Clinical, FDG and amyloid PET imaging in posterior cortical atrophy

This is one of the largest prospective studies to date of a cohort of patients with a clinical diagnosis of PCA that underwent FDG-PET and amyloid PET imaging. The study identifies important clinical, cognitive and neuroimaging features of PCA.

While the clinical features of our PCA patients are unsurprisingly similar to previously reported PCA cohorts [2, 4, 13, 31‐33], our study design allowed us to further refine the clinical syndrome of PCA. Our PCA cohort had an average age of onset of 60 years which might suggest that PCA should be considered an early onset dementia. However, we did not have any subjects presenting less than 45 years and hence PCA would not be considered a young onset dementia, based on our previous definitions [34]. Simultanagnosia was noted to be the most common clinical feature in our cohort in general, and was also the most common feature of early PCA, affecting 85 % of patients with disease duration of 4 years or less. Dysgraphia and oculomotor apraxia were also common features of PCA, and of early PCA. Poly-mini-myoclonus however, is not a feature that is considered, or typically assessed, in patients with PCA yet it was more common than some of the features typically associated with PCA, such as oculomotor apraxia and optic ataxia. We previously reported an association between poly-mini-myoclonus and PCA [35], although our previous study was retrospective in nature and clearly underestimated the frequency of this clinical feature compared to the current prospectively recruited cohort, which shows that poly-mini-myoclonus is a common feature of PCA. In contrast to simultanagnosia, oculomotor apraxia and poly-mini-myoclonus, our data suggested that optic ataxia, while relatively common in PCA, is a later feature of PCA. In fact, optic ataxia was not identified in early PCA. Anomia was also observed in our PCA cohort, as previously reported [4, 33, 35, 36], and is most likely due to involvement of the left temporal lobe. We also found that approximately 20 % of our cohort had visual hallucinations or RBD, with the frequency similar in early PCA and across the entire cohort. This suggests that hallucinations and RBD when associated with PCA are both likely to be observed early in the disease course. These two features are not frequently reported in PCA, but are suggestive of underlying Lewy body disease [7, 35], given these are classic features of dementia with Lewy bodies [37]. Acalculia, a feature of the Gerstmann’s syndrome, was striking and was also observed in early PCA. Prosopagnosia was noted to be present in some of the patients but was relatively mild when present in most instances and was most likely perceptual, as opposed to agnostic in nature. Neuropsychiatric, behavioral, executive and extrapyramidal features were noted to be absent to mild in PCA. Ideomotor apraxia was not a common feature and when present also appeared to be mild in most patients.

On FDG-PET analysis, we observed marked hypometabolism of the bilateral parietal and occipital lobes, showing involvement of these regions in more than 80 % of the PCA patients. Mild hypometabolism in bilateral frontal lobes, posterior cingulate and the bilateral visual cortex was observed in 50 % of the PCA patients. The bilateral anterior cingulate and medial frontal cortex appears to remain unaffected. These findings concur with previous PET studies [9‐12]. All 25 PCA patients were amyloid positive with very high SUVR ratios. This suggest, with high certainty that the underlying pathology in all of our PCA patients is Alzheimer’s disease. It is also possible however, although less likely, that in some of our cases amyloid deposition is co-occurring with another pathology, such as corticobasal degeneration that has been reported in a couple of PCA patients [6]. It is also unlikely that our patients suffered from Creutzfeldt-Jakob disease, given the disease durations and the fact that amyloid imaging is negative in Creutzfeldt-Jakob disease [38]. Furthermore, large amplitude myoclonic jerks, a feature of Creutzfeldt-Jakob disease, was not observed in our cohort. Our PCA patients instead showed poly-mini-myoclonus. Therefore, our cohort is likely very homogeneous, pathologically, although it is possible that in some patients there is co-existing Lewy body pathology.

Our association analyses allowed us to investigate the hypometabolic associations of the clinical features observed in PCA. Simultanagnosia was associated with hypometabolism in the right occipital lobe, posterior cingulum and visual cortex. These findings are expected as simultanagnosia generally reflects the severity of damage in the visual associative cortex without the involvement of the lateral parietal cortex [2, 39]. However, in our study it was found to be associated only with right-sided involvement. Optic ataxia was associated with involvement of the left occipital and visual cortex, while oculomotor apraxia with left parietal lobe and posterior cingulum. These results further quantify the neural correlations of oculomotor apraxia to the posterior parietal cortices. In fact, an fMRI study recently reported finding an association between the lateral occipitoparietal junction (dorsal stream) and changes in graspable stimuli and visually guided reaching to grasp [40, 41]. The optic ataxia findings suggest that occipital lobe involvement needs to be present before this PCA feature will emerge. We did not find any association between two of the features of Gerstmann’s syndrome (acalculia and agraphia) with any hypometabolic patterns, which could be due a small sample size of our cohort; however, these findings have been shown to be associated with the left parietal cortex by another group [42].

One of the important aspects of our study was the assessment of clinical and FDG-PET features of early PCA. We observed little difference in early PCA compared to the entire cohort suggesting most features of PCA will be present early in the disease course. The only exception appeared to be optic ataxia which was not a feature of early PCA. Interestingly, the presence of optic ataxia was associated with left occipital lobe involvement which was a region found to be less affected in early PCA. And although we do not have longitudinal data, we can speculate that progressive involvement likely spreads from the right parietal lobe across to the left occipital lobe, which would explain the later emergence of optic ataxia.

All 25 PCA patients had SUVR ratios consistent with beta-amyloid deposition. Beta-amyloid is one of the two major proteins required for a pathological diagnosis of Alzheimer’s disease; tau being the other. Our group analysis of PiB did not show a regional pattern of beta-amyloid deposition unique to PCA. Prefrontal, temporal and medial parietal involvement has also been observed with other Alzheimer’s disease related syndromes [28, 43, 44]. The lack of beta-amyloid deposition being more striking in the occipitoparietal regions, mirroring the pattern of hypometabolism on FDG-PET, is consistent with pathological studies demonstrating greater tau, but not beta-amyloid deposition in these regions [4]. The subjects in this study all met clinical criteria for PCA [4, 13]. All subjects also met stringent research criteria for PCA. Given that we did not have any amyloid negative subjects in our cohort, it appears that our research criteria may be more specific to underlying Alzheimer’s disease than the clinical criteria, since other PCA cohorts have found subjects with non-Alzheimer’s pathologies [5, 6]. We have previously reported two subjects who presented with aphasia but were agnostic to their visual deficits [45]. Visuospatial/perceptual deficits were strikingly abnormal on neuropsychological testings and was the most severely impairment cognitive domain in both. Such subjects remain controversial in terms of classification. In fact, both were excluded from this study as they did not meet our research criteria for PCA.

The major strength of this study is the large number of PCA cases that were prospectively recruited and underwent amyloid and FDG-PET imaging. Another strength was the assessment of PCA subjects with short disease duration allowing the assessment of early features of PCA. A limitation of the study was that not all subjects underwent FDG-PET. In addition, pathological confirmation of Alzheimer’s disease was not available. The lack of autopsy also meant we could not confirm an association between the presence of hallucinations and RBD and co-existence of Lewy body disease.