Volume 3: The Early Years, 1986-88
3. Epidemiology
Discussion
Was there delay in identifying feed as the cause of infection?
Why was the feed infectious?

3.63 This completes our summary of the epidemiological investigations up to the time that the decision was taken to recommend a ruminant feed ban. Eighteen months had elapsed since the initial identification of BSE and 11 months since Mr Wilesmith's investigations began. Over 11 further years have now passed. While there is a general consensus among the scientists that MBM was the vector responsible for the spread of the BSE epidemic, it is remarkable that controversy still exists about the origin of the BSE agent.

3.64 We consider that the origin of the BSE agent remains uncertain. With hindsight, the most likely possibility is that it is a novel TSE agent arising in either sheep or cattle. This is discussed further in vol. 2: Science. We have, however, in that volume felt it right to explain our reasons for concluding that Mr Wilesmith was correct in identifying animal feed, and more particularly the MBM component of that feed, as the vector for the epidemic of BSE. At this point we pause to consider, as part of our duty to comment on the adequacy of the response to BSE, whether Mr Wilesmith's investigation in the period up to the beginning of May 1988 was carried out with due speed and diligence.

Was there delay in identifying feed as the cause of infection?

3.65 Should Mr Wilesmith have been involved earlier? He told us that an epidemiologist likes to be involved at the outset of the emergence of a new disease, and to have an input into any circulars sent out seeking information about its incidence. ¹ When the decision was taken to ask Mr Wilesmith to assist in investigation, six cases had already been confirmed by histopathology, involving four different farms. A further 13 cases were suspected on these farms on the basis of clinical criteria. Reports of a similar clinical syndrome occurring on other farms had been received by some VICs. ²

3.66 We do not consider that Mr Wilesmith's colleagues at the CVL can be criticised for not seeking his assistance at an earlier stage. The resources of the Epidemiology Department were limited and could not be directed to every suspected new animal disease. It was not until the stage was reached at which a number of suspect cases had been identified that the skills of the epidemiologist had a significant contribution to make. Where we have been critical is in respect of the delay in reaching that stage, which resulted from the restrictions on dissemination of information about BSE in the early days (see Chapter 2).

3.67 The next matter that we have considered is whether the identification of MBM as a front-runner in causing BSE could not and should not have been made more swiftly by seeking the assistance of the rendering and animal feed trade associations and their members at the outset. We have drawn attention to the emphasis that was placed on the need for informal approaches and the preservation of confidentiality about the theory that feed was the cause of BSE. We note Mr Bradley's comment 'if we approach too early concealment of information is likely' (see paragraph 3.16 above). In so far as MAFF's reticence was attributable to a fear that an open approach might result in a refusal to provide the information sought, we have some sympathy with this. Indeed, we find that the extent of cooperation provided by UKRA and UKASTA and the leading members of those associations, while commendable, was surprising. The theory MAFF was seeking to explore was that feed compounders and renderers had been marketing products that were so unfit for their purpose that they were infecting the cattle to which they were fed. The minutes of the meeting between MAFF, GAFTA, UKASTA and UKRA on 8 April 1988 end:

UKASTA offered help in tracing the source of feed supplied to affected herds.

It was emphasised that this exercise was not intended to find a 'culprit' but to pinpoint any fault in the system and jointly seek a remedy. There would be no public statement until the proposed sampling operation was complete and results assessed.

UKASTA asked what the legal position would be if suspect plants were identified. They understood that there could be no product liability because of the lack of scientific evidence at the time the contamination occurred. However, they would welcome clarification on this point and a statement which would discourage spurious claims from farmers. ³

3.68 It seems that no response was made to the request for clarification of the legal position. MAFF was not in a position to provide reassurance. The reality was that, if the link between BSE and animal protein had been established, those who could be shown to have supplied an infective product would have been susceptible to claims for damages by those to whom the product had been sold. We do not believe that those who insured compounders and renderers would have wished their clients to provide information that might have had this result. Much later in the story the CVL's Epidemiology Department sought to investigate the processes involved in the compounding of commercial feedstuffs in order to identify possible risk factors that allowed the inclusion of ruminant-derived protein in cattle feed. The feed compounders took legal advice and, on the basis of that advice, denied access to this information. A similar reaction might well have been experienced in relation to Mr Wilesmith's enquiries in 1988, and MAFF had no statutory powers to require information. MAFF officials are not to be criticised for treading gingerly in conducting their investigation.

