The first step would be ruling out pregnancy with urine or serum estimation of human chorionic gonadotropins. Complete blood count, renal function, liver function test and urinalysis help exclude systemic disease. Endocrine evaluation should include thyroid function tests, prolactin, gonadotrophins and oestradiol. Prolactin estimation should not be performed immediately following breast examination. Borderline prolactin values should be reconfirmed by repeat testing because of the significant variability of prolactin levels across the day and its response to stressful stimuli. All patients with unexplained hyperprolactinaemia require MRI (magnetic resonance imaging). Serum prolactin above 2,000 mU/l is almost always associated with pituitary or sellar tumours and a prolactin >5,000 mU/l is virtually diagnostic of a macroprolactinoma.

When evaluating prolactin results the clinician needs to be aware of assay “interferences” due to the presence of macroprolactin and the “prolactin hook effect.” Macroprolactin is a complex of prolactin with IgG antibody found in about 1 % of normal individuals resulting in artificially elevated measurements of serum prolactin. Macroprolactin is a physiologically inactive complex and therefore not associated with clinical symptoms of hyperprolactinaemia. Re-assay of serum after polyethylene glycol treatment usually confirms the presence of macroprolactin. Large tumours producing exceedingly high concentrations of prolactin and this may saturate both the capture and signal antibodies in immunoassays. This is known as the “hook effect” and results from blockage of the capture antibody-prolactin-signal antibody “sandwich” formation, leading to misleadingly low prolactin values, very rarely within the normal range, in automated immunometric assays. Re-assay of diluted serum will resolve the problem leading to accurate determination of the prolactin level. Patients with galactorrhoea and apparently normal serum prolactin should have prolactin levels verified after sample dilution.

Whilst not in routine use, this can provide useful information in patients with mild to moderate elevations in serum prolactin that may be medication related or due to a microprolactinoma too small to be seen on MRI. The intravenous administration of a dopamine antagonist in normal subjects produces a marked rise in prolactin and only a mild increase in thyrotropin stimulating hormone (TSH). Patients with a microprolactinoma exhibit a blunted prolactin response and an exaggerated TSH response.

Pituitary imaging provides information about the size of the tumour and its relationship to the optic pathways; it is also useful in the follow up of patients with larger tumours. Imaging is best performed using MRI of the pituitary and hypothalamus with gadolinium contrast enhancement. For patients unable to tolerate or too large to fit in the MRI tunnel, computed tomography scanning of the pituitary can be useful but provides inferior resolution of the pituitary.

Indications for MRI pituitary would include:

Formal visual field testing using Goldmann perimetry should be performed if the adenoma is touching or in close proximity to optic structures.