Early stage CRC is highly heterogeneous at multiple molecular levels. However, microsatellite instability (MSI) is the only recommended biomarker to risk stratify patients (stage II). New biomarkers and clinically-applicable assays are highly necessary. These require thorough biological understanding and prospective validation for real-time application.
Stage I-III CRC patients were prospectively recruited at the INCLIVA Institute (Valencia) in collaboration with the Institute of Cancer Research (London). Clinicopathological features (such as stage, grade, sidedness, MSI, CDX2 expression, vascular/perineural invasion, Neu/Lymph ratio (NLR) and platelets/lymph (PLR)) were collected. DNA and RNA extracted from FFPE samples were assessed with a custom 29 frequently mutated CRC gene-panel (NGS) and a validated CRCAssigner/CMS subtyping assay (NanoCRC – NanoString Technologies). Plasma-based prometastatic cytokines (IL-6, IL-11, TGF-b) were measured (ELISA). Fisher’s exact test, Kruskal-Wallis and Mann Whitney tests were used for categorical/non categorical variables.
A total of 119 patients with known clinical variables were available for subtypes, mutational profile and cytokines levels analysis. NanoCRC subtypes demonstrated known significant association (p<0.05) with clinical and mutational features (CMS1 and right side, MSI high; CMS3 and RAS mutation) and newly observed significant association with CDX2 expression (low in CMS1), CTNNB1 mutation (enriched in CMS3), vascular/perineural invasion (high in mesenchymal subtypes CMS1 and 4). Between cytokines, high IL-6 levels were significantly associated (p<0.05) with CMS subtypes (high in CMS1) BRAF mutation, right sidedness, MSI status, high grade, vascular invasion, high NLR and PLR. After 18.7 months of median follow-up, disease relapse was observed in 10 patients. A multivariable analysis confirmed stage, low CDX2, vascular/perineural invasion significantly associated with relapse and borderline significance for CMS4.
LBs response-based multi-omics profiles identify diverse risk groups, predict response to adjuvant therapy and may ultimately lead to a personalised approach for early CRC.