Objective: To examine the utility of single-photon emission computed tomography (SPECT) to predict conversion from mild cognitive impairment (MCI) to Alzheimer disease (AD).
Design: Longitudinal, prospective study.
Setting: University-based memory disorders clinic.
Participants: One hundred twenty seven patients with MCI and 59 healthy comparison subjects followed up for 1-9 years.
Measurements: Diagnostic evaluation, neuropsychological tests, social/cognitive function, olfactory identification, apolipoprotein E genotype, magnetic resonance imaging, and brain Tc hexamethyl-propylene-aminoxime SPECT scan with visual ratings, and region of interest (ROI) analyses were done.
Results: Visual ratings of SPECT temporal and parietal blood flow did not distinguish eventual MCI converters to AD (N = 31) from nonconverters (N = 96), but the global rating predicted conversion (41.9% sensitivity and 82.3% specificity, Fisher's exact test p = 0.013). Blood flow in each ROI was not predictive, but when dichotomized at the median value of the patients with MCI, low flow increased the hazard of conversion to AD for parietal (hazard ratio: 2.96, 95% confidence interval: 1.16-7.53, p = 0.023) and medial temporal regions (hazard ratio: 3.12, 95% confidence interval: 1.14-8.56, p = 0.027). In the 3-year follow-up sample, low parietal (p <0.05) and medial temporal (p <0.01) flow predicted conversion to AD, with or without controlling for age, Mini-Mental State Examination, and apolipoprotein E ε4 genotype. These measures lost significance when other strong predictors were included in logistic regression analyses: verbal memory, social/cognitive functioning, olfactory identification deficits, hippocampal, and entorhinal cortex volumes.
Conclusions: SPECT visual ratings showed limited utility in predicting MCI conversion to AD. The modest predictive utility of quantified low parietal and medial temporal flow using SPECT may decrease when other stronger predictors are available.