9.11 The transmissible spongiform encephalopathies
Oxford Textbook of Public Health
The transmissible spongiform encephalopathies
Stephen Palmer, Charles Hillier, and Roland Salmon
Transmissible mink encephalopathy
Chronic wasting disease
Bovine spongiform encephalopathy
Spongiform encephalopathy in exotic ungulates
Feline spongiform encephalopathy
Possible sources of sporadic CJD
Variant Creutzfeldt–Jakob disease
Predicting the epidemic
The public health response to BSE and variant CJD
Human population surveillance of CJD
Risk assessment, risk communication, and public confidence
The transmissible spongiform encephalopathies (TSEs) are a group of invariably fatal central nervous system neurodegenerative diseases affecting both animals and humans, and include scrapie, bovine spongiform encephalopathy, kuru, Creutzfeldt–Jakob disease (CJD), and the recently described variant Creutzfeldt–Jakob disease (vCJD) (Table 1). The term TSE derives from characteristic microscopic appearances of sponge-like vacuolation of neuropil of the brain and the fact that transmission occurs naturally within species and can be induced experimentally between species by parenteral injections of infected tissue and by oral challenge.
Table 1 Animal and human spongiform encephalopathies
Although the oldest known form of TSE, scrapie in sheep, was described as early as 1732 (Pattison 1988), the nature of the agent causing the TSEs remains controversial. Certain characteristics unite the TSEs: they are caused by a highly resistant and unconventional agent, they have extremely long incubation periods, they are progressive and invariably fatal, and they cause degenerative changes in the brain without causing an inflammatory or immunological reaction.
It was traditionally believed that the TSEs were caused by ‘unconventional transmissible agents’, originally ‘slow viruses’, a term coined to emphasize their remarkable resistance to thermal and other physical factors (ultraviolet light, ionizing radiation (Alper et al. 1967), and the majority of chemical sterilization procedures (Dickinson and Taylor 1978)) that would be expected to inactivate conventional micro-organisms. This suggests that if the agents contain nucleic acid, it is either very small or extremely well protected. Treatment with nucleic acid denaturing enzymes does not alter the infectivity of the infective agent (Prusiner 1982), but infectivity is reduced by a number of procedures that modify or hydrolyse proteins. This evidence has been used to suggest that the agent does not contain nucleic acid and that the infectious agent consists solely of a protein. From the above discussion it can be appreciated that two fundamentally different categories of explanation describing the cause of these diseases have emerged. The ‘prion’ hypotheses (Prusiner 1982) (see below) propose that a particular membrane-bound host-coded protein, devoid of nucleic acid, causes disease. In contrast, the virus and the virino hypotheses predict that a yet-to-be-discovered replicating nucleic acid is the obligatory factor for transmission.
In 1982 biochemical purification of the scrapie agent from brain homogenate was achieved and produced evidence that this consisted solely of protein. The term ‘prion’ was introduced and defined as ‘small proteinaceous infectious particles that resist inactivation by procedures which modify nucleic acids’ (Prusiner 1982). Following this, the amino acid sequence of at least part of the protein was deciphered. Molecular techniques revealed that a single cellular gene was responsible for encoding prion protein (PrP), now routinely referred to as the prion protein gene (PRNP) and which is located on chromosome 20. This finding led to the recognition of two forms of the PrP, the normal cellular form, termed PrPC (where C stands for cellular), and a pathological form present only in scrapie-infected organisms, termed PrPSc or PrPRES (where superscript Sc stands for scrapie and superscript RES for resistant form). The amino acid sequence of PrP encoded by the gene of a healthy animal does not appear to differ from that encoded by a scrapie-infected animal. This suggests that the different properties of PrPC and PrPSc are due to post-translational change (Basler et al. 1986), that is, an alteration in the structure and behaviour of the protein after it has been synthesized.
Although the ‘prion’ hypothesis could account for the infectious agent, prions are as yet poorly defined, and ‘for the most part this term is used as an operational term for the transmissible agent’ (Chesebro 1998). Although it has been widely accepted throughout the scientific community, the ‘prion hypothesis’ still fails to explain the large number of different strains of disease seen in a species. Some feel the only explanation can be that the agent must contain an informational molecule (RNS/DNA). The major arguments for and against ‘prion’ versus ‘virus’ are eloquently argued by Chesebro (1998) and recommended to the reader.
Scrapie, affecting sheep and goats, was the first of the TSEs to be described and has been recorded in Europe for more than two centuries (Kimberlin 1981). Although its true prevalence is unknown in the United Kingdom, a rough estimate is 2 cases per 1000 sheep. Only Australia and New Zealand appear to be free of disease (Spongiform Encephalopathy Advisory Committee 1995). Most breeds and both sexes are affected and the age of peak incidence is about 3.5 years. The disease is characterized by rubbing of the poll and buttocks in response to pruritus, nervousness, or aggression, and hypersensitivity to sound or movement (Spongiform Encephalopathy Advisory Committee 1995).
In 1936, Cuille and Chelle formally transmitted scrapie to healthy sheep by intraocular injection of homogenized brain from clinically affected sheep. At about the same time an accidental natural experiment occurred in the United Kingdom through the use of contaminated louping-ill vaccine. The vaccine had been prepared from a suspension of brain, spinal cord, and spleen tissue from sheep and a large number of flocks were vaccinated in 1935. Two years later an epidemic of scrapie occurred in these flocks even though the louping-ill virus in the vaccine had been inactivated by the addition of formalin to the original suspension (Gordon 1946). Subsequent work by Gordon revealed resistance of scrapie agent to formalin and boiling, and showed that it would pass through 410-nm filters. In 1961 Chandler demonstrated that experimental transmission could occur between species when he successfully managed to produce a spongiform encephalopathy in mice by inoculating them with scrapie brain material, thus producing an inexpensive experimental model (Chandler 1961).
Transmissible mink encephalopathy
This disease, rare in ranch-raised mink, was first identified in Wisconsin in 1947 (Hartsough and Burger 1965), since when there have been 23 outbreaks worldwide (McKenzie et al. 1996). It was originally assumed that these sporadic outbreaks were attributable to the feeding of scrapie-infected sheep offal or carcass material, but the most recent outbreak in 1985 has raised the possibility that a previously unrecognized bovine agent might be involved (Marsh et al. 1991).
Chronic wasting disease
This disease of deer (family Cervidae) was first described in captive mule deer (Williams and Young 1980) in 1980 and then later in Rocky mountain elk (Williams and Young 1982). Cases have occurred primarily in captive animals but a few affected free-ranging animals have been identified in northeastern Colorado and southeastern Wyoming. A prevalence of 2.5 per cent has been estimated in wild deer in that area (Johnson and Gibbs 1998).
Bovine spongiform encephalopathy
Bovine spongiform encephalopathy (BSE) was first described in November 1986 in the United Kingdom (Wells et al. 1987), but it is believed that a few cases occurred as early as April 1985. A large-scale epidemic has since occurred in the United Kingdom, peaking in January 1993, and involving at least 186 000 cattle (Fig. 1). Since then the incidence has steadily fallen as predicted at the time of the feed ban (Spongiform Encephalopathy Advisory Committee 1995; Collee and Bradley 1997; Pattison 1998), although it is estimated that perhaps half a million infected animals entered the human food chain before control measures were introduced in November 1989, and a further 283 000 entered the food chain from December 1989 to December 1995. The disease has been reported in other countries, including Switzerland, France, Oman, the Falkland Islands, Denmark, Germany, Portugal, and Canada. In some of these cases affected cattle were exported from the United Kingdom (Oman, the Falkland Islands), but in other countries, notably Switzerland and France, no firm association with United Kingdom exports can be established (Bradley and Wilesmith 1993; Schreuder 1994).
Fig. 1 Confirmed cases of BSE plotted by month and year of clinical onset.
BSE affects adult cattle of both sexes. The age of peak incidence and median incubation period is 4 to 5 years. Onset of disease is insidious and signs include changes in sensation with hyperaesthesia to touch or sound, excessive nose licking and teeth grinding, apprehension, frenzy, and nervousness of doorways. Abnormalities of posture with dropped head, arched back, splayed, stiff hind limbs, and ataxia of the hind limbs are all regularly described in affected animals. There are general signs such as loss of condition and weight and reduced milk yield in dairy cows. The duration of illness can be as short as 2 weeks but can last up to 1 year, with the average clinical course being several months.
Whether BSE originated as scrapie introduced into cattle, or whether BSE has always existed in cattle, has not definitely been established, but a series of case–control studies suggested that the epidemic spread of the disease occurred through the use of proprietary cattle dietary protein supplements containing contaminated meat and bonemeal prepared from abattoir waste (Wilesmith et al. 1988). It has been suggested, although not proved, that changes in the rendering industry in the late 1970s, which had led to decreased processing times, and a change from batch to continuous methods of processing, together with the abandonment of the use of organic solvents for fat extraction, allowed the introduction of the agent into feedstuff and its epidemic amplification by recycling infected bovine material (Pattison 1998). From computer simulation of the early incidence of the epidemic and analysis of the year of birth of cases, it was estimated that the date of the first exposures were in 1980 or 1981.
BSE has been transmitted experimentally to mouse, sheep, goat, and mink by the oral route and is probably responsible for spongiform encephalopathy in the exotic ungulates (see below), domestic cats, and humans (see below) (Bruce et al. 1997). To date only one BSE strain has been recognized after experimental transmission to sheep, and this was the strain identified in nyala, kudu, the domestic cat, and humans.
Spongiform encephalopathy in exotic ungulates
Since 1986, spongiform encephalopathy has been diagnosed in captive wild animals of six species (nyala, eland, greater kudu, Arabian oryx and scimitar horned oryx, cheetah, and puma) from eight zoological institutions in the United Kingdom. Temporal and geographical features coincided with the BSE epidemic, and animals were exposed to proprietary feeds containing ruminant-derived protein or consumption of cattle carcasses unfit for human consumption, and because brain homogenate induces BSE-like changes in mice, it is likely that BSE is responsible (Kirkwood and Cunningham 1994).
Feline spongiform encephalopathy
From 1990 to May 1998, 81 cases in domestic cats were reported over a wide geographical area in the United Kingdom (Pearson et al. 1993). It is estimated that the disease was probably under-reported before 1994, when feline spongiform encephalopathy was made a notifiable disease. Since that time the number of new cases has declined so that there were only four reported cases in the United Kingdom during 1998 and only one of these would have been born after the specified bovine offal ban in pet food in 1990. It is assumed that cats became infected through consumption of commercially produced cat feed that contained the BSE agent (Pattison 1998).
