SummaryAs emphasized by Lecrubier in 1980, the major finding of the last thirty years of classical neuroleptic use in schizophrenia is not that they are antischizophrenic or antipsychotic agents, but that they act on positive symptoms whatever the cause. There is now a widely accepted attribution of this kind of therapeutic property to the post-synaptic dopaminergic blockade induced by most of the neuroleptics when administered at high doses. On the other hand, during the last decade, various authors have reported clear evidence for a disinhibitory action of some neuroleptics when used at low doses. From a clinical point of view, the literature on the therapeutic action of low doses of neuroleptics seems quite controversial. In order to assess the exact determinants for clinical activity of neuroleptics (such as patient type, selected substances, administered doses) two series of independent controlled studies were conducted.First series of studies: active drugs at low and high doses in schizophrenic patients(Tables 1 to 6)The first study was designed to assess the change of negative symptoms under low doses of neuroleptics. Sixty-two patients meeting the DSM III criteria of schizophrenia (subtypes: disorganized, catatonic or residual) were randomly assigned after a three-week washout period to six weeks’ treatment with either amisulpride (50 to 300 mg/day) or fluphenazine (2 to 12 mg/day), administered in keeping with a flexible dosage schedule. All patients had to meet the Andreasen criteria for negative symptoms (at least two of the following components had to be checked as present: anhedonia, alogia, affective flattening, avolition-apathy, attentional impairment). None of the patients presented one of the positive components to a marked degree: hallucinations, delusions, bizarre behavior, positive formal thought disorders. The evolution of symptomatology was assessed by means of the Brief Psychiatric Rating Scale (B.P.R.S.) and an ad hoc defective symptoms scale (D.S.A.S.), the sensitivity and reliability of which was previously tested in our department. The results show that the two groups (amisulpride versus fluphenazine) were initially highly comparable. Negative symptoms were severe, as evidenced by the D.S.A.S. scores, and by the presence of three items of the A6 criterion of the DSM III. The final global clinical assessment and the final D.S.A.S. scores both showed a significant improvement, with no statistically significant difference between the two treatments. Nevertheless, the scores of the “anergia” and “anxiety-depression” factors of the B.P.R.S. showed a significantly greater improvement in the amisulpride group.The second study, complementary to the first, aimed to check the efficacy of high doses of amisulpride on the productive symptoms of schizophrenia. In this case, only the B.P.R.S. was used due to the good correlation of this scale with the global severity of the positive symptomatology. After three weeks of treatment, consisting either of amisulpride at a high, flexible dosage (800 to 1200 mg/day) or of haloperidol (20 to 30 mg/day), each of the two groups of twenty patients showed a significant improvement. In particular, the “thought disorders” factor of the B.P.R.S. (which unfortunately does not correspond exactly to the “formal thought disorder” component of the Andreasen positive symptoms scale - S.A.P.S.) was greatly improved in both groups. Evolution of the other symptoms was however identical in the two groups.Second series of studies: active drugs versus placebo in schizophrenia(Figures 1 to 6)As we know, dopaminergic blocking agents are able to induce negative symptomatology. Consequently, to separate the secondary syndrome from the true deficit, a longer washout period than that previously described has to be imposed. On the other hand, the longitudinal course of schizophrenia must be taken into account for correct interpretation of changes in symptomatology; for example, patients with negative symptoms may abruptly present productive episodes, in particular during the neuroleptic withdrawal period. For optimal control of these two variables (natural history of the disease, the blunting effect of neuroleptics), 90 patients presenting either or both subtypes of schizophrenia were selected and included in a two-step, double-blind, controlled study. Patients with negative symptoms underwent a six-week washout period, after which they were treated either with low doses of amisulpride (100 or 300 mg/day) or a placebo. Patients with initial positive symptoms received mandatory high doses of amisulpride. According to the protocol, negative patients presenting productive symptoms during the washout period were to be systematically assigned to the highdose group. First results concerning 38 patients with a predominant negative symptomatology are presented here. A very clear improvement can be shown both for the SANS global mean score and for the sum of global ratings in the groups treated with low doses of amisulpride compared to the placebo group. The scores of the alogia, blunted affect and attentional impairment subscales decrease dramatically as well with the active drug.