Mullerian adenosarcoma of the uterus, ovary and peritoneum: Whole transcriptomic digital spatial profiling (DSP) in a cohort of 42 tumors

: Mullerian adenosarcoma (MA) is a rare biphasic tumor with a malignant stromal component and benign glandular epithelium with periglandular stromal condensation occurring mainly in the uterus. The diagnosis is essentially morphologic and may be difficult particularly in distinguishing MA from other biphasic tumors at the lower end of the spectrum with histologic features similar to, but falling short-off adenosarcoma; i.e., uterine atypical polyp/low-grade or well-differentiated adenosarcoma with low mitotic count and lacking significant nuclear atypia. We investigated the transcriptomic profiles of 42 biphasic mesenchymal tumors, diagnosed as MAs, by whole transcriptomic Digital Spatial Profiling (DSP) and morphology marker antibody staining with PanCK, CD45 and SYTO 13 using the GeoMx platform on tissue microarrays (TMAs). Regions of interest (ROI) were segmented using morphology marker staining for tumor stromal component/periglandular cuffing and immune infiltrated areas of illumination (AOIs). The median age at diagnosis was 46.5 (39.0-64.8) years. Sarcomatous overgrowth was present in 9 (21.4%), myometrial invasion in 11 (26.2%), heterologous elements in 5 (11.9%), and sex-cord differentiation in 8 (19.0%). At follow-up, metastases occurred in 8 (19.0%) patients; 7 (16.6%) died of disease and 5 (11.9%) were alive with disease. DSP revealed two groups of tumors with significant differences in gene expression profiles; namely, conventional MAs and biphasic tumors exhibiting similar morphologic features but insufficient for the diagnosis. In addition, we explored in-depth the molecular profiles of MAs with and without sarcomatous overgrowth as well as MAs with and without myometrial invasion, and also found differentially expressed genes and gene-pathways. Our results revealed that MAs have distinct transcriptomic signatures and confirm the existence of a spectrum of mesenchymal biphasic mullerian tumors. Given the lack of consistent mutations in MAs, transcriptomic-based biomarkers may aid in their classification and should be further investigated.