Volume 8: Variant CJD
5. Emergence of variant CJD
Discussion on the work of the CJD Surveillance Unit (CJDSU)
Could vCJD have been identified earlier?
Could PHLS have been usefully involved in surveillance?
Was there adequate communication of information to the Department of Health?
Was involvement of the CJDSU in the public debate about the safety of beef appropriate?
Is a link between BSE and vCJD established?
Some observations about vCJD that call for explanation

5.158 While the Southwood Working Party judged that the risk of transmission of BSE to humans appeared remote, they made the reasonable assumption that, if BSE did transmit, the clinical disorder would closely resemble Creutzfeldt-Jakob disease. They recommended the monitoring of cases of CJD through the neurological network with particular emphasis on occupational groups exposed to bovine tissues, and suggested that the new Consultative Committee on Research chaired by Dr Tyrrell should consider arrangements for surveillance.

5.159 Dr Will presented a paper to the Tyrrell Committee in April 1989 on the surveillance of the population for CJD based on the studies he had done as a research registrar with Professor Bryan Matthews in 1979-82. The Tyrrell Committee Report recommended CJD surveillance and a grant application based on Dr Will's report was accepted by the Department of Health, and the CJD Surveillance Unit (CJDSU) started work in May 1990.

5.160 The worst fears of officials and advisors were realised when, on 8 March 1996, the CJDSU reported to SEAC that they were of the opinion that cases in ten young people could be a new form of CJD. The circumstantial evidence of a link with BSE led to the public announcement on 20 March 1996.

Could vCJD have been identified earlier?

5.161 There were indications from 1993 onwards that the CJDSU might indeed detect atypical cases or changing patterns in the incidence of CJD. We have summarised in this chapter the referrals to the CJDSU of cases of CJD in young people, beginning with case A which came to the attention of Dr Will in 1993, and cases of CJD in farmers whose farms had a history of BSE. With hindsight, we know that the farmers and some cases in young people have not been established as vCJD. However, the majority of cases of CJD in young people referred to the CJDSU by the end of 1995 have now been established to be vCJD. Only one of the CJD patients from the UK had been diagnosed in their teens (in 1980) prior to the emergence of BSE. The concern of Dr Will and Professor Collinge that so many young patients had been identified in such as short period was noted in the minutes of the SEAC meeting of 5 January 1996. In late 1995, Dr Will had noted that three of the cases were unusual in that pathological examination had revealed extensive plaque deposition in the brain. Why was it not until March 1996 that the CJDSU concluded that there was indeed probably a new variant of CJD?

5.162 Dr Will considered factors which might have led to increased identification of young patients since 1994:

1. Ascertainment bias. Young patients might have been wrongly classified as subacute sclerosis panencephalitis in the past. However, a review of patients with this diagnosis revealed no cases of CJD.
2. An increased post-mortem rate. This could have led to an increased chance in identifying atypical cases in young patients.
3. The availability of immuno-chemistry for PrP^Sc. This technique was increasingly available in the 1990s and had great diagnostic value.

5.163 At the 23rd meeting of SEAC on 5 January 1996, Dr Will listed four definite cases and one possible case under 30 years of age, along with two confirmed cases and one suspect case aged between 30 and 40. Several days later, the possible case under 30 had been confirmed by neuropathology as had a case in a patient aged 41. In his assessment of the significance of the cases, Dr Will sought up-to-date information from the European surveillance project on CJD. This revealed that between 1993-96 there were five recorded deaths within the age range 20-39. Excluding the teenagers, this age distribution was similar to the UK distribution of cases.

5.164 Dr Will then reviewed the clinical presentation and course of illness in the young cases and this revealed:

1. a relatively prolonged duration of illness;
2. early psychiatric symptoms;
3. absence of the EEG pattern characteristic of sporadic CJD; and
4. prominent ataxia.

5.165 By the end of January, the clinical phenotype and the unusual neuropathology had been recognised by the CJDSU. Similar features had been associated occasionally in both sporadic and familial CJD. Three of the cases had been tested and shown not to have prion gene mutations. They were noted also to be homozygous for methionine at codon 129. Arrangements were made to check for prion gene mutations in the remaining patients. (These results for a total of six patients were available on 4 March 1996.)

