VOLUME 16 NO1 2006 10 TB in domestic species other than ... · GVJ Production Team Internal Communications and Stakeholder Management Team Department for Environment Food and Rural

www.gvs.gov.uk

VOLUME 16 | NO1 | 2006

GVJGovernment Veterinary Journal(formerly known as the

State Veterinary Journal)

Bovine TB Special Edition

PB 12125 Nobel House17 Smith SqaureLondonSW1P 3JR

About Defrawww.defra.gov.uk

ContentsForeword 4

1 TB policy developments 5

2 Some lessons from the history of the eradication of Bovine Tuberculosis in Great Britain 11

3 Bovine Tuberculosis in the European Union and othercountries: current status, control programmes andconstraints to eradication 19

4 Bovine TB: Modelling and predicting its distribution in GB using CTS data 46

5 The laboratory diagnosis of Bovine Tuberculosis 53

6 The introduction of pre and post-movement TB testing in Scotland for cattle from high incidence TB areas 58

7 Ante mortem diagnosis of Bovine Tuberculosis: thesignificance of unconfirmed test reactors 65

8 The BOVIGAM® assay as ancillary test to the Tuberculin skin test 72

9 Tuberculous pneumonia and BVD in housed calves 81

10 TB in domestic species other than cattle and badgers 87

Bo

vine TB

Special Ed

ition

Go

vernm

ent V

eterinary Jo

urn

alV

OLU

ME 16 |

NO

1 |2006

10493 AW Defra GVJ_03-52 29/8/06 14:27 Page 1

91VOLUME 16 | NO1 | 2006 Government Veterinary Journal

mycobacteria. However, very few catswith mycobacterial infections haveactually tested positive for FIV or feline leukaemia virus (FeLV). Incidence Between 1980 and 2003 M. bovis wasisolated from cats on 41 occasions atVLA. The incidence of tuberculousdisease in cats caused by M. bovis hasbeen sporadic in GB. However in 2005the VLA received 90 supect felinesamples, 13 of which were positive for M. bovis and 23 for M. microti-like or M. avium-complex organisms. This ismore than double the recorded incidencein previous years.

ConclusionAmong domestic animals in Great Britain, pigs, horses, sheep, cats, dogsand camelids are all considered spilloverhosts as is perhaps the case with all othermammalian species (badgers, cattle anddeer excluded). In other words, thesespecies become infected only when thechallenge level is relatively high, and theycannot sustain the infection within theirown populations in the absence ofinfected cattle or a wildlife reservoir. This does not mean that, if infected,these species cannot occasionallytransmit the disease to other animals and humans, i.e. some of them may actas amplifier hosts.

TB is particularly rare (but not unheardof) in horses and sheep, which should be considered true dead-end hosts.

Pigs, farmed wild boar, dogs, cats andcamelids should be treated as potentialamplifier hosts of TB.

AcknowledgementsMuch of this article has been edited fromthe SVS standing instructions on Bovinetuberculosis – Viper Ch 23 section Yprepared by R. de la Rua-Domenech.

4

Figure 4Popliteal LymphNode – Cat

l Information

Wyn Buick is aVeterinary Officerwith the SVS,Carmarthen.

TB in domestic species other than cattle and badgers

The Government Veterinary Journal (GVJ) is the official journal of theGovernment Veterinary Service and those who support its work or have aninterest in state veterinary medicine. It is compiled and produced by theDepartment for Environment, Food and Rural Affairs (Defra) for, and on behalf of veterinary surgeons and those who support them across all parts of Government.Its key aims are to enhance the contribution of veterinary expertise within andacross government, promote the work of Government veterinary surgeons andprovide a range of technical, factual and interesting articles in the fields of:

• Disease control • Animal welfare • Public health • Consumer protection

It is intended that the GVJ will also highlight progress in relation to the AnimalHealth and Welfare Strategy and focus on issues identified by the VeterinaryHead of Profession in Government.

Editorial Board

Editor: Linda SmithDefraBlock A GovernmentOfficesColey ParkReadingBerkshire RG1 6DT

Board: Jose BisMartyn BlissittWyn BuickAdrienne ConroyPaul GethingsAndrew GreshamAngela JonesSivam Kumar

Alex MorrowRichard RowlandLinda Smith

Production Team: Peter GreenAngela GreenawayEmily Thorne

2 Government Veterinary Journal VOLUME 16 | NO1 | 2006

Correspondence, including distribution inquiries and requests for additionalcopies should be sent to:

GVJ Production Team Internal Communications andStakeholder Management Team Department for Environment Food and Rural AffairsArea 206 1a Page StreetLondon SW1P 4PQ

Tel: 020 7904 6482 Email: [email protected]

The Government Veterinary Journal canbe viewed on the Government VeterinarySurgeons (GVS) website atwww.gvs.gov.uk. If you would like further information onany of the topics addressed in this issueplease contact your local Animal HealthOffice or the Production Team. Articles appearing in the GovernmentVeterinary Journal are Crown Copyrightand permission must be obtained fromthe Production Team before they can bereproduced elsewhere.The Government Veterinary Journal isdistributed free to its readers. It is notavailable for sale.

Website addresses included in this report are correct at time of publication. If you experience any problems with links, please contact [email protected]

Front Cover: State Veterinary Service vet recording details about the tuberculosis inspection.

The GVJ is printed on recycled paper containing 80% post consumer waste and 20% totally chlorine free virgin pulp.

10493 AW Defra GVJ_03-52 29/8/06 14:27 Page 2


Contents Foreword 4


2 Some lessons from the history of the eradication of Bovine Tuberculosis 11in Great Britain

3 Bovine Tuberculosis in the European Union and other countries: 19current status, control programmes and constraints to eradication



6 The introduction of pre and post-movement TB testing in Scotland 58for cattle from high incidence TB areas

7 Ante mortem diagnosis of Bovine Tuberculosis: the significance of 65unconfirmed test reactors




Are you interested in writing for the GVJ?Ideally, articles should be 1,000-3,000 words long, preferably with relatedillustrations. Full instructions for authors are available from any board member or the Production Team. Subject matter should be related to Government veterinarymedicine in its broadest sense, and you do not have to work for Defra or the StateVeterinary Service to be a contributor to the Journal. Articles for consideration can be submitted through any board member or the Editor (via the Production Team), a full list of which appears on the preceding page.

Contents

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 3


I am proud to present to you the first Government Veterinary Journal bovine TB specialedition. I am sure you will both enjoy it and learn from it; contributions have been made by many of the UK’s leading authorities on bovine tuberculosis. The first line of the first article sums up the importance of this special edition; bovine tuberculosis is themost difficult animal health problem we face in Great Britain today. This opening articleeloquently summarises policy in such a difficult area and includes reference to currentdisease trends, and policy changes during the last few years.

How can I summarise such a diversity of papers all on one subject? Andrew Proudpresents a historical perspective on the virtual eradication of bovine tuberculosis in days gone by. Putting the experience of Great Britain in perspective, Ricardo de la Rua-Domenech then describes the situation in other countries in a long but very useful article.

The following item, by Andy Mitchell, describes how Cattle Tracing Scheme (CTS) datacan be used for disease modelling purposes, and the laboratory diagnosis perspective isrepresented by Keith Jahans’ and Danny Worth’s article.

Closely linked to the opening policy position paper, Martyn Blissitt discusses theintroduction of pre- and post-movement testing in Scotland. The subsequent article helps explain the significance of unconfirmed TB test reactors and is of particular value to staff working in the field. Martin Vordermeier et al then explain the significance andimportance of the BOVIGAM® test.

Finally, and on a more specific basis, Bob Monies describes an on-farm incidentinvolving tuberculous pneumonia in calves, and Wyn Buick has adapted an existingdocument which originated from Ricardo de la Rua-Domenech which outlines bovinetuberculosis in domestic species other than cattle.

Putting together a TB special has been a considerable challenge and we haveattempted to provide a balance of articles ranging from the highly scientific to thehistorical. We would not, however, wish to mislead the reader. This is by no means the sum total of knowledge about bovine tuberculosis; it is, rather, a range of articleswhich address some important aspects of the disease.

I certainly believe that this edition provides significant Continuing ProfessionalDevelopment value for all readers throughout the Government Veterinary Surgeonsnetwork. We all benefit from a reminder of the science surrounding TB epidemiology and diagnosis and it should be noted that, where indicated, the authors hold anabundance of references or further reading material information which could not be included in the GVJ.

My grateful thanks to the Editorial Board and Production Team, who have done a sterling job in pulling all of the articles into publishable format. Special thanks to those from the Production Team who have moved on.

Linda Smith, Editor

Foreword

Foreword

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 4

5

TB policy developments

IntroductionBovine tuberculosis is the most difficultanimal health problem we face in GreatBritain (GB) today. The scale of thechallenge facing both Government andindustry in seeking to reverse the long-term upward trend in the disease, andpreventing its spread into new areas, is significant. In leading the changesrequired to make this happen, theGovernment is committed to developingpolicies based on the full range of evidenceavailable. We are committed to workingever more closely with our deliverypartners and stakeholders throughoutthe policy development process and informulating plans for the successfuldelivery of measures to reduce the levelof the disease. These principles areenshrined in the ‘Government strategicframework for the sustainable control of bovine tuberculosis in Great Britain’.This article sets out the key policydevelopments since publication of thestrategic framework in March 2005, with particular emphasis on England.

The scale of the problem Although bovine TB currently affects only a relatively small proportion of thenational herd, the problem is severe inthose areas where the disease isconcentrated, such as the south west andwest of England and south west Wales.

The costs of dealing with the diseasehave also escalated in recent years and in the financial year 2004-2005 we spent£90.5 million on the TB programme in

Dr DebbyReynolds

A u t h o r

Figure 1Geographicaldistribution of cattleherds with confirmednew TB incidents (red dots) in 2005

Headline bovine TB statistics for GB in 2005• 3.4% apparent prevalence at the end

of the year (herds under restrictiondue to a TB breakdown, excludingherds with overdue test restrictions)

• 4.3% incidence of confirmed newherd breakdowns throughout the year(7.8% if all breakdowns are included)

• 93.6% of cattle herds were consideredofficially TB-free at the end of the year

• 4.8 million cattle were tuberculintested in more than 43,000 herd tests

• 3,653 new TB incidents were recorded(up 9.1% on 2004). Infection wasconfirmed in 2,023 of those (up14.7% on the previous year)

• 25,755 tuberculin test reactors wereslaughtered (0.53% reactors per 100animals tested)

kCONTINUED ON p6

1

VOLUME 16 | NO1 | 2006 Government Veterinary Journal


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 5

Government Veterinary Journal VOLUME 16 | NO1 | 2006

GB. This covered expenditure on bovineTB testing and surveillance, compensationand research.

Very recently it has become apparentthat there has been a substantialreduction, in the region of 20%, in thenumber of new TB incidents in the firstfew months of 2006, compared to thesame period in 2005. However, theemerging figures should, for the time being, be treated with caution. The statistics on TB have historically gone up and down on a month by monthbasis due to changes in disease risk andthe population of cattle under test. Realchanges in disease incidence can only be established over a protracted timeseries by assessing long-term trends. At this stage it is fartoo early to drawconclusions aboutwhether thedecrease is anexaggeratedseasonal fall, orwhether it indicatesa more sustainedreduction. Thereasons for anylong-term fall, if maintained, are likely to be due to a complex combination offactors, including the introduction overrecent years of a suite of enhanced cattlecontrol measures along with a possiblereduction in transmission risks. In addition,we are giving consideration to thepotential effect that a switch in tuberculinsupply towards the end of 2005 may havehad on the number of new TB incidentsdisclosed. At the moment there is noevidence that the change in tuberculinsupply could have caused this reduction,but further investigation is requiredbefore this can be confirmed.

Current policy position – the GB TBStrategic FrameworkIn March 2005 Defra, the WelshAssembly Government and the ScottishExecutive announced a joint ten-year

Government strategic framework for the control of bovine TB in cattle andfarmed deer within GB. The frameworkwas developed following extensiveconsultation with farmers, veterinarybodies and wildlife interests. It buildsupon the Government’s previous FivePoint Plan (which focussed on publichealth protection measures; TB testingand control; development of a TBvaccine; research into transmission andspread of the disease and carrying out a badger culling trial), and applies tobovine TB the guiding principles set out in the overarching GB Animal Health and Welfare Strategy.

The strategic framework sets out a vision for a new partnership with

Government,working with itsdelivery partnersand stakeholders, to reduce theeconomic impact of bovine TB, tomaintain publichealth protectionand ensure animalhealth and welfare.

Its aim is to slow down and stop thespread of disease into areas currently freeof the disease, and achieve a sustainedand steady reduction in TB incidence inhigh incidence areas. The frameworkrecognises that TB control policies needto be tailored to reflect the regionalvariation in disease risk, and be adjustedto make best use of emerging scientificfindings. The document also sets out a process for making decisions onwhether badger culling may form part of future policy.

As part of the TB strategic framework,Government is committed to evidence-based policy development. There is anongoing programme of research toimprove our scientific understanding ofbovine TB, for instance how it is spreadand the role of badgers. The frameworkalso formally recognises the importance

6 EXOTIC DISEASES IN DOGS AND CATS – THE DACTARI SCHEMEGovernment Veterinary Journal VOLUME 16 | NO1 | 2006

The strategic framework sets out avision for a new partnership withGovernment, working with its deliverypartners and stakeholders, to reducethe economic impact of bovine TB, tomaintain public health protection andensure animal health and welfare.


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 6

VOLUME 16 | NO1 | 2006 Government Veterinary Journal 7

kCONTINUED ON p8

of other strands of evidence in relationto policy development – economics,environmental impact assessments,societal issues and the practicality ofdelivering policies.

Since the framework was published we have been working extensively withthe State Veterinary Service, private vets,representatives of the farming industry,livestock auctioneers, Local Authoritiesand wildlife groups amongst others, toshare ideas and to develop a greaterunderstanding of the practical issuesassociated with the delivery of newbovine TB measures. For example, weconvened a stakeholder sub-group, withan independent chair, to develop a detailedproposal for the introduction of statutorypre-movement testing. The eventual policywas based largely on the stakeholdergroup’s proposals except where therewere practical or legal constraints.

