An Immune-Related Prognostic Classifier Is Associated with Diffuse Large B Cell Lymphoma Microenvironment

Figure 4

Development and validation of a nomogram. (a) Nomogram combining risk score and other clinicopathological covariates. (b, c) Calibration plots indicating the correspondence between real observations and nomogram-predicted three-year (b) and five-year (c) survival probabilities of the GSE31312 cohort. (d, e) Calibration plots indicating the correspondence between real observations and nomogram-predicted three-year (d) and five-year (e) survival probabilities of the GSE10846 cohort. (f, g) ROC curves describing the accuracy of the OS nomogram. In the training cohort, the AUC was 0.799 for three-year OS (f) and 0.823 for five-year OS (g). (h, i) ROC curves describing the accuracy of the OS nomogram. In the GSE10846 cohort, the AUC was 0.751 for three-year OS (h) and 0.754 for five-year OS (i). (j, k) ROC curves integrating age and subtype with the multi-IRG classifier to predict the outcome of DLBCL in patients from the GSE32918 cohort. The AUC was 0.719 for three-year OS (j) and 0.718 for five-year OS (k).

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)