<<

. 23
( 45 .)



>>

Du 153, 1145-l 148
5 1. Brown, P., Llberskl, P P , Wolff, A , and GaJdusek,D C ( 1990) Resistanceof
scraple infectlvlty to steam autoclavmg after formaldehyde fixation and lImIted
survival after ashing at 36O™C practical and theoretical lmphcatlons J Infect
DIS 161,467472
52. Collinge, J., Palmer, M S , and Dryden, A. J (1991) Genetic predlsposltlon to
latrogenic Creutzfeldt-Jakob diseaseLancet 337, 1441,1442.
53 Brown, P , CervenhkovB, L , Goldfarb, L. G., McComble, W. R , Rubenstem,R.,
Will, R. G , et al (1994) Iatrogemc Creutzfeldt-Jakob disease* example of the
an
interplay between ancient genesand modern medicine Neurology 44,29 l-293
54. Deslys, J.-P., Marc& D., and Dormont, D. (1994) Slmllar genetic susceptlblhty m
iatrogemc and sporadic Creutzfeldt-Jakob diseaseJ Gen Vu-01 75,23-27
55 Tange, R A , Troost, D , and Llmburg, M (1990) Progressive fatal dementia
(Creutzfeldt-Jakob disease)in a patient who received homograft tissue for tym-
pamc membraneclosure Eur Arch Otorhmolaryngol. 247, 199-20 1.
56 Manuehdls, E E., Kim, J H , Mencangas, J R., and Manuehdls, L. (1985) Trans-
mlsslon to animals of Creutzfeldt-Jakob diseasefrom human blood. Lancef ii,
896,897.
57. Tatelshl, J (1985) Transmissionof Creutzfeldt-Jakob diseasefrom human blood
and urine mto mice. Lancet ii, 1074.
58 Tamal, Y , Kojuma, H , KitaJlma, R., Taguchl, F , Ohtam, Y , Kawaguchl, T , et
al. (1992) Demonstration of the transmissible agent m tissue from a pregnant
woman with Creutzfeldt-Jakob diseaseN Engl J Med 327,649
154 Brown

59 Deslys, J P , LasmCzas, C , and Dormont, D. (1994) Selection of specific strams
m latrogemc Creutzfeldt-Jakob disease Lancet 343, 848,849
60 Esmonde, T F G , W111, R. G , Slattery, J. M., Knight, R , Harries-Jones, R , de
Sllva, R , et al (1993) Creutzfeldt-Jakob disease and blood transfusion Lancet
341,205-207
61 Heye, N , Hensen, S , and Muller, N (1994) Creutzfeldt-Jakob disease and blood
transfusion Lancet 343,298
62 Dl Martmo, A , Safar, J , Cerom, M , and Gibbs, C J , Jr (1992) Purlficatlon of
non-mfectlous ganglioslde preparations from scraple-Infected brain tissue Arch
Vz,I 124, 11 l-121
63 Masters, C L , Hams, J 0 , GaJdusek, D C , Gibbs, C. J , Jr, Bernoulh, C , and
Asher, D M (1979) Creutzfeldt-Jakob disease. patterns of world wide occur-
rence and the slgmficance of famlhal and sporadic clustermg Ann Neurol 5,
177-188
64 Brown, P , Cathala, F , Raubertas, R. F., GaJdusek, D C , and Castalgne, P (1987)
The epldemlology of Creutzfeldt-Jakob disease* conclusion of a 15-year mvestl-
gatron m France and review of the world literature Neurology 37, 895-904
65 Harries-Jones, R., Knight, R , Will, R. G., Cousens, S., Smith, P G , and
Matthews, W B (1988) Creutzfeldt-Jakob disease m England and Wales, 198&
1984. a case control study ofpotentlal risk factors. J Neurol Neurosurg Psychat
51,1113-1119
66 Schoene, W C., Masters, C L , Gibbs, C J , Jr, GaJdusek, D C , Tyler, H R ,
Moore, F D., et al (1981) Transmissible spongtform encephalopathy (Creutzfeldt-
Jakob disease) Atypical chmcal and pathological findings. Arch Neural 38,47S477
67 Miller, D. C (1988) Creutzfeldt-Jakob disease m hlstopathology techmclans N
Engl J Med 318,853,854
68. Sltwell, L , Lath, B , Atack, E., and Atack, D. (1988) Creutzfeldt-Jakob disease m
histopathology techmcians N Engl J Med 318, 854
69 Gorman, D. G , Benson, D. F., Vogel, D. G , and Vmters, H V (1992) Creutzfeldt-
Jakob disease m a pathologist. Neurology 42,463
70 Weber, T , Hayrettm, T , Holdorff, B , Collmge, J., Palmer, M., Kretzschmar, H
A , et al (1993) Transmlsslon of Creutzfeldt-Jakob disease by handlmg of dura
mater Lancet 341, 123,124
71 Berger, J. R. and David, N J. (1993) Creutzfeldt-Jakob disease m a physician a
review of the disorder m health care workers Neurology 43,205,206
72. Sawcer, S. J., Yuill, G. M., Esmonde, T F. G., Estlbelro, P , Ironside, J. W., Bell,
J E , et al (1993) Creutzfeldt-Jakob disease m an mdlvldual occupatlonally
exposed to BSE. Lancet 341,642
73 Alperovltch, A , Brown, P , Weber, T , Pocchlan, M , Hofman, A, and Will, R
(1993) Incidence of Creutzfeldt-Jakob disease m Europe m 1993. Lancet 343,9 18.
74 Brown, P , Kenney, K , Little, B , Ironside, J , Will, R, Cervenikovi, L., et al
(1995) The mtracerebral distribution of mfectlous amylold protein m patients with
spongiform encephalopathy Ann Neurol , m press
Q