Recent developments in TB policy in EnglandThe pre-movement testing policy was one of a series of measures announcedby Ministers on 15 December 2005 totackle bovine TB in England and helpachieve the vision set out in the TB StrategicFramework. The key components of theannouncement were:

• the introduction of a statutoryrequirement for pre-movement testing of cattle to help reduce the risk of spreading bovine TB throughmovement of cattle;

• the introduction of a newcompensation scheme for farmerswhose cattle are affected by bovine TB;

• a public consultation on both theprinciple and method of badger culling as a measure to control bovine TB in cattle;

• a renewed commitment to pursuingthe development of a TB vaccine; and

• a commitment for extended use of a blood test based on a gamma-

interferon assay as an adjunct to theskin test to help improve diagnosis of the disease.

At the same time the IndependentScientific Group on Cattle TB (ISG)reported on the scientific findings of theproactive element of the RandomisedBadger Culling Trial, or ‘Krebs trial’(RBCT). The ISG reported that proactiveculling of badgers in response to a herdTB breakdown, as conducted in theRBCT, while possibly benefiting farmswhere badgers were removed (20%reduced incidence) worsens the diseaseincidence in neighbouring farms (30%increased incidence). These figures havebeen updated since those published inNature, 14 December 2005. The RBCTwas designed to test the effect of twodifferent approaches to badger culling,implemented under field conditions, and in a way that could be extended into a viable TB control policy option. These findings supported the ISG’s earlierconclusions, arising from the reactiveelement of the RBCT, that localisedculling of badgers is likely to beineffective in controlling TB in cattle.

The RBCT is just one component of the TB research effort put in place byDefra, and which the ISG has directedsince 1998. During the remainder of2006 and the first quarter of 2007, the

Figure 2Local veterinaryinspectors testing a cow

2


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 7

ISG will continue to undertake detailedevaluation of the results, submit furtherreports to Ministers and prepare papersfor scientific journals. The ISG expects itsfinal report – to include a full review ofthe RBCT and its findings, and otherrelated research – to be ready in March orApril 2007. That will conclude their work.

Pre-movement testingA statutory requirement for pre-movementtesting was introduced in England on 27 March 2006. The policy is beingintroduced in two phases to allowsufficient time for herd owners and theveterinary profession to adjust to the newrequirements. The current arrangements,phase one of the policy, apply to cattleover 15 months of age moving from oneand two year tested herds, unless the herdor movement meets one of a number ofexemptions. All pre-movement tests mustbe arranged and paid for by the herdowner. Routine TB surveillance tests, paidfor by the Government, qualify as pre-movement tests if animals are movedwithin 60 days of the test. Prior to theplanned implementation date for phaseone of the policy in February 2006, serious concerns were raised by farmingorganisations about the readiness of theirmembers to comply with the policy and of the adequacy of veterinary resources to carry out the increased level of testing. We listened carefully to the concerns andas a result implementation was delayed by just over a month to give industry moretime to prepare for the measure and allow a speedy independent review ofveterinary capacity and preparedness to be carried out.

The review found no evidence to supportan additional delay to the introduction ofpre-movement testing.

To further ease the transition to the new arrangements, Government agreed to pay for one pre-movement test perherd owner in England between 20 February and 30 June.

The new arrangements are beingclosely monitored and early indicationsare that the new system is settling downwell. The policy is being kept underreview to enable modifications, if necessary, prior to the plannedimplementation of phase two of thepolicy, in March 2007, when pre-movement testing will be extended to movements of cattle over 42 days old.

Related measures have been introducedin Scotland and Wales. In Scotland,compulsory pre- and post-movementtesting requirements were introducedfrom 23 September 2005, and in Walespre-movement testing was introduced on the same basis as that in England on 2 May 2006.

CompensationFollowing public consultation, a newcompensation scheme for farmers whose cattle are affected by bovine TB,brucellosis and Enzootic Bovine Leukosiswas introduced on 1 February 2006. The scheme was extended on 1 March to cover BSE. The new system wasdeveloped following the findings of anumber of independent reports showingsignificant and widespread overpaymentsunder the previous compensation system.It is more transparent than the old systemand is designed to be fairer to both cattleowners and taxpayers.

Compensation in England is nowdetermined primarily using table values,which reflect the average market price ofbovine animals in 47 different categories.The categories are based on the animal’sage, gender, type (dairy or beef) andstatus (i.e. pedigree or non-pedigree). As with pre-movement testing,Government has taken account whereverpossible of the detailed concerns raisedby stakeholders and, following consultation,made significant enhancements to thesystem. In particular, the proposed numberof table valuation cattle categories wasincreased from 29 to 47, with separatetables for commercial and pedigree cattle.



10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 8

VOLUME 16 | NO1 | 2006 Government Veterinary Journal 9

kCONTINUED ON p10

We have also set up a CattleCompensation Advisory Group, to help monitor the new compensationarrangements. The group includesrepresentatives from NFU, valuerorganisations, pedigree beef sector,pedigree dairy sector and the Meat and Livestock Commission.

Consultation on badger culling Results from the RBCT carried out inEngland, combined with other scientificevidence, including that from theRepublic of Ireland, led us to concludethat we needed to consult on whether ornot badger controls should form part ofour strategy to tackle bovine TB, and ifso, what forms of control would be mostappropriate. The resulting 12-week public consultation, which ran from 15 December 2005 to 10 March 2006,looked at the principle and method of a badger culling policy in areas of high TB incidence in cattle. The consultationdocument presented the scientific evidence,the balance of costs and benefits, andconsidered the implications of a badgercull for animal welfare and conservation.

We received over 47,000 responses tothe consultation. A report summarisingthese showed that 95% of all respondentswere opposed to a cull of badgers, butthat opinion was much more evenlydivided amongst organisations with aparticular interest in TB. Of the interestedorganisations which responded, 50%were opposed to a cull whilst 41%supported culling badgers to control the disease. The remaining 9% of

responses were neutral. A further reportsummarising the outcomes of Citizen’sPanels held to consider the issue, showedthere was an even division of opinionamongst the individuals involved. However,in group discussions the view wasmarginally in favour of a cull as part of amulti-faceted strategy and with manyconditions attached such as improvedbiosecurity and continued research intoTB vaccines for cattle and badgers.

At the time of writing, no decision on badger culling has been made.

Research and TB VaccinesThe Government’s research programme is focused on improving understanding of bovine TB in cattle and wildlife, trialingdisease control options and developingnew tools to fight the disease. The researchpriorities are improving diagnosis of thedisease, developing vaccines both forbadgers and cattle, improvedunderstanding of the epidemiology(including modeling), more in depthunderstanding of wildlife behaviour andecology and the impact of this, and cattlehusbandry on TB, and modeling of thedisease for use in cost benefit analyses of control policy options.

As well as funding a significant TBscience programme, GB researchers arecollaborating closely with researchersfrom overseas. It is critical that we workclosely with other countries, to shareexperience and apply what has beenlearned elsewhere to the GB situation.

We are committed to developing a TB vaccine for badgers and cattle. Our announcement in December 2005re-emphasised our commitment to thiswork. A substantial part of the Defraresearch programme focuses on this and over the past seven years we haveinvested more than £10.5 million in vaccinedevelopment and associated research.Although a vaccine is a long-term aim,and we do not anticipate that a new

3

Figure 3Meles meles (badger)Photographer:Richard Yarnell(2002)


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 9

vaccine will be developed and licensedwithin the ten-year timeframe covered by the TB Strategic Framework, we arepressing ahead and making real progress.

In January this year the VeterinaryLaboratories Agency (VLA) began furtherwork looking at new vaccine candidatesand delivery protocols in a naturaltransmission study in cattle. This will runin parallel with existing studies at the VLAand the Institute of Animal Health (and in collaboration with research in NewZealand and Ethiopia).

Research also continues on differentialdiagnostic tests, which are needed todistinguish vaccinated from infectedcattle, since vaccines based on BCG makecattle react to the current tuberculin testas if they were infected with Mycobacteriumbovis. Progress would need to be madeon the legislative and licensing frontbefore use in cattle would be possible.

The present programme of researchinto the use of a vaccine for badgers hasbeen underway since 1999. A three yearbadger vaccine field study led by the VLA,in collaboration with the Central ScienceLaboratory, using parenteral BCG hasrecently started to obtain safety datatogether with some efficacy data forlicensing purposes. Trapping andsampling badgers to obtain baseline data has started, and vaccination of wildbadgers is likely to start in September.

Further work on oral vaccine and baitdelivery systems for badgers has alsostarted at the VLA in collaboration with colleagues at the Central ScienceLaboratory and New Zealand, and inclose liaison with similar work in theRepublic of Ireland.

Gamma interferon blood assayDefra is already using the gammainterferon blood-based diagnostic test onan ad hoc basis in identified problem TBherds to improve test sensitivity and theremoval of more infected cattle. About14,000 tests were undertaken in 2005-2006. Preparations are now being made

for wider use of the gamma interferon test as an adjunct to the TB skin test inprescribed circumstances, and a GammaInterferon Working Group has beenestablished to prepare and deliver a policy for increased use of the test.

Engagement with interestedorganisationsThe GB strategic framework set out acommitment for an annual conference for organisations with a particular interestin bovine TB, with the aim of reaching awide audience and focussing on exchangeof information. We held our first GBconference on 6 March 2006. Theprogramme for the meeting was drawnup to provide a balanced coverage ofbovine TB issues including cattle controls,wildlife issues and research. Nearly 60organisations sent delegates, representingfarming, veterinary, wildlife and otherinterests from across GB. The outcomesof the discussions on wildlife controls arebeing considered alongside the responsesto the public consultation on badgerculling and will be taken into account in the decision-making process.

We are making progress in establishingnew bovine TB policy advisoryarrangements. Any new group will bemade up of individuals representing abroad range of stakeholder interests and will have close links with the AnimalHealth & Welfare Strategy’s EnglandImplementation Group.

ConclusionThe Government in GB is committed to working in partnership with others toachieve the vision set out in the GB TBStrategic Framework. The range of policymechanisms available for controlling TBdepends largely on achieving a betterunderstanding of the disease, how it is spread, and the effectiveness andpracticality of interventions and theoutcomes of our research programme and other evidence will help with this.


Dr DebbyReynolds is the CVO (UK),Director-Generalof Animal Health& Welfare andVeterinary Head of Profession.

l Information


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 10


Andrew J ProudA u t h o r

Some lessons from the history of the eradication of BovineTuberculosis in Great BritainThe First Half-Century – DevelopingUnderstanding and Reluctance to ActThe Public Health Act of 1875 lit the fuse, which led, via a series of delayedexplosions, to government action tocontrol bovine tuberculosis. The Actempowered Medical Officers of Healthand Inspectors of Nuisances (later knownas Sanitary Inspectors, then Public HealthInspectors and finally (we hope!)Environmental Health Officers), to seizeunfit meat. Eventually, some of them did.Spurred on by the report of a departmentalcommittee in 1888, which, whileconceding that ‘the bacilli may be foundbut rarely in the flesh’, considered the risk‘too probable to ever allow of the flesh of a tubercular animal being used forfood under any circumstances’, the moreenthusiastic inspectors began to seize the whole carcase of any ox in which any tuberculous lesion was found.

And so it was that in 1889, the firstexplosion thrust bovine tuberculosis onthe attention of an unwilling government,when a deputation of outraged butchersand cattlemen arrived at the Board ofTrade to protest. The protest had thesympathy of thinking veterinary surgeons,including the Chief Inspector of theVeterinary Department of the Privy CouncilOffice, Professor G T Brown, who wroteof ‘the frequent seizure by sanitaryauthorities of carcases of… healthy cattlein fine condition… confiscated withoutany compensation to the purchaser, who had acted throughout in good faith.’Brown revealed his sound scientific bentby referring to ‘the presumption that theuse of meat from tuberculous animals isprejudicial to public health’ (author’s italics).

The Government responded in classicstyle by setting up a Royal Commission in 1890. For the next 21 years, a series of Royal Commissions deliberated, took evidence, and even did their own research, finally coming to firmconclusions in 1911, but apart from awise response to the grievance of theinitial protesters, their findings andrecommendations, if not entirely ignored,were met with less than prompt andenthusiastic implementation. Critics ofmodern Government attitudes to foodsafety would do well to reflect that it took over a quarter of a century before an interim report of the RoyalCommission, which concluded thattuberculous milk was the cause of asignificant loss of human life, resulted in effective action to make milk safe. By contrast the speed of action and theextreme precautions taken in response to BSE are breathtaking. An Order wasmade in 1909 in response to the interimreport of 1907 but was never put intooperation. Orders made in 1913 and1914 were revoked on the outbreak ofthe first World War, but these Orders wentno further than requiring the notificationof clinical cases, veterinary investigationand slaughter with compensation.

The 1925 Order resulted in the slaughterof an annual average of some 17,000cows but, in the judgement of GowlandHopkins, did ‘nothing to reduce theincidence of disease. Nor has it done muchto protect the public from infected milk, asthe majority of cows are not reported untiltowards the end of their lactation or whenin an advanced state of the disease.’

kCONTINUED ON p12

Some lessons from the history of the eradication of Bovine Tuberculosis in Great Britain

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 11


Although meat inspection remained avoluntary pastime for Local Authoritiesuntil wartime regulation of the meatindustry in 1940, where they did carry it out, they adopted a less draconianjudgement for localised tuberculouslesions. ‘The French system’ whichM’Fadyean favoured, in his evidence to a Royal Commission in the 1890s, was essentially the same as thatembodied in legislation in 1961 inScotland and 1963in England andWales, later to becarried, virtuallyunchanged, intoEEC Directives andpresent UKlegislation; onlygeneralisedtuberculosisrequired totalrejection, a fact ofwhich some modern Meat Inspectorsseem to be unaware.

Were the Conclusions of the RoyalCommissions Right?At a time when voices are being raised inprotest against TB reactors being salvagedfor human consumption, and publicmoney is being spent on testing animalswhich are all due for slaughter within ayear or two, we might pause to askwhether the Royal Commissions wereright to support the Smithfield protesters.

