ABSTRACT & COMMENTARY
DaTSCAN to Distinguish Parkinson’s Disease from Secondary Parkinsonism
By Claire Henchcliffe, MD
Associate Professor of Neurology and Neuroscience, Weill Cornell Medical College
Dr. Henchcliffe reports she is on the speakers bureau and advisory board for GE, Teva Pharmaceutical Industries, and UCB; advisory board for Allergan and USWorldmeds; receives grant/research support from Biogen and Kaneka; and does CME program development and presentation for MedIQ.
A meta-analysis of five studies determined 86% sensitivity and 83% specificity in distinguishing Parkinson’s disease from vascular parkinsonism, and 86% sensitivity and 98% specificity in distinguishing Parkinson’s disease from drug-induced parkinsonism using DaTSCAN imaging.
Brigo F, et al. [123I]FP-CIT SPECT (DaTSCAN) may be a useful tool to differentiate between Parkinson’s disease and vascular or drug-induced parkinsonisms: A meta-analysis. Eur J Neurol 2014;21:1369-1376.
DaTSCAN imaging makes use of the dopamine transporter (DAT) ligand [123I]FP-CIT used with SPECT to provide a means to visualize the integrity of the nigrostriatal dopaminergic system. DaTSCAN has been approved in the United States to aid in evaluation of patients in whom Parkinson’s disease (PD) and certain other related neurodegenerative disorders are suspected. Brigo and colleagues now report results of a meta-analysis aiming to evaluate the usefulness of DaTSCAN in distinguishing PD, in which an abnormal scan is expected, from vascular parkinsonism (VP) and drug-induced parkinsonism (DIP), in which normal scans are expected. The meta-analysis included prospective and retrospective studies that examined DaTSCAN in cases of unclear clinical parkinsonism, in which final diagnoses included PD, VP, and/or DIP. Case-control studies were excluded. Five studies published between 2001 and 2013 met criteria for inclusion in the statistical analysis out of an initial 31 studies provisionally identified. Comparing DaTSCAN with final diagnosis, polled measures for PD vs VP yielded a sensitivity of 86.2% (95% confidence interval [CI], 81.3-90.1) and specificity of 82.9% (95% CI, 67.9-92.8%) along with a positive likelihood ratio (PLR) of 4.813 (95% CI, 1.523-15.211) and negative likelihood ratio (NLR) of 0.190 (95% CI, 0.139-0.259). Ability of DaTSCAN to diagnose PD vs DIP when compared with final diagnosis yielded a sensitivity of 86.2% (95% CI, 81.3-90.1) and specificity of 93.8% (95% CI, 69.8-99.8%), with a PLR of 5.366 (95% CI, 1.913-15.050) and NLR of 0.178 (95% CI, 0.125-0.253). There are certain limitations of the studies included. Two studies derived from a single group within a short time period, leading to the possibility that duplicate data were included. Moreover, diagnostic criteria were not reported for VP and DIP. Three studies did not report whether clinical diagnosis was made blinded to DaTSCAN results, and two studies did not report whether DaTSCAN results were interpreted while blinded to clinical diagnosis.
COMMENTARY
DaTSCAN was approved by the FDA in 2011 for use as an adjunct to clinical diagnosis in cases of parkinsonism with a suspicion for dopamine deficiency, such as PD, multiple system atrophy, or progressive supranuclear palsy. A recent study from Adler and colleagues highlighted diagnostic difficulty, and, using data from the Arizona Study of Aging and Neurodegenerative Disorders, they found that clinical diagnosis was only 88% sensitive and 68% specific for PD when compared to neuropathological diagnosis.1 Moreover, since PD and many of its mimics have an insidious onset with clinical overlap, the risk of misdiagnosis is particularly high in early disease. This low diagnostic accuracy not only directly impacts patients and their families, but also "dilutes" PD cohorts enrolled in clinical studies. Thus, it is imperative that objective and accurate diagnostic biomarkers are developed. DaTSCAN and related neuroimaging modalities hold much hope in this regard.
As clinicians gain experience in how best to employ this test in practice, this meta-analysis by Brigo and colleagues is therefore a worthwhile addition to the literature. Most importantly, it finds that DaTSCAN is likely to be helpful in distinguishing cases of PD vs VP and DIP in which there may be clinical uncertainty. In both cases, this is predicted to alter clinical decision-making. For example, in the case of DIP, it opens a different diagnostic avenue to patients who otherwise would need to taper off antipsychotic medications for further clinical evaluation. However, as the authors note, their study highlights a number of limitations in the published literature so far. Many of the studies are small, and of the five studies included in this report, two had 15 or fewer participants. There are methodological variations in scanning that are difficult to fully account for. Clinical data reported varied between studies, and three of the studies included did not report participant age, gender, or disease duration. Finally, the use of DAT imaging is limited at this time in differentiating PD from Parkinson’s plus disorders, such as multiple system atrophy or progressive supranuclear palsy, that account for many of the misdiagnosed cases seen in clinicopathologic studies. Therefore, this report should stimulate further investigations to support how best to use DaTSCAN in clinical practice.
Reference
- Adler CH, et al. Low clinical diagnostic accuracy of early vs advanced Parkinson disease: Clinicopathologic study. Neurology 2014;83:406-412.
