Review of: Khan RS et al, Lancet Psychiatry, published online 8/13/2018

Type of study: Sequential trial with open-label and randomized, double-blind comparison phases

Clozapine is often used as a last resort in schizophrenia, even though practice guidelines recommend a trial of this medication after failing 2 antipsychotics. The current study set out to test a treatment algorithm based on those guidelines in patients with first-episode psychosis.
Researchers recruited a total of 446 patients from 27 clinics in various European countries. All patients were in their first psychotic episode and had diagnoses of schizophrenia (51%), schizophreniform disorder (43%), or schizoaffective disorder (6%). To refresh your memory, schizophreniform disorder means that symptoms of schizophrenia have been present for more than a month but less than 6 months. The average age was 26; most were male (70%) and Caucasian (87%). The primary outcome was symptomatic remission on the Positive and Negative Syndrome Scale (PANSS). The trial was funded by the European Commission.

The patients were entered into a three-phase study:

• Phase 1: All 446 patients were prescribed open-label amisulpride, an antipsychotic used outside the US, for 4 weeks at ≤ 800 mg/day.


• Phase 2: Those patients who did not achieve remission on amisulpride were randomly assigned to a double-blind trial of either continuing amisulpride or switching to olanzapine (≤ 20 mg/day, mean 16 mg/day) for 6 weeks.


• Phase 3: Patients who did not respond to either amisulpride or olanzapine were treated with open-label clozapine (≤ 900 mg/day, mean 490 mg/day) for 12 weeks.

Amisulpride and olanzapine were selected for this algorithm because their effect sizes are second only to clozapine’s in schizophrenia.

Just over half (56%) of the patients remitted during the first phase of antipsychotic treatment with amisulpride. Of the 93 patients who started the second phase, about 32% remitted with either amisulpride continuation or olanzapine switch, with no significant differences between these drugs. Finally, 40 patients were left to assign to clozapine; 18 of those completed the 12-week trial, and 5 (28%) achieved remission.
Because switching to olanzapine did not yield better outcomes than continuing the first antipsychotic, the authors suggested that this second-line switch could be skipped and that patients who don’t respond to their first antipsychotic might be better served by going straight to clozapine.

TCPR’s Take
Moving clozapine up to a second-line treatment in schizophrenia is a bold suggestion. We’d like to see that tested out in a more controlled manner before changing practice guidelines. What these results do tell us is that schizophrenia recovery can take time. If patients haven’t recovered after 10 weeks, whether with one antipsychotic or two, a trial of clozapine is not unreasonable, but it’s not clearly the best option either.