In 1957, Gajdusek and Zigas described kuru, an unusual progressive neurological disorder in the Fore tribe in the eastern highlands of Papua New Guinea (Gajdusek and Zigas 1957). ‘Kuru’ means shivering or trembling, and the disease was characterized clinically by cerebellar ataxia, tremor, dysathria, dysphagia, and finally dementia, progressing to death within one year of onset. Kuru is characterized pathologically by cerebellar atrophy, with spongiform change and neuronal cell loss evident throughout the central nervous system. Seventy per cent of brains affected contain characteristic plaques comprising a central eosinophilic core with a peripheral halo of radiating filaments, particularly in the cerebellum (Ironside 1996).
Early observation of the tribe revealed that the disease affected mainly the women, adolescent children of both sexes, and children, whilst the men seemed immune to the disease (Gajdusek 1977). It appears that the disease was spread by ritualistic endocannibalism; the women and children would prepare the bodies of tribespeople who had died, including removing the brain. They would either contaminate themselves by inoculation injury through skin lacerations or rubbing conjunctiva, or by eating brain tissue. The men would confine themselves to consuming skeletal muscle of the diseased and it is thought this tissue has a much lower infectivity than brain and spinal cord. Since the cessation of these practices, kuru has progressively disappeared, first among children and then adolescents. No one born in a village since cannibalism ceased has ever developed kuru (Prusiner 1993). However, cases are still reported each year, suggesting incubation periods of over 40 years.
In 1920, Creutzfeldt reported the case of a 23-year-old woman who had died of a neurodegenerative illness, and the following year Jakob reported five cases. At least two of the six cases had clinical features suggestive of the entity we now recognize as CJD. The following four forms of CJD are recognized: sporadic, familial, iatrogenic, and variant.
Sporadic CJD, which makes up about 85 per cent of all CJD cases, occurs worldwide with a reported incidence ranging from 0.06 to 1.6 per million of the population, with mean age of death of 65 years (Table 2). The annual reported incidence in the United Kingdom is 0.5 to 1 cases per million population. Most patients present with a rapidly progressive dementia with a mean clinical course of 4.5 months (Brown et al. 1994a), although up to 10 per cent may survive 2 years or more (Brown et al. 1994a). A wide range of neurological abnormalities has been described (de Silva 1996), but the commonest features include a subacute dementia with pyramidal and extrapyramidal signs. Myoclonus is present at some stage of the disease in up to 90 per cent. Electroencephalogram recordings show a characteristic periodic discharge in 60 to 80 per cent of cases.
Table 2 National and regional CJD surveys
Macroscopic examination of the brain in sporadic CJD may reveal no significant abnormality, particularly in cases with a short clinical history, but in most cases there is evidence of cortical atrophy. The most characteristic abnormality is spongiform change in the cerebral cortex, which consists of a fine vacuolation of the neuropil of the grey matter. The severity and distribution of this varies widely from case to case. Immunocytochemical staining for detection of disease-related protein may demonstrate plaques of abnormal protein similar in pattern to cases of kuru. Clinical (WHO 1998; Belay 1999) and pathological diagnostic criteria (Budka et al. 1995; Kretzschmar et al. 1996) for sporadic CJD have been established.
Iatrogenic CJD, making up less than 1 per cent of all cases, is spread from human to human via contamination of neurosurgical instruments, and by tissue grafts (such as the cornea) from infected donors; the latter have been reported since the mid-1970s (Duffy et al. 1974; Bernoulli et al. 1977). From the mid-1980s increasing numbers of cases associated with the use of infected cadaveric dura mater grafts were identified (Brown et al. 1992), but the most dramatic example of iatrogenic CJD is the continuing epidemic of cases in young adults treated for hypopituitarism with pooled growth-hormone extract.
The clinical and pathological picture of iatrogenic cases appears to depend on the route of inoculation and also the genotype. In those cases where inoculation occurs directly into the brain (i.e. via contaminated dural grafts, surgical instruments) the incubation period is measured in months and clinically there is dementia and pathology similar to that of sporadic CJD. In the cases of the growth hormone recipients who received peripheral injections (i.e. outside the central nervous system), the incubation time is measured in years. Cerebellar signs are prominent and early, and dementia appears only later. Susceptibility is influenced by whether the individual is homozygous or heterozygous for the amino acids methianine or valine at the codon position 129 of the prion protein gene (a common human PRNP polymorphism).
Approximately 10 to 15 per cent of CJD cases are familial and are due to specific point mutations, deletions or insertions affecting the PRNP located on chromosome 20. More than 20 different mutations in the PRNP have been discovered. Patients with sporadic CJD do not have a mutation deletion or insertion of this gene.
Familial CJD encompasses a wide range of clinical phenotypes, some with atypical features including dysautonomia and isolated spastic paraparesis. In general, familial CJD has an earlier age of onset and a longer duration than sporadic CJD.
Gerstmann–Straussler–Scheinker disease, first described by Gerstmann in 1936, contributes less than 2 per cent of all known human TSEs (Masters et al. 1981). Most cases are familial, exhibiting an autosomal dominant pattern of inheritance with almost complete penetrance. The clinical duration is from 2 to 10 years, much longer than for CJD.
Linkage studies suggested that substitution of leucine for proline at PrP codon 102 in the PRNP leads to the development of clinical disease in some families with Gerstmann–Straussler–Scheinker disease (Hsaio et al. 1989).
Fatal familial insomnia was first described in 1986 (Lugaresi et al. 1986), and is a dominantly inherited condition typified by a thalamic dementia with some pathological features of CJD in which insomnia, rather than dementia, is the dominant clinical feature. Over 21 affected families have been recognized worldwide (Brown et al. 1998; Budka et al. 1998; Gambetti and Lugaresi 1998; Julien et al. 1998; Kretzschmar et al. 1998; Pocchiari et al. 1998; Will et al. 1998b). The families have a PRNP mutation at codon 178 resulting in the substitution of the amino acid asparagine for aspartic acid, which when associated with the amino acid methionine at codon position 129 (a common human PRNP polymorphism) leads to this unique clinical phenotype (Medori et al. 1992).
Fatal sporadic insomnia was first described in 1999, in a total of six subjects (Parchi et al. 1998; Mastrianni et al. 1999). All subjects had a disease clinically similar and histopathologically virtually identical to fatal familial insomnia, but none had the PRNP mutation at codon 178. All were homozygous for the amino acid methionine at the codon 129 position of the PRNP.
Possible sources of sporadic CJD
The most commonly cited hypothesis for the cause of sporadic CJD is a spontaneous conformational change in PrP, in the absence of exogenous factors. However, there is evidence that exogenous environmental factors may be important, an observation which is relevant to considering the aetiology of vCJD (Hillier et al. 2000). One observation difficult to reconcile to the prevailing hypothesis is the existence of clusters of cases of ‘sporadic’ CJD, even though some clustering may occur by chance, and that ascertainment of a first case may lead to better ascertainment of further cases (Raubertas et al. 1991). At least eight have been reported (Matthews 1975; Mayer et al. 1977; Will and Matthews 1982; Araya et al. 1983; Lechi et al. 1983). Matthews described three clusters in England. The first was of three cases in a small rural community in the Midlands over 7 years. The second involved five cases with onsets between 1969 and 1973 in four towns/villages in eastern England (separated by no more than 20 km), of whom at least three had lived locally since 1960. The third cluster involved two patients in 1965 and 1968 who lived within 250 m of each other and shared the same general practitioner. In 1980, a patient who lived mid-way between the two previous patients and within sight of both houses died of pathologically proven CJD. In Italy 11 cases of CJD, including five farmers, occurred in and around Parma, Italy, between 1975 and 1979. In Chile, four cases of histologically proven CJD in three farmers and one housewife occurred in 1982 and 1983. The four lived in small rural communities situated within a 20-km radius of the town. In addition to these clusters case-to-case transmission has been suggested in six reports.
CJD can be transmitted from person to person by injecting or implanting CJD-infected instruments, tissue extracts, and organs (Brown 1988, 1995; Creange et al. 1995; Ricketts et al. 1997). Several points relevant to a wider understanding of the epidemiology arise. Firstly, the time from inoculation to disease onset is dependent on the route of entry. Two types of cases, those involving implantation of cerebral electrodes and corneal grafts, show the shortest incubation, of about 16 months. The longest incubation time is 30 years, following subcutaneous and intramuscular injection of growth-hormone extract and of gonadotrophins. Secondly, the route of transmission predicts the clinical appearance. Those infected directly into the central nervous system develop a global dementia, whereas those infected peripherally develop a progressive cerebellar syndrome. The genotype of the prion gene does not explain this difference. That the peripheral route of infection presents as a cerebellar syndrome would support the view that vCJD is acquired peripherally by diet, for example, as cerebellar signs are an early prominent feature of vCJD but not of sporadic CJD.
The infectivity of blood for TSEs in animals is well described (Brown 1995; Ricketts et al. 1997). Nevertheless, attempts to transmit CJD by infusing units of blood from patients with CJD to the chimpanzee have failed (Brown et al. 1994a). Although experimentally difficult, transmission of sporadic CJD and human growth-hormone related iatrogenic CJD to mouse, hamster, and guinea pig, by buffy coat and whole blood, has been demonstrated on four separate occasions (Brown 1995), but only after intracerebral injection. This suggests that transmission from human to human, via blood and blood products, is difficult but theoretically possible. A number of case reports have suggested that transmission by the transfusion of blood or a blood product has occurred in humans (Klein and Dumble 1993; Patry 1998).
One case–control study has examined blood transfusion in sporadic CJD. Esmonde (1993) identified 202 definite and probable United Kingdom cases of sporadic CJD between 1980–1984 and 1990–1992. Twenty-one patients had received blood and 29 had donated blood. This frequency of blood transfusion or donation did not differ from that in age- and sex-matched controls. The clinical features in patients with a history of blood transfusion were similar to those with a history of classical sporadic CJD, and clearly distinct from CJD in recipients of human growth hormone. However, since sporadic CJD is a rare disease and the number of potentially infected units of blood is small, any difference in the rate of receipt of transfusions between cases and controls will be small and a very large study would be needed to detect such a small difference.
Several case reports and series have reported an apparent increase in the prevalence of CJD in populations of people who eat the brains of sheep or wild animals. There has been one case report of definite CJD in a Dutch man who ate occasional sheep brain and an annual feast of hog brain (Alter et al. 1975), and a case series of four patients from North America who all ate squirrel and other animal brains, including goat (Kamin and Patten 1984). A more recent case series obtained a history of eating squirrel brains in all five patients with probable or definite CJD seen over 3.5 years in a neurocognitive clinic in Western Kentucky (Berger et al. 1997). However, there are two reports of CJD in vegetarians. Matthews and Will (1981) described a case in a 62-year-old life-long vegetarian. Although she had never eaten meat, meat products, or offal, she had handled raw meat while preparing food for her husband. A study of seven patients from India (Singhal and Dastur 1983) revealed that two cases had been vegetarians but no further details are given.
