Background The human prion diseases are a group of universally fatal neurodegenerative disorders associated with the auto-catalytic misfolding of the normal cell surface prion protein (PrP). Mutations causative of inherited human prion disease (IPD) include an insertion of six additional octapeptide repeats (6-OPRI) and a missense mutation (P102L) with large families segregating for each mutation residing in southern England. Here we report for the first time the neuropsychological and clinical assessments in these two groups.
Method The cognitive profiles addressing all major domains were obtained for 26 patients (18 6-OPRI, 8 P102L) and the cortical thickness determined using 1.5T MRI in a subset of 10 (six 6-OPRI, four P102L).
Results The cognitive profiles were different in patients with the two mutations in the symptomatic phase of the disease. The 6-OPRI group had lower premorbid optimal levels of functioning (assessed on the NART) than the P102L group. In the symptomatic phase of the disease the 6-OPRI patients had significantly more executive dysfunction than the P102L group and were more impaired on tests of perception and nominal functions. There was anecdotal evidence of low premorbid social performance in the 6-OPRI but not P102L patients. Cortical thinning distribution correlated with the neuropsychological profile in the 6-OPRI group principally involving the parietal, occipital and posterior frontal regions. The small number of patients in the P102L group precluded statistical comparison between the groups.
Conclusions The 6-OPRI patients had more widespread and severe cognitive dysfunction than the P102L group and this correlated with cortical thinning distribution.
Statistics from Altmetric.com
The prion diseases or transmissible spongiform encephalopathies are a group of fatal neurodegenerative disorders affecting humans and animals.1 The central feature of prion disease is the conversion of the host encoded prion protein (PrPC) to a misfolded and aggregated state termed PrPSc.2 Human prion diseases occur in sporadic, acquired and inherited forms. All types of prion disease are characterised by progressive neurodegeneration with cognitive impairment and other progressive neurological deficits. The neurodegeneration is rapid in the case of sporadic and acquired prion disease with death typically occurring in less than 2 years from clinical onset. However, in inherited prion disease, neurological deterioration may be much slower.3
Inherited human prion disease (IPD) occurs as a result of one of more than 30 mutations in the prion protein gene (PRNP). It is highly heterogeneous in clinical onset, manifestation and duration. Some of this heterogeneity is conferred by the PRNP mutation, but variability is also seen among affected individuals within families carrying the same mutation,4–6 a phenomenon that is incompletely understood.7–9 Pathological mutations in the PRNP gene occur as point mutations, stop codon mutations and alteration of the number of octapeptide repeats in the N terminal domain of PRNP.3 7 In the UK, large kindreds with the 6-OPRI mutation, in which there are six additional repeats in the octapeptide region, and the P102L mutation, in which there is a missense mutation at codon 102 of the prion gene, have been followed-up for over two decades.10–13 The latter mutation has been shown to be responsible for the original cases described by Gerstmann, Sträussler and Scheinker but often the eponym Gerstmann–Sträussler–Scheinker disease has been extended to other mutations associated with ataxia and dementia.
Most of the literature on prion disease has primarily focused on the clinical, neurological and psychiatric features of the disease rather than on the neuropsychological features. Despite cognitive impairment being the dominant feature of most forms of prion disease, detailed neuropsychological studies have principally focused on the acquired forms of the disease.14 In variant Creutzfeldt–Jakob disease, impairments were noted in memory, executive functions (verbal fluency), speed and attention, and visual perception (face perception).14 Some verbal skills such as vocabulary, verbal reasoning and digit span were relatively preserved in most patients. This led to the conclusion that variant Creutzfeldt–Jakob disease showed features of both cortical and subcortical dementia. There is less known about the detailed neuropsychology of IPD although small studies suggest differences in profile between sporadic, acquired and inherited disease.15 16 It is noteworthy that the IPD data from these studies were analysed collectively, which might obscure differences between mutation types. It has been anecdotally reported that nine out of 11 patients with P102L mutation showed intellectual decline. Unfortunately, no formal assessment of these patients' cognitive functions has been reported to date.17
Equally, although the impaired cognition in prion diseases is associated with brain volume loss,18 19 detailed studies of neuroanatomical distribution of these changes and correlation with neuropsychology have not previously been performed. Measurement of cortical thickness is a relatively new technique for assessing the regional distribution and quantification of gray matter cortical loss which has been used in various neurodegenerative diseases to provide complementary information in the pathophysiology of these diseases.20–22 The sensitivity and specificity of cortical thickness techniques allow correlation with behavioural data, which is difficult to assess using more conventional imaging modalities.
