Regional stability of cerebral blood flow measured by repeated technetium-99m-HMPAO SPECT, implications for the study of state-dependent change

The replicability of resting state rCBF has implications for the analysis of cerebral activation protocols and the interpretation of rCBF in disease states. This study examined the stability of rCBF as measured by two resting state (TC)-T-99m-HMPAO brain SPECT scans with an emphasis on examining the contribution of specific cerebral regions to within and between subjects variance, Methods: Nine normal, medically healthy subjects underwent two Tc-99m-HMPAO brain SPECT scans under identical conditions separated by 48 hr, A reference system and semiautomated computer ROI method was used to enable accurate alignment and cortical analysis of the two scans. Results: Mean within-subject difference between Scans 1 and 2 was 2.8% (range 0%-7.8%) for the 36 cortical ROIs, The mean between-subject coefficient of variation was 10% (range 7%-15%) for these ROIs, Correlation analysis of rCBF pattern replication for ail slice levels yielded a highly significant overall consistency of pattern within subjects (Pearson r = 0.698, p = 0.0001), Variance component analysis revealed regional heterogeneity in between-subjects variance, with significantly greater variability found in frontal regions, The within-subject repeated measures variability was not significantly different across regions, Conclusion: Good within-subject 48-hr replicability indicates that individual resting state rCBF reflects fairly stable, subject-specific factors, This also justifies comparing state-dependent studies separated by a modest length of time. Although individual patterns of rCBF replicate well, the larger contribution of frontal regions to normal between-subjects variance makes evaluating the frontal effects of disease or activation more difficult.