<<

. 18
( 45 .)



>>

cases is that this type of clinical presentation, occurring mainly in patients of
middle age (Fig. l), will result in referral for a medical opinion, hospital admis-
sion, and referral to a neurologist and/or the recording of an electroencepha-
logram (EEG) and/or eventual neuropathological examination.
In countries with a developed medical system, it is extremely unlikely that
patients who become totally dependent will not be admitted to the hospital and
in France it was established that all such patients are subsequently referred to a
neurologist (7). In the United Kingdom the 10% of patients with CJD initially
admitted to the hospital under the care of a psychiatrist were all later referred
to a neurologist. Between 1990 and 1994,84% of patients with CJD underwent
Prion Disease Surveillance 121

Table 1
The Electroencephalogram in CJD (UK 1990-1994)
Atypical Untraced
Typical No EEG
Sporadic 99 68 17 4
Familial 3 2 0
Iatrogenic ;; 7 0 5
“3 codon 200 mutations, 2 msert mutations
h2 dura mater reclplents


an EEG and 70% had a postmortem. The referral of CJD cases to a national
register depends on the cooperation of neurologists and other professlonal
groups. In the United Kingdom neurologists, neurophysiologlsts, and neuro-
pathologists are regularly cn-cularized with mformation on the CJD Survetl-
lance ProJect and asked to refer any suspect case of CJD.
CJD has been referred to as the archetype exotic neurological disorder (8),
perhaps because It is both untreatable and in many casesrecogmzable from the
end of the bed Although there was almost certainly some improvement m dlag-
nostic efficiency from the 1960s to the 1970s in the United Kingdom, typical
casesare now almost certainly promptly recognized. In order to Identify a high
proportion of cases, including atypical variants, an important component of
surveillance 1sthe referral of “suspect” as well as “typical” cases and approx
50% of all referred cases (from all sources, including death certificates) are
eventually classified as “possible” CJD or “not CJD” (6).
Another source of cases 1s from neurophyslology departments Between
1990 and 1994,90% of casesof CJD underwent an EEG and 60% of defimte or
probable caseshad a characteristic tracing (Table 1). Although the great maJor-
ity of cases of CJD were identified from other sources, a small proportlon of
caseswere identified from EEG departments alone.
Neuropathological confirmation of the diagnosis of CJD is an important
component of surveillance. The accuracy of clinical diagnosis 1sover 95% m
cases with a typical clinical presentation and a characterlstlc EEG, but a sig-
nificant proportion of casesdo not undergo EEG or have a single atypical trac-
ing. Similarly, in a small proportion of casesthere may be no clinical susplcron
of CJD, for example casesreferred to nonneurologlsts. Thirdly, atypical forms
of prlon disease may be ldentlfied. This 1san important conslderatlon in rela-
tion to the putative risk of BSE because atypical forms of human prion disease
may have a very atypical clinical course from classical CJD. The first case of
CJD m a human growth hormone recipient m the United Kingdom was certl-
fied as dying of encephalitis and it was only after the ldentlficatlon of the typl-
cal pathological changes that the diagnosis was confirmed.
122 W/II

Table 2
Accuracy of Death Certification of CJD in the UKB
Decade 1960s 1970s 1980s 1990s
Dlagnostlc classlticatlonb
CJD 39 73 66 67
Other 61 27 34 33
OPercentage of certified cases classified as CJD after exatnmatlon of records
˜Definite, probable or possible CJD

Table 3
Case Ascertainment of CJD
from Death Certificates in the United Kingdom (197Gl992)
Period 1970s 198&1984 199&l 992
Percentage of total cases” certified as CJD 78 72
62
Percentage of total cases of CJDa
ldentlfled from death certificates alone 42 13 6
˜Definite and probable cases


2.2. Death Certificates
It 1s inevitable that a proportion of cases of CJD are either not referred to a
neurologist or not notified to the surveillance center. As a safety net, all death
certificates coded under the rubrics 046.1 and 33 1.9 are obtained and clmlcal
and pathological details sought m every case.
It has been claimed that death certificates alone may be an efficient source of
case ascertainment m CJD (9), but this contrasts with experience m the United
Kingdom. The accuracy of death certlficatlon of CJD in the United Kingdom can
be measured by the percentage of cases classified as CJD after exammatlon of
case notes (Table 2). Although there has been an improvement m diagnostic
accuracy since the 196Os, currently only about two-thirds of certified cases of
CJD fulfill diagnostic criteria for the condition, underlining the need to review
case records/pathology reports in order to verify the diagnosis m certified cases.
An important question m relation to case ascertainment IS whether death
certificates, assuming diagnostic verification, can be regarded as an adequate
single source of cases for epldemlologlcal studies. This 1s an important issue
because current epldemlologlcal surveillance of CJD in some countries is based
primarily on this methodology (10). In the United Kingdom, cases of CJD are
Identified from a range of sources m order to achieve as high a level of case
ascertainment as possible. The proportion of cases of CJD that would have been
Identified from death certificates 1s shown in Table 3. In the United Kingdom
Prion Disease Surverllance 123

