Alpha-synucleinopathy and neuropsychological symptoms in a population-based cohort of the elderly

Medical Research Council Cognitive Function and Ageing Study

The Medical Research Council Cognitive Function and Ageing Study (MRC CFAS) has been
fully described elsewhere 21]. It is a longitudinal, prospective, population-based cohort study initiated in 1989.
At each of five centres in England and Wales random samples of about 2,500 people,
64-years old and older (with an 82% response rate), agreed to a structured initial
interview. Structured interviews collected details of sociodemographic status, general
health, Mini Mental State Examination (MMSE) 22] and functional ability. Approximately 20% of the sample completed a more detailed
diagnostic assessment, including full mood and organicity sections of the Geriatric
Mental State (GMS) Automated Geriatric Examination for Computer Assisted Taxonomy
(AGECAT) 23] and the Cambridge Cognitive Examination (CAMCOG) 24]. Further waves of interviewing (both screening and assessment) were conducted at
two, five and nine years after baseline on the survivors. In a sixth centre 5,200
people, 65-years old and older, were interviewed using the same assessment interviews
with follow-up at one to two, three to four, five to six, and nine years later. The
CFAS Neuropathology Study is based on a brain donor programme established in 1992.
Individuals selected for the assessment interviews were approached by a trained liaison
officer for brain donation. The donation programme was discussed with the respondent
and their family or carer as appropriate. All participants in the study gave their
informed consent. This analysis is based on the burden and anatomical distribution
of AS in all donations from two MRC CFAS centres before July 2003 (Nottingham and
Cambridgeshire, n?=?208). Research was carried out after approval from the Eastern
Multi Centre Research Ethics Committee based at Papworth Hospitals NHS Trust (reference
03/5/017).

Postmortem procedure and anatomical sampling

Pathological evaluation of the Neuropathology Study cohort had previously been carried
out using the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) protocol
1] and each brain was assigned a Braak score for neurofibrillary tangles 2]. At autopsy samples of brain tissue were removed for frozen storage. The remainder
of the brain was fixed for standardised assessment using paraffin-embedded tissues.
For the evaluation of synucleinopathy a hierarchical sampling strategy was adopted
based on evaluation of the midbrain (substantia nigra), medulla (dorsal efferent nucleus
of the vagus nerve, medullary reticular formation) and amygdala to detect AS pathology.
This method of screening was selected because literature at the time of the study
indicated that no cases had been reported in which AS was present in any brain region
but absent from these three screening areas 3],25]-29]. If at least one positive ?-synuclein immunoreactive profile was found in a screening
area five additional areas were investigated (cingulate gyrus, entorhinal cortex and
hippocampus, frontal cortex, parietal cortex and temporal cortex), comprising the
DLB Consensus recommendations for the evaluation of AS 30].

Immunocytochemistry for alpha-synuclein, quantification

Microwave antigen retrieval (citric acid 10 mM pH6) of 6 ?m brain sections was followed
by immunocytochemistry using a monoclonal antibody to mouse alpha-synuclein (LB509,
Zymed Laboratories Inc; San Francisco, CA USA; dilution 1/200; incubated at 4°C overnight).
Incubation with biotinylated secondary antibody (1/250) for 1.5 hours at room temperature
was followed by ABC Elite Mouse IgG reagents (Vector Laboratories, Burlingame, CA,
USA) for one hour. Reaction product was visualised using Vector Nova Red (Vector Laboratories)
according to the manufacturer’s protocol. All sections were counterstained with haematoxylin
and mounted in DPX. Interpretation of the appearances was jointly established with
propidium iodide (PI). Acceptable inter- and intra-rater reliability was previously
reported in this study 31]. AS pathology was then rated by a single investigator (JZ). The microscopic field
with the highest density of AS was identified and the number of AS profiles was counted
at x200 magnification. Spherical and densely stained structures were considered to
be LB irrespective of intra- or extracellular localisation. Elongated AS-positive
profiles were considered to be Lewy neurites (LN). Two consecutive sections from each
block were assessed and the highest density score retained for the statistical analysis.
In the final analysis, results were dichotomised to AS present versus absent.

