Parkinsons Disease Diagnosis Accuracy

The diagnosis of PD is currently based primarily on clinical judgment. However, the variability of disease presentation, progression, and response to medications often makes the diagnosis uncertain. In a population-based study, at least 15% of patients with a diagnosis of PD did not meet strict diagnostic criteria, and approximately 20% of patients with PD who had medical attention had not been diagnosed with PD (48).

Prevalence studies of parkinsonism suggest a diagnostic accuracy of 80% after examination and application of clinical diagnostic criteria (49-51). Long-term clinico-pathologic studies evaluating the diagnostic accuracy of PD demonstrate that the diagnoses most commonly mistaken for PD are progressive supranuclear palsy (PSP) and multiple system atrophy (MSA) (52,53). However, early in the course of PD, the most common misdiagnoses include essential tremor, vascular parkinsonism, drug-induced parkinsonism, and Alzheimer's disease (54,55). It has been estimated that the diagnosis is incorrect in as many as 35% of those initially diagnosed as PD by generalists (56). In addition, symptoms of parkinsonism are relatively common in elderly subjects, making the diagnosis most challenging in this population. Subtle extrapyramidal signs on neurological evaluation are common in the elderly with recorded prevalences of 32% (57) and 35% (58). Prevalence estimates for clinically evident parkinsonism in similarly aged subjects are much lower at around 3%. Neurologists with specialized training in parkinsonism are able to make the diagnosis of PD with a higher accuracy than generalists. In a six-year follow-up study of the Deprenyl and Tocopherol Antioxidative Therapy of Parkinson's Disease (DATA TOP) cohort, the diagnosis of PD was changed in only 8% of subjects. Similarly, in a study of movement disorder specialists, sensitivity was 91% for the diagnosis of PD compared to a diagnosis based on pathology (59,60).

In vivo imaging holds the promise of improving diagnostic accuracy by providing an assessment of the nigrostriatal dopaminergic system early in disease or at the threshold of diagnosis. The reduction in DAT density in PD is region-specific (putamen > caudate) and asymmetric, consistent with both pathologic assessment of DAT loss and clinical presentation of PD. Although comparisons of imaging and pathology help to confirm the validity of the imaging studies, they also highlight questions for which imaging studies may provide unique and otherwise unobtainable data. For example, in vivo imaging studies with either 18F-DOPA/PET or [123I]P-CIT/SPECT demonstrate a reduction in ligand uptake of approximately 50% to 70% in the putamen in PD subjects (20,61). The reduction in dopamine terminal function or DAT density and the pattern of striatal degeneration is consistent with the reduction in substantia nigra pars compacta neurons of greater than 80% and of putame-nal dopamine content of 95% in PD brains (62-64). However, these imaging studies also have shown at or near the threshold of diagnosis, patients with symptoms of PD show a reduction in putamenal 18F-DOPA or DAT activity of 40% to 60% rather than 80% to 90%, as suggested by pathology studies (65-67).

Difficulty in accurately diagnosing individuals early in the course of PD clearly impacts the clinical care of individuals and may also have implications when recruiting subjects for early PD clinical trials. Two recent studies involving early, untreated parkinsonian subjects suggest that imaging may identify individuals without typical PD at the time of enrollment. In the REAL-PET study, comparing ropinirole and levodopa as initial treatments in untreated patients, 11% (21/193) of enrolled subjects had scans without evidence of reduction in 18F-DOPA uptake at baseline and after two years (68). In the Early versus Late LevoDOPA (ELLDOPA)-CIT study (69), comparing initial levodopa therapy to placebo in recently diagnosed patients, 14% (21/142) of enrolled subjects had scans without evidence of reduction in [123I]P-CIT uptake at baseline and again at nine months (19/19, 2 terminated). The uncertainty of the clinical diagnosis is an important factor in the design and critical analysis of clinical therapeutic trials. Inclusion of subjects who do not have PD increases estimates of disease frequency and confounds efficacy studies of agents that may alter the rate of progression of disease. Data from these studies underscores the difficulty in accurately diagnosing parkinsonian patients in the early stages, based solely on clinical evaluation. DAT imaging offers an objective measure of the density of the presynaptic dopaminergic neurons. Several studies have shown that DAT ligand uptake is already reduced by about 50% when compared to age-corrected controls, indicating a role for DAT and SPECT in confirming a diagnosis of PD in patients with early symptoms (70,71).

In a blinded study, 35 patients with suspected early parkinsonian syndrome (PS) were referred for DAT imaging, using [123I]P-CIT and SPECT, by community neurologists who were unsure of their diagnosis (71). In this study, PS was defined as any condition expected to have a reduction in DAT density, including PD, PSP, MSA, diffuse Lewy body disease (DLBD), striatonigral degeneration (SND), or cor-tiobasal degeneration (CBD). To evaluate the accuracy of DAT imaging as a diagnostic tool, patients were followed clinically over a six-month period. Two movement disorder experts assigned a clinical diagnosis at the time of referral. One movement disorder expert remained masked to the imaging data and assigned a clinical diagnosis at the six-month interval. The six-month clinical diagnosis served as the "gold standard" diagnosis for the study. Based on this study, the sensitivity of the [123I]P-CIT and SPECT imaging diagnosis was 92% whereas the specificity of the imaging was 100% when compared to the clinical "gold standard" diagnosis at six-month follow-up. Two subjects referred with a questionable diagnosis of PS had a diagnosis of PS by the clinical "gold standard," whereas their imaging showed no deficit of DAT uptake. [123I]FP-CIT and SPECT was performed, and subjects were followed over a two- to four-year period. In this study, the clinicians were aware of the imaging results and utilized this information in making a final diagnosis. In 9 of 33 subjects, dopaminergic neuronal degeneration was found and, in all cases, a diagnosis of PS was confirmed clinically. In 24 subjects, there was no evidence of dopaminergic neuronal degeneration, and other non-PS diagnoses were assigned in 19 of these subjects at follow-up (70). Both studies suggest that the positive predictive value of DAT imaging in the diagnosis of PS is high; however, the negative predictive value is lower. Combining data from DAT with the clinical evaluation improves the diagnostic accuracy of PS in difficult to diagnose cases. These imaging data acquired from early PD patients at the threshold of illness have clarified the natural history of the disease, providing a means to improve our diagnostic accuracy earlier in the disease process and providing further impetus to develop therapies to protect the remaining 50% of dopaminergic neurons not yet affected at disease onset.

0 0

Post a comment