Antipsychotics: Less is Again More

It is now clinical standard and recommended by all treatment guidelines to continue an antipsychotic maintenance therapy for at least 12 months after remission of the first psychotic episode. Numerous studies have shown that the risk of relapse is significantly increased, if a drug treatment is terminated prematurely. In the September issue of JAMA Psychiatry Lex Wunderink and colleagues from the Netherlands, however, question these treatment habits at least in part (Wunderink et al., JAMA Psychiatry 2013; 70: 913-920). They asked the question of whether patients possibly achieve a better long-term outcome after their first psychotic episode, if their antipsychotic is discontinued early after remission.

Between 2001 and 2002, the authors had conducted an open, randomized study in 128 patients with the first episode of a schizophrenic disorder. After six months free of symptoms, the patients were randomized into two different treatment strategies: They received either an antipsychotic maintenance therapy or antipsychotics were gradually discontinued (Wunderink et al., J Clin Psychiatry 2007; 68: 654-661). Endpoints were relapse rates, and social and vocational level of functioning. The observation period was 18 months. Significantly more patients treated in the dose reduction arm had a relapse (43% versus 21%, p = 0.011). In approximately 50% of patients who were randomized to the dose reduction arm, dose reduction, however, was not possible. In another 30% the treatment had to be resumed  because psychotic symptoms reoccurred. After all, in 20% of patients antipsychotic treatment was successfully discontinued. The functional outcome was comparable in both treatment arms.

In the current publication, the most interesting results of a follow-up after seven years have now been reported. In the follow-up examination the following parameters were recorded: symptom severity and social functioning during the past six months, relapses during the entire seven-year period, and the type and dose of antipsychotic medication during the preceding two years. Recovery was defined as symptomatic and functional remission during the past six months. These in turn were recorded using standard scales (PANSS and GSDS).

Of the original 128 study participants 103 (80.5%) could be reassessed at follow-up. Of the remaining 25, one patient had committed suicide, 18 refused to participate and 6 were lost to follow-up. Participants and non-participants did not differ in terms of their characteristics at the time of the baseline examination. The patients in the two treatment arms did also not significantly differ at baseline. Patients in the dose reduction arm, however, had more often a regular job of at least 16 hours/week at a trend level (p = 0.07).

After all, 30 patients (29.1%) had achieved recovery after seven years. Significantly more patients who had been initially treated in the dose reduction arm (21 patients, 40.4%) had achieved a recovery compared with only 9 patients (17.6%) who had originally received a maintenance treatment (p = 0.004). The incidence of symptomatic remission was not different in the two arms, but significantly (p = 0.01) more patients had achieved functional remission with the dose reduction strategy. 28.2% of patients achieved either a symptomatic or functional remission.

A shorter period of non-treatment (DUP, duration of untreated psychosis) predicted symptomatic remission after seven years. Four baseline variables predicted a functional remission after seven years: less pronounced negative symptoms, living together, better social functioning and treatment group (dose reduction).

Overall, the mean number of psychotic relapses was 1.24 over the entire period. They did not differ in the two treatment arms (dose reduction: 1.13 vs. maintenance therapy. 1.35, p = 0.42). Although the patients in which the medication was tapered/discontinued had twice as many relapses  in the first two years compared to those patients who received a maintenance therapy, the recurrence rates equaled after about three years and then – although not significantly – reversed. After seven years 67 of the 103 patients had had at least one relapse (dose reduction: 61.5%; maintenance therapy: 68.6%).

Patients who were initially treated in the dose reduction arm had also received significantly lower doses of antipsychotics in the last two years of the follow-up period than patients in the maintenance therapy arm (2.2 mg vs. 3.6 mg haloperidol equivalents, p = 0.03). Of the 17 patients who had discontinued the medication in the original study, 13 took part in the follow-up, 10 of which had been in the dose reduction arm, 3 in the maintenance treatment arm. Two patients had to restart antipsychotic treatment again, which means that 11 were antipsychotic-free in the preceding two years. After seven years three patients from each treatment group had discontinued antipsychotic medication for the last two years, so that a total of 17 patients were drug-free at follow-up. A further 11 patients from the dose reduction arm and 6 patients in the maintenance treatment arm took less than 1 mg haloperidol equivalents, so that after seven years, 33% (34 patients) of the original group of patients were without substantial antipsychotic treatment (dose reduction: 42.3 % vs. maintenance therapy 23.5%, p = 0.04). Of these 34 patients 85.3% had achieved symptomatic remission (vs. 59.4% of antipsychotic-treated patients, p = 0.08), 55.9% were in functional remission (vs. 17.4%, p <0.001) and 52.9% had achieved recovery (vs. 17.4%, p <0.001).

The study shows for the first time, that an observation period of two years is too short for assessment of treatment strategies. Periods of seven years or even more may come to quite different results and call our current clinical routines into question. Another important finding of the study is that social and vocational rehabilitation are possibly more important endpoints than symptomatic remission. Long-term future drug treatment strategies should also be directed to smoothly regulate more basic neurochemical processes rather than to influence disturbed dopaminergic neurotransmission.

This post is also available in: German

Please leave a comment!