Long‐term antipsychotic treatment of schizophrenia: does it help or hurt over a 20‐year period?

Correll et al1 argue for a positive view of the risk-benefit ratio for long-term continuous antipsychotic treatment of schizophrenia. They claim that studies of long-term outcome which show negative results are not convincing because of confounding factors. Their chief argument is that in “non-randomized, uncontrolled studies… there is a high risk of confounding by indication and reverse causation, in that greater illness severity could be the cause of continued antipsychotic treatment, rather than being the effect”1. The other argument is that long-term continuous use of antipsychotics does not involve significant morbidity from dopamine supersensitivity psychosis. Here we provide evidence which severely questions both of these conclusions, showing that they overestimate the benefits and underestimate the risks of long-term antipsychotic treatment. There are at least eight studies assessing whether schizophrenia patients improve when treated longer than two-three years with antipsychotic medications. These studies have been conducted by eight different investigator groups. They include those by Wunderink et al in the Netherlands2, our own Chicago Followup Study3, the Suffolk County study of Kotov et al4 in the US, and the long-term data provided by the Danish OPUS trial5, the AESOP-10 study in England6, the Finnish Birth Cohort Study7, the Alberta Hospital Follow-Up Study in Western Canada8, and the international follow-up study by Harrison et al9. These research programs included samples studied from 7 to 20 years. Unlike short-term studies, none of them showed positive long-term results. Correll et al quote for support a study by Ran et al10 favoring long-term use of antipsychotics for schizophrenia in China. However, there are many weaknesses in that study. In particular, the untreated group was selected from much older unmarried chronic rural uneducated patients, while the treated group consisted of younger married educated urban patients, some of whom had received only one short period of medication over the 14 year period, rather than being continuously medicated. As we have noted, one argument used to explain the negative results of long-term antipsychotic treatment is that the schizophrenia patients on antipsychotics for a prolonged period are more severely ill than those not on antipsychotics. However, there are no clear features on which everyone would agree distinguishing “more severely ill people with schizophrenia”. Nor is it always clear what “severity” means in relation to schizophrenia. One frequently used criterion for severity refers to more blatant psychotic illness. However, some episodes of blatant psychosis clear up quickly and thus these psychotic patients may not be more severely ill in every respect. Another potential criterion for severity in people with schizophrenia involves those whose disorder is more likely to be sustained over a longer period of time, or who have a poorer long-term prognosis. To control for this possible confounder, we have utilized the prognostic indices outlined by Vaillant, Stephens and Zigler. These were collected in our studies at index hospitalization. Later we compared the long-term outcome of poor-prognosis schizophrenia patients medicated with antipsychotics for 15-20 years to that of poor-prognosis patients not prescribed antipsychotics for 15-20 years. We also compared a good-prognosis sample of patients prescribed antipsychotics for 15-20 years with a good-prognosis sample of patients not prescribed antipsychotics for 15-20 years. In both comparisons, those patients not on antipsychotics for 15-20 years had fewer symptoms and better outcomes after the first 2-3 years3. An additional limitation of Correll et al's paper is that they do not fully address the evidence on dopamine supersensitivity psychosis from animals and from humans. They limit their discussion to short-term studies of psychotic relapse and the potential loss of antipsychotic efficacy, while ignoring the serious risk for the syndrome resulting from continuous long-term antipsychotic treatment. The clinical picture of dopamine supersensitivity psychosis is well defined and occurs with increasing frequency after two to three years of continuous antipsychotic maintenance use. Studies indicate that the syndrome manifests in 70% of patients with treatment resistant schizophrenia11. Other studies show that the switch to aripiprazole, mentioned by the authors, may actually unmask and intensify psychotic symptoms previously suppressed by stronger D2 antagonists12. While long-term continuous use of antipsychotics may induce the syndrome, these medications also block psychotic symptoms, which therefore remain largely unrecognized until the “breakthrough” of more severe symptoms occurs and leads to treatment resistance. While several research groups have described dopamine supersensitivity psychosis as a serious risk of long-term continuous use of antipsychotics, there has been a systematic failure to incorporate this finding into the risk-benefit ratio for continuous use of antipsychotics. The same applies to the possible negative impact of long-term antipsychotic treatment on work functioning3: the block of dopamine receptors may indeed reduce drive and motivation. Unfortunately, views about the long-term efficacy of antipsychotics are often based on the results from short-term (0-2 years) evaluations. As we have highlighted, there are at least eight major studies which fail to find better outcomes for schizophrenia patients treated on a long-term basis with antipsychotics. These negative results from multiple large well-documented long-term studies are a clear warning sign. Martin Harrow, Thomas H. Jobe Department of Psychiatry, University of Illinois at Chicago, College of Medicine, Chicago, IL, USA

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.