Cerebral Glucose Utilization in Motor Neuron Disease John

M.

Hoffman, MD; John

C. Mazziotta, MD,

emission tomography with fludeoxyglucose F 18 (18F-fluorodeoxyglucose) was used to examine regional cerebral glucose metabolism in individuals with motor neuron \s=b\ Positron

disease. Motor neuron disease involves selective loss of molarge pyramidal cells in the motor cortex, and corticospinal tract degeneration. We postulated that the local cerebral metabolic rate of glucose should correlate with this regional neuronal cell loss. Glucose metabolism values in patients with motor neuron disease were reduced compared with those of controls in several regions; however, when corrected for multiple comparisons, no significant difference was observed between patients with motor neuron disease and age-matched controls. No correlation was noted between the local cerebral metabolic rate of glucose and duration or severity of illness. Correlation between metabolic changes with objective findings on neurologic examination, including motor weakness and tendon reflexes, provided interesting results, including a decline in glucose metabolism with progressive weakness and upper motor neuron dysfunction. Moreover, in supplementary motor areas, there appears to be an increase in regional glucose metabolism as the neurologic condition deteriorates, possibly representing increased metabolic activity of the motor association cortex in response to primary loss of pyramidal cells. (Arch Neurol. 1992;49:849-854) tor neurons,

disease (MND) is heterogeneous group Motoroftheprogressive degenerative diseases clinically af¬ the neuron

a

voluntary motor system. Pathologically, ce¬ rebral cortex, pyramidal structures, brain stem, anterior horn cells, and spinal cord can be involved.1"3 Amyotrophic lateral sclerosis (ALS), a specific form of MND, effects, to varying degrees, spinal and bulbar lower motor neurons

PhD; Thomas

Accepted for publication March 23,

1992. From the Division of Nuclear Medicine and Biophysics (Drs Hoffman and Mazziotta and Messrs Hawk and Sumida) and the Department of Neurology (Drs Hoffman and Mazziotta), UCLA School of Medicine. Dr Hoffman and Mr Hawk are now with the Division of Nuclear Medicine,

Department of Radiology, Duke University Medical Center, Durham, NC.

Reprint requests to the Division of Nuclear Medicine, Box 3949, Duke University Medical Center, Durham, NC 27710 (Dr Hoffman).

Hawk;

Ron Sumida

ity, fasiculations, weakness, reflex changes, evidence of cor¬

ticospinal tract dysfunction, and intact sensory function.2 Despite extensive epidemiologie and scientific investi¬ gations, no cause of MND has been identified.10 The differential diagnosis of syndromes causing similar clini¬ cal findings includes toxin exposure (eg, mercury,11 lead, or tricreosyl phosphate poisoning), remote effects of neo¬ plasia,12 hypoglycemia,13,14 and cervical spondylosis. In their recently published excellent review, Mitsumoto et al5 discuss historical, clinical, epidemiologie, biochemical, metabolic, and immunologie theories concerning the cause of MND. With a complete neurologic examination, appro¬ priate laboratory studies, and confirmation of denervation changes (ie, electromyography and/or biopsy), indicating widespread anterior horn cell disease, the diagnosis can be made with confidence. Various therapies have been pro¬ posed; however, none has proved to be efficacious.15"19 Herein, we correlate known neuropathologic changes

observed in MND with the local cerebral metabolic rate for glucose (LCMRGlc) using fludeoxyglucose F 18 (18F-fluorodeoxyglucose) (FDG) and positron emission to¬ mography (PET). Because LCMRGlc reflects neuronal synaptic metabolic activity in healthy adults20,21 and be¬ cause MND involves selective loss of motor neurons, large pyramidal cells in motor cortex, and corticospinal tract degeneration,1,2,22 we hypothesized that reduced LCMRGlc should parallel this regional neuronal cell loss.

PATIENTS, SUBJECTS,

fecting

(LMNs) and upper motor neurons (UMNs.)4 The world¬ wide incidence is approximately 0.4 to 1.8 per 100 000, with a 1 to 1.6:1 male-to-female ratio.5 Median age at onset is 55 years, with death typically occurring 3 to 5 years later. Bul¬ bar involvement is associated with a more rapid decline.5"9 Individuals with purely spinal LMN involvement (pro¬ gressive muscular atrophy) often have a less rapid clinical deterioration. Symptoms of ALS can include weakness, cramping, and observable muscle twitching or fasiculation. With bulbar involvement, dysarthria and dysphagia are prominent problems. Death typically occurs from pulmo¬ nary complications. Clinical signs include atrophy, spastic-

C.

