PD-L1 is a biomarker that may predict the response to antiPD-1/PD-L1 immunotherapy. We evaluated the expression of PD-L1 in carcinoma cells and immune cells across histopathological and TCGA molecular subgroups of endometrial carcinoma.Our study included 842 patients with endometrial carcinoma. Direct sequencing of polymerase epsilon (POLE) exonuclease domain hot spots and conventional immunohistochemistry (MLH1, PMS2, MSH2, MSH6, p53) were conducted to identify TCGA classification-based molecular subgroups of endometrial carcinoma: POLE-mutated, mismatch repair (MMR) deficient, no specific molecular profile and p53-aberrant. Multiplex immunohistochemistry was performed to evaluate PD-L1 expression in carcinoma cells (Ca) and tumor-infiltrating immune cells (ICs). PD-L1 expression in carcinoma cells and in ICs was detected in 8.6% and 27.7% of the cases, respectively. Combined positive score (CPS) was ≥1% in 19.4% of the samples.PD-L1 positivity in carcinoma cells, ICs and CPS correlated with tumor T cell density (TILs, p<0.001). POLE-mutated and MMR-deficient tumors were more likely to present PD-L1 expressing ICs, CPS positivity and abundant TILs compared with other TCGA subgroups (p<0.001). No differences existed in Ca-PD-L1 expression (p=0.366). Within various histotypes, non-endometrioid carcinomas displayed the highest Ca-PD-L1, ICs and CPS (p<0.03). Advanced cancers showed more frequent Ca-PD-L1 positivity (p=0.016), CPS (p=0.029) and IC≥1% (p=0.037) positivity compared to early disease.In conclusion, PD-L1 expression profiles differ between molecular subclasses, histological subtypes and disease stage of endometrial carcinoma. Prospective studies are needed to explore the predictive value of various PD-L1 scoring systems within the subgroups of endometrial cancer. CPS presents methodological advantages over cell-type specific scoring systems.

BACKGROUND: Uterine leiomyomas (ULs) are the most common gynecologic tumors and affect 3 of every 4 women by the age of 50 years. The majority of ULs are classified as conventional tumors, whereas 10% represent various histopathological subtypes with features that mimic malignancy. These subtypes include cellular and mitotically active ULs and ULs with bizarre nuclei. Uterine leiomyosarcoma (ULMS), the malignant counterpart of UL, is an aggressive cancer with poor overall survival. The early diagnosis and preoperative differentiation of ULMS from UL are often challenging because their symptoms and morphology resemble one another. Recent studies have shown frequent loss of alpha-thalassemia/mental retardation syndrome X-linked (ATRX) or death domain-associated protein (DAXX) expression in ULMS, and this is often associated with an alternative lengthening of telomeres (ALT) phenotype. METHODS: To investigate ATRX and DAXX expression and the presence of ALT in UL subtypes, immunohistochemical and telomere-specific fluorescence in situ hybridization analyses were performed. The study material consisted of 142 formalin-fixed, paraffin-embedded tissue samples representing various UL subtypes and 64 conventional ULs. RESULTS: A loss of ATRX or DAXX and/or ALT was detected in 6.3% of the histopathological UL subtype samples (9 of 142). Two patients whose ULs showed either ATRX loss or ALT were later diagnosed with a pulmonary smooth muscle tumor. Pulmonary tumors displayed molecular alterations found in the corresponding uterine tumors, which indicated metastasis to the lungs. All conventional ULs displayed normal ATRX, DAXX, and telomeres. CONCLUSIONS: These results highlight the differences between conventional and histopathologically atypical ULs and indicate that some UL subtype tumors may harbor long-term malignant potential. Cancer 2018;124:4650-4656.

These data suggest that mutations in the TM or EPCR genes are not a major cause of RM, although they may exert a modifier effect in combination with other variants.

Hereditary leiomyomatosis and renal cell cancer (HLRCC) is a tumor predisposition syndrome caused by germline fumarate hydratase (FH) mutations and characterized by uterine and cutaneous leiomyomas and renal cell cancer. Currently, there is no generally approved method to differentiate FH-deficient uterine leiomyomas from other leiomyomas. Here, we analyzed 3 antibodies (S-(2-succino)-cysteine [2SC], aldo-keto reductase family 1, member B10 [AKR1B10], and FH) as potential biomarkers. The study consisted of 2 sample series. The first series included 155 formalin-fixed paraffin-embedded uterine leiomyomas, of which 90 were from HLRCC patients and 65 were sporadic. The second series included 1590 unselected fresh frozen leiomyomas. Twenty-seven tumors were from known HLRCC patients, while the FH status for the remaining 1563 tumors has been determined by copy number analysis and Sanger sequencing revealing 45 tumors with monoallelic (n = 33) or biallelic (n = 12) FH loss. Altogether 197 samples were included in immunohistochemical analyses: all 155 samples from series 1 and 42 available corresponding formalin-fixed paraffin-embedded samples from series 2 (15 tumors with monoallelic and 7 with biallelic FH loss, 20 with no FH deletion). Results show that 2SC performed best with 100% sensitivity and specificity. Scoring was straightforward with unambiguously positive or negative results. AKR1B10 identified most tumors accurately with 100% sensitivity and 99% specificity. FH was 100% specific but showed slightly reduced 91% sensitivity. Both FH and AKR1B10 displayed also intermediate staining intensities. We suggest that when patient's medical history and/or histopathologic tumor characteristics indicate potential FH-deficiency, the tumor's FH status is determined by 2SC staining. When aberrant staining is observed, the patient can be directed to genetic counseling and mutation screening.