<<

. 25
( 45 .)



>>

Again, no changes have been observed.
The third arm, to determine whether other means of transmission had
occurred, was achieved by a number of analyses in addition to those used to
determine the effects of the statutory control measures. The possiblllty of
maternal transmission was assessed by comparmg the observed incidence of
BSE m the offspring of confirmed cases with that expected as a result of expo-
sure from the food-borne source. The latter was estimated m a number of ways,
but essentially employed the annual mcidences m successive birth cohorts
(27,28). These analyses have been conducted at three monthly Intervals and
are not regarded as the ultimate means of determmmg whether or not there was
evidence of maternal transmrsston, but more as a background momtormg m an
attempt to identify this means of transmission at the earliest opportumty
Other analyses to this end included wtthm-herd mctdences in affected herds,
especially those in which the only cases were in purchased animals and there
was no evidence that the recipient herd had itself been exposed to mfectlon.
The population of beef suckler herds was of interest here because commercral
concentrates were fed much less frequently than m dairy herds, and because a
proportion of these cross-bred females are purchased from dairy herds to pro-
vide replacements for the adult herd.
The inevitable long-term monitoring of the epidemic was an important com-
ponent of the eprdemiological studies, especially m view of the national and
international interest resulting m a constant flow of questions of either a politt-
cal or scientific nature. The results and tindmgs have also been used to reassess
BSE Analysis in the UK 167

the origmal thoughts and hypotheses on the orrgm of the epidemic. Although
normal descriptive epidemiological analyses have provided the basis fat the
assessmentsand reassessments,it has also been necessary to identtfy novel
means of analysts to address specific aspects, such as the possibihty of other
means of transmission.
7. Further Analytical Studies
As described m Section 5., the best approach to determmmg whether or not
maternal transmission occurs is a cohort study. This was supplemented by con-
ducting the possible appropriate analysesto examine the incidence of BSE m off-
sprmg of confirmed casescompared with that expectedfrom the food-borne source.
The occurrence of casesof BSE m animals born after July 18, 1988, when
statutory measures were introduced to prevent further exposure from the food-
borne source, provided an opportunity for additional analytical studies The
prerequisite of these was naturally a more detailed study of each case, i.e., to
obtain descriptive epidemiological data to formulate hypotheses and, ulti-
mately, design analytical studies The outcome of these case studies indicated
that there was a clear risk of exposure for animals born m 1988, and possibly
later, from the feed-borne source, as a result of feedstuffs manufactured before
the ban still being m the food cham or on farms. The recommended “shelf hfe”
of these products is of the order of 3 mo, but they can remam fit for consump-
tion for years. Case studies of later-born animals proved to be unrewarding
with respect to an assessmentof the risk from the food-borne source. There-
fore, the feasrbrhty of conducting a case-control study was examined to assess
the risks of mfectron from feedstuffs, maternal and horizontal transmission,
and associated populatron-attributable risks.
It proved possible to design a within herd study to assessall these risks
except that from feed. More than 300 herds with homebred casesm cattle born
between November 1, 1988 and June 30, 1990 parttcrpated, and m addition to
the cases, up to four control animals were randomly selected from unaffected
homebred animals that had been born in the same herd and calving season as
each case.In order to take account of the matched study design, condtttonal logis-
tic regression was used with herd and calving seasonas the matching variable.
A simulation model was developed to determine the rates of transmission of
BSE m the population that would be consistent with the odds ratios observed
and supplement the population-attributable risk. Essentially the results indt-
cated that neither maternal nor horizontal transmission could account for the
maJortty of the cases, leaving the feed-borne source as the only known possi-
bility. This would have been at a reducing rate in successive birth cohorts smce
July 1988 because the changes in the risk of infection over time have been
estimated by calculatmg what is akin to standardized morbidity ratios for am-
168 Wilesmr th

