logo SBA


Digital archive of theses discussed at the University of Pisa


Thesis etd-08082018-113950

Thesis type
Tesi di specializzazione (4 anni)
Thesis title
Course of study
relatore Prof. Perugi, Giulio
  • electroconvulsive therapy
  • bipolar disorder
  • mixed state
  • bipolar depression
  • long term follow-up
  • recurrence/relapse
Graduation session start date
Release date
Objective: The long-term stabilization is a major challenge in the treatment of bipolar disorder (BD) and, despite continual pharmacological treatment, a high rate of relapse/recurrence has been observed. ECT is a highly effective treatment for the acute phases of depressive, manic, and mixed episodes of bipolar disorder; less information is available on long-term relapse rate after successful ECT treatment.
In a naturalistic, prospective, and observational study, we evaluated the long-term outcome and the predictors of relapse/recurrence in a sample of ECT-responder bipolar patients.
Method: During a follow-up ranging from 26 to 160 weeks, we explored the rate and the characteristics of the depressive and mixed relapse/recurrence of 70 with Bipolar Disorder according to DSM-IV-TR diagnostic criteria. Patients were selected from a population of 475 patients who received ECT between December 2008 and December 2017 at the Department of Psychiatry of the University of Pisa. All patients involved in this study were responders to ECT. They were evaluated 1 day prior to ECT and a week after the treatment termination using the Clinical Global Impression Scale (CGI), the Hamilton Rating Scale for Depression-17 item (HAM-D-17), the Brief Psychiatric Rating Scale (BPRS), the Young Mania Rating Scale (YMRS) and the Mini Mental State examination (MMSE). Clinical Global Impression-Improvement (I-CGI) sub-scale was used for the definition of response. The Longitudinal Interval Follow-up Evaluation (LIFE), administered at least every 16 weeks, was used to assess time to relapse (defined as LIFE scores >/= 5 for at least 2 consecutive weeks or as the need of hospitalization), as well as periods of response and remission. Descriptive and comparative analyses were performed and survival analysis were used to identify if length of current episode, psychotic symptoms and comorbidity could affect the time of relapse/recurrence.
Results: The mean duration of follow-up was 57.04 +/- 27.2 weeks (range, 26-160 weeks). All patients were followed up for at least 6 months; 44 patients (62.9%) for at least one year. Sixty-five patients (92.8%) spent in response more than 90% of the time; thirty-two patients (45.7%) were in remission for more than 60% of the time. Patients with depressive index episode compared with patients with mixed index episode showed a significative difference in percentage of patients in remission (83.3% vs 61.8%; p = 0.043).
Relapse rate at one-year follow-up was 37.1% (n=26); 60% of total relapses were within the first 6 months of follow-up. The remaining 40% was distributed along the whole observation period. No manic switches or episodes occurred after the ECT course and during the follow-up.
No significant differences were found either in demographic, clinical and ECT characteristics or in the type and duration of psychopharmacological treatment between the patients who relapsed and the patients who did not. Twenty-four patients (34.3%) showed a depressive relapse. Psychotic symptoms were less frequent among the patients who fulfill relapse criteria [5 (20.8%) vs 21 (45.7%); p = 0.041]. The index episode mean duration was longer in the patients with depressive relapses (10.25 +/- 10.3 vs 6.6 +/- 5.0 months; p = 0.05). Eleven patients (15.7%) showed a mixed relapse at any time during the follow-up. Patient with mixed episode have more frequently mixed relapse than patients with depressive episode (29.4% vs 2.8%; p = 0.002). Patients whit mixed relapse showed a higher burden for anxiety comorbidity (81.8% vs 42.4%, p = 0.002) in general and panic disorder in particular (81.8% vs 39%; p = 0.009. Patients with mixed relapse reported a higher number of ECT during the follow-up than patients with depressive relapses. (63.6% vs 21.1%; p = 0.020).
There was an association between time to relapse and the lifetime panic comorbidity: patients with panic disorder relapsed in a shorter time. The 70,6% of the patients in this group relapsed within 26 weeks, whereas 53,8% of the patients without panic disorder did it in the same time [Log Rank (Mantel-Cox) chi-square = 0.05]. Neither the length of index episode, nor the presence of psychotic symptoms correlates with the time of relapse.
Conclusions: In our study, electroconvulsive therapy showed a positive impact on the clinical course of severe and treatment-resistant patients with bipolar disorder, as suggested by the high number of weeks spent in remission during the follow-up period. The duration of index episode was related to higher probability of depressive relapse, while psychotic symptoms were more frequent in patients who did not relapsed. As regard mixed relapses, lifetime panic disorder comorbidity and the mixed polarity of index episode may predict the probability of relapse. Further research on clinical predictors of early relapse is needed in order to identify therapeutic strategies to improve long-term outcomes in bipolar ECT-responder patients.