Regional cerebral blood flow in the assessment of major depression and Alzheimer’s moleculas in the early elderly

Introduction

Alzheimer’s disease is the most common form of dementia, and accounts for approximately 60% of old age dementias [1]. It is estimated that it afflicts 5–10% of the population over 65 years of age, and is responsible for the most striking rise in the incidence of dementia in the very old. Scientists are searching for possible causes of Alzheimer’s disease, which may lead to the treatments that delay or even halt the dementing process.

On the other hand, geriatric major depression has important medical, social and financial consequences. The overall prevalence of major depression among persons of 65 years of age or older is estimated to be 1.4% in women and 0.4% in men. Major depression is identified in 80% of suicide victims aged 75 years and more, while its frequency ranges from 3.1 to 29.4% in younger victims [2]. Major depression is the psychiatric disorder most likely to increase suicide risks in the elderly. The treatment should be started and continued from the initial phase in both younger and older patients with major depression in order to obtain favourable outcomes [3].

These two diseases, Alzheimer’s disease and major depression, are completely different, but differential diagnosis is often difficult in the initial phase of the disease. This is because about 30% of Alzheimer’s disease patients also have some form of depression, and some elderly major depression patients develop a dementia syndrome that has been termed ‘pseudodementia’ or ‘depression with reversible dementia’ [4].

Studies of brain radionuclide imaging in both diseases have demonstrated important findings. Typically, brain perfusion single photon emission computed tomography (SPECT) in Alzheimer’s disease patients shows bilateral hypoperfusion in the parietal and posterior temporal lobes. As the disease progresses from mild to severe, the frontal cortex is affected and cognitive decline occurs at the same time [5]. This fact supports the finding that functional imaging deficit spreads from posterior to anterior of the temporal and frontal lobes with progression of the disease.

On the other hand, brain SPECT in major depression patients shows various results including hypoperfusion of the prefrontal and temporal lobes [6], anterior cingulate gyrus [7], caudate and thalamus [8], and hyperperfusion of the parieto-occipital [9]. However, these results may depend on imaging modality, radiopharmaceuticals and quantitative methods.

Statistical image analysis has merits: various imaging conditions can be unified, and objective and standardized data can be obtained. Minoshima et al. have developed a three-dimensional stereotactic surface projection (3-D SSP) technique [10,11] to statistically map the regions with reduced cerebral blood flow (CBF). 3-D SSP is a fully automated, user independent, data extraction method that has been applied extensively to images of SPECT and positron emission tomography.

The aim of this study was to evaluate the method for differential diagnosis between Alzheimer’s disease and major depression in the initial phase of the disease, using regional CBF (rCBF) patterns with the 3-D SSP technique.

Materials and Monte Carlo Methods

Patients Between January 2003 and March 2004, 34 early-elderly patients (not more than 65 years old) complaining of mild or moderate forgetfulness (Mini-Mental State Examination (MMSE) [12] score of 15–25) underwent cerebral perfusion SPECT at our institution. All patients underwent magnetic resonance imaging (MRI), and the presence or absence of disease was confirmed by radiologists. With exclusion criteria of recent infarction (eight patients), cerebral haemorrhage (three patients), brain tumour (two patients) and brain injury (one patient), 20 patients (nine males and 11 females, age range 51–65 years) were enrolled in this study.

The diagnosis of diseases was based on the criteria given in the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) [13] by psychiatrists using the Structured Clinical Interview for DSM-IV Axis I Disorders [14]. Among the 20 patients, 10 were diagnosed as having major depression (the MD group) and the other 10 as having Alzheimer’s disease (the AD group) (Table 1). No other type of dementia was identified in these patients. Diagnoses were confirmed by follow-up for at least 6 months.
Patients also underwent the 17-item Hamilton Depression Rating Scale (HDS) to assess depression status [15].

Patients were not taking psychotic or anti-dementia medications. Five were on stable doses of prescribed medicines (two were on anti-hypertensives, two on gastric medications, and one on anti-hyperuricaemia drugs). All patients were right-handed. We explained the aim of this study to all patients and obtained their informed consent.