reliance on death certificates alone for case identification would have resulted
m missing 22-38% of cases of CJD. Retrospective surveillance of CJD may
have to rely primarily on death certificates as the source of cases. In England
and Wales between 1970 and 1979, the overall incidence of CJD was 0.3 cases/
milhon (Zl) and 42% of these cases were identified from death certificates
Prospective surveillance is more efficient and in both the periods 1980-1984
(5) and 1990-1992, the incidence of CJD was higher (0.49 and 0 67, respec-
tively) and the proportion of cases identified from death certificates was sig-
nificantly lower.
In concluston, death certificates are an important safety net for the identtfica-
tion of casesof CJD. However, it is essential to verify the diagnosis m certified
casesand the use of death certificates alone as a means of caseascertainment is
unreliable if a high degree of case ascertainment 1sto be achieved.
2.3. Elecfrcphysio/cgy
The EEG was first recognized as an important aid to the diagnosis of CJD m
1954 (12) and was included as a component of the first publtshed diagnostic
criteria for CJD in 1979 (13). In systematic surveys approx 70% of cases of
CJD exhibit the typical EEG pattern (24), which consrsts of generalized
triphasic periodic complexes occurrmg at a frequency of approx l/s (Fig. 2) In
some casesof CJD, no EEG is carried out and in others only a single tracing 1s
obtained early m the course of the illness, and the chances of obtainmg a char-
acteristic record are enhanced if serial recordmgs are carried out. Rarely, the
typical EEG does not develop at any stage of the clmtcal course and there are a
number of conditions m which the EEG changes mtmtc CJD (Table 4). In the
majority of these conditions, the differentiation from CJD is usually evident on
climcal grounds. The occurrence of a typical EEG tracing in Alzheimer™s dis-
ease has been described (15), but this is clearly an exceptionally rare occur-
rence because we have identified only one such case in 25 yr of CJD
surveillance m the United Kingdom. It is likely, but unproven, that the occur-
rence of similar EEG appearances m Lewy body dementia is also exceptional.
The relative specificity of the EEG changes in CJD and the high percentage
of cases with such a tracing indicate that EEG departments are a potentially
important source for case identification in CJD. In the United Kingdom, EEG
departments are circularized and visited when possible, which has led to the
identification of the small number of cases not identified from other sources.
The EEG is also a crucial component in the diagnostic classificatton of cases m
whtch there is no pathological confirmatton of the diagnosis and, as with death
certificates, it is essentral to review EEG records.
A maJor problem with the use of the EEG m the diagnosis of CJD is that
there are no established parameters for categorizing the EEG, other than the
124 WI/l




Fig. 2. The typical EEG m CJD


general description of the typical changes, and there 1svariation in the accu-
racy of the reporting of the EEG m CJD. An important element of surveillance
is to review the EEGs m all suspectcasesin order to build up experience and to
allow independent classlftcatlon of tracings, preferably blind to the clmlcal or
Prlon Disease Surverllance 125
Table 4
Conditions in Which
a CJD-Like EEG Appearance May Occur
Hyponatremta
Hypernatremta
Hypoglycemia
Hepattcencephalopathy
Hyperammonemla
Llthmm toxlctty
Metrrzamideencephalopathy
Alzhelmer™s disease
Lewy body dementia


pathologtcal features. Inevrtably there are occasional EEG tracmgs that are
dtfficult to classify and m such casesin the United Kingdom, an independent
opinion is sought from an mdivtdual with extensive previous expertence of the
EEG m CJD. The mtsclasslfication of EEGs can lead to apparent eptdemto-
logtcal anomalies. For example, CJD was diagnosed in 16 patients m two nergh-
bormg hospitals over 2 yr on the basis of the EEG, although review of the
tracmgs and clinical features did not allow the diagnosis of even possible CJD
in any case.
The EEG IS nonetheless an important component of dtagnosts and case clas-
sification in CJD. There IS, however, variatron m the duration of the periodic
complexes and the proportion of any record with such suggestive appearances,
indicating the need for established EEG criteria for the dtagnosts of CJD
2.4. Neuropathology
The clmtcal diagnosis of CJD 1s remarkably accurate in typical cases.
Review of the clmical features in two large series of pathologtcally confirmed
CJD has demonstrated that only 10% of cases present atypically (16) and m
some of these atypical cases the diagnosis was nonetheless suspected. How-
ever, the dlagnostlc classrfication of the 30% of caseswtthout a typical EEG IS
dependent on neuropathologtcal vertficatton and chmcally atypical casesmay
only be diagnosed after hrstological examination of the brain. A high degree of
case ascertainment IS crucially dependent on obtaining a high postmortem rate
in systematic surveys of CJD and the identification of rare or unusual forms of
human prton disease may only be possible through neuropathologtcal examt-
nation. In the UK study, neurohrstological examination has been important m
the correct diagnosis of the 5% of cases with a stroke-like presentation and m
the 10% of caseswith a duratron of illness of greater than a year. Neuropatho-
logical vertficatron is an rmportant component of case ascertainment and may
126 WI/I

