Fumarate Hydratase-Deficient Leiomyoma of the Uterine Corpus: Comparative Morphologic Analysis of Protein-Deficient Tumors With and Without Pathogenic Germline Fumarate Hydratase Gene Mutations

Abstract

Deficiency of fumarate hydratase (FH) protein expression in uterine corpus leiomyomas may be attributable to either germline or somatic mutations of the FH gene, the former being definitional for the hereditary leiomyomatosis and renal cell cancer syndrome. The authors assess whether, using previously reported FH-associated morphologic features, FH protein-deficient uterine corpus leiomyomas associated with a pathogenic germline mutations of the FH gene (group 1) are distinguishable from FH protein-deficient uterine corpus leiomyomas without such mutations (and whose FH protein loss is presumed to be attributable to somatic/epigenetic inactivation or other unknown phenomena: group 2). Groups 1 and 2 were compared regarding a variety of clinicopathologic features, including 7 core “FH-associated” tumoral morphologic features: staghorn vasculature; alveolar-type edema; bizarre nuclei; chain-like tumor nuclei; hyaline cytoplasmic globules; prominent nucleoli, intranuclear inclusions, and perinucleolar halos; and prominent eosinophilic/fibrillary cytoplasm. Among 2418 patients diagnosed with uterine corpus leiomyoma during the study period, FH-associated morphologic features were reported in 1.5% (37 patients), and FH immunohistochemistry was performed in 29 (1.19%). Fourteen (48.27%) of the 29 patients showed FH protein deficiency by immunohistochemistry. Twelve patients underwent germline testing, of which 8 (66.7%) were classified as group 1 and 4 (33.3%) as group 2. FH protein-deficient tumors were larger (10.44 vs 4.08 cm, P  =  0.01) and associated with younger patients (42.05 vs 47.97, P  =  0.004) than 370 randomly selected uterine leiomyoma controls. Groups 1 and 2 showed no significant differences in patient age and tumor size. In group 1 tumors, the FH-associated morphologic features were generally present diffusely; all group 1 tumors displayed ≥5 FH-associated features, whereas all group 2 tumors displayed <5 FH-associated features (means 6.5  ±  0.53 vs 3.5  ±  1.00, P < 0.001). Notably, eosinophilic/fibrillary cytoplasm and alveolar-type edema were each significantly more prevalent in group 1 tumors than group 2 tumors (P  =  0.018 for both). No single morphologic feature was found to be completely sensitive and specific in making the distinction between group 1 and 2 tumors. Our findings suggest that groups 1 and 2 are unlikely to be morphologically distinguishable by individual morphologic features. Whether there is a combination of features that can reliably make this distinction is unclear and will require additional studies with larger cohorts.