• 
• 

Three Part Question

In [adult patient suspected to have had a seizure], should [serum prolactin level] be [measured to help differentiate if he had a seizure or not] ?

Clinical Scenario

A 55 years old man brought by ambulance for lost of consciousness without witness. Discover by his wife 5 minutes later, he woke up but still have confusion 20 minutes later in your emergency room. He is not known for any disease and doesn’t take any pill. You wonder if he had a seizure and you heard about a prolactin test that could help you with your diagnosis.

Search Strategy

A bestbet on the same topic was published in 2011, but did not include an important prospective study published later in the same year.
The website clinicaltrials.gov was searched for an ongoing trial on the topic. None was found.
MEDLINE (PUBMED) :
-("Seizures"[Mesh]) OR epilepsy Or convulsion AND "Prolactin"[Mesh] = 225 articles
- After filters : abstract available = 142
- After abstract review = 4 relevant papers
EMBASE :
- Seizure AND prolactin = 176 articles
- After filters : clinical trial, controlled clinical trial, controlled study, prospective stydy, ramdomized controlled trial, systematic review, english, human = 44
- After abstract review = 8 relevant papers (including preceding relevant papers)

Search Outcome

Relevant papers about adults = 4

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
S Ahmad, M W Beckett 2004 United Kingdom	3 cohort studies met the criteria: Laxer 1985 Trimble 1978 Anzola 1993 (total 102 patients)	meta analysis	Serum prolactin level (Prolactin test: 3 x baseline) in patients who suffered from seizure vs pseudo seizure for the diagnosis of seizure	Serum prolactin appears to be released at the onset of a generalised tonico-clonic seizure reaching a peak at 15-25 min ad 1h. Pooled sensitivity: 0.62 (IC 95% o.40-0.83) LR+ : 8.92 (IC 95% 1.31-60.91) Pooled specificity: 0.89 (IC 95% 0.60-0.98) LR- : 0.44 (IC 95% 0.22-0.91)	2/3 studies did an EEG as a diagnosis test gold standard. Small sample of patients (51, 14, 37). Very short description of studies included.
Suleyman Aydin et al. 2011	50 epileptic male of 20 years old 20 patients with psychogenic event 20 controls matched patients Intervention : Medications (carbamazepine or valproic acid) were stopped for 48 hours. Blood sample was drawn at 5 minutes, 1, 24 and 48 hours after the seizure and once in control subject	Prospective cohort study	Serum PRL level (and ghrelin and nesftatin-1) in patients who had a seizure vs healthy control patients.	In patients diagnosed with secondary generalized seizure and primary generalized seizure serum PRL concentrations were highest within 5 minutes and decreased rapidly after 1 hour compared to control9 (IC 95% 8-10) vs 1.8 (IC 95% 1.7-1.9).	Study population different from the emergency department population. No comparison with other causes of syncope. Sensibility/specificity/LR of PRL blood level are not reported Serum PRL level were drawn only once from control subject.
David K. Chen et al. 2005	1. One class I study and nine class II studies 2. Two class II studies (57 patients) Intervention : 1. Serum PRL measured 10-20 min post seizure. EEG was used as the gold standard. 2. Serum PRL during Head-up-tilt-test on patient at risk of syncope or non at risk	Systematic review	1. Serum PRL level to differentiate generalized seizures from psychogenic lost of consciousness	1. Pooled sensitivity ranged from 46.1 to 60%Pooled specificity ranged from 95.9-96.3%. PPV >99% but low NPV. PRL is useful whenmeasured at 10 minutes to differentiate generalized seizures from psychogenic lost of consciousness .	Small sample study Only 1 study class I evaluated Criterion for elevated PRL are variable
			2. Serum PRL level to differentiate generalized seizures from syncopes	2. 25 syncopals subjects showed a significant elevation of twice the baseline of PRL level compared to nonsyncopal patients. Prolactin is not useful to differentiate a seizure from a syncope.
Vipin Mishra et al. 1990	3 groups : - 35 patients with true seizures (15 generalized tonico-clonic seizures (GTC), 11 complex partial seizure (CPS) and 9 simple partial) - 20 cases of pseudoseizures - 22 matched healthy persons Intervention: Serum PRL at 15-20 minutes post seizure and at 6-8 hours post. Serum PRL were drawn once from healthy subject.	Prospective cohort study	Serum PRL level to differentiate GTC seizures from pseudoseizures.	15-20 minutes post seizure, there is a significant rise of PRL in GTC seizures and in CPS compared to pseudoseizures, healthy subject and simple partial seizure (p<0.001). At 6-8 hours post, there is no significant rise between groups.	Small sample study; Very brief details about how the study were conducted;

Comment(s)

Since the last bestbets on the same subject, a good prospective study was published. This study change the way we can answer this question now that we have better evidences. A commun weakness of these studies is that they do not include other causes of altered level of consciousness. Although it is a major concern in the emergency room, only the above meta-analysis includes other causes of altered level of consciousness when compared with the seizure group. Another weakness of those studies is they lack to report any effect on clinical decisions with the result of the prolactin level.

Clinical Bottom Line

The serum prolactin level is useful to distinguish a seizure from a pseudo seizure. The measurement must be done within one hour of the episode. Evidences are lacking if other organic causes of altered level of consciousness are suspected.

References

1. Ahmad S, Beckett MW Value of serum prolactin in the management of syncope Emergency medicine journal 2004 Mar;21(2)
2. Aydin S, Dag E, Ozkan Y, Arslan O, Koc G, Bek S Time-dependent changes in the serum levels of prolactin, nesfatin-1 and ghrelin as a marker of epileptic attacks young male patients Peptides 2011 Jun;32(6):1276-80
3. Chen DK, So YT, Fisher RS Therapeutics, Technology Assessment Subcommittee of the American Academy of N. Use of serum prolactin in diagnosing epileptic seizures: report of the Therapeutics and Technology Assessment Subcommitte Neurology 2005 Sep 13;65(5):668-75
4. Mishra V1, Gahlaut DS, Kumar S, Mathur GP, Agnihotri SS, Gupta V. Value of serum prolactin in differentiating epilepsy from pseudoseizure. J Assoc Physicians India. 1990 Nov;38(11):846-7.