The large UK 6-OPRI family was first reported in detail by Collinge in 1992,4 and updated in 2006.8 Of considerable interest was the possible presence in a number of cases of a premorbid psychological/personality disorder, consisting of aggression, irritability with short temper, antisocial behaviour and hypersexuality. Involvement of the police and charge with offences was not uncommon. These personality difficulties were longstanding and predated neurological onset in their affected relatives, leading to the suggestion that this PRNP mutation might influence neurodevelopment as well as lead to neurodegeneration in later life4 but no formal documentation of the cognitive and behavioural profiles of these patients has been reported so far.8 Here we report, for the first time, the detailed neuropsychological assessment and cortical thickness analysis comparison of 6-OPRI and P102L patients in the UK; we tested the hypothesis that premorbid cognitive abnormalities are specific to the insertional mutation.
Patient groups and ethics
Patients were referred to the National Prion Clinic at the National Hospital for Neurology and Neurosurgery, London, UK. The National Prion Clinic is a national service providing clinical assessment and genetic testing for patients considered to be at risk of any form of prion disease throughout the UK. All patients in the study had a diagnosis of 6-OPRI or P102L IPD. Details of the clinical features of these groups of patients have previously been published.8 17 They were assessed clinically and neurologically and they also underwent a neuropsychological assessment in the Neuropsychology Department at the National Hospital for Neurology and Neurosurgery. The neuropsychology data were analysed retrospectively. Six patients with the 6-OPRI mutation (three men, mean age 38.0 years) and four patients with the P102L mutation (three men, mean age 59.5 years) who had undergone a volumetric 1.5 T T1 weighted and MR brain scan as part of the MRC Prion-1 Trial23 were included in the cortical thickness analysis. Ten control subjects (mean age 52 years) were scanned using the same protocol. Ethics approval for the study was granted by the Eastern Multicentre Research Ethics Committee (MREC), and informed consent for participation in the study was given by either the patient or patient's next of kin. All MR scans were performed within 6 months of the neuropsychology and clinical assessment. A control group of 10 cognitively normal subjects (five men, mean age 52.3 years) with a volumetric T1 weighted 1.5 T MRI were included in this study.
The neuropsychological assessment was conducted at the time of the clinical and neurological evaluations. The neuropsychological battery included tests evaluating general intellectual functioning,24 memory,25–27 naming,28 literacy,29 calculation,30 perception31 and ‘executive’ functions.32–35 The patients' performance on the neuropsychological tests was classed impaired according to a variety of scoring methods (see appendix, available online only).
MRI acquisition and analysis
All subjects were examined using a GE Signa LX 1.5T MRI system (GE Healthcare, Milwaukee, Wisconsin. USA). Three-dimensional T1 weighted image data were acquired from 124 contiguous 1.5 mm thick coronal slices (inversion recovery prepared spoiled gradient echo sequence; TE 5 ms, TR 35 ms, flip angle 35°, matrix 256×128, FOV 24×24 cm). Cortical thickness reconstruction and thickness estimation was performed with the FreeSurfer image analysis suite (Fischl and Dale 11050–55), V.4.4.0 (http://www.surfer.nmr.mgh.harvard.edu/) on a 64 bit Linux centos four Cluster managed by a Sun Grid Engine. Briefly, the process involves initially generating an automatic gray matter, white matter and CSF tissue classification. The results of these segmentations were visually inspected and, if needed, manually edited by adding control points. Where manual editing was performed, the operator was blinded to the subject's diagnosis. Finally, an automatic reconstruction of the cortex was produced and cortical thickness estimated by computing the average shortest distance between the white matter boundary and the pial surface. Surface maps were generated following registration of all subjects' cortical reconstructions to a common average surface and then smoothed using a surface based Gaussian kernel of 20 mm full width half maximum. The standard FreeSurfer processing stream was used apart from two modifications. Firstly, we used locally generated brain masks for the skull stripping process. This brain mask was produced using a semiautomated segmentation procedure that involved selection of thresholds, followed by a series of erosions and dilatations, yielding a brain region separated from surrounding CSF skull and dura.36 Secondly, we modified the white matter mask by incorporating the ventricle segmentations from the FreeSurfer volume processing stream because of mislabelling of CSF in the standard white matter mask, particularly in cases where the ventricles are large. We also used the FreeSurfer subcortical volume processing stream to generate cerebellar volumes.37
Statistical analysis was performed using the statistical package for the social sciences V.11.5 (SPSS). To establish whether any differences between the two groups were statistically significant, Fisher's exact test was used for the categorical variables and the Mann–Whitney U test was used for the continuous variables. To establish any difference within a group, the Wilcoxon signed ranks test was used for continuous variables. Non-parametric tests were chosen to be conservative when we could not be certain of normality given the small sample size.
