Antipsychotics for long-term treatment – when to stop? Or better not?
A new publication in the prestigious Lancet Psychiatry seems to provide new evidence for the benefits of long-term antipsychotic maintenance treatment in schizophrenia (Hui et al., Lancet Psychiatry 2018). On closer inspection, however, the paper confuses more than it provides new insights.
The follow-up after seven years, which was published in 2013 in JAMA Psychiatry, provided some unexpected results. Wunderink et al. now investigated what had happened to the patients originally included in the study. The primary endpoint after seven years was recovery from illness. Patients initially treated in the dose reduction group had achieved 40% recovery after seven years, but only 18% of patients in the maintenance treatment group. The authors now concluded: “Dose reduction/discontinuation of antipsychotics during the early stages of remitted FEP [first episode psychosis] shows superior long-term recovery rates compared with the rates achieved with MT [maintenance treatment]. To our knowledge, this is the first study showing long-term gains of an early-course DR [dose reduction] strategy in patients with remitted FEP. Additional studies are necessary before these results are incorporated into general practice.”
One of these “additional studies” requested by Wunderink et al. was now published by Hui et al. from Hong Kong University (Hui et al., Lancet Psychiatry 2018). Here, 178 patients were followed up for ten years after the first episode of schizophrenia. What exactly was done?
These 178 patients had undergone maintenance treatment for almost two years (minimum one year) after the first episode of disease, before being randomized into two groups: in a double-blind fashion, they were either given further maintenance treatment with quetiapine 400 mg for another 12 months or placebo. After a total of three years of treatment, further treatment was then naturalistic. After ten years, the patients were finally re-examined. The primary endpoint was the presence of a “good” or “bad outcome.” The “poor outcome” was defined as having persistent positive symptoms or treatment with clozapine or death by suicide. The authors found a poor outcome in 35 of 89 patients (39%) in the discontinuation group and in 19 of 89 patients (21%) in the maintenance treatment group (p = 0.012). In the former group, four suicides occurred, in the latter two. The authors conclude: “In patients with first-episode psychosis with a full initial response to treatment, medication continuation for at least the first 3 years after starting treatment decreases the risk of relapse and poor long-term clinical outcome.”
I am convinced that the present study, unfortunately, provides no significant new evidence for how patients with schizophrenia should be treated long-term. There are two reasons.
First, before inclusion in the study, patients had been treated on average for almost two years, sometimes significantly longer. The majority of patients had been treated with second-generation antipsychotics. However, this is the only indication that the authors provide for pretreatment. Dosages on entry into the study (after two years) are given in chlorpromazine equivalents, but cumulative dosages over the first two years of treatment are not reported. A treatment over two years, however, may have decisively influenced the long-term course of the disease. The patients in the Dutch collective had been treated considerably shorter before being randomized.
Second, and more importantly, how ill were the patients in the study by Hui et al. actually? Were these patients with schizophrenia as we know them? When enrolled in the study, patients had a mean PANSS score of 33.5 points. The lowest score possible on the PANSS is 30. Even a very healthy person may have a PANSS score of 33.5 on a bad day! Personally, I cannot remember a patient with the diagnosis of “schizophrenia” who at some point would have had such a low PANSS score.
And: After ten years, both groups had a mean PANSS score of 35 points, which is still extremely low. These patients were on average healthy and fully functional. In both groups, 70% of patients were in employment! A “bad outcome”, which is so common in our daily practice of patients with schizophrenia, is something very different! The two groups also did not differ in any other secondary endpoint and showed the same social adaptation and health-related quality of life.
Unfortunately, in my opinion, the present study does not provide any new indications as to how we should treat our patients with schizophrenia in the long term.