. 11
( 45 .)


stage. In 70% of casesthat were examined, the cerebellum displayed a signifi-
cant number of small plaques (26) with a dense center when stained for amy-
loid, but with a dense peripheral rim on staining for PrPCJD. A corona of
radiating Iibrils typically is present in these classical “km-u-type plaques.” The
plaques have been demonstrated in all the layers of the cerebellum, particu-
larly in the granular layer. They also are present elsewhere in the CNS. It is of
interest that some casesof apparently sporadic CJD closely resemble kuru cases
with regard to the type and distribution of plaques within the cerebellum and
elsewhere (Fig. 7).
5.2.2. latrogenic Cases
The development of CJD in some patients previously treated with human
hormones extracted from cadaver-derived pituitary glands (14,Z7), or in occa-
sional patients who have received cornea1 or dura mater grafts (33,34), has
provided a tragic opportunity to study the pathology and pattern of PrPCJD
deposits in iatrogenic casesin whom the site of entry of the transmissible agent,
Diagnosis of Human Prion Disease 73

Fig. 14. Section from the cerebellar cortex (same case as Fig. 13) showing
immunolocalization of PrPCJD in large and complex plaques in the molecular layer.
Smaller plaques are present within the granular layer and several large plaques are
present at the interface between molecular and granular layers. The plaque burden
is much greater in this case than in the plaque-bearing case of sporadic CJD shown
in Fig. 6.

and the latent period before development of signs and symptoms of CJD, can
be accurately identified. In patients treated with dural grafts, the pattern of
PrPCJD deposition resembles that seen in classical sporadic CJD, with predomi-
nantly cortical involvement and with a clinical presentation of dementia (2). In
contrast, patients treated with peripheral injections of cadaver-derived pitu-
itary hormones first display cerebellar signs and symptoms (35) (which may be
misdiagnosed as paraneoplastic syndrome or in some instances as recurrence
of the primary CNS tumor that originally had led to their treatment with pitu-
itary hormones), and immunocytochemical studies show predominantly cer-
ebellar positivity with accumulation of kuru-type plaques (similar to Fig. 6) or
more commonly, diffuse PrPCJDdeposition in the granular layer (similar to
Fig. 8) (2, I7,3.5,36). There may be very limited PrPCJDdeposition in the
cortical gray matter in patients treated with pituitary hormone injections. It
is of interest to note that the gray matter of the spinal cord shows heavy
PrPCJDdeposition (Fig. 15) in the anterior and posterior horns and Clarke™s
column (autonomic interneurons of the thoracic spinal cord) (27). These
cases of iatrogenic prion disease with spinal cord involvement resemble a
subset of sporadic CJD cases (Fig. 11) that also show spinal cord deposits
of PrPCJD(27).
Fig. 15. Computer image of the upper lumbar spinal cord from a case of growth-
hormone-induced CJD. Immunopositivity for PrPCJD is present in the gray matter and
particularly pronounced in the dorsal horn and substantia gelatinosa on each side. Posi-
tivity is also seen in Clarke™s column, indicated by the arrow.

5.3. Fatal Familial Insomnia
This genetic disorder is associated with a point mutation at codon 178 (3 7).
The pathological substrate of the clinical signs and symptoms appears to be
thalamic gliosis (Fig. 16) and there may be very little in the way of spongiform
change in this prion disorder. No plaques are seen. PrPCJDis demonstrated by
immunocytochemical means rather inconsistently in the thalamus and predomi-
nantly associated with neurons (Fig. 17). Reactive astrocytes are PrPCJD-nega-
tive even when associated with profound neuronal loss as in the inferior olive
but focal positivity for PrPCJD may be present in the granular layer of the cer-
ebellum (Fig. 18).
5.4. The Atypical Prion Dementias
In these familial cases, there are abnormalities of the PrP genotype but no
very characteristic neuropathological abnormalities (see Chapter 3). The brain
is atrophic and there may be cortical loss and gliosis, but spongiform change is
not a feature of these diseases (10,38). The cerebellar cortex may appear
atrophic but routine staining does not demonstrate any particular abnormality
in the molecular layer (Fig. 19). However, immunostaining with antibodies to
PrPCJD reveals striking and consistent linear deposits confined to the molecular
layer (Fig. 20).
Some prion dementiasmay be particularly difftcult to diagnose on routine stain-
ing and the benefits of reliable PrPcJDimmunolocalization clearly relate not only
to diagnostic clarification but also to judging the extent of pathological change.
Diagnosis of Human Prion Disease 75

Fig. 16. Case of fatal familial insomnia showing thalamic gliosis. Astrocytes are
identified by strong immunoreactivity for glial fibrillary acidic protein. Neuronal loss
is evident but spongiform change is not a noticeable feature in this section.

Fig. 17. Occasional neurons in the thalamus in fatal familial insomnia show strong
positivity for PrPCJD.

6. Differentiation from Other Dementias
Although there should be no difficulty in differentiating classical CJD cases
from other dementias, clearly there is some neuropathological overlap and diag-
nostic difficulty in separating subtle cases of prion diseases from the range of
76 Bell

Fig. 18. Cerebellar cortex from a case of fatal familial insomnia showing focal
positivity for PrPCJD in the granular layer.

