How common are polycystic ovaries and the polycystic ovarian syndrome in women with Cushing's syndrome?

Abstract

Women with Cushing's syndrome (CS) may present with menstrual irregularity and symptoms/signs of hyperandrogenism, a phenotype similar to that of the polycystic ovarian syndrome (PCOS); however, currently there are no data on the prevalence of either polycystic ovaries (PCO) and/or PCOS in patients with CS. The aim of this study was to investigate their presence among women of reproductive age presenting with CS by analysing clinical, endocrinological and ultrasonographic features. Prospective study of all women within the reproductive age (range 18–40 years) who presented with CS between August 1994 and January 2000. Thirteen women (median age 32 years, range 18–39 years) with CS were evaluated. The diagnosis of CS was based on the presence of appropriate clinical features and an elevated serum midnight cortisol with failure to suppress 0900 hours serum cortisol to less than 50 nmol/l following a formal low-dose dexamethasone suppression test (LDDST). All women had their clinical features relevant to possible hyperandrogenism, menstrual disorder and infertility recorded, and circulating gonadotrophins, oestradiol, androgens and SHBG levels measured; ovarian ultrasonography was performed during their initial assessment. Relevant MR/CT imaging of the pituitary and/or adrenal glands was performed. Eleven women had ACTH-dependent CS [nine Cushing's disease (CD), one ectopic ACTH syndrome due to a bronchial carcinoid, one periodic CS of unknown origin); two patients had ACTH-independent CS (adrenal adenomas). All women with CS had at least one symptom/sign of hyperandrogenism (13 hirsutism, seven acne, five male-pattern alopecia). Nine women (70%) had menstrual disturbances (four oligomenorrhoea, four amenorrhoea, one polymenorrhoea) while four women (30%) had a normal menstrual pattern. Serum oestradiol levels for the group as a whole were similar to those observed in the early follicular phase of normally menstruating women; however, seven women had low oestradiol, LH and FSH levels suggestive of hypogonadotrophic hypogonadism. Serum androgen levels (testosterone, androstendione and DHEAS), even in the presence of symptoms/signs of hyperandrogenism, were within the normal reference range but SHBG levels were uniformly decreased even in women with normal menstrual cycles. There was a negative correlation between urinary free cortisol, but not mean serum cortisol, and serum oestradiol, testosterone and SHBG levels (r = − 0.8, r = − 0.86 and r = − 0.66, P < 0.02, P < 0.01 and P < 0.05, respectively), but not LH or FSH levels. Despite the fact that seven of these 13 patients lacked normal gonadotrophin stimulation, ovarian volumes of both ovaries were relatively preserved: right 7.3 ml, range 2.8–12.8 ml, and left 5.3 ml, range 2.3–13 ml. Women who were defined as oestrogen sufficient (E₂ > 140 pmol/l) had higher serum androstenedione, and lower urinary free cortisol levels, than women who were oestrogen deficient (E₂ < 140 pmol/l). Six of the 13 women (46%) had ovarian morphology suggestive of PCO, four of six oestrogen sufficient women and two of seven oestrogen deficient women. The results did not differ according to the underlying cause of CS. PCO and PCOS are common in women with Cushing's syndrome; women with Cushing's syndrome and only moderately elevated cortisol secretion maintain gonadotrophin stimulation to the ovary with normal oestradiol levels, in contrast to women with Cushing's syndrome and higher cortisol secretion who develop hypogonadotrophic hypogonadism. However, even in the latter group, high ovarian volumes were maintained and some had ovarian morphology suggestive of PCO.