Experts Square Off Across Antipsychotic Generation Gap

Are second-generation antipsychotics (SGAs) uniformly superior in efficacy and safety to first-generation antipsychotics (FGAs)?

Four leading lights in schizophrenia research sought to answer that question in a spirited debate at APA's 2005 annual meeting in May in Atlanta.

Answering in the affirmative were John Davis, M.D., a professor of psychiatry at the University of Illinois, Chicago, and Ira Glick, M.D., a professor of psychiatry and behavioral science at Stanford University School of Medicine.

They emphasized the findings of a meta-analysis of 124 randomized, controlled trials comparing efficacy data on first- and second-generation agents and 18 comparison studies of SGAs published in the June 2003 Archives of General Psychiatry. The results showed that while SGAs do not constitute a completely homogenous group, clozapine, amisulpride, risperidone, and olanzapine each showed greater efficacy than FGAs (Psychiatric News, July 4, 2003).

Davis and Glick, co-authors of the article, argued also that studies strongly suggest that the SGAs have a superior effect on negative symptoms, as well as on cognition. The result is better overall functioning in the long term, they said.

Defending the FGAs were Rajiv Tandon, M.D., director of the schizophrenia division in the Department of Psychiatry at the University of Michigan, and William Carpenter, M.D., director of the Maryland Psychiatric Research Center. They criticized much of the existing clinical-trials data comparing SGAs and FGAs for methodological problems and biases. Among them was improper dosing with FGAs. Carpenter and Tandon said putative benefits of SGAs disappeared when they were compared with haloperidol given in lower doses in combination with anticholinergic drugs.

Carpenter said that many of the industry studies looked at quality of life but did not include those data in their reports. That omission suggests that the SGAs are not making a good showing in the truly important category of functional improvement.

“None of the studies shows that we are changing functional outcome with any of these drugs,” Carpenter said. “Industry has created blockbuster markets without creating important therapeutic advantages.”

Presenting data from the June 2003 Archives report, Davis told the standing-room-only audience that much of the greater efficacy found for at least some of the SGAs (olanzapine, clozapine, amisulpride, and risperidone) was due to their greater effect on negative symptoms.

“I feel the four more effective antipsychotics have a broader spectrum of action,” Davis said. “They are doing more, and most of why they are doing better is that they are doing more for symptoms that are not helped by the conventional antipsychotics.”

Davis added that though the evidence is limited, SGAs may have a better effect on cognition. Especially intriguing, he said, is evidence from a comparison of olanzapine and haloperidol in first-episode schizophrenia suggesting the SGA was associated with less deterioration of gray matter (Psychiatric News, May 5).

Glick underscored the importance of advances against negative symptoms.“ Long-term maintenance is where the action is,” he said.“ Overall, as we manage patients over many decades, the SGAs clearly have a major advantage.... The focus is no longer on positive symptoms; it's on long-term quality of life, negative symptoms, and cognition.”

“I will go for every bit of benefit I can get,” Davis said.“ You have more flexibility with the SGAs, and ultimately you will come out with a better result. Weight gain is [a major] side effect, but the way to minimize it is to start working on it in the first episode and manage all the risk factors. If the patient is doing better [overall], you will be able to do a better job [managing weight gain].”

In response, Tandon emphasized what the Davis-Glick meta-analysis showed as well—that the SGAs are not a homogenous group and that there was significant variability in between-group differences. For that reason alone, he said, it was not possible to agree with the assertion that SGAs are“ uniformly” superior in efficacy, safety, and tolerability.

Tandon went on to say, however, that while some SGAs, particularly clozapine, may have superior efficacy with refractory patients, the evidence is much less clear with first-episode patients. “The differences in efficacy are very soft,” he said.

He also criticized as problematic the use of “last observation carried forward” (LOCF) analysis in many studies. LOCF is sometimes used in longitudinal studies to make up for treatment drop-out and attrition and to preserve sample size, but its use may result in faulty estimation of treatment effects. “Our patients are not the last observation carried forward,” Tandon said.

Carpenter echoed this criticism, saying that unless attrition is random from both arms of the comparison, LOCF is not a valid tool for analysis. Patients who do not respond in a short period to haloperidol, for instance, may drop out or switch to the comparator drug; the last observed value for their performance is carried forward as the value that will be used in the results.

“If you agree that eight weeks of Haldol beats two weeks of Haldol, then you have to accept that how long you are on treatment really matters,” Carpenter said. “So the last observation carried forward is very problematic.”

Carpenter cited a host of other methodological problems with comparison studies between FGAs and SGAs. Principal among these is the fact that all of the studies recruited patients who had failed on FGAs, ensuring a bias in the results toward efficacy with SGAs.

He also cited very high doses of FGAs in many studies and the failure to pre-treat those receiving FGAs with anti-Parkinsonian drugs. When lower doses and anti-Parkinsonian drugs are used, SGAs' advantages in efficacy disappear, Carpenter said.

Because of the large sample sizes, studies may result in fairly small effect sizes that, while statistically significant, are clinically all but negligible.

Carpenter was also highly critical of the claim that SGAs improved cognition. “It is preposterous to think of these as pro-cognition drugs when the advantages are in comparison to drugs that we know interfere with cognition,” he said.

Finally, Carpenter and Tandon stressed that weight gain associated with SGAs is a serious, potentially life-shortening side effect that cannot be minimized. Tandon noted, for instance, that the greatest cause of increased mortality among patients with schizophrenia was cardiovascular disease.

There was agreement among the debaters that SGAs are not homogenous and that the best treatment is one tailored to each patient.

“You must individualize treatment for each patient,” Glick said. “Do the maximum that you can for the patient and the family over the long run and don't overtreat. I consistently emphasize monotherapy and a minimum of concomitant therapy because of the side-effect issues.”

Carpenter said, “There is a wide range of drugs that vary mostly in their side-effect profiles. We ought to be matching up an individual patient with the side-effect profile that would be most advantageous for that particular patient.” ▪