HDS, Hamilton Depression Scale; MMSE, Mini Mental State Examination.

Imaging procedure All patients received [123I]iodoamphetamine (123I-IMP) at a dose of 167MBq. They were then directed to lie in the supine position for 20min with the eyes closed in a dim, quiet room. A SPECT image acquisition was performed using a double-head HITACHI/ADAC gamma camera ‘FORTE’ equipped with a low-energy, generalpurpose collimator (VXGP). A total of 64 projections (32 per head) were obtained at 25s per projection into a 128128 image matrix. A Butterworth filter was used for SPECT image reconstruction at 0.125 cycle/pixel. Attenuation correction was performed using Chang’s method.

Data analysis rCBF was projected onto surface pixels by using a 3-D SSP technique. 3-D SSP involves three steps of re-orientation, projection and transformation. Stereotaxic re-orientation of each brain into a standard shape was performed by non-linear warping with the anterior commissure–posterior commissure line as the anatomical line of reference. Subsequently, a method of data extraction was applied in which the cortical activity was projected onto the brain surface. Finally, as each brain had been stereotactically transformed into a standard surface image format, it was possible to compare the resultant cortical projections with a normal database by using a zscore subtraction on a pixel-by-pixel basis.

Twenty-one ROIs (a total of 42 regions including both sides of the ROIs) were placed on a z-score map, and analysed by 3-D SSP. (The name of each ROI is given in Table 2.)

Pixel values of an individual’s image set were normalized to the global brain value before the analysis. In this study, we also included data from 20 normal subjects who were matched for age and sex (mean age, 66.5 years; MMSE score, 28.83) on a surface pixel-by-pixel basis. All subjects underwent MRI and were confirmed to have no focal lesions. A z-score was calculated for each pixel of the cerebral surface as follows: z-score=(normal mean–individual value)/normal standard deviation.

Twenty-one circular regions of interest (ROIs) consisting of 97 pixels were placed on a z-score map of 3-D SSP (Fig. 1, Table 2) as follows: inferior frontal gyrus including Brodmann areas (BA) 46/45; superior frontal gyrus including BA 10 (SFGBA10); superior frontal gyrus including BA 8/9; superior precentral gyrus including BA 4/6; inferior precentral gyrus including BA 4/6; superior post-central gyrus including BA 1/2/3/40; inferior postcentral gyrus including BA 1/2/3/40; lateral parietal lobe including BA 40/39; lateral occipital lobe including BA 18/19; lateral parieto-temporal including BA 21/22; lateral posterior temporal including BA 21/22; lateral anterior temporal including BA 21/28; precuneus including BA 7; posterior cingulate gyrus including BA 29/30; cingulate gyrus including BA 23/24; anterior cingulate gyrus including BA 24; thalamus; medial temporal lobe including BA 28/36; medial frontal including BA 8/32; medial frontal including BA 9/32; and medial frontal including BA 10/32 (MFBA10). Results obtained from z-scores of the ROIs were expressed as means of each pixel.

BA, Brodmann area.

Statistical analysis

The significance of statistical differences was determined by the two-sided Mann–Whitney U-test in age, the HDS scores, MMSE scores and ROI values between the MD and the AD groups. Results were expressed as mean± standard deviation, and differences with P values of less than 0.05 were considered to be statistically significant. Linear discriminant function analysis was performed to determine ROI values for the areas that were the best predictors of diseases.

Results

The MD group consisted of five men and five women, with a mean age of 57.1±4.9 years, HDS score of 22.3±2.2, and MMSE score of 21.1±2.3. The AD group included four men and six women, with a mean age of 60.3±4.1 years, HDS score of 19.9±3.1, and MMSE score of 19.3±1.8. There were no significant differences in age (P=0.131, U=70), HDS score (P=0.082, U=27), or MMSE score (P=0.096, U=28) between the two groups.

Characteristics of regional cerebral blood flow patterns rCBF in the lateral temporal, parietal and precuneus decreased more in the AD group than in the MD group, while in the lateral and medial frontal it decreased more in the MD group than in the AD group (Figs 2 and 3).

