Guide to the Consortium to Establish a Registry for Alzheimer’s
Disease (CERAD) Form1
General Guide to the Process
The brain bank was notified as soon as possible after death and arrangements were made
for immediate dissection and retrieval of brain tissue. Once recovered, the brain was
sliced into roughly 1 cm thick slices. Slices from one half of the brain were snap frozen to
-80oC. Slices from the other half were formalin fixed for 4-6 weeks and selected areas
were processed into paraffin embedded blocks for analysis (see Table 1 for protocol).
Fixed tissue that was not blocked still exists for most cases.
Slides were produced and assessed for pathology according to the CERAD1 protocol.
Since the project began in 1985, the protocol has changed as new procedures became
available. Early cases used protocols involving specialised silver stains and congo red.
Protocols for later cases involve immunohistochemistry.
For some cases, photos were taken of gross post mortem appearance from various angles,
with a reference scale, and also of various brain slices.
Table 1 Blocks and stains taken for neuropathological assessment.
Block
H+E
Tau
Ubiquitin
Abeta
ERCF
HCF1
HCF2
FCF
TCF
PCF
OCLF
OCMF
Cing
Mid brain1
Mid brain 2
Pons
Medulla
Cerebellum
BGF 1
BGF 2
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
List of abbreviations at end of document.
A/syn
*
*
*
*
*
*
*
*
Details of variables in the CC75C CERAD neuropathological database
The following sections relate variables from the STATA database, in which data are
archived, to entries on the CERAD forms completed for each CC75C brain donor.
Completing the CERAD Form
Both the macroscopic and microscopic reports, (macroreport and microreport
respectively) are required to complete this form. It should be completed by the
neuropathologist immediately following the survey. The macroreport is used for sections
C and D of the CERAD form. The microreport is used to complete sections E, G and H of
the CERAD from.
Section C Gross examination
brw : Total brain weight
This is measured from the whole brain as fresh tissue, but can also be measured from
whole brain fixed tissue or calculated from the fixed half brain. The variable fixat
describes whether the tissue was fresh or fixed; the variables sidehemi, sidebrs and
sidecbl give details about which side of the cerebral hemispheres, brainstem or
cerebellum was fixed.
mening; records the appearance of the meninges. Any abnormality, e.g. haemorrhage,
neoplasm, inflammation, fibrosis or thickening, is recorded in a written note on the form.
atrophy: records the visible, external appearance of atrophy. This was measured by eye;
the exact size of cortical sulci was not measured. Associated variables include ventr, a
measure of degree of atrophy relative to ventricles, atrfr, atrophy of frontal cortex,
atrpa, atrophy of parietal cortex, atrte, atrophy of temporal cortex, atrhi, atrophy of
hippocampus, atroc, atrophy of occipital cortex and atrcbl, atrophy of cerebellum. Any
unusual sulcal or ventricular features were recorded in the variable unus with a written
note.
palsn: records gross appearance of the substantia nigra, a nucleus of dopamine producing
cells located in the midbrain; pallor is associated with loss of neurons or loss of neuronal
function.
pallc: records gross appearance of the locus coeruleus, a nucleus of noradrenalinproducing cells located in the dorsal pons; pallor is associated with loss of neurons or loss
of neuronal function.
The presence of other gross CNS lesions is recorded in the variable onvg, with further
description in a written note.
Section D. Cerebral Vascular Disease: Gross Findings
gvas: records gross appearance and degree of atherosclerosis of large vessels. The related
variable gvobstr records the presence of vessels from the circle of Willis and major
branches with an obstruction greater than 50% of the internal lumen. A written note
describes the degree and location of any obstructed vessels.
gveovl: records the presence of other gross vascular lesions including aneurysms and
arterio-venous malformations.
gvpvl: records the presence of any gross parenchymal lesions and begins a large interrelated section of the form. Type of vascular lesion is recorded in the related variables
gvinf for infarcts, gvlac for visible lacunes and gvhaem for haemorrhages.
