PSYCHIATRY AND PRECLINICAL PSYCHIATRIC STUDIES - ORIGINAL ARTICLE
Protein S-100 and neuron-specific enolase serum levels remain unaffected by electroconvulsive therapy in patients with depression Laura Kranaster • Christoph Janke Sonani Mindt • Michael Neumaier • Alexander Sartorius
Received: 27 February 2014 / Accepted: 22 April 2014 Ó Springer-Verlag Wien 2014
Abstract The mechanism of the reversible cognitive deficits that might occur within an electroconvulsive therapy (ECT) treatment has not been clarified in a substantial way yet. Although the data available so far do not point towards a cause due to any structural or diffuse damage, further clarification, especially of the role of S-100 seems to be necessary before robust conclusions can be drawn. Serum levels of protein S-100 and neuron-specific enolase (NSE) were analysed in 19 patients with depression, who received ECT. The sampling was adjusted for the short half-life of protein S-100. Several outcome parameters such as Hamilton Depression Rating Scale and Minimental state examination before and after the ECT, response and remission to the treatment were recorded. S-100 and NSE levels at baseline, 30 and 60 min after the third session and after the end of the ECT remained stable. S-100 and NSE levels were neither associated with antidepressant response or remission nor with alterations in the cognitive performance. Although aiming for detecting
L. Kranaster and C. Janke contributed equally. L. Kranaster (&) A. Sartorius Department of Psychiatry and Psychotherapy, Central Institute of Mental Health, Medical Faculty Mannheim/Heidelberg University, Ruprecht-Karls-University Heidelberg, J5, 68159 Mannheim, Germany e-mail: [email protected] C. Janke Department of Anesthesiology and Critical Care Medicine, Medical Centre Mannheim, Mannheim, Germany S. Mindt M. Neumaier Institute for Clinical Chemistry, University Medical Centre Mannheim, Faculty of Medicine Mannheim, University of Heidelberg, Mannheim, Germany
potential rise in these established brain damage markers, an increase due to ECT was not observed, which is in line with the previous studies concerning the safety of ECT on a cellular basis. Keywords
ECT NSE S-100 Depression
Introduction Electroconvulsive therapy (ECT) is considered as highly effective treatment option for certain psychiatric disorders such as melancholic depression, treatment-resistant depression and mania and several forms of catatonia. Unfortunately, there still exists discomfort using this treatment by a significant proportion of psychiatrists and patients, who were offered this therapy (Lauber et al. 2005; Dauenhauer et al. 2011). A common reason for this reluctance is the possible occurrence of cognitive side effects that affect mostly memory functions. In severe cases the treatment must be even prematurely paused due to these side effects. Although it is well known that cognitive problems that occur within ECT are reversible (Semkovska and McLoughlin 2010), it is feared that they are induced by neuronal or glial damage due to ECT itself. In order to provide data on the safety of ECT, basically two different approaches were used in previous studies: postmortem studies and modern imaging techniques such as magnetic resonance and diffusion weighted imaging should identify structural damages from the ECT, whereas analysing proteins, which are supposed to indicate neuronal or glial damages such as protein S-100 and neuron-specific enolase (NSE) were prone to identify possible diffuse, small microstructural brain changes.
