent
as monomers (Class1 V600E); BRAF mutations activating RAS-independent signaling as dimers
(class 2 codons 597/601); BRAF mutations activating RAS-dependent signaling with impaired kinase
activity (class 3 codons 594/596) [
30
]. Class 3 BRAF-mutated metastatic CRCs were more frequently left
sided and without peritoneal metastases compared to class 1; class 3 tumors have an overall survival
comparable to that of BRAF wt tumors; while class 1 and 2 tumors have a poorer overall survival than
BRAF wt tumors [30].
3. Comparative Analysis of the Genetic Abnormalities of Primary Metastatic CRCs and of Metastases
Several studies have performed comparative lesion sequencing of paired primary metastatic
CRCs and of corresponding metastases.
About 20% of patients with CRC already have metastases at diagnosis [
24
]. The patterns of
metastasis of colon and rectal cancer were recently explored in a very large cohort of patients (49,096,
31,285 with colon cancer and 17,811 with rectal cancer: 30% of colon cancer and 31% of rectal cancer
patients had metastases) [
31
]. Of all patients with metastatic cancer, the most common sites of metastasis
were the liver (70% in both colon and rectal cancer) and the thorax (32% in colon cancer and 47%
in rectal cancer), followed by the peritoneum for colon cancer (21%) and the bone for rectal cancer
(12%); nervous system metastases were more rare, being observed in 5% of colon cancer and 8% in
rectal cancer [
31
]; thoracic metastases were more frequent in lower tumor stages, particularly in rectal
cancer, whereas the relative frequency of liver metastases increased with tumor stages; liver metastases
were most frequently solitary metastases (in 48% of colon and 45% of rectal cancer); lung metastases
were frequently observed in association with liver metastases (73% in colon cancer and 63% in rectal
cancer) [31].
Several comparative sequencing studies have shown a high concordance in the genomic profile
between primary and metastatic CRCs. Jones and coworkers through a comparative sequencing
analysis of a small number of patients observed a high degree of concordance between primary tumors
and metastases [
32
]. Vakiani et al. reported the analysis for KRAS,NRAS,BRAF,PIK3CA, and TP53
genes of 84 CRCs in whom tumor tissue from both primary and metastatic sites was available [
33
].
The results of this analysis showed that: the frequency of KRAS,NRAS, and PIK3CA mutations
was similar in metastatic versus primary tumors; TP53 mutations were more frequent in metastatic
versus primary tumors (53% vs. 30%, respectively), whereas BRAF mutations were significantly less
frequent (1.9% vs. 7.7%, respectively) [
33
]. In a subsequent analysis, 69 CRC patients were explored
for their mutational profile by NGS in primary and metastatic tumor tissues [
33
]. The results of
this study showed that 79% of the mutations were shared between primary and metastatic tumors.
Particularly, a high degree of concordance at the level of early occurring and recurrent mutations was
observed [
33
]. No discordant mutations in KRAS/NRAS and BRAF were observed; the only private
mutations, defined as mutations observed only in the primary or the metastatic tumor, were observed
at the level of APC,PIK3CA,SMAD4 and TP53 genes [
34
]. These findings have supported the view
that genetic alterations occurring early during colorectal cancer genesis, such as APC,KRAS,NRAS,
and BRAF mutations are maintained during the process of tumor evolution up to the final level of
tumor metastases [34].
In some contrast with these studies, Vermaat et al., using next generation sequencing, showed a
high degree of mutational discordance between primary and metastatic samples, with 52% and 86% of
dissimilarities of KRAS and EGFR mutational status between paired primary and metastatic tumor
samples. Modest variability was reported for HRAS (34%), PIK3CA (19%), FLT1 (10%), NRAS (10%)
and BRAF (14%) [35].
Lim et al [
36
] performed an analysis of 34 CRC patients with liver metastases by sequencing
(whole exome and RNA sequencing) both primary tumors and metastases and showed in these patients
frequent mutations of APC (65%), TP53 (68%), KRAS (24%), TCF7L2 (21%), PIK3CA (18%), NRG1 (18%),
Biomedicines 2020,8, 414 11 of 29
FBXW7 (15%), SMAD4 (15%), CARD11 (12%), and BMI1 (9%) [
36
]. Based on the absence or presence
of mutations in liver metastases, the mutations occurring in these patients were classified into three
dierent classes: class 1, mutations shared between primary tumors and liver metastases (57.6% of
all mutations); class 2 mutations present only in primary tumors (20.9% of all mutations); class 3
mutations, detected in only liver metastases (21.5% of all metastases) [
36
]. Importantly, the frequency
of class 1 mutations was highly variable across individual patients (ranging from 25% to 92%),
thus suggesting that the presence of a clonal selection during metastasis formation is an event highly
variable among patients; a decreased clonality during metastasis formation was usually associated
with a high-mutational concordance between primary tumors and metastases, whereas an increased
clonality during metastasis formation was usually linked with low mutational concordance between
primary tumors and liver metastases [36].
Vignot et al. reported a mutational analysis by targeted NGS on surgical samples from primary
and matched metastatic tissues from 13 CRC patients [
37
]. A global concordance rate for mutations of
78% was observed between primary and metastatic tumors; this concordance raised to 90% for the 12
most recurrent mutations occurring in CRC [
37
]. On 17 pathways explored, only two pathways were
upregulated in metastatic tissues compared to primary tumors [37].
Tan and coworkers reported a detailed analysis of the mutational profile and of CNAs of 18
matched primary and metastatic tumor tissues by high-depth sequencing of over 750 cancer-associated
genes and copy number profiling, supporting a high concordance of primary tumor and metastases [
38
].
Particularly, their results showed a median of 79.3% of somatic gene mutations present both in the
primary and metastasis and 81.7% of all alterations present in both primary tumors and metastases [
38
].
Private alterations, primary-specific or metastasis-specific are observed at lower allelic frequencies [
38
].
The mutations most frequently occurring only at the level of metastases are represented by MLL3,
FAT1, and GNAS gene mutations [
38
]. Interestingly, distinct mutational signatures are observed in
shared variants and private variants [
38
]. The analysis of copy number alterations similarly showed a
conserved pattern between primary tumors and metastases: chromosomal regions of allelic imbalance
were similar in the matched primary tumor and metastasis; focal gains and losses of genes commonly
amplified or deleted in cancer were similar in the primary tumors and metastases [
38
]. These findings
supported a model of linear evolution in most CRC patients with liver-limited metastatic disease.
Several studies reported a concordant mutation profile for the main CRC driver genes, including
KRAS,TP53,APC,PIK3CA,BRAF, and NRAS between primary tumors and metastatic lesions regardless
of the temporal relationship between metastases (synchronous or metachronous) [
39
,
40
]. Only in a
minority of cases (7–15%) metastases diered from paired primary tumors [
39
,
40
]. Similarly, Jesinghaus
and coworkers have explored the mutational landscape of 24 primary MSS CRCs and of their respective
metastases: A high degree of genetic concordance of the mutations aecting the driver genes APC,
KRAS,FBXW7,PIK3CA,BRAF,SMAD4, and ACVR2A was observed; only 16% of cases displayed the
acquisition of new mutations in metastatic lesions involving the TP53,CTNNB1,PTEN and SYNE1,
all the remaining cases sharing the genetic lesions of the primary tumor with metastases, for all types
of metastases, lymph node and distant metastases [41].
Isaque and coworkers have performed a comprehensive whole-genome analysis of dierences
between metastatic lesions and their corresponding primary tumors in 12 MSS CRC patients [
42
].
This detailed analysis showed that 65% (range from 36% to 92%) of all mutation events were shared
between primary tumors and corresponding metastases, suggesting the existence of a common truncal
clone; 15% (range from 1% to 29%) were tumor-specific and 19% (ranging from 3 to 42%) were
metastasis-specific; recurrent driver mutations were equally present in primary tumors and their
matched metastases, with the exception of only metastatic TP53 mutation, absent in the corresponding
primary tumor; a number of metastasis-specific mutations were identified, including non-silent
mutations of FAT1,FGF1,BRCA2,TP53, and KDR, splice site mutations of JAK2 and 3
0
-UTR mutations
in KDR,PDGFRA, and AKT2 genes [42].
Biomedicines 2020,8, 414 12 of 29
Several studies have explored copy number profiles of paired primary and metastatic CRC.
Kawamata et al. have analyzed CNAs in paired primary and metastatic tumor samples derived from
16 patients; the CNA profile was explored and was correlated with the timing of primary and metastatic
tissue resection and with the exposure to chemotherapy [
43
]. An average copy number dierence of
22% was observed when comparing primary and paired liver metastases; the dierences observed
between metastases and corresponding primary tumors increased when considering in this analysis
post-therapy metastases; some loss of heterozygosity (LOH) events were unique either to primary
tumor samples or to metastases: those unique to primary tumors occurred more frequently in those
treatment naive, while LOH events unique to metastases occurred most frequently post-therapy [
43
].
Interestingly, events of amplification of clinically actionable genes ERBB2,FGFR1,PIK3CA, or CDK8
were observed in some patients at the level of metastases but not in the corresponding primary
CRCs [43].
Smeets and coworkers investigated the pattern of CNAs in 409 metastatic CRC patients undergoing
treatment with chemotherapy alone or chemotherapy plus bevacizumab in the context of the phase II
MoMa study [
44
]. mCRCs were clustered into three dierent subgroups according to increasing degrees
of chromosomal instability: tumors belonging to the intermediate-to-high instability subgroups have
improved outcome following treatment with chemotherapy plus bevacizumab versus chemotherapy
alone; low instability tumors, including POLE-mutated and MSI tumors, derive no further benefit from
bevacizumab [44].
The targeted therapy of metastatic CRC patients implies the exploration of the targeted biomarker
and its presence in both primary and metastatic tumors. The introduction of EGFR inhibitors for
treatment of metastatic CRC patients allowed the unique opportunity to obtain, through the analysis of
numerous clinical studies, data on the concordance of the mutational status for KRAS,NRAS,BRAF and
PIK3CA between primary tumors and metastases in more than 3500 patients [
45
]. This metanalysis
involving 61 clinical studies and data on 3565 metastatic CRCs showed: (i) a very high median
biomarker concordance for KRAS (93%), NRAS (100%), BRAF (99.4%), PIK3CA (93%); (ii) a pooled
discordance of 8% for KRAS, 8% for BRAF, and 7% for PIK3CA [
35
]. These observations further support
the maintenance of the main driver mutations in CRCs undergoing metastatic spreading [45].
4. Tumor Heterogeneity and Metastatic Evolution
Study of intratumor heterogeneity (ITH) is fundamental from both a biological and clinical
perspective, to understand the genomic changes driving the evolution of the malignant process
up to metastasis generation. Several studies have shown that CRCs display a consistent degree of
spatial intratumor heterogeneity; particularly, three types of spatial heterogeneity of CRCs have been
described: (i) ITH related to the existence of genetic dierences at the level of tumor cells within the
primary tumor; (ii) ITH related to dierences at the level of various metastatic lesions within a single
patient; (iii) ITH related to the existence of genetic dierences within the cells of a single metastatic
lesion (intrametastatic heterogeneity) [46].
An initial study by Baisse and coworkers provided evidence through multiregional sequencing
analysis of 15–20 areas within a tumor, that 67% of advanced CRCs displayed significant ITH at
the level of gene alterations and CNAs [
47
]. Jeantet and coworkers performed the analysis of the
distribution of RAS mutations in dierent areas of primary tumor, metastatic lymph nodes and distant
metastases: primary tumors displayed an intra-tumoral heterogeneity for RAS mutations in 33% of
cases; the comparative analysis of primary tumors and metastatic tumors showed an inter-tumoral
heterogeneity in 36% of cases; multiple RAS mutated subclones were observed in 28% of cases in the
same tumor [48].
Kim and coworkers have performed a multiregion analysis of the mutational spectrum and
CNAs at the level of both primary and metastatic colorectal cancer lesions from five CRC patients [
49
].
This study showed a substantial level of ITH in both primary and matched liver metastases, with 46% to
80% subclonal mutation fractions. The spatial localization of the mutations allowed their classification
Biomedicines 2020,8, 414 13 of 29
into three types: the universal mutations are those observed in all the regional biopsies, are enriched in
genes such as APC,KRAS, and TP53 and represent events occurring early during tumor evolution;
metastasis-clonal mutations are those that are regionally clonal only in the metastatic regions and may
represent genetic events involved in the development of distant metastases; primary-private mutations
are those present in primary but absent at the level of metastases; metastasis-private mutations are
those present in primary but absent in only a part of metastatic lesions and may represent events
that are acquired during the expansion of metastatic clones [
49
]. It was estimated that 20–54% of
mutations in a given sample were universal, whereas from 46% to 80% of mutations were subclonal;
among the subclonal lesions, 1–15% were metastasis-clonal, 2–41% metastasis-private, and 14–56%
primary-private [
49
]. Most CNAs containing genes involved in CRC development, such as APC,
PTEN and SMAD4 were observed in both primary and metastatic lesions, thus representing early
or universal genomic events [
49
]. In contrast, copy number changes such as chromosomal gains of
c-MYC and chromotripsis can be region-specific and may represent the source of genetic intra-tumor
heterogeneity. Finally, the inferred evolution pattern of cancer progression was as a branched evolution,
rather than as linear evolution [49].
Sveen et al [
50
] have reported high-resolution DNA copy number analysis of metastatic lesions
from 45 CRC patients; this analysis showed a pronounced variation in the level of intra-patient
inter-metastatic heterogeneity [
50
]. Interestingly, the level of intra-patient inter-metastatic heterogeneity
resulted to be a strong prognostic determinant, stronger than commonly adopted clinico-pathological
prognostic markers: patients with a high-level of heterogeneity had a three-year overall survival of
18%, compared to 66% for patients with a low-level of tumor heterogeneity [50].
Uchi and coworkers have investigated intratumor heterogeneity in CRC by analyzing samples from
distinct areas of 9 dierent primary tumors [
51
]. Multiregional exome sequencing provided evidence
about the existence of extensive intratumor heterogeneity and branched evolution. Particularly,
the analysis of the various mutations showed that they can be classified as founder, shared and unique
mutations: parental clones acquire mutations in driver genes, such as APC,KRAS and FBWX7 as
founder mutations during tumor development, whereas subclones acquire mutations in PIK3CA
mutations as progressor mutations [
41
]. The age of patients correlated with the number of founder
mutations. Similar to gene mutations, some copy number alterations occurred as founder events
(such as amplifications of 7p, 13q, 10q, 20p, and 20q), while other CNAs, such as several focal deletions,
predominantly occur as progressor CNAs [
51
]. The analysis of epigenetic intratumoral heterogeneity
showed that CIMP-H occurs early in tumor evolution [
51
]. Similar to the other genetic alterations,
some epigenomic modifications occurred as founder events, such as hypermethylation of SFRPs,
GATA4 and GATA5 genes, whereas other epigenomic modifications occurred as progressor events [
51
].