The controversy was informed by theviews of two giants of medical science.The great Koch, whose best-known postulateis still a cornerstone of bacteriology, wasalso postulating, presciently that bovinetuberculosis was caused by a separatestrain of the bacterium, but wrongly thatbovine tuberculosis was insignificant inhuman disease. M’Fadyean, however,urged that the bovine strain was significantfor human disease; nevertheless he hadthe wisdom to distinguish between therisks of contracting the disease from milkand from meat.

While he advocated the heat treatmentof milk by ‘steaming’ as early as the1890s, he supported the Smithfieldprotesters. He estimated that, while totalrejection of the meat of all tuberculousanimals would lead to the destruction of the carcases of 15-20% of all adultcattle and probably 30% of dairy cows,the French system would result in theseizure of only 0.5%.

It would have been hard for the RoyalCommissions to havemissed the area ofagreement betweenthese opposing views.No doubt the Frenchsystem of MeatInspection owedsomething to Koch’sviews, but that it wassupported by theleading advocate ofthe heat treatment

of milk, must surely have weighted thescales of the argument. It is hard tobelieve that M’Fadyean had not noticedthe obvious fact that is so commonlymissed today, that if tuberculous milk is rendered safe by minimal heattreatment, beef will be rendered safe bycooking. The Gowland Hopkins reportcertainly grasped this point: “But asmeat is generally well cooked, and as thetubercle bacillus is destroyed by exposureto comparatively low temperatures, thisis unlikely to be a serious source ofinfection. It is to infection from milk, thefood of childhood, that most of thebovine tuberculosis in human beings is attributable” (author’s italics).

It is an irony that the French method of meat inspection arose and took rootin continental Europe where, unlike,Britain, there is a tradition of eatinguncooked meat.

M’Fadyean’s estimate of prevalence maybe open to question, as may the tentativeestimate of the learned Fellows of theRoyal Society who wrote the GowlandHopkins Report of 1934. They accepted

‘The French system’... was essentiallythe same as that embodied inlegislation in 1961 in Scotland and1963 in England and Wales, later to be carried, virtually unchanged,into EEC Directives and present UK legislation; only generalisedtuberculosis required total rejection.


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 12


kCONTINUED ON p14

that 40% of dairy cows were infected and referred to several witnesses whoestimated that 40% of all cattle wouldreact to the tuberculin test, but theconclusion is inescapable that betweenthe 1880s and at least the mid 1930s the disease was endemic.

Fresh data available to Gowland Hopkinsalso endorsed M’Fadyean’s view that onlya minority of cattle developed generaliseddisease. In Edinburgh, 70% of carcasesfrom cattle with tuberculous lesions couldbe passed as fit for human consumption,20% were subject to partial rejectionwhile only 10% were totally rejected.

While the pasteurisation of all milk notarising from tuberculin tested cows hasbeen enforced since 1935, the attempt to ensure inspection of all meat duringthe fourteen years of rationing, whichbegan with the Second World War didnot approach success until towards theend of that control in 1954. Not until thedelayed implementation of the MeatInspection Regulations of 1961 (Scotland)and 1963 (England and Wales) was alllegally sold meat subject to compulsoryinspection. According to ‘Animal Health:A Centenary’, compulsory meat inspectionwas introduced in Scotland some 30years earlier than in England and Wales.

The author’s view of the effectiveness ofthis inspection is prejudiced by the accountof an Aberdonian Meat Inspector, the lateCharles Knights, who started work at theage of 12 as a part time slaughterman,during wartime regulation. He recalledthat one of his early duties was to remove the retropharyngeal lymph nodes and conceal them in the paunches.Clearly these nodes were commonlytuberculous but their total absenceescaped the notice of the inspectors!

Meat inspection was not, therefore,universal until well after the whole countryhad been declared an attested area andthe incidence of Bovine Tuberculosis hadfallen to 0.06%. The dramatic decline of human tuberculosis attributed toMycobacterium bovis accords much

better with the sudden removal of riskfrom milk after 1935 than with the slowimplementation of national, compulsorymeat inspection, which was finallyovertaken by virtual eradication. Quiteapart from the uninspected meat, whichwas reaching the consumer routinely,during and after the decade in whicheradication took place, all reactors, whichwere not sold on the open market, wereslaughtered and subjected to the ‘French’method of inspection. It beggars beliefthat, if M’Fadyean was wrong, a clearoccupational risk among slaughterman,butchers and beef-eaters would not have emerged. It also betrays a strangeinconsistency of thought, if cooking isregarded as a satisfactory primary preventivemeasure for salmonellosis in poultry meatand not as an effective second line ofdefence for tuberculosis after theelimination of the vast bulk of infection byvisual inspection and selective rejection.

Early ActionThe Tuberculosis Order of 1925 and the Milk and Dairies Order of 1926, had the combined effect of making clinicaltuberculosis notifiable and giving power to Local Authorities to undertake veterinaryinspection of dairy herds. The goodintention of this legislation was frustratedby the reluctance of farmers to reportdisease promptly and a marked, althoughnot universal tardiness on the part of theLocal Authorities, to appoint sufficientveterinary inspectors and to undertakeregular herd inspections. GowlandHopkins’ summary of responses fromLocal Authorities as to the number ofinspections undertaken indicated that manyhad no veterinary inspectors and over athird undertook no routine herd inspections!

Really effective inspection had to awaitthe amalgamation of the veterinary servicesof local and central government in 1938.In Scotland, where annual veterinaryinspection of herds was compulsory,


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 13

experience led to the opinion that annual inspection was insufficient.

Milk legislation provided for milk to besold under one of four designations, butdid not require all milk to be so designated.Table 1 sets out the requirements for thedifferent grades.

The obtuseness of the categories mayexplain Gowland Hopkins’ observation thatalthough the effect was to provide ‘milk ofa quality maintained by a system of publicinspection for those who are prepared topay for it… sales of designated milk assuch are far below the supply and show no signs of expansion.’ Nevertheless, the grading system reflected and fuelledprivate sector activity in tuberculin testingso that Gowland Hopkins could also reportthat although ‘many herds are free oftuberculosis but are not officially registeredfor the production of designated milk…under 1% of dairy cows are in herdsofficially recognised as being free oftuberculosis.’ However, at that time ‘freefrom tuberculosis’ only meant that reactorsat regular tests were immediately removed.A herd could be officially free even if itnever had a clear test!

The Gowland Hopkins Report and aQuarter Century of Spectacular ActionPossibly in response to the last annualreport of a retiring Chief VeterinaryOfficer, in 1932, the Prime Ministerappointed the Economic Advisory CouncilCommittee on Cattle Diseases, with the

president of the Royal Society, SirGowland Hopkins as chairman. It was thereport of this committee two years later,which led to effective control and virtualeradication of bovine tuberculosis in thefollowing 25 years despite interruption bythe Second World War.

The Committee had the opportunity to take evidence from witnesses who had long experience of endemic bovinetuberculosis and who were informed by the experience of other countries. It realised that the prevalence of diseasewas so high that early compulsoryeradication was not feasible but,recommended measures for reducing the risk to human health, cutting the load of bovine infection and paving the way for eventual eradication.

The members of the Committee graspedthree important principles. That thepasteurisation of milk, properly regulatedand enforced, would prevent mosttransmission to humans. That it waspossible to split the national herd (andindeed individual herds) into clean andinfected sections, since the diseaseneeded close and prolonged contact of infected with uninfected cattle fortransmission to take place. That theeffective removal of clinical cases wouldreduce the prevalence of disease, sincemost infected animals were not significantlyshedding the causal organism.

Although not formally recommendingits adoption as a government measure,

14


EXOTIC DISEASES IN DOGS AND CATS – THE DACTARI SCHEME

Table 1Comparison ofgrades of milk at the time of theGowland Hopkinsreport, 1934. Notethat the TotalBacteria figure forCertified milk is NOTa misprint; it is thefigure given in theReport, but mayhave been a misprintin that source

1


Grade Turberculin Veterinary Total Coliforms OtherTest Inspection Bacteria/cc

Certified 6 Monthly 6 Monthly <30 000 (see 0 in 0.1 ccnote below

Grade A (TB 6 Monthly 6 Monthly <200 000 0 in 0.1 cc May be bottledtested) on farm

Grade A Not required Not required <200 000 0 in 0.01 cc May be bottledon farm

Pasteurised Not required Not required <100 00 Not specified

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 14



kCONTINUED ON p16

they were clearly impressed with Bang’smethod of eradication as practised inDenmark. This depended on the fact thatcalves of infected animals were almostinvariably free from infection at birth:calves were removed from their damsimmediately and reared separately; when they calved they were milked andkept as an isolated group. As time wenton, and clinical cases were removed, the clean herd eventually replaced theinfected herd.

The immediate government response to Gowland Hopkins was threefold: rapid progress towards the effectivepasteurisation of all milk not designatedas ‘tuberculin tested’, regular veterinaryinspections of all dairy herds to detectand slaughter clinical cases and a voluntary attestation scheme withbonus paymentsfor attested herds.

During thefollowing threeyears the averagenumber of clinicalcases jumped to22,793, almostcertainly as aresult of increasedveterinaryinspection. The subsequentdecline wasaccelerated bywartime shortages which led to increasedculling of unthrifty cows and in 1945 theannual figure fell below 10,000.

While pasteurisation made animmediate impact on new humaninfections (although that only becameclear in hindsight when the mortality ratebegan to decline), the Attested HerdsScheme took much longer to make animpact. But a vocal section of the publicdid not appreciate their new safety, they complained that pasteurisation wasunnatural, spoiled the taste, raised thecream line and reduced the nutritionalvalue. Initial progress was brisk. By the

end of 1935 only 99 (out of at least aquarter of a million herds) were attested.During the next four years the totalnumber at the end of each year increasedbetween three- and fourfold, but theoutbreak of war led to severe limits beingplaced on new applicants. To nearly14,000 herds, which had gained attestedstatus by the end of 1939, were addedonly 3,100 by the end of 1944. It isdifficult to appreciate how a bonus ofone penny a gallon for attested herds, to which was added a premium for ‘TT’designated milk of 2.5 pence per gallonin 1938, translates into modern values. If one reflects that a new-graduateveterinary surgeon might have beenpleased with a ‘salary’ of £3.50 a weekand that as late as 1970 a penny a gallonwas still considered a significant incentive

to accelerate brucellosisaccreditation, one can only consider it as avery significantincentive indeed.

Modern critics ofdisease eradicationprogrammes wouldbe appalled by ascheme whichallowed herdownersto gain attested statusby selling theirreactors on the openmarket. However,

the policy was totally defensible: The unanswerable argument was that,with a suspected prevalence of 40%,purchasing on the open market wasa certain way of adding more infectedanimals to a herd which was very likelyalready infected.

In addition, the Attested Herds Scheme,for the first time, was establishing a poolof animals known to be free of tuberculosisfrom which purchasers could buy withconfidence. The effect was to redistributeinfected animals rather than eliminate

The immediate government responseto Gowland Hopkins was threefold:rapid progress towards the effctivepasteurisation of all milk notdesignated as ‘tuberculin tested’,regular veterinary inspections of alldairy herds to detect the slaughterclinical cases and a voluntaryattestation scheme with bonuspayments for attested herds.

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 15

them; once a majority of animals were inherds known to be free of tuberculosis,national eradication could be considered.

It was fortunate that Gowland-Hopkins,did not recommend delaying actionpending further surveys, trials orrefinement of testing. Had the AttestedHerds Scheme been delayed, pending the results of a trial undertaken during a 12 month period following August1938, it would, almost certainly havebeen shelved for the duration of the war.Among 364,286 cattle in 12,300 self-contained herds, selected on the basis of a herd history of little or no clinicaltuberculosis, the prevalence of reactorswas 13%. More than half the herds had no reactors, but approximately one third of the cattle were in herds with aprevalence exceeding 10%. The resultsare summarised in Table 2. Clearly theestimates of M’Fadyean and GowlandHopkins were not excessive.

The slow growth of the Attested HerdsScheme during wartime was not withoutits redeeming features. Research continued,particularly in the refinement of thetuberculin test. A variety of tuberculins of varying potency were eventuallyreplaced by a standardised purified proteinderivative (PPD), as recommended byGowland-Hopkins. By the time theScheme began, the subcutaneous testwhich depended on detecting rises in temperature in the 24 hour period

following injection, had been replaced by the double intradermal test whichnecessitated three visits to the farm. In 1940 the use of more potenttuberculins made it necessary to introducea comparative test to reduce the numberof false positives. In 1943 PPD tuberculinscame into general use and were found tobe of such high sensitivity and specificitythat the double intradermal test conferredno advantage; it was abandoned in 1947.In the following 15 years this test proveditself to be dramatically effective. Critics ofthe present test would do well to realisethat, following the replacement of‘mammalian’ (i.e. human) tuberculin by themore specific ‘bovine’ in 1975, the present test is actually a refinement of that test.

The introduction of a bonus of 4d pergallon for milk from tuberculin testedherds in October 1943 and the reopeningof the scheme to all applicants ninemonths later; set the scene for rapidacceleration after the war. Almost asmany herds were added to the attestedregister in 1945 as in the whole of theprevious five years. From then onwardsprogress was rapid so that by the end of 1950, the number of attested herdsexceeded 55,000 and included 22% ofthe bovine population of around tenmillion. A year later another 19,000 herdsincluding a million cattle had been added.

Since the geographical distribution ofattested herds was far from uniform with


Table 2Summary of Results of1937-1938 Survey

2

Total cattle Total herds Average % reactors % herdsherd size

All herds 364,286 12,300 29.6 100 100

Herds with 127,141 6,971 18.2 34.9 57no reactors

Herds with < 114,701 2,897 39.6 31.5 23.610% reactors

Herds with < 122,444 2,432 50.3 33.6 19.810% reactors


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 16


Scotland far in the lead and Englandlagging a long way behind Wales, it wasnow possible to consider eradicating the disease from designated areas. The plan was to select areas where a highproportion of herds were already attestedand which were largely self-sufficient for herd replacements and marketing. The announcement that these areas wereto be declared eradication areas precededthe declaration by two years which gavethe minority of non-attested herds theopportunity, lubricated by further incentivepayments, to gain attested status andavoid the movement restrictions andcompulsory testing which would beimposed on the declaration date.

