Exploration of tumor-infiltrating immune cells on the prognosis of
endometrial cancer
Abstract
Setting:Immune cells played a vital role in the progression of
endometrial cancer, whereas immunologic mechanism and clinical
prognostic biomarkers remained to be identified. Objective: To explore
the landscape of tumor-infiltrating immune cells in endometrial cancer
and the correlation between TIICs and EC prognosis. Methods: Data of 530
patients was downloaded from TCGA. CIBERSORTx estimated the abundance of
immune cells. Ranked data used Spearman coefficient to measure the
relationship between immune cells and clinical characteristics.
Univariate COX hazard analysis was performed to explore the relationship
between immune cells and clinical outcomes. Lasso regression was used to
screen variables. Multivariate regression was performed for stepwise
regression variable screening. Univariate and multivariate Independent
prognosis analysis was performed to validate the reliability of the
multivariate risk model. A nomographic chart was used to grade and
assess each patient’s prognosis. Results: 11 type TIICs were increased
in cancer tissues(P<0.05). Univariable regression revealed
that 5 types TIICs were significantly related with EC prognosis. TTMMD
model was constructed via lasso regression and multivariable cox
regression. Log-rank was used to perform survival analysis, and the
result was presented by Kaplan-Meier curve(p=5.13e-03). The Roc curve
was simultaneously fulfilled to validate the model’s accuracy and
predictability(AUC=0.676). The Kaplan-Meier curve of the training set
was statistically significant(p=3.807e-02). ROC curve replied the
feasibility of the model(AUC=0.678). The risk model was an independent
factor for predicting EC prognosis(p<0.001). Conclusion: 11
types TIICs were significantly associated with the prognosis of
endometrial cancers. The immune-cells-based risk model was reliable for
prognosis assessment.