5.166 By early March, Dr Will had obtained from European colleagues information on the clinical presentation, phenotype and neuropathology on young patients with CJD identified in continental Europe. The clinical features and duration of illness were variable in these cases and did not follow the pattern that had been recognised in the young cases in the UK. This was confirmed by a literature search of other young cases of CJD from around the world. It appeared probable that the CJD of young patients recognised in the UK was a new variant, unrecognised elsewhere.

5.167 By 13 March 1996, the number of UK patients under the age of 50 with the new phenotype of CJD was nine. ¹ This number excluded one case found to have a prion gene mutation. All were homozygous for methionine at codon 129. All had the characteristic phenotype with psychiatric presentation and a relatively long duration of illness. It had been confirmed by three external referees that the pathological features, including prominent plaques, constituted a novel type of CJD, previously unreported.

5.168 It is clear that since the summer of 1995, the key question in the forefront of the minds of Dr Will and his colleagues concerning the young patients with CJD was whether or not there could be a link with BSE. On 8 March 1996, Dr Will and Dr Ironside presented their reasons for concluding that cases in young people could be a new form of CJD. Review of the medical histories of the young patients revealed no common risk factor which might explain the occurrence of their condition. At their meeting on 16 March, SEAC agreed that 'They must take very seriously the possibility that this new risk factor was BSE, although it was noted that the data did not allow this conclusion to be drawn firmly'.

5.169 We believe that the presentation to SEAC on 8 March 1996 could not have been made sooner. It was necessary to establish the clinical and pathological characteristics of the condition in a sufficient number of patients to justify the conclusion that a new variant of CJD had been identified. The findings had to be checked by independent scientists and clinicians, and it had to be shown that what appeared to be a new variant was not in fact a type of CJD previously reported in young people, either in the UK or abroad, before the BSE epidemic. Furthermore, the DNA of each patient had to be analysed to exclude a disease-producing mutation of the prion gene. These were all prerequisites to the conclusion that there was a new variant of CJD which was probably linked with BSE. The time taken to establish a link with BSE does not warrant criticism. A wrong conclusion, hastily drawn, could have created unwarranted public anxiety and could have been very damaging to public confidence.

Could PHLS have been usefully involved in surveillance?

5.170 We now turn to consideration of the procedures in place for CJD surveillance. It is hard to see what alternatives to the CJDSU could have been more effective. The CJDSU was a research unit, established on the basis of a remote possibility that BSE might be transmitted to humans. It was sensible to have a dedicated unit to monitor a comparatively rare and complex group of disorders for which there was no laboratory test. The decision to place the responsibility for surveillance with a small research team of dedicated medical scientists headed by a clinical neurologist with extensive experience in CJD was entirely correct. The CJDSU was a research unit whose work deserves nothing but praise.

5.171 It is unlikely that existing institutions responsible for monitoring public health would have achieved the same result in as short a time. The OPCS (later the Office of National Statistics) is experienced in monitoring health trends but is largely dependent on death certification. While this may be sufficient for indentifying trends in relatively common causes of death, the procedure is inefficient for rare complex disorders in which diagnosis is difficult. The correct certification of causes of death is notoriously unreliable and especially so for a condition like CJD which can be mistaken for other more common neurodegenerative disorders such as Alzheimer's disease. In such circumstances, the underlying diagnosis may even be omitted and the immediate cause of death recorded, for example bronchopneumonia.

5.172 There has been much discussion about whether or not the PHLS could have provided surveillance of CJD. The PHLS was established in 1940 to monitor infectious disease in the community and to provide laboratory diagnosis for many zoonotic disorders. They have a network of laboratories throughout England and Wales and many other resources including epidemiologists, statisticians and field workers. They have access to medical registers in the NHS and close links with clinicians. However, in 1989 they did not have expertise in CJD and most importantly there was (and still is) no established laboratory test for either CJD screening or for diagnosis in suspect cases. None the less, we believe that, short of undertaking CJD surveillance themselves, the PHLS could have given considerable assistance to the CJDSU. They might have provided expertise in epidemiology based on their experience of work in the field of investigating disease outbreaks. Their field epidemiologists could have been involved in designing questionnaires, interviewing relatives and in collecting unbiased answers designed to explore aetiogical factors. They had the resources to conduct active surveillance throughout the UK through polling neurologists and other relevant groups of clinicians and compiling null returns in addition to suspect cases. They could have provided access to medical registers, hospital in-patient and out-patient statistics and an efficient system of communication within the PHLS network. They could have helped to look for potential routes of infection between cattle and humans, for example by investigating possible associations with vaccines prepared from bovine products. In these and other respects, the PHLS could have contributed to the CJDSU surveillance of CJD and their programme of epidemiological research. These comments in no way detract from the sterling work of the CJDSU team who so promptly detected the emergence of vCJD and so efficiently established the clinical and pathological characteristics of the disease. While assistance from the PHLS could have been valuable, it would not have enabled identification of vCJD at any earlier date. We do not criticise those who concluded that the task of monitoring CJD should be left to the Surveillance Unit.

