Schizophrenia Podcast

Antipsychotics in Schizophrenia: Effectiveness and Optimal Dosing

John M. Kane, MD; Stefan Leucht, MD


August 02, 2022

This transcript has been edited for clarity.

John M. Kane, MD: Hi. I'm Dr John Kane, professor and chairman of psychiatry at the Donald and Barbara Zucker School of Medicine at Hofstra/Northwell in New York, and also the Zucker Hillside Hospital in New York as well. Welcome to Medscape's InDiscussion series on schizophrenia. Today we'll discuss how effective are antipsychotic medications and what have we learned about optimal dosing?

First, let me introduce my guest, Stefan Leucht. He is professor and head of the section of Evidence-based Medicine and Psychiatry and Psychotherapy at the Technical University of Munich in Munich, Germany. Stefan is one of the most well published and frequently cited experts in the entire field of psychiatry. Stefan, it's a pleasure to have you with us today.

Let me tee off the discussion by presenting a brief case. This is a 20-year-old young male who was a university student when he experienced his first psychotic episode, leading to an admission to a psychiatric unit at a nearby hospital, where he received the diagnosis of schizophrenia. He has never received antipsychotic medication. He and his family are somewhat reluctant for him to take antipsychotic medication and they're asking, why can't he be treated with just psychotherapy?

So, Stefan, what would your advice be to the clinicians on that unit when confronted with a patient like that? How would you explain the use of antipsychotic drugs and whether they are more effective than psychotherapy?

Stefan Leucht, MD: I would recommend to the parents of this young man that antipsychotic drugs are, according to current knowledge, the only effective treatment for people with schizophrenia. Therefore, I would strongly recommend taking one. For psychotherapy, unfortunately, there is some growing evidence that maybe it can also be used in such a case, but it's definitely not in the guidelines. So there's no strong evidence there.

But I would also recommend to them to use a relatively benign antipsychotic, meaning an antipsychotic with few side effects, and to use this antipsychotic at a low dose, because we know that people with a first psychotic break generally respond very well to antipsychotic drugs.

Kane: So when you say "respond very well," what does that mean? What are the chances that he's going to get better with medication?

Leucht: Studies from Professor Kane and colleagues have shown that over a year, about 80% of the patients would achieve a remission of positive symptoms, which is quite a high number.

Kane: Yeah, that's a pretty good response. One of the papers that you published a number of years ago compared the effectiveness of medications that we used in psychiatry vs the effectiveness of medicines that are used in other branches of medicine. I wondered if you could comment on that.

Leucht: Psychotropic drugs have a very bad image. There are two reasons for that. One reason clearly is that, unfortunately, in particular, the antipsychotics have many side effects and therefore they have a bad reputation. But the other thing is that many people think that psychiatric problems are purely psychological in origin, and therefore they think, well, we believe all this is psychology and thus we need to treat these patients with psychotherapy.

Then, in the context of this debate, there are also people who think that the psychotropics are not effective at all. Therefore, we looked at meta-analyses, reviews of psychotropic medications, and looked at what the effect size and the magnitude of the difference are vs placebo treatment for the major psychotropic interventions and compared them with the effect sizes of drugs in general medicine, like painkillers or antitumor medication, or drugs for COPD — a big variety, actually, of antipsychotic and of treatments that are used in general medicine. We found that on average, the effect sizes are the same. Of course, that's a very difficult comparison to make because we compared different drugs for different indications and in different outcomes. But still, I think this was a very important message to the field that our drugs work as well as those from other medical specialties.

Kane: I think that's a very important message to try to get across, particularly to a family like this, who are skeptical. They're reluctant to accept the recommendation of an antipsychotic drug.

Let's play it out a little further. Let's say that this young man does take the antipsychotic medication and he's in the category of the good responders. You said about 80% would respond. So then, when he leaves the hospital, he's also confronting his psychiatrist and saying, I don't want to take the medicine anymore because I feel fine now. I don't think this is going to happen again. What should the clinical team do at that point?

