An article Online First and in the February edition of The Lancet Neurology reports that brain scans using positron emission tomography (PET) can identify with high accuracy which form of Parkinsonism a patient has. Such early diagnosis is essential to make sure that patients receive the correct treatment and do not receive ineffective treatments due to misdiagnosis. The article is the work of Dr. David Eidelberg, Center for Neurosciences,The Feinstein Institute for Medical Research, Manhasset, New York, and colleagues.
Parkinson's disease arising spontaneously can present with symptoms comparable to those of multiple system atrophy or progressive supranuclear palsy. The investigators aimed to assess in this study whether metabolic brain imaging combined with pattern analysis could accurately discriminate patients with different forms of Parkinsonism.
A total of 167 patients were assessed in this study. They were recruited from the New York area between 1998 and 2006. They all had parkinsonian features but uncertain clinical diagnosis. At the Feinstein Institute for Medical Research, all of the patients underwent a PET scan. The team of researchers developed an automated image-based classification procedure to tell apart individual patients with idiopathic Parkinson's disease, multiple system atrophy, and progressive supranuclear palsy. The likelihood of having each of the three diseases was calculated for each patient. A classification was made according to probability measurements. After imaging, movement disorders specialists who were unaware of the PET results assessed the patients for an average of 2.6 years before a final clinical diagnosis was made. Then, the accuracy of the initial image-based classification was evaluated and compared with the final clinical diagnosis.
Results indicated that image-based classification for idiopathic Parkinson's disease had 84 percent sensitivity. Sensitivity measures the proportion of actual positives which are correctly identified as such. For instance, it is the percentage of sick people who are identified as having the condition. They had 97 percent specificity which measures the proportion of negatives that are correctly identified. 98 percent had positive predictive value (PPV) which is the proportion of patients with positive test results who are correctly diagnosed. 82 percent had negative predictive value (NPV) which is the proportion of patients with negative test results who are correctly diagnosed. Imaging classifications were also accurate for multiple system atrophy (85 percent sensitivity, 96 percent specificity, 97 percent PPV, and 83 percent NPV) and progressive supranuclear palsy (88 percent sensitivity, 94 percent specificity, 91 percent PPV, and 92 percent NPV). Dr Eidelberg comments: "The excellent specificity and PPV of the imaging classification makes this test suitable for diagnostic use rather than as a screening tool."
In addition to those findings, the authors indicate that early and correct diagnosis is crucial to make sure that patients with the proper diagnosis are enrolled in drug trials for potentially disease-modifying drugs for the various parkinsonian disorders. Also, the authors wish to expand their work to be able to differentiate other forms of parkinsonism.
They comment: "Automated image-based classification has high specificity in distinguishing between parkinsonian disorders and could help in selecting treatment for early stage patients and identifying participants for clinical trials."
They note: "Blinded, prospective imaging studies - ideally involving multiple centres, a larger validation group, repeat imaging, and more extensive post-mortem confirmation - are needed to establish the accuracy of this pattern-based categorisation procedure."
In an associated comment, Professor Angelo Antonini, IRCCS San Camillo, Venice and Parkinson Institute, Milan, Italy, remarks: "The clinical and research relevance of these findings should not be underestimated. Neuroprotective and disease-modifying drug research is intensifying and results mostly rely on accurate early diagnosis."
He writes in conclusion: "Although imaging might be cost effective for early diagnosis, I expect that these procedures will find their natural application in the identification of suitable candidates for drug trials or complex surgical procedures (eg, deep brain stimulation, stem-cell transplantation, or foetal tissue transplantation). However, additional blinded, prospective, multicentre studies will first be needed to confirm the accuracy of this pattern-based categorisation procedure."