Justification of Functional Brain Imaging in the Assessment of Dementia
Determining the etiology of dementia has proven to be extremely difficult using clinical assessment and standard cross-sectional imaging studies of the brain. Functional brain imaging in the form of SPECT and FDG-PET have improved our diagnostic abilities, not only to determine whether abnormality exists but to subcategorize the abnormalities into different subsets of dementia.
AD = Alzheimer’s disease; MCI = mild cognitive impairment; FDG-PET = fluorodeoxyglucose positron emitted tomography; HMPAO SPET = hexamethylpropyleneamineoxime single-photon emission tomography; MMSE = mini mental state examination.
Meta-analysis of the literature regarding functional brain imaging in the dementias concluded that functional brain imaging provided critical information in the initial diagnosis of dementia as well as aiding in the differential diagnosis of the specific dementia disorders.5,6 Devous compared functional brain imaging with clinical assessment and magnetic resonance imaging (MRI). This information is summarized in Table 1. In a review by Silverman, the sensitivity for detecting AD using FDG-PET was 91.5% ± 3.5%.7
Table 2 summarizes these results and compares the results of FDG-PET with clinical evaluation based upon the report of the Quality Standard Subcommittee of the American Academy of Neurology (AAN). Additional reports appeared to document the superiority of functional brain imaging over clinical assessment and cross-sectional imaging.
Functional brain imaging studies, particularly FDGPET, appeared to be extremely important in the early phases of the dementias, particularly AD in which mild cognitive impairment (MCI) may be an early manifestation. The separation of an early Alzheimer’s process from other causes of dementia, especially vascular and psychiatric related, are extremely important in that clinical management decisions and prognosis depend upon the correct diagnosis.8,9
PET and SPECT Comparisons
Due to the wide availability of SPECT detectors, SPECT imaging of the brain was most often used for evaluating patients with neurodementia.10,11 As more information became available with regards to the utility of FDG-PET in evaluating patients with suspected neurodementia, it became apparent that the ability to separate normal controls from patients with true dementia was superior for PET when directly compared with SPECT.12,13 The overall literature now appears to indicate FDG-PET as the most accurate assessment in the evaluation of patients with suspected AD. Silverman et al. determined a sensitivity of 93% with specificity of 58% based upon the standard of clinical diagnosis in reviewing the PET literature for diagnosis of AD.6
Note the relatively normal activity throughout the frontal lobe in the Alzheimer subject (B). In the latter phases of Alzheimer’s disease, the frontal lobe may be significantly involved.
Technical Aspects of SPECT and PET
There appears to be wide-ranging variability in the ability to produce adequate images for purposes of interpretation using SPECT and PET. SPECT imaging appears to have the greatest variability due to the demanding nature of performing a SPECT study. Images are dependent on detector resolution including the choice of collimator. The numbers of counts per study also play a role in the ultimate production of a quality image.Three-detector SPECT systems are superior to single detector SPECT systems in producing higher count rate studies.
New generation PET or PET-computed tomography (CT) systems have high count-rate capabilities and excellent resolution. The variability between major manufacturers of PET or PET-CT systems is minimal with the new generation devices and produce uniformly excellent images of the brain using FDG. This allows nearly all PET centers with new generation systems to develop a high-quality neuroimaging program.
Image Patterns of Dementia Using Functional Brain Technology
A variety of displays are available to evaluate functional brain images. The standard projections of three orthogonal plains including axial, coronal, and sagittal are commonly utilized in evaluating a functional brain imaging study.These are often coupled with maximum intensity projection (MIP) images and surface-rendered images.
Figure 1 is a surface-rendered display of a normal brain and a comparison with a patient with mild to moderate AD.
The typical areas of involvement in AD are the parietal, posterior-parietal, temporal, and posterior-temporal regions. Most patients with AD also demonstrate early reductions in the posterior cingulate and pre-cuneal regions. The AD subject in this Figure 1 demonstrated mild–moderate cognitive impairment but was functioning at a relatively high level with no major supervisory care required. This pattern is often seen in the early phases of AD. The frontal lobes are within normal limits. Abnormalities of the frontal lobe are a later manifestation as the AD progresses.
AD is characterized by deposition of beta amyloid in contrast to Lewey Body dementia, which is associated with Lewey Bodies comprised of synuclein. Patients with AD may have a Lewey Body component. This is estimated to occur in approximately 30% of patients with AD.
Clinically, Lewey Body dementia often presents as a fluctuating cognitive state and in the later phases subjects exhibit visual hallucinations. Figure 2A is a subject with in an early manifestation of Lewey Body dementia. Figure 2B demonstrates a subject with a later phase of the disease. While the pattern of AD and Lewey Body disease have similarities, a major difference is the involvement of the occipital cortex in subjects with Lewey Body dementia. This is demonstrated in Figure 2, particularly Figure 2B, where the abnormalities appear to be severe.
Frontal-temporal dementia has a different pattern than that of AD or Lewey Body disease and predominantly affects the frontal and temporal lobes. Subjects with frontal-temporal dementia demonstrate tau protein deposition in regions of reduced perfusion and metabolism.
Figure 3 is a subject who initially presented with mild disinhibition and mild–moderate changes in the frontal lobe, more severe on the right than on the left.Within two years, the disease had progressed to a moderate–severe status involving both frontal lobes as noted in Figure 3B.
