OBSERVATION Functional Brain Imaging in Apraxia David A. Kareken, PhD; Frederick Unverzagt, PhD; Karen Caldemeyer, MD; Martin R. Farlow, MD; Gary D. Hutchins, PhD Background: An extensive literature describes struc- sults of the neurological examination were normal. There tural lesions in apraxia, but few studies have used func- was ideomotor apraxia of both hands (command, imita- tional neuroimaging. We used positron emission tomog- tion, and object) and buccofacial apraxia. The patient raphy (PET) to characterize relative cerebral glucose could recognize meaningful gestures performed by the metabolism in a 65-year-old, right-handed woman with examiner and discriminate between his accurate and awk- progressive decline in ability to manipulate objects, write, ward pantomime. The magnetic resonance image showed and articulate speech. moderate generalized atrophy and mild ischemic changes. Positron emission tomographic scans showed abnormal Objective: To characterize functional brain organiza- fludeoxyglucose F 18 uptake in the posterior frontal, tion in apraxia. supplementary motor, and parietal regions, the left af- fected more than the right. Focal metabolic deficit was Design and Methods: The patient underwent a neu- present in the angular gyrus, an area hypothesized to store rological examination, neuropsychological testing, mag- conceptual knowledge of skilled movement. netic resonance imaging, and fludeoxyglucose F 18 PET. The patient’s magnetic resonance image was coregis- Conclusions: Greater parietal than frontal physiologi- tered to her PET image, which was compared with the cal dysfunction and preserved gesture recognition are not PET images of 7 right-handed, healthy controls. Hemi- consistent with the theory that knowledge of limb praxis spheric regions of interest were normalized by calcrine is stored in the dominant parietal cortex. Gesture com- cortex. prehension may be more diffusely distributed. Results: Except for apraxia and mild grip weakness, re- Arch Neurol. 1998;55:107-113 APRAXIA IS a monly responsible for apraxia in both hands. Because right hemisphere lesions disorder of skilled movement not caused by are not known to induce apraxia,5,6 the weakness, akinesia, deafferentation, abnor- left hemisphere appears dominant for the mal tone or posture, movement disorders, in- tellectual deterioration, poor comprehension, storage and execution of learned move- or uncooperativeness.1 ments. Although both anterior and pos- terior lesions in the left hemisphere pro- Patients with ideomotor apraxia are un- duce apraxia, patients with parietal able to pantomime on command sym- lesions often have difficulty compre- bolic gestures (eg, saluting) or use of tools, hending gesture, and discriminating and their movements are characterized by between accurately and poorly per- errors in posture, spatial orientation, and formed pantomime gestures.7,8 Occipital joint coordination.2-4 Analogous move- lesions that disconnect visual association ment defects in the face, lips, tongue, areas from the parietal lobe may lead to From the Departments cheeks, larynx, and pharynx (eg, blow- impaired gesture imitation, but other- of Neurology (Drs Kareken ing out a match or whistling) are termed wise normal praxis.9 Although still and Farlow), Psychiatry buccofacial apraxia.1 Although a number unable to execute gesture, patients with (Drs Kareken and Unverzagt), of studies have described the structural le- frontal lesions are better able to accu- and Radiology (Drs Caldemeyer and Hutchins), Indiana sions that result in apraxia, little work has rately discriminate good from bad panto- University School of Medicine, described the functional network of brain mime, and comprehend gesture in oth- 7,8 7,10 and the Richard L. Roudebush regions that become compromised. ers. Some authors therefore propose Veterans Affairs Medical Center Lesion research suggests that left that the left supramarginal and angular (Dr Kareken), Indianapolis. intrahemispheric lesions are most com- gyri contain visuokinesthetic motor pro- ARCH NEUROL / VOL 55, JAN 1998 107 ©1998 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 SUBJECTS, DESIGN, AND glucose uptake was measured with 2-fludeoxyglucose F 18 (10-mCi intravenous bolus), and images were recon- METHODS structed to 9-mm spatial resolution. The PET images for all subjects were spatially registered in the horizontal plane PATIENT AND SUBJECTS along the anterior commissure–posterior commissure line. A whole brain proton density–weighted MRI scan of the The patient was a 65-year-old, high school–educated, right- patient’s head was acquired in the axial plane, and coreg- handed woman who presented to the Indiana Alzheimer istered with her PET image volume using an