A vertex by vertex analysis using a general linear model was performed to examine differences in cortical thickness between the patient groups and the control group. Cortical thickness, C, was modelled as a function of group, controlling for age and sex, by including them as nuisance covariates. C=β1 6-OPRI+β2 P102L+β3 controls+β4 age+β5 sex+μ+ε (where μ is the constant and ε is the error) with the contrasts of interest being the two tailed t tests between the estimates of the group parameters. Maps showing the significant differences between the disease and controls groups were generated, correcting for multiple comparisons using the false discovery rate (FDR), thresholded at a 0.01 significance level. As well as the surface maps, the FreeSurfer processing stream also generates thickness measures from cortical regions of interest, as described in Deskan and colleagues.38 From these measures, mean cortical thickness estimates for frontal, parietal, temporal, occipital, cingulate and insular cortex were generated for each hemisphere. As no significant differences in right versus left cortical thickness asymmetry in each region were identified using the paired t test, overall mean cortical thickness was calculated for each anatomical area and differences across groups were assessed using a one way ANOVA.
Finally, cerebellar volume corrected for total intracranial volume (C/TIV) was calculated for each patient and differences in group means were assessed using a one way ANOVA.
Demographic and clinical evaluation
The clinical presentations of 6-OPRI and P102L patients are generally different.4 8 13 39 Clinical details of the patients studied here are shown in tables 1 and 2 and confirm that the patients were representative of their mutation type and had a similar duration of illness, providing context for the neuropsychological profiles.
In spite of the small number of patients studied, significant differences in the psychological profile were seen between the two groups of patients. Although the age of the patients with the 6-OPRI mutation was nearly 20 years less than those with the P102L mutation, the duration of clinical illness was similar (table 1).
The P102L patients had a premorbid optimal level of functioning, as measured by the National Adult Reading Test (NART), in the average range, while the 6-OPRI patients had a premorbid optimal level of functioning in the low average range. This difference was significant (p<0.01, Mann–Whitney U test; mean difference=22.6; 6-OPRI mean=84.2, SD=11.7; P102L mean=106.9, SD=9.5). The significantly lower premorbid optimal level of functioning of the 6-OPRI patients remained, even when we excluded from the analysis those patients showing signs of acquired dyslexia (n=2). Indeed, only the NART scores of the 6-OPRI patients with comparable expressive (as measured by the Wechsler Adult Intelligence Scale-revised vocabulary subtest) and reading vocabulary were analysed further. Thus the difference in premorbid ability cannot be attributed to the fact the 6-OPRI patients were dyslexic. Furthermore, the reported difference in premorbid optimal level of functioning between the two patient groups cannot be attributed to differences in education. In fact, the significantly lower NART scores of the 6-OPRI patients remained, even when we analysed the reading performance of only those patients with 10 or 11 years in formal education in both groups (n=17, Mann–Whitney U test, p=0.003).