mals born m each month and year since the ban (31). These continued analyses
have indicated that the risk of mfectton was reduced by at least 40% tmmedt-
ately after the introduction of the feed ban and the risk of mfectton for ammals
born m December 1990 was 10% of that for animals born m December 1987.
The question of obtaining evidence for a continued risk from the feed-borne
source, albeit at a much reduced rate, was therefore raised. The magmtude of
the dose exposure m natural mfectton had been a subJect of interest throughout
the eptdemrc and the weight of evidence was that tt was a low-dose phenom-
enon (32) Evidence IS currently emerging from the expertmental oral expo-
sure of calves with 1, 10, and 100 g of brain from terminal cases, each on a
smgle occaston, and 100 g on three occasions (33), that the eptdemtologrcal
mterpretatton was correct There was, therefore, some biologtcal plaustbthty
for accidental contammatlon of cattle feedstuffs involvmg relatively small
quantmes of ruminant-derived meat and bone meal.
The posstbtltty of such contammatton was examined by recourse to the basic
descriptive eptdemtology, studying the regional proporttonal distribution of
homebred cases by 12-mo birth cohorts before and after the feed ban. This
revealed that followmg the feed ban the proportton of homebred cases m the
eastern and northern regtons of England Increased (34). This was of interest
because the national poultry and ptg populattons are largely concentrated m
these regions and the feedstuffs for these species could legally have included
ruminant protein derived from potentrally high-risk tissues (although not from
clmically suspect animals) until September 1990. More importantly, there are
a number of mills producmg commercial feedstuffs for both ruminants and
monogastrtc animals. The correlatton between the incidence of homebred cases
of BSE m animals born after October 30, 1988 m each county m England and
Wales and the ratios of adult cattle to the number of pigs, the number of poul-
try, and the combined populatton sizes of ptgs and poultry in each county was
examined as a further analysts of the descriptive eptdemtologrcal data. A highly
stattstically stgmficant correlation was found (34) Such an ecologtcal correla-
tion cannot obvtously be taken as evidence of a causal assoctatton, but tt can
provide confidence in the further assessment of the hypothesis A between-
herd case-control study was the appropriate means for further study of the prob-
lem, and prehmmary vtstts to a small sample of feed mills were made to assist
m tinahzing the study design. However, these visits provided evidence that
accidental cross-contammatton could have occurred m mills producing rations
for cattle and nonrummant species. The envisaged case-control study has yet
to be completed, but efforts have been made to ensure the enforcement of the
statutory bans on both the feeding of ruminant-derived protein to rummants
(July 1988) and the specified bovine offals (September 1990) considered likely
to contam significant titers of the BSE agent, and the legislation controlling the
BSE Analysis m the UK 169
separatton and disposal of specified bovine offals was further strengthened m
April and July 1995.

8. The Future Course of the Epidemic
All epidemtologists who become mvolved m studying diseases of national
and/or mternattonal concern will become aware of the Inherent interest of the
general human populatton m foretelling what the future holds. There is, there-
fore, a need to recognize the difference between the perception, by the media
and their attendant population, of the power of astrology and that of epidemto-
logical methods. On the other hand, there has been a specific need to estimate
the number of animals to be slaughtered as suspect casesof BSE m the commg
years, since money must be allocated to compensate the owners of animals that
are compulsorily slaughtered, and to pay for the incineration of the carcasses
of slaughtered cattle. Modeling studies to provide these estimates have, there-
fore, been necessary. A number of approaches have been used (12,33,35,36),
but an age-period-cohort approach has proved to be remarkably successful and
accurate, and the current decline in the incidence is m agreement with that
predicted Although pubhshmg estimates of the future number of caseswould
not be prudent, the present evidence is that the incidence of BSE will continue
to decline such that by the turn of the mtllenmum the incidence will be insig-
nificant. This said, there is a need to continue to study the epidemic m some
detail m order to detect any changes m the epidemtology that could result in an
undue and unexpected prolongation of the epidemic.
9. An Epidemiological Epilogue
As indicated at the beginning of this chapter, the occurrence of BSE was
unexpected. Its occurrence presented a number of difficult problems, the appar-
ent answers for some of which naturally raising further questions that increased
the level of concern about how cattle had become exposed. However, from a
personal point of view, a systematic approach involving an adherence to the
basic study plan proved to be of great help m accumulating the necessary evt-
dence and therefore confidence, especially for providmg advice for the formu-
lation of control pohctes. Important in this process was obtaining valid and
detailed enough descrtptive epidemiological data to conduct relatively stratght-
forward analyses. Although subsequent, more complex, analytical studies pro-
vided the epidemiological perspective m terms of population-attributable risks,
the descriptive epidemtological analyses provided a potent basis for hypoth-
esis formulation and a general understanding of the epidemiology.
The whole process of studymg the epidemiology was facilitated by the avail-
ability of a multi-dtsciplmary team, with pathology and eptdemiology as the
most important disciplines in the initial stages. An experience in neuro-
170 Wilesml th

