Colorectal cancer (CRC) is a heterogeneous disease and each CRC possesses a unique molecular tumor profile. The main pathways of oncogenesis are the chromosomal instability, microsatellite instability and serrated neoplasia pathway. Sessile serrated adenomas/polyps (SSA/Ps) may be the precursor lesions of CRC arising via the serrated neoplasia pathway. This has led to a paradigm shift because all SSA/Ps should be detected and resected during colonoscopy. The ability to accurately detect and resect only those polyps with a malignant potential could result in safer and cost-effective practice. Optimization of the endoscopic classification systems is however needed to implement targeted prevention methods.
Key points
- •
Colorectal cancer (CRC) is a very heterogeneous disease resulting from multiple overarching neoplasia pathways, of which the chromosomal instability, the microsatellite instability, and the serrated neoplasia pathways are the most significant.
- •
Besides conventional adenomas, serrated polyps also seem to be precursor lesions of CRC, arising via the serrated neoplasia pathway.
- •
Accurate endoscopic characterization of premalignant colonic lesions would enable the implementation of targeted treatment in the near future, resulting in effective, safe, as well as cost-effective CRC prevention.
Introduction
Colorectal cancer (CRC) is a very heterogeneous disease with regards to tumor development as well as clinical tumor behavior. Due to the presence or absence of diverse genetic and epigenetic alterations, each CRC appears to possess a unique molecular tumor profile. The oncogenesis of CRC is extensively evaluated compared to other solid tumors in the human body ( Fig. 1 ). This is partly due to its high priority because CRC has high morbidity and mortality rates in the Western world. Other factors are the slow progression of disease and the possibility to sample CRC precursor lesions, enabling the assessment of molecular changes in consecutive steps of tumorigenesis.
Until recently, conventional adenomas (further referred to as adenomas) were regarded as the only precursors of CRC, developing into cancer via the adenoma-carcinoma pathway. A multistep molecular sequence responsible for the adenoma-carcinoma pathway was first described by Fearon and Vogelstein. Although more recent literature has demonstrated that a minority of CRCs arise exactly via the pathway described by Fearon and Vogelstein, the model has long been a prominent paradigm for research concerning the origin of CRC and other solid tumors. Currently, multiple colorectal neoplasia pathways are distinguished, of which the chromosomal instability (CIN) pathway; the microsatellite instability (MSI) pathway; and the CpG island methylator pathway (CIMP), also referred to as the serrated neoplasia pathway, are the best studied. Oncogenesis occurring via these pathways may, however, not be completely separated. Crossover, or even overlap, between the pathways is likely to exist. Serrated polyps (SPs), rather than adenomas, are suggested to be the precursor lesions of CRC arising via the serrated neoplasia pathway, responsible for approximately 15% to 30% of CRC. This finding has led to an important paradigm shift in daily practice for colonoscopies because it has become apparent that adenomas as well as premalignant SP subtypes should be detected and resected to effectively prevent CRC.
Only a small percentage of polyps seem to develop into CRC, whereas most polyps remain stable over time or even regress. However, for adenomas as well as for SPs, it is unclear which molecular alterations induce the final transition toward invasive growth. As a result, it is still largely unknown which lesions will develop into CRC and which lesions will have a benign course. Profound knowledge of the colorectal neoplasia pathways is essential for each endoscopist to optimize current prevention strategies. Ongoing research could, it is hoped, lead toward the targeted resection of truly premalignant colorectal lesions in the near future.
Introduction
Colorectal cancer (CRC) is a very heterogeneous disease with regards to tumor development as well as clinical tumor behavior. Due to the presence or absence of diverse genetic and epigenetic alterations, each CRC appears to possess a unique molecular tumor profile. The oncogenesis of CRC is extensively evaluated compared to other solid tumors in the human body ( Fig. 1 ). This is partly due to its high priority because CRC has high morbidity and mortality rates in the Western world. Other factors are the slow progression of disease and the possibility to sample CRC precursor lesions, enabling the assessment of molecular changes in consecutive steps of tumorigenesis.
Until recently, conventional adenomas (further referred to as adenomas) were regarded as the only precursors of CRC, developing into cancer via the adenoma-carcinoma pathway. A multistep molecular sequence responsible for the adenoma-carcinoma pathway was first described by Fearon and Vogelstein. Although more recent literature has demonstrated that a minority of CRCs arise exactly via the pathway described by Fearon and Vogelstein, the model has long been a prominent paradigm for research concerning the origin of CRC and other solid tumors. Currently, multiple colorectal neoplasia pathways are distinguished, of which the chromosomal instability (CIN) pathway; the microsatellite instability (MSI) pathway; and the CpG island methylator pathway (CIMP), also referred to as the serrated neoplasia pathway, are the best studied. Oncogenesis occurring via these pathways may, however, not be completely separated. Crossover, or even overlap, between the pathways is likely to exist. Serrated polyps (SPs), rather than adenomas, are suggested to be the precursor lesions of CRC arising via the serrated neoplasia pathway, responsible for approximately 15% to 30% of CRC. This finding has led to an important paradigm shift in daily practice for colonoscopies because it has become apparent that adenomas as well as premalignant SP subtypes should be detected and resected to effectively prevent CRC.
