Serum PRL levels are a useful adjunct in the diagnosis of epilepsy if one is knowledgeable of the physiologic and pharmacologic effects on PRL release. PRL levels may be misleading in the evaluation of nocturnal seizures because of physiologic surges of PRL release during sleep. Samples taken when the patient awakes, which are paired with recovery levels 1 hour later, will show a decline in serum PRL that may be misinterpreted as a postictal effect, rather than the physiological change it actually represents.
Serum PRL increases with virtually all GTCS, the great majority of CPS, and a smaller percentage of SPS. There is no change in serum PRL after myoclonic, absence, or akinetic seizures. Repeated seizures or status epilepticus usually do not produce serum PRL elevations.
NES produces increases in serum cortisol levels, but most studies have demonstrated no effect on PRL. Thus, a generalized convulsive event that does not produce a rise in serum PRL almost always represents a nonepileptic event. A rise in PRL after a staring spell, which potentially represents an absence, complex partial, or psychogenic nonepileptic seizure, almost certainly reflects a CPS.
Criteria used to determine a significant elevation of serum PRL have varied widely. We require that results for an acute sample obtained 10 to 20 min after the event exceed the normal age-related range and that it is 2.5- to 3-fold higher than the recovery sample obtained at least 1 hour after the event. We do not utilize serum PRL to evaluate nocturnal events, and we interpret with caution results obtained within 90 minutes of waking. The results of serum PRL are most reliably interpreted in conjunction with a description (or video recording) of the event and with the ictal EEG. Single photon emission tomography (SPECT) and positron emission tomography (PET) scans also may help validate the diagnosis.
A single study has addressed the issues of sensitivity and specificity for postictal hyperprolactinemia (56). Yerby and colleagues analyzed data from a dozen reports that addressed serum prolactin in relation to nonepileptic and various types of epileptic seizures, deriving a sensitivity of 63% and a specificity of 91% for epileptic seizures. The predictive value of significantly elevated postictal serum prolactin in the detection of NES depends on the prevalence of NES in a given population. The probability of epileptic seizure if postictal serum prolactin is elevated ranges from >99% if the prevalence of NES is 1%, versus 88% if the prevalence of NES is 50%. Tertiary care centers often find the prevalence of NES in the range of 10 to 25%, yielding a predictive value of 95 to 98% for postictal hyperpro-lactinemia as an indication of true epileptic seizures. In contrast, a negative serum prolactin result is not highly indicative of NES.
Was this article helpful?