Early Differential
Diagnosis of Parkinsonism
with Metabolic Imaging
and Pattern Analysis


logo_udal_300px

Distinguishing Parkinson’s disease (PD) from other Neurodegenerative Diseases

Team Leaders: Andrew Feigin, MD, Chris Tang, MD, PhD, and Martin Lesser, PhD


Parkinson’s Disease

Multisystem Atrophy (MSA)


Progressive Supranuclear Palsy (PSP)

Corticobasal Ganglionic Degeneration (CBGD)


PET scans of brains of patients with various forms of “parkinsonism.” The colors mark regions of increased (red) or depressed (blue) glucose metabolism, compared to the same regions in people free of disease. The distinctions among the diseases are far clearer in the PET scans than they are in clinical manifestations.


Idiopathic Parkinson’s disease (PD) is the most common cause of “parkinsonism,” a term that loosely refers to the clinical triad of rigidity, resting tremor, and bradykinesia. These signs manifest the nigrostriatal degeneration underlying classic PD, but other conditions can produce highly similar clinical signs, especially in their early stages. These conditions include [1]:

  • Multisystem atrophy (MSA),
  • Progressive supranuclear palsy (PSP),
  • Corticobasal ganglionic degeneration (CBGD),
  • Diffuse Lewy body disease,
  • Essential tremor,
  • Drug-induced parkinsonism,
  • Vascular parkinsonism,
  • Normal pressure hydrocephalus, and even
  • Normal aging.

Because the diagnoses are entirely clinical, based on history and physical examination, these other forms of parkinsonism present a diagnostic challenge early in the disease course, when signs are mild and few (or when their presentation is atypical).

Pathological studies indicate that some 25 percent of patients believed to have PD actually suffer from a different disease [2]. Approximately 80 percent of such patients misdiagnosed as having PD turn out to have either MSA or PSP [3, 4].

An error rate of between 20 and 30 percent has significant consequences for clinical research trials as well as for patient care.

First, the prognosis for various parkinsonian disorders differs significantly: PD does not shorten lifespan, but the life expectancy of MSA and PSP patients is typically only between 5 and 10 years after diagnosis.

Second, the outcomes of certain therapies in PD patients are quite divergent from those in patients with atypical parkinsonism. Deep brain stimulation (DBS) procedures, for example, can be quite helpful in PD, but they are ineffective or even deleterious in MSA [5-7].

Third, clinical trials of potential disease-modifying agents can be compromised by unwittingly mixed patient populations: approximately 15 percent of participants diagnosed with early PD at the time of their recruitment for the trial have later been found to have other parkinsonian syndromes [8-10]. Yet diagnosis currently can be confirmed only by autopsy—far too late for researcher and patient alike.

Our goal in this project is to test a new, automated, image-based algorithm for its effectiveness in providing accurate, differential diagnosis of PD within two years following symptom onset (or perhaps even before symptoms appear in an at-risk population).

Using FDG PET and spatial covariance mapping, we have identified patterns of brain metabolism that distinguish PD from multiple system atrophy (MSA) and from progressive supranuclear palsy (PSP), the two most common atypical parkinsonian syndromes. Utilizing our novel automated differential-diagnosis algorithm, image-based classifications have been shown to provide excellent diagnostic accuracy for idiopathic PD, even early in the disease course (≤2 years from symptom onset). Metabolic imaging with automated pattern analysis is thus a promising approach for accurate early diagnosis of PD vs. MSA or PSP.

In this project, we are validating this approach in strictly defined patient populations, establishing its false-positive and false-negative rates, and determining how early the metabolic abnormalities are detectable. To do so, we are studying a population at risk for PD: individuals with REM behavior disorder, or RBD. Collaborations with Dr. Kathleen L. Poston, MD, of Stanford University and Dr. Howard Hurtig, MD, of the Udall Center at the University of Pennsylvania broaden our patient base and enable us to test the feasibility of our approach across different movement-disorder clinics and imaging centers.

The project has the following Specific Aims:
Return to Top

    • Aim 1. Test the sensitivity and specificity of our image-based automated classification in patients with unequivocal clinical diagnoses of PD, MSA, or PSP.
      Our proof-of-principle study provides compelling evidence that our novel automated FDG PET-based method can reliably distinguish PD from its most common look-alikes, MSA and PSP. Nonetheless, this method needs to be formally validated in a true prospective, blinded study. In addition, to date we have only applied this method to FDG PET scans done at our center; the study in Aim 1 assesses the applicability of our method to subjects clinically evaluated and scanned at other medical centers.

 

    • Aim 2. Assess the diagnostic accuracy of FDG PET in a real-world context by testing our method in early parkinsonian subjects with an uncertain diagnosis.
      Our proof-of-principle study strongly suggests that our automated method provides an accurate early diagnosis of PD, MSA, and PSP. In Aim 2, all subjects with parkinsonism are included (i.e., diagnoses not limited to PD, MSA, and PSP); and the data from subjects with other diagnoses at the end of the study (e.g., CBGD) will be used to identify other disease-related patterns and then added to our automated algorithm. Furthermore, as stated for Aim 1, this study assesses the applicability of our method to early parkinsonian subjects clinically evaluated, scanned, and followed at other medical centers.

