A Novel Imaging Approach for Early Diagnosis of Parkinson’s and Diffuse Lewy Body Diseases

Georges El Fakhri, Ph.D.

Harvard Medical School, Boston, MA

Grant Program:

David Mahoney Neuroimaging Program

Funded in:

December 2004, for 4 years

Funding Amount:


Lay Summary

Using Imaging for Early Diagnosis of Parkinson’s and Diffuse Lewy Body Disease

Harvard investigators have refined an imaging technique that can measure two events simultaneously in patients with Parkinson’s disease, Diffuse Lewy Body disease, and Alzheimer’s disease, to see if it can effectively diagnose each disease early.  If so, the technique could be used to diagnose, follow the natural history, and evaluate therapies for each disease.

The investigators have refined SPECT ((Single Photon Emission Computed Tomography) so that it can image two events in the brain simultaneously, rather than serially.  In Parkinson’s disease patients, the investigators will see whether they can simultaneously show a decrease in transporter proteins that usher dopamine to the synapse to be transmitted to the adjoining brain cell and an increase in the receptors in the adjoining cell that takes up the dopamine.  This excess post-synaptic activity only occurs early in Parkinson’s disease.  These events will be imaged in Parkinson’s disease patients and compared to events in healthy adults.  The investigators will verify their findings in Parkinson’s patients by comparing them to those seen the animal model of the disease.

Simultaneous dual-isotope SPECT also is anticipated to differentiate Alzheimer’s disease from Lewy Body dementia, by showing that people with Alzheimer’s disease have normal dopamine transporter protein levels but reduced blood perfusion, while people with Lewy Body dementia have decreases in both dopamine transporter proteins and blood perfusion. The investigators hypothesize that this technique will be more accurate than serial dual-isotope SPECT and PET, which cannot be used to image two simultaneous events.  If the investigators are correct, simultaneous dual-isotope SPECT could be used to diagnose, follow the natural history of, and evaluate treatments for each of these brain diseases.

Significance:  If dual-isotope SPECT imaging is more accurate than the other two imaging techniques in differentiating among Parkinson’s, Lewy Body disease, and Alzheimer’s, this could become the standard tool for making an early diagnosis among three diseases.


A Novel Imaging Approach for Early Diagnosis of Parkinson's and Diffuse Lewy Body Diseases

Simultaneous 99mTc/123I SPECT imaging makes possible the assessment of perfusion and neurotransmission, or pre- and post-synaptic neurotransmission, in perfect spatial registration and under identical physiological conditions (simultaneous dual isotope imaging is not possible with PET). Despite its obvious advantages, simultaneous imaging is not being used in the clinic. Because the emission energies of 99mTc (140 keV) and 123I (159 keV) photons are very close, discrimination between them on the basis of energy is extremely difficult; accurate simultaneous imaging of these isotopes had been considered virtually impossible. We have recently shown, using realistic computer-simulated data, that accurate separation of these two radionuclides is possible for simultaneous perfusion (99mTc-HMPAO) and neurotransmission (123I-altropane) studies by a new approach using artificial neural networks (ANN). Our next step is to extend this method to simultaneous imaging of dopamine transporters and receptors.

This is more challenging than imaging perfusion and dopamine transporter function, as both tracers concentrate in the same structures. We will generate artificial, but realistic, dual-isotope SPECT studies using computer simulated data from assumed 99mTc and 123I activity distributions based on clinical studies. The known activity distributions will serve as a gold standard for assessment of the accuracy of the dual-isotope technique. These simulation studies will be the basis for the design of an ANN-based procedure for correction of cross-contamination, scatter, and high-energy contamination. We will also extend our quantitative methodology to correct for attenuation and spatially variable, limited collimator resolution, and develop new processing methods for quantitative simultaneous dual-isotope SPECT that will exploit energy spectral and dynamic (time-varying) information.

Our approach will be evaluated by one primate study and two patient studies. We will compare simultaneous dual-isotope SPECT, with images obtained using our quantitative techniques, to single-isotope SPECT, as well as to 11C-altropane PET, on the basis of performance in classification tasks. The primate study will use 25 rhesus monkeys in whom PD will be induced by treatment with MPTP, an established model of parkinsonism; we will measure performance in discriminating between normal (pre-treatment) and early-stage PD images. One patient study will involve 25 early-stage PD patients with unilateral symptoms and 25 age- and gender-matched normal controls. We will measure performance in two tasks: discrimination between early PD and normal and discrimination between very early PD and normal. The other patient study will enroll 25 DLBD patients and 25 early AD patients, matched for age and gender; we will measure performance in discriminating between the two groups.