In October 1950 the first eradicationareas (one each in Scotland and Wales)were announced and new incentivepayments for both dairy and beef herdswere introduced. However, before theseareas were declared, voluntary attestationproceeded so well in the Scillies, theShetlands and the islands of Arran andCumbrae that these could be declaredattested areas (without a compulsoryeradication phase) in February 1951.

This policy worked well so that by theend of 1959 the whole of Scotland andWales had been declared attested areas.The last attested areas (in the Midlands,North and East of England) were declaredon 1 March 1960 and gained attestedstatus only seven months later.

The response of farmers to theannouncement of eradication areas fellinto three categories. Many decided totake advantage of the incentives and selltheir reactors on the open market; others,

either through inertia or belief that thecompensation finally payable for reactorswould be generous, waited for compulsorymeasures; a minority, especially in thelater stages of eradication, deliberatelybought cheap reactors expecting to makea large profit from the compensationwhich they would receive under compulsorymeasures. This latter category left a bittermemory in the collective consciousness ofthe farming community to such an extentthat two decades later the NFU indicatedtheir opposition to over-generouscompensation during Brucellosiseradication. Farmers who had eradicatedunder the voluntary scheme, and receivedtheir incentive payments felt threatenedwhen they found a close neighbouraccumulating reactors. The authorencountered one such case who believedthat his herd had become heavily infectedas a result. Consequently, having receivedincentive payments, he was not eligiblefor compensation and suffered heavyfinancial loss.

Despite the inevitable effect of the area eradication scheme in concentratingreactors in areas yet to be designated foreradication, some farmers must havebeen surprised to find herds which theyhad believed to be heavily infected werein fact clear. During the whole ofcompulsory eradication approximatelyhalf of all herds included in the first roundof compulsory testing had no reactors.Perhaps this led to fraud. The authortreasures a story told to him in 1967,about a veterinary surgeon in thehighlands of Scotland who arrived at hismost distant farm only to find that he had forgotten his tuberculin. He injectedall the animals with air and was acutelyembarrassed to find that they all showedlarge reactions at the mammalian sitethree days later. During the years whencompensation was fixed at 75% of thevalue of the reactor up to a ceiling relatedto average values, there was little motive

Figure 1Advanced case ofgeneralised TB

1 kCONTINUED ON p18


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 17


Andrew Proudwas formerly aVeterinary Officer(VO) at theGloucesterAnimal HealthDivisional Office,and retired in2005.

l Information

for such fraud. Nevertheless the authorencountered one such case where 18reactors were faked. It is somewhatgratifying to report that the miscreantdied in prison while serving a double life sentence – but for other offences. The overall prevalence of reactors was17.5% and the herd range was 0.5-30.5%.It is hard to escape the conclusion thattuberculosis is not a disease, which spreadseasily or rapidly between and within herds.

The progress of attestation wasreflected in the decline of clinical cases.Having fallen below 10,000 during 1945,it had fallen below 400 before the firsteradication areas were declared, and toonly 28 in 1960. Not all of these suspectclinical cases, of course, were confirmed.

For an area to be declared attested,total disappearance of reactors was notrequired but only a fall to a very lowprevalence. In 1960, when the wholecountry was declared attested, theprevalence was 0.19%. By 1964 it hadfallen to 0.06% from which level it wasnot destined to rise significantly until thesuccess of government policy to protectbadgers showed its unfortunate by-product. The clear lesson from the historyof the eradication programme is thatbovine tuberculosis is readily susceptibleto simple control measures where cattleare the sole reservoir population.

Space does not permit a considerationof the history subsequent to the declarationof national eradication. Although aninteresting study, the need for it is lessurgent as there are many people stillliving who remember most of it and aconsiderable literature is readily available.

Learning the Lessons of HistoryTo the lessons explicit or implicit in the foregoing text should be addedparticular emphasis in three areas.

Those who claim that tuberculosis isnow endemic in the national herd or in the South West, are clearly wrong.Currently prevalence is nowhere near

the 20-40% of the half century whichpreceded the area eradication scheme.

The methods of control developed andrefined during the first half of the 20thcentury worked very well and allowedrapid reduction of prevalence to negligiblelevels. There is no reason to suppose thatin the absence of a protected wildlifereservoir population they would no longerwork today. With the same proviso, thecontrol of the disease by isolation –within farms, of farms, and of designatedareas – is as feasible today as it was in the 1940s and 50s. In those days it wasunderstood that double fencing betweenfarms was an adequate safeguard andthat tying cows tail to tail, as opposed tonose to nose, in the shippon significantlyreduced risk. While not able to quote acontemporary source, the author’s clearmemory is that the two yard spacerequired between the fences during theBrucellosis Incentives Scheme provokedthe comment from older farmers thatonly one yard had been required duringtuberculosis voluntary attestation.

The lack of urgency attached to theslaughter of reactors until modern timesdid not appear to hinder the progress oferadication; considerable economies couldbe affected by learning from ProfessorBang’s methods. If an unseemly haste in removing reactors was not necessarywhen the only risk was cattle to cattletransmission, it is still less appropriate in dealing with disease acquired fromanother source, which cannot be removed.

Bibliographical NoteThe Gowland Hopkins Report, HMSO1934 and Animal Health: A Centenary,HMSO 1965, were the main sources ofthe information given above, but readersof the latter work should note that thehistory of tuberculosis eradication is to be found concentrated in two entirelyseparate parts of the book and scatteredpiecemeal throughout a third.


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 18


IntroductionBovine tuberculosis (TB) caused byinfection with Mycobacterium bovis(M. bovis) is a zoonotic disease thataffects cattle worldwide. Althoughdomestic cattle (Bos taurus and B. indicus)are the natural hosts of the organism,nearly all warm-blooded animals appearmore or less susceptible to the infection.Given the right conditions, the infectioncan become self-sustaining in some ofthese species, which then constitutemaintenance hosts to M. bovis in additionto domestic cattle. Public health concernsin the first half of the 20th century led to compulsory pasteurisation of milk andcampaigns to eradicate M. bovis from thecattle population in developed countries.Whilst the zoonotic risk posed by M. bovis has been largely controlled inthe developed World, the success of TBeradication campaigns in cattle has beenmixed. Despite the limitations of existingscreening tests, bovine TB has beeneffectively eliminated in many countriesand regions. This has been achievedthrough the application of long-termsystematic programmes of tuberculin skin testing and removal of reactors,coupled with repeat testing and culling of infected herds, slaughterhousesurveillance, cattle movement restrictionsand occasional slaughter of entire herdswith intractable breakdowns. Morerecently, deployment of ancillary in vitrodiagnostic assays alongside the skin testshas enhanced the detection of infectedcattle and helped resolve problems ofnon-specific reactivity to bovinetuberculin. In other developed countries

the traditional ‘test and slaughter’approach has been instrumental inreducing the incidence of bovine TB, but complete control of the disease indomestic cattle and farmed deer hasbeen hampered by wildlife reservoirs and other factors. The problem of bovineTB in wildlife is not unique to the BritishIsles. In a survey conducted in 2000 bythe Office International des Épizooties(OIE), 22% of countries reported that TB had been identified in wildlife within theprevious 10 years. In the majority ofthose countries the role of infectedwildlife in the maintenance and spread of M. bovis is not well defined. However,the experience accumulated over the last30 years has highlighted the difficultyand controversy in controlling TB in cattleonce the disease becomes established ina wildlife reservoir. Additionally, endemicM. bovis infection in livestock and wildlifeconstitutes a serious threat to publichealth in many developing countries. This is exacerbated by the pandemic ofhuman immunodeficiency virus (HIV) infection, socio-political unrest,inadequate veterinary, financial anddiagnostic support to conduct test andslaughter campaigns, lack of access topasteurised milk and close human-to-livestock interactions in pastoralistcommunities. This paper reviews thecurrent global situation of bovine TB,focusing on the prospects for eradicationof this disease in the European Union andother parts of the World.

Bovine Tuberculosis in the EuropeanUnion and other countries:current status, control programmes and constrains to eradication

Ricardo de la Rua-Domenech

A u t h o r

kCONTINUED ON p20

Bovine Tuberculosis in the European Union and other countries

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 19

20 EXOTIC DISEASES IN DOGS AND CATS – THE DACTARI SCHEME

The European Union (EU)

Historical background and legislative frameworkThe first attempts to control TB in cattle in Europe and other industrialisedcountries began in the late 19th century,following Robert Koch’s identification ofthe TB bacillus and the discovery of a linkbetween certain cases of human TB andconsumption of raw cows’ milk. In theyears that followed, control measuresconsisted of clinical examination of herds,bacteriological examination of milk andvoluntary slaughter of tuberculous dairycows. These measures were intended toprotect public health, rather than aserious attempt at reducing the ratherhigh prevalence of bovine TB in thenational herd. Screening of cattle withtuberculin-based tests was not widelyadopted in those early days. The controlof TB in cattle intensified in the 1930s,but it was only after World War II thatEuropean governments began to phasein national or regional TB eradicationcampaigns based on regular tuberculinherd testing, slaughter of test reactorsand restriction of infected herds. By 1960 participation in TB eradicationprogrammes had become compulsoryacross Europe (e.g. GB in 1950, France in 1955, Ireland in 1957 and NorthernIreland, Poland and former Czechoslovakiain 1959) except in the Southern Europeancountries, where it took a little longer to introduce similar schemes (e.g. Italy in 1977). As expected, each country tailored its programme to suit the localconditions. For instance, the cervicalsingle intradermal test (SIT) was adoptedas the primary screening test for TB invirtually all the continental Europeancountries. By contrast, the singleintradermal comparative tuberculin(SICCT) test performed with both bovine and avian tuberculin remains theprimary test in Portugal, GB (since 1947),the Republic of Ireland (1954) andNorthern Ireland (1958). Elsewhere in

Europe, the SICCT is only used for re-testing of SIT reactors and inconclusivereactors.

The first attempt to co-ordinate bovineTB control efforts across Europe wasintroduced on 26 June 1964, date onwhich Council Directive 64/432/EECcame into force setting out specificanimal health requirements for the intra-Community trade of cattle. In relation toTB, the aim of this Directive was tofacilitate trade of cattle between the EUMember States whilst minimising the riskof spreading the disease by movementsof infected animals. This Directive laiddown the criteria for herds, regions andcountries to achieve (and maintain)‘officially tuberculosis-free’ (OTF) status.Trade in bovine animals (including Asianwater buffalo [Bubalus bubalis] andNorth American bison [Bison bison]) for breeding and production purposescan only take place out of OTF herds. In brief, Directive 64/432/EEC states thata bovine herd is OTF if all the animalsover six months of age have passed two skin tests at six-month intervals.Although herds must be testedsubsequently at yearly intervals tomaintain OTF status, the intervalbetween tests may be extended as the national or regional incidence ofconfirmed herd breakdowns declines.When the annual percentage of infected herds in an Member State (or a region of it) has not exceeded 0.1%for six consecutive years and all bovinesslaughtered are subject to officialveterinary meat inspection, thatgeographical unit can be declared OTFand dispense with routine skin testing of cattle. The technical annexes of thisDirective also lay down the technique and standard interpretation of the twoprimary screening tests approved for usein bovine animals in the EU (the SIT andSICCT test). Directive 64/432/EEC hasbeen amended on several occasions since its enactment. Most recently,Commission Regulation (EC)



10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 20

No 1226/2002 of 8 July 2002 re-definedthe rules for standardisation andcalibration of avian and bovinetuberculins and allowed, for the firsttime, the deployment of the gammainterferon (γ– IFN) blood assay(BOVIGAM®) as an ancillary parallel testfor TB in cattle.

A second important legal initiative wasthe promulgation of Council Directive77/391/EEC, which laid down EU-widemeasures for the eradication of bovineTB, brucellosis and leucosis. Under this

Directive, non-OTF Member State areexpected to draft programmes in order toaccelerate the eradication of bovine TB.The European legislation also provides forfinancial support of national eradicationprogrammes from the EU budget(commonly known as the ‘EU VeterinaryFund’). In 2005, TB eradicationprogrammes in Cyprus, Greece, Italy,Poland, Portugal and Spain were all co-financed under that scheme.


1

2004 2004 2003 2002 2001 2000 1999cases/100,000

population No. of cases

Austria (OTF) <0.1 4 4 4 5 0 0

Belgium (OTF) 0.05 5 5 2 2 0 0

Cyprus 0.14 1 (1) - - - - -

Denmark (OTF) <0.1 2 (1) 1 2 4 12 2

Finland (OTF) 0 0 0 0 0 0 0

France (OTF) - - - - - - 22

Germany (OTF) <0.1 51 - - - - 64

Greece 0 0 0 0 0 93 -

Hungary 0 0 - - - - -

Ireland <0.1 2 6 7 3 1 8

Italy <0.1 5 1 4 0 0 -

Latvia 0 0 - - - - -

Lithuania 0 0 0 - - - -

Portugal - - - - 0 - -

Slovenia 0 0 - - - - -

Spain <0.1 4 6 2 3 5 2

Sweden (OTF) <0.1 4 (2) 5 7 5 5 2

Netherlands (OTF) - - - 8 10 13 19

United Kingdom <0.1 8 30 20 30 21 40

EU-Total 86 58 56 62 150 159

Current OTF status (as of 2005) is indicated. Imported cases are noted in parentheses and included in the total number of reported cases.No data provided by Czech Republic (OTF), Estonia, Luxembourg (OTF), Malta, Poland and Slovakia (OTF).

Table 1Total number of caseshuman TB caused byM. bovis infectionreported in the EUover the period 1999-2004 (source: data as reported byMember States to the European FoodSafety Authority)


kCONTINUED ON p22

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 21



Human M. bovis infection in the EUDespite the harmonisation of TB testingrules for intra-Community trade of cattleand TB controls in milk, dairy productsand meat, there is still room forimprovements in the reporting of human M. bovis infections in the EU. The notification systems for human TBare not standardised across the EU: many Member State do not distinguishbetween the different species ofMycobacterium that cause TB in humans and, even in those countries that undertake speciation of humanmycobacterial isolates, this may not be standard practice. As a result, theincidence and overall trend of human M. bovis infections cannot be reliablyestimated in the EU. In 2004 a total of 83cases of human TB caused by M. boviswere recorded across the Member Statethat reported this data (Table 1), but theactual number of infected persons is

probably higher than that. Nevertheless,the risk of humans contracting TB fromanimals in the EU is considered very low, even in the non-OTF states. Mostincidents of M. bovis infection in humansarise from reactivation of latent disease inelderly or immunocompromised people,or infections acquired outside the EU.

