Posterior cortical atrophy (PCA) is manifests mainly as early visual dysfunction in the context of neurodegeneration of posterior cortical regions. Benson [1] first described this syndrome and proposed that it be independently classified. The main clinical features, besides space perception (visuospatial) and object perception (visuoperceptual) deficits, include alexia, Balint syndrome (simultanagnosia, ocular apraxia, and visual ataxia), Gerstmann’s syndrome (acalculia, agraphia, digital agnosia, and left and right confusion), with relative preservation of memory and insight [2]. PCA usually occurs from the mid-50s to early 60s and may account for 5 to 10% of cases diagnosed as early-onset Alzheimer’s disease (EOAD) [3].

Neuropathological studies have indicated demonstrated that most cases of PCA are identical to Alzheimer’s disease [4–6]. In some rare instances, corticobasal degeneration, Lewy bodies dementia and prion disease cause PCA [4, 5]. In cases of Alzheimer’s disease pathology, PCA has also been recognized and described in consensus criteria for atypical Alzheimer’s disease [7], as well as a “visual variant of Alzheimer’s disease” [8]. For a clearer understanding of the pathogenesis in AD, it is crucial to distinguish the cognitive features of PCA from those of AD. Moreover, with the multiple pathologies that underlie the PCA syndrome [5], pharmacological treatment and behavioral interventions for PCA may differ from those of AD.

Several studies have evaluated clinical differences between PCA and AD, listing core clinical and cognitive features for PCA [4, 6, 9–11]. PCA also has significant features in the form of visuospatial and visuoperceptual symptoms. However, there are some inconsistencies exist among the cognitive features described in previous studies. Mendes et al. [6] found that verbal fluency was untouched by PCA, but Tang-wai et al. [4] did not. Charles and Hillis [9] and McMonagle et al. [10] found the visual constructional ability of PCA patients was significantly more impaired than in typical AD, which contradicted findings by Mendes. McMonagle et al. [10] and Kas et al. [11], unlike other research teams, found working memory to be more prominently impaired in PCA than in AD.

The origins of the discrepancies in these findings may lie in the use of different neuropsychological tests, different research contexts, different populations, or differences in the duration of disease. Past studies have generally not included information regarding the pathology underlying PCA. This has had several consequences. First, clear differentiation of PCA from EOAD might benefit from the application of a broader set of neuropsychological measures. Second, verification of pathology is necessary to validate neuropsychological differences. That verification is particularly important, because cognitive testing may be more readily available and less expensive than other diagnostic techniques (e.g., PET imaging). The use of these neuropsychological indicators may be efficacious in the diagnosis of PCA, particularly early diagnosis.

The five studies that we found were all performed on Euro-American participants, and it seems feasible that cultural differences may affect the manifestations of PCA. China, with its own cultural characteristics, has a population estimated at more than 1.4 billion people, a substantial proportion of whom fall into the age group most likely to be affected by PCA and EOAD. For this reason, it is important, as both a practical medical concern and a potentially valuable theoretical perspective, to examine neuropsychological factors in Chinese PCA and EOAD patients. We further note that many Chinese people may have limited access to or limited ability to afford sophisticated diagnostic techniques (e.g., imaging), so a quick preliminary screening process would be beneficial.

For this reason, we compared cognitive manifestations of PCA and EOAD in Chinese patients during early stages of the disease. We used a series of neuropsychological tests to identify cognitive features of PCA that distinguish it from EOAD.

We compared the cognitive manifestations between PCA and EOAD in Chinese patients. All patients had disease durations under 4 years. The results showed PCA patients to have greater impairment on visuospatial and visuoperceptual tests than EOAD patients, while PCA patients were significantly impaired in Rey-Osterreith complex figure copying, writing, and reading tests, with symptoms of visuo-constructive agnosia, alexia, and agraphia. However, both PCA and EOAD patients had impairments of calculation, episodic memory, working memory, executive function, picture naming, and verbal fluency in the present study. The difference in PCA and EOAD in cognitive features could help the differentiation in clinical diagnosis during early stages of the disease.