An integrated view of the various parameters of intratumor genetic/epigenetic heterogeneity allowed
the reconstruction of each CRC’s life history. A typical example is given by one of these nine patients:
in this patient, the initial founder mutations, APC,KRAS and FBWX7 mutations, were observed at the
level of the parental clone; this initial parental clone subdivided into two subclones, one characterized
by the acquisition of a focal MYC amplification and the other one by several shared CNAs, such as
20p amplification and 1p deletion. At the subsequent steps of tumor evolution, the two subclones
branched into minor subclones, a process accompanied by accumulation of progressor mutations and
methylation alterations. These events caused the development of a consistent degree of intratumor
heterogeneity, extended also at the level of transcriptome heterogeneity [
51
]. Interestingly, these authors
have performed a comparative analysis of ITH in early and advanced CRCs, providing evidence that
early tumors acquire more subclonal driver mutations compared to advanced tumors: in early CRCs
50% of driver mutations were branch mutations, while only 22% mutations were branch mutations in
advanced colorectal cancers [52].
Some studies have explored ITH of CRCs using deep sequencing techniques. Thus, Wei et al.
performed a high depth multiregional wide exome sequencing in 28 tissues from four CRC patients
with matched primary and metastatic tumors. This study provided several interesting findings to
Biomedicines 2020,8, 414 14 of 29
better understand the process of CRC metastasization: metastatic tumors exhibited less intratumor
heterogeneity than primary tumors; primary and metastatic tumors dier significantly based on
the analysis of allelic frequency of the various mutations; all metastatic tumors inherited multiple
genetically distinct subclones from primary tumors, thus suggesting a possible polyclonal seeding
mechanism for metastasis [53]. In one of these patients, both lymph nodes and lung metastases were
analyzed, showing a completely dierent genetic landscape in these two dierent metastatic sites;
according to this finding, it was suggested that parallel metastatic dissemination to distant organs is
independent of lymph nodes [
53
]. Suzuki et al. have shown a variable level of ITH using deep-targeted
NGS followed by ultra-deep amplicon sequencing through the analysis of 4 dierent CRC patients
investigated at the level of various tumor regions; dierent tumor regions shared mutations in driver
genes, such as APC,KRAS and TP53. However, in addition, many mutations were observed only
at subclonal levels and in many instances their detection was only revealed by an ultra-high-depth
sequencing approach [54].
Very interestingly, Oh and coworkers performed a study of intratumor heterogeneity on a large
set of patients across 8 dierent tumor types by targeted deep sequencing; using this technique, a ITH
index was determined showing that CRCs are among the tumors with the highest ITH index [
55
].
In this study, CRC patients of all tumor stages were included showing that ITH index was already
high in 40% of stage I patients and moderately increased with tumor stage progression, with a high
ITH index in 55% of stage IV CRCs [
55
]. The presence of high ITH index was clearly associated with a
decreased progression-free survival (PFS) in stage I-III patients, but not in stage IV patients [55].
It is important to note that intratumor heterogeneity is not dictated only by genetic mechanisms,
but also by phenotypic heterogeneity/plasticity apparently unrelated to genetic determinants. A notable
example is provided by a study by Kreso et al., based on the analysis of serially expanded CRC clones
from patient samples, remaining genetically stable during serial transplantation; in spite this stability,
reproducible dierences in the functional fates and response to chemotherapy of individual CRC
cells, suggesting that
in vivo
dynamic changes of CRCs are not dictated by genomic changes [
56
].
These observations support the view that, in addition to the well-known mechanisms of tumor
heterogeneity driven by genetic diversity, other diversity-generating processes exist within a genetic
clone, seemingly related to epigenetic diversity, variability of tumor microenvironment and multiple
external factors aecting gene expression [56].
5. Liver Metastases
The liver is the most frequent metastatic site for CRC, with 60% of CRC patients developing
colorectal liver metastases (CLMs). CLMs can be surgically removed or therapeutically ablated and
these procedures may significantly improve the survival of these patients. Recent studies have explored
a possible link between genomic features and outcomes of metastatic CRC undergoing CLM resection.
Initial studies have suggested that mutations in KRAS and BRAF are associated with a poor outcome
after CLM resection, whereas mutations in NRAS,TP53,PIK3CA, and SMAD4 were shown to be
potential prognostic factors after CLM resection [57].
More recent studies have shown that the analysis of co-mutation status is more predictive of
outcome after CLM removal [
57
]. Thus, it was shown that RAS/TP53 double-mutant metastatic
CRC with predominant location in right colon of primary tumors, corresponding to 31% of patients,
displayed a shorter five-year overall survival (12%), compared with 55% overall survival of TP53
wild-type [57].
The presence of V600E BRAF mutations observed in 5.1% of metastatic CRC patients, but not
non-V600E BRAF mutations was associated with worse prognosis (reduced survival and frequent and
rapid recurrence) after resection of CLMs [
58
]. Interestingly, V600E BRAF mutations had a stronger
association with overall and disease-free survival than KRAS mutations [58].
Datta and coworkers explored a large group of 935 patients with metastatic CRC and showed
that co-alteration of oncogenic TP53 with either KRAS,NRAS, or BRAF mutations was associated with
Biomedicines 2020,8, 414 15 of 29
significantly worse survival compared to alterations in either gene group alone [
59
]. Interestingly,
RAS/BRAF-TP53 co-mutated CRCs were associated with worse survival in patients with liver and lung,
but not with peritoneal surface metastases Moreover, co-altered BRAF/RAS-TP53 were significantly
associated with the development of extra-hepatic metastatic sites [
59
]. Similar conclusions were
reached by Kawaguchi et al. who analyzed the possible relationship between somatic gene mutation
profile and outcome in 507 metastatic CRC patients who underwent CLM resection: BRAF,RAS,TP53,
and SMAD4 mutations were significantly associated with overall survival, coexisting mutations in
RAS,TP53, and SMAD4 were associated with negative outcome (reduced OS and RFS) than coexisting
mutations in any two of these genes and mutations in one or more of these genes [60].
Smith et al. recently reported the results of a retrospective study on 370 metastatic CRC patients
who underwent either colorectal liver hepatectomy followed by hepatic arterial infusion (HAI)
chemotherapy or HAI and systematic therapy (patients with unresectable metastases); 34.8% of these
patients have extrahepatic disease and 65.2% have liver-restricted disease [
60
]. Concurrently mutated
RAS/BRAF and SMAD4 were associated with negative survival in resectable patients, while concurrent
RAS/BRAF and TP53 mutations were associated with worse survival in unresectable patients [61].
Leung et al. have developed a highly multiplexed single-cell DNA sequencing to trace the
metastatic lineage of two CRC patients with matched liver metastases [
62
]. In the first patient,
a monoclonal seeding was observed, in which a single clone of tumor cells acquired a large number of
mutations before developing the capacity to migrate to the liver and to develop an advanced, metastatic
tumor; in the second patient, a polyclonal mechanism of seeding was observed, in which two clones
that have diverged from the primary tumor metastasize to the liver [
62
]. Interestingly, the single-cell
sequencing approach allowed to show the existence in one of the two patients of a rare subpopulation
of diploid cells that carried a heterozygous mutation in APC gene, but not associated with other
somatic mutations; these cells were diploid and seemingly represent the initial tumorigenic cells and
remained present in the advanced tumor representing 2.6% of tumor cells [
62
]. A second unexpected
finding was observed in the second patient and consisted in the detection of a small independent
subpopulation of diploid tumor cells that harbored a completely dierent set of mutations than the
main tumor lineage [62].
6. Lymphatic Metastases
Other studies have explored the process of CRC metastasization, focusing on the mechanisms of
spreading of cancer cells from the primary tumors to regional lymph nodes. Lymph node metastasis
associates with negative outcomes in CRCs and the presence of tumor cells in regional lymph nodes
defines stage III disease and the need for adjuvant chemotherapy and lowers the 5-year survival
compared to stage II disease without lymphatic lymph nodes metastasis [63].
Naxerova et al. have explored the evolutionary relationship between primary tumor, lymph node
and distant metastases in CRC: through the study of 213 biopsy samples from 17 patients, these authors
have used somatic variants in hypermutable DNA regions to reconstruct phylogenetic trees of tumor
metastatic evolution [
64
]. This analysis provided evidence about the existence of two dierent pathways
of lymph node and distant metastases generation in CRC patients. In fact, the genetic distances
between lymph node metastases, distant metastases and corresponding primary tumors were measured
showing that for the majority (73%) of lymph node metastases the distance with respect to the primary
tumor was shorter than the distance with respect to distant metastases; distant metastases (69% of
cases) had shorter distance to the primary tumor than to lymph nodes metastases [
64
]. In line with
these observations, reconstruction of phylogenetic evolutionary tumor trees allowed to establish that
in 35% of cases lymphatic and distant metastases have a common origin from the same subclone of the
primary tumor (either they originate both from the primary tumor or, alternatively, distant metastases
originate from lymph node metastases). In contrast, in 65% of cases, there is evidence of a distinct
origin of lymphatic and distant metastases, as supported by the evidence of genetically dierent
Biomedicines 2020,8, 414 16 of 29
alterations, thus indicating that in these patients primary tumors harboring multiple subclones at
dierent stages of evolution have seeded genetically distinct metastases [64].
Ulintz and coworkers have explored the clonal origin of lymph node metastasis in CRC. Thus,
they have investigated multiple tumor regions and cancer-containing lymph nodes from 7 CRC patients,
providing evidence that: (i) for each patient, the primary tumor regions and matched lymph node
metastases were polyclonal and the clonal populations diered from one node to another; (ii) in a part of
CRC patients, the cancer cells present in a given lymph node originated from multiple distinct regions
of a primary tumor, while in other cases these metastatic cells originate from a single geographic region
of the primary tumor; (iii) lymph node metastases contain subclones originated early or late during
tumor development [
65
]. According to these findings, a model of lymph node metastatic spreading in
CRCs involving multiple waves of seeding from the primary tumor over time was proposed [65].
Hu et al. have recently characterized the evolutionary dynamics of metastatic seeding by analyzing
exome sequencing profiles from 118 biopsies derived from 23 patients with CRC with metastases to
liver or brain [
66
]. Particularly, these authors performed multi-region sequencing on the primary
tumor and paired metastasis to build phylogenetic trees. The results of this study indicate that the
genomic divergence between the primary tumor and paired metastases is low: mutations in KRAS,
TP53,SMAD4,TCF7L2,FN1,ERLF3, and ATM were highly concordant between primary tumors
and metastases and 70% of highly frequent gene mutations were shared by both lesions, a finding
similarly observed in liver and brain metastases; among the genes that tended to be private to the
primary tumors or to metastases the most frequent were SYNE1 and APOB. Somatic copy number
alterations were generally concordant. Some putative oncogenes, such as PIK3CA,GNAS,SRC,
FXR1,MUCA,GPC6, and MECOM were more recurrently amplified in metastases than in primary
tumors [
56
]. Interestingly, the analysis of genetic data relative to large sets of CRC patients allowed
to define the existence of metastasis-associated early driver gene modules present in early tumors
and characterized by modules of tumor cells exhibiting CRC drivers (combinations of APC,KRAS,
TP53, or SMAD4) associated with potential metastasis-associated genes, such as TCF7L2,AMER1,
or PTPRT [
56
]. Interestingly, PTPRT mutations in combination with canonical CRC drivers are almost
exclusively found in metastatic CRC patients [
66
]. The simulation of spatial tumor growth under
selective or neutral growth evolutionary modes, coupled with the evaluation of the patterns of subclonal
divergence at the level of dierent tumor regions allowed to establish whether a given tumor is driven
by positive selective selection (either strong or weak) or by neutral evolution. The development
of a spatial computational model of tumor progression and statistical inference framework to time
dissemination in a patient-specific fashion, allowed to suggest that the capacity to seed metastasis is a
property inherent to cancer cells originated early during tumor development (81% of cases), when the
tumor bulk is clinically undetectable [
66
]. The analysis of a large set of public databases provided
evidence that the large majority (90%) of metastatic primary CRCs displayed subclonal selection,
thus suggesting that the metastatic clone possesses a consistent selective growth advantage. However,
only a lower proportion (33%) of stage I-III CRCs displayed patterns of tumor evolution compatible
with subclonal selection. Importantly, this observation suggests that type of tumor evolution may be
dependent on disease stage or disease aggressiveness [
66
]. As mentioned above, driver mutations
were usually not enriched in metastases; however, the stratification of CRC patients according to the
profile of tumor evolution (early dissemination vs. late dissemination) showed a higher frequency of
private driver mutations in metastases evolving under selection conditions compared to those evolving
neutrally, thus suggesting that in these patients additional subclonal driver mutations may occur
during the development of some metastases [66].
The same authors very recently reported the analysis of whole-exome sequencing data from 457
paired primary tumors and metastases derived from 136 patients with colorectal, breast and lung cancer:
this study involved the analysis of 39 metastatic CRC patients, including both untreated and treated
metastases [
67
]. The results of this study provided several interesting findings: (i) the mutational
burden (single nucleotide variation and CNAs) was highly concordant between primary and metastatic
Biomedicines 2020,8, 414 17 of 29
tumors; (ii) metastases displayed a slight increase in the number of clonal single nucleotide mutations
and fewer subclonal nucleotide variants, supporting the existence of an evolutionary bottleneck during
metastasis; (iii) a high percentage (84%) of clonal drivers in each primary CRC tumor and metastasis
was shared, while the fraction of subclonal drivers was 20%; (iv) among the three cancers investigated,
CRC had the highest prevalence of primary tumor-private subclonal drivers; (v) driver mutations
present in metastases are enriched in the trunk of the phylogenetic mutational tree; (vi) treatment
induced a dramatic increase of the frequency of private clonal drivers across all the three cancers,
including CRC (78% of metastasis-private clonal driver mutations), thus suggesting that therapy selects
a minor micro metastatic subclone; (vii) a small number of driver genes that were more frequently
amplified or deleted in metastases compared to primary tumors (such as amplification of RAC1 or
deletion of FAT1 and ALB genes); (viii) polyclonal seeding was common in untreated lymph node
metastases and distant metastases, but was less frequent in treated distant metastases [
67
]. The low
number of metastasis-private clonal mutations is consistent with early metastatic seeding [67].
