Inter- and intra-tumor heterogeneous immune presentation in primary uterine carcinosarcoma determined by digital spatial profiling

Abstract

Uterine carcinosarcoma (UCS) is a rare gynecologic cancer with unfavorable prognosis. Its biphasic histologic components are known to associated with outcomes. Dissecting the tumor heterogeneity within UCS is crucial to discover molecular biomarkers for patient stratification. The aim of this study was to utilize spatial profiling technology to analyze immune presentation in different histological components within UCS tumors. The study included 23 UCS patients whose tumor sections were subjected to nanoString GeoMx Digital Spatial Profiling (DSP) to profile 30 immune-related proteins in 491 segments. We identified distinct cluster-labeled subtypes showing varied immune marker expressions. The immune presentations within the carcinoma and sarcoma components showed significant differences. Markers indicating M2 macrophages (CD45+CD163+) and epithelial-mesenchymal transition (EMT) (SMA, fibronectin) were highly expressed in the sarcoma components. The immune presentation subtypes within each histological component also differed between patients with recurrence in 5 years and without recurrence over 5 years. The cluster-labeled subtype with high Neutrophil/PMN-MDSC marker expression (CD66b) was associated with longer survival, whereas the subtype with high NK cell marker expressions (CD56) was linked to poor prognosis. Our result suggested that immune features within the carcinoma component of UCS tumors would determine patient outcomes and should be considered as a strategy for patient stratification.