Diagnostic, prognostic, and predictive biomarkers—from genomics to multi-omics integration—are shaping how physicians diagnose, stratify risk, and select therapies for patients with colorectal cancer, based on a narrative review published in the Journal of Personalized Medicine by Alfonso Agüera-Sánchez, of Universidad de Murcia, Spain, and colleagues.
“Established biomarkers such as KRAS/NRAS and BRAF mutations, MSI [microsatellite instability]/dMMR [deficient DNA mismatch repair] status, and CEA [carcinoembryonic antigen] are now cornerstones of personalized CRC [colorectal cancer] management, guiding therapy choices and surveillance, while a wide range of emerging biomarkers—from DNA methylation panels and circulating miRNAs [microRNAs] to [immune-based prognostic scores] and microbiome signatures—are moving toward translation,” the researchers wrote.
Diagnostic Applications
For diagnosis and early detection, the review highlighted “intensive” research into noninvasive and cost-effective biomarkers—including stool- and blood-based assays such as DNA methylation markers, circulating tumor DNA (ctDNA), exosomal biomarkers, and microbiome analysis—to complement or enhance existing screening approaches.
As examples, the researchers cited a US Food and Drug Administration–approved multitarget stool DNA test that combines hemoglobin detection with tumor-derived DNA alterations to achieve approximately 92% sensitivity for CRC, as well as an emerging next-generation cell-free DNA assay with approximately 83% sensitivity. However, they noted that “the main challenge remains the sensitivity in asymptomatic early-stage CRC, when the concentration of tumor DNA in blood is very low.”
The discussion also covered RNA-based biomarkers, including circulating miRNAs and long noncoding RNAs, along with exosome-based protein markers and gut microbiome profiling. The researchers emphasized a trend toward multianalyte tests and precision diagnostic models that include integrative biomarker panels in combination with existing methods, with one example being the combination of transfer RNA–derived small RNA fragments with CEA and Cancer Antigen 19-9 recently achieving 86% sensitivity and 84% specificity for CRC detection.
Prognostic Applications
The researchers further reviewed prognostic biomarkers in CRC, spanning both tumor-intrinsic molecular features and circulating biomarkers, and highlighted how multi-omics integration is being explored to develop composite prognostic models that integrate genomic data, transcriptomic subtypes, and microbiome or metabolomic profiles to refine risk groups. One analysis of an integrated prognostic signature—built on transcriptomic and other omics features—reported time-dependent areas under the curve exceeding 0.90 for 1- to 3-year survival, they noted.
“The consensus from emerging literature is that no single biomarker can capture the full complexity of tumor behavior,” the researchers explained. “Therefore, holistic models incorporating diverse data [genetic, epigenetic, immunologic, etc.] are likely to provide the most robust prognostic stratification.”
Predictive Applications
The researchers wrote that predictive biomarkers are well established in the metastatic setting, with RAS mutations and MSI guiding anti-EGFR antibody and immunotherapy selection, respectively. In reviewing emerging approaches, they noted that liquid biopsy monitoring of ctDNA can detect emergent RAS resistance mutations a few months before radiologic progression. They also cited evidence showing that patients who initially responded and then experienced disease progression with anti-EGFR therapy could benefit from rechallenge if they had no RAS/BRAF mutations in ctDNA after a “drug holiday,” with objective responses of approximately 30%.
The discussion also covered alterations occurring in smaller subsets of patients; HER2-targeted therapies have achieved response rates of approximately 30% to 50% in heavily pretreated patients, while NTRK gene fusion–positive tumors demonstrated response rates exceeding 70% with TRK inhibitors across clinical trials. Additional predictive factors under investigation include PIK3CA mutations, PTEN loss, tumor sidedness, and gut microbiome composition.
Beyond guiding initial treatment selection, biomarkers are increasingly integrated into response monitoring to assess ongoing therapy and detect resistance or relapse earlier than conventional imaging, the researchers noted. They wrote that serial evaluation of circulating markers such as CEA, ctDNA, and other analytes “holds great promise for personalizing therapy duration and sequence.”
Challenges and Future Directions
The researchers concluded their review by contextualizing the current state and future trajectory of biomarker development in CRC. They acknowledged both the progress achieved with established markers and the challenges that remain, including the need for large-scale validation trials, assay standardization across platforms, and addressing issues of access and cost-effectiveness. They emphasized that the path forward includes multi-omics integration, adaptive precision oncology approaches, advances in noninvasive screening technologies, and deeper characterization of the tumor microenvironment to guide immunotherapy strategies.
“Overall, the current landscape of CRC biomarkers is rich and rapidly evolving. Translating these advances into routine care will require coordinated efforts in terms of validation, standardization, and clinician and patient education. Yet the trajectory is clear: biomarker-driven approaches are poised to refine CRC screening, prognostication, and treatment. By tailoring interventions to the molecular features of each tumor and adapting those interventions as the tumor evolves over time, we move closer to truly personalized CRC care, with improved survival and quality of life as the ultimate goals,” they concluded.
The researchers reported no conflicts of interest.
Source: Journal of Personalized Medicine
