Headaches are often evaluated with an imaging study that may sometimes demonstrate a pituitary tumor, but there is not always a causal relationship. Therefore, the treatment of pituitary tumors does not always result in resolution of headaches, and headaches alone are not usually an indication for pituitary surgery. However, a change in headache pattern may be related to the enlargement of a pituitary tumor and the sudden development of a severe headache can be seen in the setting of pituitary apoplexy (sudden bleeding into a pituitary tumor), and this may be an indication for urgent surgery.
In contrast to the insidious, subacute development of headaches in most patients with pituitary tumors, patients with pituitary apoplexy may experience acute, severe headaches, perhaps associated with signs and symptoms of meningeal irritation (stiff neck, photophobia), CSF pleocytosis, or occulomotor paresis. Routine CT scans of the head occasionally skip the sella, hence the presence of blood or a mass within the sella may not be detected and patients can be misdiagnosed with meningitis or aneurysm. Because pituitary apoplexy represents a neurosurgical emergency, MRl should be used in patients with symptoms suggestive of this disorder. A subacute form of pituitary apoplexy has also been reported. Patients with subacute pituitary apoplexy experience severe and/or frequent headaches over weeks to months and have heme products within the sella on MRI scans.
In most instances, headaches are not attributable to direct effects of the pituitary tumor, and indirect causes must be considered. Generally, indirect effects of pituitary tumors are caused by reduced secretion of pituitary hormones, and are manifested by promotion of "vascular" headaches (e.g. migraine). The major exception to this rule relates to the potential for acromegalic patients to develop headaches secondary to cervical osteoarthritis. Vascular headaches may be exacerbated in association with disruption of normal menstrual cyclicity and impaired gonadal steroid secretion (e.g. from hyperprolactinemia or gonadotropin deficiency). Hyperprolactinemia, hypothyroidism and hyperthyroidism may also have direct effects independent of gonadal hormones. Headaches are common in acromegaly, and in the majority of cases the etiology is not well understood.
Bromocriptine or other dopamine agonists occasionally trigger severe headaches. When this occurs, it is important to recognize that bromocriptine has been reported as a cause of pituitary apoplexy, and it may be necessary to perform an MRI or CT to rule out infarction or hemorrhage within the pituitary. Once it is established that the patient is not infarcting the pituitary, it is generally safe to treat the headaches symptomatically (not with an ASA containing drug) and consider alternative therapies for the prolactinoma if the problem remains severe.
Pituitary tumor patients with vascular headaches are generally quite responsive to standard prophylactic migraine drugs (e.g. tricyclic antidepressants, verapamil, beta-blockers). It is best to begin therapy with very low-dose medication (e.g. 10 mg of amitriptyline at bedtime) and resist the impulse to escalate the dose rapidly to higher levels. Often patients have an excellent response to 10-30 mg of a tricyclic antidepressant, although it may take up to six or more weeks to reach the ultimate benefit. The choice of tricyclic antidepressant should be based upon the desired side effects (e.g. either more sedation or less sedation) Serotonin-selective antidepressants are generally less effective for headaches than tricyclics, although some patients do respond nicely to these agents. In some cases it may be necessary to use combination therapy (e.g. verapamil plus a tricyclic).
Although "abortive" therapies for headache such as cafergot or imitrex are probably effective in patients with pituitary tumors, I believe that it is prudent to avoid these drugs in patients with macroadenomas because of the potential for precipitating pituitary apoplexy.