Only a small percentage of polyps seem to develop into CRC, whereas most polyps remain stable over time or even regress. However, for adenomas as well as for SPs, it is unclear which molecular alterations induce the final transition toward invasive growth. As a result, it is still largely unknown which lesions will develop into CRC and which lesions will have a benign course. Profound knowledge of the colorectal neoplasia pathways is essential for each endoscopist to optimize current prevention strategies. Ongoing research could, it is hoped, lead toward the targeted resection of truly premalignant colorectal lesions in the near future.
From adenoma to carcinoma
Histopathologically, every adenoma can be subdivided based on the severity of dysplasia (low-grade or high-grade) and presence and proportion of a villous component (tubular, tubulovillous or villous adenoma). It is generally accepted that adenomas with high-grade dysplasia and/or villous components harbor an increased risk to develop into cancer compared with tubular adenomas with low-grade dysplasia. Studies that show the natural course of adenomas are, however, scarce. Several decades ago, Muto and colleagues demonstrated in an observational study that less than 5% of small tubular adenomas with low-grade dysplasia will eventually transform into CRC. This percentage increased significantly for adenomas with high-grade dysplasia and/or a villous component. In a more recent study, Brenner and colleagues demonstrated that the 10-year cumulative risk of an advanced adenoma (≥25% villous component, high-grade dysplasia or size ≥10 mm) to develop into cancer was increased from 25.4% at age 55 years to 42.9% at age 80 years in women, and from 25.2% at age 55 years to 39.7% at age 80 years in men. In this study, risk estimates were based on CRC incidence and the prevalence of advanced adenomas per age group. In an endoscopic study, Hofstad and colleagues demonstrated that small adenomas (6–9 mm) left in situ may partly regress or partly increase in size, whereas 25% of polyps appeared unchanged after 3 years. In a computed tomography–colography study, Pickhardt and colleagues demonstrated that small adenomas with a prominent villous component or high-grade dysplasia show a more rapid growth than small tubular adenomas with low-grade dysplasia. Polyps were acquired via polypectomy after a mean surveillance of 2.3 years. These studies strengthen the hypothesis that advanced adenomas possess a higher malignant potential than nonadvanced adenomas.
The Chromosomal Instability Pathway
The CIN pathway is the most evaluated sequence of colorectal carcinogenesis, arising from adenoma precursor lesions. It is suggested that around 70% of CRCs arise via the CIN pathway. CIN refers to structural changes or numerical gains or losses of chromosomes that result in variability in the karyotype of individual cells (aneuploidy) and subsequent loss of heterozygosity (LOH) of genes. Karyotype copy defects during cell mitosis are responsible for chromosome instability. The exact underlying cause of aneuploidy is unknown. The total dwell time of tumor development via the CIN pathway is assumed to be approximately 10 to 15 years.
Initiation of adenoma development
An important initiating mutation for early adenoma development is the mutation of the adenomatous polyposis coli (APC) tumor suppressor gene, which has a role in the Wnt pathway. Familial adenomatous polyposis (FAP) has served as a valuable model to evaluate the exact function of the APC gene. FAP patients have a germline mutation in the APC gene, which results in the phenotypic expression of hundreds of adenomas throughout the colorectum. The inactivation of APC will result in accumulation of β-catenin and a subsequent stimulation of T-cell factor-1 (TCF-1) and lymphoid enhancing factor-1 (LEF-1) transcription factors. This will result in increased proliferation of colorectal cells. The Wnt pathway can also be triggered via a gain-of function mutation in β-catenin, which is present in up to 50% of CIN-tumors without an APC mutation.
A second prevalent mutation in early adenoma development is the Kirsten-rat sarcoma 2 viral oncogene homolog (K-RAS) proto-oncogene mutation, which is involved in the mitogen-activated protein kinase (MAPK) cascade. Point mutations are found in 30% to 40% of CRCs. RAS proteins act as molecular switches that control intracellular signal transduction and are important for cell proliferation and apoptosis. A K-RAS mutation can result in resistance to inhibition of surface receptors, such as the epidermal growth factor receptor, and will permit the cell to evade apoptosis.