The distinction between patients with Parkinson's disease who have concomitant Alzheimer disease (generally known as the Lewy Body Variant of AD [LBV-AD]) and those with diffuse Lewy Body Disease (DLBD) is very difficult to make at the time when cognitive impairment is first evident. The origin of the cognitive impairment in these patients is also unclear. Assessing brain perfusion and neurotransmission simultaneously using simultaneous dual isotope SPECT would yield valuable insight into the two diseases and, potentially, make it possible to discriminate between the two dementia categories at an early stage. Such studies would also allow us to relate the quantitative estimates of brain perfusion and pre-synaptic dopamine function to quantitative measures of motor and cognitive function.

The overall goal of the study will be to determine how well measures of dopamine function and blood perfusion derived from quantitative simultaneous dual-tracer SPECT correlate with clinical manifestations of PD and measures of cognitive ability, both global and particular.

The ultimate goal of this work is to optimize quantitative dual isotope SPECT imaging for the purpose of detecting the earliest disease-related changes in the brain. This will lead to an improved understanding of the natural history of neurological diseases such as PD and DLBD. Furthermore, simultaneous dual-isotope SPECT imaging has the potential to become a convenient, safe, and accurate technique for monitoring disease progression or response to therapy.

Because the emission energies of photons from the two radio-isotopes that are attached to the tracers, 99mTc and 123I, are very close, discrimination between them on the basis of energy is extremely difficult. We will extend a method we have recently developed that makes possible separation of these two isotopes to the simultaneous imaging of blood perfusion and post-synaptic dopamine function in PD patients. Our methodology will also address correction of the effects of other physical processes, such as attenuation and limited and variable spatial resolution, that compromise the accuracy of the images.

We will apply this technique to simultaneous imaging of pre-synaptic dopamine function (123I-altropane) and perfusion (99mTc-HMPAO) in 40 PD patients with cognitive impairment and 20 PD patients without cognitive impairment. Subjects will have a UPDRS and a brief neuropsychological battery, as well as simultaneous 123I-altropane / 99mTc-HMPAO SPECT scans. The overall goal of the analyses will be to determine, using correlation analysis, how well measures of dopamine function and blood perfusion derived from quantitative simultaneous dual-tracer SPECT correlate with clinical manifestations of PD, as measured by the UPDRS, and measures of cognitive ability, both global (the Blessed score) and particular (the tests in the neuropsychological battery, including tests of memory, conceptualization, spatial function, and attention).

We have demonstrated the feasibility and utility of dual isotope SPECT for differential diagnosis of idiopatic Parkinson’s disease (IPD) and multiple system atrophy (MSA). Simultaneous 99mTc-ECD/123I-FP-CIT studies were performed in 9 normal controls, 5 IPD, and 5 MSA patients. Projections were corrected for scatter, cross-talk, and high-energy penetration using artificial neural networks and five dimensional factor analysis, and iteratively reconstructed while correcting for patient-specific attenuation and variable collimator response. Perfusion and dopamine transporter (DAT) function were assessed using voxel-based statistical parametric mapping (SPM2) and volume of interest (VOI) quantitation. DAT binding potential (BP) and asymmetry index (AI) were estimated in the putamen (PUT) and caudate nucleus (CN). Striatal BP was lower in IPD (55%) and MSA (23%) compared to normal controls, and in IPD compared to MSA (p<0.05). AI was greater for IPD than for MSA and controls both in CN and PUT (p<0.05). There was significantly decreased perfusion in the left+right nucleus-lentiformis in MSA compared to IPD and controls (p<0.05). Our work shows that dual-isotope studies are both feasible in and promising for the diagnosis of parkinsonian syndromes.

This work has led to NIH funding for the principal investigator under a jointly funded RO1 by the NIA and NIBIB (EB005876: “Quantitative Simultaneous Dual Isotope SPECT”), as well as grant funding from the American Heart Association under Grant in Aid (GIA655909T: “Absolute Quantitation of Myocardial Blood Flow Using Rb-82 PET”).

Selected Publications

El Fakhri G., Sitek A., Zimmerman R.E., and Ouyang J.  Generalized five-dimensional dynamic and spectral factor analysis. Med Phys. 2006 Apr;33(4):1016-24.

El Fakhri G., Ouyang J., Zimmerman R.E., Fischman A.J., and Kijewski M.F.  Performance of a novel collimator for high-sensitivity brain SPECT. Med Phys. 2006 Jan;33(1):209-15.

El Fakhri G., Habert M.O., Maksud P., Kas A., Malek Z., Kijewski M.F., and Lacomblez L.   Quantitative simultaneous (99m)Tc-ECD/123I-FP-CIT SPECT in Parkinson’s disease and multiple system atrophy.  Eur J Nucl Med Mol Imaging. 2006 Jan;33(1):87-92.