Original Article
Impact of negative lymph nodes on colon cancer survival and exploring relevant transcriptomics differences through real-world data analyses
Abstract
Background: The prognostic role and underlying heterogeneity of negative lymph nodes (NLNs) on colon cancer is not well understood. The purpose of this study was to construct NLN-based prognostic models and reveal relevant mechanisms affecting NLNs by analyzing omic data.
Methods: This inception cohort study included 314,398 colon cancer patients from the US Surveillance, Epidemiology, and End Results (SEER) database. Receiver operating characteristic (ROC) curve was used to determine the cut-off of NLNs. Nomograms were constructed and validated using SEER data and the Cancer Genome Atlas (TCGA) data, respectively. The differentially expressed genes (DEGs) were analyzed using edgeR. Enrichment analyses were performed by Metascape.
Results: Multivariate analysis confirmed the high NLN had improved cancer-specific survival (CSS) and overall survival (OS) compared to low NLN [hazard ratio (HR) =0.610, 95% confidence interval (CI), 0.601–0.620] for CSS and (HR =0. 682, 95% CI, 0.674–0.690) for OS. Nomograms were established for CSS and OS with the c-statistic 0.790 (95% CI, 0.788–0.792) for CSS and 0.734 (95% CI, 0.732–0.736) for OS. High NLN was associated with less B cell (P=0.002) and macrophage infiltration (P<0.0001), high microsatellite instability (MSI) (OR =4.325, P=0.001), and hypermutation (OR =4.285, P=0.001; high vs. low). Transcriptomics analysis demonstrated histone modifiers were the most significant different biological processes between the high and low NLN group.
Conclusions: The NLN-based models can aid in personalized risk stratification for colon cancer. This study postulates that high NLN may represent a biological subtype with less macrophage infiltration, high MSI status, hypermutation, and histone modifier gene enriched expression, and thus warrants further investigation.
Methods: This inception cohort study included 314,398 colon cancer patients from the US Surveillance, Epidemiology, and End Results (SEER) database. Receiver operating characteristic (ROC) curve was used to determine the cut-off of NLNs. Nomograms were constructed and validated using SEER data and the Cancer Genome Atlas (TCGA) data, respectively. The differentially expressed genes (DEGs) were analyzed using edgeR. Enrichment analyses were performed by Metascape.
Results: Multivariate analysis confirmed the high NLN had improved cancer-specific survival (CSS) and overall survival (OS) compared to low NLN [hazard ratio (HR) =0.610, 95% confidence interval (CI), 0.601–0.620] for CSS and (HR =0. 682, 95% CI, 0.674–0.690) for OS. Nomograms were established for CSS and OS with the c-statistic 0.790 (95% CI, 0.788–0.792) for CSS and 0.734 (95% CI, 0.732–0.736) for OS. High NLN was associated with less B cell (P=0.002) and macrophage infiltration (P<0.0001), high microsatellite instability (MSI) (OR =4.325, P=0.001), and hypermutation (OR =4.285, P=0.001; high vs. low). Transcriptomics analysis demonstrated histone modifiers were the most significant different biological processes between the high and low NLN group.
Conclusions: The NLN-based models can aid in personalized risk stratification for colon cancer. This study postulates that high NLN may represent a biological subtype with less macrophage infiltration, high MSI status, hypermutation, and histone modifier gene enriched expression, and thus warrants further investigation.