Eradication of bovine TB in the EU:progress and constraintsTables 2 and 3 summarise the situation ofbovine TB in the 25 Member State of theEU at the time of writing, according tothe most recent information supplied bythe national veterinary authorities to theEuropean Commission and the OIE.Figure 1 shows the classification of EUMember State according to their officialTB status as at 30 September 2005. The success of systematic testing andslaughter programmes has so far resultedin the recognition of 11 Member States

1

IRL

UK

E

P

F

B

NL

L

D

I

SI

A

CZ

PL

SK

HU

EL

LT

LV

EE

FIN

S

MT CY

DK

Figure 1Map of the EuropeanUnion showing theclassification ofMember States (orregions of them)according to theirofficial TB status as at30 September 2005

Light green: OfficiallyTB free (OTF) MemberStates or regions (inItaly only)

Orange: Non officiallyTB free MemberStates

White: Non EUcountries

Table 2Descriptive TBstatistics for cattle in Member States of the EU recognisedas officially TB free(OTF) in 2005 by the EuropeanCommission. Data for 2004, except whereotherwise indicated(adapted fromReviriego andVermeersch 2005)

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 22

BOVINE TUBERCULOSIS IN THE EUROPEAN UNION AND OTHER COUNTRIESBovine Tuberculosis in the European Union and other countries


N.A

. dat

a no

t av

aila

ble

1.D

ata

for

2003

2.Ex

plan

atio

n: a

) No

rout

ine

test

ing

(sla

ught

erho

use

surv

eilla

nce

only

), b)

Tes

ts o

nce

a ye

ar (e

xclu

ding

bov

ines

und

er s

ixw

eeks

in m

ost

case

s), c

) Tes

ts e

very

tw

o ye

ars,

d) T

ests

eve

ry t

hree

yea

rs, e

) Tes

ts e

very

fou

r ye

ars,

f) o

ther

(see

det

ails

).3.

i.e. p

re-m

ovem

ent

test

ing

4.Bu

lls in

art

ifici

al in

sem

inat

ion

stat

ions

2

Tota

l nu

mb

er o

fR

ou

tin

e sk

in

Nu

mb

er o

f N

umbe

r of

cat

tle

Nu

mb

er

Off

icia

lly

Infe

cted

ex

isti

ng

bo

vin

ete

stin

gsk

in t

ests

wit

h s

usp

ect

of

cult

ure

-TB

fre

eh

erd

sca

rrie

d o

ut

TB le

sio

ns

po

siti

ve(O

TF) h

erd

sb

efo

resu

bm

itte

d f

or

anim

als

intr

od

uct

ion

his

tolo

gic

al a

nd

o

f ca

ttle

into

bac

teri

olo

gic

alth

e h

erd

s3ex

amin

atio

ns

Cou

ntry

OTF

Her

dsA

nim

als

Inte

rval

Num

ber

Num

ber

%N

umbe

r%

sinc

ebe

twee

nof

of

her

dsof

her

dsro

utin

e an

imal

ste

sts2

test

ed

Aus

tria

119

9995

,401

2,08

8,27

3N

.A80

3N

.A51

095

,401

100

00

Belg

ium

2003

42,5

532,

807,

915

(a)

045

0,00

013

446

42,5

4999

.99

80.

01

Cze

ch20

0427

,806

1,42

8,32

9(b

) for

bul

ls32

2,49

42,

778

330

27,8

0610

00

0Re

publ

ic(c

) for

oth

erca

ttle

ove

r 24

mon

ths

Den

mar

k119

8028

,270

1,75

9,19

2(f

) bul

ls a

nd2,

400

01

028

,270

100

00

catt

lein

tend

ed f

orex

port

Finl

and

1995

22,8

8296

9,14

0(a

)81

440

110

22,8

8210

00

0

Fran

ce1

2000

277,

291

17,8

73,9

75(a

) 3.

6 m

illio

n86

0,00

015

144

277,

213

99.9

778

0.03

(b)

(in 5

7,00

3(c

) he

rds)

(d)

(e)

Ger

man

y19

9723

5,89

714

,028

,087

(a)

6,78

417

586

4423

5,88

699

.99

00.

01

Luxe

mbo

urg1

1997

1,61

918

9,67

4(a

)0

00

01,

619

100

00

Net

herla

nds1

1995

60,8

15N

.A(a

)0

01

060

,815

100

00

Slov

akia

2005

11,3

5256

3,77

1(c

)32

1,12

70

280

11,3

5210

00

0

Swed

en19

9529

,456

1,66

6,16

4(a

)44

51

00

29,4

5610

00

0

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 23



3

Tota

l nu

mb

er o

fR

ou

tin

e sk

in

Erad

icat

ion

Nat

ion

al a

nim

al T

BO

ffic

ially

Nat

ion

al h

erd

inci

den

ceex

isti

ng

bo

vin

ete

stin

gp

rog

ram

me

inci

den

ce (r

eact

ors

TB F

ree

(OTF

)(n

ew r

eact

or

her

ds

per

100

cat

tle

test

edh

erd

sd

ivid

ed b

y h

erd

s te

sted

in t

he

year

)d

uri

ng

th

e ye

ar)

Cou

ntry

Her

dsA

nim

als

Inte

rval

bet

wee

nN

umbe

r of

Co-

finan

ced

%Y

ear o

f las

t%

%Tr

end

over

the

perio

dro

utin

e te

sts3

anim

als

by th

e EU

TB in

cide

nt19

99-2

004

test

edin

200

5

Cyp

rus

363

60,0

00(g

) Las

t tes

ting

Non

e in

Yes

019

2810

0 (p

resu

med

)0

No

rece

nt b

reak

dow

nsca

mpa

ign

2004

, W

orki

ng to

war

dsin

200

0-20

01.

7,00

0O

TF s

tatu

sRe

sum

ed a

nnua

l in

200

5te

stin

g in

200

5

Esto

nia1

9,18

024

8,16

0(b

)14

9,81

3N

o0

1986

100

(pre

sum

ed)

0N

o re

cent

bre

akdo

wns

Wor

king

tow

ards

OTF

sta

tus

Gre

ece

37,7

7887

0,66

3(b

)68

7,12

2Y

es0.

99O

ngoi

ng95

.20

0.74

Stab

le

Hun

gary

32,7

2176

6,00

0(c

)N

.A.

No

<0.

0120

0499

.99

<0.

01D

eclin

ing.

Wor

king

tow

ards

OTF

sta

tus

Irela

nd12

4,41

46,

992,

264

(b)

9.5

mill

ion

No

0.33

Ong

oing

96.9

05.

72Sl

owly

dec

reas

ing

Italy

187,

098

6,45

6,99

3(b

)3.

8 m

illio

nY

es (f

or0.

21O

ngoi

ng90

.90

0.62

Stab

le w

ith w

ide

(plu

s 24

0,00

0no

n-O

TFre

gion

al v

aria

tions

buff

alo)

regi

ons)

in p

reva

lenc

e5

Latv

ia71

,799

376,

540

(b)

166,

767

No

<0.

0119

8910

0 (p

resu

med

)0

No

rece

nt b

reak

dow

nsW

orki

ng t

owar

dsO

TF s

tatu

s

Lith

uani

a19

5,22

691

6,71

5(b

)51

2,86

3N

o0

2001

100

0D

eclin

ing.

Wor

king

tow

ards

OTF

sta

tus

Mal

ta2

281

18,4

17N

.A.

N.A

.N

o0

2001

100

(pre

sum

ed)

0N

o re

cent

bre

akdo

wns

Wor

king

tow

ards

OTF

sta

tus

Pola

nd88

2,76

15,

649,

362

(d)

1.7

mill

ion1

Yes

0.03

Ong

oing

99.9

90.

04St

able

Port

ugal

77,8

551,

119,

545

(b)

958,

3061

Yes

0.09

Ong

oing

N.A

.0.

17St

able

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 24

as OTF (Table 2). OTF status does notnecessarily imply biological freedom fromM. bovis and sporadic TB breakdownscontinue to be reported in some OTFstates. Of the ten Eastern and SouthernEuropean countries that joined the EU on 1st May 2004, eight had not achievedOTF status at the time of writing thisarticle (Table 3 and Figure 1). Nevertheless,the outlook for these new Member Stateis quite favourable, as TB incidence hasremained at very low levels for severalyears. By contrast, bovine TB remains a significant animal health problem in the UK, Ireland, Spain, Italy and (to a lesser degree) Greece and Portugal. The reasons for this vary from country to country, but are generally linked to the presence of reservoirs of infection inwild or domestic mammals, extensive useof outlying, rented or common grazing,high stocking densities, incompletetesting coverage, lack of compliance withscheme requirements or a combinationof these factors.

The comprehensive testing regimes that have eradicated TB in manyEuropean countries have failed to yieldsimilar results in the British Isles. In GB,the highly successful herd attestationscheme launched in 1950 reduced thenumber and incidence of test reactorcattle from nearly 15,000 (16.2 reactorsper 10,000 cattle tests) in 1961 to 569(2.3 reactors per 10,000 tests) in 1982(Figure 2). This remarkable progress cameto a halt in the mid 1980s, when thesituation began to gradually worsen to apoint where GB now sustains one of thehighest incidences of bovine TB in the EU(Table 3). In Northern Ireland the controlof bovine TB has mirrored the experiencein GB. A significant reduction in bovineTB incidence was observed after theintroduction of a compulsory testingscheme in 1959. Disease incidencefluctuated in the 1970s and mid 1980s.This was then followed by a sustained


kCONTINUED ON p26

Bovine Tuberculosis in the European Union and other countriesSl

oven

ia46

,041

470,

807

(g) G

over

nmen

t40

2,51

4N

o0

2002

100

(pre

sum

ed)

0D

eclin

ing.

Wor

king

pres

crib

es a

ges

tow

ards

OTF

sta

tus

and

type

s of

ca

ttle

elig

ible

for

test

ing

each

yea

r

Spai

n17

1,19

06,

328,

241

(b)

4.8

mill

ion

Yes

0.47

Ong

oing

96.5

41.

11D

eclin

ing

UK

(GB)

93,1

658,

924,

000

(b) 2

5.6%

,4.

6 m

illio

nN

o0.

43O

ngoi

ng97

.79

(at

the

6.87

Incr

easi

ng s

tead

ily(c

) 12.

8%,

end

of 2

004)

(3.6

)4La

rge

regi

onal

(d) 0

.7%

,va

riatio

ns in

inci

denc

e (e

) 60.

9%

of h

erds

UK

(NI)

28,5

681,

677,

583

(b)

1,86

5,67

1N

o0.

81O

ngoi

ng91

.87

9.17

Gra

dual

incr

ease

acce

lera

ted

in 2

001

and

2002

. Fol

low

ed

by a

slo

w d

ownt

urn

acce

lera

ted

in 2

005

3

Table 3Descriptive TB statisticsfor cattle in MemberStates of the EU notrecognised as OTF by the EuropeanCommission in 2005.Data for 2004, exceptwhere otherwiseindicated (based ondata from Reviriego and Vermeersch 2005,Pavlik et al. 2005 and TB eradicationprogrammes submittedby Member States to the EuropeanCommission)

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 25

Figure 2Total number (left y-axis) and incidence(right y-axis) of skintest reactor cattleslaughtered in GBevery year between1956 and 2005(based on officialdata published in theCVO Annual Reports– Animal Health GB)

rise from 1987 to 2002 similar to thetrend in GB, although both herd andanimal-level incidence have fallen inNorthern Ireland over the period 2003-2005. In the Republic of Ireland thebovine TB eradication programme madeconsiderable progress in its early years. In the late 1950s and early 1960s thenumber of cattle culled as reactorsexceeded 100,000 each year, but by1965 this number had fallen to around23,000. In contrast to the situation in theUK, the number of reactors identified inIreland since the mid-60s has remainedfairly static at between 20,000 and40,000 reactors per year, despite annualtesting of all herds.

M. bovis has been isolated fromdomestic animals other than cattle, wilddeer, several species of carnivores andsmall mammals. However, the Eurasianbadger (Meles meles) has been identifiedsince the mid 1970s as a true maintenancehost and the principal wildlife reservoir ofM. bovis in the British Isles, where itremains a protected species. Nowadays,endemic M. bovis infection within badgerpopulations in parts of the UK andIreland is a major impediment to theeradication of bovine TB, although the

relative contribution of infected badgersto the incidence of TB in the nationalcattle herds is hard to quantify. A seriesof badger culling strategies have been inoperation both in GB and Ireland sincethe mid 1970s and early 1980srespectively, but their success in helpingcontrol TB in cattle has been variable.Large-scale field trials have taken place inboth GB and the Republic of Ireland toassess the impact of long-term badgerremoval operations on the incidence ofTB in cattle. However, it is not yet clearhow the lessons from the most recenttrials will translate into cost-effective,socially acceptable and sustainablebadger management policies. In the longterm, both the UK and Ireland arecommitted to the development of aneffective badger vaccine whilst retainingor enhancing the existing cattle controls.In the interim, Ireland continues toimplement a national programme ofreactive badger culling on affected farmswhere introduced cattle have been ruledout as the cause of a TB breakdown. This badger culling effort is more intensivein those areas of Ireland with persistentTB incidence in cattle (Figure 3).