Leucht: I would recommend to the clinical team to continue with the antipsychotics for at least a year. Why not longer? Well, the course of schizophrenia is very different, and we know that about 20% of the patients with a first psychotic episode will not have a second one. So you have a small chance, a comparatively small chance of 20%, that you will not need the antipsychotics anymore. Therefore, I would not recommend a longer period, but 80% of patients with the first episode will have a second one. In particular, the best predictor for that is stopping the antipsychotic. Therefore, I would clearly recommend continuing with an antipsychotic drug for at least 1 or 2 years.

Kane: How would you communicate benefit-to-risk ratio? What are the chances that he would relapse if he stopped taking the medicine after he left the hospital? What are the chances that he would relapse compared with the chances of him relapsing while taking the medication?

Leucht: The chances of relapse are at least two times higher without medication vs with medication. And usually somebody who has had a first break of schizophrenia loses at least half a year until doing all the rehab and necessary rehabilitation, before being able to restart work or restart at university. So I think the stakes are very high to have another one. Therefore, I think that overall, the benefit-to-risk ratio is in favor of continuing antipsychotics.

Kane: So if he were to stop the antipsychotic medication a year or two later, you're suggesting that that might be feasible, but isn't there a really high risk that he would relapse at that point? I understand that maybe 20% of patients might not have another episode, but if 80% of patients do, what are the chances that he would relapse, even though he's been well for a year?

Leucht: I think that the relapse risk stays high. Also, after 2 years, we do not have good data — at least, no good randomized data. But we also have data from big registries like that in Finland, where they could look at nationwide data and find that even when people with a first psychotic break stop antipsychotics after 8 years, the risk is still very high that they will have a second rehospitalization. So the risk stays high.

However, schizophrenia is very heterogeneous. Not each case is the same, and we must not forget the side effects of antipsychotics. The side effects are a main driver for patients to say, Well, you know, I feel somewhat sedated, I'm not the same. I want to try this out.

Then I think it's always an individual decision. How dangerous was the first episode, for example? Was the patient suicidal? Was she or he aggressive? This would then go toward a recommendation to stay [on treatment] longer. But sometimes the episodes are comparably mild. Often they can be stopped quickly if they are realized early enough. I think individual decisions are also there. For example, is there family around who would recognize that the patient is getting unwell again? All of these are factors that need to be taken into account.

Kane: What's your take on the data that seem to suggest that with each relapse, the response is not as good as it was before?

Leucht: That's actually a very interesting discussion, John, because I think this is really going in two directions. On the one hand, we have these data showing that maybe the continuous use of antipsychotics leads to brain volume loss and things like that. I think that the data are pretty good that this happens. Of course, we can debate, how large is it? Is it really clinically important? But I think that everybody would be scared to learn that maybe your brain shrinks a little bit if you take antipsychotics for a long time. So that's one side of the coin.

The other side of the coin is what you just mentioned: There's now two or three replications [of data] finding that if patients stop medication after the first episode and have a second episode — I don't have the exact numbers in mind — but a comparatively smaller proportion will go into remission again. This is what people sometimes call the toxic effects of a relapse. I don't like this term, because what does toxic mean? It implies something like an immunologic process or something. But these are the two sides of the coin.

Kane: I think it's a very hard thing to study. There are some studies that suggest that prolonged periods of psychosis are also associated with brain changes. So we have the effect of the drug, and we have the effect of the disease. It's a very hard balance to achieve. But as you said, it's an individual decision. And we have to try to help our patients make an informed decision in the shared decision-making process. It's a tough choice; when one is faced with an illness like that, it involves a lot of tough decisions.

Let's go back to our case and play out a different scenario. Originally, the patient was in the group of people who responded well. What if he were in the 10% or 20% of people who don't get a good response even in their very first episode? What would you do at that point?

Leucht: Recently there was a study that tried to either keep the patient on the same antipsychotic or randomize them to another antipsychotic. This was the OPTiMiSE study, published in The Lancet Psychiatry, but unfortunately, there was no difference between them. In this case, the drugs were amisulpride or olanzapine. So it's really a very difficult decision.

I think I would still try another antipsychotic. Maybe before that, I would go to the upper dose range. I said that I would initially start such a patient on a low dose. I then would probably go up to the upper range of the officially licensed dose range, and then switch and try something with a different receptor binding profile. I think this would be my first approach.

Kane: And then if that doesn't work?

Leucht: And if that doesn't work following the guidelines, we are already there and then trying clozapine, which is the only drug that we have good evidence for that is more efficacious than other ones.

Kane: Just another question about the individual who doesn't respond well. I know you've published some data suggesting that we can learn a lot by observing what happens in the first couple of weeks, and if we don't see very much improvement in the first couple of weeks, that's not a good sign. It might indicate that the person is going to be in the poor responder category. Can you talk about that for a minute?

Leucht: Very often, we want to titrate the doses slowly up. But those patients who have not responded to 2 weeks of treatment with a full dose of an antipsychotic are very unlikely to then be in remission, for example, after 6 weeks. Therefore, actually we may already be at a full dose at 2 weeks. Probably clinically speaking, I would always say 3 weeks, for example, because it takes some time until we get to the dose and so on and so forth. Then we might already think about doing something else.

That dose increase has only been tested once, in a trial on lurasidone. It showed that indeed, most patients had not responded at 2 weeks to 80 mg of lurasidone, but the dose was increased to 160 mg. A substantial proportion of more patients responded to this dose of lurasidone, but that's actually the only study of this kind. Usually, dose increase does not seem to be very helpful if patients are already receiving an effective dose.

Kane: Do you think there's any role for therapeutic drug monitoring in situations like that?

Leucht: Oh, yes, I think so. Unfortunately, we do not have good evidence for the plasma level ranges in which we should titrate the drugs. But if a patient has not responded, I would get a plasma level just to know where the patient approximately is in terms of plasma levels. The best established plasma levels are probably for clozapine. We say that if a patient does not respond — again, the variability is quite enormous — but if patients don't respond, we should try to get the clozapine level up to more than 350-400 ng/mL.

Kane: I think you're right in what you said a minute ago, that even for those drugs where we don't have carefully established therapeutic ranges, we can still learn something from a blood level because if it's undetectable or very low, it tells us that either the patient isn't taking the medicine or they're a rapid metabolizer and we might need to raise the dose. Or, you know, in the case of clozapine, they're someone who's a heavy smoker. There are a lot of issues, but I think it's a strategy that seems to be grossly underutilized. People seem to be reluctant to do blood levels. Maybe it's inconvenient for the patient and for the clinician, but it's something that we should probably take more advantage of.

Leucht: I fully agree, John, but another aspect is also the cost. It's quite costly if you do it often.

Kane: That kind of introduces another topic, which is adherence. Certainly when I was a medical student, there wasn't much discussion about adherence. I always assumed that if I wrote a prescription for a patient, he or she would take it. And then I learned, as I began to treat more patients myself, that was a false assumption. We don't always know whether our patients are taking their medicine or not because they don't always tell us. Even if we ask them, they may not want to offend us or they may be embarrassed, or they may not even remember themselves how much medicine they've missed. What's your take on the use of the long-acting injectable (LAI) formulations?

Leucht: I'm generally favorable. I think they're very good options. Also in terms of convenience, for most of them, you only need to get a shot every month — some even every 3 months. So, I think that can be very useful and also very convenient for the patient. You come in about once per month, for example. That's probably what many people would like to do anyhow. And you can speak to your doctor or to your treatment team. So they're definitely a very good option.

