Optimal DaTQUANT Thresholds for Diagnostic Accuracy of Dementia with Lewy Bodies (DLB) and Parkinson's Disease (PD).

Background: Quantitative thresholds are helpful to define an abnormal DaT SPECT in patients with suspected nigrostriatal degenerative diseases (NSDD). The optimal DaTQUANT threshold for diagnostic accuracy of DaT SPECT across combined movement and cognitive disorder populations has been previously described. Methods: We established optimal DaTQUANT thresholds that enhance the discrimination between dementia with Lewy bodies (DLB) and non-DLB dementia types, as well as between Parkinsonian syndromes (PS) and conditions not characterized by nigrostriatal degeneration (non-PS). Results: Data from a total of 303 patients were used in this retrospective analysis. Posterior putamen of the more affected hemisphere (MAH) was shown to be an accurate single-variable predictor for both DLB and PS and was comparable to the most accurate multi-variable models. Conclusions: Automated quantification with DaTQUANT can accurately aid in the differentiation of DLB from non-DLB dementias and PS from non-PS. Optimal thresholds for assisting a diagnosis of DLB are striatal binding ratio (SBR) ≤ 0.65, z-score ≤ -2.36, and a percent deviation ≤ -0.54 for the posterior putamen of the MAH. Optimal posterior putamen thresholds for assisting a diagnosis of PS are SBR ≤ 0.92, z-score ≤ -1.53, and a percent deviation ≤ -0.33, which are similar to our previously reported posterior putamen threshold values using a blended patient pool from multiple study populations.