The Impact of DaTscan in the Diagnosis of Parkinson Disease

Abstract

Purpose: The aim of this study was to evaluate the impact of DaTscan in a heterogeneous group of patients with movement disorders as well as the degree of confidence in scan findings between different readers.

Procedures: A retrospective evaluation of consecutive patients who underwent DaTscan during 1 year was performed. The patients’ demographics, symptoms, duration, clinical diagnosis, and medications were collected. The scan findings were categorized by 2 blinded observers on a semiquantitative scale as follows: 0, normal; 1, mild; 2, moderate; 3, marked; and 4, absent uptake for each of the caudate heads and putamina separately. A correlation of the scan findings with the clinical symptoms and diagnosis as well as interobserver agreement was performed. Disagreement was considered when a difference greater than 2 in more than 1 area of the basal ganglia was recorded. Descriptive statistics and [kappa] test for interobserver agreement were used for data analysis.

Results: Fifty-seven patients were included (mean age, 63.4 years; 29 men, 28 women). Clinical diagnosis of Parkinson disease (PD) was certain in 26 and uncertain in 31 patients. DaTscan was markedly abnormal in 24 (92%) of 26 patients with certain clinical diagnosis of PD and normal in the remaining 2 (8%). In 31 patients with uncertain diagnosis, 15 (48%) had markedly abnormal scans, 5 (16%) had mild abnormalities, and 11 (36%) had normal scans. Each of the sensitivity and positive predictive value of DaTscan in patients who had certain clinical diagnosis of PD (26 patients) is 92%. Interobserver agreement occurred in 52 (91%) of 57 scans and disagreement in 5 (9%) of 57 ([kappa] = 0.82). There was also a good correlation with laterality of symptoms in 32 (82%) of 39 positive studies.

Conclusions: Markedly abnormal DaTscan is confirmed as the diagnostic pattern for PD. This pattern helps confirm the diagnosis in patients with unclear clinical diagnosis. Good interobserver agreement is easily obtained in reading DaTscans.