Endocrine Abstracts

Androgen excess in women is most commonly caused by polycystic ovary syndrome (PCOS), but sinister ovarian and adrenal pathology requiring immediate action needs to be excluded. Here we examined whether the severity of androgen excess indicates the likely underlying pathology in women evaluated for androgen excess.

We included all women undergoing assessment of serum DHEAS, androstenedione (D4) and testosterone (T) by liquid chromatography-tandem mass spectrometry at the University Hospital Birmingham from 2012-2016. All patients with at least one androgen increased above the reference range were phenotyped by clinical notes review, including ascertainment of an underlying diagnosis likely to cause androgen excess.

A complete serum androgen profile was available in 1205 women; at least one of the three androgens was increased above the sex- and age-specific reference range in 378 patients (31.4%; 303 pre-and 75 postmenopausal). Recorded diagnoses included: PCOS (n=293=76%); congenital adrenal hyperplasia (CAH; n=18=5%), adrenocortical carcinoma (ACC; n=15=4%), ovarian hyperthecosis (OHT; n=7=2%), Cushing’s disease (n=7=2%), ovarian tumour (n=2=0.5%), adrenocortical adenoma (n=2=0.5%), and unknown (n=22=6%). Increased androgens were divided into 3 concentration ranges according to severity. In premenopausal women, PCOS was the most likely diagnosis at any level of androgen excess (prevalence 75-97%), except at Level 3 T (>5 nmol/L) and Level 3 A4 excess (>16.5 nmol/L), where CAH and ACC were more prevalent. In postmenopausal women, PCOS was the commonest diagnosis only at Level 1 T excess; ACC and OH were the most prevalent diagnoses at Level 3 excess (T>5 nmol/L, D4>13 nmol/L, DHEAS>20 mmol/L).

Patterns and severity of hyperandrogenism predict the likelihood of non-PCOS pathology. CAH and ACC should be considered in Level 3 T or A4 excess in premenopausal women, and ACC and OHT in Levels 2 and 3 androgen excess in postmenopausal women. These data provide clinical guidance for the need of further work-up to identify non-PCOS causes of androgen excess.