Bovine Spongiform Encephalopathy
Methods of Analyzing the Epidemic in the United Kingdom

John W. Wilesmith


1. Introduction
It seems unlikely that anyone could have foretold the Interest, controversy,
and concern that the occurrence of bovme spongiform encephalopathy (BSE)
m cattle m Great Britain would cause when the author was commissioned to
investigate its eptdemiology in the sprmg of 1987. The epidemic has proved to
be the largest food-borne epidemic of a transmissible spongiform enceph-
alopathy (TSE), and has been the subject of detailed scrutmy epidemiologi-
tally, by international agencies and governments, and by representatives of all
types of media worldwide. During the course of the epidemic an attempt has
been made to decipher the epidemiology of BSE with the usual objectives, the
most nnportant being to provide a model of causation such that necessary statu-
tory controls could be identified and enacted to protect animal and human
health, both in Great Britam and abroad.
In this chapter a brief history of the epidemiological studies of BSE in Great
Britain is described. Inherent m these studies are the problems of a protracted
incubation period, the absence of a valid diagnostic test in the live animal, the
novelty of the disease m cattle, and the usual difficulties of conducting epide-
miologrcal research as outlined by Rothman (I). Hopefully, these hurdles are
not used as excuses and the following may be of help in investigating diseases
with similar characteristics that may well occur in the future.
2. The Identification of BSE in Great Britain
BSE was first identified m the south of England m November 1986 (2) as a
result of what could be termed the background surverllance of ammal disease
From Methods m Molecular Medrane Pnon Diseases
Edlted by H Baker and R M Rtdley Humana Press Inc , Totowa, NJ

155
156 Wilesmith

m Great Britain Although this IS a passive rather than active system tt mvolves
a relatively htgh degree of active communication between animal keepers and
then- vetermary surgeons, who m turn seek help from the network of Veterr-
nary Investtgatton (VI) Centres, whose staff may seek spectahst advtce from
the Central Vetermary Laboratory (CVL) More specifically, BSE was tdentr-
tied as a result of the occurrence of multiple cases of unusual neurologtcal
disease m adult cattle m large dairy herds and the informal exchange of infor-
matron between herd owners about unusual cases of disease. An mtttal dtfti-
culty was securing sufftctently well-fixed brain material for htstologtcal
exammatton from suspect cases of BSE. This, however, did not delay the tden-
tificatton of the disease unduly and fortunately electron mtcroscoptsts at CVL
were experienced m the detection of scrapte-associated fibrtls (SAF) as a result
of a then recently completed study of SAFs m sheep scraple (3). The tdentlti-
canon of SAFs m bovine brains wtth spongrform encephalopathy provided
some confidence that the new disease was a TSE (2).
In retrospect, tt IS clear that the national animal disease survetllance system
identified BSE at a relatively early stage of the epidemic This IS probably a
result, m part, of the absence of a federal structure and the national mfrastructure
of the State Vetermary Services VI Centres, which mtrmstcally has a high degree
and rapid rate of commumcatlon and the required expertise m neuropathology

