Pituitary apoplexy is one of the most dramatic emergencies in endocrinology — it causes sudden severe headache (thunderclap onset, maximal at onset), often with diplopia (CN III, IV, VI — passing through the cavernous sinus immediately lateral to the pituitary), visual field defects, and profound adrenal insufficiency (the posterior pituitary is usually spared, but anterior pituitary destruction causes acute cortisol deficiency). The headache pattern can mimic subarachnoid haemorrhage. The key GP action is: immediate hydrocortisone 100 mg IM (acute adrenal crisis prevention) + 999. Any patient with a known pituitary adenoma who develops acute severe headache with or without visual symptoms should be treated as pituitary apoplexy until proved otherwise. The importance of prolactin sample conditions: prolactin is a stress-hormone that rises in response to venepuncture, exercise, physical examination (especially breast examination — prolactin released by nipple stimulation), and food. A single elevated prolactin result in a patient who was anxious, had a difficult venepuncture, had a breast examination beforehand, or ate immediately before the test may be a false positive. The standard is: two fasting morning samples with a resting period of 30 minutes before venepuncture (a dedicated 'prolactin canula' technique — insert a cannula, leave patient at rest for 30 minutes, then draw blood without further stress). Many apparent hyperprolactinaemia cases are artefactual.

The hypothyroidism-prolactin connection is one of the most commonly missed causes of apparent hyperprolactinaemia — TSH stimulates both TSH and prolactin release (via TRH — thyrotropin-releasing hormone), so primary hypothyroidism can cause clinically significant hyperprolactinaemia (prolactin typically <2000 mIU/L). Treatment of the hypothyroidism with levothyroxine normalises the prolactin completely. This is why TSH must always be checked before investigating hyperprolactinaemia further. Macroprolactinaemia is the most important cause of false-positive hyperprolactinaemia — it affects approximately 10–40% of patients referred for hyperprolactinaemia investigation and is caused by immunoglobulin binding to prolactin molecules, creating large biologically inactive complexes that are measured by immunoassay but cannot bind to prolactin receptors in tissues (therefore no clinical effect). The macroprolactin complex is too large to be filtered by the kidney and accumulates in the serum. A patient with macroprolactinaemia has a high measured prolactin but no symptoms (no amenorrhoea, no galactorrhoea, no infertility, no headache). The PEG precipitation test identifies macroprolactin — it should be requested for any mildly elevated prolactin (<2000 mIU/L) without clinical symptoms before arranging MRI pituitary.

The prolactin cannula technique is the gold standard for accurate prolactin measurement — a sample taken immediately after the stress of venepuncture will contain stress-released prolactin in addition to the baseline. The correct protocol: (1) insert cannula at the start of the consultation; (2) allow patient to sit quietly and rest for 20–30 minutes; (3) draw blood at the end of the rest period through the cannula without any new needle insertion. This eliminates the venepuncture stress response. Many GP practices do not follow this protocol — a single non-fasting sample taken at the time of a stressful consultation will frequently show falsely elevated prolactin. If this protocol is impractical, an alternative is to request the laboratory to perform a PEG precipitation test on any prolactin result between 400–2000 mIU/L before escalating to MRI. The prolactin level ranges that guide management: <400 mIU/L = normal; 400–1000 = borderline, likely physiological/pharmacological (repeat, check TSH, PEG test); 1000–2000 = significant, investigate (PEG, TSH, drug review, MRI if no cause); >2000 = strongly suggests prolactinoma (MRI mandatory); >5000 = macroprolactinoma likely (MRI + urgent endocrinology); >10,000 = almost certainly macroprolactinoma (endocrinology urgently).

The prolactin level at which macroprolactinoma vs stalk compression can be distinguished is clinically important — a prolactin result above 5000 mIU/L almost always indicates a prolactinoma (the pituitary cells secreting prolactin, not just compression releasing the inhibition on normal cells). Stalk compression by non-prolactin-secreting tumours typically produces prolactin levels below 3000 mIU/L. This matters for treatment: prolactinomas respond to cabergoline (which directly suppresses prolactin secretion from the tumour and causes tumour shrinkage); non-functioning adenomas with stalk compression do not respond to cabergoline and require surgical decompression. Getting the diagnosis right prevents both under-treatment (missing a cabergoline-responsive prolactinoma) and inappropriate treatment (giving cabergoline to a patient with a non-functioning adenoma requiring surgery).