L. Kranaster et al.
Consistently, no definite signal abnormalities or structural damages were found in the postmortem and imaging studies (Coffey et al. 1991; Devanand et al. 1994; Szabo et al. 2007). Although the data from the previous protein studies also did not reveal any relevant change, the results should be interpreted with more cautions and less certainty. The first limitation in the previous studies is the small sample size ranging from 7 to 14 patients, which results in a higher likelihood of undetected differences before and after treatment (Berrouschot et al. 1997; Agelink et al. 2001; Arts et al. 2006; Palmio et al. 2010). There are other problems with the interpretation of the S-100 data, because S-100 contains a short estimated half-life between 30 (Ghanem et al. 2001) and 97 min (Townend et al. 2006). Unfortunately, this was not taken into account in most previous studies (Agelink et al. 2001; Stelzhammer et al. 2013). Besides being an established marker for glial damage in micromolar serum concentrations, S-100 is known to act as a trophic factor for serotoninergic neurons, and plays a role in axonal growth and synaptogenesis in nanomolar concentrations (Arolt et al. 2003; Busnello et al. 2006), which complicates the interpretation of S-100 data in patients treated with ECT, because an elevation of S-100 might indicate a certain glial damage or induced synaptogenesis as a part of the antidepressant pathway induced by the ECT. Another point worth mentioning is that some types of seizures have been reported to increase serum concentrations of brain biomarkers. For example, it has been shown very recently that the glial marker S-100 is increased in patients with seizures as compared to patients with sleep disorders. In contrast, NSE was reported not to be discriminative (Kacinski et al. 2012). This difference was interpreted as a potential secretion of S-100 rather than being the sign of a micro trauma. Also, CJD may be associated with increases in S-100 (Williams et al. 2011). In contrast, S-100 was reported to be unaltered in children with febrile seizures (Atici et al. 2012; Mikkonen et al. 2012). In sum, the mechanism of the reversible cognitive deficits that might occur within an ECT has not been clarified in a substantial way yet. However, the data available so far do not point towards a cause due to any structural or diffuse damage. In our opinion further clarification, especially on the complex role of S-100, not so much on NSE with a half-life of about 20 h (Ingebrigtsen and Romner 2003) and no proven neurotrophic effects is necessary before robust conclusions can be drawn (Ullrich et al. 2013). In this study we aimed to contribute to this question by taking blood samples before and after the treatment and adjusted to the estimated half-life time of S-100. Our main hypothesis was that ECT does not increase brain marker concentrations in the serum thereby, deliver no indication for tissue damage or brain barrier functional disturbance by the ECT.
Methods The present prospective study has been approved by the appropriate ethics committee, was carried out in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki) and written informed consent was obtained from all participants before inclusion in the study. Inclusion criteria were the diagnosis of a major depressive episode, age above 18 years and the clinical decision for an ECT. The following data were documented for each patient: age, sex, body mass index (BMI), Hamilton Depression Rating Scale (HDRS; 21 items version) and Mini-mental state examination (MMSE) before and after the ECT, age, uni- or bipolar disorder, existing clinical diagnosis of Alzheimer’s disease before ECT, response to treatment, defined as at least 50 % reduction of baseline HDRS and remission, defined as HDRS \8. All ECT sessions were performed in the Department of Psychiatry, and Psychotherapy, Central Institute of Mental Health, Mannheim, Germany during 2012 and 2013. A Thymatron IV device (Somatics, LLC. Lake Bluff, IL, USA) was used for ECT. Anesthetics were either thiopental or ketamine, chosen according to clinical experience or preferences of the anesthesiologist or the psychiatric supervisor. Seizure threshold in all patients was titrated during the first treatment, and energy was subsequently increased if patients did not respond clinically or if seizures were insufficient during the ECT course (i.e. usually motor response time \20 s and EEG seizure activity \30 s). Blood samples for determining S-100 and NSE were taken before the first ECT session started as baseline, 30 and 60 min after the third session for the measurement of acute effects and between 1 day and 1 week after the end of the ECT for the measurement of cumulative and chronic effects. Samples were centrifuged after the adequate clotting time and then stored at -80 °C until analysis. Serum S-100 and NSE assays were performed using a Cobas E 411 immunoassay technique (Roche Diagnostics, Basel, Switzerland). Statistics were performed using STATAÒ (StataCorp, Texas 77845, USA, version 11) at a significance level of 0.05 (two-tailed). Because not all the parameters were normally distributed, non-parametric methods were used: Friedman test for differences in S-100 and NSE at different time points, Wilcoxon-signed rank sum test for comparison of pre- and post-treatment characteristics, such as HDRS and MMSE and Spearman’s correlation and logistic regression to evaluate correlations between S-100 and NSE levels and other parameters. No corrections for multiple testing were made.