CJD and scrapie exist together in many countries, but the incidence of CJD in countries where there is no scrapie, such as Australia, New Zealand, and Chile (Table 2), is no different from that in scrapie-endemic countries. In France, a detailed study looking at regional distribution of CJD and scrapie-infected flocks found there to be no relationship (Chatelain et al. 1981). The recent case of a 60-year-old Italian man and his cat (Zanusso et al. 1998) is of interest. The man developed CJD in November 1993 and died 3 months later. His cat developed a neurological disease at the same time and was killed at the time of her owner’s death. Strain-typing suggested that the man was a case of sporadic CJD. The cat had the same strain, distinct from the feline spongiform encephalopathy strain, described in the United Kingdom in the early 1990s (Pearson et al. 1993). Zoonotic transmission would require that the cat had infected her owner whilst presymptomatic. A common source for both cat and man would seem more probable. The circumstances would be hard to explain without an exogenous source of disease. Matthews et al. (1979) described definite or probable CJD in four patients, all of whom kept ferrets for pets. One of the cases, who had also kept ferrets 30 years previously, had definitely been bitten by a ferret 2 years before disease onset. After his death the ferret was killed and examined histologically, but no spongiform change was found. In addition, the brain of the ferret was inoculated into four species of monkey and one cat, all of whom remained well. No such studies were performed with the other cases.
Case reports and series focus on two main occupational groups; health professionals and those who are exposed to farm animals, in particular sheep and cattle. There are at least 26 reports of sporadic CJD in health care workers worldwide (Maytenyi 1991; Berger and Noble 1993; Weber et al. 1993; Bobowick et al. 1973). These include seven physicians, including a neurologist, the head of an intensive-care unit, two neurosurgeons, an orthopaedic surgeon, and a pathologist, four dentists, nine nurses, three nursing assistants, and two histopathology technicians.
A wide range of other occupations have been reported among cases of sporadic CJD (Brown et al. 1979; Masters et al. 1979). Several case series quote an excess number of farmers and farmers’ wives (Kovanen and Haltia 1988; Masullo et al. 1988). In an Italian study (Masullo et al. 1988), the incidence was three times the expected. An analysis of epidemiological surveillance data in the United Kingdom from 1970 to 1996 (Cousens et al. 1997a) revealed a statistically significant excess of cases among dairy-farm workers and their spouses, and among people with greater degrees of contact with live cattle infected with BSE. No such excess was found in abattoir workers, butchers, or meat cutters.
To date there have been eight case–control studies of sporadic CJD in humans (Table 3). Such studies are particularly difficult to perform for two main reasons, the necessity of using proxy respondents for cases, and the long incubation period and therefore long recall period. The quality of these studies has been criticized particularly for control selection. Respondent-nominated controls, such as used in the first American study, may have lead to an overrepresentation of controls with similar occupations and social backgrounds (Masters et al. 1979). Selecting controls from relatives may obscure inherited risk factors for disease. Selecting from hospitals has the potential to mask associations with other medical and health care procedures, through overmatching, even though individuals with mental or neurological disease were generally excluded.
Table 3 Case–control methods
In one of the most important England and Wales studies in 1980 to 1984 (Harries-Jones et al. 1988), two control groups were selected, hospital patients with neurological disease (one group) and those with non-neurological disease (the other). Suitable controls were selected by questioning ward staff and accepting the first offered patient. The selection criteria used by staff were not described. In the 1993 to 1995 European study (van Duijn et al. 1998) the researchers simply state that control participants were recruited from the hospital where the patient who had sporadic CJD had been diagnosed and, apart from the exclusion of patients with dementia, no further information is given on selection. On stronger ground was the Australian study, where three community controls for every case were selected at random by using a random-dialling telephone survey, stratified by age, sex, and urban or rural residence.
Information bias is a major potential threat to the validity of these case–control studies since a spouse or close relative has to provide data for cases but, in many studies, the controls respond on their own behalf.
Despite these deficiencies, and given that the studies span Europe, Asia, Australia, and North America, it is of note that similar categories of risk factors do emerge: consumption of undercooked meat and offal products, contacts with various small biting animals, and working in, or receiving health care, notably surgical procedures (Table 4). It is surprising that the United Kingdom government’s Spongiform Encephalopathy Advisory Committee (SEAC), when assessing the zoonotic potential of BSE, did not apparently pay much attention to these studies (Hillier and Salmon 2000). Significant associations with offal have consistently been dismissed as statistical artefacts.
Table 4 Significant exposures from case–control studies
As part of the National Surveillance Unit’s work, a rolling case–control study has been ongoing since 1990 (National CJD Surveillance Unit 1995, 1997, 1998). Relatives of cases are interviewed by a clinician from the unit and a control is chosen who was a patient at the same hospital matched for sex and aged within 4 years of the case. An attempt is made to interview a relative of the control rather than the control him- or herself. However, use of hospital controls can lead to overmatching of exposures related to hospital admission. Selection of controls by ward staff could also introduce bias. Time available for interview is short and there is no attempt to validate histories of exposure by checking with other friends or relatives and by a systematic confirmation of exposure histories such as following up reported sources of foods eaten.
A total of 206 case–control pairs of sporadic CJD between 1990 and April 1997 were analysed (National CJD Surveillance Unit 1998). Significant associations were found with lifetime consumption of beef, venison, and veal. Also, dose–response relationships were found for beef, venison, and veal consumption. A significant association has been found for consumption of brain as well, with a strong dose–response, but not for other animal products.
Interpretation of these associations is particularly difficult since the study design means that data are particularly subject to bias arising from the fact that relatives of cases know both the diagnosis and publicity about the probable source of infection. To assess this bias researchers exploited the fact that relatives of cases were usually interviewed before the death of the case and therefore before pathological confirmation. Dietary histories for confirmed cases were compared with histories from patients who, though suspected initially to have CJD, were not subsequently confirmed. A comparison was made of 206 cases and 80 ‘non-cases’. The dose–response association for beef completely disappeared. For brain consumption, the odds ratios for eating it less than once a year, and for more than once a year, were raised (1.6 and 1.4) but the trend was not statistically significant (p = 0.37). These data suggest that the significant associations found in the case–control study are strongly influenced by information bias.
Variant Creutzfeldt–Jakob disease
In April 1996, the National CJD Surveillance Unit reported 10 cases of CJD in young people in the United Kingdom. It had a distinct clinical appearance characterized by an extended illness duration, a younger age at death, and predominantly psychiatric symptoms in the early stages including withdrawal, delusions, hallucinations, aggression, and depression (Zeidler et al. 1997a,b; National CJD Surveillance Unit 1998; Will et al. 1999, 2000). Neurological indications included cerebellar signs, dementia, involuntary movements, myoclonus, akinetic mutism, chorea, upgaze palsy, and dystonia. Electroencephalography changes characteristic of sporadic CJD with generalized bi/triphasic periodic complex (Brown et al. 1986) have not been reported in vCJD. The peak age of onset or sporadic CJD lies between 60 and 70 years of age, and in Britain only four cases of CJD had occurred in people under 35 between 1970 and the end of 1989. In contrast, six of the first ten cases of vCJD were younger than 30. Only fourteen cases of CJD under 30 years had ever previously been reported in the world (Will et al. 1996). Cases had no family history of prion disease and had no exposure to known transmitted CJD.
In addition to the United Kingdom cases of vCJD, one case occurred in a 26-year-old Frenchman who may have acquired infection from injecting the anabolic steroid bovine growth hormone somatotrophin, as an aid to body-building, prior to the July 1992 ban on its use.
The most striking and consistent neuropathological feature that helps separate vCJD from sporadic CJD is the presence of numerous large ‘florid’ amyloid plaques surrounded by spongiform change extensively distributed throughout the cerebral and cerebellar cortex, with smaller numbers in the basal ganglia, thalamus, and hypothalamus. Immunocytochemical staining for PrP shows strong staining of these plaque-like lesions. However, these lesions are not pathognomonic of this disorder (Ironside and Bell 1997) and have been occasionally reported in cases of iatrogenic CJD (Takashima et al. 1997).
Lymphoreticular tissues (68 tonsils, 64 spleens, and 40 lymph nodes) were obtained at autopsy from patients with ‘prion’ diseases, and from neurological and non-neurological controls, as well as tonsil samples from patients with suspected ‘prion’ disease. In lymphoreticular tissue, PrPSc was detected only in samples from individuals with confirmed vCJD. It was also detected from tonsil biopsy samples from all eight patients whose disease turned out to be vCJD. It was not detected in samples from patients with other forms of prion diseases, including iatrogenic and familial CJD (Hill et al. 1999). Tonsillar biopsy appears to be both an extremely specific and sensitive test for the antemortem diagnosis of vCJD and so far represents the only means of diagnosing vCJD before death without brain biopsy (Petersen 1999).
Prion strains can be typed using Western-blot protein analysis and strain typing links vCJD to BSE. Patterns of banding of protein fragments of PrPSc are determined by the degree of protein glycosylation. In humans four patterns have been recognized, designated Types 1–4. Most sporadic CJD is Type 2, most iatrogenic CJD is Type 3, but all vCJD is Type 4, the same pattern seen in BSE-infected cattle and in mice and macaques experimentally infected with BSE.
Unlike familial CJD, in which more than 20 different coding mutations have been found to be associated with the development of disease, no pathogenic mutations have been detected in the PRNP in cases of vCJD. However, in addition to the pathogenic mutations, there is a common polymorphism at codon 129 of the PRNP that results in the substitution of valine (V) for methionine (M) (Goldfarb et al. 1989). The codon 129 genotype is known to influence susceptibility to both sporadic CJD (Palmer et al. 1991) and iatrogenic CJD (Collinge et al. 1991; Brown et al. 1994b), and the disease phenotype of familial CJD can also be modulated by the codon 129 polymorphism (Goldfarb et al. 1992). All cases of vCJD so far studied have been homozygous for methionine (MM) at codon 129, which occurs in only around 37 per cent of normal controls and 79 per cent of sporadic cases of CJD. Interestingly, cattle are uniformly homozygous for methionine at codon 129 in the PRNP (Schätzl et al. 1997). If, as has been suggested, the ability to transmit disease from one species to another depends on the likeness of the infecting ‘prion’ to the hosts ‘prion’, then the MM genotype in humans might confer an increased susceptibility to bovine ‘prions’. However, it would be premature to conclude that individuals who possess the codon 129 genotypes MV or VV cannot develop vCJD (Hill et al. 1997).