7. Eect of Therapy on Mutational Landscape of Metastatic CRC
The targeted therapy of metastatic CRC patients implies the exploration of the targeted biomarker
and its presence in both the primary and the metastatic tumors. The introduction of EGFR inhibitors
for treatment of metastatic CRC patients allowed the unique opportunity to obtain, through the
analysis of numerous clinical studies, data on the concordance of the mutational status for KRAS,
NRAS,BRAF and PIK3CA between primary tumors and metastases in more than 3500 patients [
45
].
This meta-analysis involving 61 clinical studies and data on 3565 metastatic CRCs, showed: (i) a
median biomarker concordance for KRAS (93.7%), NRAS (100%), BRAF (99.4%), and PIK3CA (93%);
(ii) a pooled discordance of 8% for KRAS, 8% for BRAF, and 7% for PIK3CA [
45
]. These observations
further support the maintenance of the main driver gene alterations in CRCs undergoing metastatic
spreading [
45
]. The detection of KRAS mutations in metastatic CRC is important because implies a
negative prognosis and a poor response to standard chemotherapy [68].
An important example of the therapy-driven eects on the genomic alterations of metastatic CRC
derives from the analysis of patients developing resistance to therapies based on EGFR inhibitors.
EGFR inhibitors are eective in a subset of KRAS wild-type metastatic CRCs; however, after an initial
response, the development of secondary resistance mechanisms cause disease relapse, thus limiting
the clinical benefit of this treatment: The analyses of metastases of patients who developed resistance
to EGFR inhibitors showed more rarely the emergence of KRAS amplification and more frequently
the acquisition of secondary KRAS mutations; in these patients, KRAS mutant alleles were detectable
in the blood circulating tumor DNA 10 months before the radiographic documentation of disease
progression [
69
]. These observations suggest that EGFR-targeted therapy exerts a selective eect
on CRCs either inducing the expansion of pre-existing KRAS-mutant subclones or favoring the
development of new KRAS alterations [
69
]. Another mechanism of secondary resistance to EGFR
blockade is represented by novel alterations of ectodomain of EGFR [
70
]. The study of individual
patients has shown that dierent metastatic biopsies from the same patient with CRC display genetically
distinct mechanisms of resistance to EGFR blockade: thus, in some patients, it was documented that
distinct resistance mechanisms emerge in dierent metastases in the same patient and can drive
lesion-specific responses to dierent targeted therapies [70].
Genetic mechanisms of primary resistance to EGFR inhibitors among KRAS wild-type CRC patients
are represented by NRAS mutations,
V600E
BRAF mutations, MET amplification, ERBB2 amplification,
PIK3CA mutations at the level of exon 20, mutations in FGFR1, PDGFRA, and MAP2K1, and homozygous
deletions of PTEN [71].
Using xenografts derived from hepatic metastases of CRC patients, amplification of ERBB2
was identified as a potential therapeutic target in cetuximab-resistant CRCs [
72
]. These preclinical
observations supported a clinical study (HERACLES) evaluating trastuzumab and lapatinib in
metastatic CRC patients with amplified ERBB2 refractory to standard cares: in 33 patients,
Biomedicines 2020,8, 414 18 of 29
24.2% objective responses were observed with durable clinical benefit lasting >24 months in responding
patients [
72
]. Although ERBB2 blockade was eective, most of responding patients relapse [
73
]. A recent
study explored the mechanisms of tumor evolution responsible for relapse to HER2 blockade. In fact,
the analysis of circulating tumor DNA allowed to define organ and metastases-private evolutionary
patterns and high-levels in intra-patient molecular heterogeneity, defining lesion-specific evolutionary
trees and potential pharmacologic vulnerabilities [74].
8. Models of CRC Progression and Evolution
The study of tumor heterogeneity is a fundamental tool to analyze and to define the molecular
and cellular mechanisms responsible for the development of CRC and have provided a consistent
contribution to the development of current theories to explain CRC development.
Two dierent models have been proposed in the time to explain the origin and development of
CRC metastasis: one suggesting a common origin for both the primary tumor and metastases and
the other hypothesizing a completely independent genesis of metastases and of the primary tumor.
The sequencing data of matched primary tumors and metastases have strongly supported the existence
of a common ancestor of both the primary tumor and of the corresponding metastases.
The development of CRC from a common ancestor implies two dierent models to explain
metastasis evolution: the parallel progression model suggests that the dissemination of metastasizing
tumor cells occurs during early stages of primary tumor and the primary tumor and metastases evolve
separately thereafter. The linear progression model implies the occurrence of metastases as a sequential
event occurring during primary tumor development.
8.1. Somatic Mutations in Normal Colonic Epithelium
Colon epithelium is organized in crypts, composed by about 2000 cells, representing the tissutal
units. The main function of crypts consists in providing an ecient system of renewing of the
short-lived colonic epithelium, through the dierentiation of intestinal stem cells, located at the base of
the crypts; these stem cells stochastically replace one another through a biologic process of neutral drift,
thus ensuring that all stem cells and dierentiated cells present in a crypt derive from a single ancestral
stem cell. As a consequence of this hierarchical organization of the intestinal crypt, somatic mutations
in these ancestor stem cells are present in all the stem cells composing the crypt; these stem cells are
considered the cells of origin of CRCs [
75
]. A recent study explored somatic mutational landscape in
normal colorectal epithelium through whole-genome sequencing of normal colorectal crypts from 42
individuals [
76
]. Signatures of multiple mutational processes were detected, with some signatures
being ubiquitous, while other ones observed in some individuals, in some crypts. Driver mutations
were observed in about 1% of normal colorectal crypts in middle-aged individuals [
76
]. Among the
driver mutations detected in normal crypts there are AXIN2,STAG2,PIK3CA,ERBB2,ERBB3,FBXW7
mutations [
66
]. A dierent pattern of mutations was observed in normal crypts compared to those
observed in CRCs: ERBB2 and ERBB3 mutations are common in normal colon but rare in CRCs (1%),
whereas mutations in driver genes mutations in APC,KRAS and TP53 are common in CRCs, but are
rare among normal crypts (one in 14) [
66
]. These observations strongly suggest a major oncogenic
potential to APC,KRAS, and TP53 mutations promoting the conversion to colorectal adenoma (CRA)
and CRC, whereas mutations in ERBB2 and ERBB3 confer higher like hoods of crypt colonization by
stem cells [
76
]. No significant dierence was observed in the frequency of driver mutations between
individuals who had CRC and those who did not [
76
]. According to these findings, it was concluded
that CRAs and CRCs are rare outcomes of a pervasive process of neoplastic change occurring at the
level of morphologically normal colorectal epithelium [76].
The investigation of individuals with inflammatory bowel disease provided evidence that the
repeated inflammatory cycles aecting the colonic epithelium induce a 2.4-fold increase of the average
rate of colonic crypts aected; the mutations observed in IBD non-neoplastic epithelium mostly
involve ARID1A,FBXW7,PIGR, and ZC3H12A genes in the IL17 and Toll-like receptor pathways [
77
].
Biomedicines 2020,8, 414 19 of 29
Mutations in KRAS,APC and TP53 are rare in non-dysplastic tissues from IBD patients. At variance
with the normal colon, where clonal expansions are limited to the crypts, in IBD epithelium, frequent
widespread millimeter-scale clonal expansions were observed [
77
]. The dierences in driver landscape
of IBD colon, suggest that there are dierent selection mechanisms in the colitis-aected colon and that
somatic mutations potentially play a causal role in IBD pathogenesis [77].
Nicholson and colleagues have analyzed stem cell dynamics in normal human colon to define the
eciency of clone fixation within the epithelium and the rate of subsequent lateral expansion [
78
].
The process of mutant clone fixation within colonic crypts takes years, due to the time required for the
mutated intestinal stem cells to replace neighbors cells to populate the entire crypt; crypt fission allows
the lateral expansion of mutant clones: this process is rare for neutral mutations (0.7% per year); biases in
both fixation and expansion of stem cells increases age-related pro-oncogenic burden; pro-oncogenic
mutations modify the stem cell turnover and accelerate fixation and clonal expansion by crypt fission
to generate high mutant allelic frequencies with age [78].
8.2. Mutational Landscape in the Progression from Colorectal Adenomas to Colorectal Cancers
Several studies have compared the spectrum of genetic alterations in CRAs and in CRCs.
In an initial study, Jones et al. have performed an analysis of the mutations observed in benign,
invasive and metastatic colorectal tumors and reached the conclusion that more selective mutational
events are required for the transition of a benign adenoma into a CRC than those required for the
acquisition of metastasizing properties by a CRC [
32
]. The results of this study supported a classical
model of colorectal tumorigenesis, characterized by the progressive acquisition of mutational events
through various clinical stages of tumor progression: the tumor process is initiated by the acquisition
of a mutation into a gene of the Wnt pathway (mostly APC mutations) with consequent formation
of a small adenoma; mutations constitutively activating KRAS/BRAF pathway are required for the
proliferation of the small adenoma and for its transformation into a large adenoma; subsequent
acquisition of mutations at the level of genes controlling the PIK3CA, TGF-
β
and TP53 pathways is
required for the transformation of a benign adenoma into a CRC; only few metastasis-specific mutations
are acquired during the transition of a CRC from an invasive condition to a metastatic status [32].
APC loss of function is a key event in the colon carcinogenesis and represents the first event in
the tumor initiation. This conclusion was directly supported through sequencing studies on colon
adenomas. Nikolaev and coworkers have performed an exome sequencing analysis of 24 human colon
polyps, derived from 22 individuals with no family history of predisposition to cancer. The mutational
profiles observed at the level of the cancer-driver genes APC,CTNNB1 and BRAF genes allowed
to subdivide polyps into three dierent groups: the group 1 with APC mutations, included the
majority of polyps, mostly corresponding to colon adenomas: All the observed mutations introduced
premature stop codon and none of these polyps retained a normal APC allele, due to the presence
of two APC mutations or a single APC mutation associated with loss of heterozygosity; the group
2 with CTNNB1 mutations included only a few minority of polyps: the CTNNB1 mutation was
homozygous, due to concomitant LOH; the group 3 with BRAF mutations included polyps with
serrated histology: BRAF mutations were heterozygous [
79
]. Adenomas with CTNNB1 or BRAF
mutations did not display mutations in other cancer-driver genes, whereas adenomas with APC
mutations showed additional cancer-driver mutations (at the level of KRAS,NRAS,GNAS,AKT1,
SOX9 and TP53 genes), whose number correlated with the degree of dysplasia and invasiveness [79].
In addition to cancer-driver gene mutations, many passenger mutations were observed in colon
adenomas [
69
]. According to the rate of single nucleotide substitutions, it was suggested the existence
of a mutator phenotype in colon adenomas [79].
Lin and coworkers have reported the results of a whole-exome sequencing and targeted sequencing
study on 149 colon adenocarcinoma samples, corresponding to 134 conventional adenomas (CADs)
(104 non-advanced and 30 advanced) and 14 serrated adenomas (SSAs). No significant dierences in
the mutation rates were found between CNADs and SSAs (1.5 and 1.7 mutations/Mb, respectively) [
80
].
Biomedicines 2020,8, 414 20 of 29
As it is expected, the gene most frequently mutated in CNADs was APC, while BRAF was the gene
most recurrently mutated in SSAs [
70
]. In addition to APC, four genes were frequently mutated in
CADs: CTNNB1 (catenin beta 1), KRTAP4-5 (keratin-associated protein 4-5), GOLGA8B (golgin A8
family member B) and TMPRSS13 (transmembrane protease, serine 13) [
80
]. The biological role of
GOLGA8B,TMPRSS13 and KRTAP4-5 in the development of colon adenomas and in their progression
to CRC remains largely unknown. The comparison of the mutational profile observed in non-advanced
CADs, advanced CADs and CRCs showed that: PIK3CA and SMAD mutations are absent in CADs;
APC mutations are increasing from non-advanced to advanced CADs; KRAS and TP53 mutations
are progressively increasing in the progression from non-advanced to advanced CADs and then to
CRCs [
80
]. The identification of some CRC-specific mutated genes, absent in CADs, provides a tool for
distinguishing between adenomas and CRCs and supports the view that some mutational events are
essential for the transition from benign adenomas to CRCs [80].
Lee and coworkers reported the mutational profiling by whole-exome sequencing of 12 high-grade
colon adenomas (HGCAs, 11 non-hypermutated and 1 hypermutated). This analysis showed that total
numbers and spectrum of somatic mutations detected in HGCAs were not consistently dierent from
those observed in CRCs [
81
]. The most recurrent gene alterations observed in these tumors consisted
in mutations of APC,KRAS,SMAD4,ERBB4,AMER1, and TP53 genes, copy number loss of SMAD4,
and copy number gain of GNAS and ARID2 genes [
81
]. The peculiar finding of this study was related
to the observation that mono-allelic inactivation of SMAD4 may occur in HGCAs.
Druliner and coworkers have recently reported the analysis of cancer-adjacent polyps (CAPs) and
cancer-free polyps (CFPs): CAP cases included matched, distant normal colon epithelium, the polyp
(residual polyp of origin) and the corresponding cancer that arose from the polyp, whereas CFP cases
include matched, distant normal colonic epithelium and colon adenoma (polyp) [
82
]. The mutational
spectrum of CAPs and CFPs was explored by wide exome sequencing; the majority of the top 10
genes involved in CRC tumorigenesis had a mutational frequency higher in CAPs than in CFPs: TP53,
FBXW7,PIK3CA,KIAA1804,SMAD2, and SMAD4 were almost exclusively mutated in CAPs [
82
]. Thus,
the CAPs displayed an increased number of genetic variants as compared to the CFPs and the genes
preferentially or exclusively mutated in the CAPs were enriched for cancer pathways [
82
]. Some genes,
GREM1,IGF2,CTGF, and PLAU displayed significant changes between CFPs and CAPs [82].
