SAT-223 Insights into Ovarian Hyperandrogenism: Lessons from Two Unusual Cases

Abstract

Introduction: The majority of cases with postmenopausal hyperandrogenism have a benign etiology such as ovarian hyperthecosis, hilus cell hyperplasia, ovarian or adrenal tumors. Determining adrenal versus ovarian OR tumorous versus nontumorous causes of hyperandrogenism can be tricky. Here we report two very different cases associated with ovarian hyperthecosis. Case 1: The 73 year old obese female with a history of prediabetes, hysterectomy and breast cancer presented with severe hirsutism and virilization over 5 years. Testosterone levels per LC/MS were between 200-350 ng/dL and rose further during a 2d LDDST (Low Dose Dexamethasone Suppression Test), while cortisol and DHEA-S suppressed appropriately. Imaging was negative for an adrenal or ovarian mass. Based on the biochemical data, we suspected a tumorous ovarian source of androgen excess despite negative imaging findings. The patient underwent salpingo-oophorectomy which revealed small bilateral steroid cell tumors (0.7-0.9 cm) in the setting of bilateral focal hyperthecosis and hilus cell hyperplasia. Postoperative testosterone levels fell to 14 ng/dL. Case 2: The 53 year old obese female with a history of diabetes and PCOS presented with progressive male pattern hair loss, acne and acanthosis nigricans over 3-5 years. Her testosterone levels per LC/MS were between 100-150 ng/dL. During LDDST, cortisol and DHEA-S suppressed appropriately, testosterone increased further. Imaging was again negative for an adrenal or ovarian mass. Following this workup, the patient was treated with a GnRH analogue which resulted in complete testosterone suppression, regression of acanthosis nigricans and scalp hair regrowth. She later underwent salpingo-oophorectomy which showed only a few scattered, minute clusters of luteinizing theca cells in both ovaries, no tumor was identified. Lessons: These cases illustrate that imaging often fails to detect small ovarian tumors, stressing the need for additional diagnostic markers such as LDDST. Based on their case series, Kaltsas et al. suggested that lack of testosterone suppression during LDDST ruled out a nontumorous androgen source (1). However, this was not true for case 2 presented here: Testosterone remained non-suppressed during LDDST, yet she did not have an androgen-producing tumor, but mild hyperthecosis. Notably, her testosterone levels were <150 ng/dL. In summary, we propose that testosterone levels >150 ng/dL PLUS absent testosterone suppression during LDDST might be suggestive of androgen-producing tumors. Furthermore, therapy with GnRH can be an alternative to surgery in cases with ovarian hyperandrogenism.