Up to October 2000 there have been 83 confirmed cases of vCJD in the United Kingdom (Table 5), one case from the Republic of Ireland, and two cases reported from France. Detailed demographic data have been made available on 52 of the vCJD cases that have occurred in the United Kingdom (National CJD Surveillance Unit 1999). Of the 52 cases, 28 were women. The median age at onset of disease was 28 years and the median age at death 29 years (compared with 65 years for the median age of death for sporadic CJD). The median duration of illness was 14 months (range 7 to 38) which is long when compared with 4.5 months in sporadic CJD. The median delay between onset of disease and confirmation of vCJD was 15 months (range 7 to 32 months). This has not decreased over time. Since the first case of vCJD, with onset in early 1994, there is now some evidence that the incidence of vCJD is increasing.
Table 5 Monthly Creutzfeldt–Jakob disease statistics
Figure 2 shows the geographical distribution by place of residence at onset, of the 52 cases of vCJD with onset in the United Kingdom. Comparison of vCJD incidence rates in four northernmost regions (northwest, Yorkshire and Humberside, northern, Scotland) with those further south (southwest, southeast, Wales, West Midlands, East Midlands, East Anglia) according to place of residence in 1991 suggests that the incidence in the ‘North’ of the United Kingdom is twice that in the ‘South’. There is no explanation as yet for this observation.
Fig. 2 The geographical distribution, by place of residence at onset, of the 52 cases of vCJD with onset in the United Kingdom.
Initial concern over a cluster of four cases in east Kent led to speculation that there was a causal relationship between vCJD and a cattle-rendering plant in the vicinity (Chandrakumar 1998), but detailed investigation has suggested that the occurrence of four or more cases in a population of 1.5 million (approximately the population of east Kent) is not unexpected (Cousens et al. 1999). More recently, concern has been raised over a cluster of five cases in Leicestershire that are currently under investigation. Analysis of 48 vCJD cases by place of residence in 1991, as the likely time of peak exposure, provided no evidence of local clustering.
Cases of vCJD have not been the subject of intensive epidemiological field investigation, as might have been expected of a new disease cluster, but cases have been enrolled in the ongoing case–control study run by the CJD Surveillance Unit (National CJD Surveillance Unit 1996). Up until 1997, only hospital-based controls were recruited, but since then a decision has been made to change the study design to recruit four community controls per case, selected from general medical practices. The most recent data (National CJD Surveillance Unit 1999) present 51 cases but no community controls. Cases have been compared with 27 non-matched ‘non-cases’. Cases were reported to have eaten beef more frequently than controls, but not significantly so. Interestingly, 51 per cent of the vCJD cases were reported to have eaten burgers, meat pies, and sausages likely to have contained mechanically recovered meat (MRM), compared to 41 per cent of controls. However, even if cases really have had a 10 per cent greater exposure to MRM, a case–control study would need to have 254 cases and 1016 controls to have an 80 per cent power to detect an increased odds ratio of 1.5 at the 95 per cent level of significance, a result reflecting the high exposure rate of the normal population to MRM.
Predicting the epidemic
Several attempts have been made to predict the size of the eventual vCJD epidemic. The first of these was by Cousens (1997b) based on the first 14 reported cases. In any calculation of this sort several fundamental assumptions have to be made. In the case of Cousens’ predictions these were as follows.
vCJD is due to exposure to the BSE agent.
Until 1989 the number of people newly infected with the BSE agent each year was proportional to the number of cases of BSE in cattle with onset in that year.
Individuals infected with the BSE agent develop vCJD after a long and variable incubation period (mean incubation periods of 10, 15, 20, and 25 years were used for these calculations).
There were few (< 5) or no cases of vCJD before 1994.
Cousens used two sets of predictions for the number of BSE cases. The first set assumes that the degree of under-reporting of BSE changed little over time. The second set assumes that under-reporting was greatest early in the epidemic and reduced after BSE was made a notifiable disease.
Three models for predicting epidemic size were used, based on the distribution of incubation periods. These were lognormal, gamma, and Weibull. Assuming that there were fewer than five cases with onset before 1994, the Weibull model does not provide a good approximation to the incubation-period distribution.
Using the lognormal incubation-period distribution model the least number of cases expected would be 75 and the worst scenario would be 80 000 cases. Using the gamma incubation-period distribution model, the ‘best’ scenario would be 88 and the worst would be 13 000.
The actual number of confirmed cases with clinical disease onset in 1996, 1997, and 1998 were 10, 10, and 13 respectively (Ghani et al. 1998). These numbers would exclude Cousens’ predicted worst-case scenarios and suggest that, based on the assumptions given above, the final size of the epidemic may well lie in the range of 75–450 cases.
Ghani et al. (1998) constructed a mathematical model to predict the magnitude of the vCJD epidemic based on data up to the end of 1997 (23 cases) and used the following assumptions to calculate a large number (5 million) of different scenarios. The following assumptions were made.
vCJD is due to exposure (orally) to the BSE agent.
Incubation-period distribution of vCJD is unimodal for the prion gene codon 129 methionine homozygous genotype (other genotypes were not considered in their calculations).
There was no under-reporting of vCJD cases prior to 1995.
There has been no human-to-human transmission.
The two epidemiological factors that had the greatest impact on the estimated epidemic size and duration were the average number of humans infected by one infectious bovine and the incubation-period distribution. From their calculations, the smallest epidemic would consist of just zero to 200 cases assuming that, on average, only 0.0001 humans are infected by one maximally infectious bovine. The worst-case scenario would be between 2 million and 10 million cases, assuming that over 100 persons are infected by one maximally infectious bovine. Ghani et al. (1998) address the issue of when it will be possible to narrow the range of possible outcomes. If the incidence of vCJD remains very low for the next 3 to 5 years, a small epidemic becomes more likely since it will become increasingly probable that the epidemic peak has been reached.
Based on the assumptions given above, and on the actual annual incidence of vCJD deaths in 1995 to 1999, Ghani et al. (2000) have reassessed their initial predictions using extensive scenario analyses. From the vCJD incidence in 1999 and the observed stable age structure of cases, they substantially reduce the upper limit of possible cases to 136 000.
A study by Cohen et al. (1999) attempted to predict the start date of the BSE epidemic and the consequent implications on any vCJD epidemic. Cousens has already made predications of the future vCJD epidemic based on the first human exposure to BSE taking place between 1983 and 1985 (Cousens 1997b). Cohen et al. applied an age-cohort model to BSE data between 1985 and 1996 to estimate the earliest possible date of the first unrecognized BSE cases. According to the models, BSE cases may have occurred as early as 1980, which implies a much smaller epidemic than that predicted by Cousens.
The public health response to BSE and variant CJD
The epidemic of BSE in British cattle raised immediate concerns about its zoonotic potential and eventually in 1988 a working party was established under the chairmanship of Sir Richard Southwood (Table 6) (Southwood 1989).
Table 6 BSE: selected chronology of events, 1986–98
The main task for this and subsequent committees was to decide whether the BSE agent would be transmitted to humans and what the possible pathways were. Thus the distribution of the agent in incubating and affected cattle needed to be known, as well as the ways in which products were processed and where they ended up. This proved to be immensely complex, since products of cattle such as gelatin may be used in cosmetic, pharmaceutical, and medical products. Southwood quickly recommended that BSE should be made a notifiable disease and that the practice of feeding ruminant protein to other ruminants should be banned. The order was made on 14 June 1988 but not implemented until 18 July 1988 in order to allow stocks of meat and bonemeal feed to be cleared. It is now clear that enforcement of this ban was inadequate (Patterson and Painter 1997). Cases continued to be seen in cattle born after the ruminant feed ban, and a case–control study suggested that this was due to ongoing feed contamination, possibly through cross-contamination of cattle feed by ruminant protein prepared in the same feed mills as for non-ruminants. (Only in March 1996 was the use of mammalian meat and bonemeal for any farm animal terminated.) Affected cattle were required to be slaughtered and carcasses incinerated or buried. However, until February 1990 compensation to farmers was set at only 50 per cent of the value of the cow, with the consequence that possibly infected animals would be sold for meat before a definite diagnosis was made.
In the United Kingdom in November 1989 certain bovine tissues (termed specified bovine offal (SBO)) believed to be potentially infectious, such as brain, spinal cord, intestines, tonsil, thymus, and spleen from cattle over 6 months of age, were banned from the human food chain. However, once again inadequate enforcement of the ban led to some of this material continuing to enter the food chain, particularly in the form of MRM (Patterson and Painter 1997). MRM from the vertebral column, which could contain fragments of spinal cord, was eventually banned only in December 1995. Therefore it seems probable that the British public was potentially exposed to the agent of BSE in presymptomatic cattle, via foods such as beefburgers and pies which use MRM, right up with the end of 1995.
The recognition of new variant CJD in 1996 led to further stringent control measures. In March 1996 the European Union banned the export of any British beef. In April 1996, in the United Kingdom, cattle over 30 months old were banned from entering the food chain. In September 1996 a prohibition on heads of sheep and goats was introduced, and in December 1997 beef on the bone was banned.
Human population surveillance of CJD
In parallel with the veterinary- and food-control measures introduced by MAFF in the United Kingdom, population surveillance of CJD was reinstated in May 1990 as recommended by the Southwood Committee (Southwood 1989). The United Kingdom government decided to give this responsibility to a new, clinically based unit in Edinburgh rather than use the existing communicable disease-control structures of the Public Health Laboratory Service, Communicable Disease Surveillance Centre or Scottish Centre for Infection and Environmental Health. The initial aim of the project was to identify changes in the pattern of CJD that might be the result of a new agent (BSE) entering the human population. The passive reporting system was based on receiving clinical reports mainly from neurologists followed by in-depth clinical and neuropathological investigation. In 1995 a new variant of CJD was identified; the National CJD Surveillance Centre was alerted by the very young age of cases compared with sporadic CJD and their unusual histopathological appearance.
The National CJD Surveillance Unit has produced detailed annual reports including numbers of suspected cases referred to them, and the clinical outcome. As of October 2000 (www.doh.gov.uk/cjd/stats/), 1261 possible cases of CJD have been referred since 1990, of which 639 have been confirmed. Of these 635 cases, 476 (74 per cent) have been sporadic cases, 36 (6 per cent) iatrogenic, 28 (4 per cent) familial, 15 (2 per cent) GSS, and 83 (13 per cent) vCJD. Trends in mortality since 1970 have been analysed, showing an increase in all age groups, a phenomenon seen in other European countries which is probably the result of improving surveillance and case ascertainment, and reporting.
In order to establish whether vCJD was really a new condition and not due to improved surveillance, a retrospective review of death certificates has been carried out (Majeed et al. 1998). A total of 1485 deaths from selected neurological disorders were identified in people aged 15 to 44 years in England from 1979 to 1996. Clinical records were traced for nearly half of the cases and reviewed. No new cases of sporadic or vCJD were identified, giving a strong measure of confidence in the completion of the surveillance system. A similar study in Wales also found no case of vCJD between 1985 and 1995, supporting the view that vCJD was a new disease entity rather than the result of better ascertainment.
