Colorectal cancer (CRC) is a heterogeneous disease with complex molecular underpinnings. In recent years, research has elucidated a variety of molecular features that influence both the prognosis and treatment response of patients with CRC. These biomarkers not only help stratify risk but also guide therapeutic decisions, particularly in the era of precision oncology. This article outlines the most clinically relevant molecular biomarkers in CRC, their prognostic and predictive significance, current guideline recommendations, and the ongoing developments in molecular classification.
1. Overview of Prognostic and Predictive Molecular Markers
Molecular alterations in CRC are critical not only for understanding tumor biology but also for personalizing treatment strategies. Certain molecular features such as mismatch repair deficiency (dMMR)/microsatellite instability (MSI), RAS and BRAF mutations, and emerging gene expression profiles have been found to independently impact prognosis, regardless of tumor stage. Some of these markers also predict response to specific targeted therapies.
Key molecular markers with both prognostic and predictive relevance include:
- dMMR/MSI – Predicts lack of benefit from fluoropyrimidine monotherapy and indicates potential benefit from
- BRAF mutations – Associated with poor prognosis in microsatellite stable tumors and predicts response to BRAF
- RAS mutations (KRAS, NRAS) – Predict lack of efficacy of EGFR-targeted therapies and are being targeted with KRAS G12C
2. Guideline Recommendations
Joint guidelines from the American Society for Clinical Pathology (ASCP), College of American Pathologists (CAP), Association for Molecular Pathology (AMP), and American Society of Clinical Oncology (ASCO) recommend the following:
- MMR testing for all CRC patients for prognostic stratification and Lynch syndrome risk assessment.
- Extended RAS testing (KRAS and NRAS exons 2, 3, and 4) prior to considering EGFR inhibitor
- BRAF V600E mutation analysis in dMMR tumors with MLH1 loss to evaluate for sporadic vs hereditary origins (Lynch syndrome risk).
However, BRAF V600E is not currently recommended as a predictive marker for EGFR inhibitor therapy due to conflicting data.
3. Mismatch Repair Deficiency (dMMR) and Microsatellite Instability (MSI)
Definition and Prevalence
- dMMR occurs in approximately 15–20% of sporadic CRCs and most cases of Lynch
- Characterized by accumulation of replication errors, leading to MSI, which is
defined as instability in ≥30% of microsatellite loci.
Prognostic Significance
- dMMR/MSI-H tumors are associated with better prognosis in early-stage disease, despite poor
- The prognostic advantage may not extend to metastatic CRC, where MSI-H is less common (3.5%) and may coincide with adverse features such as BRAF
Predictive Value
- Adjuvant fluoropyrimidine chemotherapy appears less effective in MSI-H stage II
- MSI-H/dMMR predicts responsiveness to immune checkpoint inhibitors in both locoregional and metastatic
Hereditary vs Sporadic dMMR Tumors
- Lynch syndrome-associated dMMR tumors exhibit higher neoantigen loads and somatic mutations than sporadic tumors, potentially influencing immune response and
- The implications for differential treatment response remain under
4. RAS and BRAF Mutations
KRAS and NRAS
- KRAS mutations (codons 12 and 13) occur in 12–75% of
- Associated with inferior prognosis, particularly codon 12 mutations in node- positive
- NRAS mutations, though less common, also correlate with poor
BRAF Mutations
- V600E mutation is present in ~10% of CRCs and linked to poor prognosis, especially in non-MSI-H tumors.
- BRAF-mutant tumors are resistant to anti-EGFR
- Non-V600E BRAF mutations (~2.2%) tend to occur in younger patients, are less aggressive, and have better survival outcomes than V600E or wild-type
Clinical Implications
- Routine testing for RAS and BRAF mutations is recommended for all CRCs beyond stage I.
- Results guide therapeutic choices, including EGFR and BRAF-targeted
5. Other Molecular Markers Under Investigation
Numerous other genes and biomarkers have been studied, though their clinical application remains unproven due to variability in methodology and inconsistent results:
- Tumor suppressor genes: TP53, SMAD4, LOH at chromosomes 1p, 5q, and
- Oncogenes: c-
- Apoptosis regulators: Bcl-2, BAX,
- Growth factors/receptors: TGF-α/β,
- Cell cycle regulators: p21,
- Angiogenesis:
- Adhesion molecules: CD44, E-
- Matrix modulators: MMPs, urokinase-type plasminogen
- MicroRNA, circulating tumor DNA (ctDNA), and epigenetic markers (CIMP, methylation)
- Soluble collagen IV and 18q deletions
Standardization of assays and robust multivariate analyses are needed before these markers can be integrated into clinical prognostication.
6. Gene Expression Profiles and Prognostic Molecular Tools
Gene expression-based tools such as the 12-gene Oncotype DX Colon Cancer Assay have shown promise in predicting recurrence in stage II and III CRCs. However, large prospective trials are needed for validation before clinical adoption.
7. Molecular Classification of CRC
Gene expression analysis has identified four consensus molecular subtypes (CMS), each with distinct biology and clinical implications:
| Subtype | Features | Prognosis | Associated Pathways |
| CMS1 (MSI-like) | dMMR/MSI-H, CIMP+,
BRAF mutations |
Good | Immune infiltration, high
mutational load |
| CMS2
(Canonical) |
CIN+, WNT/MYC
activation |
Intermediate |
Low immune activity |
| CMS3
(Metabolic) |
KRAS mutations, metabolic dysregulation | Intermediate |
Low immune activity |
| CMS4
(Mesenchymal) |
Fibroblast-rich, TGF-β activation |
Poor |
Angiogenesis, inflammation, immunosuppression |
The CMS classification correlates with tumor microenvironment and influences response to therapy. CMS4 tumors exhibit poor outcomes and are characterized by stromal cell-driven inflammation and fibrosis, not tumor cell epithelial-to- mesenchymal transition as previously thought.
8. Future Directions
Efforts continue to integrate molecular data into clinical practice to improve prognostic accuracy and therapeutic targeting. Emerging areas of interest include:
- Integration of CMS with TNM
- Exploration of ctDNA for early relapse
- Tailored immunotherapy approaches based on tumor microenvironment and molecular
- Prospective validation of molecular assays in large clinical
Conclusion
The molecular landscape of colorectal cancer is complex and dynamic. Understanding key molecular alterations—including MSI status, RAS and BRAF mutations, and gene expression profiles—is essential for accurate prognosis and guiding targeted therapy. Although significant progress has been made, the clinical utility of many emerging biomarkers awaits further validation. Future incorporation of molecular classification into routine staging systems holds the potential to refine patient stratification and optimize personalized treatment strategies in CRC.