In a recent study, Cross and coworkers have mapped the evolutionary landscape of CRAs and
CRCs through the study of multi-targeted whole genome sequencing on 2–16 regions from 9 CRAs and
15 CRCs [
83
]. The mutational frequency (single nucleotide alterations) was similar in CRAs and CRCs;
the burden of driver mutations was similar in CRAs and CRCs. Individual driver gene mutations
were detected at similar frequencies across CRAs and CRCs, with the exception of TP53, which was
more commonly mutated in CRCs than in CRAs [
83
]. Intra-tumor heterogeneity and phylogenetic
analyses suggest that CRCs occupy sharper fitness peaks that CRAs: 56% of CRA single nucleotide
alterations (SNAs) were subclonal, while only 45% of CRC SNAs were subclonal. The phylogenetic
trees of CRAs have shorter trunks and longer branches/leaves than those of CRCs; CRAs were more
heterogeneous than CRCs, suggesting that the former occupy a broader fitness peak than the latter
ones [
83
]. The analysis of non-synonymous mutations to synonymous mutations on the branches/leaves
of CRCs relative to their trunks, but not of CRAs, possibly suggesting a possible positive subclonal
selection in CRAs; these findings suggest that subclonal selection is absent/weak at the level of
established CRC [
83
]. The driver gene alterations can be subdivided into tier 1 mutations (mutations
or gene alterations playing a defined role in CRC pathogenesis) and tier 2 mutations (gene alterations
of uncertain pathogenic role or pan-cancer genes): tier 1 driver mutations were very frequently clonal
in both CRAs (80%) and CRCs (89%); tier 2 driver mutations were less frequently clonal in CRAs
(47%), compared to CRC (80%) [
83
]. The analysis of copy number alterations showed some remarkable
dierences between CRAs and CRCs: Adenomas had fewer CNAs than CRCs and the overall average
proportion of the genome disrupted by CNAs was lower in adenomas (40%) than in CRCs (72%) [
83
].
Driver CNAs in CRC involve losses of chromosomes 5q (APC), 17p (TP53) and 18q (SMAD4): 17p loss
Biomedicines 2020,8, 414 21 of 29
occurred more frequently in CRCs than in CRAs, whereas loss of 5q and 18q occurred at similar
frequencies in CRAs and CRCs [
83
]. 78% of CN gains were subclonal in CRAs compared to 48% in
CRCs; 57% and 27% of CN losses were subclonal in CRAs and CRCs, respectively [
83
]. The evolution
of CRC involves either a punctuated or a more gradual CNA acquisition [
83
]. Finally, the analysis
of few MSI
+
CRCs indicate that these tumors evolve in a similar way to MSS CRCs, with a higher
mutational burden and with a more limited evidence of subclonal selection [
83
]. The ensemble of
these observations suggests that CRAs can harbor mutations in any CRC driver gene and driver
acquisition does not necessarily involves selective sweeps, inducing stepwise evolution of the tumors,
as supported by the finding that subclones with additional driver mutations do not replace subclones
lacking these driver mutations, but co-exist in dierent areas of the tumors [84].
8.3. The Classical Linear Progression Model and the Big Bang Model
The classical linear progression model implies that a CRA is initiated by two genetic alterations at
the level of the APC gene and progresses to invasive CRC through a progressive, stepwise acquisition
of additional genetic alterations involving driver gene mutations such as KRAS and TP53 and deletion
of chromosome 18q4 [
74
]. The evolutionary dynamics of this process is governed by a series of
progressive selective sweeps to fixation, each involving the progressive development of subclones
exhibiting increasing fitness, due to the acquisition of new driver mutations [32].
Using early index-lesion sequencing and a mathematical model helping to translate the mutational
events into distance of time, it was estimated a shorter time required for the development of metastases
from advanced CRC (1.8 yeas) than for the development of an advanced CRC from a colon adenoma
(17 years) [32].
In 2015, Sottoriva and coworkers proposed the “Big Bang” model of human colorectal tumor
evolution, based on the assumption that these tumors are genetically heterogeneous from their
initiation and subsequent genetic alterations are changes of their original ancestral cancer-driving
alterations [
85
]. Several observations support the Big Band model: (i) Intratumor heterogeneity is a
“constitutive” property of CRCs arising from their initiation and increasing with their progressive
growth, not significantly influenced by events of clonal selection; this spontaneous propensity to
intra-tumor heterogeneity predisposes the CRCs to a branched phylogeny pattern of growth. (ii) Marked
clonal expansions or selective sweeps are rare events at the level of CRCs at an advanced stage of
tumor development. (iii) Both universal and private genetic alterations originate early during tumor
development a become widespread during tumor progression, thus becoming the dominating elements
in the genetic structure of developed CRCs. (iv) Aggressive subclones are present in the primary tumor
and remain rare; however, these subclones have a relative fitness advantage that contributes to fuel
resistance to drug treatments and may become dominant under these circumstances [85].
Several observations directly support this theory. In fact, Kang et al [
86
] have explored the
mutational heterogeneity of colorectal adenomas and reached the conclusion that these tumors display
the presence of private mutations in dierent parts of the same tumor. This consistent intratumor
heterogeneity originates from the first tumor divisions [
86
]. Sievers have investigated the mutational
landscape of small colorectal polyps and showed that these tumors carried 0–3 driver pathogenic
mutations, the most frequent being APC,KRAS,TP53,BRAF,FBXW7, and BRAF mutations [
87
].
About 31% of small polyps display two or more pathogenic mutations, with variable allelic frequencies,
a finding supporting the presence of multiple tumor cell populations [87].
The large majority of driver mutations are clonal and arise before the start of tumor expansion,
thus explaining the existence of only a minimal driver gene heterogeneity among untreated CRC
metastases [
88
]. The rarity of subclonal driver mutations supports the view that subclones may
dier by the selective presence of passenger mutations progressively accumulating during growth:
these subclones have similar fitness and occupy dierent tumor regions, thus generating ITH and their
size is mainly dependent on the timing of their generation during the process of tumor evolution [
89
].
Biomedicines 2020,8, 414 22 of 29
Additional evidence in favor of the “Big Bang” model tumor growth comes from additional recent
studies. Thus, Williams et al [
90
] have explored whether the subclonal mutant allele frequencies of
a part of cancers of dierent origin follow a model of tumor evolution based on simple power-law
distribution, as predicted by neutral growth. This analysis provided evidence that other cancers,
such as stomach, bladder, lung and cervical cancers, as well as CRCs, follow a model of tumor evolution
bye neutral growth [
90
]. In these malignancies, after an initial single tumor expansion, characterized
by the formation of multiple heterogeneous subclones that, in spite their genetic heterogeneity, initially
grow at comparable rates, without overtaking one another; thus, in these tumors, all clonal selection
events occur at a very early stage of tumor development and not in late-developing subclones,
thus resulting in the generation of numerous passenger mutations, involved in the generation of
intra-tumor heterogeneity [90].
It is commonly believed that passenger mutations have no role in cancer development. However,
many passenger mutations fall within protein-coding genes and, although individually weak,
these mutations tend to accumulate during tumor progression evading negative selection mechanisms,
and in their collective burden, alter the course of tumor progression [91,92].
Lineage tracing experiments in human colorectal adenomas further support the “Bing Bang”
theory of colony cancer development [
93
]. These experiments led to the identification of multipotential
stem cells within human colorectal adenomas, responsible for the development and maintenance of
these tumors. The study of methylation patterns of non-expressed genes, as well as the analysis of
genetic lesions in micro dissected individual crypts from colonic adenomas were used to characterize
clonal evolution of these tumors [
93
]. The analysis of individual crypts within each adenoma showed
that adenomatous crypts are clonal populations maintained by multipotential stem cells; individual
crypts from each adenoma display dierent methylation patterns; intratumor clones present in some
colonic adenomas are epigenetically homogeneous [
93
]. The results of this study were compatible
with a model of colorectal adenoma evolution not based on continual steady growth but on an initial
burst of tumor growth, followed by relative quiescence; the tumor clones form at the initial stages
of tumor development but not sweep through the tumor and are present as localized with divergent
intraclone methylation patterns. Rare subclones are generated later during tumor development,
exhibit homogeneous methylation patterns, and are localized at the level of focal regions of the
tumor [93].
Studies of the spatial distribution of genetic alterations within a tumor by phylogeography, an
approach that combines tumor phylogeny or the ancestral relationships of tumor subclones with their
spatial physical locations in the tumor, allows to visualize how tumors spread [
94
]. The spatial analysis
of private mutations in early CRCs, combining multiregional sequencing with mathematical multiscale
models showed the existence of spatial mutation patterns in these tumors, supporting the existence of
early colorectal tumor cell mobility, a tumor cell property required for generating ITH [95].
The analysis of epigenetic ITH into CRCs analyzing opposite tumor sides showed evidence of
little ITH or stepwise selection during tumor development, suggesting that the epigenome observed in
various tumor regions reflects that of its founder cells; despite epigenomic conservation, RNA expression
displayed significant variation between individual tumor regions, seemingly due to mechanisms of
continue adaptation related to phenotypic plasticity [96].
Saturation microdissection and targeted deep resequencing have shown that CRCs are jigsaw
arrayed in millimeter-wide columns sharing common phenotypes rather than being arranged
horizontally by phenotype [
97
]. Most of the large subclones thus identified shared both invasive
and superficial phenotype; subclones with invasive phenotypes arose from both early and late
phylogenetic branches [
97
]. This pattern of phylogeography is consistent with single tumor expansions
by founder cells possessing all the driver mutations required to sustain tumor growth rather than a
stepwise mechanism involving progressive invasions by a minority of subclones at various levels of
progression [
97
]. Particularly, on 11 CRCs analyzed in this study, two out of 11 displayed private driver
Biomedicines 2020,8, 414 23 of 29
mutations, while nine in 11 did not have private driver mutations, showing evidence of multiclonal
invasion, and invasive and metastatic subclones originate early during tumor development [97].
In conclusion, the analysis of the genetic heterogeneity observed at the level of CRCs is compatible
with a “Big Bang” expansion model, characterized by an early phase of tumor growth consisting in a
single cell expansion; this initial tumor expansion generates a large number of early-arising clones,
coexisting within the tumor for long periods of time for the absence of a selective pressure [
98
]. This weak
selection was insucient to determine large clonal expansions in short times. This finding supports
the view that the large part of tumor heterogeneity is generated early during tumor development, at a
stage where the tumor is still undetectable at clinical level [98].
9. Conclusions
About half of CRCs develop metastases and metastatic spreading is the main cause of CRC-related
death. The dynamics and the molecular processes remain largely unknown. Several recent studies
have shown that systemic spread can occur early in CRC development. Recent studies have reported
a detailed analysis of the genomic landscape of metastatic CRC patients underlying the molecular
heterogeneity of these patients and the possibility to identify some therapeutic targets in these patients.
The study of molecular evolution of CRCs suggest that these tumors may evolve either through a
process of subclonal selection or neutral evolution.
A better understanding of the cellular and molecular processes governing CRC metastasis
spreading will be necessary to improve the outcome of metastatic CRC patients.
Although the survival rate of patients with metastatic CRC patients improved in the last years,
the response to current treatments and prognosis of patients bearing KRAS,NRAS, and BRAF mutations
remain still poor. Therefore, there is an absolute need to identify these patients and to discover new
improvements for therapeutic vulnerabilities and to formulate rational prospective personalized
therapies aiming to improve their survival chances.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Dekker, E.; Tanis, P.J.; Valengels, J.; Kass, P.M.; Wallace, M.B. Colorectal cancer. Lancet
2019
,394, 1467–1480.
[CrossRef]
2.
Mullert, M.F.; Ibrahim, A.; Arends, M.J. Molecular pathological classification of colorectal cancer.
Wirchows Arch. 2016,469, 125–134. [CrossRef] [PubMed]
3.
Walther, A.; Houlston, R.; Toulinson, I. Association between chromosomal instability and prognosis in
colorectal cancer: A meta-analysis. Gut 2008,57, 941–950. [CrossRef] [PubMed]
4.
The Cancer Genome Atlas Network. Comprehensive molecular characterization of human colon and rectal
cancer. Nature 2012,487, 330–337. [CrossRef] [PubMed]
5.
Zhang, B.; Wang, J.; Wang, X.; Zhu, J.; Liu, Q.; Shi, Z.; Chambers, M.C.; Zimmerman, L.J.; Shaddox, K.F.
Proteogenomic characterization of human colon and rectal cancer. Nature 2014,513, 382–387. [CrossRef]
6.
Ried, T.; Meijer, G.A.; Harrison, D.J.; Grech, G.; Franch-Esposito, S.; Bria, R.; Carvalho, B.; Campos, J.
The landscape of genomic copy number alterations in colorectal cancer and their consequences on gene
expression levels and disease outcome. Mol. Asp. Med. 2019,69, 48–61. [CrossRef]
7.
Guinney, J.; Dienstmann, R.; Wang, X.; De Reynies, A.; Schilker, A.; Soneson, C.; Marisa, L.; Roepman, P.;
Nyamundanda, G.; Angelino, P.; et al. The consensus molecular subtypes of colorectal cancer. Nat. Med.
2015,21, 1350–1356. [CrossRef]
8.
Okita, A.; Takahashi, S.; Ouchi, K.; Inuoe, M.; Watanabe, M.; Endo, M.; Honda, H.; Yamada, Y.; Ishioka, C.
Consensus molecular subtypes classification of colorectal cancer as a predictive factor for chemotherapeutic
ecacy again st metastatic colorectal cancer. Oncotarget 2018,9, 18698–18711. [CrossRef]
Biomedicines 2020,8, 414 24 of 29
9.
Mooi, J.K.; Wirapati, P.; Asher, R.; Lee, C.K.; Savas, P.; Price, T.J.; Towsend, A.; Hardingham, J.; Buchanan, D.;
Williams, D.; et al. The prognostic impact of consensus molecular subtypes (CMS) and its predictive eects
for bevacizumab benefit in metastatic colorectal cancer: Molecular analysis of the AGITG MAX clinical trial.
Ann. Oncol. 2018,29, 2240–2246. [CrossRef]
10.
Stintzing, S.; Wirapati, P.; Lenz, H.J.; Neureiter, D.; Fisher von Weikersthal, L.; Decker, T.; Kiani, A.; Kaiser, F.;
Al-Batran, S.; Heingtes, T.; et al. Consensus molecular subgroups (CMS) of colorectal cancer (CRC) and
first-line ecacy of FOLFIRI plus cetuximab or bevacizumab in the FIRE3 (AIO KRK-0306) trial. Ann. Oncol.
2019,30, 1796–1803. [CrossRef]
11.