pathology was naturally of importance and in the case of eptdemtology the
avatlabthty of a team capable of handhng, vahdatmg, and analyzing large sets
of data was probably crucial, m addition to the general experience m dealing
with national veterinary epidemtologtcal problems.
In the vetermary field, eptdemtologists have the advantage of utdtzmg the
results of experimental studies m the species of interest, a benefit not available
to human epidemtologtsts. Interestingly, because of the protracted mcubatton
period, laboratory-based studies were not of tmmedtate help m formulatmg
hypotheses of causality A number of these are now commg to frumon and will
be of enormous interest m attempting to piece together some of the more com-
plex parts of this epidemiological Jig-saw, but it seemsthat none will alter the
general theses An exception 1s perhaps the evidence of an absence of any
genetic component m the susceptibility of cattle to the BSE agent The results
of the initial epidemiologtcal studies dtd not indicate that this was a significant
factor, but molecular genetic studies (3 7-39), the results of the mtttal parenteral
exposure of calves using bram homogenates, and a population approach using
statistical models of possible modes of inheritance of susceptibility confirmed
that the genotype of cattle was not a risk factor that required constderatton m
the design of eptdemtologtcal studies. This 1s m contrast to studies of sheep
scrapie. If the Judgement m the early stages that this was not a factor had been
wrong, then a number of the analytical studies would have been flawed. In the
outcome, this did not prove to be a problem and, m fact, the study of the epide-
miology of BSE has been relatively simple without such confoundmg factors.
As indicated m Section 5.2., all researchers have resorted to the published
literature to formulate then ideas, but there had been few, tf any, rigorous stud-
tes of the epidemtology of naturally occurrmg TSEs m animals, notably on
sheep scrapie. This statement is perhaps a little harsh because since the heyday
of studies of natural sheep scrapie eptdemtologtcal methodology has devel-
oped, and molecular genetic studies of sheep have provided a stimulatmg, addt-
ttonal tool for epidemtologtcal studies that can take account of all known risk
factors. This said, there has to be some concern m taking account of, and
extrapolating from, the results of experimental studies m laboratory animals,
notably inbred strains of mice and hamsters, neither of which are naturally
susceptible species. This 1sa somewhat controversial point, but there has to be
some doubt, and therefore word of cautton, in using the results of such studies,
especially with respect to the spectes-barrier phenomenon. An element ofjudg-
ment is required m assessingthis polemical view, but the translation of results
from the studies m laboratory animals to naturally susceptible species will
always be difficult and contenttous.
Finally, but of some considerable importance, the early identification of BSE
and the subsequent mvestigation of the epidemtology of BSE was only
BSE Analysis in the UK 171

