Yan-Qiu Wang1, Yi Wen1, Ming-Min Wang1, Yi-Wei Zhang1, Zhi-Xu Fang2. 1. Department of Pediatrics, The First People's Hospital of Chongqing Liang Jiang New Area, Chongqing, 401121, China. 2. Department of Neurology, Children's Hospital of Chongqing Medical University, Chongqing, 400014, China; Ministry of Education Key Laboratory of Child Development and Disorders, Chongqing, 400014, China; National Clinical Research Center for Child Health and Disorders, Chongqing, 400014, China; China International Science and Technology Cooperation Base of Child Development and Critical Disorders, China; Chongqing Key Laboratory of Pediatrics, China. Electronic address: fangsteve1992@163.com.
Abstract
OBJECTIVE: Psychogenic non-epileptic seizures (PNES) are conversion disorders with functional neurological symptoms that can resemble epileptic seizures (ES). We conducted a systematic review to obtain an overview of the value of prolactin (PRL) levels in the differential diagnosis between PNES and ES. METHODS: We searched PubMed, EMBASE, and Cochrane Library databases for studies published up to June 4th, 2020. Published studies were included if they fulfilled the following criteria: original research on PRL changes after ES and PNES. By applying Bayes' theorem, we calculated the predicted values of PRL with pretest probabilities of 90 % and 75 % in ES. RESULTS: Sixteen studies were included in this review. All the studies showed that PRL levels increase after ES, especially 10-20 min after ES, when the elevation was most obvious. In studies where capillary PRL level measurements were included, the median sensitivity in the diagnosis of ES (all epileptic seizure types), generalized tonic clonic seizures (GTCS), focal impaired awareness seizures (FIAS), and focal aware seizures (FAS) was 67.3 %, 66.7 %, 33.9 %, and 11.1 %, respectively. The median specificity in the diagnosis of ES was 99.1 %. By using Bayes' theorem, when we used the median specificity and sensitivity for predictive value calculation, assuming a pretest probability of 90 %, a positive PRL measure was highly predictive (99 %) of all types of ES, and negative predictive values were all below 30 %. When we used the lowest specificity and sensitivity for predictive value calculation, assuming a pretest probability of 75 %, ES and GTCS had positive predictive values of 77.2 % and 81.0 %, respectively; the negative predictive values of PRL in ES and GTCS were 26.2 % and 29.6 %, respectively. CONCLUSIONS: The use of PRL could be a useful adjunct to differentiate GTCS from PNES. However, PRL levels are of limited use for differentiating FIAS or FAS from PNES, and a negative PRL measure is not predictive of PNES.
OBJECTIVE: Psychogenic non-epileptic seizures (PNES) are conversion disorders with functional neurological symptoms that can resemble epileptic seizures (ES). We conducted a systematic review to obtain an overview of the value of prolactin (PRL) levels in the differential diagnosis between PNES and ES. METHODS: We searched PubMed, EMBASE, and Cochrane Library databases for studies published up to June 4th, 2020. Published studies were included if they fulfilled the following criteria: original research on PRL changes after ES and PNES. By applying Bayes' theorem, we calculated the predicted values of PRL with pretest probabilities of 90 % and 75 % in ES. RESULTS: Sixteen studies were included in this review. All the studies showed that PRL levels increase after ES, especially 10-20 min after ES, when the elevation was most obvious. In studies where capillary PRL level measurements were included, the median sensitivity in the diagnosis of ES (all epileptic seizure types), generalized tonic clonic seizures (GTCS), focal impaired awareness seizures (FIAS), and focal aware seizures (FAS) was 67.3 %, 66.7 %, 33.9 %, and 11.1 %, respectively. The median specificity in the diagnosis of ES was 99.1 %. By using Bayes' theorem, when we used the median specificity and sensitivity for predictive value calculation, assuming a pretest probability of 90 %, a positive PRL measure was highly predictive (99 %) of all types of ES, and negative predictive values were all below 30 %. When we used the lowest specificity and sensitivity for predictive value calculation, assuming a pretest probability of 75 %, ES and GTCS had positive predictive values of 77.2 % and 81.0 %, respectively; the negative predictive values of PRL in ES and GTCS were 26.2 % and 29.6 %, respectively. CONCLUSIONS: The use of PRL could be a useful adjunct to differentiate GTCS from PNES. However, PRL levels are of limited use for differentiating FIAS or FAS from PNES, and a negative PRL measure is not predictive of PNES.