3.69 Mr Meldrum has, however, informed us that, as far as he was concerned, any fear that an open approach might result in a refusal to provide information sought 'did not come into the decision to make informal approaches to the rendering and feed industries'. ⁴ His explanation was: 'When I was asked about these "off the record" meetings I commented that it was the general agreement of the meeting on 15 December 1987 that they should be "off the record"', because:

We were still considering the hypothetical, indeed the possibility that meat and bone meal could be the cause of BSE; and whilst those investigations were being taken forward it seemed wise to keep them confidential . . . When you are dealing with a very early investigation or some very early reports, it was better to keep the investigation to yourself and release the information in a proper and reasoned and planned way when it becomes substantive. ⁵

3.70 When asked what the result would have been if, in December 1987, the State Veterinary Service had gone on record in its dealings with the feed industry, disclosed that MBM was suspected as the cause of BSE, and asked for assistance, he replied:

I think in one word confusion, because in December [1987] we were still at a very early stage in the investigations. If we had made that information public in a disorderly way I think it would have caused significant confusion both to the renderers, feed industry and to farmers, not least to the public at large. It would not have been very helpful. ⁶

3.71 We have not found the logic of these comments easy to follow and feel that they may be indicative of a general approach to the management of information.

3.72 While, with hindsight, we are inclined to think that a direct approach to UKASTA in November or December might have saved some time, we are not disposed to criticise the information-gathering technique adopted by Mr Meldrum and his colleagues. It resulted in Mr Wilesmith's obtaining the information that he was seeking and making a correct diagnosis as to how it was that cattle were contracting BSE.

3.73 On 9 and 10 September 1991 a Visiting Group consisting of two distinguished epidemiologists, Professor Roger Morris and Professor R A Robinson, conducted a review of the Epidemiology Department of the CVL. In their report, they described the Department's investigation into the cause of BSE, of which we have so far only considered the first phase, as 'a careful and valid epidemiological approach to identifying the vehicle of transmission of BSE, ie meat and bone meal contaminated with a scrapie-like agent'. ⁷ They gave high commendation to the work of Mr Wilesmith and his team:

We believe the staff of the Epidemiology Department should be recognised as having moved quickly and effectively to define the nature and course of the epidemic, and to identify the mechanism of transmission in the absence of almost all the diagnostic aids which would normally be called on in such an investigation. This problem ranks among the major disease puzzles of the century, and has been solved by application of epidemiological techniques far faster and more definitively than virtually any other equivalent problem in human or veterinary medicine. It will prove to be a landmark example which will be used for many years to demonstrate how epidemiological methods can be applied to unknown diseases.

It is important that Department staff take pride in their achievement in this area, because enormous progress was achieved with minimal resources as the epidemic unfolded. ⁸

3.74 When giving evidence to us, Mr Wilesmith described his achievement in more modest terms:

It really was by exclusion of all those other things quite simply that we were left with - [it] had to be the meat and bone meal . . . It was not anything more clever than that, I just excluded all those other things. ⁹

3.75 Mr Wilesmith accurately describes the process that led him to conclude that MBM was the source of infection with BSE. We would, however, echo the commendation of the Visiting Group. It took Mr Wilesmith about six months to identify the vector of the BSE epidemic, by a process of elimination, and gain sufficient confidence in the identification to recommend appropriate remedial action. In May 1988 his conclusions were still tentative to a degree. His belief was, and it was a logical belief, that the common source of the outbreak of BSE must be some novel element affecting the nature or the quality of some of the MBM that was being incorporated into cattle feed. What that element was remained unclear and, until it was identified, further epidemiological research was needed. In the meantime the achievement of the modest Epidemiology Department of the CVL in identifying the probable source of infection in the space of six months constituted an impressive response to the challenge of BSE.

Why was the feed infectious?

3.76 Mr Wilesmith had concluded that the cases of BSE that were breaking out all over the country could not have been caused by a single point source of infection. He thought that each case was an index case and that there was a common source of infection. The probable source was scrapie-contaminated MBM. Scrapie-infected sheep must, however, have been a constituent of the MBM which went into cattle feed for decades. Why had it only started to transmit to cattle in about 1982? And why was this transmission only occurring in Great Britain when scrapie-infected sheep were undoubtedly being rendered down and incorporated in cattle feed in other parts of the world?

