From the Editor

A decade after the CATIE study, the focus has shifted from effectiveness to neuroprotection

Author and Disclosure Information

This past September, exactly 10 years after publication of the primary findings of the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study.



This past September, exactly 10 years after publication of the primary findings of the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study1—namely, that effectiveness (defined as all-cause discontinuation) was not different across first-generation antipsychotics (FGAs) and second generation antipsychotics (SGAs)— a new meta-analysis by Vita et al2 of differences in cortical gray-matter change between those 2 classes of antipsychotics offers a reminder: The clinical focus of the CATIE study overlooked impor­tant neurobiological and neuropro­tective differences between FGAs and SGAs.

How drastically 1 decade can change the scientific perspective! Vita et al’s meta-analysis and meta-regression encompassed all 18 MRI studies of cortical gray matter in patients with schizophrenia.2 Earlier studies (published between 1983 and 2014) had lumped together patients who were receiving an FGA and those receiving an SGA, and authors reported overall reduction in cortical gray matter with prolonged antipsy­chotic treatment.

Remarkable findings emerge
When Vita et al2 analyzed FGA- and SGA-treated patients separately, how­ever, they found a significant reduc­tion in cortical gray matter in the FGA group but not in the SGA group. In fact, while higher daily dosages of FGAs were associated with greater reduction in cortical gray matter, higher dosages of SGAs were associated with lower cortical gray matter reduction and, in some samples, with an increase in vol­ume of cortical gray matter.

The researchers hypothesized that the differential effects of FGAs and SGAs might be attributable to the neu­rotoxicity of typical FGAs and the neu­roprotective effect of atypical SGAs.

The key neurobiological difference between FGAs and SGAs reported by Vita et al2 was not addressed in the CATIE study, leading, at that time, to a rush to judgment that all antipsy­chotics are the same. This conclusion emboldened managed-care organi­zations to mandate use of older (and cheaper) generic FGAs instead of newer (and more expensive) SGAs— most of which have become available as generic equivalents since the CATIE study was completed.

Investigators in the CATIE study— of which I was one—cannot be blamed for not focusing on neurotoxicity and neuroprotection; those data were not on the psychiatry’s radar when the CATIE study was designed in 1998. The major focus was on whether SGAs (new on the scene in the late 1990s) were more efficacious, safe, and tol­erable (that is, more effective) than FGAs.

In fact, the first study reporting that SGAs stimulated neurogenesis (in ani­mals) was published in 2002,3 when the CATIE study was more than half complete. Research into the neuropro­tective properties of SGAs then grew rapidly. In fact, the principal investi­gator of the CATIE study conducted a head-to-head comparison of FGA haloperidol and SGA olanzapine in a sample of first-episode schizophrenia patients4; over 1 year of follow-up, it was determined that patients in the haloperidol-treated group exhibited significant brain volume loss on MRI but those in the olanzapine-treated group did not. This study was pub­lished in 2005—the same year the CATIE study was published!

SGAs offer neuroprotection
Over the past decade, the neuroprotec­tive effects of SGAs5 and the neurotoxic effects of FGAs6 have been studied intensively, revealing that SGAs have multiple neuroprotective effects. These effects include:
• stimulation of the production of new brain cells (neurons and glia), known as neurogenesis5,7,8
• an increase in neurotrophic factors, such as nerve growth factor (NGF) and brain-derived neurotrophic factor (BDNF),9 which are found at a significantly low level in patients with psychosis10
• reversal of phencyclidine (PCP)-induced changes in gene expression11
• neuroprotection against ischemic stroke12-14
• reversal of PCP-induced loss of dendritic spines in the frontal cortex15
• prevention of oligodendrocyte damage caused by interferon gamma-stimulated microglia16,17
• reversal of loss of dendritic spines in the prefrontal cortex induced by dopamine depletion18
• an anti-inflammatory effect19,20
• protection against β-amyloid and hydrogen peroxide-induced cell death21
• protection against prefrontal cortical neuronal damage caused by dizocilpine (MK-801)22
• reversal of a PCP-induced decrease in the glutathione level and alteration of antioxidant defenses23
• protection of cortical neurons from glutamate neurotoxicity.24

One reason why SGAs are neuro­protective, but FGAs are not, can be attributed to their receptor profiles. FGAs block dopamine D2 receptors far more than serotonin 2A recep­tors, whereas SGAs do the opposite: They block 5-HT2A receptors 500% to 1,000% more than they block D2 recep­tors. This difference is associated in turn with a different neurobiological and neuroprotective profiles, such as a decrease or an increase in BDNF.25,26

Neither similar nor interchangeable
Since publication of the findings of the CATIE study, the primary investiga­tor has proposed that neuroprotection can be a therapeutic strategy to prevent neurodegeneration and neurodeterio­ration associated with schizophrenia.27 Given the preponderance of data show­ing that SGAs have numerous neuro­protective properties but FGAs have many neurotoxic effects,6 the message to psychiatric practitioners, a decade after the CATIE study, is that the 2 gen­erations of antipsychotic agents are not really similar or interchangeable. They might have similar clinical effective­ness but they exert very different neu­robiological effects.

Next Article: