Research Article: Can cognitive assessment really discriminate early stages of Alzheimer’s and behavioural variant frontotemporal dementia at initial clinical presentation?

Date Published: August 9, 2017

Publisher: BioMed Central

Author(s): Sophia Reul, Hubertus Lohmann, Heinz Wiendl, Thomas Duning, Andreas Johnen.


Neuropsychological testing is considered crucial for differential diagnosis of Alzheimer’s disease (AD) and behavioural variant frontotemporal dementia (bvFTD). In-depth neuropsychological assessment revealed specific dysfunctions in the two dementia syndromes. However, a significant overlap of cognitive impairments exists in early disease stages. We questioned whether a standard neuropsychological assessment at initial clinical presentation can delineate patients with AD versus bvFTD.

In a retrospective approach, we evaluated and compared how cognitive profiles assessed at initial clinical presentation predicted the diagnosis of later verified AD (n = 43) and bvFTD (n = 26). Additionally, the neuropsychological standard domains memory, language, visuospatial skills, executive functions, praxis and social cognition were subjected to stepwise discriminant analysis to compare their differential contribution to diagnosis.

Regardless of diagnosis, a percentage of patients presented with major deterioration in a wide range of cognitive domains when compared with age-matched normative data. Only few significant differences were detected on the group level: Patients with AD were relatively more impaired in the verbal recall, verbal recognition, figure copy, and surprisingly in the executive subdomains, set shifting and processing speed whereas bvFTD was characterised by more deficits in imitation of face postures. A combination of tests for verbal recall, imitation of limb and face postures, and figure copy showed the greatest discriminatory power.

Our results imply that the contribution of a standard neuropsychological assessment is limited for differential diagnosis of AD and bvFTD at initial presentation. In contrast to current clinical guidelines, executive functions are neither particularly nor exclusively impaired in patients with bvFTD when assessed within a standard clinical neuropsychological test battery. The significant overlap of bvFTD and AD concerning the profile of cognitive impairments questions current neuropsychological diagnostic criteria and calls for revision, regarding both the degree and the profile of cognitive deficits.

Partial Text

Behavioural variant frontotemporal dementia (bvFTD) and Alzheimer’s disease (AD) are two of the most common dementia syndromes affecting people under the age of 65 [1]. A correct and early differential diagnosis is crucial for disease management and treatment [2]. In patients with suspected dementia, a comprehensive neuropsychological examination is an essential diagnostic element besides history taking, evaluation of regional brain atrophy on magnetic resonance imaging (MRI), and a cerebrospinal fluid (CSF) biomarker profile. Clinically, AD is characterised by progressive cognitive decline, especially in episodic memory, spatial perception and working memory, deriving from temporal and medial parietal lobe atrophy [3]. In contrast, patients with bvFTD predominantly present with severe behavioural deterioration, personality changes and social dysfunction, associated with prominent frontal and anterior temporal lobe atrophy [4]. Besides the behavioural conspicuities, a typical cognitive profile described as “executive deficits with relative sparing of memory and visuospatial functions” (page 2460) represents one of the six diagnostic criteria [4]. The neuropsychological item has been described as highly sensitive for bvFTD [5]. Therefore, testing for cognitive dysfunction is an important and auxiliary element in the diagnostic process. However, recent evidence has revealed that patients with bvFTD also show a range of cognitive deficits similar to those found in patients with AD, especially in early stages of the disease [6, 7]. In particular, standard memory tests have failed to reliably discriminate the two diseases [8]. Due to such clinical overlap, the differentiation of bvFTD and AD in early disease stages remains challenging [6, 9, 10]. To address the issue of partly overlapping cognitive impairments in AD and bvFTD, researchers in several recent studies did in-depth investigations of neuropsychological domains such as memory, executive function, praxis and social cognition [8, 11–21]. These studies succeeded in delineating the two diseases by investigating single cognitive domains with extensive but often time-consuming neuropsychological test batteries. However, it remains unclear whether a standard neuropsychological examination as usually applied in specialised memory clinics actually contributes to correct differentiation between AD and bvFTD and which of the aforementioned neuropsychological domains has the highest value for differential diagnosis.

Using an innovative approach, we examined the usefulness and predictive diagnostic ability of naive standardised neuropsychological testing at initial clinical presentation for the differential diagnosis of early stages of bvFTD and AD. For this purpose, we retrospectively evaluated the neuropsychological performance of patients with validated dementia subtype diagnosis based on typical MRI atrophy pattern, CSF dementia biomarker profile and clinical follow-up presentation (Fig. 1). Because researchers in previous studies have described typical cognitive profiles for AD and bvFTD pertaining to single cognitive domains and when using in-depth assessments [11–21], we aimed to investigate typical patterns of impairment in patients with bvFTD in a standard neuropsychological assessment employed in clinical neuropsychological routine.

Patients with bvFTD present with significant impairments in a broad range of cognitive domains at initial clinical presentation. In contrast to current clinical guidelines [4], executive function is neither particularly nor exclusively impaired when assessed using a standard clinical neuropsychological test battery. This result calls for a revision of the neuropsychology item in current diagnostic criteria for bvFTD, including (a) an elaboration and operationalisation of the proposed executive dysfunction and (b) deterioration in other cognitive domains, such as memory, visuoconstruction, praxis and social cognition. We found a significant overlap of bvFTD and AD concerning the profile of cognitive impairments, complicating early differential diagnosis of these dementia subtypes in clinical practice, especially in individual cases. Our results imply that the contribution of a standard neuropsychological assessment to differential diagnosis at initial clinical presentation is limited and argues for the additional use of more specialised and expansive test batteries. Especially, the domains social cognition, attention, praxis and executive function may benefit from more sensitive assessments to display their diagnostic power. Because neuropsychological testing is an economically worthwhile and non-invasive diagnostic procedure, further studies need to be done to investigate specific neuropsychological assessments for differential diagnosis between bvFTD and AD. Whenever such specialised neuropsychological diagnostics cannot be provided, findings of our discriminant analysis indicate that the standard tasks for verbal recall and figure copy as well as imitation of face and imitation of limb postures provide a relatively good efficacy for differential diagnosis between bvFTD and AD. Thus, an employed standard test battery should at least cover the domains of memory, visuoconstruction and praxis.




Leave a Reply

Your email address will not be published.