Was there adequate communication of information to the Department of Health?

5.173 We now turn to consider whether the communication of information from the CJD Surveillance Unit to the Department of Health was adequate. The only matter that gave us concern in this regard was Dr Will's reply to Mr Skinner on 23 February 1996. In his reply to Mr Skinner on 23 February 1996, Dr Will stated that 'There is a remote theoretical risk that BSE in cattle might cause disease in humans' and 'that this is likely to be negligible provided statutory measures are fully enforced'. He added that he agreed with the statement that 'there is currently no scientific evidence that BSE can be transmitted to humans and that eating beef causes CJD'. Dr Will stressed to us that at this time there were 'a whole range of issues that had to come together' before he reached a stage at which he considered there was sufficient evidence to make a scientific judgement that there had been a change. This range of issues included genotyping results, reaching more conclusions about the neuropathology and the clinical phenotype and other issues. The word 'currently' in Dr Will's reply to Mr Skinner did not preclude the materialisation of any new evidence. In the event, that new evidence was announced only four weeks later.

5.174 We have given our reasons above (paragraph 5.170) for concluding that the CJD Surveillance Unit could not have informed SEAC of the emergence of a new version of CJD earlier than 8 March 1996. It was not until about 13 March that Dr Will was sure that the novel clinical phenotype had not been reported outside the UK, that the distinct neuropathological findings had been confirmed by independent pathologists, and that prion protein gene mutations were not involved. Until this information had been assembled Dr Will did not feel in a position to make a scientific judgement that there was evidence that BSE could be transmitted to humans.

5.175 However, we note that some inkling of what was to come is to be found in the minute dated 1 March from Mr Eddy to Mr Haddon, Mr Meldrum, Mr Taylor and Dr Render, in which it is indicated that the Department of Health had been informed that the results were beginning to look rather firmer and that there was a new sub-population of the disease emerging. We do not think that Dr Will was able to go this far on 23 February and we consider that his presentation to SEAC on 1 February had given Mr Skinner a fair picture of the degree to which his thinking had progressed. In these circumstances, we do not think that Dr Will can be blamed for continuing to be cautious in his reply to Mr Skinner on 23 February 1996. ²

Was involvement of the CJDSU in the public debate about the safety of beef appropriate?

5.176 The difficulties that arose from involvement of the CJDSU in the public debate are illustrated by the Parliamentary Food and Health Forum on 20 February 1996 at which Dr Will was an invited speaker. At that meeting Dr Will found himself obliged to deal with the question whether the age distribution of CJD appeared to have changed in the past year. During his talk, he stated that current evidence did not suggest any definite change in the pattern of CJD that could be attributable to BSE and that there was no overall change in the age-specific incidence of CJD in the UK. He was recorded in a note of the meeting as stating that current information was reassuring and that any risk to man from BSE might not be evident for some years. Four weeks later, results of key investigations had been completed and different conclusions had been drawn.

5.177 Dr Will stated to us that his view at this time was that there was no overall change in the age specific incidence of CJD in the UK. He stated that this view was based on the statistical analysis provided by Dr Gore on 5 February 1996. Dr Gore's calculations were based on the first of four young cases of confirmed CJD (aged 16, 19, 29 and 30 years) which had occurred during the 5-year-period between 1990-96. When considered in relation to the three young cases (aged 35, 36 and 38 years) occurring in the previous quinquennium, the probability of observing four cases was not expected. No account was taken of the younger ages of the four cases, nor of the fact that by 20 February 1996, eight cases under the age of 40 were known. Statisticians approach these matters by attempting to assess the likelihood that the cases in question may represent no more than random events. If the probability of the cases in question being a random event is less than 5 per cent then they would normally be thought to have statistical significance. The probability of eight cases occurring by chance was now 0.03, ie within the 5 per cent probability to which statistical significance is generally ascribed. It was therefore not strictly correct to conclude that there had been no statistically significant change in the age distribution of cases in the UK.