Concern over the emergence of vCJD in the United Kingdom led the United States Centers for Disease Control and Prevention to set up an active CJD surveillance in April–May 1996 in the four established Emerging Infections Program sites and Atlanta (Schonberger 1998). The average annual death rate was stable from 1991 to 1995 at 1.2 per million population. No variant cases were identified and since then ongoing review of United States mortality data of CJD cases in younger adults has failed to identify variant cases.
Risk assessment, risk communication, and public confidence
Major issues of public policy have been highlighted by the BSE crisis (Maxwell 1999). The initial risk assessment of the threat to human health was based erroneously on the evidence that scrapie, though prevalent in the United Kingdom for centuries, had not seemingly been transmitted to humans, and that when experimentally transmitted to other animals the infectivity of scrapie via the oral route was low. The Southwood Report in 1989 concluded that ‘it was most unlikely that BSE would have any implications for human health’. Patterson and Painter (1997) commented that:
In their main report to Parliament in February 1989, the Southwood Committee acknowledged the paucity of evidence available to them. They advised that cattle were likely to prove a ‘dead end host’ for the disease agent and that it was ‘most unlikely that BSE would have any implications for human health’. With the exceptions of baby foods and medicinal products no action was taken to reduce human exposure to high titre bovine tissues (such as brain and spinal cord) from preclinically infected cattle. This was despite prior health warnings about the consumption of BSE infected bovine brain, and concern that the inability to identify BSE infected cattle was a major threat to human disease control and to cattle exports from the UK.
Southwood did say that if his assumptions proved wrong ‘the implications would be extremely serious’. However, Tyrrell, the first chairman of SEAC, set up in 1989, was even more convinced of the lack of risk and wrote that ‘any risk as a result of eating beef or beef products is minute. Thus we believe that there is no scientific reason for not eating British beef and that it can be eaten by everyone’ (Maxwell 1999).
Clearly, events have proved that judgement to be seriously in error and to lack the caution which the state of knowledge at the time required. Lack of evidence of a risk is to be distinguished from evidence of no risk. It is of great significance that up until 1997 the work of SEAC, like other government advisory committees, was not shared in any formal way with the public health profession or the public, contributing to the sense of distrust of MAFF and government advice (Anonymous 1996b). Further, its composition in the critical early years has been criticized for a lack of public-health experts who might have offered a more balanced approach to assessing population risk and the necessity to gain public confidence (Gore 1996). One editorial in the Lancet said:
No sooner had a terse statement been made in the House of Commons … than the lay and professional media swept into action, filling the information gap with variable accuracy. This reaction, condemned by some, was understandable since the scientific evidence that underpinned the announcement and subsequent advice had not been made publicly available … The conclusions they reach and the recommendations they make are made public in short, bald statements that are distributed by the commissioning department. Only then might the data be offered for consideration for publication in a peer-reviewed journal; too late, perhaps, to contradict the baseless bluster that has already been given free space to shape opinion and policy. (Anonymous 1996b)
The identification of vCJD and the government statement that they were probably due to BSE led to a collapse of public confidence in both the United Kingdom and Europe, and a consequent collapse of the beef market (Palmer 1996). Just 14 weeks before the announcement of vCJD the official view of the Department of Health was that the risk to humans from eating beef was only theoretical. At the end of 1995 the report of two cases in young people was accompanied by a suggestion that this was ascertainment bias or a chance finding. The account on March 20 by SEAC with eight more young cases stated that ‘on current data and in the absence of any credible alternative the most likely explanation at present is these cases are linked to exposure to BSE before the introduction of the SBO ban in 1989. This is cause for great concern.’ But in the press release issued the same day by Stephen Dorrell, Secretary for Health, the statement is also made that ‘There remains no scientific proof that BSE can be transmitted to man by beef.’ ‘No scientific proof’ and ’cause for great concern’ do not sit easily together and may have increased anxiety levels and decreased trust. Mr Dorrell’s statement said in one paragraph that ‘these new findings suggest that there may have been an association between eating bovine products’ and CJD, and in the next paragraph states ‘there remains, however, no scientific evidence that BSE can be transmitted by beef’. If the latter is true, then the expert scientific committee on whose advice government policy is based had no scientific evidence for its ‘great concern’! There followed a robust appeal by at least one public-health advocate to ‘have done with misleading the profession, the public, and the press with unqualified “no evidence of” statements. All evidence must be quantified’ (Gore 1996).
The government, in an attempt to regain public confidence in its advice on food safety, announced its intention to set up a Food Standards Agency. Also, and in response to pressure from patients’ families, an open enquiry into BSE was initiated under the chairmanship of Lord Phillips to ‘establish and review the history of the emergence and identification of BSE and variant CJD and of the action taken in response to it up to 20 March 1996’ and ‘to reach conclusions on the adequacy of that response, taking into account the state of knowledge at the time.’ This enquiry has broken new ground in the way in which access to information has been allowed. All proceedings have been conducted in public and all evidence has been made available as literal transcripts on the internet (http://www.bse.org.uk/) within a few hours. Confidential government minutes of papers have been called as evidence and all these are available on the World Wide Web. Inquiry lawyers have sieved the huge volume of documentary evidence and have even compared draft documents and minutes with final versions to seek explanations for significant omissions. The key conclusions of the Phillips Inquiry are as follows.
BSE has caused a harrowing fatal disease for humans.
A vital industry has been dealt a body blow, inflicting misery on tens of thousands for whom livestock farming is their way of life. They have seen over 170 000 of their animals die or be destroyed, and the precautionary slaughter and destruction within the United Kingdom of very many more.
BSE developed into an epidemic as a consequence of an intensive farming practice—the recycling of animal protein in ruminant feed. This practice, unchallenged over decades, proved a recipe for disaster.
At the heart of the BSE story lie questions of how to handle hazard—a known hazard to cattle and an unknown hazard to humans. The government took measures to address both hazards. They were sensible measures, but they were not always timely nor adequately implemented and enforced.
The rigour with which policy measures were implemented for the protection of human health was affected by the belief of many prior to early 1996 that BSE was not a potential threat to human life.
The government was anxious to act in the best interests of human and animal health. To this end it sought and followed the advice of independent scientific experts, sometimes when decisions could have been reached more swiftly and satisfactorily within government.
At times officials showed a lack of rigour in considering how policy should be turned into practice, to the detriment of the efficacy of the measures taken.
At times bureaucratic processes resulted in unacceptable delays in giving effect to policy.
The government introduced measures to guard against the risk that BSE might be a matter of life and death not merely for cattle but also for humans, but the possibility of a risk to humans was not communicated to the public or to those whose job it was to implement and enforce the precautionary measures.
The government did not lie to the public about BSE. It believed that the risks posed by BSE to humans were remote. The government was preoccupied with preventing an alarmist over-reaction to BSE because it believed that the risk was remote. It is now clear that this campaign of reassurance was a mistake. When, on 20 March 1996, the government announced that BSE had probably been transmitted to humans, the public felt that they had been betrayed. Confidence in government pronouncements about risk was a further casualty of BSE.
Cases of new variant CJD (vCJD) were identified by the National CJD Surveillance Unit, and the conclusion that they were probably linked to BSE was reached as early as was reasonably possible. The link between BSE and vCJD is now clearly established, though the manner of infection is not clear.
Of particular note for the public health management of the epidemic were the following conclusions of the Phillips Inquiry, released in November 2000.
The original risk assessment for human health was erroneously based on the view that BSE was introduced into cattle via feed derived from sheep infected with conventional scrapie. Since scrapie was not considered a human health hazard the potential threat to humans from BSE was underestimated.
There were significant communication problems within and between MAFF and the Department of Health, leading to significant delays in implementing control measures in relation to mechanically recovered meat, medicines, cosmetics, and occupational guidance. For example, ‘By the end of 1987 MAFF officials had become concerned as to whether it was acceptable for cattle showing signs of BSE to be slaughtered for human consumption. However, the Department of Health was not asked to collaborate with MAFF in considering the implications that BSE had for human health. It should have been.’
Communication with the public ‘failed to explain that the views expressed were subject to proper observance of the precautionary measures which had been introduced to protect human health against the possibility that BSE might be transmissible. These statements conveyed the message not merely that beef was safe but that BSE was not transmissible. The impression thus given to the public that BSE was not transmissible to humans was a significant factor leading to the public feeling of betrayal when it was announced on 20 March 1996 that BSE was likely to have been transmitted to humans.’
The Department of Health and MAFF were taken to task for overly relying on scientific committees.
The Southwood Working Party considered that all reasonably practicable precautions should be taken to reduce the risks that would exist should BSE prove to be transmissible to humans. However, they did not make this plain in their Report and did not recommend that the possible risks from eating animals incubating BSE but not yet showing signs of the disease (“subclinical cases”) called for any precautions, other than a recommendation that manufacturers should not include ruminant offal and thymus in baby food. This was a shortcoming in their Report.
Because of a failure to subject the Southwood Report to an adequate review, MAFF and DoH failed to identify this shortcoming. Concern about the food risks posed by subclinical cases was, however, expressed by some scientists, by the media and by the public. With the agreement of DoH, MAFF reacted by announcing in June 1989 that those categories of offal of cattle most likely to be infectious (SBO) were to be banned from use in human food. The introduction of this vital precautionary measure was commendable. However, this ban was presented to the public in terms that underplayed its importance as a public health measure.
Careful consideration was given by MAFF and DoH in 1989 to the terms of the human SBO ban, with one important exception. During the consultation process, concerns were raised about the practicality of ensuring the removal of all the spinal cord during abattoir processes, and about the practice of mechanical recovery of scraps left attached to the vertebral column for use in human food (‘mechanically recovered meat’ or MRM). However, MAFF officials discounted these concerns without subjecting them to rigorous consideration—in particular no advice was sought as to the minimum quantity of spinal cord that might transmit the disease in food.
MAFF gave detailed consideration to spinal cord and MRM in 1990. A lengthy paper was submitted to SEAC, the Government’s new expert advisory committee on TSEs. Unhappily, as a result of a breakdown in communications, MAFF officials understood that the members of SEAC were not concerned about the inclusion in human food of an occasional scrap of spinal cord, so that no action was called for. In fact the advice of some, at least, of the members of SEAC was premised on the false assumption that spinal cord could readily be removed from the carcass in its entirety, and would be so removed. This was one of a number of occasions that has given rise to lessons for the future about the proper use of expert committees by the Government.
