Abstract
In conclusion, this meta-analysis revealed that TMB can be used as a potential predictive biomarker of colorectal cancer patients receiving ICI therapy. Nevertheless, this finding is not stable enough. Therefore, many more randomized controlled trials are needed to prove that TMB is reliable enough to be used clinically to predict the efficacy of immunotherapy in colorectal cancer. And the most relevant biomarker remains to be determined when TMB high overlaps with other biomarkers like MSI and TILs.
Introduction
Immune Checkpoint Inhibitors (ICIs) have revolutionized cancer treatment, yet their efficacy is limited to a specific subset of patients. In colorectal cancer (CRC), the standard biomarker for predicting ICI response is deficient mismatch repair or high microsatellite instability (dMMR/MSI-High). However, this status only applies to approximately 5% of metastatic CRC patients. Furthermore, response is not guaranteed within this group, and some patients without this status (microsatellite stable, or MSS) still derive benefit from immunotherapy.
Oncologists require more precise biomarkers to identify the broader population of CRC patients who can benefit from ICIs. Tumor Mutational Burden (TMB)—the number of nonsynonymous mutations per megabase—has emerged as a strong predictor of ICI efficacy in other cancers like melanoma and non-small cell lung cancer. This study investigates whether TMB can serve as a similarly reliable, independent biomarker for CRC.
Methods
This systematic review and meta-analysis followed PRISMA guidelines. The authors searched PubMed and Ovid MEDLINE through May 2021 for clinical trials or cohort studies reporting TMB and ICI outcomes in CRC.
Statistical analysis involved pooling Hazard Ratios (HR) for Overall Survival (OS) and Odds Ratios (OR) for Objective Response Rate (ORR). Heterogeneity was assessed using the $I^2$ statistic, and sensitivity analyses were performed by iteratively removing studies to check result stability. Subgroup analyses were also conducted based on cohort size, geography, sequencing method, and TMB cutoffs.
The final analysis synthesized data from 13 studies (2015–2021) involving 5,062 patient samples. The dataset primarily comprised retrospective cohort studies (11), with one prospective cohort and one randomized controlled trial (RCT). Sequencing methods varied, including Whole Exome Sequencing (WES) and various targeted Next-Generation Sequencing (NGS) panels.
Results
The meta-analysis indicates that TMB-high status is associated with improved clinical outcomes in CRC patients treated with ICIs.
Overall Survival (OS)
Patients in the TMB-high group demonstrated a 32% lower risk of death compared to the TMB-low group (Pooled HR = 0.68, 95% CI: 0.51–0.92, p = 0.013). However, significant statistical heterogeneity was observed ($I^2$ = 82.7%, p < 0.001). Sensitivity analysis showed that the exclusion of a single large study (Zaidi et al.) reduced heterogeneity but also rendered the OS finding statistically non-significant, suggesting the result is somewhat fragile.
Objective Response Rate (ORR)
The association with tumor shrinkage was even more pronounced. TMB-high patients had an ORR 19.25 times higher than TMB-low patients (Pooled OR = 19.25, 95% CI: 10.06–36.82, p < 0.001). This result showed much lower heterogeneity ($I^2$ = 22.8%).
The Impact of Methodology
The predictive power of TMB varied significantly based on the measurement protocol. Subgroup analysis revealed that studies using a TMB cutoff of less than 28 muts/Mb reported a stronger association with OS than those using higher thresholds. Additionally, the choice of NGS panel mattered; studies utilizing the MSK-IMPACT panel showed consistent OS benefits with zero heterogeneity, whereas other panels showed high heterogeneity.
Discussion and Conclusion
While TMB shows promise as a predictive biomarker for ICI efficacy in CRC, several hurdles remain. The high heterogeneity across studies highlights the lack of standardization in TMB detection methods, sample sources (tissue vs. blood), and numerical cutoffs.
Future clinical implementation requires large-scale, prospective randomized controlled trials to validate these findings. Research must focus on standardizing testing platforms and establishing universal cutoff strategies. Furthermore, investigating how TMB interacts with other biomarkers—such as MSI status, POLE mutations, and Tumor Infiltrating Lymphocytes (TILs)—will be essential for developing more robust predictive models for immunotherapy in colorectal cancer.
Summary updated