Temporal lobe rating scale: application to Alzheimer's disease and frontotemporal dementia.(Statistical Data Included)

Abstract

Objectives--Temporal lobe atrophy as assessed by MRI can be measured in several ways. Volumetric measurements are quantitative but very time consuming and require extensive training to perform, so are not easily transferable to clinical practice. Visual rating scales, by contrast, are quick and widely applicable. Although medial temporal lobe atrophy is well described in Alzheimer's disease (AD), it is uncertain how early these changes can be detected and whether they discriminate AD from other neurodegenerative diseases, most notably frontotemporal dementia (FTD). The objectives were (1) to develop a widely applicable temporal lobe rating scale, and (2) to characterise and quantify the patterns of temporal lobe atrophy in AD versus temporal and frontal variants of FTD.

Methods--The temporal lobe assessments were made using an established hippocampal rating scale extended to incorporate additional temporal regions. This was firstly validated with volumetric analysis and then applied to 30 probable AD, 30 FTD (consisting of 17 temporal variant (semantic dementia) and 13 frontal variant) and 18 control coronal MRI images. Results--Bilateral hippocampal atrophy was found in 50% of the patients with AD. Contrary to expectations, patients with semantic dementia also had hippocampal atrophy, which for the left side exceeded that seen in AD; other regions (temporal pole, parahippocampal gyrus, and lateral temporal lobe), spared in AD, were severely atrophied in this group. The patients with frontal variant FTD occupied an intermediate position and were largely indistinguishable from AD.

Conclusions--Hippocampal atrophy is, therefore, not specific for AD. Semantic dementia can be distinguished from AD, by the presence of severe bilateral atrophy of the temporal pole, parahippocampal and lateral regions. These findings have implications for the differential diagnosis of dementias.

(J Neurol Neurosurg Psychiatry 2001;70;165-173)

Keywords: frontotemporal dementia; temporal lobe atrophy; Alzheimer's disease

The quantification of brain atrophy on MRI has been attempted using many different methods, including rating scales, linear and volumetric measurements of regions of interest, and coregistration techniques (for review see Fox [1]). Each method has advantages and disadvantages. Volumetric analysis provides an accurate and detailed measure of a predetermined circumscribed area. It is, however, a laborious, and labour intensive technique which needs extensive training. Rating scales, by contrast, are quick, and can be performed on large numbers of scans in a clinical setting, the disadvantage being that there is a loss of accuracy compared with volumetric analysis. We describe the development of a new temporal lobe scale which has been applied to patients with Alzheimer's disease (AD) and patients with frontotemporal dementia (FTD). These disorders represent the two most common causes of dementia in younger patients. [2] This scale has been validated against detailed volumetric measures of multiple temporal lobe regions.

The differentiation of AD from other causes of dementias is of practical and theoretical importance. Here we consider the practical clinical issues. Atrophy of the medial temporal lobe in patients with a diagnosis of AD, has been established using various methods, including subjective rating scales, [3, 4-6] measures of hippocampal width and volumes [7-10] and, more recently, measures of entorhinal cortical volume. [11] The major impetus for the development of methods of temporal lobe assessment has been to improve the accuracy of early diagnosis in AD, yet it is unclear whether medial temporal lobe atrophy is specific for AD or can also be a feature of other dementias [12] as few studies to date have compared patients with AD with those with other forms of degenerative disease. One notable exception is a pair of studies by Frisoni et al who compared AD and FTD using firstly linear measurements [13] and subsequently volumetric techniques, [14] and demonstrated medial temporal lobe atrophy. In both of these s tudies, however, patients with progressive aphasia or prominent temporal lobe pathology were excluded.

The classification of the non-Alzheimer focal lobar degenerations is debated, and there has been a plethora of terminologies. [15] In keeping with previous studies from our group [16-19] and in line with international consensus criteria [20] we have subdivided patients with FTD into those with a predominant frontal presentation, and those with semantic dementia (progressive fluent aphasia), a temporal presentation. Patients with progressive non-fluent aphasia have not been included in this study. The frontal variant of FTD (fvFTD) presents with progressive and insidious behavioural disturbance and personality change. [18] The label semantic dementia has been adopted to describe the temporal variant of FTD [21, 22] because it captures the essential disorder of impoverished general knowledge about objects, people, and the meanings of words. The distribution of atrophy in this syndrome, which can easily be mistaken clinically for AD, has not been systemically investigated. Visual inspection of brain MRI has sug gested that the hippocampal complex is preserved in semantic dementia compared with AD and although pathological studies suggest hippocampal involvement occurs eventually, it is uncertain whether this can be detected in [15]

The principal aim of the current study was to develop a reliable, but quick, method of temporal lobe assessment using MR images which might have wider applications in diagnosis. We developed a new temporal lobe rating scale which incorporates a well validated assessment of the hippocampus [3] but extended this methodology to other temporal lobe structures. This method was validated against volumetric measures and assessed for reproducibility.

The temporal lobe rating scale was then applied to larger groups of patients with AD and both presentations of FTD--namely, semantic dementia and fvFTD--matched for overall disease duration. The aim was to characterise and quantify the pattern of atrophy in a range of carefully defined temporal lobe regions in patients presenting with FTD and mild to moderate stage AD. Based on prior clinical and neuropsychological studies …