Progression of adenomas
A late event in colorectal tumorigenesis is a mutation in the p53 tumor suppressor gene, located on chromosome 17p. p53 is often defined as the guardian of the genome, playing an essential role in its stability. A mutation in p53 or 17p allelic loss has been reported in 4% to 26% of adenomas, in approximately 50% of adenomas with high-grade dysplasia, and in 50% to 75% of adenocarcinomas. This distribution of results has led to the belief that functional inactivation of p53 protein might be associated with the transition from adenoma to carcinoma, although it is unclear whether this mutation is absolutely required.
LOH of chromosome 18q is a genetic alteration that occurs in approximately 70% of colorectal adenocarcinomas. Many tumor suppressor genes are located on chromosome 18q of which deleted in colorectal carcinoma (DCC), SMAD 4, and SMAD 2 are important in regulation of controlled cell proliferation and apoptosis. 18q loss is often co-occurring with the loss of p53. Combined loss occurs in 65% of CRC.
Co-occurring events
Many other molecular events are described in the CIN pathway to CRC. For example, a mutation in the PI3K pathway is often seen, which can eventually result in accelerated cell growth. Also, microRNAs may play a role in the tumorigenesis of CIN-derived CRCs. MicroRNAs are noncoding RNAs that regulate protein expression by inhibition of messenger RNA translation. The amount of microRNAs involved in CRC pathogenesis is large and expanding. MicroRNAs can be upregulated or downregulated in CRC, operating like oncogenes and tumor suppressor genes.
The Microsatellite Instability Pathway
The second major colorectal neoplasia pathway in which adenomas are involved is the MSI pathway. Around 15% of CRCs possess genetic instability due to MSI. Of these tumors, approximately 20% have a hereditary cause and arise in patients with a germline mutation in one of the mismatch repair genes, the Lynch syndrome. The other 80% of MSI CRC seem to arise due to the hypermethylation of the promoter of the human MutL homologue (hMLH) 1, a specific mismatch repair gene. The latter is a sporadic epigenetic alteration, which may exemplify a distinct pathway to colon cancer and is suggested to play an important role in the serrated neoplasia pathway (see later discussion). The pathophysiology of mismatch repair (MMR) system dysfunction and subsequent elevated cancer risk does, however, also account for tumors with a sporadic cause of MMR dysfunction. Tumor development via the hereditary MSI pathway is assumed to be faster compared with CIN oncogenesis, resulting in CRC within approximately 3 to 5 years.
Pathophysiology of MMR dysfunction
Microsatellites or simple sequence repeats are repeated nucleotide sequences of DNA made of base-pairs, prone to mutations and base-pair substitutions during replication. The MMR system identifies and corrects errors that occur during the replication of DNA polymerase. MMR dysfunction will result in instability in the length of microsatellites, causing frame-shift mutations. The mutation rate in the DNA of a colorectal mucosal cell is increased by a 100-fold when the MMR system is disabled, facilitating accelerated tumorigenesis. A mutation in one of the MMR genes is an important factor in oncogenesis, but is not singly sufficient to develop CRC.
Other molecular alterations in the hereditary microsatellite instability pathway
Activation of the Wnt pathway is an important factor in early adenoma development in the hereditary MSI pathway, although not always observed. However, compared with the CIN pathway, Wnt pathway activation is more often due to a mutation in β-catenin than in the APC gene. In addition, alterations in the MAPK pathway seem to play a role in CRC development in patients with Lynch syndrome. K-RAS is often mutated in these patients. Another role is reserved for transforming growth factor-β, which is an inhibitor of epithelial cell growth. More than 80% of tumors derived via the hereditary MSI pathway harbor a mutation in this gene.
Transition to CRC seems to occur by influence of a mutation in the human BAX gene, which is normally activated by p53. A mutation in the BAX gene will result in an escape from intrinsic cell apoptosis, equal to a p53 gene mutation. A study demonstrated that 54.2% of Lynch syndrome CRCs showed a BAX-gene mutation, while only 4.2% of tumors showed a p53 mutation.
Serrated polyps, an alternative route to colorectal cancer
Historically, colonic polyps were mainly subdivided into adenomas and hyperplastic (or metaplastic) polyps (HPs). Whereas adenomas were considered premalignant, HPs were regarded as harmless, not harboring any neoplastic potential. Research from the last decades, however, has changed these views. The recent classification of the World Health Organization has reorganized HPs into an overarching group of SPs, histopathologically characterized by a saw-tooth infolding of the crypt epithelium. SPs are subdivided into HPs, sessile serrated adenomas/polyps (SSA/Ps) and traditional serrated adenomas (TSAs). HPs can be further subdivided into a microvesicular-type HP (MVHP), a goblet cell–type HP, and a mucin-poor–type HP. It is suggested that SSA/Ps are the main precursors of CRC developing via the serrated neoplasia pathway, responsible for approximately 15% to 30% of CRCs. TSAs are also considered premalignant; however, oncogenesis is indistinct and TSAs seem to be very rare in asymptomatic patient cohorts.