Not until 1995 was action taken in relation to MRM . . . Up to 1995, MRM was a potential pathway to the infection of humans with BSE, not merely because of the risk of inclusion of the occasional portion of spinal cord, but because the material recovered by the MRM process included dorsal root ganglia. These were peripheral nervous tissues which were not thought to be infectious at the time, but which have since been demonstrated to be infectious in the late stages of incubation.
Formal scientific risk assessment has continued. For example, the United Kingdom government’s Advisory Committee on Dangerous Pathogens, together with SEAC, has reviewed the guidance for those who work with the agents of transmissible spongiform encephalopathies (Advisory Committee on Dangerous Pathogens 1996, 1998). A particular priority was to review advice to those in industrial workplaces such as slaughterhouses and carcass-dressing plants (Box 1). This was published in June 1996 (Advisory Committee on Dangerous Pathogens 1996), and in March 1998 revised guidance for those working in laboratories and hospitals, and handling animals, was published (Advisory Committee on Dangerous Pathogens 1998). However, a significant change in approach has been taken. Since 1997 a concentrated effort to conduct risk assessment more openly has led to the use of the World Wide Web to provide information on the agenda of SEAC, its advice, and government response.
Box 1 General basic protective measures for those working with BSE-infective animals
Adhere to safe working practices and take extra care to avoid or minimize the use of tools and equipment likely to cause cuts, abrasions, or puncture wounds
Where use of such equipment is unavoidable, wear suitable protective clothing (e.g. chain-mail gloves when using knives in the abattoir)
Cover existing cuts, abrasions, and skin lesions on exposed skin with waterproof dressings
If cuts or puncture wounds occur, encourage the wound to bleed, and then wash thoroughly with soap and water and cover with a waterproof dressing
Use face protection (chiefly for eyes and mouth) if there is a risk of splashing, for example a visor or fixed screen
If splashed in the eyes or face, wash with running water
Take steps to avoid the generation of aerosols and dusts
Wash hands and exposed skin before eating, drinking, smoking, taking any medication, using the telephone, or going to the toilet
Wash down contaminated areas and equipment regularly with hot water and detergent
Wash protective clothing thoroughly after use and store separately from other clothing; alternatively, use disposable clothing
Use non-penetrative methods of stunning cattle if possible
Avoid the use of reciprocating saws if the carcass has to be split through the spine
Use low-pressure hoses to clean areas contaminated with specified bovine material
Control exposure to dusts arising from greaves by use of engineering controls and/or standard personal protection measures
For carcasses destined for disposal, we recommend acceleration of the development of alternative dressing techniques to avoid exposure of the spinal cord
Based on Advisory Committee on Dangerous Pathogens/SEAC (1998)
The task for scientists estimating risks on usually limited evidence is considerable, especially so when the consequences of a wrong judgement may be grave, and particularly when conducted under the gaze of the media (www.doh.gov.uk/cjd/, 4 December 2000). At times the precautionary principle has been followed (such as the banning of beef on the bone) even when the estimated risk is minuscule. However, where the social and economic cost of such action is potentially huge, the precautionary principle is unrealistic. For example, of particular concern has been the discovery of PrPSc in lymphoid tissue and postmortem material from a patient with vCJD, and in appendix tissue removed 8 months before clinical onset, calling into question the safety of blood, blood products, and reusable surgical instruments (Collinge 1999). In September 1997 the United Kingdom Committee on Safety of Medicines recommended that blood products possibly obtained from confirmed CJD cases should be withdrawn. In November 1997 SEAC asked the Department of Health to commission an independent risk assessment on human-to-human transmission of vCJD via blood. Preliminary action has already been taken to remove white blood cells from all transfusion blood. An even more difficult issue is the potential for transmission via contaminated reused sterilized instruments, since prions are highly resistant to conventional sterilization procedures. Use of new instruments for every patient will have huge cost implications.
The BSE epidemic will continue to have major repercussions for the public’s health. Firstly, the scale of the vCJD epidemic will become clear in the next few years, and there will be an increasing threat from person-to-person spread, both horizontally through surgical equipment, and vertically from mother to infant. Secondly, the question of whether there are exogenous risk factors for classical CJD remains open and new research may implicate other zoonotic exposures. Thirdly, the biology of prions will develop significantly. Will other prion-related diseases be identified? Fourthly, the procedures for public policy development, scientific risk assessment, and communication will be adapted to accommodate the recommendations of the Phillips report (http://www.bseinquiry.gov.uk/).
Advisory Committee on Dangerous Pathogens (1996). BSE (bovine spongiform encephalopathy): background and general occupational guidance. HSE Books, London.
Advisory Committee on Dangerous Pathogens/SEAC (1998). Transmissible spongiform encephalopathy agents: safe working and the prevention of infection. HMSO, London.
Akai, J., Ishihara, O., and Higuchi, S. (1989). Creutzfeldt–Jakob disease in Japan: an epidemiological study done in a select prefecture between 1976 and 1986. Neuroepidemiology, 8, 32–7.
Alper, T., Cramp, W.A., Haig, D.A., and Clarke, M.C. (1967). Does the agent of scrapie replicate without nucleic acid? Nature, 214, 764–6.
Alter, M., Frank, Y., Doyne, H., and Webster, D.D. (1975). Creutzfeldt–Jakob disease after eating ovine brains? New England Journal of Medicine, 292, 927.
Anonymous (1996a). Creutzfeldt–Jakob disease in Canada. Canadian Communicable Disease Report, 22-8, 57–60.
Anonymous (1996b). Less beef, more brain. Lancet, 347, 915.
Araya, G., Gálvez, S., Cartier, L., and Gajdusek, D.C. (1983). A spatiotemporal clustering of Creutzfeldt–Jakob disease in Chile. Revista Chilena de Neuropsiquiatria, 21, 291–5.
Basler, K., Oesch, B., Scott, M., et al. (1986). Scrapie and cellular PrP isoforms are encoded by the same chromosomal gene. Cell, 46, 417–28.
Belay E.D. (1999). Transmissible spongiform encephalopathies in humans. Annual Review of Microbiology, 53, 283–314.
Berger, J.R. and Noble, J.D. (1993). Creutzfeldt–Jakob disease in a physician: a review of the disorder in health care workers. Neurology, 43, 205–6.
Berger, J.R., Weisman, E., and Weisman, B. (1997). Creutzfeldt–Jakob disease and eating squirrel brains. Lancet, 350, 642.
Bernoulli C., Siegfried, J., Baumgartner, G., et al. (1977). Danger of accidental person-to-person transmission of Creutsfeldt-Jakob disease by surgery. Lancet, i, 478–9.
Bobowick, A.R., Brody, J.A., Matthews, M.R., Roos, R., and Gajdusek, D.C. (1973). Creutzfeldt–Jakob disease: a case–control study. American Journal of Epidemiology, 98, 381–94.
Bradley, R. and Wilesmith, J.W. (1993). Epidemiology and control of bovine spongiform encephalopathy (BSE). British Medical Bulletin, 49, 932–59.
Brown, P. (1988). The clinical neurology and epidemiology of Creutzfeldt–Jakob disease, with special reference to iatrogenic cases. Ciba Foundation Symposium, 135, 3–23.
Brown, P. (1995). Can Creutzfeldt–Jakob disease be transmitted by transfusion? Current Opinion in Haematology, 2, 472–7.
Brown, P., Cathala, F., and Gajdusek, D.C. (1979). Creutzfeldt–Jakob disease in France: III. Epidemiological study of 170 patients dying during the decade 1968–1977. Annals of Neurology, 6, 438–46.
Brown, P., Cathala, F., Castaigne, P., and Gajdusek, D.C. (1986). Creutzfeldt–Jakob disease: clinical analysis of a consecutive series of 230 neuropathologically verified cases. Annals of Neurology, 20, 597–602.
Brown, P., Cathala, F., and Raubertas, R. (1988). Creutzfeldt–Jakob disease in France: epidemiologic data from 1968–1982. In Unconventional virus diseases of the central nervous system (ed. L.A. Court, D. Dormant, and P. Brown). Commissariat a l’Energie Atomique, Fontenay-aux-Roses.
Brown, P., Preece, M.A., and Will, R.G. (1992). ‘Friendly fire’ in medicine: hormones, homografts, and Creutzfeldt–Jakob disease. Lancet, 340, 24–7.
Brown, P., Gibbs, Jr., C.J., Rodgers-Johnson, P., et al. (1994a). Human spongiform encephalopathy: the national institutes of health series of 300 cases of experimentally transmitted disease. Annals of Neurology, 35, 513–29.
Brown, P., Cervenáková, L., Goldfarb, L.G., et al. (1994b). Iatrogenic Creutzfeldt–Jakob disease: an example of the interplay between ancient genes and modern medicine. Neurology, 44, 291–3.
Brown, P., Cervenáková, L., and Powers, J.M. (1998). Fatal familial insomnia around the world: FFI cases from the United States, Australia, and Japan. Brain Pathology, 8, 567–70.
Bruce, M.E., Will, R.G., Ironside, J.W., et al. (1997). Transmissions to mice indicate that ‘new variant’ CJD is caused by the BSE agent. Nature, 389, 498–501.
Budka, H., Aguzzi, A., Brown, P., et al. (1995). Neuropathological diagnostic criteria for Creutzfeldt–Jakob disease (CJD) and other human spongiform encephalopathies (prion diseases). Brain Pathology, 5, 459–66.
Budka, H., Almer, G., Hainfellner, J.A., and Jellinger, K. (1998). Fatal familial insomnia around the world: the Austrian FFI cases. Brain Pathology, 8, 554.
Chandler, R.L. (1961). Encephalopathy in mice produced by inoculation with scrapie brain material. Lancet, 1378–9.
Chandrakumar, M. (1998). An investigation of the local cases of Creutzfeldt–Jakob disease in East Kent. Report, East Kent Health Authority.
Chatelain, J., Cathala, F., Brown, P., Raharison, S., Court, L., and Gajdusek, D.C. (1981). Epidemiologic comparisons between Creutzfeldt–Jakob disease and scrapie in France during the 12-year period 1968–1979. Journal of the Neurological Sciences, 51, 329–37.
Chesebro, B. (1998). BSE and prions: uncertainties about the agent. Science, 279, 42–3.
Cohen, C.H. and Valleron, A.-J. (1999). When did bovine spongiform encephalopathy (BSE) start? Implications of the prediction of a new variant Creutzfeldt–Jakob disease (nvCJD) epidemic. International Journal of Epidemiology, 28, 526–31.
Collee, J.G. and Bradley, R. (1997). BSE: a decade on—Part 2. Lancet, 349, 715–21.
Collinge, J. (1999). Variant Creutzfeldt–Jakob Disease. Lancet, 354, 317–23.
Collinge, J., Palmer, M.S., and Dryden, A.J. (1991). Genetic predisposition to iatrogenic Creutzfeldt–Jakob disease. Lancet, 337, 1441–2.