Both the 6-OPRI and P102L patients had impairment in verbal IQ, performance IQ and full scale IQ when compared with optimum level of functioning calculated from the NART. Verbal IQ was significantly higher than performance IQ within each group (p<0.05, Wilcoxon signed ranks test). Patients with the 6-OPRI mutation were significantly more impaired than the P102L patient group in verbal IQ, performance IQ and full scale IQ (p<0.05, Mann–Whitney U) (figure 1).
The percentage of patients with impairment in memory, nominal, visual perception, visuospatial and frontal executive functions are shown in figure 2. The 6-OPRI group showed greater global cognitive impairment. Patients were impaired in virtually all areas of cognition, with frontal ‘executive’ functions most affected. Visuoperceptual, visuospatial and nominal functions were also impaired in more than half of the patient sample. In contrast, the P102L group showed cognitive impairment affecting predominantly memory and frontal ‘executive’ functions. Notably the frontal ‘executive’ impairment in the P102L group was much less common than in the 6-OPRI group. Interestingly, visuospatial and nominal impairments rarely occurred and visual perception impairments were absent. Statistical analysis confirmed these observations. Significantly more 6-OPRI patients demonstrated impairment in executive functions (p<0.01), visual perception (p<0.01), visuospatial functions (p<0.01) and nominal skills (p<0.05) when compared with the P102L patients. There were no significant differences between the two groups in terms of memory impairment.
The considerable difference in terms of current and premorbid IQ functioning in the two groups may have masked potential differences in patient performance on focal cognitive tasks. To address the imperfect IQ match, we selected three 6-OPRI patients and three P102L patients who were matched as far as possible in terms of current IQ functioning and we analysed their cognitive performance. Two of the three 6-OPRI patients presented with executive dysfunction and one patient presented with selective visual memory impairment. By contrast, none of the P102L patients presented with executive dysfunction. However, memory functions were impaired in all three patients.
There was evidence of premorbid poor social performance in the 6-OPRI patients compared with those with P102L. These data are summarised in the supplementary appendix (available online only).
Cortical thickness and cerebellar volume analysis
We compared cortical thickness between the two patient groups and a healthy control group. In the 6-OPRI patients there was symmetrical thinning of the cortex in the parietal, occipital and posterior frontal regions (figure 3, table 3). The areas of greatest thinning in the parietal lobe were the precuneus, demonstrating a reduced cortical thickness of 32%, inferior parietal lobule 24% and supramarginal gyrus 24%. The areas of greatest reduction in the occipital lobe were the lingual (22%) and pericalcarine gyri (20%), and in the frontal lobe, the paracentral gyrus (25%) was most affected. The temporal lobe, apart from the posterior superior and middle temporal gyri, the cingulate and insular cortex were relatively unaffected.
The P102L group was too small to reliably compare with the 6-OPRI group but there was a trend towards cortical thinning in the parietal and occipital lobes (table 3), particularly involving the inferior parietal lobe (11%) and in the occipital lobe, the cuneus (13%) and pericalcarine cortex (15%) (table 3).
The small numbers of P102L patients precludes reliable conclusions on cerebellar volume in this group but the volume was reduced compared with the 6-OPRI and control groups combined (C/TIV in P102L patients=8.0%±1.4; combined 6-OPRI and control group of 8.9%±1.0%; p=0.095).
We have documented for the first time the cognitive profiles associated with the two different mutations of PRNP. The cognitive profile of the 6-OPRI group was characterised by marked intellectual decline and widespread cognitive impairment. In the P102L group, the cognitive profile showed predominantly, although not exclusively, intellectual decline less marked than in the 6-OPRI and memory impairment. Interestingly, frontal ‘executive’, visuospatial and nominal skills were significantly more impaired in the 6-OPRI than in the P102L group. Visuoperceptual impairment was present in almost 60% of the 6-OPRI patients but almost absent in the P102L group. When we analysed a small cohort of 6-OPRI and P102L patients matched for IQ, we continued to observe differences in their cognitive profiles. The 6-OPRI patients showed predominantly executive dysfunction while the P102L patients showed predominantly memory impairment.