M. bovis infection in wild deer was



30,000

25,000

20,000

15,000

10,000

5,000

0

1956

1959

1962

1965

1968

1971

1974

1977

1980

1983

1986

1989

Year

Reac

tors

sla

ught

ered

60

50

40

30

20

10

0

Reac

tors

per

10,

000

test

s

1992

1995

1998

2001

2004

Reactors slaughtered

Animal incidence (reactors per 10,000 animal tests)

2

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 26


Figure 3Map of Irelandrepresenting the‘chronic’ TB areas(marked green).These areas compriseroughly 30% of theIrish agricultural landwhich yielded 70%of all standardinterpretation reactorcattle during theperiod 1998-2000.Badger culling in thechronic TB areasdoes not take placeover more than 60%of the agriculturalland, whereaselsewhere in Irelandbadger removal islimited to 20% ofthe land (source:James O'Keeffe,Department ofAgriculture andFood, Republic of Ireland)

kCONTINUED ON p28

first recorded in the UK in 1985 and theorganism has since been isolated fromfive of the six species of free-living deer in the UK (Figure 4). Although TB isnotifiable in deer, there is no evidence ofwidespread infection throughout the UK.Maintenance of infection and onwardtransmission to other species is most likelyto occur when TB pathology is severe, theprevalence of infection is high, the speciesis abundant and infected individuals (ortheir excretions) encounter susceptibleanimals repeatedly or for prolongedperiods. In general, wild deer in the BritishIsles do not fulfil these conditions, thuspresenting a lower risk to cattle thanbadgers. Even in those areas of Britainwhere TB is endemic in badgers, prevalenceestimates for deer are considerably lower.Furthermore, the relatively limited socialinteraction of deer and their free-livingstatus act to limit transmission opportunities.However, the potential for wild deer to actas localised TB reservoirs for cattle existswhere their populations are poorly

managed or artificially maintained abovethe natural carrying capacity. This could behappening in some populations of red(Cervus elaphus) and fallow (Dama dama)deer in South-West England. Based onanecdotal evidence, wild deer could alsoplay a role in the maintenance of bovineTB in some regions of Ireland, but how(and how often) this may occur is still unclear.

Information about M. bovis infections in wild mammals in the rest of Europe israther patchy. Surveys of TB in wildlifehave been few and rather localised. With the exception of Finland andSweden, TB in wildlife is not notifiable inEU Member State. Furthermore, in manyMember State responsibility for bovine TB eradication programmes and wildlifeconservation policies rests with differentGovernment departments with competingagendas. All this makes it difficult tocorrelate any data on the presence of TB in wildlife with the incidence of thedisease in cattle. In general, badgerpopulations are estimated to be at muchlower densities in continental Europe thanin the British Isles. Furthermore, thisspecies is huntable small game in severalEU Member State such as Austria, Finland,France, Latvia and Sweden. Outside theBritish Isles sporadic M. bovis infection ofbadgers has only been documented inSwitzerland in the 1950s. Rather than amaintenance host of M. bovis, badgers inthat country were considered a spillover


3

4

Figure 4Red deer (Cervuselaphus) stag in theScottish highlands(photograph suppliedby Charlie Wilson,RDS National WildlifeManagement Team,Exeter). M. bovisinfection in deer was first confirmedin Britain nearly 75 years ago. The bacterium isresponsible forsporadic severebreakdowns of TB in farmed red andfallow deer. M. bovishas also been isolatedfrom most species of wild deer in NewZealand and Europe.In continentalEurope, TB appearsto be an importantre-emerging diseasein wild deer and wild boar

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 27


host that became infected by scavengingcarcases of tuberculous roe deer(Capreolus capreolus). Switzerland hasbeen officially free of bovine TB since1959 and the disease has only beenconfirmed in five cattle herds between1991 and 2003. Once eradicated fromcattle, cases of TB in the Swiss wild deerand badger populations were no longerreported.

M. bovis has been isolated fromcarcases of red and fallow deer, wild boar (Sus scrofa) (Figure 5), hare (Lepuseuropaeus), Iberian lynx (Lynx pardinus,an endangered carnivore) and fox (Vulpesvulpes) collected in surveys of wildlife killedon national parks and private estates inCentral, Southern and Western Spain. In those regions it is quite common forherds of native Iberian pig and cattlebreeds (including fighting bulls) to bereared in intensive free-range breedingsystems. The incidence of TB in cattle has traditionally been high and difficult to control in those regions, where skintesting coverage is more patchy than in the predominantly dairy herds ofNorthern Spain. Furthermore, M. bovisinfection is widespread among wildungulates in Central, South and WesternSpain, where red deer and wild boarpopulations have been artificially increasedin recent decades to satisfy the demandsof a thriving commercial game huntingindustry. Geographical and temporalclustering of M. bovis strains isolatedfrom cattle, domestic pigs, goats andwildlife is suggestive of interspecies

transmission within shared ecosystems.Based on the high prevalence of infection,repeated isolations over several huntingseasons, ecological interactions andpathological presentation, red deer and wild boar have been identified as potential maintenance hosts of TBin Spain. There are two additionalchallenges to the eradication of bovineTB from that country: a reportedly highprevalence of infection in goatherds andthe seasonal movement of beef herds tomountain common grazing (transhumance).M. bovis infection of wild boar could alsoconstitute a problem in some regions ofPortugal, but the data currently availableis scarce and does not allow a proper riskassessment of this species.

Wildlife does not appear to be asignificant factor in the persistence of TB in cattle in Italy. M. bovis infectionwas identified in wild boar carcasescollected in Liguria and other regions in Northern Italy in the late 1980s andthroughout the 1990s. The confirmationthat most M. bovis isolates from wildboar were of the same genotype as thosefrom cattle reactors in the same regions is suggestive of inter-species infection.Furthermore, a number of unusual TB recrudescences in cattle herds (onoccasions after whole herd slaughter),also point towards a wildlife reservoir.However, in line with the experience of other countries, the incidence ofbovine TB in wild boar has reduced as the prevalence of cattle infectioncontinues to fall in Northern Italy.

France attained OTF status inDecember 2000, at a time when cattlewere believed to be the sole reservoir ofM. bovis in that country. The bacteriumhad never been isolated in Frenchwildlife, with the exception of two wildboars. However, over the period 2001-2005 M. bovis infections were discoveredby accident in wild boar and deer in fourgeographically distant regions of France.Epidemiological surveys have beenconducted in some of these to establish



5

Figure 5Wild boar (Sus scrofa)foraging in theAbruzzo region of Central Italy(photograph suppliedby Charlie Wilson,RDS National WildlifeManagement Team,Exeter). Wild boarpopulations in Europeare increasing innumber andexpanding in range.Tuberculosis causedby M. bovis wasdocumented in free-ranging wild boar for the first time inGermany in the1930s. The infectionhas been latelyrecorded acrossFrance, Northern Italy, Spain, Germanyand other CentralEuropean countries,occasionally with ahigh prevalence oflesions in head andpulmonary lymphnodes. Althoughgenerally regarded as a spillover host,more studies arerequired to elucidatethe role of wild boarin the epidemiology of bovine TB incontinental Europe

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 28

29

the extent of the wildlife reservoir (Table4). It is still unclear why bovine TB is a re-emerging disease in French wildlifewhilst almost eradicated in cattle. The discovery of these foci has coincidedwith alleged increases in wild ungulatepopulations in France over the last 25 years. Some have speculated thatinfection remained latent in some wildlifereservoirs throughout the cattleeradication campaign and that expandingwild deer and boar populations, coupledwith more intensive wildlife diseasesurveillance, have led to the recentdiscovery of TB in those species.

Over the period 1999-2001, there havebeen reports of bovine TB confirmed byculture in free-ranging red deer from anatural park in the northern Alpsstraddling Austria and Germany.Molecular epidemiological data and theabsence of TB in local cattle herds suggestthat M. bovis had been circulating in the local wild deer population for someyears. In the northern German region ofMecklenburg-Vorpommern tuberculouslesions were found in 102 (1.4%) of7,419 carcases of free-ranging wild boarkilled during the period 1982 to 1988.

Of the carcases with lesions, 55.6%yielded M. bovis. The alleged source of infection were TB breakdowns inneighbouring cattle herds. In otherCentral European Member States,where the TB eradication programmes in cattle have reduced the incidence of TB to very low levels, M. bovis has beenisolated in several wildlife species. In theperiod 2000-2004, M. bovis infectionswere diagnosed in wild boar (n=14) and red deer (n=6) in Hungary, and inEuropean bison (Bison bonasus) (n=14)and roe deer (n=2) in Poland. The trueepidemiological significance, if any, ofthese sporadic cases of bovine TB in wildungulates in Central Europe is unclear.

Finally, Sweden declared bovine TBfreedom in 1958 under its own nationaleradication programme. However, M. bovis infection re-emerged in 1991 in the farmed deer population after theimportation, four years earlier, ofinfected fallow deer from the UK. With atleast 13 deer herds infected and facedwith impossibility to trace all potentiallyinfected deer due to lack of farm records,


kCONTINUED ON p30


4

Region or Species Year of Apparent prevalence Self-sustaining infection inDépartement affected discovery of infection wildlife (maintenance hosts)?

Haute Normandie Wild boar, 2001-2002 28% and 14% of Probable(Brotonne Forest) Red deer and hunted boar and

neighbouring deer, respectivelycattle herds

Animals of all agegroups affected

Respiratory and digestive lesions

Haute Corse Wild boar 2003-2004 Low prevalence Unlikely

Bourgogne Wild boar 2003-2004 Low prevalence Unlikelyand Red deer

Pyrénées Wild boar 2005 Very low prevalence UnlikelyAtlantiques

Table 4Location andepidemiologicalsignificance of recentM. bovis isolatesfrom wild ungulatesin France

There is no evidenceat present of M. bovis infection in badgers or othersusceptible wildlifespecies in France

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 29


a voluntary scheme was put in place toallow deer farmers to regain OTF statusthrough repeat tuberculin testing and long-term slaughterhouse surveillance.The outbreak was apparently containedwith no evidence of spillover onto other species.

Russia and other countries of theformer USSRApart from the three Baltic States thatjoined the EU in 2004 (Table 3), there isvery little reliable data on the incidenceand prevalence of bovine TB in thecountries of the former Soviet Union,where 74 million head of cattle were kept in 2003. Systematic tuberculin herdtesting and culling of reactors began inRussia in the late 1950s. This led to agradual reduction in animal and herdincidence of bovine TB. Over the period1996-2003, Russia reported a total of2,012 TB breakdowns to the OIE, withthe number of breakdowns falling from712 in 1996 to 87 in 2003. All othercountries in the formed Soviet Uniondeclared nil or very low incidences of TB herd breakdowns in 2003, exceptArmenia, Kyrgyzstan and Turkmenistan(no data available). Epidemiologicalinvestigations carried out in the 1960s and 1970s in various zones ofKazakhstan (Central Asia) revealedbovine tuberculin reactivity and M. bovis

infections in cats, dogs, sheep, goats andBactrian camels.

AustraliaBovine TB was probably introduced inAustralia early in the 19th century withthe arrival of cattle from Britain. By the20th century, the disease was wellestablished in cattle herds throughoutthe country and posed a serious publichealth risk. The various States graduallyintroduced test and slaughter programmesfrom the early 1900s. Although thesewere initially voluntary, from the 1940sthe testing of dairy herds supplying rawmilk became compulsory. By the late1960s TB in dairy herds was largely undercontrol, but little progress had beenachieved in beef herds outside theintensively farmed coastal strips of thesouthern parts of Australia. Faced withthe prospect of losing vital exportmarkets, on 1 July 1970 the State andFederal governments launched aneradication campaign with financialsupport of the beef and dairy industries.The Brucellosis and TuberculosisEradication Campaign (BTEC) relied onsystematic and repeat skin testing ofcattle herds using the caudal fold SIT(Figure 6), slaughter of reactors andmovement controls. This wassupplemented with slaughterhousemonitoring and traceback of carcaseswith confirmed TB. In the latter stages of the campaign, entire groups or herds suspected of being infected wereslaughtered out.

The in vitro γ−IFN test (BOVIGAM®) wasdeveloped in Australia in the late 1980sfor the diagnosis of TB in cattle incombination with the caudal fold SIT.However, by the time the γ−IFN test hadbeen optimised and validated for use inthe field, bovine TB had already beenbrought under control by a conventionalpolicy of skin testing and slaughter ofreactors. So, this test played a minor rolein the successful Australian BTEC


Figure 6 Tuberculin skintesting of cattle in Australia using the caudal fold test(source: Dr BrianRadunz, Departmentof Business, Industryand ResourceDevelopment,Northern Territory,Australia).


6

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 30

31

programme and was never routinelyapplied. It was, nevertheless, accreditedas an official diagnostic test for bovine TB in Australia in 1991 and subsequentlyadopted as an ancillary test for TBdiagnosis in other parts of the world.

One of the challenges the BTECprogramme had to face was the largeferal population of Asian water buffalo in the vast coastal floodplains of theNorthern Territory. These animals hadbeen introduced from south-east Asia inthe 1800s and became well adapted totheir new habitat. By the early 1960s theprevalence of bovine TB in feral buffalohad reached 16.4% and they wereregarded as asignificant wildlifereservoir of M. bovisfor domestic cattle.Between 1984 and 1989 the feral population of 300,000 waterbuffalo in theNorthern Territorywas extensively culled as part of the BTEC programme. ‘Bush’ areas across the country were also destocked ofuncontrolled cattle and water buffaloliving in a feral state, which could not bemustered for skin testing. By contrast,feral pigs (Sus scrofa) in Australia wereconsidered an end host to M. bovisand not systematically culled. The highprevalence of TB recorded in feral pigs in the Northern Territory throughout the 1970s declined ostensibly after theeradication of the infection in cattle and the culling of feral water buffalo.

Australia finally became an OTF countryaccording to OIE criteria on 31 December1997. The following year, routine skintesting of cattle herds was replaced by intensive slaughterhouse-basedsurveillance (the National GranulomaSubmission Programme). All meat plantsare encouraged to submit for laboratoryexamination at least one suspicious TBlesion for every 1,000 cattle slaughtered

with two or more permanent incisor teeth.With M. bovis infection diagnosed in onlyfour of 23,000 granulomas submitted inthe period 2000-2004, Australia’s objectiveto achieve biological TB freedom byDecember 2006 appears well within reach.

New Zealand

TB eradication programme in cattleand farmed deer: history, features and trendsBovine TB was introduced in NewZealand in the early 19th century withthe cattle of European settlers. A test and slaughter programme was rolled

out on a voluntarybasis in 1945. This becamecompulsory for alldairy herds in 1961and for beef herdsin 1968. The firstcase of M. bovisinfection everrecorded in

farmed deer was diagnosed in 1978 in New Zealand. TB spread within the New Zealand farmed deer industry (thelargest in the world) through movementsof untested infected stock and capturedinfected feral deer. By the early 1980s itwas recognised that farmed deer couldact as maintenance hosts of M. bovis,resulting in severe and prolonged TBbreakdowns and were a possible sourceof infection for other species. A voluntarytest and slaughter scheme for farmeddeer was thus introduced in 1983 and became compulsory in late 1989.