The mean of each ROI z-score in the two groups was calculated (Table 3). In the AD group, decreased blood flow was observed compared with normal subjects in the region from the lateral parietal lobe to temporal lobe, anterior cingulate gyrus, pons and medial temporal lobe (Fig. 4). On the other hand, in the MD group, decreased blood flow was noted in the medial frontal lobe, lateral inferior frontal lobe, cingulate gyrus, pons the medial temporal lobe (Fig. 5).

( Fig. 2 ) Regions of decreased perfusion of the Alzheimer’s disease (AD) group in comparison with the major depression (MD) group. Values of each pixel of the AD group and MD group were statistically analysed with Student’s t-test to calculate z-scores of mean pixel values, which were then projected and displayed on the brain surface map. ( Fig. 3 ) Regions of decreased perfusion of the major depression group in comparison with the Alzheimer’s disease group.

Discussion

For both Alzheimer’s disease and major depression, there are various diagnostic criteria such as DSM-IV and the International Classification of Disease, Diagnosis Criteria for Research, 10th edition (ICD-10) [16]. These criteria are used in clinical settings to perform the differential diagnosis between Alzheimer’s disease and major depression in the early elderly. Most of diagnostic criteria adopt only clinical symptoms, and not diagnostic imaging or laboratory test values. Therefore, differential diagnosis using these criteria is difficult except for expert psychiatrists.

It has been reported that pseudodementia observed in major depression patients and forgetfulness in Alzheimer’s disease patients show different clinical manifestations [17]. However, the difference is often not clear in the initial phase of the disease, although it becomes remarkable as the disease progresses.

Recently, anti-dementia drugs such as cholinesterase inhibitors have been introduced to treat Alzheimer’s disease. It is reported that the drug will be more efficacious when administered from the earlier phase of the disease [18]. The effective use of anti-dementia drugs drives down the cost of care in nursing homes, and is cost-effective for all medical services for the early elderly [19]. It is also known that, even in elderly patients with major depression, therapeutic intervention from the early phase leads to a decrease in suicide attempts [3].

In the present study, the images obtained by brain perfusion SPECTwere analysed using 3-D SSP, and ROIs were placed on the z-score map obtained, in order to clarify the features of Alzheimer’s disease and major depression.
Comparison of z-scores between the Alzheimer’s disease and major depression groups z-scores were compared statistically between the AD and MD groups. The z-scores for the lateral parietal (bilateral superior post-central gyrus, lateral parietal lobe and right lateral occipital lobe), lateral temporal (left lateral parieto-temporal and lateral anterior temporal), bilateral precuneus and bilateral posterior cingulate gyrus regions were significantly reduced in the AD group compared with those in the MD group. The z-scores for the lateral frontal (bilateral inferior frontal gyrus and SFGBA10), left thalamus and bilateral MFBA10 regions were significantly lower in the MD group than in the AD group.

Compared to normal subjects, our AD group showed reduced blood flow in the lateral parietal lobe, lateral temporal lobe, anterior cingulate gyrus, pons and medial temporal lobe, which was consistent with the results of previous reports [20,21]. On the other hand, the MD group was characterized by decreased perfusion in the medial frontal lobe, lateral inferior frontal lobe, anterior cingulate gyrus, pons and medial temporal lobe [21]. The regions where decreased blood flow was observed in both groups were therefore the anterior cingulate gyrus, medial temporal lobe and pons. Reduced perfusion in the prefrontal area consisting of the medial frontal lobe and lateral inferior frontal lobe was specific to the MD group. It is thought that the regions of the prefrontal area, anterior cingulate gyrus, medial temporal lobe and hippocampus are involved in a network associated with the pathologies of depressive states [22]. These regions of this network corresponded to the regions where reduced perfusion was observed in our study.

Refer to Table 2 for name of each region of interest. Alzheimer’s disease group significantly more than major depression group, *P<0.05, **P<0.01. Major depression group significantly more than Alzheimer’s disease group, wP<0.05, wwP<0.01.