Protein S-100 and neuron-specific enolase serum levels remain unaffected
Results We included 19 patients (8 male, 11 female) with a mean age of 66.1 (SD 14.7; min 22, max 83) and mean BMI of 23.9 (SD 3.5) into this prospective study. All patients completed the study. Mean baseline HDRS was 28.7 (SD 6.6), whereas the mean HDRS after the ECT was 8.5 (SD 6.1) (p [ 0.001). We found a response rate of 89.5 % and a remission rate of 52.6 %. Mean baseline MMSE was 26.3 (SD 3.1) and mean MMSE after the treatment was 26.5 (SD 2.6) (p = 0.89). Other demographic and clinical characteristics are displayed in Table 1. Protein S-100 S-100 protein level at baseline was 0.053 ng/ml (SD 0.021), 0.055 ng/ml (SD 0.019) and 0.055 ng/ml (SD 0.023) 30 and 60 min, respectively after the third ECT session, and 0.056 ng/ml (SD 0.017) after the end of the ECT. There was no serum level difference between any of these points (Friedman = 2.4; p = 0.50). Baseline S-100 serum levels were not associated with demographic (age, sex, BMI) or clinical (uni-/bipolar, duration of illness or episode, comorbid dementia) features or psychopathological scores (baseline HDRS or MMSE), and it neither predicts antidepressant response or remission nor cognitive decline (defined as worse MMSE score after the treatment compared to the baseline MMSE score). A higher S-100 baseline level compared to the S-100 concentration, 30 min after the third ECT session was associated with the prior diagnosis of Alzheimer’s disease (Coef -175.4, SE 88.8, p = 0.048, 95 % CI -349.4 to -1.4). S-100 serum concentration after the completion of the ECT and the pre-/post-treatment concentration differences
Table 1 Demographic and clinical data of the 19 patients Age
Duration of current episode (in month)
Duration of illness (in years)
HDRS at baseline
HDRS after the treatment
MMSE at baseline
MMSE after the treatment
Cognitive worsening (%)
Number of ECT sessions
Response rate Remission rate
were neither associated with the number of total ECT sessions that were performed, state of response or remission, nor with D (pre-/post-treatment difference) HDRS or DMMSE, nor with cognitive decline or prior diagnosis of Alzheimer’s disease. Neuron-specific enolase serum NSE serum level at baseline was 14.5 ng/ml (SD 3.2), 14.2 ng/ml (SD 3.1) and 13.9 ng/ml (SD 5.0) 30 and 60 min, respectively, after the third ECT session and 13.8 ng/ml (SD 3.2) after the end of the ECT. There was no serum level difference between any of these points (Friedman = 3.1; p = 0.38). NSE concentration at baseline was negatively correlated with the baseline MMSE score (Spearman’s rho = -0.68; p = 0.005), but was not associated with the clinical diagnosis of dementia (p = 0.38) or with age (p = 0.52). Additionally, baseline NSE was neither associated with sex, BMI, type of depression (uni- or bipolar) nor with duration of illness or episode and severity of illness (baseline HDRS score). NSE at baseline also did not predict remission or response to ECT. The difference between the baseline and the end of treatment (D) NSE serum levels was negatively correlated with age (Spearman’s rho = -0.49; p = 0.04), but not with sex, the total numbers of ECT session within the treatment, state of response or remission after the ECT, DHDRS, DMMSE or cognitive decline or prior diagnosis of dementia (Alzheimer’s disease).
Discussion In this prospective study, serum levels of S-100 and NSE in patients with major depression, who were treated with effective ECT, were analysed in order to quantify the change in these proteins that are considered as ‘‘brain damage markers’’. No differences of serum concentrations of S-100 and NSE between baseline, 30 and 60 min after the third ECT session and after the completion of the treatment were found. We neither found any macromolar changes of S-100 during the ECT, which would have represented glial damage, nor a correlation of S-100 and the total number of ECT sessions. Our results on S-100 strengthen the already existing data that ECT produces no glial damage Agelink and colleagues did not find any change in serum S-100 concentration in patients suffering from therapy resistant major depression or schizodepressive psychosis that were treated with bifrontotemporal ECT (Agelink et al. 2001). In line with this, another group reported a similar negative finding in serum and cerebrospinal fluid (Zachrisson et al.