The New Zealand bovine TB eradicationprogramme (known as Pest ManagementStrategy [PMS]), is administered by anincorporated society of stakeholders fromthe dairy, beef and deer farminginterests, meat industry and localgovernment, known as the AnimalHealth Board (AHB). Its activities are


kCONTINUED ON p32


Faced with the prospect of losingvital export markets, on 1 July 1970the State and Federal governmentslaunched an eradication campaignwith financial support of the beef and dairy industries.

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 31


funded by the Crown and Regionalgovernments, a levy on the slaughter ofcattle, grants from the agriculturalindustry and the salvage from theslaughter of reactors. The AHB’s aim isthe eradication of TB from New Zealand,but its medium-term operationalobjective by 2013 is to reduce thenumbers of TB-infected cattle and deerherds to 0.2% annual period prevalence(APR: the number of herds classified asinfected at the start of the financial year,together with herds found infectedduring the financial year, divided by totalnumber of herds). From the mid 1990s,New Zealand has achieved a sustaineddownturn in numbers of infected cattleand deer herds (Table 5 and Figure 7)through a combination of controlmeasures targeting reservoirs of

infection in both farmed and wildmammal populations:

• A test and slaughter programme for cattle and farmed deer,supplemented by routineslaughterhouse inspection;

• Herd- and area-based movementrestrictions and controls, including pre-movement TB testing;

• Sustained TB vector control, designedto eliminate infection from wildanimals in ‘Vector Risk Areas’ as wellas preventing incursions of infectedwildlife into ‘Vector Free Areas’ (see below);

• Buffer zones around populations of infected vectors and cattle and deer herds.



Measures used to Cattle Deerevaluate progress

1997-1998 2004-2005 1997-1998 2004-2005

Primary animal tests 4,775,672 5,307,138 552,129 756,782

Number (%) test positive 8,115 7,200 6,778 9,717(0.17%) (0.14%) (1.23%) (1.28%)

Number (%) serially re-tested 5,207 6,779 5,828 8,744(64%) (94%) (86%) (90%)

Percentage positive to serial re-test 16% 7.1% 3.3% 3.9%and slaughtered

Number of reactors slaughtered 3,722 1,073 1,142 1,318(% of all animals tested) (0.08%) (0.02%) (0.2%) (0.17%)

Number (%) of reactors with 1,824 307 196 125visible TB lesions (49%) (28.6%) (17.2%) (10%)

Number of tuberculous animals 722 249 156 214found during routine slaughter (0.03%) (0.01%) (0.04%) (0.05%)(% of the national kill)

Annual period prevalence of TB 0.03% 0.005% 0.03% 0.023%in animals*

Percent of TB animals from VRAs 83% 91% 74% 97%

* Based on the number of reactors slaughtered plus the number of lesioned non-reactors found at routine slaughter, divided by the population of the species.

5

Table 5Descriptive TBanimal statistics for New Zealandcattle and deer in1997-1998 and2002-2003. Thenumber of infectedcattle and deerherds also fell in the years ending 30 June 2003 and2004, continuingthe downward trend in numbers of affected herdsachieved over theprevious nine years(source: AnimalHealth Board, New Zealand)

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 32

33

Primary TB screening of cattle herds inNew Zealand is by the caudal fold SIT.Accredited lay technicians undertake thevast majority of routine skin tests in cattleand about 60% of deer tests. Cattle andfarmed deer herds are categorised aseither ‘Infected’ or ‘Clear’, together withan integer denoting the number of yearsthat a herd has maintained its status (e.g.‘I5’ or ‘C8’). Herd testing takes placeannually in declared movement controlareas (‘DMCAs’, where herd periodprevalence exceeds 1%) and triennially in clear (‘C2’ or better) herds elsewhere. All cattle and deer over three months ofage moving out of or within a DMCA(except those going directly to slaughter)must be negative to a caudal fold SITundertaken in the 60 days beforemovement. The γ–IFN blood test wasadopted in New Zealand in the mid1990s as an ancillary test for TB in cattle. As in other countries, this test is employed in very specific situations,

representing a small component of theoverall testing effort (0.47% of the totalof 5.33 million cattle tests administeredin 2004-2005). In low TB incidence areas the γ–IFN test is an ancillary test of SIT-positive animals (serial testing) and has all but replaced the SICCT in this role. Inexplosive and chronic herd breakdowns(‘problem’ herds), the γ–IFN test is usedas an ancillary test for skin test-negativeanimals to speed up the detection ofinfected animals (parallel testing).

Constraints to the eradication of TBfrom New ZealandWildlife reservoirs of M. bovis constitute amajor impediment to the eradication ofbovine TB in the New Zealand cattle herd.However, a fundamental difference withthe situation in the UK and Ireland is thatthose reservoirs are non-native speciesand regarded as ‘introduced pests’.Furthermore, agriculture is New Zealand’s


kCONTINUED ON p34


Figure 7Evalution of thenumbers of infectedcattle (green bars)and deer (white bars)herds in New Zealandrelative to thecumulative area ofthe country underpossum controloperations (orangeline), from June 1991to April 2005 (source:Dr Paul Livingstone,Animal Health Board,New Zealand)

1,400

1,200

1,000

800

600

400

200

0

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

Year ending June

Num

ber o

f Inf

ecte

d H

erds

1,600

1,800

7.00

6.00

5.00

4.00

3.00

2.00

1.00

0.00

8.00

9.00

2003

2004

2005

7

Vec

tor C

ontr

ol A

rea

(Mill

ions

Ha)

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 33


largest industry, representing 60% of the country’s export earnings. All thismakes the culling of wildlife (for livestockdisease control purposes) a more sociallyacceptable proposition than it is in theUK. The brushtail possum (Trichosurusvulpecula) is a small, herbivorousmarsupial imported from Australia in the19th century, whose total population inNew Zealand now exceeds 60 million. M. bovis infection was first described in possums in 1967 and today thisspecies is considered a maintenance hostof M. bovis and its main wildlife reservoir in New Zealand. The ferret (Mustelaputorius) is also an important TB reservoirwhere its population density is high,although its status as a true maintenanceor merely spillover-amplifier host is stilldebatable. Both possums and ferrets are

regarded as ‘TB vectors’ and the zones ofwhere TB has been identified in thesespecies are known as TB Vector RiskAreas (VRAs). As at June 2005, 39.5% of New Zealand’s land area was classifiedas a VRA (Figure 8). More than 90% ofinfected cattle and deer herds are locatedin VRAs, even though such areas containonly 20% and 32% of New Zealand’scattle and deer herds, respectively. Therole of possums, ferrets and other wildlifein the epidemiology of bovine TB in NewZealand is summarised in Table 6. All thisphilosophy is reflected in New Zealand’snational TB vector control programme,which is systematically applied topossums in the VRAs and accounts formore than half of the AHB’s expenditure.By contrast, other wild mammals (ferrets,feral pigs and wild deer, in decreasing


Figure 8Map of New Zealandshowing theMovement ControlArea overlaid onVector Risk (possumcontrol) and VectorFree Areas, as atJanuary 2005(source: Dr PaulLivingstone, Animal HealthBoard, New Zealand)

Table 6Wild mammals ofepidemiologicalsignificance in thetransmission of M. bovis infection in New Zealand

N

100

Kilometres

0 200

Surveillance Areaoverlaid on Vector Free Area

Vector Risk Area

Movement Control Areaoverlaid on Vector Risk Area

8


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 34


Bovine Tuberculosis in the European Union and other countriesW

ild d

eer

Bru

shta

il p

oss

um

Fera

l fer

ret

Fera

l pig

(mai

nly

red

, bu

t al

so

fallo

w a

nd

sik

a)

1967

• Tr

ue m

aint

enan

ceho

st.

• H

ighl

y su

scep

tible

to

infe

ctio

n.•

Maj

or w

ildlif

e ve

ctor

of T

B fo

r do

mes

ticliv

esto

ck in

New

Zeal

and.

• M

ain

fact

or in

the

expa

nsio

n of

vec

tor-

risk

(end

emic

) ar

eas.

• 60

-70

mill

ion

(up

to 2

5/ha

).•

Ave

rage

poi

ntpr

eval

ence

of

5%

(but

as

high

as

60%

).

• Sy

stem

atic

cul

ling

in V

RAs

thro

ugh

ava

riety

of

met

hods

. •

Dev

elop

men

t of

TB

ora

l vac

cine

.

Yea

r of

firs

t re

port

of M

. bov

isin

fect

ion

Epid

emio

logi

cal

stat

us a

s ho

sts

toM

. bo

vis

Dis

trib

utio

n an

dap

pare

nt

prev

alen

ce o

fin

fect

ion

Type

of

inte

rven

tion

1982

• C

erta

inly

a s

pillo

ver

(am

plifi

er)

host

and

poss

ibly

a m

aint

enan

ceho

st.

• La

rges

t kn

own

fera

lfe

rret

pop

ulat

ion

in t

hew

orld

(3-

8/km

2in

nat

ive

gras

slan

ds,

rare

r in

fore

sts

and

wet

ter

area

s).

• Pr

eval

ence

as

high

as

32%

.

• C

ulle

d on

ly if

epid

emio

logi

cal a

ndsu

rvey

info

rmat

ion

indi

cate

s th

at t

hey

coul

d be

act

ing

assi

gnifi

cant

TB

vect

ors.

1990

s

• Sp

illov

er (

ampl

ifier

) ho

stat

wor

st.

Gen

eral

lyco

nsid

ered

a d

ead-

end

(not

a v

ecto

r).

• Be

com

e in

fect

edth

roug

h di

rect

con

tact

with

(or

by

scav

engi

ngca

rcas

ses

of)

infe

cted

poss

ums

or f

erre

ts.

• V

ery

sens

itive

sen

tinel

s of

M. b

ovis

in t

heen

viro

nmen

t.

• 12

-43/

ha in

unh

unte

dpo

pula

tions

.•

31%

pre

vale

nce

inso

me

surv

eys,

but

alw

ays

in a

ssoc

iatio

nw

ith in

fect

ed p

ossu

ms.

• C

ulle

d on

ly if

epid

emio

logi

cal a

ndsu

rvey

dat

a in

dica

teth

at t

hey

coul

d be

actin

g as

sig

nific

ant

TB v

ecto

rs.

1970

• Sp

illov

er (

ampl

ifier

) ho

st.

• C

ould

beh

ave

as m

aint

enan

ce h

ost

only

in a

reas

of

very

hig

h de

erde

nsity

.•

Wild

dee

r di

sper

sing

ove

r lo

ngdi

stan

ces

seed

infe

ctio

n in

pre

viou

sly

unin

fect

ed w

ildlif

e po

pula

tions

, w

hen

scav

enge

rs f

eed

on in

fect

ed c

arca

ses.

• U

sefu

l wild

life

sent

inel

s to

sur

vey

for

M. b

ovis

infe

ctio

n in

pos

sum

s in

larg

ebu

sh a

reas

whe

re t

here

are

no

catt

leor

far

med

dee

r.

• A

s hi

gh a

s 37

% in

som

e ar

eas,

bu

t al

way

s in

ass

ocia

tion

with

foc

i of

TB

in p

ossu

ms.

• D

ensi

ties

in N

Z ge

nera

lly b

elow

th

e th

resh

old

requ

ired

for

self-

sust

aini

ng in

fect

ion.

• C

ulle

d if

epid

emio

logi

cal a

nd s

urve

yin

form

atio

n in

dica

tes

that

the

y co

uld

be a

ctin

g as

sig

nific

ant

TB v

ecto

rs.

• Ta

rget

ed c

ullin

g in

som

e op

erat

ions

,to

pre

vent

est

ablis

hmen

t an

d sp

read

of in

fect

ed w

ild d

eer

into

VFA

s.

6

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 35


order of probability) are targeted by ad-hoc control operations only ifepidemiological and survey informationindicates that they could be acting assignificant vectors of TB. A variety ofmethods of vector control are used (e.g. aerial application of poisoned baits, ground application of toxins andtrapping), depending on environmentalconsiderations and cost-effectiveness.Vector control operations have reducedpossum populations to very low densitieswithin VRAs and this effort has beencredited with the sustained decline innumbers of infected cattle and deerherds since 1996 (Table 5 and Figure 7).

Vaccination ofcattle and farmeddeer against TB inNew Zealand wouldonly be consideredas a last resort inspecific areas andwould be based ona decision torelinquish, at least temporarily, theeradication objective in those areas. Bycontrast, vaccination of possums hasalready been trialled in the field andwould offer the potential to furtherreduce the rate of spread of infection inintractable wildlife TB hot spots where alow possum population density hasalready been achieved through poisoning.The development of an oral bait vaccine for possums containing a lipid formulationof BCG (an attenuated strain of M. bovis)is well advanced.

CanadaAt the turn of the 20th century theprevalence of bovine TB in Canadiancattle was estimated at about 4%. An eradication programme wasestablished in Canada in 1923. By 1961,animal prevalence of bovine TB had fallen to 0.11% through regular skinherd testing and slaughter of reactors.

From 1978 the thrust of the programmeshifted from routine on-farm testing toroutine slaughter inspection of cattle,with depopulation of infected premises.This includes slaughterhouse surveillanceof Canadian cattle slaughtered in USAabattoirs (more than one million animalsin 2002). In 1989 the TB eradicationprogramme was extended to includefarmed bison and cervids, resulting in the detection of 37 infected cervid herds through November 2003.Currently, traceback investigations on all histologically suspect granulomas are supplemented by targeted on-farm testing of cattle herds in special

eradication areas(see below). Tocompensate for the relatively smallnumbers of adultbison and deerconsigned toslaughter, all herds of captive

bison and deer are routinely tested,usually at three-yearly intervals.

