Text version
Hi and hello from Boston, Massachusetts. Oliver Freudenreich, or Dr. F, is here for the Psychopharmacology Institute. In this Quick Take, I am going to talk about a clinical antipsychotic dose reduction trial for schizophrenia that was published online in Lancet Psychiatry. The trial was conducted in England and goes by the name of RADAR—which stands for Research Into Antipsychotic Discontinuation and Reduction. It is a very apt name, as we will see, which is not a given for acronym trial names that can feel quite contrived.
What clinical question were the authors trying to answer with RADAR? They tried to figure out whether people with schizophrenia who had more than 1 episode of psychosis could be taken off their antipsychotic or reduce their antipsychotic: (1) Without increasing the relapse risk and (2) with benefit regarding overall psychosocial function. How did they conduct the study? The RADAR study design was a randomized trial of patients with the aforementioned schizophrenia spectrum diagnoses who had several relapses of the illness. Let me emphasize this: These were patients who had a confirmed history of relapses. They were not first-episode patients who only had 1 episode of illness. This is important, as we know that a significant minority of first-episode patients—around 20%–30%—only have 1 episode of illness, and they do not seem to need ongoing treatment to prevent a second episode. RADAR was randomized but open, meaning that patients and clinicians knew their assigned treatment, but the assessors were blinded to group assignment. Any study that uses such an open design is, of course, susceptible to certain biases. It is a weakness. The duration of the study was 2 years.
The 2 treatment arms were a maintenance treatment and a dose reduction treatment arm. Dose reduction was done very slowly, with dose adjustments only occurring every 2 months. People who reached a low enough dose could then decide to discontinue it. Social functioning was chosen as the primary outcome to reflect outcomes that are important to patients and society, and the Social Functioning Scale was used—a very established scale in schizophrenia trials. The main secondary outcome was severe relapse, defined as requiring admission to a hospital, which is an important real-world outcome variable. Statistically, the study was set up as a so-called noninferiority trial, meaning that the assumption was that the dose reduction was at least not worse than simply continuing on maintenance treatment.
Let me emphasize the strength of this trial: The authors did 2 important things when they designed this trial. One, they made the dose reduction very gradually over many months with a lot of flexibility to avoid withdrawal phenomena—including psychosis from discontinuing the antipsychotic too rapidly. In many trials that looked at relapse risk in schizophrenia, the medication is typically removed over days or a few weeks. Second, they added a measure of social function, which is important, as an adverse effect of dopamine blockade may be impairment in motivational or cognitive systems, which gets in the way of rehabilitation or work. Other than medical concerns from maintenance treatment with antipsychotics or any psychotropic, this is, in my mind, a valid concern that we tend not to talk with our patients about as much as maybe we should.
Now, let’s look at the main study results together: 253 patients were randomly assigned—126 to the reduction group and 127 to the maintenance group. The mean age was 46. The majority was male. Most were diagnosed with schizophrenia. This sample looks reasonably typical for community dwellings who refer their patients to me. What do you think the authors found? Do you think they were able to take people off their antipsychotics successfully or lower the dose compared with not doing anything with the antipsychotic? Well, you be the judge. These are the 3 main findings. One, there was no psychosocial benefit from dose reduction or discontinuation. Tow, people in the dose reduction group had about double the relapse risk compared with the maintenance group. Looking at the outcome after 2 years, 25% in the reduction group had at least 1 severe relapse compared with 13% in the maintenance group. If you take any relapse over these 2 years, 41% in the reduction group had a relapse vs 22% in the maintenance group. Again, this is a rule of thumb: It is about double the risk of relapse if you lower the dose or stop your antipsychotic. Three, other secondary outcomes showed no difference between the groups at 24 months, including symptoms, quality of life, adverse effect scales, body weight, and employment. Let me also add that only 27%, compared with 10% in the maintenance group, actually stopped their antipsychotic. The dose reduction was only from 300 mg–200 mg chlorpromazine equivalents. So, they did not actually succeed in significantly reducing the dose.
You can always look at clinical trial results as either a half-full or half-empty glass. I certainly understand when people point out that a significant number of people in the discontinuation arm in this study did successfully lower their antipsychotic or even came off their medications. However, I would argue that the dose was not lowered by that much—33% at the end of 2 years—and there was no benefit from doing this for medical or social variables. So, why even take the risk of a low-dose approach?
The main limitation of the study is its short-term follow-up of 2 years. It may simply not be long enough for the brain to heal itself—if you accept my choice of metaphorical language here—from this dopamine blockade or to translate into psychosocial benefits. There was also the problem of COVID-19, which started during the conduct of this trial after the recruitment period. Follow-up visits were online, and there were lockdowns, which may have limited employment opportunities, such as one of the desired psychosocial outcomes.
I think we have reached the limits of trials where we cannot personalize risks better, which means individualizing risks with biomarkers. We have basically taken the clinical method as far as we can. In this case, we have basically identified clinically a group of patients with remitting, relapsing psychotic illness where we know that removing the antipsychotic is risky or, vice versa, where treatment prevents relapse. Now, those patients had a recurrence risk as a defining part of their neurobiology, and reducing this risk requires the use of dopamine blockade. Given the wide range of clinical severity of schizophrenia and people’s values, some people may do okay with no treatment or much less treatment, which has potentially very significant advantages regarding function and quality of life, even though in this short trial, it was not seen.