3. Case Definition of BSE
Previous research on TSEs m other species had indicated that tt was highly
unlikely that there would be a dtagnostlc test for the presence of the abnormal
PrPSC m the live animal in the foreseeable future. Although the chronic clmlcal
course and ultimate clmrcal signs presented by cases of BSE appeared to be
pathognomomc, tt was apparent that the initial clmtcal signs were not. Two
options were therefore available m selecting a valid diagnostic method for the
case definmon of BSE, and both were pathologtcal The first was the electron
mlcroscoprcal (EM) exammatton of fresh brain tissue for SAFs and the second
was a htstologtcal exammatton of fixed brain. The mtttal compartson of these
two dragnosttc methods was made dtfficult by the competmon for the same ana-
tomical areas of the brain, since the prevalence of SAFs was likely to be great-
est at sites where htstologtcal changes are most marked. This compartson was
made more difficult because the htstological findings in sheep scrapte indicated
a vartatton m the dtstrtbution of lessons (3).
The outcome of this contest was the decrston to conduct a htstologtcal examt-
nation of 64 neuron groups of the brains of suspect cases of BSE (4) during the
course of the mittal eptdemtologtcal study. Eventually the opportumty was
taken to examine the posstbihty of htstologrcal examination of a reduced num-
ber of brain sections for routme diagnosis, especially because of the apparent
BSE Analysis m the UK 157




7 ALTERATIONS/REFINEMENTS
TO HYPOTHESES AND CASE
DEFINITION
Fig. 1. Dynamics of the epidemlologxal researchplan


remarkable uniformity of the distribution of bram lesions (5,6). This resulted
m the present method of dtagnosls based on a single section, which 1s possrble
because the uniform distrlbutton of lesions was confirmed (7). Subsequent stud-
tes of the validity of EM exammattons for the presence of SAF have indicated
that hrstological examination 1s the preferred method of diagnosis (8). The
exception IS the exammation of autolysed bram tissues, for which EM examr-
natton has proved to be superior (9).
The validity of the hrstologrcal examination has been the subject of a contm-
ued reassessment as described in Section 6. This has been a necessary aspect of
the research effort, and has been extended to assessments of the more novel
methods of dtagnosrs, such as immuno-cytochemtstry. These are in progress,
but raise the issue of the detection of infection rather than disease. Studres of
the pathogenesis of BSE (IO), still m progress, should facilitate the validation
of these and other methods of diagnosis. In the meantime, the hlstologlcal
examination of the brain stem remains as the routine method of confirming a
clmrcal diagnosis.