iterative pro- Disease Center in Indianapolis. Approximately 4 years ear- cedure that minimizes the least-squared difference be- lier, she experienced the insidious onset of difficulty speak- tween the MRI and PET image volumes. Magnetic resonance ing and writing. These symptoms gradually progressed, and images were unavailable for the controls. The regions of she became unable to properly manipulate silverware and interest analyses were performed on the PET images by one small items (eg, buttons or jewelry). She gradually be- of us (D.A.K.), normalizing hemispheric regions of inter- came unable to perform her household chores, including est by the average of left and right calcrine cortex. management of family finances. At the time of her clinic Five regions of interest were delineated: anterior fron- visit, she had difficulty finding and articulating words, but tal (FA) (corresponding to the dorsolateral prefrontal cor- continued to read and understand television programs. tex), posterior frontal (FP) (corresponding to premotor and Seven healthy control subjects from the Indiana Alz- primary motor regions), parietal, caudate-putamen (CP), heimer Disease Center–PET database were selected for im- and supplementary motor area (SMA). All regions were di- aging comparison (mean age, 69 years; age range, 56-79 vided into left (L) and right (R). The central sulcus was ap- years). The controls had no evidence of cerebral or neu- proximated by a line just posterior to the thalamus as viewed ropsychological abnormality. Informed consent was ob- on the midsagittal image (Figure 1, A, plane a). Al- tained from all subjects after the nature of the procedure though this landmark confounds slightly the actual divi- had been fully explained. sion between frontal and parietal cortex, it was a reliable landmark. The division between FA and FP was defined MATERIALS AND PROCEDURE on the axial image with a line through the center of the 2 caudate heads. This was an easily identifiable, reliable land- Clinical Examinations mark, and constitutes the approximate division between premotor cortex (Brodmann area 6) and the cortex ante- The patient underwent neurological and neuropsychologi- rior to it28 (Figure 1, A, plane c). Insular cortex was ex- cal examinations (Table 1 and Table 2). Apraxia exami- cluded from FP and cingulate cortex from FA. For the pa- nation involved testing gesture production and compre- rietal regions of interest, the cuneus and lateral occipital hension. To test production, the patient was asked to gyri were excluded in slices inferior to the parieto- pantomime on command 10 intransitive gestures (waving occiptial sulcus (Figure 1, A, plane d). Tracing of frontal goodbye, hitchhike, salute, beckon “come here,” signal stop, and parietal regions of interest began on an axial plane just finger to lips for “shsh,” finger tapping, “ok” sign, 2 “ok” dorsal to the thalamus (Figure 1, C, plane e). On the coro- signs linked together, and the “peace” or “V” for victory nal images this plane passed through both sylvian fissures sign), 8 transitive gestures (opening a door with a key, flip- (Figure 1, B, plane e). Tracings were made of the cortical ping a coin, opening a ketchup bottle, stirring coffee with ribbon with a computer mouse on the axial images at each a spoon, and using a screwdriver, hammer, scissors, and of 5 planes dorsal to the thalamus, with an interplane in- handsaw), and 8 buccofacial movements (cough, sniff, puff terval of 5 pixels (<10 mm) (Figure 1, C). out cheeks, wink, stick out tongue, blow a kiss, blow out a The SMA was approximated with a right triangle on match, and suck on a straw). To test gesture comprehen- the sagittal image (Figure 1, C, shaded area). The vertical sion, the patient was asked to discriminate 4 target ges- line of the right angle (posterior boundary) was placed at tures from foils (using a hammer, waving goodbye, shoot- the anterior-posterior division line, and the horizontal line ing a basketball, and flipping a coin) and 4 precise target (ventral boundary) was placed at the midpoint between the gestures from the same gestures performed clumsily and corpus callosum and the vertex of the cerebrum. The hy- with a body part as object (brushing teeth and using a ham- potenuse followed the cortex along the edge of the inter- mer, scissors, and key). hemispheric sulcus. As defined, SMA probably includes most of Brodmann area 6 and at least some of Brodmann area 8 Imaging (see reference 28). Two planes of SMA (one on the mesial surface of the interhemispheric sulcus, one 4-mm lateral) The PET studies were performed on a whole body PET to- were traced in each hemisphere and averaged. Tracings of mograph (Siemens 951/31R, 6-mm full-width half- CP were performed