Fig. 19. Section of the cerebellar cortex from a case of atypical prion dementia with
a mutation in the prion gene but no evidence of spongiform change. In routine stain-
ing, the cerebellum appears somewhat atrophic but otherwise unremarkable.

other dementias in which spongiform change, amyloid plaques, or neurotibril-
lary tangles, or simply neuronal loss and extensive gliosis, are common fea-
tures (2,3,.5,8,9,32,39-41). These difficulties have been alluded to both in
Chapter 3 and in this chapter. Very occasional cases have been reported in
which CJD does appear to coexist with other diseases(3,5,41) and we have had
Diagnosis of Human Prion Disease 77

Fig. 20. Section from the case of atypical prion dementia shown in Fig. 19, stained
for PrPCJD. Linear deposits are clearly identified in the molecular layer, whereas other
layers of the cortex are negative. Sections from this case, and permission to photo-
graph the staining results (Figs. 19 and 20), were kindly provided by P. Lantos.

project in the last 4 yr. There
two cases of this type in the national surveillance
are a number of factors that may be taken into consideration when trying to
finalize a diagnosis. These include the clinical and family history, EEG and
neuroimaging findings, the general pathology and neuropathology findings,
and the results of immunocytochemical investigation. In many casesinforma-
tion regarding the PrP genotype will be a useful adjunct in achieving diagnosis.
The advent of antibodies that will specifically detect different forms of amy-
loid, particularly PA4 amyloid and PrPCJD,has been very helpful. Transmissi-
bility would also help to separate prion diseasesfrom other dementias (221, but
until recently it has not been clear that all the prion diseases are clearly trans-
missible. In any case this is not a practical possibility in all casesbecause of the
cost and time involved in transmission experiments. In this regard, reliable
PrPCJDimmunolocalization is a most useful development.
7. Role of PrP lmmunocytochemistry in Diagnosis
Confidence in the reliability of PrPCJD immunolocalization has come slowly
in view of the potential pitfalls in its interpretation (see the following). In cases
currently under examination that have been fixed in formalin for ˜4 wk and
with routine processing of tissue blocks after immersion in 96% formic acid
78 Bell

for 1 h, there 1s no difficulty m demonstrating PrPCJD by lmmunocytochemls-
try. Occasional difficulty 1s encountered m archival cases or m cases referred
from elsewhere, m which classical CJD neuropathology, including confluent
sponglform change, 1s widespread but mununocytochemlcal results are nega-
tive. This we have attributed to variation m fixation time or processing sched-
ules. Nevertheless, the consistency with which the abnormal protein 1s localized
m nearly all disease cases; the posltlvlty with all the different antIbodies and
the abolltlon of posltlvlty when preabsorbed antibodies are used, the high level
of conformity between reports from different studies and the negativity of con-
trol cases, all suggest that PrP immunocytochemlstry has now graduated to
take Its place as a useful diagnostic procedure. The topographical dlscrepan-
cles between conventlonal pathology of prlon diseases and the PrPCJD deposl-
tlon raise some interesting questions but do not interfere with the diagnostic
usefulness of the technique, provided these discrepancies are recognized Reh-
able PrPCJD lmmunolocahzatlon will allow us to better define the mcldence of
these interesting diseases (IO), particularly those diseases that are very rare or
that show atypical neuropathologlcal changes, and also has a useful role m
separating prlon diseases from other neurodegeneratlve diseases The mterest-
mg relationships between the deposition of PrPCJD, clinical signs, and symp-
toms, and PrP genotype require further study before the diagnostic apphcatlons
of PrPCJD lmmunocytochemlstry are apparent m the full range of prton diseases.

8. Role of PrP lmmunocytochemistry
in Furthering the Understanding of Prion Diseases
If PrPCJD can be localized accurately within CNS tissues, the first step 1s to
delineate the cell types in which it 1spresent at light microscope and ultrastruc-
tural level. Clearly some of the deposits are extracellular and these are best
exemplified by the presence of plaques. Whether these deposits commence as
mtracellular accumulations that are actively extruded or become extracellular
as a result of cell breakdown 1s not clear (42). However, current lmmuno-
locahzatlon studies suggest that at least some PrP CJD deposits are mtracellular
or closely perlcellular (2,23). We are currently mvestlgatmg the ultrastructural
lmmunolocahzatlon of PrPCJD. Double lmmunostammg at the light microscope
level will also help to ldentlfy the cells in which, or around which, PrPCJD
deposition occurs.
The mapping of PrP CJDlocahzatlon may throw light on the spread of PrPCJD
wlthm the CNS. The finding that the spinal cord 1s involved m all cases of
latrogemc CJD, m which the agent has been introduced peripherally, and m
some cases of sporadic CJD, suggests that the spinal cord may be one pomt of
initial entry and accumulation of PrPCJD in the CNS (27). Current interest IS
focusing on possible accumulatron of PrPCJD m peripheral nerves and m auto-
Diagnosis of Human Prion Disease

