[Are convulsions necessary for the antidepressive effect of electroconvulsive therapy: outcome of repeated transcranial magnetic stimulation].

Sismotherapy (ST) brings about numerous neurobiological changes, particularly changes in neuromediators and their receptors, second messengers, neuropeptides and neurotropic factors, a number of which are hypothesized to play a role in the pathophysiology or therapeutics of affective disorders (M. Fink). What is not yet known is which of these mechanisms is crucial for the psychotropic and anticonvulsant effects of ST. However, it is clear that the effects of ST tend to be relatively acute, and do not attack the deep-seated abnormalities that are the underlying causes of recurrences of affective disorders. This is corroborated by the fact that in animals, most of the effects of ECS on catecholamines and their receptors (and on receptors for benzodiazepines or neuropeptides such as TRH) tend to be relatively transient, and in most cases have been found to represent compensatory adaptations to the induced motor convulsions. However, recent preclinical data using attenuation, and clinical findings using reiterated transcranial magnetic stimulation (rTMS), suggest that it may not be necessary to provoke a clonic convulsion in order to achieve the beneficial psychotropic and anticonvulsant effects of ST. In rodents receiving stimulation to the cerebellar tonsil, seven daily subacute low-frequency sessions (stimulation at 1 Hz for 15 minutes) produced clear improvement in clonic convulsions and in post-discharge thresholds, together with durable inhibition of convulsions when stimulation was resumed (Weiss et al., 1995). Stimulation at 1 Hz for 15 minutes was more effective than stimulation at 10 or 20 Hz in attenuating convulsions. Although reiterated ECS also induced an anti-triggering effect, this dissipated rapidly over five days (Post et al., 1984). It is of great interest that recent publications have shown that rTMS at 10 or 20 Hz to the left frontal cortex, administered to patients suffering from refractory depression (George et al., 1995) or to patients (hospitalised or not) with milder degrees of depression (Pasquale-Leon et al., 1996), had a moderate or marked antidepressant effect. In these studies, rTMS showed few unwanted effects (other than mild pain in some patients, due to contraction of the temporal muscles); it did not induce motor convulsions, and did not, as such, appear to be associated with the memory loss described in subjective accounts or in preliminary neuropsychological tests (Little and Kimbrell et al., 1996). The optimal frequencies, durations and positions for rTMS to maximise its antidepressant effect still remain to be determined. However, the first controlled and open studies have tended to show that (because of the capacity of rapid magnetic fluxes to produce sub-convulsant electrical discharges that are relatively localised in the brain), rTMS may be found to be a clinically useful antidepressant model. This would suggest the possibility that some of the neurochemical changes induced by the clonic convulsions of ECS could be directly induced by stimulation at the very edge of the threshold (but still below it); this would open up the hope that one day these endogenous neurochemical processes could be identified and exploited in an optimal way for therapeutic purposes.