The relationship of regional frontal hypometabolism to executive function: A resting fluorodeoxyglucose PET study of patients with epilepsy and healthy controls

Abstract

Executive dysfunction is common in patients with frontal lobe damage and may depend on the location of pathology within the frontal lobes. However, it is unclear how specific brain regions contribute to different aspects of executive functioning. Eighteen patients with frontal lobe epilepsy, 10 patients with juvenile myoclonic epilepsy, and 14 controls completed a series of tests that measure a broad range of executive functions. Resting fluorodeoxyglucose positron emission tomography scans were collected and regional cerebral rates of glucose uptake values were regressed on test scores. Results revealed that frontal lobe metabolic values were strong predictors of executive functioning in patients with epilepsy, but not in healthy controls. However, nonfrontal regions also contributed unique variance on several measures, suggesting that (1) a network of frontal and nonfrontal regions subserve many executive functions and (2) resting hypometabolism can be a useful predictor of executive dysfunction in patients with epilepsy.

Introduction

Neuropsychological testing is a routine part of the presurgical evaluation of patients with intractable epilepsy. Such evaluations are designed to assist with seizure localization and lateralization, as well as to provide an estimate of current cognitive ability. These evaluations have traditionally focused on clarifying the nature of the relationship between temporal lobe dysfunction and impairments in language and memory in patients with temporal lobe epilepsy (TLE). As a result, the contribution of cortical brain regions to other cognitive domains has been largely ignored.

In recent years, the importance of evaluating executive functioning in patients with epilepsy has been recognized and proven useful for differentiating patients with frontal lobe epilepsy (FLE) from those with TLE [1], [2]. That is, greater impairments are observed in patients with FLE relative to those with a nonfrontal seizure focus, presumably because of the importance of prefrontal regions in executive functioning [3], [4]. The most commonly assessed executive functions include verbal and nonverbal fluency, cognitive flexibility, response inhibition, and working memory/attention. Although there is some evidence from lesion and neuroimaging studies that different areas within the prefrontal cortex mediate different executive skills (see Table 1), the literature is mixed and consistent relationships have not been established in patients with epilepsy or healthy controls.

To date, even less is known about the relationship between regional brain metabolism and performance on neuropsychological measures of executive functioning [5]. Previous studies have demonstrated that regions of brain dysfunction indicated by fluorodeoxyglucose positron emission tomography (FDG-PET) are well-correlated with neuropsychological performance in patients with epilepsy undergoing presurgical evaluations [6], [7], [8], [9], [10], [11]. Furthermore, there is a growing body of literature demonstrating a relationship between regional resting hypometabolism and cognitive dysfunction in other neurological disorders [5], [12], [13]. To our knowledge, no studies have examined the contribution of frontal and nonfrontal brain regions to a wide variety of executive functions in patients with epilepsy and healthy controls. Including nonfrontal regions is important because there is strong evidence that many tests of executive functioning recruit a widespread network of brain regions within the left and right hemispheres [14], [15], [16], [17]. In addition, patients with epilepsy often exhibit diffuse brain dysfunction on structural imaging, even when the origin of the seizures appears to be focal [18]. Understanding the relationship between regional brain hypometabolism and executive dysfunction is important in that it may help to guide treatment decisions in patients with epilepsy (i.e., surgery), especially in those with a frontal lobe seizure focus.

This study represents an exploratory investigation of the relationships between interictal regional cerebral rates of glucose uptake (rCMRGlc) and specific neuropsychological tests of executive functioning in patients with FLE, juvenile myoclonic epilepsy (JME), and healthy controls. Patients with JME were selected for the study because they represent a group of epilepsy patients who have been shown to exhibit mild executive dysfunction [19], [20], as well as frontal lobe hypometabolism, on PET [21], but whose cognitive and metabolic dysfunction is not as extensive as that seen in patients with focal FLE [22]. Several broad hypotheses were made with respect to clinicometabolic relationships within the groups. Based on previous literature on patients with frontal lobe dysfunction, we hypothesized that frontal lobe hypometabolism would be the best predictor of executive dysfunction in the FLE group, but that nonfrontal brain regions (i.e., limbic, including anterior cingulate) would also contribute to test scores on some measures (e.g., Wisconsin Card Sorting Test). We also predicted that left frontal lobe metabolism would be the strongest predictor of executive functions with a high verbal component (e.g., verbal fluency; digit span), whereas right frontal metabolism would be the strongest predictor of measures with a high visuospatial component (e.g., figural fluency, spatial span). However, material specificity has not always been demonstrated in patients with frontal lobe dysfunction (see [23]). Therefore, we predicted that a combination of left and right frontal lobe regions would contribute to performance on some measures. It was also predicted that frontal lobe hypometabolism would be associated with performance to a lesser degree in patients with JME, whereas these relationships were not expected to emerge in controls. The latter hypothesis was based on previous literature demonstrating a relationship between resting PET and cognitive performance in patients with neurological disease, but not in healthy controls [13].