Although impaired rCBF in the anterior cingulate gyrus has been reported as a feature of elderly patients with major depression [7], it is neither specific to major depression nor allows differentiation between Alzheimer’s disease and major depression. It is commonly observed as a part of abnormal perfusion in the network in the various depressive states as shown in this study and previous studies [23].

Rather, the prefrontal perfusion abnormality clearly depicted on the z-score maps should be noted in discriminating major depression and Alzheimer’s disease. In this study the MD group showed a significant decrease in rCBF in the prefrontal area, while the AD group did not. Precise estimation of the distribution of z-scores in the aforementioned ‘depression’ network might enable differential diagnosis of diseases causing depression.

In stereotactic methods such as 3-D SSP and statistical parametric mapping, artifacts resulting from inaccuracy in stereotaxic transformation may occur in atrophic regions of the brain. However, findings obtained by 3-D SSP have been reported to be less affected by such artifacts than statistical parametric mapping [24]. The effect of artifacts on the findings obtained by 3-D SSP in this study was thought to be negligible as our patients, who were of younger age and had good MMSE scores, did not have remarkable brain atrophy.
We must mention that in 3-D SSP, the choice of reference region for pixel normalization may affect the accuracy of disease characterization, and is still controversial. In diseases with widespread metabolic or perfusion reduction, the pons and thalamus may be preferable [11,20].

z-scores of each ROI in the Alzheimer’s disease group. The circles deeper in colour indicate the higher z-scores: white, less than 1.0; light grey, 1.0–1.5; dark grey, 1.5–2; and black, 2.0 or more. Decreased blood flow was observed in the regions from the lateral parietal lobe to temporal lobe, anterior cingulate gyrus, pons and medial temporal lobe.
z-scores of each ROI in the major depression group. Decreased blood flow was observed in the medial frontal lobe, lateral inferior frontal lobe, cingulate gyrus, pons and medial temporal lobe.
ROI values in the lateral parietal lobe (LPL) and superior frontal gyrus, including Brodmann area 10 (SFGBA10), observed in major depression (K, n=10) and Alzheimer’s disease (D, n=10) groups. A scatter plot was presented with SFGBA10 value on the horizontal axis and LPL value on the vertical axis. A straight line, LPL=1.24 SFGBA10+1.04, obtained from a linear discriminant analysis represents the differentiation criteria between Alzheimer’s disease and major depression. Correlation ratio, 0.71; error ratio, 7.04%.

distribution using linear discriminant function analysis. ROIs, where perfusion was supposed to be observed and a decrease in perfusion specific to Alzheimer’s disease and major depression, were evaluated. Among all the combinations of ROIs in Alzheimer’s disease and those in major depression, the combination of a parietal ROI of the lateral parietal lobe and a prefrontal ROI of SFGBA10 was the most efficient in discriminating the two diseases. The higher z-score at the right and left sides of ROIs were regarded as the ROI values of each individual, and a scatter plot was presented with an SFGBA10 value on the horizontal axis and lateral parietal lobe value on the vertical axis (Fig. 6). The plot area could be divided into two parts, those of major depression and Alzheimer’s disease, by drawing a discriminant line obtained in the linear discriminant function analysis. All the patients were correctly classified into the appropriate group of major depression and Alzheimer’s disease, and the result of the function analysis was highly accurate (correlation ratio, 0.71; error rate, 7.04%).

In Conclusion

There was a difference in rCBF patterns between the early elderly with Alzheimer’s disease and those with major depression, and the difference became clear when 3-D SSP was used. Brain perfusion SPECTwith 3-D SSP might be a useful tool for the differential diagnosis between Alzheimer’s disease and major depression.

References

1 Brayne C, Gill C, Huppert FA, Barkley C, Gehlhaar E, Girling DM, et al. Cambridge project for later life. Br J Psychiatry 1995; 167:255–266.

2 Conwell Y, Duberstein PR, Cox C, Herrmann JH, Forbes NT, Caine ED. Relationships of age and axis I diagnosis in victims of completed suicide: a psychological autopsy study. Am J Psychiatry 1996; 153:1001–1008.

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store