Objective: To establish if repeat SPECT imaging, using automated quantitative analysis, aids diagnosis in suspected Lewy Body Disorders (LBD).

Background: Patients with suspected LBD are usually diagnosed on clinical criteria. Guidelines recommend use of imaging in cases of diagnostic uncertainty. A proportion who clinically appear to have LBD have normal scans. Use of automated software to aid reporting has become common. Here we used quantitative analysis using Hermes BRASS and examined whether repeat imaging could aid diagnosis.

Method: Case notes for all patients who had 2 DaTscans between 17/6/2014 and 6/2/2019 were reviewed. Indications for scan, result, clinical diagnosis, interval between scans and subsequent diagnosis was recorded. Data were analysed to establish if a repeat scan changed diagnosis. Semi-quantitative analysis using Hermes BRASS analysis was performed on all scans.

Results: Of 1097 patients who underwent DaT, 12 Patients (8 male) had 2 scans. Mean age was 80 (SD 7.7). At baseline 7 patients had Parkinsonism, 5 only reported neuropsychiatric features suggested of DLB. At baseline 10 scans did not show dopaminergic deficit visually and on BRASS criteria (z>-3 at all sites), 1 corresponded with a basal ganglia infarct and 1 was consistent with dopaminergic deficit (visual uptake deficit and/or abnormal BRASS, z≤-3 at any site). Mean interval between scans was 22 months (SD 9.6). 9 were repeated due to new or worsening Parkinsonism, 3 for worsening hallucinations. In 10 cases the follow-up remained normal. Where the initial scan was suggestive of a dopaminergic deficit, follow-up confirmed this. In 2 cases where the initial scan was normal repeat showed dopaminergic deficit based on BRASS criteria.  Our findings warrant further exploration: 1) patients with a final diagnosis of drug induced parkinsonism (n=3) showed an increase in BRASS z-score (mean Δz=+1.1 ± 0.5 (SD) over all sites); 2) two patients with a final diagnosis of DLB had decreased z-score (mean Δz=-0.85 ± 0.4 (SD) over all sites); 3) Three patients, working diagnosis non-LBD (normal DaTscans), showed decreases in z-score suggesting possible misdiagnosis.

Conclusion: Changes in BRASS z-score following repeat scan have potential for improving the accuracy of the technique with positive Δz confirming DIP and negative Δz confirming LBD

« Back to 2019 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/does-automated-analysis-of-repeat-datscan-imaging-aid-diagnosis-in-suspected-lewy-body-disorders/