achieved by the presence and use of the infrastructure of the State Veterinary
Service in Great Bntam. The inherent surveillance network identified the dls-
ease at the earliest opportunity, and the national structure of the veterinary field
service facilitated both the continued monitoring of the epidemic and assisted
in obtaining the necessary data for studies. Epidemiologlcal studies are heavily
dependent on the provision of valid data, and the contrlbutlon made by field
staff of the State Veterinary Service m providing this has been exemplary.
References
1 Rothman, K J (1986) Modern Epzdemzologv Little, Brown and Co Boston/
Toronto, pp. 7-2 1
2 Wells, G A H , Scott, A C , Johnson, C T , Gunmng, R F , Hancock, R D ,
Jeffrey, M , and Bradley, R (1987) A novel progressive sponglform encepha-
lopathy m cattle. I/et Ret 121,419,420.
3. Dawson, M , Mansley, L M., Hunter, A R , Stack, M. J., and Scott, A C (1987)
Comparison of scrapie associated fibril detectlon and histology m the diagnosis of
natural sheep scrapie. Vet Ret 121,591.
4 Wells, G A H , Hawkins, S A C , Hadlow, W J and Spencer, Y I (1992) The
dlscovety of bovine sponglform encephalopathy and observations on the vacuolar
changes, m Prton Diseases zn Humans andAnimals (Prusmer, S B., Collmge, J ,
Powell, J , and Anderton, B , eds ), Elhs Horwood, London, pp 256-274
5. WIlesmIth, J W and Wells, G A H (1991) Bovine sponglform encephalopathy,
m Transmlsslble Sponglform Encephalopathles Current Topics zn Mcroblologv
and Immunology, vol 172 (Chesebro, B W , ed.), Springer-Verlag, Berlm/Heldel-
berg, pp 2 l-38
6 Wells, G. A. H and Wilesmith, J W (1989) The dlstrlbutlon pattern of neuronal
vacuolatlon m bovme sponglform encephalopathy (BSE) IS constant. Neuropathol
Appl Neurobiol 15, 591
7 Wells, G A H , Hancock, R. D., Cooley, W A., Richards, M S , Higgins, R J., and
David, G. P (1989) Bovine spongiform encephalopathy. dlagnostlc significance of
vacuolar changes m selected nuclei of the medulla oblongata. Vet Ret 125,52 l-524
8 Wells, G A. H., Scott, A C , Wilesmith, J W , Simmons, M M , and Matthews,
D (1994) Correlation between the results of a hlstologlcal exammatlon and the
detectlon of abnormal bram flbrlls m the diagnosis of bovme sponglform
encephalopathy Res Vet Set. 56, 346-35 1.
9. Scott, A C., Wells, G A H , Chaplin, M. J , and Dawson, M. (1992) Bovine
sponglform encephalopathy. detection of fibrlls m the central nervous system IS
not affected by autolysls. Res Vet Ser. 52,332-336.
10 Wells, G A H , Dawson, M , Hawkins, S. A C , Green, R B , Dexter, I , Francis,
M. E., Simmons, M. M., Austin, A. R., and Hongan, M. W (1994) Infectlvlty m
the Ileum of cattle challenged orally with bovine spongiform encephalopathy Vet.
Rec. 135,40,41.
11 Hill, A B. (1965) The environment and disease. association or causation? Proc
R Sot Med 58.295-300
172 Wilesmith
12 Wilesmith, J W (1993) BSE eptdemtologtcal approaches, trials and trtbulattons
Prev Vet Med 18,33-42
13 Wilesmith, J W , Wells, Cl. A H , Cranwell, M P , and Ryan, J B M (1988)
Bovme spongiformencephalopathy:epidemiologicalstudtes.Vet Ret 123,63%644
14 Wtlesmith, J W , Ryan, J B M , Hueston, W D , and Homville, L J (1991)
Bovine spongiform encephalopathy. descriptive eptdemtological features 1985%
1990 Vet Ret 130,9&94
15 Wtlesmtth, J W (1991) Eptdemiology of bovine spongtform encephalopathy
Semm Vu-o1 2,239-245
16 Wtlesmtth, J W (1994) Bovine spongtform encephalopathy and related dtseases
an eptdemtologtcal overview N Z Vet J 42, l-8
17 Jeffrey, M and Wells, G. A H. (1988) Spongiform encephalopathy m a nyala
(Tragelaphus angusl) Vet Path01 25,398,399
18 Brookmeyer, R and Gail, M H (1988) A methodfor determmmgshort-termpredlc-
ttonsand lower bounds the sizeof the AIDS epidemicJ Am Stat Sot 83,30 l-308
on
19 Wooldrtdge, M J A (1995) A study of the incubation period, or age at onset, of
the transmissiblespongiform encephalopathies/prtondiseasesPh D Thesis, Urn-
verstty of London
20 Wilesmith, J W , Ryan, J B M , and Atkinson, M J (1991) Bovme spongiform
encephalopathy eptdemtological studieson the ortgm Vet Ret 128, 199-203
21 Schlesselman, J J (1982) Case-Control Studies Design, Conduct, Analysis
Oxford Umversity Press,New York.
22. Homvtlle, L J , Wilesmtth, J W , and Richards, M. S (1995) An investigation of
risk factors for casesof bovine spongtform encephalopathy born after the mtro-
duction of the “feed ban” Vet Ret 136,3 12-3 18
23 Brown, P (1994) Vertical transmission of priori diseases Hum Reprod 9,
1796,1797
24 Homvtlle, L J (1995) A revtew of the epidemiology of scrapiem sheep.Rev Scl
Tech Off Int Ep1.z (submitted for publication)
25 Wilesmith, J W. (1992) Bovine spongiformencephalopathy.a brief epidemiography,
1985-1991, m Prlon Dzseases ofHumans and Ammals (Prusmer, S B., Collmge,

<<

. 25
( 45 .)



>>