gvinfd: records the presence of infarcts greater than 10mm in diameter with the related
variables gvnol recording the number of lesions and gvz1 and gvz2 recording the size in
mm of the largest infarct. The related variable i1, i2, i3, i4, i5, i6 and i7 describe the
anatomical and arterial distribution related to the infarcts recorded, i.e. which large
arteries supply the infarcted areas.
gvlacd: records the presence of lacunes, also labelled as cystic infarcts less than 10mm in
diameter. The related variables l1, l2, l3, and l4 describe the location of the lacunes; l1
corresponds to basal ganglia or thalamus, l2 corresponds to cerebral white matter, l3
corresponds to brainstem and l4 corresponds to any other location with detail given in a
written note.
gvoi: records the presence other infarcts less than 10 mm. The related variables oi1, oi2,
oi3, oi4, oi5 describe the location.
gvhaemp: records the presence of gross haemorrhages in the brain, with related variables
h1 recording the number of small (less than 5mm diameter) haemorrhages, h2 recording
the number of medium (6-10mm diameter) and h3 recording the number of large (greater
than 10mm diameter) haemorrhages. The related variables h4 to h18 form a 3 x 5 table to
record the number and location of haemorrhages by size. The top row of the table
corresponds to h4, h5 and h6 with h18 in the bottom right hand corner
Section E. Microscopic Vascular Findings
mvas: records the presence or absence of severe microvascular atherosclerosis
mvals: records the presence or absence of severe microvascular arteriolosclerosis
mvvr: records the presence or absence of severe Virchow-Robin space expansion. The
related variables describe VR space location, mvvrc refers to cortical VR spaces, mvvrw
to spaces in white matter and mvvrd as spaces in the deep grey matter of the basal
ganglia or thalamus.
mvpg: records the presence or absence of perivascular gliosis. The related variable,
mvpgl describes the location of gliosis as in white matter, grey matter or both.
mvomd: records other microvascular disease e.g. vascular malformation and vasculitis,
described with a written note
mvvl: records the presence or absence of microvascular lesions and begins a section of
the form detailing 1) microinfarcts and 2) white matter pallor.
1) mvmdm: records the presence or absence of microinfarcts. The related variables,
mvhi1, mvhi2, mvte1, mvte2, mvfr1, mvfr2, mvpa1, mvpa2, mvoc1, mvoc2, mvdg1,
mvdg2, record presence or absence in particular brain areas and whether the
microinfarcts are in white matter, grey matter or both.
2) wmp: records the presence or absence of areas of white matter pallor. The related
variables wmpoc (occipital), wmppa, (parietal) wmpfr, (frontal) wmpte, (temporal) and
wmpdw (deep white), describe the location of white matter pallor seen.
Axonal loss, cell loss and demyelination appear as areas of pallor in H/E slides under the
microscope as there is less tissue to absorb any stain, hence the name; white matter pallor
may be a marker for loss of conductivity and connectivity between different brain areas,
affects are dependent on extent of pallor and brain area affected.
Section G microscopic evaluation of hippocampus, neocortex and other selected
regions
Evaluation of the hippocampus and entorhinal cortex (subcortical)
The hippocampus and entorhinal cortex were assessed from anterior and posterior
regions. The highest grade of pathology was recorded on the CERAD form, regardless of
location or specific anatomical region. The Cornus Ammonis 1, CA1, region of the
hippocampus and the trans- region from the entorhinal cortex tend to dominate as they
generally have a greater burden of pathology.
nphi and nperc record the density of neuritic plaques in the hippocampus and entorhinal
cortex respectively. The pathology is graded as none = 0, sparse (one or two plaques per
section) =1, moderate (several plaques per section) = 3 and severe (many plaques per
section) = 5. Plaque density is referenced to images in CERAD1 Handbook.
adhi and aderc record the density of amyloid beta protein deposits in the hippocampus
and entorhinal cortex respectively. The pathology is graded as none = 0, sparse (one or
two deposits per section) =1, moderate (several deposits per section) = 3 and severe
(many deposits per section) = 5. Plaque density is referenced to images in CERAD1
Handbook.
nfthi and nfterc record the density of tau reactive neurofibrillary tangles in the
hippocampus and entorhinal cortex respectively. The pathology is graded as none = 0,
sparse (one or two affected neurons per section) =1, moderate (several affected neurons
per section) = 3 and severe (many affected neurons per section) = 5. Plaque density is
referenced to images in CERAD1 Handbook.
