Melengestrol acetate-induced exuberant endometrial decidualization in Goeldi's marmosets (Callimico goeldii) and squirrel monkeys (Saimiri sciureus)

Gross appearance, histopathology, electron microscopy, immunohistochemistry, and bacteriology of uteri from 17 melengestrol acetate (MGA)-implanted Goeldi's marmosets (Callimico goeldii), three MGA-implanted squirrel monkeys (Saimiri sciureus), and four nonimplanted Goeldi's marmosets were evaluated. The marmosets were first implanted at 1-7 yr of age with 88-267 mg/kg MGA, and the squirrel monkeys were implanted at 5-12 yr of age with 109-376 mg/kg MGA. Implants remained in marmosets for 2-7 yr and in squirrel monkeys for approximately 3.5 yr. Three MGA-implanted marmosets were ovariohysterectomized because of clinical signs related to uterine lesions. Other samples were obtained via laparotomy or postmortem. In all MGA-implanted marmosets, the uteri were grossly enlarged, firm, and distended with desiccated, friable material. Histologically, uterine lumina contained massive amounts of necrotic debris merging peripherally with proliferative endometrial stromal cells. The endometrium consisted of multinodular crowded sheets, cords, and papillary projections of large stromal cells with proliferations often thicker than the distended myometrium and with obliteration of glands and normal architecture. The stromal cells had large, irregular, vesiculated nuclei with moderate numbers of often bizarre mitotic figures, abundant and vacuolated cytoplasm, often indistinct cytoplasmic borders, and occasional intercellular bridges. There was random invasion of underlying myometrium with occasional remote, deep, isolated nests of cells. Lesions were often accompanied by severe infiltrates of primarily neutrophils. Biopsies of all nonimplanted marmosets were histologically normal. Immunohistochemical examination revealed multifocal immunoreactivity for vimentin within proliferative stroma and no reactivity for cytokeratin and desmin. The proliferative cells ultrastructurally had features consistent with decidualized endometrial stroma. Aerobic bacterial cultures of eight implanted marmosets were negative. All implanted squirrel monkeys had similar lesions but with less necrotic debris and suppuration and only slight invasion. The lesion was termed exuberant decidualization, and we concluded that MGA implantation caused the lesion in the marmosets and probably caused the similar lesion in the squirrel monkeys. The use of MGA for long-term contraception in Goeldi's marmosets should be discouraged because it will almost certainly result in exuberant decidualization and may result in clinically apparent disease. The use of MGA in squirrel monkeys should also be discouraged, and any implanted animals should be closely monitored. The development of nonsteroidal birth control measures should be pursued.