An exosome-based liquid biopsy signature for pre-operative identification of lymph node metastasis in patients with pathological high-risk T1 colorectal cancer

Background According to current guidelines, more than 70% of patients with invasive submucosal colorectal cancer (T1 CRC) undergo a radical operation with lymph node dissection, even though only similar to 10% have lymph node metastasis (LNM). Hence, there is imperative to develop biomarkers that can help robustly identify LNM-positive patients to prevent such overtreatments. Given the emerging interest in exosomal cargo as a source for biomarker development in cancer, we examined the potential of exosomal miRNAs as LNM prediction biomarkers in T1 CRC. Methods We analyzed 200 patients with high-risk T1 CRC from two independent cohorts, including a training (n = 58) and a validation cohort (n = 142). Cell-free and exosomal RNAs from pre-operative serum were extracted, followed by quantitative reverse-transcription polymerase chain reactions for a panel of miRNAs. Results A panel of four miRNAs (miR-181b, miR-193b, miR-195, and miR-411) exhibited robust ability for detecting LNM in the exosomal vs. cell-free component. We subsequently established a cell-free and exosomal combination signature, successfully validated in two independent clinical cohorts (AUC, 0.84; 95% CI 0.70-0.98). Finally, we developed a risk-stratification model by including key pathological features, which reduced the false positive rates for LNM by 76% without missing any true LNM-positive patients. Conclusions Our novel exosomal miRNA-based liquid biopsy signature robustly identifies T1 CRC patients at risk of LNM in a preoperative setting. This could be clinically transformative in reducing the significant overtreatment burden of this malignancy.

Automated summary

According to current guidelines, a large number of T1 CRC patients undergo unnecessary radical operations due to the lack of accurate biomarkers for identifying those with lymph node metastasis (LNM). This study found that a panel of exosomal miRNAs can predict LNM in T1 CRC patients with high accuracy, and a risk-stratification model including pathological features further improves the specificity.