vaphi records the severity of vascular amyloid deposits in the brain parenchyma of the
hippocampus. The pathology is graded as none = 0, sparse (one or two affected vessels
per section) =1, moderate (several vessels per section) = 3 and severe (many affected
vessels per section) = 5.
vamhi records the severity of vascular amyloid deposits in the meninges of the
hippocampus. The pathology is graded as none = 0, sparse (one or two affected vessels
per section) =1, moderate (several vessels per section) = 3 and severe (many affected
vessels per section) = 5.
vahhi records the severity of an haemorrhages associated with vessels affected by
amyloid deposition. The pathology is graded as none = 0, sparse (one or two affected
vessels per section) =1, moderate (several vessels per section) = 3 and severe (many
affected vessels per section) = 5.
gvdhi and gvderc record the presence or absence of granulovacuolar degeneration in the
hippocampus and entorhinal cortex respectively.
hbhi and hberc record the severity of Hirano bodies in the hippocampus and entorhinal
cortex respectively. The pathology is graded as none = 0, sparse (one or two affected
neurons per section) =1, moderate (several affected neurons per section) = 3 and severe
(many affected neurons per section) = 5.
snlhi and snlerc record the presence or absence of severe neuronal loss in the
hippocampus and entorhinal cortex respectively.
sghi and sgerc record the presence or absence of severe gliosis in the hippocampus and
entorhinal cortex respectively.
othi and oterc record the presence or absence of any other pathology in the hippocampus
and entorhinal cortex respectively. There should be a written note to give details of this.
pbhi and pberc record the severity of Pick bodies in the hippocampus and entorhinal
cortex respectively. The pathology is graded as none = 0, sparse (one or two affected
neurons per section) =1, moderate (several affected neurons per section) = 3 and severe
(many affected neurons per section) = 5.
lbhi and lberc record the severity of alpha synuclein Lewy bodies in the hippocampus
and entorhinal cortex respectively. The pathology is graded as none = 0, sparse (one or
two affected neurons per section) =1, moderate (several affected neurons per section) = 3
and severe (many affected neurons per section) = 5.
updhi, upderc, ubiinchip and ubiincerc record the presence of ubiquitin positive dots
and inclusions in the hippocampus and entorhinal cortex. These were recorded in early
cases but not recorded in later cases and usually are filled with 9 for unknown in the
STATA database.
Evaluation of the neocortex (frontal, temporal, parietal and occipital cortices)
npfr, npte, nppa, and npoc record the severity of tau reactive neuritic plaques in the
frontal, temporal, parietal and occipital cortices respectively. The pathology is graded as
none = 0, sparse (one or two plaques per section) =1, moderate (several plaques per
section) = 3 and severe (many plaques per section) = 5. Plaque density is referenced to
images in CERAD1 Handbook.
adfr, adte, adpa, and adoc record the severity of amyloid beta protein- reactive plaque
deposits in the frontal, temporal, parietal and occipital cortices respectively. The
pathology is graded as none = 0, sparse (one or deposits per section) =1, moderate
(several deposits per section) = 3 and severe (many deposits per section) = 5. Plaque
density is referenced to images in CERAD1 Handbook.
nftfr, nftte, nftpa, and nftoc record the severity of neurofibrillary tangles in the frontal,
temporal, parietal and occipital cortices respectively. The pathology is graded as none =
0, sparse (one or two affected neurons per section) =1, moderate (several affected
neurons per section) = 3 and severe (many affected neurons per section) = 5. Plaque
density is referenced to images in CERAD1 Handbook.
vapfr, vapte, vappa, and vapoc record the severity of vascular amyloid deposits in the
brain parenchyma of frontal, temporal, parietal and occipital cortices respectively. The
pathology is graded as none = 0, sparse (one or two affected vessels per section) =1,
moderate (several vessels per section) = 3 and severe (many affected vessels per section)
= 5.