Collins, S., Law, M.G., Fletcher, A., Boyd, A., Kaldor, J., and Masters, C.L. (1999). Surgical treatment and risk of sporadic Creutzfeldt–Jakob disease: a case control study. Lancet, 353, 693–7.
Cousens, S.N., Zeidler, M., Esmonde, T.F., et al. (1997a). Sporadic Creutzfeldt–Jakob disease in the United Kingdom: analysis of epidemiological surveillance data for 1970–1996. British Medical Journal, 315, 389–96.
Cousens, S.N., Vynnycky, E., Zeidler, M., Will, R.G, and Smith, P.G. (1997b). Predicting the CJD epidemic in humans. Nature, 385, 197–8.
Cousens, S.N., Linsell, L., Smith, P.G., and Will, R.G. (1999). Cluster of vCJD in Kent and its importance. Lancet, 353, 18–21.
Creange, A., Gray, F., Cesaro, P., et al. (1995). Creutzfeldt–Jakob disease after liver transplantation. Annals of Neurology, 38, 269–72.
Davanipour, Z., Alter, M., Sobel, E., Asher, D.M., and Gajdusek, D.C. (1985a). Creutzfeldt–Jakob disease: possible medical risk factors. Neurology, 35, 1483–6.
Davanipour, Z., Alter, M., Sobel, E., Asher, D.M., and Gajdusek, D.C. (1985b). A case control study of Creutzfeldt–Jakob disease: dietary risk factors. American Journal of Epidemiology, 122, 443–51.
Davanipour, Z., Alter, M., Sobel, E., Asher, D.M., and Gajdusek, D.C. (1986). Transmissible virus dementia: evaluation of a zoonotic hypothesis. Neuroepidemiology, 5, 194–206.
de Silva, R. (1996). Human spongiform encephalopathy: clinical presentation and diagnostic tests. In Prion diseases (ed. H.F. Baker and R.M. Ridley), pp. 15–33. Humana Press, Totowa, NJ.
Dickinson, A.G. and Taylor, D.M. (1978). Resistance of scrapie agent to decontamination. New England Journal of Medicine, 229, 1413–14.
Duffy, P., Wolf, J., Collins, G., DeVoe, A.G., Streeten, B., and Cowen, D. (1974). Possible person-to-person transmission of Creutzfeldt–Jakob disease. New England Journal of Medicine, 290, 692–3.
Esmonde, T.G.F., Will, R.G., and Slattery, J.M. (1993). Creutzfeldt–Jakob disease and blood transfusion. Lancet, 341, 205–7.
Gajdusek, D.C. (1977). Unconventional viruses and the origin and disappearance of kuru. Science, 197, 943–60.
Gajdusek, D.C. and Zigas, V. (1957). Degenerative disease of the central nervous system in New Guinea: the endemic occurrence of ‘kuru’ in the native population. New England Journal of Medicine, 257, 974–8.
Galvez, S., Masters, C., and Gajdusek, D.C. (1980). Descriptive epidemiology of Creutzfeldt–Jakob disease in Chile. Archives of Neurology, 37, 11–14.
Galvez, S., Cartier, L., Monari, M., Araya, G. (1983). Familial Creutzfeldt–Jakob disease in Chile. Journal of the Neurological Sciences, 59, 139–47.
Gambetti, P. and Lugaresi, E. (1998). Fatal familial insomnia around the world: conclusions of the symposium. Brain Pathology, 8, 571–5.
Ghani, A.C., Ferguson, N.M., Donnelly, C.A., Hagenaars, T.J., and Anderson, R.M. (1998). Epidemiological determinants of the pattern and magnitude of vCJD epidemic in Great Britain. Proceedings of the Royal Society of London, 265, 2443–52.
Ghani, A., Ferguson, N.M., Donnelly, C.A., and Anderson, R.M. (2000). Predicted vCJD mortality in Great Britain. Nature, 406, 583–4.
Goldfarb, L.G., Brown, P., Goldgaber, D., et al. (1989). Patients with Creutzfeldt–Jakob disease and kuru lack the mutation in the PRIP gene found in Gerstmann-Straussler syndrome, but they show a different allele mutation in the same gene. American Journal of Genetics, 45, A189.
Goldfarb, L.G., Petersen, R.B., Tabaton, M., et al. (1992). Fatal familial insomnia and familial Creutzfeldt–Jakob disease: disease phenotype determined by a DNA polymorphism. Science, 258, 806–8.
Gordon, W.S. (1946). Advances in veterinary research: louping ill, tick-borne fever and scrapie. Veterinary Record, 58, 516–21.
Gore, S.M. (1996). Bovine Creutzfeldt–Jakob disease? Failure of epidemiology must be remedied. British Medical Journal, 312, 791–3.
Hainfellner, J.A., Jellinger, K., Diringer, H., et al. (1996). Creutzfeldt–Jakob disease in Austria. Journal of Neurology, Neurosurgery, and Psychiatry, 61, 139–42.
Harries-Jones, R., Knight, R., Will, R.G., Cousens, S., Smith, P.G., Matthews, W.B. (1988). Creutzfeldt–Jakob disease in England and Wales, 1980–1984: a case control study of potential risk factors. Journal of Neurology, Neurosurgery, and Psychiatry, 51, 1113–19.
Hartsough, G.R. and Burger, D. (1965). Encephalopathy of mink. 1. Epizootiologic and clinical observation. Journal of infectious diseases, 115, 387–92.
Hill, A.F., Desbruslais, M., Joiner, S., et al. (1997). The same prion strain causes vCJD and BSE. Nature, 389, 448–50.
Hill, A.F., Butterworth, R.J., Joiner, S., et al. (1999). Investigation of variant Creutzfeldt–Jakob disease and other human prion diseases with tonsil biopsy samples. Lancet, 353, 183–9.
Hillier, C.E.M. and Salmon, R.L. (2000). Is there evidence for exogenous risk factors in the aetiology and spread of Creutzfeldt disease? Quarterly Journal of Medicine, 93, 617–31.
Holman, R.C., Khan, A.S., Kent, J., Strine, T.W., and Schonberger, L.B. (1995). Epidemiology of Creutzfeldt–Jakob disease in the United States, 1979–1990: analysis of national mortality data. Neuroepidemiology, 14, 174–81.
Hsiao, K., Baker, H.F., Crow, T.J., et al. (1989). Linkage of a prion protein missense variant to Gerstmann-Sträussler syndrome. Nature, 338, 342–5.
Ironside, J.W. (1996). Human prion diseases. J Neural Transm, 47 (Supplement), 231–46.
Ironside, J.W. and Bell, J.E. (1997). Florid plaques and new variant Creutzfeldt–Jakob disease. Lancet, 350, 1475.
Johnson, R.T. and Gibbs, Jr., C.J. (1998). Creutzfeldt–Jakob disease and related transmissible spongiform encephalopathies. New England Journal of Medicine, 339, 1994–2004.
Julien, J., Vital, C., Delisle, M.B., and Géraud, G. (1998). Fatal familial insomnia around the world: the French FFI cases. Brain Pathology, 8, 555–8.
Kahana, E., Alter, M., Braham, J., and Sofer, D. (1974). Creutzfeldt–Jakob disease: focus among Libyan Jews in Israel. Science, 183, 90–1.
Kamin, M. and Patten, B.M. (1984). Creutzfeldt–Jakob disease: possible transmission to humans by consumption of wild animal brains. American Journal of Medicine, 76, 142–5.
Kimberlin, R.H. (1981). Scrapie. British Veterinary Journal, 137, 105–12.
Kirkwood, J.K. and Cunningham, A.A. (1994). Epidemiological observations on spongiform encephalopathies in captive wild animals in the British Isles. Veterinary Record, 135, 296–303.
Klein, R. and Dumble, L.J. (1993). Transmission of Creutzfeldt–Jakob disease by blood transfusion. Lancet, 341, 768.
Kondo, K. (1985). Epidemiology of Creutzfeldt–Jakob disease in Japan. In Creutzfeldt–Jakob disease (ed. T. Mizutanti and H. Khiraki), pp. 17–30. Wishmura/Wiigate/Elsevier, Amsterdam.
Kondo, K. and Kuroiwa, Y. (1982). A case control study of Creutzfeldt–Jakob disease: association with physical injuries. Annals of Neurology, 11, 377–81.
Kovanen, J. and Haltia, M. (1988). Descriptive epidemiology of Creutzfeldt–Jakob disease in Finland. Acta Neurologica Scandinavica, 77, 474–80.
Kretzschmar, H.A., Ironside, J.W., DeArmond, S.J., and Tateishi, J. (1996). Diagnostic criteria for sporadic Creutzfeldt–Jakob disease. Archives of Neurology, 53, 913–20.
Kretzschmar, H.A., Giese, A., Zerr, I., et al. (1998). Fatal familial insomnia around the world: the German FFI cases. Brain Pathology, 8, 559–61.
Lechi, A., Tedeschi, F., Mancia, D., et al. (1983). Creutzfeldt–Jakob disease: clinical, EEG and neuropathological findings in a cluster of eleven patients. Italian Journal of Neurological Sciences, 4, 47–59.
Lugaresi, E., Medori, R., Montagna, P., et al. (1986). Fatal familial insomnia and dysautonomia with selective degeneration of thalamic nuclei. New England Journal of Medicine, 315, 997–1003.
Lundberg, P.O. (1998). Creutzfeldt–Jakob disease in Sweden. Journal of Neurology, Neurosurgery and Psychiatry, 65, 836–41.
McKenzie, D., Bartz, J.C., and Marsh, R.G. (1996). Transmissible mink encephalopathy. Virology, 7, 201–6.
Majeed, A., Lehmann, P., Kirby, L., and Coleman, M.P. (1998). Mortality from dementias and neurodegenerative disorders in people aged 15–64 in England and Wales in 1975–1996. British Medical Journal, 317, 320–1.
Majtenyi, C. (1988). Study on the cases of Creutzfeldt–Jakob disease in Hungary. In Unconventional virus diseases of the central nervous system (ed. L.A. Court, D. Dormant, and P. Brown). Commissariat a l’Energie Atomique, Fontenay-aux-Roses.
Marsh, R.F., Bessen, R.A., Lehmann, S., and Hartsough, G.R. (1991). Epidemiological and experimental studies on a new incident of transmissible mink encephalopathy. Journal of General Virology, 72, 589–94.
Masters, C.L., Harris, J.O., Gajdusek, D.C., Gibbs, Jr., C.J., Bernoulli, C., Asher, D.M. (1979). Creutzfeldt–Jakob disease: patterns of worldwide occurrence and significance of familial and sporadic clustering. Annals of Neurology, 5, 177–88.
