Is ovarian and adrenal venous catheterization and sampling helpful in the investigation of hyperandrogenic women?

Abstract

OBJECTIVE: To audit our practice of performing ovarian and adrenal venous catheterization and sampling in hyperandrogenic women who fail to suppress their elevated androgen levels following a 48-h low-dose dexamethasone suppression test (LDDST). We considered the technical success rate of catheterization, the extra information obtained in addition to the standard biochemical tests and imaging findings, and the impact of sampling on management decisions. DESIGN: A retrospective analysis of the results of all ovarian and adrenal venous catheterizations performed at St Bartholomew's Hospital, London, in the years 1980-1996. PATIENTS AND METHODS: Baseline ovarian and adrenal androgens were measured in all women presenting with symptoms and signs of hyperandrogenism. Those patients who failed to suppress their elevated testosterone (T), androstenedione (A4) and/or dehydroepiandrosterone-sulphate (DHEAS) levels following a LDDST to within the normal range or to less than 50% of the baseline value were investigated further with adrenal computed tomography (CT), ovarian ultrasound, and ovarian and adrenal venous catheterization and sampling. RESULTS: Results were available in 38 patients. The overall catheterization success rate was: all four veins in 27%, three veins in 65%, two veins in 87%. The success rate for each individual vein was: right adrenal vein (RAV) 50%, right ovarian vein (ROV) 42%, left adrenal vein (LAV) 87% and left ovarian vein (LOV) 73%. Eight patients were found to have tumours by means of imaging (adrenal CT and ovarian ultrasound), three adrenal and five ovarian, seven of which underwent operation. In six of these patients the clinical presentation was suggestive of the presence of a tumour; in addition, the combination of imaging findings allowed the detection of suspicious adrenal and ovarian masses in all eight cases. The five patients with ovarian tumours had serum testosterone levels > 4.5 nmol/l. In a further eight patients, laparotomy was performed based on a combination of diagnostic and therapeutic indications; in two of these patients the catheterization results were suggestive of an ovarian tumour. All these eight patients were shown histologically to have polycystic ovarian syndrome (PCOS), and no occult ovarian tumour was identified. None of the patients with nontumourous hyperandrogenism had a baseline testosterone level in excess of 7 nmol/l (median 4.4 nmol/l, range 2.5-7 nmol/l). CONCLUSIONS: Our results suggest that ovarian and adrenal venous catheterization and sampling should not be performed routinely in women presenting with symptoms and signs of hyperandrogenism, even if they fail to suppress their elevated androgen levels to a formal 48-h LDDST. All patients presenting with symptoms and signs of hyperandrogenism and elevated androgen levels, and where the suspicion of an androgen-secreting tumour is high, should have adrenal CT and ovarian ultrasound imaging to detect such a tumour. Venous catheterization and sampling should be reserved for patients in whom uncertainty remains, as the presence of a small ovarian tumour cannot be excluded on biochemical and imaging studies used in this series alone. Its use should be restricted to units with expertise in this area.