vamfr, vamte, vampa, and vamoc record the severity of vascular amyloid deposits in the
meninges of the frontal, temporal, parietal and occipital cortices respectively. The
pathology is graded as none = 0, sparse (one or two affected vessels per section) =1,
moderate (several vessels per section) = 3 and severe (many affected vessels per section)
= 5.
vahfr, vahte, vahpa, and vahoc record the severity of haemorrhages associated with
vascular amyloid deposits in the frontal, temporal, parietal and occipital cortices
respectively.
snlfr, snlte, snlpa, and snloc record the presence or absence of neuronal loss in the
frontal, temporal, parietal and occipital cortices respectively.
sgfr, sgte, sgpa, and sgoc record the presence or absence of severe gliosis in the frontal,
temporal, parietal and occipital cortices respectively.
otfr, otte, otpa, and otoc record the presence or absence of other pathology in the frontal,
temporal, parietal and occipital cortices respectively. An accompanying written note
should provide further details.
pbfr, pbte, pbpa, and pboc record the presence or absence of Pick bodies in the frontal,
temporal, parietal and occipital cortices respectively. The pathology is graded as none =
0, sparse (one or two affected neurons per section) =1, moderate (several affected
neurons per section) = 3 and severe (many affected neurons per section) = 5.
lbfr, lbte, lbpa, and lboc record the presence or absence of alpha synuclein reactive
Lewy bodies in the frontal, temporal, parietal and occipital cortices respectively. The
pathology is graded as none = 0, sparse (one or two affected neurons per section) =1,
moderate (several affected neurons per section) = 3 and severe (many affected neurons
per section) = 5.
updfr, updte, updpa updoc, ubiincfro, ubiinctem, ubiincpar and ubiincocc record the
presence of ubiquitin positive dots and inclusions in the frontal, temporal, parietal and
occipital cortices respectively. These were recorded in early cases but not recorded in
later cases and usually are filled with 9 in the STATA database.
Evaluation of other selected regions and nuclei
nlsn, nlnb, nldr, nllc and nldv record the severity of neuronal loss in the substantia
nigra, nucleus basalis, dorsal Raphe nucleus, locus coeruleus and dorsal vagus nucleus
respectively. The pathology is graded as none = 0, sparse = 1, moderate = 3 and severe =
5.
glsn, glnb, gldr, gllc and gldv record the severity of gliosis in the substantia nigra,
nucleus basalis, dorsal Raphe nucleus, locus coeruleus and dorsal vagus nucleus
respectively. The pathology is graded as none = 0, sparse = 1, moderate = 3 and severe =
5.
pisn, pinb, pidr, pilc and pidv record the severity of pigment incontinence in the
substantia nigra, nucleus basalis, dorsal Raphe nucleus, locus coeruleus and dorsal vagus
nucleus respectively. The pathology is graded as none = 0, sparse = 1, moderate = 3 and
severe = 5.
otsn, otnb, otdr, otlc and otdv record the severity of other pathology in the substantia
nigra, nucleus basalis, dorsal Raphe nucleus, locus coeruleus and dorsal vagus nucleus
respectively. There should be an accompanying written note with further details.
lbsn, lbnb, lbdr, lblc and lbdv record the severity of alpha synuclein reactive Lewy
bodies in the substantia nigra, nucleus basalis, dorsal Raphe nucleus, locus coeruleus and
dorsal vagus nucleus respectively. The pathology is graded as none = 0, sparse (one or
two affected neurons per section) =1, moderate (several affected neurons per section) = 3
and severe (many affected neurons per section) = 5.
nftsn, nftnb, nftdr, nftlc and nftdv record the severity of tau reactive neurofibrillary
tangles in the substantia nigra, nucleus basalis, dorsal Raphe nucleus, locus coeruleus and
dorsal vagus nucleus respectively. The pathology is graded as none = 0, sparse (one or
two affected neurons per section) =1, moderate (several affected neurons per section) = 3
and severe (many affected neurons per section) = 5.
npsn, npnb, npdr, nplc and npdv record the severity of tau reactive neuritic plaques in
the substantia nigra, nucleus basalis, dorsal Raphe nucleus, locus coeruleus and dorsal
vagus nucleus respectively. The pathology is graded as none = 0, sparse = 1, moderate =
3 and severe = 5.