Becht, E.; de Reyni
è
s, A.; Giraldo, N.A.; Pilati, C.; Buttard, B.; Lacroix, L.; Selves, J.; Saut
è
s-Fridman, C.;
Laurent-Puig, P.; Fridman, W.H. Immune and stromal classification of colorectal cancer is associated with
molecular subtypes and relevant for precision immunotherapy. Clin. Cancer Res.
2016
,22, 4057–4066.
[CrossRef] [PubMed]
12.
Isella, C.; Brundu, F.; Bellomo, S.E.; Galimi, F.; Zanella, E.; Porporato, R.; Petti, C.; Fiori, A.; Orzan, F.;
Senetta, R.; et al. Selective analysis of cancer-cell intrinsic transcriptional traits defines novel clinically
relevant subtypes of colorectal cancer. Nat. Commun. 2017,8, 15107. [CrossRef] [PubMed]
13.
Tieng, F.; Baharudin, R.; Abu, N.; Yunos, R.I.; Lee, L.H.; Mutalib, N.S. Single cell transcriptomic in colorectal
cancer-current updates on its application in metastasis, chemoresistance and the roles of circulating tumor
cells. Fron. Pharmacol. 2020,11, 135. [CrossRef] [PubMed]
14.
Li, H.; Courtois, E.; Sengupta, D.; Tan, Y.; Chen, K.H.; Goh, J.; Kong, S.L.; Chua, C.; Hon, L.K.; Tan, W.S.;
et al. Reference component analysis of single-cell transcriptomes elucidates cellular heterogeneity in human
colorectal tumors. Nat. Genet. 2017,49, 708–718. [CrossRef] [PubMed]
15.
Uhlitz, F.; Bischo, P.; Sieber, A.; Obermayer, B.; Blanc, E.; Luthen, M.; Sawitzki, B.; Kamphues, C.; Beule, D.;
Sers, C.; et al. A census of cell types and paracrine interactions in colorectal cancer. bioRxiv
2020
. [CrossRef]
16.
Zhang, Y.; Zheng, L.; Zhang, L.; Hu, X.; Ren, X.; Zhang, Z. Deep single-cell RNA sequencing data of individual
T cells from treatment-naïve colorectal cancer patients. Sci. Data 2019,6, 131. [CrossRef]
17.
Zhang, L.; Li, Z.; Skrzypczynska, K.M.; Fang, Q.; Zhang, W.; O’Brien, S.; He, Y.; Wang, L.; Zhang, Q.; Kim, A.;
et al. Single-cell analyses inform mechanisms of myeloid-targeted therapies in colon cancer. Cell
2020
,
181, 442–459. [CrossRef]
18.
Kamal, Y.; Schmit, S.L.; Hoehn, H.J.; Amos, C.I.; Frost, H.R. Transcriptomic dierences between primary
colorectal adenocarcinomas and distant metastases reveal metastatic colorectal cancer subtypes. Cancer Res.
2019,79, 4227–4241. [CrossRef]
19.
Lieu, C.H.; Golemis, E.A.; Serebriiskii, I.G.; Newberg, J.; Hemmerich, A.; Connelly, C.; Messersmith, W.A.;
Eng, C.; Eckardt, S.G.; Frampton, G.; et al. Comprehensive genomic landscapes in early and later onset
colorectal cancer. Clin. Cancer Res. 2019,25, 5852–5858. [CrossRef]
20.
Dienstmann, R.; Salazar, R.; Tabernero, J. Molecular subtypes and the evolution of treatment decisions in
metastatic colorectal cancer. ASCO Educ. Book 2018,38, 231–238. [CrossRef]
21.
Xie, Y.H.; Chen, Y.X.; Fang, J.Y. Comprehensive review of targeted therapy for colorectal cancer. Signal
Transduct. Target. Ther. 2020,5, 22. [CrossRef] [PubMed]
22.
Zehir, A.; Benayed, R.; Shahm, R.K.; Syed, A.; Middha, S.; Kiom, H.R.; Srinivasan, P.; Gao, J.; Chakravarty, D.;
Sevlin, S.M.; et al. Mutational landscape of metastatic cancer revealed from prospective clinical sequencing
of 10,000 patients. Nat. Med. 2017,23, 703–713. [CrossRef] [PubMed]
23.
Priestley, P.; Baber, J.; Lolkema, M.P.; Steeghs, N.; de Brujn, E.; Shale, C.; Duyvesteyn, K.; Haidari, S.;
van Hoeck, A.; Onstenk, W.; et al. Pan-cancer whole-genome analyses of metastatic solid tumours. Nature
2019,575, 210–216. [CrossRef] [PubMed]
24.
Yaeger, R.; Chatila, W.; Lipsyc, M.; Hechtman, J.; Cercek, A.; Sanchez-Vega, F.; Jayakumaran, G.; Middha, S.;
Zehir, A.; Donoghue, M.; et al. Clinical sequencing defines the genomic landscape of metastatic colorectal
cancer. Cancer Cell 2018,33, 125–136. [CrossRef]
25.
Belardinilli, F.; Capalbo, C.; Malapelle, U.; Pisapia, P.; Raimondo, D.; Milanetti, E.; Yasaman, M.; Liccardi, C.;
Paci, P.; Sibilio, P.; et al. Clinical multigene panel sequencing identifies mutational association patterns in
metastatic colorectal cancer. Front. Oncol. 2020,10, 560. [CrossRef]
26.
Barras, D.; Missiagnia, E.; Wirapati, P.; Sieber, O.M.; Jorissen, R.N.; Love, C.; Molloy, P.L.; Jones, I.T.;
McLaughglin, S.; Gibbs, P.; et al. BRAF V600E mutant colorectal cancer subtypes based on gene expression.
Clin. Cancer Res. 2017,23, 104–115. [CrossRef]
Biomedicines 2020,8, 414 25 of 29
27.
Jones, J.C.; Renfro, L.A.; Al-Shamsi, O.; Achrock, A.B.; Rankin, A.; Zhang, B.Y.; Kasi, P.M.; Voss, J.S.; Leal, A.D.;
Sun, J.; et al.
Non-V600
BRAF mutations define a clinically distinct molecular subtype of metastatic colorectal
cancer. J. Clin. Oncol. 2017,35, 2624–2630. [CrossRef]
28.
Yao, Z.; Yaeger, R.; Rodrik-Outmezguine, V.S.; Tao, A.; Torres, N.M.; Chang, M.T.; Drosten, M.; Zhao, H.;
Cecchi, F.; Hembrough, T.; et al. Tumours with class 3 BRAF mutants are sensitive to the inhibition of
activated RAS. Nature 2017,548, 234–238. [CrossRef]
29.
Cremolini, C.; Di Bartolomeo, M.; Amatu, A.; Antoniotti, C.; Moretto, R.; Berenato, R.; Perrone, F.;
Tamborini, E.; Aprile, G.; Lonerdi, S.; et al. BRAF codons 594 and 596 mutations identify a new molecular
subtype of metastatic colorecxtal cancer at favorable prognosis. Ann. Oncol.
2015
,26, 2092–2097. [CrossRef]
30.
Schrippa, M.; Biason, P.; Lonardi, S.; Pella, N.; Pina, M.S.; Urbano, F.; Antoniotti, C.; Cremolini, C.; Corallo, S.;
Pietrantonio, F.; et al. Class 1, 2, and 3 BRAF-mutated metastatic colorectal cancer: A detailed clinical,
pathologic, and molecular characterization. Clin. Cancer Res. 2019,27, 3954–3961. [CrossRef]
31.
Riihimaki, M.; Hemminki, A.; Sundquist, J.; Hemminki, K. Patterns of metastasis in colon and rectal cancer.
Sci. Rep. 2016,6, 29765. [CrossRef] [PubMed]
32.
Jones, B.; Chen, W.; Parmigiani, G.; Diehl, F.; Beerenwinkel, N.; Antal, T.; Traulsen, A.; Nowak, M.A.;
Siegel, C.; Velculescu, A.; et al. Comparative lesion sequencing provides insights into tumor evolution.
Proc. Natl. Acad. Sci. USA 2008,105, 4283–4288. [CrossRef]
33.
Vakiani, E.; Janakiraman, M.; Shen, R.; Sinha, R.; Zeng, Z.; Shia, J.; Cercek, A.; Kemeny, N.; D’Angelica, M.;
Viale, A. Comparative genomic analysis of primary versus metastatic colorectal carcinomas. J. Clin. Oncol.
2012,30, 2956–2962. [CrossRef] [PubMed]
34.
Brannon, A.R.; Vakiani, E.; Sylvester, B.E.; Scott, S.N.; McDermott, G.; Shah, R.H.; Kania, K.; Viale, A.;
Oschwald, D.M.; Vacic, V.; et al. Comparative sequencing analysis reveals high genomic concordance
between matched primary and metastatic colorectal cancer lesions. Genome Biol.
2014
,13, 454. [CrossRef]
[PubMed]
35.
Vermaat, J.S.; Nijman, I.J.; Koudjis, M.J.; Gerritse, F.L.; Scherer, F.J.; Mokry, M.; Roessingh, W.M.; Lansu, N.;
de Brujin, E.; van Hillegersberg, R.; et al. Primary colorectal cancers and their subsequent hepatic metastases
are genetically dierent: Implications for selection of patients for targeted treatment. Clin. Cancer Res.
2012
,
18, 688–699. [CrossRef] [PubMed]
36.
Lim, B.; Mun, J.; Kim, J.H.; Kim, C.W.; Roh, S.A.; Cho, D.H.; Kim, Y.S.; Kim, S.Y.; Kim, J.C. Genome-wide
mutation profiles of colorectal tumors and associated liver metastases at the exome and transcriptome levels.
Oncotarget 2015,26, 22179–22190. [CrossRef] [PubMed]
37.
Vignot, S.; Lefebvre, C.; Frampton, G.M.; Stephens, P.J.; Soria, J.C.; Spano, J.P. Comparative analysis of
primary tumour and matched metastases in colorectal cancer patients: Evaluation of concordance between
genomic and transcriptomic profiles. Eur. J. Cancer 2015,51, 791–799. [CrossRef] [PubMed]
38.
Tan, I.B.; Malik, S.; Ramnarayanan, K.; McPherson, J.R.; Ho, D.L.; Suzuki, Y.; Ng, S.B.; Yan, S.; Lim, K.H.;
Koh, D.; et al. High-depth sequencing of over 750 genes supports linear progression of primary tumors
and metastases in most patients with liver-limited metastatic colorectal cancer. Genome Biol.
2019
,16, 32.
[CrossRef]
39.
Kim, K.P.; Kim, J.E.; Hong, Y.S.; Ahn, S.M.; Chun, S.M.; Hong, S.M.; Jang, S.J.; Yu, C.S.; Kim, J.C.; Kim, T.W.
Paired primary and metastatic tumor analysis of somatic mutations in synchronous and metachronous
colorectal cancer. Cancer Res. Treat. 2017,49, 161–167. [CrossRef]
40.
Fujiyoshi, K.; Yamamoto, G.; Takahashi, A.; Arai, Y.; Yamada, M.; Kakuta, M.; Yamaguchi, K.; Akagi, Y.;
Nishimura, Y.; Sakamoto, H.; et al. High concordance rate of KRAS/BRAF mutations and MSI-H between
primary and corresponding metastases. Oncol. Rep. 2017,37, 785–792. [CrossRef]
41.
Jesinghaus, M.; Wolf, T.; Pfarr, N.; Muckenhuber, A.; Ahadova, A.; Warth, A.; Goeppert, B.; Sers, C.; Kloor, M.;
Endris, V.; et al. Distinctive spatiotemporal stability of somatic mutations in metastasized microsatellite-stable
colorectal cancer. Am. J. Surg. Pathol. 2015,39, 1140–1147. [CrossRef] [PubMed]
42.
Isaque, N.; Abba, M.L.; Huaser, C.; Patil, N.; Paramasivan, N.; Huebschmann, D.; Leupold, J.H.;
Balusabramanian, G.P.; Kleinheinz, K.; Poprak, U.H.; et al. Whole genome sequencing put forward
hypotheses on metastasis evaluation and therapy in colorectal cancer. Nat. Commun.
2018
,9, 4782. [CrossRef]
[PubMed]
Biomedicines 2020,8, 414 26 of 29
43.
Kawamata, F.; Patch, A.M.; Nones, K.; Bond, C.; McKeone, K.; Pearson, S.A.; Homma, S.; Liu, C.; Fennell, L.;
Dumenil, T.; et al. Copy number profiles of paired primary and metastatic colorectal cancers. Oncotarget
2018,9, 3394–3405. [CrossRef]
44.
Smeets, D.; Miller, I.S.; O’Connor, D.P.; Das, S.; Moran, B.; Boeckx, B.; Gaiser, T.; Betge, J.; Barat, A.; Klinger, R.;
et al. Copy number load predicts outcome of metastatic colorectal cancer patients receiving bevacizumab
combination therapy. Nat. Commun. 2008,9, 4112. [CrossRef] [PubMed]
45.
Bhullar, D.S.; Barriuso, J.; Mullamitha, S.; Saunders, M.P.; O’Dwyer, S.T.; Aziz, O. Biomarker concordance
between primary colorectal cancer and its metastases. EBioMedicine 2019,40, 363–374. [CrossRef]
46.
Molinari, C.; Marisi, G.; Passardi, A.; Matteucci, L.; De Maio, G.; Ulivi, P. Heterogeneity in colorectal cancer:
A challenge for personalized medicine? Int. J. Med. Sci. 2018,19, 3733. [CrossRef]
47.
Baisse, B.; Bouzorene, H.; Saraga, E.P.; Bosman, F.T.; Benhattar, J. Intratumor genetic heterogeneity in
advanced human colorectal adenocarcinoma. Int. J. Cancer 2001,93, 346–352. [CrossRef]
48.
Jeantet, M.; Tougeron, D.; Tachon, G.; Cortes, U.; Archambaut, C.; Fromont, G.; Karayan-Tapon, L. High intra-
and inter-tumoral heterogeneity of RAS mutations in colorectal cancer. Int. J. Mol.
2016
,17, 2015. [CrossRef]
49. Kim, T.M.; Jung, S.H.; An, C.H.; Lee, S.H.; Baek, I.P.; Kim, M.S.; Park, S.W.; Rhee, J.K.; Lee, S.H.; Chung, Y.J.
Subclonal genomic architectures of primary and metastatic colorectal cancer based on intratumoral genetic
heterogeneity. Clin. Cancer Res. 2015,21, 4461–4472. [CrossRef]
50.
