Is the FOXL2 gene mutation involved in the development of granulosa cell tumours in equines?

Introduction and aim: Granulosa-theca cell tumour (GTCT) is a significant cause of infertility in mares, representing for 85% of ovarian tumours and approximately 2.5% of all neoplasms in horses [1]. In women, granulosa cell tumours (GCT) comprise about 5% of ovarian neoplasms and 70% of ovarian sex cord-stromal tumours. Human GCTs are categorized into adult and juvenile types, distinguished by clinicopathologic features. The adult type, the most common, occurs in peri- or post-menopausal women, while the juvenile type, affects prepubertal girls and young women [2]. In the adult type, a somatic "missense" mutation in the FOXL2 gene (402C→G; C134W) is present in 97% of cases but is absent in the juvenile type [2]. Adult GCTs exhibit a variety of histological patterns from well-differentiated to poorly differentiated. Well-differentiated patterns include microfollicular, macrofollicular, trabecular, insular, solid tubular, and empty tubular. "Call-Exner bodies” are typical. Adult GCTs contain round to oval cells with characteristic "coffee-bean" nuclei, few mitotic figures, mild nuclear atypia, scant cytoplasm and often luteinized cells. Juvenile GCTs share the macroscopic appearance with the adult type but differ significantly histologically, exhibiting a follicular or solid diffuse pattern with larger luteinized granulosa cells containing hyperchromatic or markedly bizarre nuclei and high mitotic activity [3]. This study aimed to explore if this categorization is possible in equines. Methods: 12 cases of equine GTCT were obtained by unilateral ovariectomy, using a flank-assisted laparoscopic technique. Ovaries were classified based on their macroscopic and microscopic appearance. For histopathology specimens were fixed in formalin and embedded in paraffin blocks. Slides were evaluated by experienced veterinary and human pathologists. GTCTs were classified according to the most prevalent pattern [4]. Additionally, a molecular test for the FOXL2 gene mutation was performed. The DNA was extracted using a commercially available kit, from 12 samples of formalin-fixed paraffin-embedded (FFPE) tissue. PCR was used to amplify the DNA region of interest before Sanger sequencing. Results: The median age of the mares was 11.6 years (3-25 years). All mares presented infertility problems; 8/12 exhibited stallion-like behaviour, while the remaining mares showed anoestrous or abnormal oestrous behaviour. All cases showed unilateral ovarian enlargement and a hypotrophic contralateral ovary. Macroscopically, the neoplastic ovary had an ovoid or spherical shape with a smooth surface, enclosed by a fibrous capsule. The cut surface revealed a multicystic structure in 9 cases and a solid structure in 3 cases. Cystic structures contained serous or serohaematic material, surrounded by solid connective septa of varying thickness and a greyish-white or yellowish colour. Histological patterns included follicular (macrofollicular, microfollicular, or macro-microfollicular) in 6/12 cases, solid (insular, tubular, and diffuse) in 2/12 cases, and mixed in 4/12 cases. The cell population was characterized by neoplastic granulosa cells in all cases and luteinized cells in 12/12 cases. Additionally, 6/12 cases presented "Sertoli-like" elements and 12/12 cases exhibited a fibro-thecal component. Call-Exner bodies were present in only 1/12 cases. Mitotic activity was low in all cases. The FOXL2 amplification and sequencing demonstrated a missense point mutation at position C135W, with a cytosine for guanine, in only one case (8.3%). Discussion and conclusions: Comparing the morphological features observed in these cases with the two subtypes of the tumour in women, pathologists agreed that mare cases exhibit greater homology with the adult type of the huma