The clinical descriptions of families and individuals having mutations in PGRN
have occupied the clinical spectrum of FTLD - cognitive impairment (executive dysfunction and/or aphasia) and behavioral changes, variably accompanied by parkinsonism.1-5
This likely represents the core phenotypic features, but in this family, prominent early memory impairment was a common clinical feature. Four members of this family presented with amnestic complaints and followed a course typical of AD, and this remained their final diagnosis in two. Another member of this family presented with prominent memory impairment thought to represent Alzheimer disease, although the patient later followed a course more suggestive of FTD. Although this early memory impairment was confirmed by detailed neuropsychological testing in only two individuals, the available medical records and family accounts provided ample detail regarding other individuals' early clinical courses. Personality change was an early finding in only three affected individuals. Similarly, executive dysfunction was an early finding in only two or three individuals. Thus, a majority of affected members presented with prominent early or initial memory impairment, while a distinct minority presented with early or initial behavioral/executive dysfunction.
Prominent early memory impairment is characteristic of amnestic MCI and Alzheimer's disease, and is an exclusionary criterion for the diagnosis of FTD.14
Generally, the neurofibrillary tangles in the mesial temporal lobes characteristic of Alzheimer's disease form the pathological substrate for this type of prominent memory impairment. In this family, three individuals having a clinical presentation and subsequent clinical course typical of AD were found to have FTLD-U plus NII as the pathological substrate. This clinical phenotype is distinct from that seen in other described kindreds having mutations in PGRN
associated with FTLD-U NII neuropathology,5
and is very atypical in association with familial FTLD neuropathology in general.
Our data confirm the phenotypic heterogeneity observed within other kindreds. Several members had primarily amnestic difficulties, one member followed an early course typical of PPA, another followed the typical course of FTDP, and one presenting with amnestic complaints evolved into a clinical picture of FTD within one year. Similar heterogeneity has been reported both within families and among families with identical mutations in MAPT
Parkinsonism has been a predominant phenotype in other kindreds,16
but it was not a dominant clinical finding in this kindred. When present, parkinsonism developed late in the clinical course and was generally appreciated on examination (rather than tremor, falls and bradykinesia being noted by family members). Whether parkinsonism is more predominant among certain PGRN
mutations remains to be seen, similar to parkinsonism being the predominant phenotype in the N279K mutation in MAPT
The mean age of onset in generation II (GII) was 75.8 years, while that in generation III (GIII) was 60.7 years, suggesting an earlier onset in the GIII generation. Plus, a number of patients in both generations were evaluated early in their course (and in two cases presymptomatically) by a behavioral neurologist. One could argue that a heightened suspicion of a familial disorder would lead the GIII subjects to possibly seek medical attention earlier in their course than those in the GII generation, but we find it difficult to explain a 15 year difference in the mean age of onset based purely on this possibility. The fact that behavioral neurologists evaluated individuals early in the course in both generations also argues against this. Individuals in GII and GIII carry the same PGRN mutation, possibly implying that other genetic or environmental factors may account for the observed difference in age of onset.
In some families with PGRN
mutations, the same cerebral hemisphere was maximally affected.18, 19
The reasons for this remain unclear. In this kindred, no hemispheric predilection was present, with widely variable patterns of atrophy.
The MRI abnormalities typically reported in sporadic FTD and familial FTD have been frontotemporal cortical abnormalities varying from symmetric to markedly asymmetric atrophy.20-22
Most descriptions have not reported subcortical white matter signal changes. One individual in this kindred had striking subcortical white matter signal change adjacent to the regions of maximal cortical atrophy, and in one other case, a milder degree of subcortical signal changes was present. MRI scans have been reported in few cases of PGRN
mutations and it is not yet possible to determine whether these signal changes are sensitive or specific for detecting PGRN
mutation-positive FTLD. Volumetric analysis of a small number of individuals having FTLD-U neuropathology suggested more severe and widespread frontal, temporal and also parietal atrophy among those having PGRN
Brain tissue was examined in six cases, all of which demonstrated features of FTLD-U with NII. As has been previously reported, the presence of NII is a feature of FTLD-U with PGRN
mutations, although it is not specific for this diagnosis.24, 25
Case II.1 had mixed FTLD-U plus NII, vascular and Alzheimer's disease pathologies. A recent study suggested that immunohistochemical staining for ubiquitin and/or TDP-43 may be a useful adjunct to identify FTLD-U pathology in the presence of additional pathological processes,24, 26
and this technique was used to confirm the presence of FTLD-U pathology in Case II.1.
As in previously reported kindreds, the pattern of inheritance was autosomal dominant. Penetrance is high, although not 100%,4
with at least one asymptomatic carrier identified.
The presence of prominent, early anterograde amnesia in this family with a mutation in PGRN and pathological frontotemporal lobar degeneration demonstrates the limitations of clinical-pathological correlations and diagnostic criteria in neurodegenerative disease. Anterograde amnesia is the prototypical manifestation of the Alzheimer's disease, and the preservation of new learning is one of the diagnostic criteria for the FTLD syndromes. At least in the case of this unusual family, purely clinical diagnostic criteria would fail to properly identify the pathological substrate of their illness. In contrast, most of the MRI scans in our patients showed a variety of features that were distinctly of a non-Alzheimer type. Clinical diagnostic criteria could be enhanced by adding explicit inclusion and exclusion imaging features for both AD and the FTLD syndromes.