Hypofrontality in schizophrenia: distributed dysfunctional circuits in neuroleptic-naïve patients

Summary

Background

There have been reports that patients with schizophrenia have decreased metabolic activity in prefrontal cortex. However, findings have been confounded by medication effects, chronic illness, and difficulties of measurement. We aimed to address these problems by examination of cerebral blood flow with positron emission tomography (PET).

Methods

We studied 17 neuroleptic-naïve patients at the early stages of illness by means of image analysis and statistical methods that can detect abnormalities at the gyral level.

Findings

An initial omnibus test with a randomisation analysis indicated that patients differed from normal controls at the 0·06 level. In the follow-up analysis, three separate prefrontal regions had decreased perfusion (lateral, orbital, medial), as well as regions in inferior temporal and parietal cortex that are known to be anatomically connected. Regions with increased perfusion were also identified (eg, thalamus, cerebellum, retrosplenial cingulate), which suggests an imbalance in distributed cortical and subcortical circuits.

Interpretation

These distributed dysfunctional circuits may form the neural basis of schizophrenia through cognitive impairment of the brain, which prevents it from processing input efficiently and producing output effectively, thereby leading to symptoms such as hallucinations, delusions, and loss of volition.

Introduction

In 1974, Ingvar and Franzen coined the term hypofrontality to denote the relative decrease they found in their calculation of a ratio of frontal to post-central blood flow in schizophrenia patients.¹ Since then, hypofrontality in schizophrenia has been studied by a range of methods: measurement of regional blood flow with xenon by use of cortical probes;2, 3 measurement of cerebral blood flow and glucose metabolism by positron emission tomography (PET);4, 5, 6, 7 neuropathological studies of tissue taken at necropsy;⁸ anatomical measurements by magnetic resonance (MR) imaging;9, 10 animal models of working memory;¹¹ and multiple techniques of cognitive psychology and neuropsychology.12, 13 Although these studies have broadly suggested prefrontal dysfunction in schizophrenia, the findings are not consistent, and Gur and Gur¹⁴ have argued that the concept of hypofrontality should be questioned.

There are several difficulties in the study of hypofrontality in schizophrenia¹⁵—most importantly, perhaps, the anatomical extent, complexity, and functional diversity of the frontal cortex itself, which includes both neocortex and more primitive limbic cortex, and carries out functions that range from working memory to regulation of eye movements. Through lesion studies, three functional areas have been identified (orbital, dorsolateral, and medial¹⁶), though these subdivisions are an oversimplification; the term hypofrontality is thus too general, since it fails to reflect the diversity of specialised regions in frontal cortex.

The effects of stage of illness (ie, first episode vs chronically ill) are contested. Early studies of chronically ill patients generally reported decreases in frontal flow in schizophrenia during rest or a specific cognitive task1, 4, 5, 17 (though some were negative).14, 18 Studies of first-episode or never-treated patients are less common; some have supported hypofrontality,¹⁵ whereas others have found no differences.6, 19 A few influential studies have suggested the opposite finding—namely, that first-episode patients have increases in frontal flow.20, 21 If true, these findings may suggest that metabolic activity in the prefrontal cortex decreases over time as a result of a degenerative process, toxic effects of schizophrenia, or effects of medications to treat the illness.

Medication effects on cerebral blood flow are a confounding factor that cannot easily be separated from stage of illness. The results of Berman,²² Weinberger,¹⁷ and their colleagues suggest that hypofrontality was unrelated to treatment in chronically ill patients. By use of a more powerful longitudinal design in a sample of five drug-naïve patients, Szechtman and colleagues²¹ found that 1 year of treatment increased metabolism in the basal ganglia, but did not affect frontal metabolism. However, when they used a cross-sectional design to compare a larger group of patients who had been receiving medication for 1 year with a group who had been treated for an average of 7 years, Szechtman and colleagues found relative hypofrontality in the chronically treated patients, which suggests that long-term neuroleptic treatment may produce a decrease in frontal metabolic tone.

Cognitive activity during image acquisition may also have substantial impact on the assessment of hypofrontality. Most early studies that supported hypofrontality involved examination of patients in a “resting state”.2, 6, 7 But the word “rest” is misleading, since the brain does not become inactive or empty of thought in the absence of specific experimental tasks or instructions; on the contrary, patients report after scans that when at “rest” they typically recalled past experiences or made future plans, both of which require prefrontal activity.²³ Other investigators tried to “stress test” the prefrontal cortex by use of cognitive challenges.15, 17, 24, 25 Although this approach has the advantage of standardising mental activity, the hypofrontality reported is difficult to interpret because schizophrenic patients typically carry out the tasks less well than healthy volunteers. There is thus a “chicken and egg” problem: does the decrease in frontal blood flow or metabolism result from an intrinsic neural deficit, or from lack of motivation and poor task performance? Other investigators have attempted to answer this question by matching task performance in patients and controls.²⁵

Symptoms may also influence measurements of cerebral blood flow or metabolism. For example, negative symptoms have been related to hypofrontality;1, 7, 15 and a study of a small group of patients who experienced auditory hallucinations during scanning indicated flow increases in multiple brain regions.²⁶

The concept of hypofrontality raises important questions about the fundamental nature of schizophrenia—its neural substrates, its course, and its treatment. Do specific regions of the prefrontal cortex have abnormal metabolism or blood flow? And, if so, is this abnormality related to symptoms, cognitive state, or treatment effects? Most importantly, are toxic medication or toxic illness effects on the prefrontal cortex a valid concern?

To eliminate confounding factors that arise from treatment and chronicity, we focused on a sample of neuroleptic-naïve patients studied early in the course of their illness, and compared them with a group of healthy volunteers closely matched for sex, age, and parental education. To eliminate the confounding effect of poor task performance, we studied them during the resting state. To gauge the relation between symptoms and blood flow, we measured symptoms just before the scan. To assess whether specific subregions of the prefrontal cortex are selectively dysfunctional, we analysed data by image analysis and statistical techniques that enabled us to identify abnormalities at the gyral level.