Sveen, A.; Loes, I.M.; Alagaratnam, S.; Nilsen, G.; Høland, M.; Lingjærde, O.C.; Sorbye, H.; Berg, K.C.;
Horn, A.; Angelsen, J.H.; et al. Intra-patient inter-metastatic genetic heterogeneity in colorectal cancer as a
key determinant of survival after curative liver resection. PLoS Genet. 2016,12, e1006225. [CrossRef]
51.
Uchi, R.; Takahashi, Y.; Niida, A.; Shimamura, T.; Hirata, H.; Sugimachi, K.; Sawada, G.; Iwaya, T.;
Kurashige, J.; Shinden, Y.; et al. Integrated multiregional analysis proposing a new model of colorectal cancer
evolution. PLoS Genet. 2016,12, e1005778. [CrossRef] [PubMed]
52.
Saito, T.; Niida, A.; Uchi, R.; Hirata, H.; Komatsu, H.; Sakimura, S.; Hayashi, S.; Nambara, S.; Kurada, Y.
A temporal shift of the evolutionary principle shaping intratumor heterogeneity in colorectal cancer.
Nat. Commun. 2018,9, 2884. [CrossRef] [PubMed]
53.
Wei, Q.; Ye, Z.; Zhong, X.; Li, L.; Wang, C.; Myers, R.E.; Palazzo, J.P.; Fortuna, D.; Yan, A.; Waldman, S.A.;
et al. Multiregion whole-exome sequencing of matched primary and metastatic tumors revealed genomic
heterogeneity and suggested polyclonal seeding in colorectal cancer metastasis. Ann. Oncol.
2017
,
28, 2135–2141. [CrossRef] [PubMed]
54.
Suzuki, Y.; Ng, S.B.; Chua, C.; Leow, W.Q.; Chng, J.; Liu, S.Y.; Ramnarayanan, K.; Gan, A.; Ho, D.L. Multiregion
ultra-deep sequencing reveals early intermixing and variable levels of intratumoral heterogeneity in colorectal
cancer. Mol. Oncol. 2017,11, 124–139. [CrossRef] [PubMed]
55.
Oh, B.; Shin, H.T.; Yun, J.W.; Kim, K.T.; Kim, J.; Bae, J.S.; Cho, Y.B.; Lee, W.Y.; Park, Y.A.; Im, Y.H.; et al.
Intratumor heterogeneity inferred from targeted deep sequencing as a prognostic indicator. Scient. Rep.
2019
,
9, 4542. [CrossRef] [PubMed]
56.
Kreso, A.; O’Brien, C.A.; van Galen, P.; Gan, O.I.; Notta, F.; Brown, A.; Ng, K.; Ma, J.; Wienholds, E.;
Dumont, C.; et al. Variable clonal repopulation dynamics influence chemotherapy response in colorectal
cancer. Science 2013,339, 543–548. [CrossRef]
57.
Chun, Y.S.; Passot, G.; Yamashita, S.; Nusrat, M.; Katsonis, P.; Loree, J.M.; Conrad, C.; Tzeng, C.W.; Xiao, L.;
Aloia, T.A.; et al. Deleterious eect of RAS and evolutionary high-risk TP53 double mutation in colorectal
liver metastases. Ann. Surg. 2019,269, 917–923. [CrossRef]
58.
Margonis, G.; Buettner, S.; Andreatos, N.; Kim, Y.; Wagner, D.; Sasaki, K.; Beer, A.; Schwartz, C.; Loes, I.M.;
Smolle, M.; et al. Association of BRAF mutations with survival and recurrence in surgically treated patients
with metastatic colorectal liver cancer. JAMA Surg. 2018,153, e180996. [CrossRef]
59.
Datta, J.; Smith, J.J.; Cjhatila, W.K.; McAulie, J.C.; Kandoth, C.; Valdani, E.; Frankel, T.L.; Ganesh, K.;
Wasserman, I.; Lipsyc-Sharf, M.; et al. Co-altered Ras/B-raf and TP53 is associated with extremes of
survivorship and distinct patterns of metastasis in patients with metastatic colorectal cancer. Clin. Cancer Res.
2020,26, 1077–1085. [CrossRef]
60.
Kawaguchi, Y.; Kopetz, S.; Newhook, T.E.; De Bellis, M.; Chey, J.N.; Chun, Y.S.; Tzeng, C.W.; Aloia, T.A.;
Vautthey, J.N. Mutation status of RAS, TP53, and SMAD4 is superior to mutation status of RAS alone for
predicting prognosis after resection of colorectal liver metastases. Clin. Cancer Res.
2019
,25, 5843–5851.
[CrossRef]
Biomedicines 2020,8, 414 27 of 29
61.
Smith, J.J.; Chatila, W.K.; Sanchez-Vega, F.; Datta, J.; Connell, L.C.; Szeglin, B.C.; Basunia, A.; Boucher, T.M.;
Hauser, H.; Wasserman, I.; et al. Genomic stratification beyond Ras/B-Raf in colorectal liver metastasis
patients treated with hepatic arterial infusion. Cancer Med. 2019,8, 6538–6548. [CrossRef] [PubMed]
62.
Leung, M.L.; Davis, A.; Gao, R.; Casasent, A.; Wang, Y.; Sei, E.; Vilar, E.; Maru, D.; Kopetz, S.; Navin, N.E.; et al.
Single-cell DNA sequencing reveals a late-dissemination model in metastatic colorectal cancer. Genome Res.
2017,27, 1287–1299. [CrossRef] [PubMed]
63.
Chang, G.J.; Rodrigiuez-Bigas, M.A.; Skibber, J.M.; Mayer, V.A. Lymph nodes evaluation and survival after
curative resection of colon cancer: Systematic review. J. Natl. Cancer Inst. 2007,99, 433–441. [CrossRef]
64.
Naxerova, K.; Reiter, J.G.; Bratchel, E.; Lennerz, J.; Van de Wetering, M.; Rowan, A.; Cai, T.; Clevers, H.;
Swanton, C.; Nowak, M.A.; et al. Origins of lymphatic and distant metastases in human colorectal cancer.
Science 2017,357, 55–60. [CrossRef]
65.
Ulintz, P.J.; Greenson, J.K.; Wu, R.; Fearon, E.R.; Hardiman, K.M. Lymph node metastases in colon cancer are
polyclonal. Clin. Cancer Res. 2017,24, 2214–2224. [CrossRef] [PubMed]
66.
Hu, Z.; Ding, J.; Ma, Z.; Sun, R.; Soane, J.A.; Shaer, J.S.; Suarez, C.J.; Bergho, A.S.; Cremolini, C.; Falcone, A.;
et al. Quantitative evidence for early metastatic seeding in colorectal cancer. Nat. Genet.
2019
,51, 1113–1122.
[CrossRef] [PubMed]
67.
Hu, Z.; Li, Z.; Ma, Z.; Curtis, C. Multi-cancer analysis of clonality and the timing of systemic spread in paired
primary tumors and metastases. Nat. Genet 2020,52, 701–708. [CrossRef] [PubMed]
68.
Garcia-Carbonero, N.; Martinez-Useros, J.; Li, W.; Orta, A.; Perez, N.; Carames, C.; Hernandez, T.; Moreno, I.;
Serrano, G.; Garcia-Foncillas, J. KRAS and BRAF mutations as prognostic and predictive biomarkers for
standard chemotherapy response in metastatic colorectal cancer: A single institutional study. Cells
2020
,
9, 219. [CrossRef]
69.
Misala, S.; Yaeger, R.; Hober, S.; Scala, E.; Janckraman, M.; Liska, D.; Valtorta, E.; Schiavo, R.; Buscarino, M.;
Siravegna, G.; et al. Emergence of KRAS mutations and acquired resistance to anti EGFR therapy in colorectal
cancer. Nature 2012,486, 532–536. [CrossRef]
70.
Bertotti, A.; Papp, E.; Jones, S.; Adle, V.; Anagnostou, V.; Lupo, B.; Sausen, M.; Phallen, J.; Hrubau, C.A.;
Tokheim, C.; et al. The genomic landscape of response to EGFR blockade in colorectal cancer. Nature
2015
,
526, 263–267. [CrossRef]
71.
Russo, M.; Siravegna, G.; Blazkowsky, L.S.; Corti, G.; Crisafulli, G.; Ahronian, L.G.; Mussolin, B.; Kwak, E.L.;
Buscorino, M.; Lazzari, L.; et al. Tumor heterogeneity and lesion-specific response to targeted therapy in
colorectal cancer. Cancer Discov. 2016,6, 147–153. [CrossRef] [PubMed]
72.
Bertotti, A.; Milgiardi, G.; Galimi, F.; Sassi, F.; Torti, D.; Isella, C.; Cor
à
, D.; Di Nicolantonio, F.; Buscarino, M.;
Petti, F.; et al. A molecularly annotated platform of patient-derived xenografts (“xenopatients”) identified
HER2 as an eective therapeutic target in cetuximab-resistant colorectal cancer. Cancer Discov.
2011
,
1, 508–523. [CrossRef]
73.
Sartore-Bianchi, A.; Trusolino, L.; Martino, C.; Bencardino, K.; Lonardi, S.; Bergamo, F.; Zagonel, V.; Leone, F.;
Depetris, I.; Martinelli, E. Dual-targeted therapy with trastuzumab and lapatinib in tretament-refractory,
KRas codon 12/13 wild-type, HER2-positive metastatic colorectal cancer (HERACLES): A proof-of-concept,
multicentre, open-label phase 2 trial. Lancet Oncol. 2016,17, 738–746. [CrossRef]
74.
Siravegna, G.; Lazzari, L.; Crisafulli, G.; Sartore-Bianchi, A.; Mussolin, B.; Cossingena, A.; Martiono, C.;
Lanman, R.B.; Nagy, R.J.; Fairalough, S.; et al. Radiologic and genomic evolution of individual metastases
during HER2 blockade in colorectal cancer. Cancer 2018,34, 148–162. [CrossRef]
75.
Barker, N.; Ridgway, R.A.; van Es, J.H.; van de Wetering, M.; Begthel, H.; van den Born, M.; Danenberg, E.;
Clarke, A.R.; Sansom, O.J.; Clevers, H. Crypt stem cells as the cells-of-origin of intestinal cancer. Nature
2009
,
457, 608–611. [CrossRef] [PubMed]
76.
Lee-Six, H.; Olafasson, S.; Ellis, P.; Osborne, R.J.; Sanders, M.A.; Moore, L.; Georgakopoulos, N.; Torrente, F.;
Noorani, A.; Goddard, M.; et al. The landscape of somatic mutation in normal colorectal epithelial cells.
Nature 2019,574, 532–537. [CrossRef]
77.
Olafsson, S.; McIntyre, R.E.; Coorens, T.; Butler, T.; Jung, H.; Robinson, P.S.; Lee-Six, H.; Sanders, M.A.;
Arestang, K.; Dawson, C.; et al. Somatic evolution in non-neoplastic IBD-aected colon. Cell
2020
,182, 1–13.
[CrossRef]
Biomedicines 2020,8, 414 28 of 29
78.
Nicholson, A.M.; Olpe, C.; Hoyle, A.; Thorsen, A.S.; Rus, T.; Colomb
é
, M.; Brunton-Sim, R.; Kemp, R.;
Marks, K.; Quirk, P.; et al. Fixation and spread of somatic mutations in adult human colonic epithelium.
Cell Stem Cell 2018,22, 909–918. [CrossRef]
79.
Nikolaev, S.I.; Sotiriou, S.K.; Pateras, J.S.; Santoni, F.; Sougioultzis, S.; Edgren, H.; Almusa, H.; Robyr, D.;
Guipponi, M.; Saarela, J.; et al. A single-nucleotide substitution mutator phenotype revealed by exome
sequencing of human colon adenomas. Cancer Res. 2012,72, 6279–6289. [CrossRef]
80.
Lin, S.H.; Raju, G.; Hu, C.; Ye, Y.; Gu, J.; Chen, J.S.; Hildebrandt, M.; Liang, H.; Menter, D.G.; Morris, J.
The somatic mutation landscape of premalignant colorectal adenoma. Gut 2017, in press. [CrossRef]
81.
Lee, S.H.; Jung, S.H.; Kim, T.M.; Rhee, J.K.; Park, H.C.; Kim, M.S.; Chang, H.A.; Lee, S.H.; Chung, H.J.
Whole-exome sequencing identified mutational profiles of high-grade colon adenomas. Oncotarget
2017
,
8, 6579–6588. [CrossRef]
82.
Druliner, B.R.; Wang, P.; Bae, T.; Baheti, S.; Slettedahl, S.; Mahoney, D.; Vasmastzis, N.; Xu, H.; Kim, M.;
Bockol, M.; et al. Molecular characterization of colorectal adenomas with and without malignancy reveals
distinguishing genome, transcriptome and methylome alterations. Scient. Rep.
2018
,8, 3161. [CrossRef]
[PubMed]
83.
Cross, W.; Kanc, M.; Mustonen, V.; Temko, D.; Davis, H.; Baker, A.M.; Biswas, S.; Arnold, R.; Chegwiddent, L.;
Gatenbeee, C.; et al. The evolutionary landscape of colorectal cancer tumorigenesis. Nat. Ecol. Evol.
2018
,
2, 1661–1672. [CrossRef] [PubMed]
84. Fearon, E.R.; Volgstein, B. A genetic model for colorectal tumorigenesis. Cell 1990,61, 759–767. [CrossRef]
85.
Sottoriva, A.; Kang, H.; Ma, Z.; Graham, T.A.; Salomon, M.P.; Zhao, J.; Marjoram, P.; Siegmund, K.; Press, M.F.;
Shibata, D.; et al. A Big Bang model of human colorectal tumor growth. Nat. Genet.
2015
,47, 209–216.
[CrossRef]
86.
Kang, H.; Salomon, M.P.; Sottoriva, A.; Zhao, J.; Toy, M.; Press, M.F.; Curtis, C.; Marjoram, P.; Siegmund, K.;
Shibata, D. Many private mutations originate from the first few divisions of a human colorectal adenoma.
J. Pathol. 2015,237, 355–362. [CrossRef]
87.
Sievers, C.K.; Zou, L.; Pickhardt, P.J.; Matkowskyj, K.A.; Albrecht, D.M.; Clipson, L.; Bacher, J.W.; Pooler, B.D.;
Moawad, F.J.; Cash, B.D.; et al. Subclonal diversity arises early even in small colorectal tumours and
contributes to dierential growth fates. Gut 2017,66, 2131–2140. [CrossRef]
88.
