• bipolar disorder
• catatonia
• electroconvulsive therapy
• mood episodes
• mood stabilizer

Objective: We prospectively evaluated the short-term outcome and the predictors of response to electroconvulsive therapy (ECT) in a large sample of patients with a bipolar disorder in different phases of the illness.
Method: From January 2006 to May 2011, we performed an analysis using data obtained from 500 of 522 consecutive patients with Bipolar Disorder according to DSM-IV-TR diagnostic criteria, who were treated with ECT at the Department of Psychiatry of the University of Pisa. All patients were evaluated prior to and after the ECT course using the Hamilton Depression Rating Scale-17 (HDRS-17), Young Mania Rating Scale (YMRS), Brief Psychiatric Rating Scale (BPRS), Clinical Global Improvement Impression (CGI) scale and Bush-Francis Catatonia Rating Scale (BFCRS). The CGI subscale “global improvement”, final HDRS-17 and YMRS total scores were used to identify non-responder, responder, and remitters in depressive and mixed patients. Outcome in catatonic patients was evaluated with CGI-I and BFCRS final scores. Descriptive analysis were conducted in manic patients, due to the small size of the sample.
Results: At the end of the ECT course, 94 out of 295 depressed patients (31.86%) were considered non-responders, 103 patients (34.92%) responders, and 98 patients (33.22%) remitters. Among 197 mixed patients, 55 patients (27.9%) were defined non-responders, 82 patients (41.6%) were responders, and 60 patients (30.5%) were remitters. At the end of the ECT course, 21 out of 26 catatonic patients (80.8%) were classified as responders (GCI ≤ 2) and 5 patients (19.2%) were classified as non-responders. As expected, at the end of the ECT trial, the CGI-S, HDRS-17, BPRS and BFCRS (for catatonic group) scores were significantly lower in remitters than in responders and non-responders in all groups. In mixed and depressive patients, the length of current episode and baseline YMRS total mean score were statistically significant predictors of non-response versus remission. In catatonic group, the mean number of previous mood episodes was significantly greater in responders than in non-responders.
Conclusions: The main finding of this study is that, in our sample of patients with severe and drug-resistant bipolar disorder, ECT appeared to be an effective and safe treatment for all phases of the illness. We observed high rates of response in bipolar depression, mania and mixed state, approximately two-thirds of the cases, with less than 30% of non-responders. In catatonic patients ECT was effective in more than 80% of the cases. With an appropriate monitoring the treatment was very safe with a very low incidence of adverse events. According to existing literature, our data demonstrates that ECT-induced mania is virtually non-existent and long-term mood destabilization, very unlikely. However, data on ECT in long-term prophylaxis are lacking and randomized controlled trials of continuation and maintenance ECT in BD are not available. Despite these limitations, ECT should be considered a mood-stabilizing treatment, clearly superior to pharmacological treatment considering its efficacy in acute phases and in relapse-recurrence prevention.