5.178 Although the numbers had just achieved formal significance levels, we believe that Dr Will had reason to be cautious about the implications of his findings. Before coming to a conclusion that there might be a link with BSE, it was necessary to make sure that the eight cases had a consistent phenotype. Dr Will was correct in waiting for the results of genetic analysis on 4 March 1996 and for the independent expert review of the pathological material. He had to be sure that similar cases had not occurred outside the UK during the same period. It was this that provided the circumstantial evidence of a link with the BSE epidemic. All the pieces of the jigsaw puzzle were not in place until mid-March 1996 and Dr Will did not feel confident to suggest the existence of a link with BSE until they were.

5.179 Dr Will has explained his reticence at the Food and Health Forum meeting about his concerns about CJD in young patients. He believed that it would be premature to make a public announcement before he was satisfied that all aspects had been fully examined and the results peer-reviewed. An announcement of a change in numbers of young patients over such a short period might have precipitated a crisis of some magnitude based on the unsupported assumption that the changes were due to BSE. It was imperative that other explanations had to be rigorously investigated before coming to such a conclusion.

5.180 Undoubtedly the recognition of an increasing number of young people affected with CJD in late 1995 and early 1996 placed the CJDSU in a difficult position if asked publicly about a possible link with BSE before investigations on the patients had been completed. The position was particularly acute for Dr Will on 20 February. Although he was concerned about the cases of CJD in young people, he felt he could not publicly refer to the change in the number of young patients.

5.181 We have some sympathy for Dr Will in the situation in which he was placed. The CJDSU are to be commended on the sound work that they had done in identifying the new variant of CJD. It is unfortunate that Dr Will found himself in a position where he was required to speak in public about the risk that BSE might cause disease in humans. With hindsight, it might have been better for him to have refrained from making any comment, and from participating in the Food and Health Forum, until his investigations were complete. Given the position in which he found himself, we do not criticise Dr (now Professor) Will.

Is a link between BSE and vCJD established?

5.182 As set out in more detail in vol.2: Science (Chapter 4), the link between BSE and vCJD, suggested on the circumstantial evidence that the two conditions were associated in both time and space almost exclusively in the UK, was soon supported by experimental and biochemical results. Chief among these were the strain typing results in mice which revealed in 1997 that BSE and vCJD had similar incubation periods and lesion profiles in the brain. Type 4 glycosylation patterns were found in both, and studies using transgenic mice in which the murine prion gene had been replaced by the bovine gene strongly suggested that the infective agent was identical in both conditions. Present evidence seems overwhelming.

Some observations about vCJD that call for explanation

Why does vCJD seem to be a disease of young people?

5.183 In sporadic CJD the median age at death for the period 1995-99 was 65 years. ³ The median age at death in vCJD, based on the experience to 31 December 1999, is 29 years. ⁴ The median age at onset in vCJD is 28, with the youngest case aged 14 years while the oldest case was aged 53 years. ⁵ We have identified two main theories about the age of current cases that have led to much speculation. These are diet and the use of vaccines.

5.184 In the case of diet it has been suggested that various types of processed meat including beefburgers, sausages and meat pies may have contained beef brain and spinal cord as ingredients. No tissues from BSE-affected animals should have entered human food after the introduction of the slaughter and compensation policy in August 1988. Brain and spinal cord from cattle incubating BSE ought not to have entered human food after the SBO ban in 1989. However, spinal cord may have done so, along with dorsal root ganglia, as contaminants of mechanically recovered meat (MRM) obtained from bovine vertebral columns up to December 1995. MRM might also have been included in baby foods, although to what extent is uncertain. We refer in vol. 2: Science (Chapter 4) to a study in which the consumption of processed beef products was examined in relation to age in a group of individuals aged 15 years and over. In the case of beefburgers, the age distribution fell steeply with age, and was to some extent correlated with age distribution of vCJD patients. The consumption of beef, sausages and meatpies was not age related. The suggestion is that the risk of infection is greater in children and young adolescents because their diet includes more beefburgers than older individuals.

