ORIGINAL ARTICLE Antipsychotic Drug Effects on Brain Morphology in First-Episode Psychosis Jeffrey A. Lieberman, MD; Gary D. Tollefson, MD, PhD; Cecil Charles, PhD; Robert Zipursky, MD; Tonmoy Sharma, MD; Rene S. Kahn, MD, PhD; Richard S. E. Keefe, PhD; Alan I. Green, MD; Raquel E. Gur, MD, PhD; Joseph McEvoy, MD; Diana Perkins, MD, MPH; Robert M. Hamer, PhD; Hongbin Gu, PhD; Mauricio Tohen, MD, DrPH; for the HGDH Study Group Background: Pathomorphologic brain changes occur- Participants: Patients with first-episode psychosis (DSM- ring as early as first-episode schizophrenia have been ex- IV) and healthy volunteers. tensively described. Longitudinal studies have demon- strated that these changes may be progressive and Interventions: Random allocation to a conventional an- associated with clinical outcome. This raises the possi- tipsychotic, haloperidol (2-20 mg/d), or an atypical an- bility that antipsychotics might alter such pathomorpho- tipsychotic, olanzapine (5-20 mg/d). logic progression in early-stage schizophrenia. Main Outcome Measures: Brain volume changes as- Objective: To test a priori hypotheses that olanzapine- sessed by MRI. treated patients have less change over time in whole brain gray matter volumes and lateral ventricle volumes than Results: Of 263 randomized patients, 161 had baseline and haloperidol-treated patients and that gray matter and lat- at least 1 postbaseline MRI evaluation. Haloperidol-treated patients exhibited significant decreases in gray matter vol- eral ventricle volume changes are associated with changes ume,whereasolanzapine-treatedpatientsdidnot.Amatched in psychopathology and neurocognition. sample of healthy volunteers (n=58) examined contempo- raneously showed no change in gray matter volume. Design: Longitudinal, randomized, controlled, mul- tisite, double-blind study. Patients treated and followed Conclusions: Patients with first-episode psychosis exhib- up for up to 104 weeks. Neurocognitive and magnetic ited a significant between-treatment difference in MRI vol- resonance imaging (MRI) assessments performed at weeks ume changes. Haloperidol was associated with significant 0 (baseline), 12, 24, 52, and 104. Mixed-models analy- reductions in gray matter volume, whereas olanzapine was ses with time-dependent covariates evaluated treatment not. Post hoc analyses suggested that treatment effects on effects on MRI end points and explored relationships be- brain volume and psychopathology of schizophrenia may tween MRI, psychopathologic, and neurocognitive out- beassociated.Thedifferentialtreatmenteffectsonbrainmor- comes. phology could be due to haloperidol-associated toxicity or greater therapeutic effects of olanzapine. Setting: Fourteen academic medical centers (United States, 11; Canada, 1; Netherlands, 1; England, 1). Arch Gen Psychiatry. 2005;62:361-370 TRUCTURAL BRAIN ABNOR- ume”) changes were progressive over malities have been exten- time and related to the illness course and sively and consistently treatment outcome of patients with first- described in patients with episode,4-9 chronic,10,11 and childhood- schizophrenia.1-3 This patho- onset12,13 schizophrenia. These findings Smorphologic finding has been most suggest that although schizophrenia commonly demonstrated as brain vol- may arise from a neurodevelopmental ume differences involving the ventricu- diathesis, its pathophysiology may be lar system and cortical and subcortical progressive after the onset of illness.14,15 gray matter regions in patients with They also raise the question of what schizophrenia compared with matched role medication may have in mitigating Author Affiliations are listed at the end of this article. healthy volunteers. Longitudinal studies schizophrenia-associated pathomorpho- Group Information: A list of using high-resolution magnetic reso- logic changes or, alternatively, contrib- the members of the HGDH nance imaging (MRI) to examine brain uting to such changes. Preclinical stud- Study Group appears on page structure have found that MRI volume ies have suggested the possibility of 368. (hereafter referred to simply as “vol- specific atypical antipsychotic drugs hav- (REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, APR 2005 WWW.ARCHGENPSYCHIATRY.COM 361 ©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 ing pharmacologic properties that could produce neu- bilizers were not allowed in the first 12 weeks of the study but rotrophic, neurogenetic, or neuroprotective effects.16-22 could be added if clinically indicated thereafter. We addressed these questions in a controlled trial of first-episode psychosis, in which patients were random- EFFICACY ASSESSMENTS ized to either a conventional antipsychotic (haloperi- dol) or an atypical antipsychotic (olanzapine). We hy- Patients were assessed by MRI at weeks 0 (baseline), 12, 24, pothesized that (1) olanzapine-treated patients would have 52, and 104. Patients who dropped out were assessed by MRI less change over time in whole brain gray matter (WBGM) until the point at which they dropped out. A very small num- volumes and lateral ventricle volumes than haloperidol- ber of subjects (12) missed a scheduled MRI assessment but treated patients; (2) decreases in gray matter volumes and did not drop out and were assessed at the next scheduled MRI increases in lateral ventricle volumes over time would be assessment. All MRI studies were performed with a 1.5-T MRI system. Six of the 8 imaging sites used Signa scanners (Gen- associated with less improvement on measures of psy- eral Electric Co, Milwaukee, Wis), and 2 sites used a Gyro- chopathology and neurocognitive functioning; and (3) scan scanner (Philips Medical Systems, Best, the Nether- caudate nuclei volumes would be differentially affected lands). Two imaging centers studied subjects from more than by treatment, increasing in response to haloperidol and 1 clinical site. Each subject laid supine with his or her head with little or no change in response to olanzapine. situated in the head holder for the radiofrequency coil. Two image-intensity standards were placed on either side of the sub- ject’s head. The imaging protocol included 3-dimensional T1- METHODS weighted, inversion recovery–prepared spoiled gradient- recalled acquisition in steady state images (0.94ϫ0.94ϫ1.50 This longitudinal study was conducted from March 1, 1997, mm, axial direction) and contiguous proton density and T2- to July 31, 2001, at 14 academic medical centers (11 in the United weighted fast spin-echo images (0.94ϫ0.94ϫ3.00 mm, axial States, 1 in Canada, 1 in the Netherlands, and 1 in England). slicing direction). Quality-control scans were performed twice a month on each MRI system with standardized imaging phan- PATIENTS toms. The imaging component of the study was centrally co- ordinated (by one of us [C.C.] at Duke Imaging and Analysis Patients who presented to clinical services (inpatient, emer- Laboratory, Durham, NC) and included site training, MRI sys- gency, and outpatient) for evaluation and treatment of psy- tem monitoring, and centralized data analysis. The coordinat- chotic symptoms were enrolled in the study if they met the fol- ing center remained blinded throughout the trial. Every data lowing inclusion and exclusion criteria: age 16 to 40 years; onset set was processed by means of an automated, atlas-based, mul- of psychotic symptoms before age 35 years; diagnosis of schizo- tichannel brain-tissue segmentation program that generates de- phrenia, schizophreniform, or schizoaffective disorder accord- tailed maps of gray matter, white matter, and cerebrospinal ing to DSM-IV criteria (as assessed with the Structured Clini- fluid.24 This processing includes a bias-field correction that ad- cal Interview for DSM-IV, Research Version23); previous justs for intensity inhomogeneities in the data sets. On the ba- antipsychotic drug treatment of more than 16 cumulative weeks, sis of Talairach coordinates, a 3-dimensional brain atlas was or treatment with clozapine at any time in the patient’s life- divided into 16 discrete boxes (parcellated volumes) for re- time; no current substance dependence (except caffeine and nico- gional measurements. The image sets were aligned to the atlas tine) by DSM-IV within 1 month before study entry; no cur- before the automated, atlas-based, multichannel segmenta- rent indication of serious suicidal risk; female subjects not tion was applied. Volumes of gray and white matter and cere- pregnant or nursing; premorbid IQ of 70 or more; no require- brospinal fluid were extracted from each of the 16 parcellated ment of concurrent treatment with anticonvulsants, benzodi- volumes. Anatomically guided combinations of these volumes azepines (except as allowed for agitation and control of extra- approximate the frontal, temporal, parietal, and occipital lobes. pyramidal symptoms), antidepressants, psychostimulants, or A rater-guided connectivity-based masking method was used other antipsychotic drugs at study entry; and no contraindica- to separate the lateral ventricles from voxel-based cerebrospi- tion for neuroimaging per current regulations from the local nal fluid segmentation. Caudate volumes were obtained with regulatory agency (eg, metal prostheses). Each patient (or a pa- manual outlining. Rigorous standardization and quality- tient’s authorized legal representative) had to understand the control procedures were used, and reliability of measure- nature of the study and sign an informed consent document. ments across sites was