Dr. Ongur, over the past few decades, the dominant  theory of schizophrenia has been the dopamine hypothesis, but that has been changing lately. Can give us a quick review of the original hypothesis and then go on to some of the alternative neurochemical theories?
Sure. The dopamine hypothesis evolved from some serendipitous observations in the early 1950s that the drugs that seemed to calm down psychotic patients were blocking the dopamine receptors. Furthermore, if you stimulate the dopamine system using psychotomimetics drugs like cocaine and amphetamines—you can either produce psychotic symptoms or worsen psychotic symptoms in patients with schizophrenia.

And how has the dopamine hypothesis held up over time?
Our thinking has changed. We no longer see a dopamine abnormality as being central. For example, the most effective of all antipsychotics, clozapine, is a pretty weak dopamine blocker. If clozapine is so great but it doesn’t actually interact with the dopamine system that much, what else is it doing that is effective? As it turns out, there is very little evidence that the dopamine system is truly abnormal in schizophrenia, or if it is abnormal, this may well be the consequence of completely different brain pathology. If so, blocking dopamine receptors may not really be addressing the cause of schizophrenia, but it may be a palliative treatment. As an analogy, imagine being in a room that is too hot because there is a big fire in the fireplace. If you can’t see the fireplace, you would not know that this is the source of the heat, and you might open a window to cool it down. The room cools down a bit, and you might say, “the cause of the heat in the room is the windows,” but in fact the actual cause is the fire, and opening the window is only palliative. Similarly, blocking dopamine receptors helps alleviate schizophrenia but the root cause of the psychotic symptoms may have little to do with dopamine.

Are there any likely alternative hypotheses?
The research has been slowly progressing. Many neuroscience researchers have become interested in glial cells and their relationship to glutamate and GABA. This has been one of my research interests as well.

What are glial cells, and how are they related to glutamate and GABA?
Glial cells are the cells in between the neurons, and we used to think that they served merely as structural support for neurons. But it turns out that they do a lot of important things, one of which is that they clear out glutamate and GABA from the synapse. Glutamate is the brain’s main excitatory neurotransmitter and GABA is the main inhibitory transmitter—the two act like the gas pedal and the brake pedal of the brain, respectively. Both of these chemicals are very useful in the synapse where neurons are communicating with one another. But if they start to diffuse out of that area they can be quite dangerous, especially glutamate, which can damage neurons and ultimately kill them. There are some neurodegenerative conditions like amyotrophic lateral sclerosis (ALS or Lou Gehrig’s disease) in which the pathological role of glutamate is well established.

So how are these glial cells relevant in psychiatric symptoms?
In postmortem studies of people with mood disorders,  here appears to be a deficit of glial cells, and there are similar abnormalities in schizophrenia. So you can imagine a situation in which glial cells are either sick or dying or they never formed appropriately, so they are not doing their job. This allows excess glutamate to accumulate in the brain, leading to neurodegeneration.

Are there any medications under development that target this mechanism?
Eli Lilly has a glutamatergic drug, mGlu2/3, that looked promising in a small single-center trial two years ago(Patil ST et al., Nature Medicine 2007;13:1102-1107) but which performed more poorly in a recent larger multi-center trial (Eli Lilly press release, March 29 2009, http://bit.ly 41ca2c). We are told that other similar drugs are in various stages of development.

Recently, there has been much talk about the treatment of negative symptoms and cognitive symptoms in schizophrenia. Are the atypicals particularly effective for such symptoms?
Unfortunately, as it turns out, atypicals are probably not better for cognitive or negative symptoms than first generation antipsychotics. There was some initial excitement based on several studies suggesting that atypicals might be better. But the problem was that many of those studies were done by the manufacturers of the atypicals, which has been shown to bias outcomes in clinical trials. Even the independent studies were poorly blinded and poorly controlled, so there was the possibility for bias and error on the part of the experimenter. Once the dust started settling, it became clearer that any advantage of atypicals was an artifact of how studies were designed. The most important information has come from the CATIE study, which was NIMH-funded. This was a large unbiased trial with several antipsychotics and subjects were given cognitive testing before and after starting medication. My interpretation of the results was that any improvements in cognition were no more than what would happen if you simply had a practice effect. 

What do you mean by a “practice effect”?
If you take a test and then take the same type of test again a few weeks later, chances are you will perform better simply because you know what to expect. The testing environment is the same, the paper and pencil are the same, and the types of questions are the same. So even if you have not become any smarter or clearer cognitively, you will improve by a small amount based on this practice effect, and this is the level of improvement seen in the CATIE trial. 

And was the level of cognitive improvement equivalent in the patients assigned to atypicals vs. typical antipsychotics?
That was the real kicker of this study. The atypicals were no better at cognitive enhancement than perphenazine (Trilafon) which was the first generation comparator in the CATIE trial. In fact, numerically perphenazine looked slightly better than the atypicals, but this was not statistically significant. Most people have come around to the idea that the early studies were all biased—not necessarily intentionally—and now we are realizing that the effect on cognitive dysfunction and negative symptoms is not really there.

Among the atypical antipsychotics, are there any that stand out in terms of efficacy?
In the first phase of the CATIE, patients stayed on Zyprexa longer than they stayed on the other medications, although this may have been an artifact the research design and many potential issues have been discussed in relation to this. Interestingly, in the second phase of the CATIE, where patients who had stopped taking medication were re-randomized to a new medication, Zyprexa did not maintain its superiority. So even if the first phase results were valid, they do not generalize to other trials, including within the CATIE. In my opinion olanzapine is a perfectly good medication. Is it head and shoulders better than everything else we have? Not necessarily, and it causes so much morbidity with the weight gain and hypercholesterolemia that we use it much more sparingly than we used to. In the inpatient unit at McLean, we usually try other medications if we are concerned about weight gain or related issues. In our experience, medications that are strong dopamine blockers (first generation drugs, as well as risperidone) are probably comparable to olanzapine—maybe a little worse in terms of efficacy, but their side effect profiles are more tolerable in many cases.

What about clozapine?
Clozapine is an underutilized drug—unlike the other antipsychotics, there is very good, solid data that clozapine is helpful for negative symptoms, such as the amotivation syndrome and alogia (impoverished speech production). Typical medication algorithms reserve clozapine as a last resort after two other medications have failed, and in clinical practice we typically see patients who have been on six or seven medications and are still doing poorly before the family will agree to a clozapine trial. This is understandable, because of the weight gain and the need for monitoring. Nonetheless, I wouldn’t be surprised if some of the guidelines begin to emphasize clozapine more. It is rare to see dramatic clinical responses in schizophrenia, but those patients who have the diagnosis and yet who present as completely healthy—in terms of the way they are dressed, the way they talk and so on—those are the people who are likely to be on clozapine.

Thank you, Dr. Ongur.