LAUSR

Rapid antidepressant response and novel therapeutic targets are all the rage now in depression treatment. But while novelty is exciting, and may eventually lead to better treatments, there is no substitute for proven reliability when it comes to the urgent treatment of seriously depressed patients. Electroconvulsive therapy (ECT) provides that, as well as reasonably rapid response in many patients. Birkenh€ager et al., in this issue (1), have added valuable data to the evidence base that shows early improvement predicts favorable outcome from a course of ECT. Their data also confirm that lack of early improvement should not at all discourage continuation of the ECT course, as most patients will improve, given an adequate course of ECT. Importantly, their data also remind us that ECT starts to act rapidly, often in patients who have been ill for years. Birkenh€ager et al. (1) report remission rates of 73%. These are high, but bitemporal ECT was used exclusively, and the population included was particularly ECT-responsive: older patients, two thirds of whom had a psychotic depression (Table 1). Comparable remission rates were reported in the PRIDE-study, using ultrabrief pulse unilateral ECT in older patients, of whom 11.7% had the psychotic subtype of depression (2). The MODECTstudy by the European research-consortium ResPECT, in older patients treated with brief pulse unilateral ECT, yielded a remission rate of 66.4% (3). Somewhat lower remission rates were reported in the Irish EFFECT-DEP-trial, comparing brief pulse BT and RUL ECT in younger patients, 21% with psychotic symptoms (4) (Table 1). ECT is a fast-acting antidepressant. A substantial decrease in depression severity is often observed after the first session (5). In a large study, patients experienced a 10.6 HAM-D point reduction, on average, after the first ECT session (6). A third of patients achieve remission at or before week 2 (i.e., six treatments) (7), and older patients with a psychotic depression have a high chance of achieving complete remission within 4 ECT sessions (8). There are even rare cases of remission after a single ECT (9). Rapid response with ECT suggests a mechanism of action other than the recently discovered neurotrophic effects of ECT; while some authors have suggested a placebo effect, we believe this is not an adequate explanation in such a severely ill population. Rather, it is likely that powerful biological mechanisms are set in motion at the first treatment. Importantly, ECT has also been shown to result in a rapid decrease in suicidal ideation (10). The search for predictors of ECT response seems to be the quest for the holy grail. Despite the high probability of good outcome with ECT in appropriately chosen patients, it would still be extremely helpful to be able to tell patients more precisely the likelihood that they will benefit from ECT. The corollary, prediction of ECT non-response, would also be very useful (11). Recently, two meta-analyses have been published, showing that older age, the presence of psychotic symptoms, and shorter episode duration predict a better outcome (12, 13). The impact of age is most probably mediated by the presence of psychomotor and psychotic symptoms (14). The role of treatment-resistance remains controversial. Recently, machine learning techniques have been applied to neuroimaging data in ECT patients to develop response prediction models (15). While still in its infancy, this application of a sophisticated data analytic technique to the rapidly expanding field of neuroimaging findings is very promising. The above clinical and neurobiological characteristics aim at selecting the most appropriate Table 1. Remission rates and patient characteristics in modern ECT-trials