Combination Analysis of Pittsburgh Compound‐B Positron Emission Tomography and ¹⁸F-Fluorodeoxyglucose Positron Emission Tomography for Alzheimer's Disease with the t Test

Introduction: For diagnosis of Alzheimer's disease (AD), carbon-11 (¹¹C)-labeled Pittsburgh compound‐B positron emission tomography (PiB-PET) and ¹⁸F-fluorodeoxyglucose PET (FDG-PET) were evaluated by visual interpretation and quantitative measurement. We performed t tests to examine the relationship between abnormalities found on PiB-PET and those found on FDG-PET for diagnosing AD.

Methods: A total of 164 participants—58 normal controls (the NC group), 60 patients with mild cognitive impairment (the MCI group), and 46 with AD (the AD group)—underwent both PiB-PET and FDG-PET in Japan’s Alzheimer’s Disease Neuroimaging Initiative (J-ADNI). The findings were analyzed with t tests. SPM12 and PMOD were used for this analysis. The standardized uptake value of each imaging study was calculated and then normalized to cerebellar values. Comparison with the PiB-negative NC group for each voxel was performed with t tests. The mean t values in the regions of interest were compared to determine the significantly abnormal regions in the AD and MCI groups. Each examination was evaluated with receiver operating characteristic (ROC) curves and the corresponding areas under the curves (AUCs).

Results: In PiB-PET, a significant increase in PiB accumulation was observed in the gyrus rectus, angular gyrus, and posterior cingulate in both the AD group (t_means = 5.87, 5.14, and 5.14, respectively) and the MCI group (t_means = 3.77, 3.60 and 3.11, respectively). On the other hand, in FDG-PET, a significant decrease in FDG accumulation in the posterior cingulate gyrus, anterior cingulate gyrus, and hippocampus were observed in both the AD group (t_means = 0.93, 0.83, and 0.81, respectively) and the MCI group (t_means = 0.67, 0.50, and 0.50, respectively). In many regions, t values in PiB-PET was larger than those in FDG-PET. ROC analysis of the NC and AD groups’ values in the posterior cingulate gyrus yielded AUCs of 0.70 for PiB-PET and 0.83 for FDG-PET. In the same region of interest, for the comparison between the NC and MCI groups, the AUCs was larger for FDG-PET (AUC = 0.71) than for PiB-PET (AUC = 0.51). In contrast, the AUC for the comparison between MCI and AD was larger for PiB-PET (AUC = 0.70) than for FDG-PET (AUC = 0.61). Conclusion: In the posterior cingulate, anterior cingulate, and angular gyri, abnormalities were observed on PiB-PET and FDG-PET. The results of PiB-PET and FDG-PET in the hippocampus and parahippocampal gyrus did not correspond. We suggest that FDG-PET can be used complementarily PiB-PET to differentiate healthy persons, patients with mild cognitive impairment, and patients with AD.