Several studies have compared the molecular profile of conventional adenomas, SPs, and CRCs, and detected a high correlation between SPs and a subset of CRCs. This has led to the proposal of a colonic mucosa→MVHP→SSA/P→SSA/P with dysplasia→CRC sequence. In this pathway, it is yet unclear if SSA/Ps could also directly originate from colonic mucosa without HPs as an intermediate stage. The natural course of SSA/Ps, as well as HPs, however, is indistinct and no proper studies have been performed that demonstrate the actual cancer risk of individual SPs. In one study, subjects diagnosed with a single SSA/P were prospectively followed for a median of 7.2 years. In total, 12.5% of these subjects developed cancer, compared with 1.8% for matched subjects diagnosed with an HP or tubular adenoma. In a recent study based on genome-wide profiling of gene expression, a subset of CRCs showed close molecular similarities to SSA/P and had an unfavorable prognosis. Therefore, it seems important that all SSA/Ps are detected and radically resected during colonoscopy to prevent CRC.
The Serrated Neoplasia Pathway
The existence of the serrated neoplasia pathway as a distinct sequence to CRC was first described approximately 15 years ago and is characterized by hypermethylation and subsequent silencing of tumor-suppressor genes as well as a mutation in the BRAF-oncogene. Much is learned from patients with serrated polyposis syndrome to gain a better understanding of the molecular changes found in the serrated neoplasia pathway. However, the total dwell time of tumorigenesis is largely unknown. A relatively rapid development of CRC is suggested in the literature, comparable with carcinogenesis in patients with Lynch syndrome. A relatively slow progression of disease is also proposed based on a cross-sectional study using computerized algorithms, showing a median transition time of 15 years from SSA/P to CRC.
Key features of the serrated neoplasia pathway
The key feature of oncogenesis via the serrated neoplasia pathway is the hypermethylation of CpG islands on the promoter regions of tumor suppressor genes, resulting in subsequent silencing of these genes. The CIMP is a common way to describe the methylation status of a lesion, which can be expressed as a CIMP-high, CIMP-low, or CIMP-negative phenotype. In 2 recent studies, the CIMP-high phenotype was found in 20% to 26% of SPs, which increased to 63% for proximal SSA/Ps. The CIMP-high phenotype was detected in approximately 1% of adenomas, indicating an association between SPs and CIMP-high CRC. For this reason, the serrated neoplasia pathway is also often referred to as the CIMP pathway. It is also plausible that a small subset of CIMP-high CRCs arise from adenomas rather than from SPs. Similarly, some SPs may develop into CRCs without CIMP.
Hypermethylation and subsequent silencing of hMLH1 is an important step for oncogenesis in CIMP-high tumors and will lead to a dysfunction of the MMR system and subsequent MSI, comparable with a germline mutation in hMLH1 as seen in Lynch syndrome. Although silencing of MLH1 is a major and accelerating step in oncogenesis via the serrated neoplasia pathway, it is not obligatory. Jass showed that, in a cohort of subjects with CRC, 12% of tumors had a CIMP-high MSI phenotype, whereas 8% had a CIMP-high microsatellite stable (MSS) phenotype, both highly correlated to SPs.
Other well-described molecular alterations are the silencing of tumor suppressor genes p16 and IGFB7. P16 is a cell cycle inhibitor and dysfunction will result in senescence of the lesion and progression from nonadvanced to more advanced precursor lesions. IGFBP7 functions downstream of p53 and silencing of this gene can mimic the inactivation of p53 associated to the progression from advanced precursor lesions to CRC.
A mutation in the BRAF-oncogene seems to be another key feature in the serrated neoplasia pathway. BRAF has a role in the MAPK cascade and is regulated by the activity of K-RAS. A point mutation in BRAF will result in uncontrolled cell proliferation, comparable with the effect of a K-RAS mutation, often seen in adenomas. In 2 large recent studies the presence of a BRAF mutation was found in 50% to 72% of MVHPs, 70% to 80% of SSA/Ps, and only 1% of adenomas. Another study showed a strong correlation between the presence of a BRAF mutation and a CIMP-high phenotype in CRC. A BRAF mutation was found in 77% of CIMP-high tumors, 18% of CIMP-low, and 0% of CIMP-negative CRCs. These results strengthen the hypothesis of a distinct serrated neoplasia pathway arising from SPs and resulting in CIMP-high, BRAF-mutant CRCs.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
