Correlations Between the Distribution of Cortical Hypometabolism and Clinical Manifestations in Patients with West Syndrome

Abstract

Purpose: To examine the relation between the distribution of cortical hypometabolism and clinical manifestations in patients with West syndrome (WS). Methods: Serial positron emission tomography (PET) with [18]fluorodeoxyglucose was performed in 32 patients with WS, first at the onset of spasms and later at age 10 months or 3 months after the initial treatments. We excluded patients with diffuse or multiple abnormalities on magnetic resonance imaging (MRI). The age at onset of spasms ranged from 2 to 26 months, and ages at te end of follow‐up ranged from 2 to 6 years. Fifteen patients were diagnosed with cryptogenic WS and 17 with symptomatic WS. We classified the distribution of cortical hypometabolism into following groups: (A) occippitotemporal areas, (B) temporoparietal areas, (C) frontotemporal areas, (D) diffuse, and (E) none, or (a) unilateral and (b) bilateral. We compared the distribution of cortical hypometabolism with the age at onset of spasms, outcome of seizures, and psychomotor development. The presence of psychomotor retardation was determined by clinical observations and a low Tsumori‐Inage developmental quotient score (DQ < 80). Chugani et al. reported that patients with bitemporal hypometabolism had poor developmental and seizure outcome, and that the majority were autistic. Therefore we particularly focused on patients with focal cortical hypometabolism in the temporal lobes bilaterally. Results: Initial PET scans demonstrated diffuse or focal hypometabolism in 24 patients. The second PET showed hypometabolism in only 10 of these patients. Hypometabolism was present on the second PET in two additional patients with normal PET findings at the onset of the disease. The number of patients with each distribution of hypometabohm at the first PET scan were as follows: six patients in group A, eight patients in B, five patients in C, five patients in D, and eight patients in E, or nine patients in group a and 15 patients in b. The mean age at onset of spasms in each group was 4.0 months in group A, 8.3 in B, 13.4 in C, 10.4 in D, and 4.8 in E, or 4.7 in group a, and 11.2 in b. The age at onset of spasms in patients with occipitotemporal hypometabolic areas was significantly lower than that in patients with frontotemporal hypometabolic areas. The age at onset of spasms in patients with unilateral hypometabolic areas was also significantly lower than that in those with bilateral hypometabolic areas. There were no significant correlations between the distributions of hypometabolism on the first scan and seizure outcome or psychomotor development. The existence of hypometabolic areas on the second PET scan was positively correlated with the poor developmental and seizure outcome. Nine patients had focal hypometabolism in bitemporal areas on the first scan. In five of these nine patients, hypometabolism disappeared on the second scan. The outcome of seizures and psychomotor development was poor in patients who had bitemporal hypometabolism after the initial treatment, as Chugani et al. reported. However, four of the five patients in whom bitemporal hypometabolism disappeared on the second scan had no epileptic seizures after the initial treatment and psychomotor development were almost normal. Conclusions: There was a correlation between the distribution of hypometabolic areas and the age at onset of spasms. It is possible that functional abnormality in the cerebral cortex undergoing maturational changes has an important role in the generation of infantile spasms. However, the cortical hypometabolism revealed by PET changes with clinical symptoms. To predict the outcome in WS, it is important to follow up the PET findings after the initial treatments.