Each brain was allocated to a pathological category as shown in Table 1. These categories were based on the presence or absence of neurofibrillary tangles
(NFT) and AS. NFT pathology was dichotomised on the basis of Braak stage (Low – stages
I-III versus High – stages IV-VI). AS was divided into three categories based on previous
work in this cohort:

1. Cases that conform to the hierarchy of anatomical spread of AS implicit in the
DLB Consensus protocol for the evaluation of AS 30], which also implies that these cases conform to the Braak staging hypothesis proposed
for AS 3]. Some analyses are based on the stratification of this group into those with DLB
‘limbic’ and ‘neocortical’ disease and those with AS confined to midbrain/brainstem
structures;

2. Cases that conform to ‘amygdala-predominant’ AS implying that amygdala AS is prominent
and that involvement of neocortical, brainstem and midbrain structures is limited
and patchy 31];

3. Cases with an atypical AS distribution, predominantly those with mild to severe
cortical involvement in the absence of AS in one of the lower hierarchic anatomical
sites (for example, substantia nigra, locus ceruleus, dorsal vagus nucleus). For example,
this category includes cases with prominent AS involvement observed in the substantia
nigra and frontal cortex but not in other investigated areas, or cases with prominent
AS involvement observed in the vagus nucleus and parietal cortex only. These also
include six cases corresponding to the ‘cerebral-form of DLB’ 32] where AS was observed in the neocortex predominantly.

Table 1. Pathology subgroups based on ?-synucleinopathy (AS) and severity of neurofibrillary
tangle (NFT) pathology

Previous findings in this cohort showed that there was no association between these
different AS staining patterns and Braak stage for NFT 31]. Thirty-two brains were excluded from the final analysis due to missing pathological
or clinical data.

Dementia and other outcomes

Dementia was defined using a combination of the AGECAT algorithm, death certificates
and informant interviews, including retrospective informant interviews after death
(RINI). Ten percent of participants were classified as having unknown dementia status
on the basis that both the time between the last interview and death was too long
to exclude incident dementia (over six months) and that data collected during informant
interviews were inconclusive. The presence or absence of hallucinations (visual and
auditory) or PD was based on self-reported symptoms and informant interview data.

Specific neuropsychological measures

General cognitive function was assessed using the MMSE. A score of 21 or less was
used to indicate the likely presence of dementia. The CAMCOG neuropsychological battery
includes seven areas of cognition: orientation, language (comprehension and expression),
memory (learning, recent and remote), attention/calculation, praxis, abstract thinking
and perception. Cut-off scores for each domain subscales were taken as values falling
below the 25th percentile cut-point in the whole of the non-demented MRC CFAS population.
Results are presented for a total CAMCOG score (all seven areas) and for CAMCOG areas
of attention/calculation, language (comprehension and expression), abstract thinking
and perception. CAMCOG cut-off points were 83 out of a maximum score of 103, 5 out
of 8 (attention/calculation), 25 out of 30 (language), 5 out of 8 (abstract thinking)
and 6 out of 8 (perception). Informant and interview data were analysed for the presence
or absence of visual hallucination, auditory hallucination and a self-reported diagnosis
of Parkinson’s disease.

Statistical analysis

MRC CFAS has released data at defined points. For this study, version 8.2 of the core
MRC CFAS dataset and version 3.3 of the neuropathology dataset were used. All statistical
work was conducted in Stata Software version 9.0 (Stata Corporation, College Station,
TX, USA). Odds ratios and Fischer’s exact (chi squared) tests were applied to estimate
the strength of any association between two observations. High P-values are associated with a higher likelihood that there is no significant difference
between compared groups. All P-values are two sided.