4. The Initial Epidemiological Studies
4.1. The Basic Plan
Much has been written about eptdemiologtcal concepts and methods of ept-
demiological research. These essentially philosophical treatises are important,
but a more pragmatic description of the sctentrfic methodology IS simply to
develop and identify causal hypotheses and test their validity. This naturally
results m a dynamic process m conductmg epidemiologrcal research. The basic
plan, in its simplified form, for this as applied to BSE is shown in Fig. 1 Despite
WilesmIth
the absence of a complete understandmg, it has been necessary throughout this
process to make causal inferences m order that control measures to protect
pubhc and animal health could be implemented m a timely manner. Therefore,
overlaymg the basic plan was a systemattc means to asstst m differenttatmg
causal from noncausal assoctattons. The most attractive set of standard criteria
is that proposed by Hill, which are strength, consistency, spectfictty, temporal-
ity, btologtcal gradient, plausibthty, coherence, experimental evidence, and
analogy (2 I). These have been discussed m some detatl by Rothman (I).
Expertise m veterinary eptdemtology and some experience in balancing the
need for rapid results with the requirement to carry out rigorous medium- to
long-term studies are needed to estabhsh such a plan. An element of sound
epidemiological Judgment is therefore required along the way.
4.2. Identification of the Causal Hypotheses and the Means
of Their Investigation in the Initial Epidemiological Study
Despite the histological findmgs of a spongtform encephalopathy and the
detectton of SAFs by electron microscopy, there was a degree of uncertainty
about the possrble etiology of the disease. The ettologtcal hypotheses were not
restricted to BSE bemg a member of the TSEs. This was important because
alternative hypotheses could have abounded and, indeed, have been contmu-
ally proffered. Also, even if BSE proved to be a TSE, other factors could be
component, rather than sufficient, causes. Those investigated can be summa-
rized as a scrapie-like agent, an mtoxtcation, or a purely genetic dtsease (12,23).
For the first of these the posstble sources or vehicles of mfectton were mvestt-
gated as far as was possible These were imported cattle, contact with sheep,
contact with wtldltfe, and contaminated biological products, mcluding
feedstuffs
Having established these basics, the next step was to decide the most appro-
prtate design for the mmal epidemiologtcal study. There were three obvtous
options: an abattoir survey, a random survey of farms, and detailed case stud-
ies of affected herds and animals These needed to be considered with the other
desirable objectives of the study, which mcluded determinmg whether BSE
was a truly new disease, and if It was, the time when the first cases occurred;
securing a sufficient number of brams to establish a case definition; obtammg
an improved descrtptton of the clmtcal signs and the clinical course; and col-
lecting the essential descriptive epidemiologtcal data. In the outcome, the
evaluatton of the three options was not protracted. An abattotr survey has the
inherent difficulty of tracmg animals back to their last and natal herds. Thts,
together with the fact that a large proportion of cases would not be clinically
acceptable for slaughter and the incidence and prevalence of the disease
appeared to be low (23), made this an meffctent option. Stmilarly, for the last
BSE Analysis in the UK

reason a national survey of herds would clearly have been an inefficient
method. The remaining option, involving detailed case studies, was therefore
chosen, but required that the ascertainment of suspect cases be maximized.
This was achieved by reversing the operation of the surveillance network and
raising awareness of the disease and, most importantly, disseminating mfor-
mation about what was known of the clinical signs. The resulting voluntary
notifications of animals suspected of having BSE provided the cases for this
study. An evaluation of the availability of the required documented data and
information and the development of a questionnan-e and lines of questioning
was the next step. This was achieved by the author vismng farms that had
suspect cases and interviewing herd owners and workers using a proforma
questionnaire. Fortunately, herd owners were very cooperative and It became
clear that there would be sufficient documented data to determine the neces-
sary life histories of the cases and the parent herd. A standard questionnaire
was therefore developed together with a list of potential sources of data from
the range of farm records encountered during the feaslblllty stage.
4.3. Assessment of the Initial Causal Hypotheses
and the Formulation of Additional Hypotheses
The target for the mitral study was to obtam the detailed epldemiologlcal
data from 200 cases.This was achieved by December 1987, having commenced
in May of that year. It resulted m a relatively large information base and insti-
gated the development of the BSE computer-based database (14)
The assessmentof the causal hypotheses proved to be relatively straightfor-
ward (12). Essentially, intoxication either from veterinary therapeutic and pro-
phylactic products or agricultural chemicals could be ruled out since no
particular product or generic chemical compound emerged as a common fac-
tor. There was also considerable evidence against BSE being merely of genetic
origin; for example, a wide range of breeds were affected. In examining the
TSE hypothesis, there was no possibility that BSE had been introduced with
the importation of cattle. Cases had occurred m totally closed herds with no
contact with other cattle. Contact with sheep was also a highly improbable
reason for the infection of cattle, because only a small proportion of cases
occurred on farms that maintained breeding flocks or, indeed, any type of flock
Similarly, the presence of species of wildlife, such as deer, was not identified
as a common factor. Although new bovine vaccines had become available dur-
ing the period of Interest, these and existing vaccines together with other bio-
logical products were recorded at very low rates of usage.
This process of ehmmation left the possibility that animal protein, m the
form of meat and bone meal, included in commercial cattle feedstuffs, or some
other unidentified medium, was the vehicle of infection for a scrapie-like agent.
160 Wilesmith

The first step was to double-check the feeding histories obtained from the
affected farms and then seek the formulation of the commercial feedstuffs con-

<<

. 23
( 45 .)



>>