 

  • Aim 3. Determine whether there are presymptomatic changes in brain metabolism and dopaminergic function in subjects with REM behavior disorder (RBD).
    It has been reported that as many as two-thirds of PD patients had symptoms of RBD prior to the onset of motor symptoms of PD, which suggests that RBD may be a harbinger of PD (and perhaps other neurodegenerative disorders, most notably MSA and Lewy body dementia) [11,12]. Because the number of patients with an inherited form of PD is so low, we believe that individuals with RBD offer the best chance to observe preclinical changes in metabolic brain activity.In Aim 3, we are:

      • A) Assessing regional brain metabolism (with FDG PET) and dopamine transporter (DAT) binding (with FPCIT PET) in 20 patients exhibiting idiopathic RBD. To do so, we quantify the expression of the patterns related to PD, MSA, and PSP in the FDG scans and correlate the resulting individual-subject scores with the reported length of time since RBD onset. We also apply the automated differential algorithm to the data from these subjects. We assess the integrity of the dopaminergic system by comparing caudate/putamen DAT binding in the RBD population to that in previously scanned age-matched healthy controls as well as in PD patients.

     

      • B) Performing a new network analysis on a combined group of the RBD subjects and age-matched controls to identify novel brain networks associated with RBD.

     

    • C) Tracking changes in brain metabolism and DAT binding over time in RBD patients, with repeat FDG and FPCIT PET scanning after two years and again after four years. In addition to tracking the change in previously identified PD covariance patterns, we will use a new within-subject network modeling approach to identify specific progression-related patterns in the preclinical period. All subjects are examined for signs of parkinsonism at baseline and at follow-ups utilizing the Unified Parkinson’s Disease Rating Scale (UPDRS; [13]).

Return to Top

Key Personnel

Team Leader: Andrew Feigin, MD

Team Leader: Chris Tang, MD, PhD

Team Leader: Martin Lesser, PhD

Co-Investigator: Vijay Dhawan, PhD

Co-Investigator: Martin Niethammer, MD, PhD

Site Principal Investigator: Howard Hurtig, MD, Professor of Neurology, Clinical Core Leader of the Penn Udall Center for Parkinson’s Research, University of Pennsylvania, Philadelphia, Pennsylvania

Site Principal Investigator: Kathleen L. Poston, MD, Assistant Professor of Neurology and Neuroscience, Stanford University School of Medicine, Stanford, California

Site Principal Investigator: Steven Frucht, MD, Professor of Neurology, Director of Movement Disorders in the Robert and John M. Bendheim Parkinson and Movement Disorders Center at Mount Sinai Medical Center, New York

 

Literature Cited

  1. Samii A, Nutt JG, Ransom BR. Parkinson’s disease. Lancet 2004;363:1783-93.
  2. Hughes AJ, Daniel SE, Kilford L, Lees AJ. Accuracy of clinical diagnosis of idiopathic Parkinson’s disease: a clinico-pathological study of 100 cases. J Neurol Neurosurg Psychiatry 1992;55:181-4.
  3. Hughes AJ, Ben-Shlomo Y, Daniel SE, Lees AJ. What features improve the accuracy of clinical diagnosis in Parkinson’s disease: a clinicopathologic study. 1992. Neurology 2001;57:S34-8.
  4. Hughes AJ, Daniel SE, Ben-Shlomo Y, Lees AJ. The accuracy of diagnosis of parkinsonian syndromes in a specialist movement disorder service. Brain 2002;125:861-70.
  5. Lambrecq V, Krim E, Meissner W, Guehl D, Tison F. [Deep-brain stimulation of the internal pallidum in multiple system atrophy]. Rev Neurol (Paris) 2008;164:398-402.
  6. Chou KL, Forman MS, Trojanowski JQ, Hurtig HI, Baltuch GH. Subthalamic nucleus deep brain stimulation in a patient with levodopa-responsive multiple system atrophy. Case report. J Neurosurg 2004;100:553-6.
  7. Tarsy D, Apetauerova D, Ryan P, Norregaard T. Adverse effects of subthalamic nucleus DBS in a patient with multiple system atrophy. Neurology 2003;61:247-9.
  8. Parkinson Study Group. Dopamine transporter brain imaging to assess the effects of pramipexole vs levodopa on Parkinson disease progression. Jama 2002;287:1653-61.
  9. Fahn S, Oakes D, Shoulson I, Kieburtz K, Rudolph A, Lang A, Olanow CW, Tanner C, Marek K. Levodopa and the progression of Parkinson’s disease. N Engl J Med 2004;351:2498-508.
  10. Jankovic J, Rajput AH, McDermott MP, Perl DP. The evolution of diagnosis in early Parkinson disease. Parkinson Study Group. Arch Neurol 2000;57:369-72.
  11. Schenck CH, Callies AL, Mahowald MW. Increased percentage of slow-wave sleep in REM sleep behavior disorder (RBD): a reanalysis of previously published data from a controlled study of RBD reported in SLEEP. Sleep 2003;26:1066; author reply 7.
  12. Iranzo A, Molinuevo JL, Santamaria J, Serradell M, Marti MJ, Valldeoriola F, Tolosa E. Rapid-eye-movement sleep behaviour disorder as an early marker for a neurodegenerative disorder: a descriptive study. Lancet Neurol 2006;5:572-7.
  13. Fahn S, Elton R. Unified Parkinson’s Disease Rating Scale. In: Fahn S, Marsden CD, Calne DB, eds. Recent Developments in Parkinson’s Disease. Florham Park: MacMillan Health Care Information; 1987. p. 153-63.

Return to Top