nomic nerves. Although the results to date are negative m these ttssues,tt may
simply reflect the fact that the abnormal PrP is transported along nerves and
accumulates only once it reaches the CNS. Accumulatton of abnormal PrP m
peripheral tissues has been demonstrated in muscle cells m cases of mclusion
body myositis (43). How this condmon relates to PrP mutations and other
abnormalities m muscle ttssue is not clear. It 1sof interest that some transgemc
mace who express high levels of PrP transgenes do develop a severe general-
ized skeletal muscle dtsorder that on pathologtcal examination proves to be
necrotizing and associated with PrP accumulation (44). Whether PrPCJD be can
demonstrated in spleen and lymphoid tissues, as has been reported m animals
(24), awaits further mvestigatron. It is too early to say whether patterns of
PrPCJD deposition can yet be linked to the putative route of entry of the agent to
the CNS, but the stmtlartty of some casesof sporadic CJD to kuru raises mter-
estmg questions. PrPCJDm-nnunolocalizatton adds an extra dimension to the
study of drsease distributron wtthin the CNS that is particularly important m
the mvesttgatton of human cases for any possible links with the outbreak of
BSE. If BSE has had a srgmticant effect on human disease, it may be that the
pattern of pathology m human cases will come to resemble that seen m BSE
(19) and that there may be a stgnificant shaft of pathology, and associated
PrPCJDimmunolocahzation, to the brainstem.
9. Image Analysis Studies PrPCJD lmmunolocalization
In order to achieve an ObJective assessmentof PrPCJDload and distribution,
we have actively pursued image analysis techniques that allow observer-mde-
pendent assessmentof the correlatton of PrPCJDdeposits with the degree of
spongiform change and with other parameters, such as the extent of colocal-
izmg ghosts (45). This IS achteved by semr-Interactive or automatic assess-
ment of nnrnunocytochemtcally stained slides. Figures 11 and 15 represent
computer images of PrPcJDtmmunolocalization in the spmal cord in cases of
sporadic and growth-hormone-associated CJD. These methods show potential
for mappmg large numbers of sections from a range of cases m the different
subsetsof prion dtseasesand may add further useful information on the patho-
logical variants to be linked to the different genotypes of prton dtseases.
10. Pitfalls in PrPCJD lmmunocytochemistry
Most of these have already been alluded to m this chapter. To summartze,
the maJor difficulties surround the crossreacttvtty between antibodies that do
not differentiate between PrPC and PrPCJD Strategies to minimize the risks of
tmmunolocahzing PrPc and of restricting the reactton to PrPCJD employed
m most published studies and are included m our protocols (Table 2). In prac-
tice, there does not seem to be a difficulty wtth crossreacttvtty when CNS tis-
80 Bell

sue sections are examined m adult subJects.However, the applicatron of these
techmques to non-CNS ttssues, or to embryonic and fetal trssues, IS still at an
experimental stage and will benefit from further understanding of the normal
function of PrPC. The discordance that is sometimes present between pathol-
ogy changes and PrPCJDlocahzatron may be genuine or may be a problem of
sensitivity. It IS certain that autoclave pretreatment srgmficantly increases
vtsuahzatron of PrPCJDdistrrbutron m disease cases. However, accumulatron
of abnormal PrP has already occurred m symptomatrc prron diseasesand sensr-
ttvtty of techniques may be an issue m presymptomatic or short duration cases
since the temporal relationship between PrPCJD accumulation, onset of cellular
neurodegeneratrve changes, and chmcal signs and symptoms 1snot understood.
The fact that some long-fixed archival material does not respond posmvely
with rmmunocytochemrcal techniques for PrPCJDdoes suggest that laboratory
processmg of tissues has some effect on the likelihood of vrsuahzation. Some
very practical diffcultres are associated with the prevrous practice of rmmersron
of tissue m phenol that was undertaken m an attempt to decontaminate prron
disease cases(46). Phenol has a deleterious interaction with formic acid and the
two pretreatments should not be used m conJunctron. This IS largely hrstorrcal
since phenol immersion has been shown to be ineffective m decontaminatton.
However, pretreatment with formrc acid leads to some technical problems m its
own right; sections taken from tissue that has been exposed to formic acid for
any length of time are difficult to handle m the laboratory since their adherence
to glass slides IS poor unless the sltdes are pretreated with very effective bonding
agents However, the advantages of formic acid m enhancing not only the VISU-
ahzatron of PrPCJDbut also some other antigens outweigh this disadvantage.
Occastonal false positive stammg, or what 1sassumed to be false positive
staining, 1sobserved m blood vessel walls both m the subarachnord space and
wtthm CNS tissue This may be present with one or other PrP antibody and is
an mconststent finding.
Despite these drfficulties, PrP immunocytochemrstry has assumed a cred-
ible place m the dragnostrc and screntrfic investrgatron of human casesof prton
disease. Consensus m the hands of many investigators lends weight to the reh-
ability of these techmques and when the results are correlated with classical
hrstopathology, chmcal findings, genetic analysis, and other techniques, such
as Western blotting and unmunoblotting, they may contribute to significant


. 11
( 45 .)