Patients and

AND METHODS

Subjects

We performed 10 FDG-PET studies in seven patients (six male and one female) aged 37 to 70 years (mean[±SD], 57.3±12.8 years). Three men underwent a repeated study approximately 1 year after the initial FDG-PET examination. Six of the seven sub¬ jects had classic ALS with both UMN and LMN involvement. One subject had strictly LMN findings, and three subjects had associ¬ ated bulbar involvement (Table 1). All subjects were otherwise healthy and without other sub¬ stantial medical illnesses. The patients with MND all underwent thorough neurologic examinations. Standard grading of motor strength,23 tendon reflexes, and corticospinal tract function was used. Patients had electromyographic evidence of MND, a normal

myelogram, and/or a spinal magnetic resonance imaging scan,

and were diagnosed independently by two neurologists. In four of the seven subjects, computed tomographic or magnetic reso¬ nance imaging brain scans were normal. None of the subjects were receiving medication at the time of the FDG-PET study. Pa¬ tients with MND were compared with an age-matched control population of 11 neurologically normal subjects aged 37 to 69 years (mean [±SD], 56.7±11.85 years; six male and five female). None of the patients with MND showed evidence of respiratory compromise or nutritional abnormalities, such as dehydration or ketosis at the time of the FDG-PET study. PET Studies Subjects were placed in the recumbent position with their eyes and ears open in a room with dim fluorescent indirect lighting and

Downloaded From: http://archneur.jamanetwork.com/ by a UQ Library User on 06/20/2015

Table

Study No./Age, y/Sex 1/64/M 2/65/M 3/40/M 4/41/M 5/69/M 6/70/M 7/37/M 8/66/F 9/62/M 10/59/M

1.—Neurologie

Status in

Subjects With

Upper Motor Neuron Involvement

Bulbar (Cranial Nerve) Involvement

Yes

No

Yes

No

Yes

No

Yes

No

Yes

No

Yes

No

Yes

Yes

No

No

Yes

Yes

Yes

Yes

"Three of the seven total patients were studied twice. All tSee text for explanation of severity rating scale.

patients had a normal nutritional status;

background fan noise from the tomograph. During the 40-minute FDG uptake period, the subjects remained quiet, and verbal in¬ teractions were avoided. All intravenous lines were placed 10 to

15 minutes before the injection of the tracer. "Arterialized" blood sampling was used to obtain the FDG plasma in¬ put function.24 All subjects were studied in accordance with the policies of the UCLA Human Subject Protection Committee. Fluorine-18 was produced in the UCLA cyclotron, and FDG was synthesized by a semiautomated method.25 The injected in¬ travenous dose of FDG was 222 to 370 MBq. Scans were obtained with the NeuroECAT tomograph (CTI, Knoxville, Tenn). The full width at half-maximum image resolution was 9.5 mm in the im¬ age plane and 12.4 mm in the axial plane. Two to three million counts per image plane were obtained.26 Attenuation correction was performed with use of the geometric method, as has been previously described.27,28 A total of 12 cross-sectional tomographic planes were obtained parallel to the canthomeatal line with 8-mm venous

center-to-center spacing. Metabolic rates for glucose

were determined for specific neuroanatomic regions, as previously described.24,29,30 The LCMRGlc of regions of interest (ROIs) for subcortical, cortical, and whitematter zones known to be part of the motor system were deter¬ mined.30 Structures that spanned more than one tomographic plane (eg, thalamus and caudate nucleus) were assigned threedimensional values for LCMRGlc by the following scheme. Both the average planar metabolic rate and cross-sectional area of the region in each tomographic plane were determined. The threedimensional metabolic rate for glucose was weighted by the crosssectional area of a given plane according to the following equation:

MRW=2)(MR)¡(A),/A,2 where MRW indicates three-dimensional weighted (by area) met¬ abolic rate for a structure; MR¡, the planar metabolic rate value of a structure in plane i; and A¡, the cross-sectional area of the struc¬ ture in plane i ( is the number of planes that include the struc¬ ture). With this technique, weighted three-dimensional LCMRGlc values were obtained for all neuroanatomic structures of interest. Weighted normalized values were obtained by dividing the weighted LCMRGlc of each neuroanatomic structure by the ipsi¬ lateral cerebral hemispheric weighted LCMRGlc. Weighted nor¬ malized values were determined for 17 different bilateral ROIs

(Table 2).

