Clinical differentiation between frontotemporal dementia and Alzheimer's disease : psychometric, behavioral, neuroimaging and neurophysiological information

Abstract

Frontotemporal dementia (FTD) initially affects the anterior regions of the brain, and gradually spreads to other cerebral areas. Behavioral alterations are described as the hallmark of FTD, whereas cognition is mostly found to be relatively spared. Alzheimer's disease (AD) starts in the posterior brain areas, and successively involves even other regions. Typical for AD are cognitive deficits, but behavioral and emotional changes have also been reported. With progression, the clinical distinctions between the diseases may be blurred, and it is uncertain to what extent and by which means they are possible to clinically differentiate. FTD is in focus of the thesis. The general aim of the thesis was to investigate FTD and AD from a differential diagnostic point of view, with the use of preclinical, neuropsychological (psychometric tests), behavioral, volumetric, and radiologic information.

In study I it was found that is possible to differentiate FTD from AD with the help of small subsets of behavioral observations (5 items), and neuropsychological tests (5 tests). The hit rate of the set of behavioral items was very high, 97%. These abnormal behaviors were more pronounced in FTD than in AD. The hit rate for the psychometric predictors was slightly lower, 90%. In this set of psychometric predictors was included a test of verbal fluency. This function was particularly low in the FfD group as compared to the AD group.

The topic for study II were the earliest signs of FTD and AD. The results indicated that clear behavioral and cognitive differences existed between the FTD and AD patients prior to their first clinical visit. The earliest signs of FTD were changes in behavior, whereas the first symptoms of AD were cognitive, particularly memory deficits. Disinhibition was the most prominent early behavioral alteration in FTD, and it was associated with right-sided frontotemporal atrophy. Language dysfunctions was commonest in the left-sided group and loss of executive functions most frequent in the FTD group with bilateral frontotemporal degeneration.

In study III was investigated the relationship between behavior, cognition, and loss of regional frontotemporal brain volumes in FTD patients. The study suggested that the frontal lobes have a limited importance for cognition, and that there is a relation between behavior and cognition. The correlations between behavior, cognitive functions and regional volume loss did not show any consistent patterns, outlining that thinking is such a complex process that it demands cooperation from different parts of the brain in both hemispheres. This multiregional dependency of thinking is probably reflected in neuropsychological tests.

In study IV it was found that FTD patients were marked by a pathological EEG, namely by an absence of an increase in slow quantitative EEG (qEEG) activities, and a decrease in fast activities. AD patients were characterized by an increase in slow qEEG frequencies and a smaller decrease in fast activities. Neuropsychological measures were better predictors of FTD versus AD than qEEG measures, but the most efficient predictor was a model combining neuropsychological tests and qEEG. The classification accuracy for this combination of modes of investigation amounted to 93.3%.

Study V showed a typical pattern of anterior cortical hypoperfusion in FTD, and of posterior cortical blood flow reduction in AD patients. Calculations of likelihood ratio (LR) disclosed that regional cerebral blood flow (rCBF) measures added more to the pretest probability than the psychometric measures. The best rCBF predictor of FTD versus AD was the perfusion in the left anterior cingulate cortex, where the LIZ was 11.9, which is very high.

When comparing the efficiency of the five modes of investigation for the differentiation between FTD and AD, the behavioral observations and rCBF measurements appears to be equally useful. Neuropsychological tests may be useful, provided that the patient groups are enough dissimilar. The EEG method must be used in combination with neuropsychological tests to contribute to classification accuracy.