3.77 Mr Wilesmith suspected that the answer might lie in the rendering process. In July 1988, with the help of Mr Martin Atkinson, who was then Deputy Regional Veterinary Officer for Midland and Western Region, he developed a questionnaire for a proposed survey of all rendering plants. This sought information from the rendering plant about methods and, in particular, about the raw materials and time/temperature of the cooking process on a year-by-year basis, between 1980 and 1987. As an afterthought, questions about the use of solvent extraction were added (see vol. 13: Industry Processes and Controls for a description of solvent extraction).

3.78 The survey was carried out by three Veterinary Officers, Mr David Mouat, Mr Andrew Proud and Mr Paddy Grant, who gave oral evidence to the Inquiry. ¹⁰ Their voluminous Survey Reports are in evidence. ¹¹

3.79 In a paper published in the Veterinary Record on 17 December 1988 Mr Wilesmith advanced the following comments on the cause of BSE:

There was no clear or single explanation why, in 1982, cattle apparently became first exposed to a transmissible agent sufficiently to result in clinical disease. A number of factors have been identified which when combined are undoubtedly significant in the occurrence of this epidemiological phenomenon. These include: a dramatic increase in the sheep population in Great Britain which commenced in 1980 and has continued (MAFF 1988); a probable increase in the prevalence of scrapie infected flocks(J. W.Wilesmith unpublished data); the greater inclusion of sheep heads in material for rendering; the greater inclusion of casualty and condemned sheep in material for rendering as a result of the reduction in the number of knacker's yards; the introduction of continuous rendering processes during the 1970s and 1980s which may have resulted in the rendering of animal material at a lower temperature and, or, a shorter time than previously and the decline in the practice of using hydrocarbon solvents and terminal heat treatment for fat extraction since the mid 1970s (MMC 1985).

These factors provide a possible explanation for a change in the exposure of cattle to sheep-derived protein and the scrapie agent. ¹²

3.80 When, however, he came to analyse the results of the rendering survey in January 1989, he concluded that the change from batch rendering to continuous rendering processes could not have been a relevant factor. The change had not taken place during the relevant period and the potential for heat penetration and thus an inactivating effect on the scrapie-like agent was probably greater under continuous rendering than under the batch processing systems. The cessation of the use of solvent extraction seemed, however, a more likely explanation for the development of infectivity in MBM in around 1981-82.

3.81 In a paper published in the Veterinary Record on 2 March 1991 Mr Wilesmith included a graph which, he suggested, showed:

The cessation of hydrocarbon solvent extraction of fat from meat and bone meal is consistent with the temporal estimate of the onset of exposure. ¹³

3.82 He went on to explain that the rendering survey indicated that the rendering process had 'seldom if ever' produced sufficient heat to deactivate the most heat-resistant strain of scrapie. He continued:

This suggests that cattle may have always been potentially exposed to a scrapie-like agent via meat and bone meal and the occurrence of BSE cannot therefore be explained simply by a change from total inactivation of a scrapie-like agent in meat and bone meal to partial inactivation, but the extent of the partial inactivation may well have been reduced by the move away from solvent extraction. The use of hydrocarbon solvent extraction was an additional, and in the main, terminal process whose removal has not resulted in a change in the basic heat treatment of material which is the application of dry heat, albeit in a fat rich environment, rather than moist heat. The solvent extraction process did, however, include the application of steam to the low fat content meat and bone meal. This component of the solvent extraction process therefore represented not only an additional application of heat, but a more effective heat treatment with respect to the inactivation of the scrapie agent. This process may not have totally eliminated a scrapie-like agent, but could have reduced the titre sufficiently to prevent clinical disease occurring during the normal lifetime of cattle, as the incubation period of the scrapie-like infections is related to the dose of agent received . . .

The cessation of this hydrocarbon solvent extraction process may have increased the contamination of meat and bone meal with a scrapie-like agent in two ways. First, the loss of any additive effects on inactivation by the combination of heat plus solvents and, secondly, the loss of an additional moist heating stage. This represents a biologically plausible hypothesis to explain how cattle became exposed to a scrapie-like agent sufficient to result in clinical disease. The magnitude of the stepwise reduction in the proportion of meat and bone meal produced by solvent extraction represents a major and sudden change in the potential exposure of cattle to a scrapie-like agent and is consistent with the abrupt change in the effective exposure as indicated by the epidemic curve and the predicted time of onset of this exposure in 1981/82. ¹⁴