Statistical

Motor Neuron Disease*

Analysis

It was hypothesized that the severity of MND symptoms would

correlate with progressive neuronal degeneration and inversely with the LCMRGlc. To test this hypothesis, a severity rating scale was devised, in which 0 indicated normal; 1, signs in

neurologic

Severityt

in several

patients,

Duration of Motor Neuron Disease, y

disease

progressed slowly.

two extremities; 2, "more pronounced" signs than ob¬ served in 1; 3, same as 2 plus bulbar involvement or all extrem¬ ities involved; 4, more pronounced than 3 and difficulty with ambulation; 5, requires assistance with transfers; 6, nonambulatory; 7, nonambulatory with bulbar and/or respiratory involve¬ ment; 8, bedridden with profound loss of motor function; 9, ven¬ tilator dependent with minimal motor function; and 10, ventilator dependent with no motor function. In this study, severity ratings ranged from 2 to 7. In patients studied on two occasions, disease progressed during the 1-year follow-up period. A Spearman Rank-Order Correlation analysis was used to determine the effects of severity of disease as well as duration of disease on the LCMRGlc for the 17 ROIs. Other statistical studies performed in¬ cluded a two-tailed Student f test with Bonferroni correction to compare the mean regional LCMRGlc in patients with MND with that of the control population. Analysis of variance was used to examine possible relationships between LCMRGlc and objective neurologic signs, such as motor strength and tendon reflexes. For the purposes of this analysis, each extremity was given a motor to the weakest muscle group in that strength value extremity. Likewise, the tendon reflex value corresponding to the most pathologic reflex noted in the entire extremity was used. It should be noted that even though the most pathologic muscle group or reflex was used to grade the entire extremity in our sub¬ jects, each extremity was fairly uniform in its involvement (ie, motor strength and deep tendon reflexes). As previously noted (Table 1), all but one subject had evidence of UMN involvement, which was defined as increased reflexes, extensor plantar responses, or increased tone. one or

corresponding

RESULTS FDG-PET Typical images from studies numbered 1 and 10 (from Table 1) are shown in Fig 1. Weighted normalized cerebral metabolic rates of glucose for the patients with MND and controls are shown in Table 2. Values for the to¬ tal number of studies (n=10) rather than individuals (n=7) are shown. The LCMRGlc values in patients with MND were different from those in controls for the frontal white matter, right superior cerebellum, right precentrai gyrus, left middle temporal gyrus, head of caudate, and dentate of the cerebellum. Following a Bonferroni correction for

multiple comparisons (P=.05/17 ROIs=.003), however, there were no significant differences between weighted

normalized LCMRGlc values in patients with MND and age-matched controls. Analyses performed to correlate metabolic changes with objective findings on neurologic examination are summa¬ rized in Fig 2. The precentrai gyrus and middle frontal gy-

Downloaded From: http://archneur.jamanetwork.com/ by a UQ Library User on 06/20/2015

Table 2.—Cerebral Metabolic Rate of Glucose* Brain

Region

Right

MND Left

Control Left

1.02: :0.05

1.03: :0.03

0.96: 0.05

1.01: 0.04

0.98: :0.10

1.01: 0.08

0.95: 0.11

0.99: 0.09

1.15: :0.13

1.06: 0.14

1.11: 0.14

1.08: 0.14

1.12: :0.08

1.07: 0.07

1.10: 0.08

1.07: 0.05

1.11: 0.08

1.06: 0.03

1.03: 0.07

1.00: 0.05

0.97: :0.10

0.94: 0.11

0.95: 0.09

0.95: 0.09

1.04: :0.05

1.04: 0.07

1.02: 0.06

1.02: 0.05

1.02: :0.07

0.92: 0.05

1.00: 0.11

0.92: 0.05

1.20: :0.05

1.29: 0.05

1.25: 0.06

1.33: 0.04

1.22: 0.08

1.20: 0.14

1.20: 0.10

1.21: 0.10

1.23: 0.09

1.28: 0.06

1.26: 0.09

1.30: 0.06

0.85: 0.13

0.97: 0.04 0.94: 0.04

MND

Control

Right

Precentrai gyrus Lateral superior frontal gyrus Medial superior frontal gyrus Middle frontal gyrus Inferior frontal gyrus Superior parietal lobule Inferior parietal lobule Middle temporal gyrus Head of caudate Body of caudate Thalamus Vermis-cerebellum Dentate-cerebellum

0.81 ±0.09

0.93±0.04

0.81: 0.11

Superior lobule cerebellum

0.72±0.09

0.80±0.3

0.70: 0.9

0.81: 0.05

0.95: 0.08

0.93: 0.07

0.71: 0.09

0.78: 0.04

Midbrain Pons 0.77±0.03

0.70±0.65

0.72: 0.03 Frontal white matter 'Values (mean±1 SD) are weighted normalized (see text for explanation). MND indicates motor neuron disease. A significant (P

Cerebral glucose utilization in motor neuron disease.

Positron emission tomography with fluorodeoxyglucose F 18 (18F-fluorodeoxyglucose) was used to examine regional cerebral glucose metabolism in individ...
2MB Sizes 0 Downloads 0 Views