Discussion
Recent studies have shown that metabolic pathways play an important role
in regulating tumor progression [8,9]. The survival and
proliferation of cancer cells depends on metabolic reprogramming
[10]. Many studies have reported the possibility of metabolic
pathways as tumor-targeted treatments. Specific metabolic activities can
directly affect the transformation process or proliferation process,
which is the biological process of tumor growth [11]. Abdel-Wahab
and other reports pointed out that controlling glucose metabolism may be
a new way to inhibit cancer progression [12,13]. Recent studies have
shown that microbial metabolites, such as secondary bile acids, can
promote cancer. The metabolism of intestinal microbes related to cancer
and diets rich in fat and meat, and extracellular metabolism can promote
cancer progression [14]. However, the basic mechanism of metabolism
in COAD has not been fully elucidated, which hinders the targeted
therapy of metabolism. Therefore, the discovery of new molecular markers
related to the prognosis of COAD is very important. In this study, based
on LASSO COX regression analysis, we identified 16 PRMG in the TCGA-COAD
and GSE40976 data sets to construct a prognostic model for COAD patients
and determine the risk score. The prognostic model is accurate and
accurate. Kaplan–Meier analysis proved that the risk score model can
predict the overall survival rate of COAD. Univariate and multivariate
regression analysis confirmed that risk score is an independent
prognostic factor for COAD. The AUC curve of the gene confirms that the
risk score has a good prognostic value in predicting overall survival.
The C-index of the nomogram was 0.732. DCA shows that the nomogram
prediction model has a higher clinical benefit rate than the TNM staging
system. Many enrichment analysis pathways are concentrated in metabolic
pathways. In addition to metabolic pathways, the high-risk group shows
some cancer-related pathways, such as antigen processing and
presentation, basal transcription factors, endometrial cancer,
glycolysis gluconeogenesis, erbb signal pathway, and
glycosylphosphatidylinositol gpi anchor biosynthesis . These results
show that these genes are closely related to metabolic pathways and
reveal the potential role of metabolic pathways in COAD.
Target genes are important members of metabolic pathways and can serve
as therapeutic targets for cancer. Prognosis prediction is very
important for selecting clinical treatment options for cancer patients.
Several studies have explored prognostic biomarkers and found that gene
expression profiles play a crucial role in the prognosis of cancer
[15]. Although our screening of these genes related to cancer
prognosis is rarely reported, these genes can reflect the status of
cancer driver genes related to their upstream and downstream to a
certain extent. The genes we screened are rich in a variety of
cancer-related pathways. Based on these results, we concluded that the
risk score can accurately predict the survival of patients with COAD,
perhaps because the score can reflect the multi-level status of COAD. We
constructed a nomogram to predict individualized clinical outcomes. The
nomogram generates a graphical statistical prediction model that assigns
scores to each factor, including age, gender, and clinical stage,
covering important factors that affect clinical outcomes. In addition to
traditional clinicopathological characteristics (such as age, gender,
TNM staging), risk scores based on genetic markers can also be
incorporated into the predictive nomogram model to predict clinical
outcomes. The nomogram is a stable and reliable quantification of
personal risk by combining clinical characteristics and risk scores. Our
nomogram includes risk scores and clinicopathological characteristics,
which can well predict patients with colon adenocarcinoma at 1, 2, and 3
years survival rate. The calibration curve for predicting OS indicated
that the nomogram-predicted survival closely corresponded with actual
survival outcomes. We constructed 16 metabolic gene models based on TCGA
and GEO to predict the prognosis of COAD patients. The risk score based
on 16 genes may be a promising independent prognostic biomarker.
However, these are not yet clear. How genes play their roles in the
mechanism, therefore, more research is needed to explore the impact of
metabolic enzymes on survival. The study has limitations. First of all,
this is a retrospective study. Therefore, information including
recurrence time, treatment records and detailed pathological staging
cannot be obtained. Second, although the model has been validated in all
cohorts, it still needs more samples for further confirmation before
clinical application.