PCA patients had more severe symptoms visuospatial and visuoperceptual symptoms than EOAD patients. These results are consistent with earlier reports [2, 11, 29]. Visuospatial and visuoperceptual dysfunction arises from dysfunction of the dorsal streams (occipito-parietal pathway) and the ventral streams (occipito-temporal pathway) [30]. These higher visual dysfunctions are core clinical and cognitive features that define the PCA syndrome. However, PCA lacks standard neuropsychology test profiles for diagnosis or further clinical trials. These cognitive tests evaluating visual function may become a very quick-acting tool that could be used to distinguish PCA from EOAD.

We found that a significantly higher proportion of PCA patients scored lower than EOAD patients in figure copying tests. Figure copying is mainly related to visuo-constructive function, spatial processing, and visual neglect in bilateral parietal occipital lobe function [31, 32]. The figure copying deficit in our patients may be associated with dorsal stream lesioning [30]. Our findings confirm the results obtained in a previous study [9], which revealed that visuo-constructive impairment is more prominent in PCA patients than in EOAD patients. Our results also indicate that no patients with PCA had normal scores in Rey-Osterreith complex figure copying task, indicating the visuo-constructive function distortion. This test of Rey-Osterreith complex figure copying has been reported to have a 100% sensitivity and specificity with respect to distinguishing PCA from typical AD [9].

PCA patients were more impaired in writing tests than EOAD patients, suggesting that agraphia is an important cognitive feature of PCA. The impairment manifested in the form of incomplete word (stroke omissions and iteration), word substitution, picture drawing and changes in calligraphy [33, 34], in both Euro-American and Chinese PCA patients. These errors might suggest limb apraxia or visual disorientation [34, 35] usually associated with damage in the left angular gyrus and nearby regions in the parietal and occipital lobes [36]. However, in our study, another prominent presentation on writing disorders is characterized by unintelligible characters (61%) and structurally impaired characters (89%) in Chinese patients, presumably due to impairment in visuo-constructive functioning. The Chinese writing system has some unique features: It uses square-shaped characters, and it differs from alphabetic writing systems in the visual features of its orthography. Chinese characters are commonly referred to as logographs and involve combination and construction [37]. It suggests that Chinese scripts would reflect visuo-constructional ability more sensitively than alphabetic scripts, possibly accounting for PCA disrupting more visually complex logographic writing systems than alphabetic system [35]. The visuo-constructional impairment on writing test of Chinese PCA patients would be a good identification tag for PCA. A cross-cultural study recently reported that alphabetic or syllabic writing systems exhibit a static pattern, not changing over time over the last three millennia [38]. If logographic writing systems follow a similarly static pattern, we might be able to develop the writing test of Chinese characters into a sensitive diagnostic neuropsychological tool usable by individuals who do not speak or read Chinese. Because writing impairments are nonvisual disorders, clinicians might dismiss them as patient complaints. Testing the writing abilities of patients may uncover evidence of subtle impairments in visual cognition. Moreover, the writing error profile in Chinese patients was unique and may facilitate the diagnosis of PCA.