L. Kranaster et al.
2000). These results could be mainly replicated by Palmio et al., but this group described a small to moderate rise in S-100 levels in four out of ten patients, which was interpreted as glial cells activation as part in mediating the antidepressant effects of ECT (Palmio et al. 2010). A similar interpretation was given in a report about chronic electroconvulsive shock treatment in mice, where an increase of S-100 in the cerebrospinal fluid was found (Busnello et al. 2006). In fact, there is only one human study in which a small, but significant, increase in S-100 1 h post-ECT was reported (Arts et al. 2006). However, the increase was neither associated with the performance in cognitive tests nor with subjective cognitive impairment and the increase was only evident in female patients. Additionally, patients with prior diagnosis of Alzheimer’s disease showed a decrease of S-100 30 min after the third ECT session, which might point towards an involvement of S-100 in the known lack of response to antidepressant therapy in Alzheimer’s disease (Banerjee et al. 2011). However, due to size and composition of the sample our data are not suitable for further assumptions concerning S-100 and antidepressant actions in neurodegenerative disorders. As in all previous published reports (Berrouschot et al. 1997; Agelink et al. 2001; Palmio et al. 2010), we did not find any differences in NSE serum levels during the course of ECT either—regardless of the total number of ECT sessions—suggesting that there is indeed no relevant neuronal damage through this kind of treatment. Independent from the treatment itself, we found a clear association between higher NSE concentrations at baseline and lower baseline MMSE scores. This has not been described in the context of major depression yet, but might display at least a partial mechanism of cognitive impairment in depression or undetected dementia covered by the depressive episode. An important limitation of our ‘‘serum’’ approach is that the peripheral concentrations of S-100 and NSE—although they are established brain damage markers—might not correlate with the situation in the central nervous system of these proteins in an adequate manner (Casmiro et al. 2005; Kleindienst et al. 2010). Of course more direct measurements such as cerebrospinal fluid or in vivo analysis of specific brain regions would have been more accurate and valuable, however, hardly practicable in patients. In sum, the results of our prospective study corroborate most previous studies concerning the safety of ECT on a cellular basis, meaning that ECT warrants the neuronal and glial integrity and that cognitive side effects by ECT cannot be explained by a disruption of cellular structures. Our results have additional value because our sample has a moderate size, which should have found relevant differences with a higher probability than other studies with lower sample sizes. Additionally, our study has taken into
account that S-100 has a short half-life and by choosing adequate intervals between treatment and sampling we could further reduce the risk of false negative results. Another strength of our study is the variety of outcome parameters such as HDRS, MMSE, and response and remission rates.
References Agelink MW, Andrich J, Postert T, Wurzinger U, Zeit T, Klotz P, Przuntek H (2001) Relation between electroconvulsive therapy, cognitive side effects, neuron specific enolase, and protein S-100. J Neurol Neurosurg Psychiatry 71(3):394–396 Arolt V, Peters M, Erfurth A, Wiesmann M, Missler U, Rudolf S, Kirchner H et al (2003) S100B and response to treatment in major depression: a pilot study. Eur Neuropsychopharmacol 13(4):235–239 Arts B, Peters M, Ponds R, Honig A, Menheere P, van Os J (2006) S100 and impact of ECT on depression and cognition. J ECT 22(3):206–212 Atici Y, Alehan F, Sezer T, Tuygun N, Haberal A, Yazici AC, Karacan CD (2012) Serum S100B levels in children with simple febrile seizures. Seizure 21(3):175–177 Banerjee S, Hellier J, Dewey M, Romeo R, Ballard C, Baldwin R, Bentham P et al (2011) Sertraline or mirtazapine for depression in dementia (HTA-SADD): a randomised, multicentre, doubleblind, placebo-controlled trial. Lancet 378(9789):403–411 Berrouschot J, Rolle K, Kuhn HJ, Schneider D (1997) Serum neuronspecific enolase levels do not increase after electroconvulsive therapy. J Neurol Sci 150(2):173–176 Busnello JV, Leke R, Oses JP, Feier G, Bruch R, Quevedo J, Kapczinski F et al (2006) Acute and chronic electroconvulsive shock in rats: effects on peripheral markers of neuronal injury and glial activity. Life Sci 78(26):3013–3017 Casmiro M, Maitan S, De Pasquale F, Cova V, Scarpa E, Vignatelli L, Group NSES (2005) Cerebrospinal fluid and serum neuronspecific enolase concentrations in a normal population. Eur J Neurol 12(5):369–374 Coffey CE, Weiner RD, Djang WT, Figiel GS, Soady SA, Patterson LJ, Holt PD et al (1991) Brain anatomic effects of electroconvulsive therapy. A prospective magnetic resonance imaging