Abstract / Summary
Pilomatrix-like high-grade endometrioid carcinoma (PiMHEC) is an aggressive variant of endometrioid carcinoma with basaloid high-grade morphology, shadow cells, and aberrant β-catenin expression. Its cytologic features remain poorly defined. We describe cytologic findings and histologic, immunohistochemical, and molecular correlates. Five histologically confirmed PiMHECs with peritoneal washing specimens were identified retrospectively. Four contained malignant cells on review and were included in the cytomorphologic analysis. Concurrent cervicovaginal liquid-based cytology was available in two cases. Corresponding surgical specimens, immunostains, and targeted next-generation sequencing results were reviewed. Peritoneal washings were malignant in all four cases; both cervicovaginal specimens were negative. Tumor cells occurred as loose single cells, sheets/trabeculae, papillary and three-dimensional clusters, and palisaded clusters. High-grade cells were small to medium sized with scant cytoplasm, high nuclear-to-cytoplasmic ratios, round to angulated nuclei, coarse to vesicular chromatin, variable nucleoli, mitoses, and apoptotic debris. Two cases showed only low-grade papillary structures without high-grade tumor cells in the peritoneal washings and had only focal mild nuclear pleomorphism, while keratinization, dyskeratotic cells, or overt keratin pearl formation was present in two cases and correlated with abrupt keratinization and shadow-cell formation in tissue sections. The basaloid component showed aberrant nuclear/cytoplasmic β-catenin and diffuse CDX2 expression, with loss of ER, PR, and PAX8. All tumors were mismatch repair proficient, p53 wild type, microsatellite stable, POLE/TP53 wild type, and harbored CTNNB1 exon 3 mutations. Three patients developed recurrent or metastatic disease. PiMHEC can show recognizable cytologic clues in serous fluid specimens. A high-grade basaloid/discohesive carcinoma with peripheral palisading, brisk apoptosis or mitotic activity, and keratinization/keratin pearl formation should prompt consideration of PiMHEC and correlation with histology, β-catenin/CDX2 immunophenotype, and CTNNB1 status.