The antipsychotic switching dilemma for hyperprolactinaemia deserves specific attention — antipsychotic-induced hyperprolactinaemia is common (approximately 40–70% of patients on first-generation antipsychotics), and the long-term consequences of sustained hyperprolactinaemia from antipsychotics include: osteoporosis (from hypogonadism), sexual dysfunction, infertility, and possible breast cancer risk (oestrogen-dependent; prolactin has growth-promoting effects on breast tissue). The evidence base: amisulpride and risperidone are the most potent prolactin-raising antipsychotics; aripiprazole is a partial D2 agonist that has negligible prolactin-raising effect and can even reduce prolactin when added to other antipsychotics (combination strategy without compromising antipsychotic efficacy). Quetiapine also has low prolactin-raising potential. GPs who identify patients on antipsychotics with symptomatic hyperprolactinaemia (amenorrhoea, galactorrhoea, sexual dysfunction, bone density concerns) should raise this with the prescribing psychiatrist — switching to aripiprazole or quetiapine is a clinically meaningful quality-of-life intervention that requires psychiatric team decision-making.

Cabergoline vs bromocriptine — the evidence clearly favours cabergoline for prolactinoma management: cabergoline normalises prolactin in approximately 90% vs 70% for bromocriptine, restores ovulation in 80% vs 60%, causes tumour shrinkage in 80–90% vs 60–70%, and has far better tolerability (once or twice weekly dosing vs BD/TDS for bromocriptine; much lower rate of nausea, postural hypotension, and gastrointestinal side effects). The cardiac valvulopathy risk from cabergoline is principally established at the much higher doses used for Parkinson's disease (10–20 mg/week) — at the lower doses used for prolactinoma (typically 0.5–2 mg/week), the risk is low but not zero, and echocardiographic monitoring is recommended by endocrinology guidelines after cumulative doses reach significant levels (approximately 3 years of treatment). Bromocriptine remains the preferred choice in women planning pregnancy because it has a much longer safety record in early pregnancy (established since the 1970s) — cabergoline is usually stopped when pregnancy is confirmed (in microprolactinoma patients) and restarted postpartum if required. The decision to stop or continue dopamine agonist during pregnancy in macroprolactinoma requires specialist input.

The fertility restoration rate with prolactin normalisation is one of medicine's most satisfying treatment outcomes — approximately 80% of women with hyperprolactinaemic amenorrhoea and infertility will resume spontaneous ovulation and achieve pregnancy within 12 months of prolactin normalisation with cabergoline. This means that for most women with hyperprolactinaemia-related infertility, the treatment is cabergoline alone, not IVF. The sequencing is important: normalise prolactin first and observe for 3 months before adding ovulation induction — many women need only the prolactin-normalising treatment and nothing further. Adding clomifene or gonadotrophins while prolactin is still elevated will be ineffective because elevated prolactin suppresses the hypothalamic-pituitary-ovarian axis at a level upstream of where these drugs act.

The cabergoline withdrawal trial (attempting to stop cabergoline after 2+ years of normalised prolactin and stable MRI) is an important management option that is underutilised — approximately 20–30% of microprolactinoma patients achieve sustained remission after cabergoline withdrawal (prolactin remains normal without treatment). The predictors of successful withdrawal include: normalised prolactin for at least 2 years on treatment, no visible tumour on MRI (or dramatic tumour size reduction), and low pre-treatment prolactin. NICE clinical guideline and Endocrine Society guidelines recommend discussing withdrawal after 2 years of stability. The protocol: reduce cabergoline dose gradually by 50% every 3 months while monitoring prolactin, aiming for eventual cessation. If prolactin rises to above normal on dose reduction → continue treatment. If prolactin remains normal off treatment at 3 and 6 months → remission confirmed. This approach prevents patients from taking a lifelong drug when they may not need it.

The macroprolactinaemia explanation to patients requires care — patients who have been told their prolactin is 'high' and then are told 'actually it's a form of prolactin that doesn't matter' often feel confused or dismissed. The clear explanation is: 'There are different forms of prolactin in your blood. Standard blood tests measure all of them together, but one form — a larger version — attaches to other proteins and can't actually reach the cells that would respond to it. You have more of this inactive form, which makes the total reading look high. But your body is working perfectly normally — this is a characteristic of your blood chemistry, not a disease, and you don't need any treatment or follow-up unless you develop symptoms.' This explanation validates the patient's concern, explains why the test result is not as alarming as it appears, and prevents unnecessary anxiety about a finding that requires no action.