Note the mild involvement of the frontal lobe greater on the right than on the left. At this stage, the patient exhibited mild disinhibition with occasional inappropriate social episodes. In approximately 24 months, the disease had progressed to a moderate to severe level (see Figure 3B) and was accompanied by a greater degree of disinhibition.
Comparisons of SPECT and PET functional brain imaging has been discussed with increasing sensitivity noted for FDG-PET in the detection of AD. Figure 4 demonstrates a comparison of SPECT using Tc-99m ECD (Figure 4A) with FDG-PET (Figure 4B). This patient was suspected of having AD and was referred for a SPECT study, which showed only minimal changes in the parietal cortex. The FDG-PET carried out two months later demonstrated marked diminished parietal, posterior-parietal, and posterior-temporal metabolism. The findings are considerably more marked with PET than with SPECT.
Figure 5 shows the axial views of the same subject demonstrating a marked decrease in the posteriorparietal regions seen with FDG-PET when compared with the mild changes seen with ECD SPECT.
PET-CT Fusion Imaging
With the increasing number of PET-CT systems currently installed, as well as successful fusion programs using a software approach for integrating images, it is now possible to perform functional brain imaging with direct correlation with cross-sectional imaging. Figure 6 is a PET-CT study performed using a new generation PET-CT system. The upper row is the FDG-PET, the middle row is the CT and the lower row represents the PET scan fused to the CT scan. Findings demonstrate the typical lack of CT findings in the parietal and posterior parietal regions, while the FDG-PET images demonstrate findings that are consistent with the clinically suspected diagnosis of early AD with no evidence for a vascular process.
New Directions in Functional Imaging of Dementia
A recently approved clinical trial will evaluate the optimum combination for imaging patients suspected of having AD. This project is being sponsored by the National Institutes on Aging (NIA) and is called the Alzheimer Disease Neuroimaging Initiative (ADNI).
The project will recruit approximately 800 participants to determine the best combination of examinations including MRI and serial FDG-PET studies. The project will focus on distinguishing mild memory decline due to normal aging from impairment caused by early AD. In addition, the project will try to determine whether the dementia of AD can be distinguished from other dementias. In addition, the initiative will attempt to determine the best approach for following AD, particularly with regards to the efficacy of drugs used to change the course of the disease process.
It has been demonstrated that beta amyloid is present in the cortex of patients with AD.There is some controversy over the significance of the beta amyloid as to whether or not the primary cause of AD is secondary to a series of reactions that are caused by amyloid. Some scientists speculate that amyloid is a by-product of other processes, possibly inflammation, and by itself is not toxic.
A diagnostic strategy attempting to image amyloid deposition has been accomplished using highly specific amyloid targeting agents. A molecular approach has been successful using derivatives of congo red and thioflavin. The University of Pittsburgh researchers Drs Klunk and Mathis have developed a molecule labeled with carbon- 11 having the formulation 2-(4-methylaminophenyl)-6- hydroxybenzothiazole, otherwise known as Pittsburgh Compound-B or PIB. This compound has been successfully demonstrated to bind to amyloid in transgenic mice and recently in humans.14 Patients with AD demonstrate high concentrations of PIB when compared with normal controls. Scientists at the University of California, Los Angeles, have successfully imaged beta amyloid using 18- fluourine-dimethyl-amino-dicyano-naphthalene propene (FDDNP). This compound has been demonstrated to bind to the neurofibrillary tangles as well as the beta amyloid plaques.15
Note the marked decreases in the posterior-parietal regions on the fluorodeoxyglucose positron emitted tomography (FDG-PET) when compared with the ethyl cysteinate dimer (ECD SPECT).
The use of PET to detect the distribution of these agents will play an important role in diagnosis as well as the monitoring of pharmaceuticals that are being developed to lower the concentration of amyloid within the brain. Other agents are being developed along these lines and show promise.
While the current molecular imaging approach mainly involves the use of radiotracers, it is entirely possible that agents linked to paramagnetic substances will be used with MRI. In addition, high Tesla magnets may also yield information not currently available using standard approaches of MRI.
At this time, FDG-PET is well documented as a clinically important clinical test in subjects with suspected dementia and has been proven to have a high degree of sensitivity and specificity, especially when compared with structural imaging techniques and clinical evaluation by competent neurologists.
In the upper row, the FDG-PET demonstrates biparietal reductions in metabolism more marked on the left than the right.The CT scan in the middle row demonstrates normal cortical structure in the parietal and posterior-parietal regions. Fusion imaging may be valuable in attempting to differentiate a vascular etiology (prior infarction) from a neurodementia.
FDG-PET has also been demonstrated to have efficacy in differentiating a variety of different dementias including frontal-temporal dementia from AD as well as multi-infarct dementia and others. FDG-PET has been demonstrated to be cost-effective and can be carried out with high quality in most facilities with new generation PET or PET-CT systems.
The use of functional brain imaging to diagnose and differentiate dementias is becoming increasingly important during an era where therapeutic agents are being prescribed.The separation of the dementias such as frontal temporal dementia, AD, and multi-infarct dementia have significant meaning in determining the appropriate therapies from which the patient is most likely to receive maximum benefit.
Functional brain imaging has been shown to have an extremely high sensitivity and moderate specificity in diagnosing AD, which is the most common of the neurodementias. Quantitative programs are emerging that will aid the interpreting physician in correctly evaluating the FDG-PET scans.
The ease of performing high-quality PET studies along with excellent reproducibility and consistency in diagnosing AD has resulted in an increasing rate of utilization of functional brain imaging for evaluating patients with suspected neurodementia.