On the other hand, also understand that not all the patients want that, and I think we should be open there, because if we get too much into this paternalistic thinking, although it's a reality, what you say is that many patients don't take the drugs as we would like them to. Of course, it's the right of everybody to take medication or not to take the medication. Because if you think too much in that direction, then we somehow also stigmatize patients a little bit. Right? It's like saying, well, I believe that following the data, following the evidence, you have a 50% chance of not taking your medications, and definitely there are patients who take the medication. This is also what most of the randomized trials comparing depot formulations with orals show — at least the patients who consent in these trials, they seem to be very compliant, because we do not find that much of a difference between LAIs and oral formulations in randomized controlled trials.

Kane: I think there are obviously many issues. In terms of convenience for the patient, if I'm confronted with making a decision every single day about taking medicine and that reminds me that I have an illness, that that may be more uncomfortable in some ways than taking and receiving an injection once a month.

Leucht: Absolutely. I think for many people, it will just be much more convenient. You know, you don't have to think about it. You just get your shot and, as you say, it can be less stigmatized.

Kane: I think we have to destigmatize nonadherence, because we don't want this to be a pejorative implication: Oh, you're a bad patient. You're not taking your medicine. I think it's human nature to have difficulty taking medicine. So we need to remind our patients that we all struggle with this issue. Even your doctor probably struggles with adherence at times. We're just trying to find a way to make it more convenient and more assured.

I also want to go back to the dose issue. You mentioned earlier when we were talking about the case that in a situation with a first-episode patient, we often use, very low doses because they've never been exposed to medicine before and they may respond well to low doses. What about in the patient who's been ill for longer? How do you go about deciding what is the best dose or what is the minimum effective dose?

Leucht: This is a very interesting debate on many levels. But generally speaking, I would say for most of the second-generation antipsychotics, the companies who develop these drugs got the dose ranges pretty well. For example, olanzapine, 5-20 mg, probably most patients respond pretty well somewhere in this range. For risperidone, the initial doses suggested were up to 16 mg, and it was found out that more than 6 mg of risperidone usually only produces more side effects, but not better efficacy. So staying in the predetermined dose range, which is also on the label of a drug, is usually correct, and going beyond usually will not produce more efficacy.

What many clinicians also do in our hospital sometimes is that they think, oh, this is an acutely ill patient or this patient will not go with, say, olanzapine, 20 mg, which is the maximum according to the label. They think that this patient needs 30 or 40 mg. There's no good evidence suggesting that this is any helpful.

Kane: And what about in the maintenance phase?

Leucht: The maintenance phase is also quite interesting. We recently published two papers. One was in the American Journal of Psychiatry in 2020, where we looked at the acute-phase dose and found that in terms of risperidone equivalents, or if you recalculate the drugs and convert to so-called risperidone equivalents, about 5 mg of risperidone is already sufficient and higher doses on average will not be more effective.

And then we did the same type of study, which is called a dose-response meta-analysis (a technical term in maintenance treatment), and our hope was that we would find that this maximum effective dose for maintenance treatment would be lower — like 3 mg, because we want to reduce the side effects, that's the that's the idea behind it. But we did not find that. So interestingly, again, the most effective dose was 5 mg risperidone equivalent. In other words, you might say "what made you well keeps you well" also applies to the antipsychotics. This does not mean that lower doses of 5 mg risperidone are inefficacious. They are also effective, but less effective than keeping a patient on such a standard dose — with again, a lot of interindividual variability. The confidence intervals were huge, meaning that there are patients who do well with lower doses, but there's also other patients who will need higher doses for maintenance treatment.

Kane: I guess this is another potential role for therapeutic drug monitoring if it can help us understand what the right dose is for a given patient.

Leucht: Definitely. At some stage, most patients or many patients will want to lower the dose. I've met many different patients. There are patients who just do not want to take antipsychotics. Then there are other patients who have even combinations of antipsychotics. I have a young woman in mind who told me, well, I've had two or three of these psychotic episodes. I'll take the side effects [of the drugs]. I just don't want to have another one. Keep me on this combination therapy.

