Construction of immune score and its prognostic value in invasive lobular carcinoma of the breast using computational pathology analysis.

Abstract

Previous studies have shown that high level of TILs in invasive lobular carcinoma (ILC) is associated with poor prognosis, contrary to that in TNBC and HER2-positive breast cancer. The densities of six immune cell markers and three immune checkpoints in the ILC microenvironment were detected by computational pathology analysis. Then, the LASSO cox regression model was used to construct an immune score (IS) and further evaluate its prognostic value. In our ILC cohort, the low density of CD4, CD8, CD20, CD56, CD68, FOXP3, PD-1, and PD-L1 had significantly longer disease-free survival (DFS) and overall survival (OS); however, the low density of CTLA-4 was associated with shorter DFS and OS. Based on this, an IS was constructed, and patients with low-IS had significantly prolonged DFS (p < 0.0001) and OS (p < 0.0001). Multivariate analysis revealed that IS was an independent prognostic indicator for DFS and OS. Further analysis showed that IS may increase the prognostic value of TNM stage. We further explored the prognostic role of CD68 and FOXP3 in the transcriptional level and the corresponding ISm in the METABRIC dataset, and found that low proportion of CD68 and FOXP3 and their ISm were associated with longer OS, and ISm was also an independent prognostic factor for OS. IS was a promising biomarker to distinguish the prognosis in ILC patients.