Reading ability was far more impaired in PCA patients than in EOAD patients, suggesting that alexia is a neuropsychological marker suitable for differentiating PCA from EOAD. This symptom appears often (90%) and early in PCA patients. In our study, PCA patients presented with missing words (94%), getting lost on the page (86%), and getting lost from one line to the next (67%). These clinical complaints in PCA have been attributed to visual disorientation [2, 24]. Some PCA patients misread similar characters or pronounced only half or part of the character, which may be due to visual neglect, especially considering the compound structure of many Chinese characters. Our PCA patients also exhibited another type of alexia: visual alexia (failure to recognize Chinese characters), also called pure alexia. This is thought to result from the destruction of the visual word form system or from deprivation of visual input. Visual form processing as part of the visual “what” pathway can be identified letters or graphemes [39, 40]. The visual word form can then trigger retrieval of the character’s meaning, grammatical features, pronunciation, and other characteristics. Again, English and Chinese writing systems are remarkably different: First, structurally, a Chinese character is composed of strokes and components, which is different from English, which is built on single alphabetic characters. Second, the process of reading Chinese characters involves orthographic and phonological conversion, but reading English involves morphemes and phonetic conversion [41]. Thus, in an alphabetic system, patients with pure alexia often use a style of compensatory reading known as letter-by-letter reading [42], but this does not work in Chinese system. The visual dysfunction disrupted the conversion of the visual word form to the orthographic, which may leads to pure alexia in Chinese PCA patients [43]. In the future, we plan to investigate the relationship between the visual form processing of Chinese characters and the ability in PCA patients to read as an indicator for pathogenesis of alexia.

We found episodic memory impairment in both PCA and EOAD patients during the early stages of the disease, as indicated by AVLT, but the two conditions did not differ from one another. Memory impairment may therefore be expected to yield poor results when used to distinguish PCA from EOAD. Our results are consistent with those of Charles and Hillis [9] and Ahmed et al. [44], who found memory to be impaired in both PCA and typical AD patients. Kas et al. [11], however, found that PCA patients performed better than typical AD patients on an episodic memory test, even during the early stages of PCA (≤ 3 years since onset). Additionally, the diagnostic criteria for PCA suggest that episodic memory is relatively preserved. This discrepancy in the research literature may be due to different durations of disease, different sample sizes, different memory tests, or some combination of those elements. Because reliability of results depends on the accuracy of diagnosis, our entire sample of PCA patients underwent imaging (FDG PET or SPECT) to ensure sufficient accuracy. In PET amyloid imaging, the majority of our PCA patients had AD pathology with elevated beta-amyloid deposition and diffusion in the cortex. These common pathological bases suggest a reason that both PCA and AD have memory impairment [44].

Despite the evident contrasts between PCA and EOAD patients in the present research, we urge caution in the interpretation of our results. The sample sizes were relatively small, which was unavoidable, owing to the relative rarity of PCA and the difficulties in recruiting participants during the early stage of symptom onset. As result of the small samples involved, it is possible that true differences exist between the PCA and EOAD groups, but that we did not find such differences because the small sample sizes give the experiment limited statistical power. However, we also emphasize that, along with meeting clinical criteria, neuroimaging verified that all our PCA patients had been accurately diagnosed. Most of the PCA patients also underwent PIB-PET to confirm diagnosis by AD pathology. Although nine PCA patients lacked pathological results, for minimize confounding and bias, we followed up those patients for at least 12 months to support AD diagnosis by excluding Lewy body disease, corticobasal degeneration, or prion disease. The PCA and EOAD groups were also matched for disease duration, which may have minimized the clinical differences between PCA and EOAD. Another constraint of our study is neuropsychology measures. Some neuropsychology measures (e.g., SSST, reading and writing tests) were brief and quick for practitioner, but they may result in underestimation of the extent of impairment, and result in ceiling effects in our data (e.g., a bunching of scores at the upper level in control group). Ceiling effects would also tend to obscure true differences between groups. In the future, our studies will systematically examine participants using neuropsychological tests on calculation, reading ability, and writing ability in PCA.

PCA patients present significant cognitive impairments including visuospatial, visuoperceptual, visuo-constructive, writing, and reading. As a diagnostic battery, those tests could be used to reliably distinguish PCA from EOAD. Tests of handwriting and reading appear to have a particularly high specificity for identifying PCA in Chinese-speaking patients. It would be practical to incorporate this battery of tests into existing cognitive screening for PCA, and it could be used to establish a brief and routine PCA neuropsychological examination for improving the early diagnosis of PCA.