Nevertheless, I think many people will want to reduce the doses, and there I would always do this very carefully and apply plasma level measurements just to know where the patient approximately is in terms of plasma levels, I think it's very useful.

I also think that LAIs are very useful even if you want to reduce the dose, because then you can go down and you know that the patient is covered and will always have at least some coverage. And this will be more assured than with using oral medication, I think.

Kane: A minute ago, you mentioned also concomitant therapy. You know that some patients are taking more than one antipsychotic drug. I know that when confronted with the scenario that we were talking about earlier of a patient who hasn't responded well, there are a lot of clinicians that would rather prescribe two antipsychotics than prescribe clozapine. What's your sense of the data on concurrent antipsychotic drugs? Is there any reason to use more than one drug?

Leucht: The randomized data, again, a little bit sobering in this regard. I think, at least when we look at double-blind studies, there's no evidence that combining in the acute phase is really effective. Again, this was interesting recently. Jari Tiihonen's group had a study of the Finnish registries population in which they found that a combination treatment of antipsychotics was more effective in preventing a relapse. So it's a difficult area, and many people do it.

What my take on that is that actually, the rationale is not very clear. All the currently available antipsychotics are mainly dopamine blockers, or dopaminergic blockers at least. The dopamine agonists are partial agonists — not really blockers, but they are still dopaminergic. So why would it make sense to combine two drugs with the primary mechanism of dopamine? This doesn't make a lot of sense to me.

Exceptions are, again, something like clozapine, for example, which only binds to dopamine receptors with 40% in an effective dose. Therefore, in terms of clozapine resistance, adding a purely dopaminergic drug like aripiprazole, for example, might be useful. But again, the data are not so clear about that, unfortunately.

What I usually tell clinicians is that I understand the patient is not doing better. We need to try something. The evidence is not good, but try it. If it works. Okay, keep going. If it doesn't work, stop it again and reduce the polypharmacy.

Kane: I think a message that's very important is that it's one thing to do a therapeutic trial as long as you have an endpoint in mind and you're actually measuring that, and then you can stop it if it's not working.

We've covered a lot of territory. To summarize, you've emphasized that antipsychotic drugs are very effective, that they're effective in the acute treatment of schizophrenia, and they're also highly effective in preventing relapse. But there is still some uncertainty as to if and when to stop the antipsychotic drug and what the consequences of that might be, and that clinicians need to recognize that dosage is important. We don't want to give a higher-than-necessary dose because that might be associated with more side effects. They need to understand the data that exist on what the most appropriate doses are. And you're also suggesting that there's not a lot of evidence supporting the use of concomitant therapy and that we need to think more about clozapine. We need to think more about therapeutic drug monitoring.

Thank you so much for being with us. And I want to thank the audience for joining Medscape's InDiscussion series on schizophrenia, and we look forward to future conversations. Thanks very much.


Toward a Phenomenological Psychotherapy for Schizophrenia

APA Releases New Practice Guideline on Treatment of Patients with Schizophrenia

Pharmacological Treatments for First-Episode Schizophrenia

Putting the Efficacy of Psychiatric and General Medicine Medication Into Perspective: Review of Meta-analyses

The Nature of Relapse in Schizophrenia

20-Year Nationwide Follow-Up Study on Discontinuation of Antipsychotic Treatment in First-Episode Schizophrenia

Amisulpride and Olanzapine Followed by Open-Label Treatment With Clozapine in First-Episode Schizophrenia and Schizophreniform Disorder (OPTiMiSE): A Three-Phase Switching Study


Lurasidone Dose Escalation in Early Nonresponding Patients With Schizophrenia: A Randomized, Placebo-Controlled Study

Dose-Response Meta-Analysis of Antipsychotic Drugs for Acute Schizophrenia

Examination of Dosing of Antipsychotic Drugs for Relapse Prevention in Patients With Stable Schizophrenia: A Meta-analysis

Association of Antipsychotic Polypharmacy vs Monotherapy With Psychiatric Rehospitalization Among Adults With Schizophrenia

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.