Reiter, J.G.; Makahon-Moore, A.P.; Gerold, J.M.; Heyde, A.; Attieh, M.A.; Kohutek, Z.A.; Tokheim, C.J.;
Brown, A.; De Blasio, R.M.; Niyazov, J.; et al. Minimal functional driver gene heterogeneity among untreated
metastases. Science 2018,361, 1033–1037. [CrossRef]
89.
Williams, M.J.; Werner, B.; Heide, T.; Curtis, C.; Barnes, C.; Sottoriva, A.; Graham, T.A. Quantification of
subclonal selection in cancer from bulk sequencing data. Nat. Genet. 2018,50, 895–903. [CrossRef]
90.
Williams, M.J.; Werner, B.; Barnes, C.; Graham, T.A.; Sottoriva, A. Identification of neutral evolution across
cancer types. Nat. Genet. 2016,48, 238–244. [CrossRef]
91.
McFarland, C.D.; Korolev, K.; Kryukov, G.; Sunyaev, S.R.; Mirny, L.A. Impact of deleterious passenger
mutations on cancer progression. Proc. Natl. Acad. Sci. USA 2013,110, 2910–2913. [CrossRef] [PubMed]
92.
Mc Farland, C.D.; Mirny, L.A.; Korolev, K.S. Tug-of-war between driver and passenger mutations in cancer
and other adaptive processes. Proc. Natl. Acad. Sci. USA 2014,111, 15138–15143. [CrossRef]
93.
Humphries, A.; Cereser, B.; Gay, L.J.; Miller, D.S.; Das, B.; Gutteridge, A.; Elia, G.; Nye, E.; Jeery, R.;
Poulsom, R.; et al. Lineage tracing reveals multipotent stem cells maintain human adenomas and the pattern
of clonal expansion in tumor evolution. Proc. Natl. Acad. Sci. USA
2013
,110, E2490–E2499. [CrossRef]
[PubMed]
94.
Shibata, D. Visualizing human colorectal cancer intratumor heterogeneity with phylogeography. iScience
2020,23, 101304. [CrossRef]
95.
Ryser, M.D.; Min, B.H.; Siegmund, K.D.; Shibata, D. Spatial mutation patterns as markers of early colorectal
tumor cell mobility. Proc. Natl. Acad. Sci. USA 2018,115, 5774–5779. [CrossRef] [PubMed]
96.
Ryser, M.D.; Yu, M.; Grady, W.; Siegmund, K.; Shibata, D. Epigenetic heterogeneity in human colorectal
tumors reveals preferential conservation and evidence of immune surveillance. Sci. Rep.
2018
,8, 17292.
[CrossRef]
Biomedicines 2020,8, 414 29 of 29
97.
Ryser, M.D.; Mallo, D.; Hall, A.; Hardman, T.; King, L.M.; Tatischev, S.; Sorribes, K.; Maley, C.C.; Marks, J.R.;
Hwang, E.S.; et al. Normal barriers to invasion driving human colorectal tumor growth. Nat. Commun.
2020
,
11, 1280. [CrossRef] [PubMed]
98.
Sottoriva, A.; Barnes, C.P.; Graham, T.A. Catch my drift? Making sense of genomic intra-tumor heterogeneity.
Biochim. Biophys. Acta 2017,1867, 95–100.
Publisher’s Note:
MDPI stays neutral with regard to jurisdictional claims in published maps and institutional
aliations.
©
2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
... First, in most cases, the presence of the target, or of eventual resistance-related mutations, is assessed in the primary tumor and not in the metastases, which actually represent the neoplastic diseases to be treated. Therefore, targeted therapy's inefficacy may be due to a dissimilarity in the genetic abnormalities between the primary CRC and the matched metastases, as reported in several studies [18]. The discordance between the primary tumor and the metastases may be due to a subclonal evolution during metastasization, or to the genetic heterogeneity of the primary tumor [18]. ...
... Therefore, targeted therapy's inefficacy may be due to a dissimilarity in the genetic abnormalities between the primary CRC and the matched metastases, as reported in several studies [18]. The discordance between the primary tumor and the metastases may be due to a subclonal evolution during metastasization, or to the genetic heterogeneity of the primary tumor [18]. A study analyzing matched samples showed that the genetic alterations in lymph node metastases reflect those found in the invasive front rather than in the main tumor mass of primary CRC, suggesting that the assessment of molecular targets should be preferentially carried out in this part of the primary tumor [19]. ...
... A study analyzing matched samples showed that the genetic alterations in lymph node metastases reflect those found in the invasive front rather than in the main tumor mass of primary CRC, suggesting that the assessment of molecular targets should be preferentially carried out in this part of the primary tumor [19]. However, discordant alterations may also be present among the different metastases [18]. ...
Article
Full-text available
According to the Global Cancer Statistics 2020, colorectal cancer (CRC) represents the third most frequent malignancy worldwide, and is the second in terms of mortality [...]
... Previous studies have addressed the mutational landscape in primary and metastatic CRCs and found that genomes of metastases are essentially not different from those of primary tumors. [8,9]. Only a few studies investigated gene expression profiles in primary tumors and metastases utilizing distinct comparative models to identify prognostic metastasis signatures and biomarkers [8][9][10][11]. ...
... [8,9]. Only a few studies investigated gene expression profiles in primary tumors and metastases utilizing distinct comparative models to identify prognostic metastasis signatures and biomarkers [8][9][10][11]. In these analyses, transcriptomic differences in cellular programs, such as downmodulation of epithelial-mesenchymal transition (EMT) and differential expression of a few single genes, were described [9][10][11]. However, most of these investigations were performed on samples of primary tumors and metastases; as both were obtained prior to any chemotherapeutic therapy, they thus do not reflect the massive influence of cytotoxic agents on the expression levels of numerous genes in metastases. ...
... Only a few studies investigated gene expression profiles in primary tumors and metastases utilizing distinct comparative models to identify prognostic metastasis signatures and biomarkers [8][9][10][11]. In these analyses, transcriptomic differences in cellular programs, such as downmodulation of epithelial-mesenchymal transition (EMT) and differential expression of a few single genes, were described [9][10][11]. However, most of these investigations were performed on samples of primary tumors and metastases; as both were obtained prior to any chemotherapeutic therapy, they thus do not reflect the massive influence of cytotoxic agents on the expression levels of numerous genes in metastases. ...
Article
Full-text available
Most metastatic colorectal cancer (mCRC) patients succumb to refractory disease due to secondary chemotherapy resistance. To elucidate the molecular changes associated with secondary resistance, we recruited 64 patients with mCRC and hepatic metastases before standard first-line chemotherapy between 2014 and 2018. We subjected DNA from primary tumor specimens (P), hepatic metastasis specimens after treatment (M), and liquid biopsies (L) taken prior to (pre), during (intra), and after (post) treatment to next generation sequencing. We performed Nanostring expression analysis in P and M specimens. Comparative bioinformatics and statistical analysis revealed typical mutational patterns with frequent alterations in TP53, APC, and KRAS in P specimens (n = 48). P and pre-L (n = 42), as well as matched P and M (n = 30), displayed a similar mutation spectrum. In contrast, gene expression profiles classified P (n = 31) and M (n = 23), distinguishable by up-regulation of immune/cytokine receptor and autophagy programs. Switching of consensus molecular subtypes from P to M occurred in 58.3% of cases. M signature genes SFRP2 and SPP1 associated with inferior survival, as validated in an independent cohort. Molecular changes during first-line treatment were detectable by expression profiling rather than by mutational tumor and liquid biopsy analyses. SFRP2 and SPP1 may serve as biomarkers and/or actionable targets.
... Accumulating lines of evidence indicate that genetic factors may play a crucial role in CRC metastasis (5). However, CRC metastases are mechanistically heterogeneous, and the heterogeneity may answer for the poor prognosis in clinics. ...
... In addition, it was reported that 65% of somatic mutations originated from a common progenitor, in which 15% were tumor-specific and 19% were metastasis-specific (15). Alternatively, some studies paid more attention to the metastasis-specific alterations (5,16). A previous study suggested that targeted therapy of colorectal liver metastases would be more effective on the basis of the genetic properties of metastasis rather than those of the primary tumor since there was a significant genetic difference (17). ...
Article
Full-text available
Metastasis is the main fatal cause of colorectal cancer (CRC). Although enormous efforts have been made to date to identify biomarkers associated with metastasis, there is still a huge gap to translate these efforts into effective clinical applications due to the poor consistency of biomarkers in dealing with the genetic heterogeneity of CRCs. In this study, a small cohort of eight CRC patients was recruited, from whom we collected cancer, paracancer, and normal tissues simultaneously and performed whole-exome sequencing. Given the exomes, a novel statistical parameter LIP was introduced to quantitatively measure the local invasion power for every somatic and germline mutation, whereby we affirmed that the innate germline mutations instead of somatic mutations might serve as the major driving force in promoting local invasion. Furthermore, via bioinformatic analyses of big data derived from the public zone, we identified ten potential driver variants that likely urged the local invasion of tumor cells into nearby tissue. Of them, six corresponding genes were new to CRC metastasis. In addition, a metastasis resister variant was also identified. Based on these eleven variants, we constructed a logistic regression model for rapid risk assessment of early metastasis, which was also deployed as an online server, AmetaRisk (http://www.bio-add.org/AmetaRisk). In summary, we made a valuable attempt in this study to exome-wide explore the genetic driving force to local invasion, which provides new insights into the mechanistic understanding of metastasis. Furthermore, the risk assessment model can assist in prioritizing therapeutic regimens in clinics and discovering new drug targets, and thus substantially increase the survival rate of CRC patients.
... Некоторые классификации РТК основываются на транскрипционных профилях опухолей. К наиболее характерным молекулярным особенностям РТК относят нарушения сигнальных путей MAPK и WNT, хромосомные дисбалансы в локусах 1p, 5q, 17p, 18p, 18q, 20p и 22q, мутации в онкогенах KRAS, NRAS или BRAF, активацию ферментов PI3-киназ и повреждения гена TP53 [4][5][6][7]. Согласно данным недавних исследований, на молекулярный патогенез РТК также влияет микробиом кишечника [8,9]. ...
Article
Full-text available
Molecular genetic diagnostics is an essential element to plan for management of colorectal cancer (CRC) patients. The choice of systemic treatment for CRC is impossible without molecular testing of the tumor. For instance, the assessment of the KRAS and NRAS genes is mandatory for consideration of anti-EGFR agents. Tumors with BRAF V600E mutation are characterized by aggressive behavior, the necessity of intensive cytostatic regimens, as well as by sensitivity to combination therapy with BRAF and EGFR inhibitors. Inactivation of the DNA mismatch repair, the MUTYH gene or DNA polymerase epsilon (POLE) leads to an excessive tumor mutational burden; these CRC types are highly immunogenic and therefore respond to immune checkpoint inhibitors. Some colorectal carcinomas are characterized by overexpression of the HER2 oncogene, which make them sensitive to corresponding target therapies. There are CRCs with clinical signs of hereditary predisposition, which require germline genetic testing. Nowadays the molecular diagnosis of CRC is being seriously modified due to worldwide implementation of the next-generation sequencing (NGS) and hypersensitive variants of polymerase chain reaction, for example, droplet digital polymerase chain reaction (ddPCR). Non-invasive liquid biopsy is an example of another highly useful innovation that has growing importance for CRC screening, control of surgical intervention efficacy and monitoring of the disease course.
... Comparative analysis of primary CRC and corresponding metastatic sites revealed maintenance of the main driver mutations in both liver and lung metastases, some of which are approved for CRC therapy, such as RAS, BRAF and MSI [17][18][19]. In contrast, genomic [20][21][22], transcriptomic [23] and proteomic [24] profiling identified molecular differences between primary tumor, liver and lung metastases that might have potential therapeutic implications for specific metastatic sites. ...
Article
Full-text available
Considering the biology of CRC, distant metastases might support the identification of high-risk patients for early recurrence and targeted therapy. Expression of a panel of druggable, metastasis-related biomarkers was immunohistochemically analyzed in 53 liver (LM) and 15 lung metastases (LuM) and correlated with survival. Differential expression between LM and LuM was observed for the growth factor receptors IGF1R (LuM 92.3% vs. LM 75.8%, p = 0.013), EGFR (LuM 68% vs. LM 41.5%, p = 0.004), the cell adhesion molecules CD44v6 (LuM 55.7% vs. LM 34.9%, p = 0.019) and α2β1 (LuM 88.3% vs. LM 58.5%, p = 0.001) and the check point molecule PD-L1 (LuM 6.1% vs. LM 3.3%, p = 0.005). Contrary, expression of HGFR, Hsp90, Muc1, Her2/neu, ERα and PR was comparable in LuM and LM. In the LM cohort (n = 52), a high CD44v6 expression was identified as an independent factor of poor prognosis (PFS: HR 2.37, 95% CI 1.18–4.78, p = 0.016). High co-expression of CD44v6/α2β1 (HR 4.14, 95% CI 1.65–10.38, p = 0.002) and CD44v6/PD-L1 (HR 2.88, 95% CI 1.21–6.85, p = 0.017) indicated early recurrence after hepatectomy, in a substantial number of patients (CD44v6/α2β1: 11 (21.15%) patients; CD44v6/PD-L1: 12 (23.1%) patients). Dual expression of druggable protein biomarkers may refine prognostic prediction and stratify high-risk patients for new therapeutic concepts, depending on the metastatic location.
Preprint
Full-text available
Cancer is one of the deadliest diseases developed through tumorigenesis and could be fatal if reached at metastatic phase. The novality of present investigation is to explore the prognostic biomarkers in hepatocellular carcinoma (HCC) that could develop glioblastoma multiforme (GBM) due to metastasis. The analysis was done using RNA-Seq datasets for both HCC (PRJNA494560 and PRJNA347513) and GBM (PRJNA494560 and PRJNA414787) from Gene Expression Omnibus (GEO). This study identified 13 hub genes found to be overexpressed in both GBM and HCC. Promoter methylation study showed these genes to be hypomethylated. Validation through genetic alteration and missense mutations resulted in chromosomal instability leading to improper chromosome segregation causing aneuploidy. A 13-gene predictive model was obtained and validated using KM plot. These hub genes could be prognostic biomarkers and potential therapeutic targets, inhibition of which could suppress tumorigenesis and metastasis.