3.83 By way of conclusion Mr Wilesmith suggested the following hypothesis:

As suggested previously, the epidemiological data are not consistent with a hypothesis involving the emergence in the sheep population of a novel strain of scrapie which is pathogenic for cattle. Epidemiological results obtained so far are consistent with the hypothesis that the occurrence of BSE in cattle has been the result of an increase in exposure to the scrapie agent, via ruminant derived protein in feedstuffs. This implies that cattle have always been susceptible to the scrapie agent but their exposure has, in the past, been insufficient to result in a detectable incidence of clinical disease. The findings are also consistent with an increase in exposure of cattle, as a result of the change in the rendering process, to a cattle 'adapted' strain of a scrapie-like agent which had been present in the cattle population for some time. Again in this case the infecting dose would previously have been insufficient to result in clinical disease occurring during the normal lifetime of cattle at a detectable incidence. ¹⁵

3.84 This theory has survived almost to the present day, as one possible explanation for the emergence of BSE. When we first learned of this theory, it seemed to us unlikely that there had been a change, or a combination of changes, in rendering that were peculiar to Great Britain and yet so nearly universal within Great Britain that they permitted an almost simultaneous transmission of scrapie to cattle throughout the country, or, alternatively, the emergence of a widespread disease that had hitherto lain latent in cattle. The theory was, it seemed to us, only viable as the least unlikely explanation for the emergence of an extended common source epidemic. Thus we were anxious to explore whether epidemiological modelling made it possible to postulate that there may have been a single point source for the BSE epidemic. Having investigated this in Phase 2 of the Inquiry hearings, we have concluded that it is indeed a possibility. We have consequently reached the following conclusions as a matter of probability, though not of certainty.

3.85 The source of BSE is not a long-established scrapie agent which would have given rise to an extended common source epidemic. BSE is a novel and unique TSE which probably started as a consequence of a genetic mutation in a single animal and which has been propagated as a consequence of recycling by rendering and incorporation in cattle feed. This would constitute a point source for the epidemic.

3.86 The reasons for these conclusions are developed in detail in vol. 2: Science. Shortly summarised, they are:

1. BSE is a single strain, which differs from any known strain of scrapie.
2. Scrapie has many strains. It is hard to conceive of near simultaneous transmission to cattle all over the country resulting in a single strain of BSE.
3. Increases in the British sheep population or in the incidence of scrapie do not impress as a plausible factor. The sheep/scrapie content of individual batches of rendering must always have varied widely. The comments made by Mr Peter Carrigan (Director of a major tripe manufacturer) were cogent. ¹⁶
4. Use of solvent extraction was far from universal prior to 1980.
5. Tests have shown that solvent extraction has no significant effect on deactivation of the scrapie agent.

3.87 We do not consider that Mr Wilesmith is to be criticised for concluding that the emergence of BSE was attributable to an extended common source or that that source was scrapie. The sophisticated techniques that have demonstrated that BSE could have had a single point source went beyond the expertise of the CVL's modest Epidemiology Department. Furthermore, we question whether any epidemiology department, in the early days of BSE up to the end of 1989, would have concluded that it might have had a single point source.

3.88 In his statement to the Inquiry Mr Wilesmith referred to a review of the epidemiology - in which he was assisted by Professor Robert Curnow, an applied statistician, and the strain typing results of Dr Moira Bruce - which established that all BSE cases that were tested had a single strain, including one case that had most likely been infected in the early 1980s. This had led Mr Wilesmith to alter his views on the likely source of infection for cattle during the early years of the clinical epidemic. He explained:

Thus, while I still subscribe to the sheep scrapie hypothesis as the explanation for the origin of BSE (the origin is, of course, the subject of continuing research), it appears that the BSE strain was selected quite rapidly by cattle and that the clinical epidemic of BSE is essentially due to the recycling of infected cattle tissues via MBM. ¹⁷

3.89 The erroneous conclusion that the cases identified up to 1988 were index cases (ie, first cases in a specified group), probably infected from scrapie as a common source, affected estimates that were made of the future incidence of the disease. Mr Wilesmith felt unable to estimate the effect of the recycling of BSE itself and contemplated the possibility that this might not be significant. Had it been recognised that the cases being identified were the product of recycling BSE, which could well have been in progress for over a decade, the probability of a rapid escalation of clinical cases would surely have been appreciated. As it is, at the time that MAFF introduced a ruminant feed ban, it was believed, on the basis of Mr Wilesmith's advice, that cases of cattle infected before the ban might well continue at a constant rate (approximately 60 clinical cases a month). In fact, the reality was that the rate at which cattle were being infected was destined to produce clinical cases at a rate of over 3,000 a month.