Brain atrophy has been well described in all forms of prion disease.18 40–42 However, the distribution of the atrophy has rarely been described apart from one study in presymptomatic P102L gene carriers demonstrating early parietal atrophy and marked progressive cerebellar loss on MRI.19 For the first time, we used cortical thickness analysis. Given the small numbers of patients studied, our analysis must in general be considered preliminary. However, this does indicate a significant cortical thinning in the parietal and occipital lobes in patients with the 6-OPRI mutation, specifically in the precuneus, inferior parietal lobule, supramarginal gyrus and lingula. This distribution of cortical damage relates well to the clinical symptoms and some of the cognitive impairments documented in patients with the 6-OPRI mutation. Apraxia is an important early feature and is generally associated with lesions to the dominant parietal lobe and specifically the supramarginal gyrus.43 Visuoperceptual and visuospatial impairments, known to be sensitive to right parietal damage, were also quite common in this patient population. Similarly, executive functions and naming were markedly affected in the 6-OPRI patients and this correlates well with the frontal lobe and posterior temporal lobe cortical thinning. Conversely, apraxia and perceptual impairments were virtually absent in the P102L patients who only showed a trend towards thinning of the parietal lobes and where thinning was present, the inferior parietal lobe was most affected.
P102L patients presented with ataxia and showed a tendency to smaller cerebellar volumes compared with the control patients and the 6-OPRI group. The small number of patients and consequent increased variability of our P102L group may have contributed to the lack of significant cortical thinning and cerebellar volumes compared with our other patient groups. Our P102L patients were also less severely affected and older than our 6-OPRI patients, although they had a similar duration of illness. Future longitudinal hypothesis driven studies correlating behavioural, pathological and neuroimaging data are needed for further evaluation.
It has previously been noted that 6-OPRI patients tend to show behavioural difficulties, attain a low level of educational standard and have employment difficulties.4 8 The present study formally documents a striking abnormality in the premorbid level of functioning for the 6-OPRI group and shows this to be significantly lower than the P102L group. The patient with the highest level of intellectual functioning scored just within the average range in the 6-OPRI group, compared with the highest functioning P102L patient who scored in the high average range. It is also interesting that the study showed a significant difference in years of education between the two groups. Furthermore, the qualitative occupational records suggested that the 6-OPRI patients had difficulties in obtaining and maintaining employment. Importantly, there were still differences in premorbid function when educational factors were controlled. The 6-OPRI kindred is extremely large (over 100 affected individuals over eight generations) such that most pairs of patients tested in this study are separated by over 10 meioses and thus share little in the way of environmental factors or genetic material, other than the mutation at PRNP.
There are at least two possible hypotheses for the notably lower social/intellectual functioning recorded for the 6-OPRI patients. A first hypothesis suggests that the 6-OPRI mutation itself results in neurodevelopmental abnormalities predating the neurodegeneration of the later illness.4 An alternative hypothesis is the social drift theory. This suggests that as patients become ill at an early age, their ability to parent and support offspring becomes interrupted, leading to behaviour disturbances, poor educational attainment and employment difficulties. However, we found significant differences in premorbid intellectual function even after we matched our patients for level of education. Further research exploring this issue would be useful.
The present study provides the first documentation of the clinical, cognitive and neuroanatomical profiles associated with 6-OPRI and P102L pathology. The clinical utility of cortical thickness techniques have not been widely evaluated but can potentially provide important complementary information to the clinical and cognitive characteristics of genetic mutations and may assist in early recognition and diagnosis.
We are grateful to all the patients and their families for generously contributing to this report and to the neurology colleagues who referred patients to the National Prion Clinic. Ray Young assisted with illustrations.
Funding This work was funded by the MRC and the Department of Health England. It was undertaken at UCLH/UCL who received a proportion of funding from the Department of Health's NIHR Biomedical Research Centres funding scheme. MC was funded by the TSB grant M1638A.
Correction notice This article has been amended since it was published Online First. The address of the first author's affiliation has been corrected and in figure 3 the word 6-OPRI was omitted. Also a paragraph from the fourth page has been moved into the legend of figure 3.
Competing interests None.
Ethics approval The study was approved by the Eastern Multicentre Research Ethics Committee (MREC).
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.