Here is my clinical bottom line from the RADAR trial: Trying to lower and even stop an antipsychotic given for maintenance treatment of chronic schizophrenia, even if done very gradually, exposes patients to an increased relapse risk, roughly doubling the risk of relapse, including getting hospitalized, and you take this risk without getting any clear benefit from it. The authors honestly state in their abstract that “At 2-year follow-up, a gradual, supported process of antipsychotic dose reduction had no effect on social functioning.” This is a really important message, I think.
How does this trial then inform clinical care? Despite not showing a clear effect on psychosocial functioning and medical outcomes, I would say, given the possible clear downsides of antipsychotic maintenance treatment, I always discuss long-term plans with patients that include the possibility of a dose reduction over the years, including perhaps coming off, and the benefits from doing this but also the risks. Many patients and families will not want to risk relapse. Others may consider it. Good follow-up is key for those patients who are willing to take the increased risk of relapse. RADAR provides you with some numbers to discuss these risks. Involving families or group home staff is key for a successful and safe dose reduction. I would only consider dose reduction if there is a collaboration with people involved in the patient’s life, as you need eyes and ears in the community to help monitor patients. Patient care demands, on the other hand, a discussion of this option, as there may be more to life than relapse intervention. I am adding this sentence with some hesitation because I trained at a time when relapse prevention was the absolute gold standard, but over the last decade or so, we have come to a more nuanced understanding, including the nonmedical risk of dopamine blockade.
I look forward to the long-term outcomes from this trial in 5 years and 10 years, if I am still practicing psychiatry then.
Abstract
Antipsychotic Dose Reduction and Discontinuation Versus Maintenance Treatment in People With Schizophrenia and Other Recurrent Psychotic Disorders in England (the RADAR trial): An Open, Parallel-Group, Randomised Controlled Trial
Joanna Moncrieff, Nadia Crellin, Jacki Stansfeld, Ruth Cooper, Louise Marston, Nick Freemantle, Glyn Lewis, Rachael Hunter, Sonia Johnson, Thomas Barnes, Nicola Morant, Vanessa Pinfold, Ruth Smith, Lyn Kent, Katherine Darton, Maria Long, Mark Horowitz, Robert Horne, Victoria Vickerstaff, Mithilesh Jha, Stefan Priebe
Background: Maintenance antipsychotic medication is recommended for people with schizophrenia or recurrent psychosis, but the adverse effects are burdensome, and evidence on long-term outcomes is sparse. We aimed to assess the benefits and harms of a gradual process of antipsychotic reduction compared with maintenance treatment. Our hypothesis was that antipsychotic reduction would improve social functioning with a short-term increase in relapse.
Methods: RADAR was an open, parallel-group, randomised trial done in 19 National Health Service Trusts in England. Participants were aged 18 years and older, had a diagnosis of recurrent, non-affective psychotic disorder, and were prescribed an antipsychotic. Exclusion criteria included people who had a mental health crisis or hospital admission in the past month, were considered to pose a serious risk to themselves or others by a treating clinician, or were mandated to take antipsychotic medication under the Mental Health Act. Through an independent, internet-based system, participants were randomly assigned (1:1) to gradual, flexible antipsychotic reduction, overseen by treating clinicians, or to maintenance. Participants and clinicians were aware of treatment allocations, but assessors were masked to them. Follow-up was for 2 years. Social functioning, assessed by the Social Functioning Scale, was the primary outcome. The principal secondary outcome was severe relapse, defined as requiring admission to hospital. Analysis was done blind to group identity using intention-to-treat data. The trial is completed and has been registered with ISRCTN registry (ISRCTN90298520) and with ClinicalTrials.gov (NCT03559426).
Findings: 4157 people were screened, of whom 253 were randomly allocated, including 168 (66%) men, 82 (32%) women, and 3 (1%) transgender people, with a mean age of 46 years (SD 12, range 22-79). 171 (67%) participants were White, 52 (21%) were Black, 16 (6%) were Asian, and 12 (5%) were of other ethnicity. The median dose reduction at any point during the trial was 67% in the reduction group and zero in the maintenance group; at 24 months it was 33% versus zero. At the 24-month follow-up, we assessed 90 of 126 people assigned to the antipsychotic dose reduction group and 94 of 127 assigned to the maintenance group, finding no difference in the Social Functioning Scale (β 0·19, 95% CI -1·94 to 2·33; p=0·86). There were 93 serious adverse events in the reduction group affecting 49 individuals, mainly comprising admission for a mental health relapse, and 64 in the maintenance group, relating to 29 individuals.
Interpretation: At 2-year follow-up, a gradual, supported process of antipsychotic dose reduction had no effect on social functioning. Our data can help to inform decisions about the use of long-term antipsychotic medication.
Download PDF and other files
Reference
Moncrieff, J., Crellin, N., Stansfeld, J., Cooper, R., Marston, L., Freemantle, N., Lewis, G., Hunter, R., Johnson, S., Barnes, T., Morant, N., Pinfold, V., Smith, R., Kent, L., Darton, K., Long, M., Horowitz, M., Horne, R., Vickerstaff, V., … Priebe, S. (2023). Antipsychotic dose reduction and discontinuation versus maintenance treatment in people with schizophrenia and other recurrent psychotic disorders in England (the RADAR trial): An open, parallel-group, randomised controlled trial. The Lancet Psychiatry, 10(11), 848-859.