cblag records the presence of agonal changes in the cerebellum (i.e. changes in the
cerebellum due to the dying process)
Section H Assessment of Neuropathological Findings
cerad records the overall CERAD score of age against tau reactive neuritic plaques. For
CC75C, with all participants over the age of 75, this is relatively straight forward. The
pathology is scored as none = 0 (No plaques seen in any slide from the neocortex),
uncertain evidence of AD = 1 (One or sparse neuritic plaques in any slide from the
neocortex), suggestion of AD = 3 (Moderate density of neuritic plaques seen in any slide
from the neocortex), Indicative of AD = 5 (severe density of neuritic plaques seen in any
slide from the neocortex).
bsnft records the overall Braak stage for the case according to [2]. This is more difficult
and requires assessment of the severity of neuropathology in many areas. Although the
staging is generally applicable, many cases do not fit the stages neatly and some
subjectivity is involved.
bsap records the overall Braak stage according to amyloid beta protein reactive plaques.
This is not used.
Section I: Neuropathological Diagnoses (blind to clinical information with regard to
dementia).
npd1 – npd25 record the neuropathological diagnoses according to the table below.
Code
Npd1
Npd2
Npd3
Npd4
Npd5
Npd6
Npd7
Npd8
Npd9
Npd10
Npd11
Npd12
Npd13
Npd14
Npd15
Npd16
Npd17
Npd18
Npd19
Npd20
Npd21
Npd22
Npd23
Npd24
Npd25
Neuropathological Diagnosis
Normal brain, clinically insignificant
Alzheimer type pathology – cortical
Alzheimer type pathology – subcortical
Alzheimer type pathology – brainstem
Lewy body type pathology – cortical
Lewy body type pathology – subcortical
Lewy body type pathology – brainstem
Vascular Disease – large vessel disease – multiple large infarcts
Vascular Disease – large vessel disease – single large infarct (non strategic)
Vascular Disease – large vessel disease – strategic infarct
Vascular Disease - small vessel disease – multiple small infarcts
Vascular Disease - small vessel disease – multiple small infarcts inc strategic area
Vascular Disease - small vessel disease – white matter pallor
Vascular Disease - small vessel disease – V-R space expansion
Vascular Disease - haemorrhage – intra-cerebral
Vascular Disease - haemorrhage – extra-cerebral
Vascular Disease - haemorrhage – other (specify)
Pick’s disease (with Pick bodies)
Lobar atrophy (without Pick bodies)
CJD spongiform encephalopathy
Huntingdon’s disease
Leukoencephalopathy (specify)
Tumour – primary – specify
Tumour – secondary – specify
Tumour – other – specify
The neuropathological diagnoses are then ranked in order of significance; r1 is related to
dia1, r2 is related to dia2 etc to r6 is related to dia6. Diagnoses may share a rank
Notes on STATA variables that are not used or discontinued
The variables trans1 – nbm8 in the STATA database were used in early cases, up to
about L69. They were not used in later cases. It is not known what they were used for or
what they represent.
The diagnostic variables in the STATA database, fdia1 – fdia4 appear to have a few
values in the early cases up to L120 but were not used at all in later cases. It is not known
what these values represent.
References
1. Consortium to Establish a Registry for Alzheimer’s Disease.
http://cerad.mc.duke.edu/Default.htm
2. Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. Acta
Neuropathol. 1991;82(4):239-59.
List of abbreviations
AD
BGF1
BGF2
CC75C
CERAD
CING
ERCF
FCF
HCF1
HCF2
OCLF
OCMF
PCF
TCF
Alzheimer's Disease
basal ganglia area 1 formalin fixed
basal ganglia area 2 formalin fixed
Cambridge City over 75 Cohort
Consortium to Establish a Registry for Alzheimer's Disease
cingulate
entorhinal cortex formalin fixed
frontal cortex formalin fixed
hippocampus anterior formalin fixed
hippocampus posterior formalin fixed
occipital lateral cortex formalin fixed
occipital medial cortex formalin fixed
parietal cortex formalin fixed
temporal cortex formalin fixed