study. Arch Gen Psychiatry 48(11):1013–1021 Dauenhauer LE, Chauhan P, Cohen BJ (2011) Factors that influence electroconvulsive therapy referrals: a statewide survey of psychiatrists. J ECT 27(3):232–235 Devanand DP, Dwork AJ, Hutchinson ER, Bolwig TG, Sackeim HA (1994) Does ECT alter brain structure? Am J Psychiatry 151(7):957–970 Ghanem G, Loir B, Morandini R, Sales F, Lienard D, Eggermont A, Lejeune F (2001) On the release and half-life of S100B protein in the peripheral blood of melanoma patients. Int J Cancer 94(4):586–590 Ingebrigtsen T, Romner B (2003) Biochemical serum markers for brain damage: a short review with emphasis on clinical utility in mild head injury. Restor Neurol Neurosci 21(3–4):171–176 Kacinski M, Budziszewska B, Lason W, Zajac A, Skowronek-Bala B, Leskiewicz M, Kubik A et al (2012) Level of S100B protein, neuron specific enolase, orexin A, adiponectin and insulin-like growth factor in serum of pediatric patients suffering from sleep disorders with or without epilepsy. Pharmacol Rep 64(6):1427–1433 Kleindienst A, Meissner S, Eyupoglu IY, Parsch H, Schmidt C, Buchfelder M (2010) Dynamics of S100B release into serum and
Protein S-100 and neuron-specific enolase serum levels remain unaffected cerebrospinal fluid following acute brain injury. Acta Neurochir Suppl 106:247–250 Lauber C, Nordt C, Falcato L, Rossler W (2005) Can a seizure help? The public’s attitude toward electroconvulsive therapy. Psychiatry Res 134(2):205–209 Mikkonen K, Pekkala N, Pokka T, Romner B, Uhari M, Rantala H (2012) S100B proteins in febrile seizures. Seizure 21(2):144–146 Palmio J, Huuhka M, Laine S, Huhtala H, Peltola J, Leinonen E, Suhonen J et al (2010) Electroconvulsive therapy and biomarkers of neuronal injury and plasticity: serum levels of neuron-specific enolase and S-100b protein. Psychiatry Res 177(1–2):97–100 Semkovska M, McLoughlin DM (2010) Objective cognitive performance associated with electroconvulsive therapy for depression: a systematic review and meta-analysis. Biol Psychiatry 68(6): 568–577 Stelzhammer V, Guest PC, Rothermundt M, Sondermann C, Michael N, Schwarz E, Rahmoune H et al (2013) Electroconvulsive therapy exerts mainly acute molecular changes in serum of major depressive disorder patients. Eur Neuropsychopharmacol 23(10): 1199–1207
Szabo K, Hirsch JG, Krause M, Ende G, Henn FA, Sartorius A, Gass A (2007) Diffusion weighted MRI in the early phase after electroconvulsive therapy. Neurol Res 29(3):256–259 Townend W, Dibble C, Abid K, Vail A, Sherwood R, Lecky F (2006) Rapid elimination of protein S-100B from serum after minor head trauma. J Neurotrauma 23(2):149–155 Ullrich H, Kranaster L, Sigges E, Andrich J, Sartorius A (2013) Neuron specific enolase and serum remain unaffected by ultra high frequency left prefrontal transcranial magnetic stimulation in patients with depression: a preliminary study. J Neural Transm 120(12):1733–1736 Williams R, Cresswell F, McClure M, Lane R (2011) Creutzfeldtjakob disease as a cause of cognitive decline and seizures in the elderly: diagnostic pointers and strategy for investigation. Case Rep Med 2011:719583 Zachrisson OC, Balldin J, Ekman R, Naesh O, Rosengren L, Agren H, Blennow K (2000) No evident neuronal damage after electroconvulsive therapy. Psychiatry Res 96(2):157–165
Studies in recent years have indicated that neuroimmunological events and immune activation may have a place in the etiology of depression. It has been suggested from data that there is a causal relationship between activation of the immune system an
Electroconvulsive therapy (ECT) is effective in treatment-resistant depression (TRD). It may act through intracellular process modulation, but its exact mechanism is still unknown. Animal research supports a neurotrophic effect for ECT. We aimed to i
Depressive disorder is frequently accompanied by changes in psychomotor activity and disturbances of the sleep-wake cycle. The chronobiological effects of electroconvulsive therapy (ECT) in patients with treatment-resistant depression (TRD) are large
Recent structural imaging studies have described hippocampal volume changes following electroconvulsive therapy (ECT). It has been proposed that serum brain-derived neurotrophic factor (sBDNF)-mediated neuroplasticity contributes critically to brain
To investigate whether total sleep deprivation (TSD) affects circulating concentrations of neuron-specific enolase (NSE) and S100 calcium binding protein B (S-100B) in humans. These factors are usually found in the cytoplasm of neurons and glia cells
Neuroactive steroids (NAS) are neuroactive molecules that have been shown to be associated with various psychiatric disorders. There are some inconclusive findings about the alteration in neuroactive steroid levels after the treatment of depression a
We examined potential changes in brain-derived neurotrophic factor (BDNF) serum levels and promoter methylation of the BDNF gene in 11 patients with treatment-resistant major depressive disorder during a series of electroconvulsive therapy (ECT). Blo