Impact of seizure duration in maintenance electroconvulsive therapy
*Jan Di Pauli, MD, Regional Hospital Rankweil, Department of Psychiatry I, Valdunastraße 16, A-6830 Rankweil, Austria. Email: firstname.lastname@example.org
Acute electroconvulsive therapy has anticonvulsive effects. The aim of the present study was to investigate the anticonvulsive efficiency of maintenance electroconvulsive therapy. Records of patients treated with maintenance electroconvulsive therapy were screened retrospectively, and the changes in seizure duration were measured. The patients were subdivided into responders and non-responders. Responders had no significant seizure duration changes within the first week of maintenance electroconvulsive therapy. In contrast, the seizure duration of the non-responder group increased significantly within the first week. It was concluded that the early increase in seizure duration at constant energy could be predictive of relapse.
IN ELECTROCONVULSIVE THERAPY (ECT) a suprathreshold stimulus must be given to achieve a therapeutic seizure.
Because of intra-individual variation in seizure threshold, the titration technique can be used to determine the patients' individual seizure threshold. Treatment effects and side-effects are related to the given energy.1 The usual practice is to titrate to seizure threshold at the first ECT session and to stimulate at 250% of the seizure threshold.2 ECT is given two or three times a week. In this setting ECT has anticonvulsive efficiency. This is reflected by a rise in seizure threshold and a decrease in seizure duration. If ECT treatment is discontinued seizure threshold decreases again within 6 months to baseline.3
Often maintenance electroconvulsive therapy (MECT) is given after a successful series of acute ECT.4 One schedule of MECT involves the prolonging of treatment intervals from once a week to once in a month.5
Little is known about changes of seizure threshold in MECT, therefore the aim of the present study was to investigate the change of seizure duration as a function of seizure threshold in MECT.
The records of patients undergoing MECT in the year 1999/2000 were screened retrospectively. Twenty-two patients were evaluated (13 female). Considering the clinical outcome, the patients were subdivided into responders versus non-responders to MECT; the responders included nine patients undergoing monthly MECT, the non-responders 10 patients who discontinued the weekly MECT because of relapse; the remaining three patients stopped the biweekly therapy on their own decision. Nine patients had a diagnosis of major depression and, of these, three were responders, four were non-responders, and two discontinued the biweekly ECT.
Two responders suffered from a bipolar disorder, three patients from a schizoaffective disorder (one responder, two non-responders) and eight patients had a diagnosis of schizophrenia (four responders, three non-responders and one discontinuation of the MECT).
Age, gender, diagnoses, handiness, stimulation side, medication, anaesthetics, stimulation charge and number of ECT sessions were documented. The ECT device was a Mecta SR (Mecta, Lake Oswego, OR, USA). To determine the seizure threshold the titration technique was used in the acute phase. We measured seizure duration recorded on electroencephalogram (EEG) and motor activities, assessed by the cuff technique. For the present study the seizure duration was evaluated at seven defined measuring points: at the last acute ECT and subsequently at the first and last weekly MECT, the first and last biweekly ECT and at the first monthly and the last MECT to study cut-off. The clinical outcome was assessed from information given on the patient record and the global clinical impression scale documented in the patient records.
The stimulus charge remained stable over the MECT period.
For statistical analysis we used SPSS Version 8 (SPSS, Chicago, IL, USA) and performed an anova with one between-subjects factor (responders vs non-responders) and Dunnett T (two-tailed) post hoc test. P < 0.05 was defined as significant.
All patients filled out an informed consent and the study was approved by the local Ethics Committee.
Between responders and non-responders a different pattern of seizure duration change could be observed. In the non-responder group the EEG duration increased significantly from the initial mean value of 33 ± 13.47 s to 47.9 ± 15.45 s within 1 week (d.f. = 2, F = 4.677, P = 0.02), while in the responder group the EEG duration slightly decreased, although not significantly, from the initial mean value of 50.22 ± 26.18 s to 47.56 ± 25.09 s.
The limitations of the present report are as follows. The study was retrospective, the number of patients was small and heterogeneous in gender, diagnosis, age and medication. Nonetheless the data suggest that in a course of MECT an increase of seizure duration at constant energy within the first week is associated with a negative prognostic factor. In general seizure duration has a non-linear relationship to seizure threshold.6 The relationship is an inverse u-curve, with a strong increase of seizure duration when stimulus dose increases above seizure threshold that reaches a peak and decreases with even higher doses. In acute ECT we adjusted the stimulus energy to ensure certain criteria of seizure quality such as seizure duration, tachycardia, suppression, synchronicity of spikes and slow waves and their amplitude.5 In MECT we worked with constant energy. Because it was assumed that stimulation was above seizure threshold and the seizure duration increase in MECT occurred under constant stimulus energy, we suggest that the seizure duration changes in the present study reflect changes in seizure threshold.
Sackeim emphasized that increase of seizure threshold, as an anticonvulsive effect, is correlated with good outcome.7 It could be assumed that the early decrease of seizure threshold, reflected as an increase in seizure duration, is responsible for relapse in MECT. The present results are in contrast to those of other studies that could not find a relationship between seizure threshold and outcome.8,9
Jarvis et al. found no increase in seizure duration, but in that study all efforts inclusive of pharmacological augmentation were made to maintain a seizure duration of at least 25 s, so the effect of increase could be distorted.10 Wild et al. were able to detect an increase of seizure duration in MECT within 60 days. Their study also suggested that the patients with an increase in seizure duration tend to relapse. Five of six patients relapsed. Four of them had an increase in seizure duration.11
Despite the aforedescribed limitations we emphasize that patients who have an early increase in seizure duration in MECT should receive particular attention.