Masters, C.L., Gajdusek, D.C., and Gibbs, Jr., C.J. (1981). Creutzfeldt–Jakob disease virus isolations from the Gerstmann–Straussler syndrome with an analysis of the various forms of amyloid plaque deposition in the virus-induced spongiform encephalopathies. Brain, 104, 559–88.
Mastrianni, J.A., Nixon, R., Layzer, R., et al. (1999). Prion protein conformation in a patient with sporadic fatal insomnia. New England Journal of Medicine, 340, 1630–8.
Masullo, C., Pocchiari, M., Neri, G., et al. (1988). A retrospective study of Creutzfeldt–Jakob disease in Italy (1972–1986). European Journal of Epidemiology, 4, 482–7.
Matthews, W.B. (1975). Epidemiology of Creutzfeldt–Jakob disease in England and Wales. Journal of Neurology, Neurosurgery and Psychiatry, 38, 210–13.
Matthews, W.B. and Will, R.G. (1981). Creutzfeldt–Jakob disease in a lifelong vegetarian. Lancet, ii, 937.
Matthews, W.B., Campbell, M., Hughes, J.T., and Tomlinson, A.H. (1979). Creutzfeldt–Jakob disease and ferrets. Lancet, i, 828.
Maxwell, R.J. (1999). The British Government’s handling of risk: some reflections on the BSE/CJD crisis. In Risk communication and public health (ed. P. Bennett and K. Calman), pp. 95–107. Oxford University Press.
Mayer, V., Orolin, D., and Mitrová, E. (1977). Cluster of Creutzfeldt–Jakob disease and presenile dementia. Lancet, ii, 256.
Maytenyi, K. (1991). Creutzfeldt–Jakob disease in the last 5 years in Hungary. European Journal of Epidemiology, 7, 457–9.
Medori, R., Tristschler, H.-J., LeBlanc, A., et al. (1992). Fatal familial insomnia, a prion disease with a mutation at codon 178 of the prion protein gene. New England Journal of Medicine, 326, 444–9.
National CJD Surveillance Unit, and the Department of Epidemiology and Population Sciences London School of Hygiene and Tropical Medicine (1995). Creutzfeldt–Jakob disease surveillance in the United Kingdom: fourth annual report, pp. 1–53.
National CJD Surveillance Unit, and the Department of Epidemiology and Population Sciences London School of Hygiene and Tropical Medicine (1996). Creutzfeldt–Jakob disease surveillance in the United Kingdom: fifth annual report, pp.1–67.
National CJD Surveillance Unit, and the Department of Epidemiology and Population Sciences London School of Hygiene and Tropical Medicine (1997). Creutzfeldt–Jakob disease surveillance in the United Kingdom: sixth annual report, pp. 1–45.
National CJD Surveillance Unit, and the Department of Epidemiology and Population Sciences London School of Hygiene and Tropical Medicine (1998). Creutzfeldt–Jakob disease surveillance in the United Kingdom: seventh annual report, pp. 1–53.
National CJD Surveillance Unit, and the Department of Epidemiology and Population Sciences London School of Hygiene and Tropical Medicine. (1999). Creutzfeldt–Jakob disease surveillance in the United Kingdom: eighth annual report, pp. 1–51.
Neugat, R.H., Neugat, A.I., Kahana, E., Stein, Z., and Alter, M. (1979). Creutzfeldt–Jakob disease: familial clustering among Libyan-born Israelis. Neurology, 29, 225–31.
Palmer, M.S., Dryden, A.J., Hughes, T., and Collinge, J. (1991). Homozygous prion protein genotype predisposes to sporadic Creutzfeldt–Jakob disease. Nature, 352, 340–2.
Palmer, S.R. (1996). The BSE Crisis—an assessment. Eurohealth, 2, 10–12.
Parchi, P., Capellari, S., Chin, S., et al. (1998). Creutzfeldt–Jakob disease (CJD) after blood product transfusion from a donor with CJD. Neurology, 50, 1872–3.
Pattison, I.H. (1988). Fifty years with scrapie: a personal reminiscence. Veterinary Record, 123, 661–6.
Pattison, J. (1998). The emergence of bovine spongiform encephalopathy and related diseases. Emerging Infectious Diseases, 4, 390–4.
Patterson, W.J. and Painter, M.J. (1997). Bovine spongiform encephalopathy and new variant Creutzfeldt–Jakob disease: an overview. Communicable Diseases and Public Health, 2, 5–13.
Pearson, G.R., Wyatt, J.M., Henderson, J.P., and Gruffydd-Jones, T.J. (1993). Feline spongiform encephalopathy: a review. Veterinary Annual, 33, 1–10.
Petersen, R.B. (1999). Antemortem diagnosis of variant Creutzfeldt–Jakob disease. Lancet, 353, 163–4.
Pocchiari, M., Ladogana, A., Petraroli, R., Cardone, F., and D’Alessandro, M. (1998). Fatal familial insomnia around the world. Recent Italian FFI cases. Brain Pathology, 8, 564–6.
Prusiner, S.B. (1982). Novel proteinaceous infectious particles cause scrapie. Science, 216, 136–44.
Prusiner, S.B. (1993). Genetic and infectious prion diseases. Archives of Neurology, 30, 1129–53.
Raubertas, R., Brown, P., Cathala, F., and Brown, I. (1991). The question of clusters in Creutzfeldt–Jakob disease. American Journal of Epidemiology, 129, 146–54.
Ricketts, M.N., Cashman, N.R., Stratton, E.E., and El Saadany, S. (1997). Is Creutzfeldt–Jakob disease transmitted in blood? Emerging Infectious Diseases, 3, 155–63.
Schätzl, H.M., Wopfner, F., Gilch, S., von Brunn, A., and Jäger, G. (1997). Is codon 129 of prion protein polymorphic in human beings but not in animals? Lancet, 349, 1603–4.
Schonberger, L.B. (1998). New variant Creutzfeldt–Jakob disease and bovine spongiform encephalopathy. Emerging Infectious Diseases, 12, 111–21.
Schreuder, B.E.C. (1994). Animal spongiform encephalopathies—an update. Part II. Bovine spongiform encephalopathy (BSE). Veterinary Quarterly, 16, 182–92.
Singhal, B.S. and Dastur, D.K. (1983). Creutzfeldt–Jakob disease in Western India. Observations in 7 patients. Neuroepidemiology, 2, 93–100.
Southwood, R. (1989). Report of the Working Party on Bovine Spongiform Encephalopathy. HMSO, London.
Spongiform Encephalopathy Advisory Committee (1995). Transmissible spongiform encephalopathies: a summary of present knowledge and research. 1, 1–95. HMSO, London.
Stratton, E.E., Ricketts, M.N., and Gully, P.R. (1997). The epidemiology of Creutzfeldt–Jakob disease in Canada: a review of mortality data. Emerging Infectious Diseases, 3, 63–4.
Takashima, S., Tateishi, J., Taguchi, Y., and Inoue, H. (1997). Creutzfeldt–Jakob disease with florid plaques after cadaveric dural graft in a Japanese woman. Lancet, 350, 865–6.
Tsuji, S. and Kuroiwa, Y. (1983). Creutzfeldt–Jakob disease in Japan. Neurology, 33, 1503–6.
van Duijn, C.M., Delasnerie-Laupretre, N., Masullo, C., et al. (1998). Case–control study of risk factors of Creutzfeldt–Jakob disease in Europe during 1993–1995. Lancet, 351, 1081–5.
Weber, T., Tumani, H., Holdorf, B., et al. (1993). Transmission of Creutzfeldt–Jakob disease by handling of dura mater. Lancet, 341, 123–4.
Wells, G.A.H., Scott, A.C., Johnson, C.T., et al. (1987). A novel progressive spongiform encephalopathy in cattle. Veterinary Record, 121, 419–20.
Wilesmith, J.W., Wells, G.A.H., Cranwell, M.P., and Ryan, J.B.M. (1988). Bovine spongiform encephalopathy: epidemiological studies. Veterinary Record, 123, 638–44.
Will, R.G. and Matthews, W.B. (1982). Evidence for case-to-case transmission of Creutzfeldt–Jakob disease. Journal of Neurology, Neurosurgery and Psychiatry, 45, 235–8.
Will, R.G., Matthews, W.B., Smith, P.G., and Hudson, C. A retrospective study of Creutzfeldt–Jakob disease in England and Wales 1970–1979. II: epidemiology. Journal of Neurology, Neurosurgery, and Psychiatry 1986, 49, 749–55.
Will, R.G., Ironside, J.W., Zeidler, M., et al. (1996). A new variant of Creutzfeldt–Jakob disease in the United Kingdom. Lancet, 347, 921–5.
Will, R.G., Alperovitch, A., Poser, S., et al. (1998a). Descriptive epidemiology of Creutzfeldt–Jakob disease in six European countries, 1993–1995. Annals of Neurology, 43, 763–67.
Will, R.G., Campbell, M.J., Moss, T.H., Bell, J.E., and Ironside, J.W. (1998b). Fatal familial insomnia around the world: FFI cases from the United Kingdom. Brain Pathology, 8, 562–3.
Will, R.G., Stewart, G., Zeidler, M., Macleod, M.A., and Knight, R.S.G. (1999). Psychiatric features of new variant Creutzfeldt–Jakob disease. Psychiatric Bulletin, 23, 264–7.
Will, R.G., Zeidler, M., Stewart, G.E., et al. (2000). Diagnosis of new variant Creutzfeldt–Jakob disease. Annals of Neurology, 47, 575–82.
Williams, E.S. and Young, S. (1980). Chronic wasting disease of captive mule deer: a spongiform encephalopathy. Journal of Wildlife Disease, 16, 89–98.
Williams, E.S. and Young, S. (1982). Spongiform encephalopathy of Rocky Mountain elk. Journal of Wildlife Disease, 18, 465–71.
WHO (World Health Organization) (1998). Global surveillance, diagnosis and therapy of human transmissible spongiform encephalopathies: report of a WHO consultation, pp. 1–30. WHO, Geneva.
Worthington, J.M. (1995). Epidemiology of Jakob-Creutzfeldt disease in Australia 1970–80. Australian and New Zealand Journal of Medicine, 25, 243–4.
Zanusso, G., Nardelli, E., and Rosati, A. (1998). Simultaneous occurrence of spongiform encephalopathy in a man and his cat in Italy. Lancet, 352, 1116–17.
Zeidler, M., Stewart, G.E., Barraclough, C.R., et al. (1997a). New variant Creutzfeldt–Jakob disease: neurological features and diagnostic tests. Lancet, 350, 903–7.
Zeidler, M., Johnstone, E.C., Bamber, R.W.K., et al. (1997b). New variant Creutzfeldt–Jakob disease: psychiatric features. Lancet, 350, 908–10.