5.185 Vaccines have been identified as a possible source of infection because, although they do not directly contain bovine ingredients, tissue culture media such as foetal calf serum or bovine serum albumin or, less commonly, bovine brain and spleen, may be used during their production. We discuss in vol. 7: Medicines and Cosmetics steps taken to close this possible pathway of infection. We also discuss in that volume tests undertaken by the NPU to ascertain whether bovine blood and serum could transmit infectivity to mice. The results of this particular piece of research were negative, which provides some reassurance about the safety of vaccines. Nevertheless, these tests were not conclusive and the possibility that such material may have transmitted infection calls for consideration:

1. During childhood, vaccines are routinely administered at three stages: namely infancy, the 4-5-year-old age-group and the 10-13-year-old age-group. A starting-point in considering whether vaccines are implicated in transmission is to look at whether any known cases of vCJD might be attributable to childhood vaccination at these ages. It has in the past been suggested that this is unlikely, in particular because of the widely held belief that the earliest cases of BSE did not occur until the early 1980s. As the majority of patients with vCJD were born before 1980 (the youngest in 1985), some have assumed that vaccination in childhood would have been completed in most cases well before vaccines could possibly have been contaminated.
2. However, current epidemiological evidence (see vol. 2: Science) is consistent with the possibility that there was a small number of cases of BSE as early as 1970-72, and that these were followed by three successive waves of the disease in 1975-77, 1981-84 and 1986-87, as a result of recycling of BSE in MBM. Batches of vaccines manufactured during the 1970s cannot therefore be ruled out as a source of infection merely because of their date of manufacture. Patients with vCJD born before 1960 are unlikely to have been infected by childhood vaccination, but the possibility of infection via vaccinations in adulthood may merit consideration.
3. These possibilities should be taken into account in an analysis of the specific batches of vaccines administered to victims of vCJD, and consideration given to whether there is any common batch or other factor.

5.186 An ad hoc Working Party of the Committee on Safety of Medicines which considered BSE-related issues associated with the use of seedlots ⁶ in vaccines noted that all vaccines in current use are prepared using bovine materials sourced from BSE-free countries (mostly Australia and New Zealand). UK-sourced bovine material was used in a number of vaccines prepared from 1980 to 1989 and distributed up to 1992. Other vaccines and seedlots were prepared prior to 1980, but the sources of bovine ingredients in these have not been identified. The Working Party noted that both human and cattle vaccines prepared with UK-sourced bovine materials have been widely distributed throughout Western Europe without apparently being associated with outbreaks of either vCJD or BSE. This is taken as a strong argument against the possibility that vaccines have been vectors of either disease.

5.187 It will be apparent that some of the assumptions made by the Working Party of the CSM are open to question for reasons we have set out in our Report. We hope that government will look at the topic again in the light of our comments.

Is there evidence of clustering of vCJD cases?

5.188 The 1999 Annual Report of CJDSU draws attention to the geographical distribution of cases of vCJD (see Figure 5.1, in which the data have been further updated) and shows that if the UK is divided into 'North' and 'South' regions, the rate per million individuals aged between 16 and 54, is 2.57 for the 'North' region and 1.30 per million individuals for the 'South'. An early suggestion of a cluster of four cases in Ashford, Kent was considered not to have significance. ⁷ More recently, a group of four deaths from vCJD in and around the village of Queniborough, Leicestershire, plus another probable case from the same area, has aroused intense media interest. Both the CJDSU and PHLS are making detailed local enquiries. We hope that the opportunity to investigate such clusters will provide information on common factors and lead to a firm conclusion about the source of infection and its route to affected patients. Such knowledge will improve understanding of the disease and hopefully lead to strategies for treatment.

Figure 5.1: Geographical distribution of places of residence of vCJD cases at onset of symptoms

Why have all cases of vCJD tested to date been homozygous for methionine at codon 129 on the prion protein gene?

5.189 This is presumably because this particular genetic polymorphism is a susceptibility factor to BSE infection. We know that 70 per cent of sporadic CJD patients are similarly homozygous for methionine at codon 129. However, 39 per cent of the general population shares this genotype, of which only a tiny number have succumbed to CJD of any type. This suggests that methionine homozygosity at codon 129 is only one of a number of susceptibility factors. Other yet undiscovered genes outside the prion protein gene locus might also confer susceptibility. If these postulated genes have a major effect on susceptibility this could lead to familial clustering. Alternatively, it might indicate that only a small proportion of the susceptible population was exposed to BSE infection in the 1970s and 1980s, as might be expected with only limited distribution of the agent in food or vaccines.