be crucial to the recogmtton of atypical cases. The current postmortem rate m
the Umted Kingdom IS approx 70% of all suspect cases.
Practrcal difficulties may be encountered in obtaming postmortem m CJD
because of concerns about the potential risks to personnel m pathology depart-
ments Gutdelmes to mmrmtze any potential rusks during postmortems m CJD
have been published (I 7) and the eptdemrologtcal evidence does not suggest
that there IS a stgmficant rusk to postmortem personnel desptte the absence of
any precauttons m relation to CJD prror to the 1970s Screntrtic evidence can-
not, however, preclude the possrbthty of a risk and anxieties have been hetght-
ened by the tone of some recent media coverage A pragmatic approach to the
problem of postmortem in CJD is to establish dedicated neuropathology labo-
ratories for the condmon and to Identify a network of neuropathology depart-
ments where postmortem, limtted If necessary to examination of the bram, can
be carried out.
Brain bropsy IS fraught with practrcal and ethical problems Most rmpor-
tantly, such a procedure may be difficult to Justify rf rt IS unhkely to be of any
benefit to the pattent The biopsy may be obtained from a region of the brain
unaffected by the pathologrcal process and rt may be difficult to reach any
definite conclusron from exammatron of small portions of trssue The current
gurdelmes m the United Kingdom state that any neurosurgtcal mstruments used
m a potential case of CJD must be destroyed postoperatrvely.

3. Case Definition
The consistent application of dragnostrc criteria IS a szne qua non of eptde-
mrologlcal research Crrterra for the diagnosis of CJD were first proposed m
1979 (13) but have had to be adapted to take account of screnttfic develop-
ments Amyotrophtc “CJD”, m whrch there IS a combmation of dementia and a
motor neuron disease-like Illness, IS no longer regarded as a prton disease m
view of the almost uniform failure to demonstrate laboratory transmisstbrhty
(28) and the absence of detectable PrP on Western blotting m thus condmon
(29). Crlterra for a dtagnosrs of famthal CJD have always been problematic
The strict requirement of confirmed or probable CJD in another family mem-
ber may result m an underestimation of famlhal cases, whereas allowing any
form of dementia m other family members as the crtterron for famthal CJD
may result m an overestlmatron of famrltal cases. Advances m the molecular
biology of prton diseases now allows an accurate classificatton of famihal
cases, provided DNA IS obtained for analysts from incident cases.
In the Umted Kingdom the systematic use of DNA analysts m CJD has dem-
onstrated that about 12% of cases are associated with PrP gene mutations, dou-
bling the estimate of the frequency of familial cases from previous surveys
(I I). Approxrmately one-thud of these cases have no evident family history
127
Prion Disease Surveillance
and the identification of two cases of FFI m the United Kingdom was depen-
dent on the molecular biological data. On the other hand, three cases of CJD
have been identified with a clear family history of dementia and no mutation of
the PrP gene.
The systemattc analysts of DNA in mctdent casesof CJD has led to major
ethical dilemmas and a marked variation in approach to this problem from
country to country. In the United Kingdom informed consent from a relative of
the patient is an essential prerequisite before proceeding to DNA analysts.
Although the genetic issues m CJD are similar to those m other dominantly
inherited diseases, such as Huntington™s chorea, there are specitic problems
with systematic screening carried out m the course of CJD survetllance. Of
particular importance are the small proportton of cases m which a mutation of
the PrP gene 1sidentified despite the absence of any suggestive family history.
In such cases the failure to obtain informed consent leads to extraordmartly
difficult ethical problems and the alternattve approach, m which no results of
DNA testing in CJD are divulged, becomes increasingly dtfficult to justify m
view of the possibility of providing genetic advice to other family members,
including prenatal testing (20,21).
Another important development since the formulatton of the origmal diag-
nostic criteria has been the occurrence of iatrogenic CJD (22), particularly in
relation to human pituitary-derived hormones and human dura mater grafts In
human-growth-hormone recipients, the clinical diagnostic criteria for CJD
would not allow such casesto be classified as variants of CJD, which has been
considered in the formulation of new diagnostic criterta. Through the Euro-
pean Community (EC) a grant was awarded in 1993 for the coordmatton of
CJD surveillance m a number of countries m Europe: France, Germany, Italy,
the Netherlands, Slovakia, and the United Kingdom. One of the initial tasks m
this collaboratton was to review the diagnostic criteria for CJD. The amended
criteria are listed in Table 5.
3.1. A typical Cases of CJD
It is possible that the application of strict dtagnostic criteria for CJD will be
self-fulfilling and that atypical cases will be missed. There is remarkable con-

<<

. 18
( 45 .)



>>