Article
Full-text available
Colorectal cancer is one of the most prevalent cancer types. Although there have been breakthroughs in its treatments, a better understanding of the molecular mechanisms and genetic involvement in colorectal cancer will have a substantial role in producing novel and targeted treatments with better safety profiles. In this review, the main molecular pathways and driver genes that are responsible for initiating and propagating the cascade of signaling molecules reaching carcinoma and the aggressive metastatic stages of colorectal cancer were presented. Protein kinases involved in colorectal cancer, as much as other cancers, have seen much focus and committed efforts due to their crucial role in subsidizing, inhibiting, or changing the disease course. Moreover, notable improvements in colorectal cancer treatments with in silico studies and the enhanced selectivity on specific macromolecular targets were discussed. Besides, the selective multi-target agents have been made easier by employing in silico methods in molecular de novo synthesis or target identification and drug repurposing.
Article
This study was designed and conducted to clarify the impact of RNF128 expression on malignant biological behaviors of colorectal cancer (CRC) cells and the underlying mechanism. The expression of RNF128 in CRC tissues was analyzed using mRNA sequencing data of TCGA database and was validated by Western blot assay. The experimental studies on biological functions of RNF128 in vitro were conducted to assess its impact on the proliferation, apoptosis, and metastasis of CRC cells. Furthermore, tumor xenograft models in nude mice were established to investigate the relationship between RNF128 expression and tumor growth in vivo. The expression levels of both RNF128 mRNA and protein were significantly increased in CRC tissues (p < .001). The knockdown of RNF128 markedly suppressed the malignant phenotype of HCT116 and SW480 cells in vitro, including cell growth, antiapoptosis, migration, and invasion (p < .001). On the other hand, knockdown of RNF128 exerted a remarkable effect on the growth inhibition of tumor xenografts in vivo (p < .001). Further investigation revealed that RNF128 knockdown lead to a significant decrease in the expression of p‐AKT and p‐PI3K protein. More importantly, the proliferative, antiapoptotic, metastatic abilities of RNF128‐knockdown cells were markedly increased by 740 Y‐P treatment (p < .001). These findings further suggested that PI3K/AKT signaling pathway played a key role in RNF128‐mediated aggressive phenotype of CRC cells. RNF128 functions as a tumor promoter in the pathogenesis of CRC via regulating PI3K/AKT pathway, and it could be a valuable target for CRC treatment.
Article
Molecular genetic diagnostics is an essential element to plan for management of colorectal cancer (CRC) patients. The choice of systemic treatment for CRC is impossible without molecular testing of the tumor. For instance, the assessment of the KRAS and NRAS genes is mandatory for consideration of anti-EGFR agents. Tumors with BRAF V600E mutation are characterized by aggressive behavior, the necessity of intensive cytostatic regimens, as well as by sensitivity to combination therapy with BRAF and EGFR inhibitors. Inactivation of the DNA mismatch repair, the MUTYH gene or DNA polymerase epsilon (POLE) leads to an excessive tumor mutational burden; these CRC types are highly immunogenic and therefore respond to immune checkpoint inhibitors. Some colorectal carcinomas are characterized by overexpression of the HER2 oncogene, which make them sensitive to corresponding target therapies. There are CRCs with clinical signs of hereditary predisposition, which require germline genetic testing. Nowadays the molecular diagnosis of CRC is being seriously modified due to worldwide implementation of the next-generation sequencing (NGS) and hypersensitive variants of polymerase chain reaction, for example, droplet digital polymerase chain reaction (ddPCR). Non-invasive liquid biopsy is an example of another highly useful innovation that has growing importance for CRC screening, control of surgical intervention efficacy and monitoring of the disease course.
Article
Molecular genetic analysis is an integral part of colorectal cancer (CRC) management. The choice of systemic therapy for CRC is largely based on the results of tumor molecular testing. Evaluation of the KRAS and NRAS gene status is mandatory for consideration of anti-epidermal growth factor receptor (EGFR) therapy. Tumors with the BRAF V600E substitution are characterized by aggressive behaviour, may require intensified cytotoxic regimens and benefit from combined BRAF and EGFR inhibition. The inactivation of DNA mismatch repair (MMR), or MUTYH gene, or DNA polymerase epsilon results in excessive tumor mutational burden; these CRCs are highly antigenic and therefore sensitive to immune checkpoint inhibitors. Some CRCs are characterized by overexpression of the HER2 oncogene and respond to the appropriate targeted therapy. There are CRCs with clinical signs of hereditary predisposition to this disease, which require germline genetic testing. Liquid biopsy is an emerging technology that has the potential to assist CRC screening, control the efficacy of surgical intervention and guide disease monitoring. The landscape of CRC molecular diagnosis is currently undergoing profound changes due to the increasing use of next generation sequencing.
Article
Full-text available
Inflammatory bowel disease (IBD) is a chronic inflammatory disease associated with increased risk of gastrointestinal cancers. We whole-genome sequenced 446 colonic crypts from 46 IBD patients and compared these to 412 crypts from 41 non-IBD controls from our previous publication on the mutation landscape of the normal colon. The average mutation rate of affected colonic epithelial cells is 2.4-fold that of healthy colon, and this increase is mostly driven by acceleration of mutational processes ubiquitously observed in normal colon. In contrast to the normal colon, where clonal expansions outside the confines of the crypt are rare, we observed widespread millimeter-scale clonal expansions. We discovered non-synonymous mutations in ARID1A, FBXW7, PIGR, ZC3H12A, and genes in the interleukin 17 and Toll-like receptor pathways, under positive selection in IBD. These results suggest distinct selection mechanisms in the colitis-affected colon and that somatic mutations potentially play a causal role in IBD pathogenesis.
Article
Full-text available
Phylogeography combines ancestry with location and can be translated to intratumor heterogeneity (ITH) to visualize how tumors spread. ITH is common in human tumors, with many genetic and phenotypic differences between regions. The roles of ITH in progression are uncertain because many subclones lack discernable driver mutations. ITH can be visualized by mapping mutations onto microscopic sections, where subclones are directly associated with phenotypes, especially the deeper areas with the more invasive cells that confer worst clinical outcomes. Instead of a stepwise hierarchy where subclones segregate by phenotype with later branching subclones in more invasive areas, multiple subclones share superficial and invasive phenotype and are jigsaw arrayed in vertical columns. Phylogeography shows that both early and late subclones extend from the surface to the invasive front, suggesting that founder cells start with phenotypic plasticity and essentially all the drivers necessary to rapidly grow into large invasive tumors.
Article
Full-text available
Metastasis is the primary cause of cancer-related deaths, but the natural history, clonal evolution and impact of treatment are poorly understood. We analyzed whole-exome sequencing (WES) data from 457 paired primary tumor and metastatic samples from 136 patients with breast, colorectal and lung cancer, including untreated (n = 99) and treated (n = 100) metastases. Treated metastases often harbored private ‘driver’ mutations, whereas untreated metastases did not, suggesting that treatment promotes clonal evolution. Polyclonal seeding was common in untreated lymph node metastases (n = 17 out of 29, 59%) and distant metastases (n = 20 out of 70, 29%), but less frequent in treated distant metastases (n = 9 out of 94, 10%). The low number of metastasis-private clonal mutations is consistent with early metastatic seeding, which we estimated occurred 2–4 years before diagnosis across these cancers. Furthermore, these data suggest that the natural course of metastasis is selectively relaxed relative to early tumorigenesis and that metastasis-private mutations are not drivers of cancer spread but instead associated with drug resistance. Analysis of whole-exome sequencing data from paired primary tumors and metastases from patients with breast, colorectal and lung cancer identifies clonal remodeling associated with therapy and few metastasis-private clonal mutations, consistent with early metastatic seeding.
Article
Full-text available
Extensive molecular characterization of human colorectal cancer (CRC) via Next Generation Sequencing (NGS) indicated that genetic or epigenetic dysregulation of a relevant, but limited, number of molecular pathways typically occurs in this tumor. The molecular picture of the disease is significantly complicated by the frequent occurrence of individually rare genetic aberrations, which expand tumor heterogeneity. Inter- and intratumor molecular heterogeneity is very likely responsible for the remarkable individual variability in the response to conventional and target-driven first-line therapies, in metastatic CRC (mCRC) patients, whose median overall survival remains unsatisfactory. Implementation of an extensive molecular characterization of mCRC in the clinical routine does not yet appear feasible on a large scale, while multigene panel sequencing of most commonly mutated oncogene/oncosuppressor hotspots is more easily achievable. Here, we report that clinical multigene panel sequencing performed for anti-EGFR therapy predictive purposes in 639 formalin-fixed paraffin-embedded (FFPE) mCRC specimens revealed previously unknown pairwise mutation associations and a high proportion of cases carrying actionable gene mutations. Most importantly, a simple principal component analysis directed the delineation of a new molecular stratification of mCRC patients in eight groups characterized by non-random, specific mutational association patterns (MAPs), aggregating samples with similar biology. These data were validated on a The Cancer Genome Atlas (TCGA) CRC dataset. The proposed stratification may provide great opportunities to direct more informed therapeutic decisions in the majority of mCRC cases.
Article
Full-text available
Colorectal cancer (CRC) is among the most lethal and prevalent malignancies in the world and was responsible for nearly 881,000 cancer-related deaths in 2018. Surgery and chemotherapy have long been the first choices for cancer patients. However, the prognosis of CRC has never been satisfying, especially for patients with metastatic lesions. Targeted therapy is a new optional approach that has successfully prolonged overall survival for CRC patients. Following successes with the anti-EGFR (epidermal growth factor receptor) agent cetuximab and the anti-angiogenesis agent bevacizumab, new agents blocking different critical pathways as well as immune checkpoints are emerging at an unprecedented rate. Guidelines worldwide are currently updating the recommended targeted drugs on the basis of the increasing number of high-quality clinical trials. This review provides an overview of existing CRC-targeted agents and their underlying mechanisms, as well as a discussion of their limitations and future trends.
Article
Full-text available
Intra-tumoral heterogeneity (ITH) could represent clonal evolution where subclones with greater fitness confer more malignant phenotypes and invasion constitutes an evolutionary bottleneck. Alternatively, ITH could represent branching evolution with invasion of multiple subclones. The two models respectively predict a hierarchy of subclones arranged by phenotype, or multiple subclones with shared phenotypes. We delineate these modes of invasion by merging ancestral, topographic, and phenotypic information from 12 human colorectal tumors (11 carcinomas, 1 adenoma) obtained through saturation microdissection of 325 small tumor regions. The majority of subclones (29/46, 60%) share superficial and invasive phenotypes. Of 11 carcinomas, 9 show evidence of multiclonal invasion, and invasive and metastatic subclones arise early along the ancestral trees. Early multiclonal invasion in the majority of these tumors indicates the expansion of co-evolving subclones with similar malignant potential in absence of late bottlenecks and suggests that barriers to invasion are minimal during colorectal cancer growth.
Article
Full-text available
Colorectal cancer (CRC) is among the most common cancer worldwide, a challenge for research, and a model for studying the molecular mechanisms involved in its development. Previously, bulk transcriptomics analyses were utilized to classify CRC based on its distinct molecular and clinicopathological features for prognosis and diagnosis of patients. The introduction of single-cell transcriptomics completely turned the table by enabling the examination of the expression levels of individual cancer cell within a single tumor. In this review, we highlighted the importance of these single-cell transcriptomics analyses as well as suggesting circulating tumor cells (CTCs) as the main focus of single-cell RNA sequencing. Characterization of these cells might reveal the intratumoral heterogeneity present in CRC while providing critical insights into cancer metastasis. To summarize, we believed the analysis of gene expression patterns of CTC from CRC at single-cell resolution holds the potential to provide key information for identification of prognostic and diagnostic markers as well as the development of precise and personalized cancer treatment.
Article
Full-text available
KRAS mutation is a confirmed predictive biomarker for anti-EGFR monoclonal antibody therapy response for metastatic colorectal cancer. However, its prognosis impact and the predictive potential for first-line standard chemotherapy remains unclear. On the other hand, V600E mutation is the most frequent and studied mutation in the BRAF gene, and it has been associated with a poor outcome of patients and a low response to anti-EGFR treatment. Thus, the aim of this study is to evaluate the role of KRAS and BRAF mutations as prognosis factors and predictive biomarkers for 1st line standard chemotherapy in metastatic colorectal cancer. KRAS mutations and BRAF V600E mutations exhibited a poor outcome (p = 0.021 and p < 0.0001, respectively). Cox multivariate analysis showed that the presence of liver metastasis (HR = 1.595; 95% CI: 1.086–2.343; p = 0.017), KRAS mutation (HR = 1.643; 95% CI: 1.110–2.431; p = 0.013) and BRAF V600E mutation (HR = 5.861; 95% CI: 2.531–13.570; p < 0.0001) were statistically significant co-variables for progression-free survival. Interestingly, patients with KRAS mutations were associated with a poor response to first line standard chemotherapy (p = 0.008). In contrast, the BRAF V600E mutation did not have any impact on the first line standard chemotherapy response (p = 0.540). Therefore, in the present study, we provide new insight on the role of KRAS and BRAF, not only as prognosis biomarkers, but also as first line standard chemotherapy response biomarkers in metastatic colorectal cancer.
Article
Undergraduate researchers are the next-generation scientists. Here, we call for more attention from our community to the proper training of undergraduates in biomedical research laboratories. By dissecting common pitfalls, we suggest how to better mentor undergraduates and prepare them for flourishing careers.
Article
Single-cell RNA sequencing (scRNA-seq) is a powerful tool for defining cellular diversity in tumors, but its application toward dissecting mechanisms underlying immune-modulating therapies is scarce. We performed scRNA-seq analyses on immune and stromal populations from colorectal cancer patients, identifying specific macrophage and conventional dendritic cell (cDC) subsets as key mediators of cellular cross-talk in the tumor microenvironment. Defining comparable myeloid populations in mouse tumors enabled characterization of their response to myeloid-targeted immunotherapy. Treatment with anti-CSF1R preferentially depleted macrophages with an inflammatory signature but spared macrophage populations that in mouse and human expresses pro-angiogenic/tumorigenic genes. Treatment with a CD40 agonist antibody preferentially activated a cDC population and increased Bhlhe40⁺ Th1-like cells and CD8⁺ memory T cells. Our comprehensive analysis of key myeloid subsets in human and mouse identifies critical cellular interactions regulating tumor immunity and defines mechanisms underlying myeloid-targeted immunotherapies currently undergoing clinical testing.