5.190 It has been observed in iatrogenic CJD due to human growth hormone contamination, that the age of onset is earlier in methionine homozygotes than in valine homozygotes. This raises the possibility that cases of vCJD with other genotypes at codon 129 will emerge in due course in older patients who have been incubating it for longer. We note that the CJDSU is alert to this possibility, and also to the possibility that the clinical phenotype in such cases might differ from present cases of vCJD.

Why have all patients with vCJD so far investigated shown signs of high levels of infectivity in tonsils and other parts of the lymphoreticular system, whereas cases of sporadic and iatrogenic CJD have not?

5.191 It has been postulated that this observation implies infection via the gastrointestinal tract in vCJD, while iatrogenic CJD is due to parental inoculation, and sporadic CJD arises by somatic mutation or by toxic conversion of PrP^C. It has also been suggested that infections of the tonsils and gastrointestinal tract due to other pathogens may increase susceptibility to TSE agents by increasing their absorption through inflamed tissues. Studies on the incidence of past disease in subjects and controls might shed light on this possibility.

Is occupation a risk factor in vCJD?

5.192 One of the original proposals in the CJD surveillance project was to monitor occupational groups exposed to BSE-affected cattle and their products. Such groups include farmers, veterinarians, slaughtermen and butchers. This part of the project was given a low priority by the Tyrrell Committee and was not implemented. It was felt that rather than set up longitudinal study of a fixed number of individuals in each group, together with matched controls, it would be adequate to take an occupational history of each CJD case at the time of referral.

5.193 From 1990 to 1996, the CJDSU had referred to it four farmers affected with CJD who were known to have had cases of BSE on their farms. Assuming a total of 155,000 dairy farmers in the UK, ⁸ the number of observed CJD cases is significantly higher than expected from population estimates. Counting only those farmers with affected cattle, the probability of observing four or more confirmed cases of CJD is estimated at less than one in 10,000. ⁹ In addition, two farmers' wives were known to have CJD from farms in which clinical BSE had not been reported (although preclinical cases of BSE on these farms might have been expected).

5.194 The affected farmers were aged between 54 and 64 and had signs and symptoms typical of sporadic CJD. Two had EEG changes typical of the sporadic disease and all four had type 2 glycosylation patterns. Three farmers were homozygous for methionine at codon 129 and the fourth was a valine homozygote. None conformed to the phenotype characteristic of vCJD. The findings remained unexplained, although a European collaborative study showed a similar increased incidence in deaths due to CJD in farmers in several member states. It was noted that unexpected numbers of affected individuals occurred in other occupational groups, such as the clergy, but numbers in each occupation remained small.

5.195 Among occupational groups exposed to BSE, farmers remain unusual in having such an excess over the incidence of CJD for the population as a whole. No cases of CJD have been reported amount veterinarians exposed to BSE. Four people in the meat industry (butchers, abattoirs, rendering plants, etc) have been reported to have vCJD. ¹⁰ The present evidence has been accepted by some as reassuring in that such occupations may not pose as serious a risk as might have been expected.

How many people are likely to succumb to vCJD?

5.196 The last available figure from the CJDSU before going to press put the total number of definate and probable cases of vCJD at 82, as of 4 September 2000. Of these 69 were confirmed deaths due to vCJD, five were deaths probably due to vCJD, and eight were probable cases of vCJD still alive. The number of patients per year is shown in Figure 5.2 (updated from the 1999 Annual report of the CJDSU). More recently an upward trend has been estimated, showing that deaths have been increasing by 33 percent per year - consistent with an emerging epidemic. ¹¹ However, due to uncertainty about the length of the incubation period, the routes of infection, genetic susceptibility and resistance, and about the environmental factors which might operate, predictions of the size of the epidemic have encompassed a wide range from a hundred to tens of thousands of individuals. A recent immunohistochemical study ¹² on a random sample of over 3,000 tonsil and appendix specimens has failed to detect any containing PrP^Sc. This negative finding gives little comfort as the number of samples tested is relatively small. Furthermore, as the authors of the paper on this study noted, if it is assumed that their test were able to detect infection in the last 75 per cent of the vCJD incubation period, then the upper bound on epidemic size is reduced from several million cases to about 150,000 cases. However, if their tests were only able to detect infection in the last 50 per cent of the incubation period, their results do not reduce the previously reported uncertainty in epidemic size. Therefore, at the time this Report goes to press, it is too early to predict the outcome on the basis of the present number of cases but this will become clearer with time. ¹³

Figure 5.2: Cases of vCJD by date of onset, notification, death and confirmation