Memory Impairment at Initial Clinical Presentation in Posterior Cortical Atrophy

INTRODUCTION

Proposed diagnostic criteria for posterior cortical atrophy (PCA) [1, 2] emphasize core visuospatial and visuoperceptual deficits, not attributable to ocular disease. Predominant atrophy is in the parieto-occipital cortex, with relative sparing of the temporal regions. The most common underlying pathology is Alzheimer’s disease (AD) [3 –5]. Episodic memory is thought to be relatively preserved in PCA. It is suggested that anterograde memory is not impaired in the early stages, but gradually deteriorates with progression of disease [1, 6]. As a result, memory has rarely been examined systematically in PCA. Where memory assessments have been reported, the findings have been variable. Comparing PCA with typical AD (tAD), in which the hallmark feature is one of early and profound episodic memory impairment, McMonagle et al. [1] found that PCA patients scored lowest on measures of visuoconstruction, as expected, but in addition there was no significant difference compared with tAD on measures of memory. PCA patients achieved less than half the maximum score on memory testing, similar to tAD performance. Similarly, Charles and Hillis [7] found that memory, as assessed by a verbal list learning task, was impaired in both PCA and tAD patients, although tAD patients showed greater impairment on delayed recall. Conversely, Kas et al. [8] found that PCA patients performed significantly better than tAD patients on an episodic verbal memory test, on both free and cued recall components. The variable findings could be due to differences in disease duration, however Kas et al. [8] divided PCA patients into short (≤3 years) and long (>3 years) disease duration groups, and reported no significant differences in neurological or neuropsychological symptoms between the two PCA groups. Each of these studies also administered different memory tests, and in different sample sizes, which may also explain the discrepant findings.

Misdiagnosis in PCA is common due to the relative rarity of presentation and unusual symptoms. Patients frequently describe a plethora of referrals and procedures undertaken by opticians and ophthalmologists in the belief that an ocular abnormality accounts for the visual deficits experienced. Equally common are patients presenting with an accompanying mood disorder which is then wrongly considered to be the primary diagnosis [9]. The presence of mild or even more prominent memory impairment early in PCA may assist in the recognition and diagnosis of PCA and prevent delays in diagnosis.

The aim of this study was to examine memory performance in PCA patients at initial clinical presentation to a cognitive disorders clinic, compared to early-onset AD (EOAD) patients with a typical memory impaired presentation.

MATERIALS AND METHODS

RESULTS

DISCUSSION

Diagnostic criteria for PCA have emphasized salient visual deficits in the absence of ocular disease, and preserved or only mildly impaired episodic memory. Consequently, memory has rarely been systematically examined in PCA and the few studies that have included memory assessment report variable findings [1 , 8]. The results of this study show that memory is often impaired in PCA patients at clinical presentation. 66.7% of PCA patients had impaired memory compared to healthy controls, suggesting that early memory impairment is common, although not ubiquitous in PCA. Furthermore, while PCA patients could be characterized by salient and significant deficits in visuospatial features compared to controls and EOAD patients, memory was also impaired in PCA. The mean scores were indistinguishable between the two patient groups. Early impairment in memory alongside the defining visual disturbances is consistent with the handful of past studies that have included simple memory measure in the neuropsychological assessment [1, 7]. However, the reasons for impairment are not clear. We propose three possible explanations.

Firstly, impairment may be driven by the primary pathology in the parietal cortex, specifically affecting attentional networks underlying memory. In task-based neuroimaging studies, the posterior parietal cortices are frequently activated in response to episodic memory retrieval [12]. A meta-analytic review [13] identified the parietal lobes as being more consistently activated during recollection, identified in 90% of studies reviewed, compared to more typically considered episodic memory regions such as the medial temporal lobe. Neuropsychological studies of patients with bilateral parietal lobe lesions have found that patients are impaired on tests of recollection, such as autobiographical memory retrieval [14] or remember/know paradigms [15], and provide few ‘remembered’ responses, and lack of specificity and richness of freely recalled autobiographical events.

Secondly, PCA patients may have more widespread pathology extending to medial temporal memory-related structures. Tang-Wai et al. [2] described a group of PCA patients in detail to form the basis of currently used diagnostic criteria, and found relatively preserved medial temporal lobe structures, which could be appreciated antemortem on MRI scans. Conversely, Migliaccio et al. [6] found that regions of grey matter reduction in PCA included the hippocampus bilaterally. Studies examining cortical thickness patterns have also reported reductions in the medial temporal lobe [16]. The majority of our patients presented within 2-3 years of symptom onset, and imaging was consistent with salient parieto-temporal atrophy. However, more detailed imaging analysis may have highlighted more extensive deficits.

A third possibility is that verbal memory impairment may be attributable to disruption of language networks in the left hemisphere. A recent study described subtle language impairments in PCA characterized by anomia, poor phonemic fluency, and slow speech rate, resembling the linguistic profile most consistent with logopenic aphasia [17]. Language was not assessed in detail at clinical presentation in the PCA patients reported in this study. However, it is worth noting the PCA patients assessed in Crutch et al. [17] had an average symptom duration of 6 years, more than twice that of the majority of PCA patients in this study, and thus it is suggested that language impairment may be a feature of more advanced stages of PCA.

A small percentage of PCA patients scored within the control range. These patients were comparable to the impaired PCA group in age, education and, importantly, in symptom duration. Preserved memory in this small group may be due to individual variation in disease expression, or due to the insensitivity of the ACE-R to subtle memory impairment.

There are limitations to this study. The ACE-R is a brief screening tool and was used to demonstrate that memory impairment can be detected in PCA with a broad measure at presentation. The ACE-R is likely, therefore, to underestimate the extent of memory impairment. It is also unable to provide detailed information as to the nature of episodic memory impairment in PCA and EOAD. Further investigation with detailed neuropsychological testing of memory function in PCA is therefore essential. However, the results support the conclusion that memory impairment can be detected at initial presentation with a simple screening instrument, and with a relatively short average symptom duration compared to other studies (for example, [8, 17]). Despite differences in the predominant clinical symptoms in PCA and EOAD, the concomitant and significant memory impairment early in the clinical course may reflect the common underlying AD pathology. A similar finding has been reported in an earlier study, examining patients with logopenic aphasia (lv-PPA), characterized by impaired repetition of sentences and naming and in which the underlying pathology is AD in the majority of patients. Flanagan et al. [18] found that lv-PPA patients were significantly impaired in memory, and did not differ from a tAD group, suggesting that identifying specific impairments in memory may supplement diagnosis of patients with pathology.

These findings have clear clinical implications. Current diagnostic criteria emphasize the relative preservation of memory in PCA, yet the findings provide evidence of a detectable memory deficit early in the course of disease. Further study is warranted to determine how diagnostic criteria may need to be revised. Diagnostically, the presence of memory impairment may assist in the early identification of PCA, helping to differentiate from visual disturbances caused by ocular or other problems. Improving our understanding of more focal presentations of AD is imperative to facilitate accurate prediction of AD pathology in life, and support the development of target therapies for AD. Studies suggest that variant clinical presentations of AD are not so infrequent, representing a third of patients with early-onset AD, and at least 5% of late onset cases [19]. In terms of intervention and management, while it is entirely appropriate to centre care and advice on the critical visual impairments experienced by patients, it is also important to consider problems arising from memory impairment. The literature on tAD cites extensive examples of the pervasive difficulties memory impairment incites, e.g., in restricting activities of daily living [20]. Memory impairment therefore must be considered in assessment and management of PCA.