There are two known (butgeographically and epidemiologicallyseparate) reservoirs of TB in wildungulates in Canada. M. bovis infectionis endemic in the free-ranging woodbison (Bison bison athabascae)population in and around Wood BuffaloNational Park (WBNP). This is Canada’slargest national park, straddling northernAlberta and the Northwest Territories.Although WBNP was established in 1922to protect the threatened wood bisonpopulation, the translocation between1925 and 1928 of infected plains bison(Bison bison bison) from anothernational park in Alberta resulted inhybridisation of the wood bison and theintroduction of bovine TB and brucellosis(eventually detected in 1937 and 1956respectively). Because of its relativeremoteness, the 4,700-strong wood


M. bovis infection was first describedin possums in 1967 and today thisspecies is considered a maintenancehost of M. bovis and its main wildlifereservoir in New Zealand.

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 36


bison herd in WBNP poses a risk primarilyto the healthy neighbouring wild bisonpopulations. There has been no evidenceof spillover of TB to other wildlife ordomestic livestock. Plans wereannounced in 1990 to depopulate theWBNP, but were abandoned due topublic opposition from the aboriginaland wildlife protection organisations.A series of multi-stakeholder andadvisory committees and researchprogrammes were set up in order todevelop a management plan for thedisease reservoir in the WBNP, but thesewere eventually dissolved due to lack ofcontinuity in funding and stakeholderinterest. An interimstrategy for thecontainment of TBand brucellosiswithin the park waspublished in June2004. This includesthe creation ofbison-free bufferzones around the park, an educationprogramme and enhanced surveillanceof commercial bison and cattle in theagricultural area west and south of the park.

Between 1997 and 2005 bovine TB was diagnosed in approximately 30 free-ranging elk, or wapiti (Cervus elaphusmanitobensis) in and around the RidingMountain National Park, situated in the middle of an agricultural area insouth-western Manitoba. M. bovisinfection is now considered endemic inthe park’s wild elk population, with somespillover detected in white-tailed deer(Odocoileus virginianus) and wolves(Canis lupus). Intensive skin testing of cattle herds within a special TBeradication area around the park hasconfirmed infection in six cattle herds inthe same period. This has been attributedto direct and indirect contact of cattlewith infected wild deer. Although both

the prevalence of infection and the deer density in the region appear to be low, infection is perpetuated by thecongregation of wild deer in unnaturallyhigh numbers at feed sites. In 2001,federal and provincial authoritiesinstituted a TB management plan for thisarea, consisting of one to three-yearlytesting of cattle herds around the park,improved TB surveillance in wild deer,erection of deer-proof fencing, legislationto deter deer baiting and management of the wild deer population by extendedhunting seasons and wolf conservation.

Further challenges to the eradication ofTB in Canada arise from sporadic TB

incidents in exoticspecies atzoologicalcollections andincompleteslaughterhousesurveillance ofculled adult cattle,bison and deer.

Nevertheless, at the end of 2004 thiscountry was nearing the completeeradication of bovine TB from cattle andfarmed bison. According to domesticstandards for cattle and farmed bison(harmonised with those of the USA since1 January 2003) all ten provinces andterritories of Canada, except Manitoba,are officially TB free. The TB incidence incattle and farmed bison herds is very low.During the ten-year period from June1993 through November 2003, M. bovisinfection was confirmed in 12 herds ofcattle and captive bison in eight separateoutbreaks affecting five of Canada’sprovinces. This figure includes the cattlebreakdowns around Riding MountainNational Park. A further breakdown wasreported in 2004.

kCONTINUED ON p38


According to domestic standards for cattle and farmed bison all ten provinces and territories ofCanada, except Manitoba, areofficially TB free.

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 37


United States of America

Brief history of the TB eradicationprogramme in the USABovine TB used to be the most prevalentinfectious disease of livestock in the USA.In the early part of the 20th century M. bovis infection was also responsiblefor a significant proportion of TB cases inthe human population. In 1917 the PlantHealth Inspection Service (APHIS) of theUS Department of Agriculture (USDA)launched a state-federal TB eradicationprogramme to eradicate bovine TB. At that time, the national reactor rate in the USA was 5% of all cattle tested. In 1940 this fellbelow 0.5% of allthe cattle tested inevery county. There was someregression duringWorld War II, but reactor rateshad fallen to 0.06%by 1969. As TBprevalence declined, the emphasis of theprogramme switched from area testingto slaughterhouse surveillance andinfected herd depopulation, as in Canadaand Australia. In 1987 regulations weretightened to reduce the risk of infectionfor cattle feedlots importing steers fromMexico. Most States and cattle herds inthe USA are now officially TB free and thereactor rate in cattle is currently less than0.02%. Endemic foci of TB in domesticcattle are confined to New Mexico,southern Texas along the Mexicanborder, northeast Michigan and, morerecently, northwest Minnesota. Thesefour areas comprise the vast majority ofherds under restriction and have resistedTB eradication efforts for several years.The persistence of infection in at leastone of these endemic foci is linked to the presence of a wildlife reservoir of M. bovis (see below). Meanwhile, herdbreakdowns continue to occur sporadically

in accredited-TB free States at a rate ofone to five per annum, with no obviousgeographical pattern. The majority ofthese breakdowns are detected throughslaughterhouse surveillance and herdtrace-back.

Main features of the TB eradicationprogrammeTB surveillance by skin testing, includingperiodic tests of cattle herds, plays a relatively minor role in the USA. Fewer than one million skin tests perannum are conducted in a national cattle herd of approximately 99 millionhead (compare that with the 4.8 million

animal tests carriedout in GB in 2005 in a population ofnine million cattle). In general, routinetesting of cattle,farmed bison andcaptive cervid herdsis limited to the fewStates or zones

where a TB eradication programme isunderway to attain accredited TB freestatus. The remaining tests are conductedfollowing trace-back of tuberculous cattledetected at slaughter and prior tointerstate or international movement.Testing of captive cervids, bison andgoats is also contemplated for herdswishing to attain or maintain TBaccredited status.

The USDA’s Uniform Methods andRules (akin to the SVS TB Chapter), setout the minimum standards for themaintenance of TB-free accredited herds,the application and interpretation ofofficial TB tests in cattle, farmed bison,goats and captive cervids and therequirements for interstate trade ofcattle. Briefly, initial screening of cattle,bison or goats not known to be at risk isby the caudal fold SIT. The SICCT test isonly applied for increased specificity onanimals that react to the caudal fold SIT.



Endemic foci of TB in domestic cattle are confined to New Mexico, southern Texas along the Mexican border, northeastMichigan and, more recently,northwest Minnesota.

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 38

39

The cervical SIT with double strengthtuberculin is the recommended test inexposed cattle and bison herds and theofficial test for routine use in captivecervids. Whilst skin tests are applied andinterpreted in bison just as in cattle,special facilities are essential to safelyhandle bison and minimise the stress onthe animals during testing. Although skintesting in the USA is normally undertakenby federal, State or accredited veterinarians,Federal and State technicians mayperform routine screening tests underdirect veterinary supervision. The γ-IFNtest has been licensed in the USA for usein cattle and bison since December 2001.This assay is nowroutinely used as areplacement for (orin parallel with) theSICCT when re-testing SIT reactorsin herds presumedto be free frominfection. In infected herds, the γ-IFN test is used in parallel with the caudalfold or cervical SIT as an alternative todepopulation. Whole-herd slaughter isthe option of choice if M. bovis infectionis confirmed in a cattle or farmed bisonherd. When this is not feasible, the herdis quarantined until it passes two tests at60 days intervals, followed by oneadditional test 180 days later. Herds withunconfirmed TB breakdowns (non-visiblelesion reactors only and no M. bovisisolated) can be released from quarantineafter a single negative test of the entireherd 60 days after slaughter of thereactors. After release from quarantine,the herd is marked forward for fiveannual whole-herd tests.

Because of the very low incidence ofthe disease, bovine TB surveillance in theUSA relies primarily on the examinationof suspicious granulomas in slaughteredcattle, coupled with tracebackinvestigations. Slaughterhousesurveillance for TB accounts for 95%

of all infected herds detected every year and is the responsibility of federalveterinarians and meat inspectors,supported by State meat inspectors in the smaller plants. In 2000, after a steadydecline in the number of granulomasubmissions, the USDA launched a plan for enhanced TB surveillance in the 40 large regional slaughterhousesprocessing 94% of all adult cattle. Meat inspectors are expected to submitat least five suspicious granulomas per10,000 head of adult cattle killed in aplant. Inspectors receive cash awards for positive submissions and for eachinfected herd successfully back traced.

The sustainedincrease in thenumber ofgranulomas since2001 has beenattributed to thisprogramme.Despite these

successes, however, herd tracebackinvestigations only achieve a 50-70%success rate because of the lack ofcentralised cattle identification andtracing systems in the USA.

The USDA ranks states and zones into five possible stages as they makeprogress towards TB eradication. Thesestages are generally based upon theincidence of the disease in cattle herds.For a State or zone to be in the topcategory (‘Accredited TB free’ status)there must have been no confirmed TBbreakdowns on holdings other thanfeedlots for at least five years. The Statein question must also have a set ofstringent regulations governing livestockdealers, maintain surveillance of cattlethroughout the supply chain and requirerecords to be kept to allow animaltraceability. Accredited TB free State orzone status must be renewed annually. IfM. bovis infection is confirmed in a herdwithin an accredited-free State or zone,


kCONTINUED ON p40


In infected herds, the γ-IFN test isused in parallel with the caudal fold or cervical SIT as an alternative to depopulation.

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 39


this status can be retained as long as theherd in question is depopulated and anepidemiological investigation concludesthat there was no evidence of TB spread.If two or more infected herds aredetected within a 48-month period, theState or zone is downgraded one level to‘modified accredited advanced’ status. As of January 2006, all States had‘accredited TB free’ status for cattle andbison except Michigan, Minnesota, NewMexico and Texas. These four States (or parts of them) were classified by the USDA as ‘Modified AccreditedAdvanced’ due to endemic TB foci or (in the case of Minnesota) recent TBbreakdowns ofuncertain origin.Federal rulesprescribe that allbreeding cattle 18months of age andolder shipped out ofthose zones must beskin tested within60 days ofmovement. This is without prejudice ofany additional controls imposed by theimporting States.

TB in wildlife in the USAAs described above, bovine TB has beennearly eradicated from cattle and captivebison in the USA. TB in farmed deer wasidentified as a potential problem in theUSA in 1991, when an outbreak of TB in captive elk in Canada was back tracedto stock imported from the USA. Thisincident led to the incorporation ofcaptive cervids into the USDA TBeradication scheme in 1994. There arefive possible State or zone designationsfor TB in farmed deer, which do notnecessarily coincide with the status forcattle and bison. According to thissystem, all States in the USA are currently‘Modified Accredited’ for captive cervids.Between 1991 and 2004, TB was disclosedin 41 captive cervid herds in 18 states

(30 of which were depopulated), but theincidence of TB in captive deer in the USAhas been declining since 2000. In wildcervids, however, TB is an emergingdisease that represents the greatestimpediment to the complete eradicationof TB from livestock in the USA.

Since 1994, an outbreak of bovine TB has affected cattle herds and free-ranging white-tailed deer in a 13-countyarea in the north-eastern lowerpeninsula of Michigan. This resulted inthe loss of the State’s accredited-free TB status in 2000. Between the spring of1995 and 1 January 2005, more than1,050,000 cattle in over 17,000 herds

have undergoneskin testing inMichigan. Infectionhas beenconfirmed in 33cattle herds andone captive deerherd, but noinfected farmswere identified in

the 2004-2005 testing season.Transmission from tuberculous free-ranging deer was the likely origin for amajority of these incidents. Between1995 and 2005, 509 of 138,000carcases of free-ranging white-taileddeer examined by the State of Michigantested positive for M. bovis. DNA typingof M. bovis isolates from deer, cattle andother mammals have showed that aunique strain is involved. This was thefirst self-sustaining outbreak of bovineTB described in free-ranging NorthAmerican cervids and has become themost intensively examined example of TBin wild deer worldwide. M. bovisinfection has also been reported in twodeer hunters, four of 1,400 culled elkand in a variety of wild carnivores at a very low prevalence (except in thescavengers of deer carcases such ascoyotes Canis latrans). The low numbersof tuberculous non-cervid wildlife, the



There is a strong association between the risk of TB breakdown on cattle farms and husbandrypractices that favour deer access to cattle housing and feeding areas.

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 40

41

variety of species involved and thegeographic spacing between cases areindicative of spillover from free-rangingdeer, rather than endemic TB. Infectionhas not yet been disclosed in the NorthAmerican badger (Taxidea taxus).American authors have estimated thatM. bovis infection in cattle spilled overonto the white-tailed deer population innorth-eastern Michigan in the first halfof the 20th century, when TB in cattlewas still quite prevalent. AlthoughMichigan’s cattle herds were declaredaccredited TB free in 1979, it is thoughtthat the prevalence of M. bovis infectionin wild deer was maintained at lowvalues until it reached a detectable levelin the mid-1990s, following a dramaticgrowth of the deer population. There is a strong association between the riskof TB breakdown on cattle farms andhusbandry practices that favour deeraccess to cattle housing and feedingareas. It appears that environmental

(feedstuffs) contamination is a keymechanism of M. bovis transmissionbetween white-tailed deer and cattle.White-tailed deer are a native wildlifespecies highly valued by the Michiganpublic and the hunting fraternity inparticular. The practice of deer feedingand baiting by hunters, farmers andwildlife enthusiasts became popular in Michigan in the 1970s and 1980s to help deer survive during harsh winters.Deer gathering in unnaturally highnumbers at feeding sites, increase therisk of deer-to-deer transmission of M. bovis (Figure 9). In 1995, the StateGovernment passed legislation banningthe feeding and baiting of deer acrossMichigan. Grants have been provided for farmers to erect deer-proof fencingaround farms and feed storage areas.These measures, combined withextended deer hunting seasons andunlimited hunting quotas have reduced


kCONTINUED ON p42


Figure 9Free-ranging white-tailed deercongregating at an artificial winterfeed site in Michigan, USA (photographreproduced from the BovineTuberculosis in Michigan Web site at:www.michigan.gov/bovinetb)

9

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 41


the wild deer population by half since1995. Considerable effort has also goneinto educating cattle farmers and deerhunters and engaging the variousstakeholder groups. The apparentprevalence of M. bovis infection inMichigan’s wild deer has fallen by 64%since 1997, but complete elimination ofTB from cattle and wildlife may takeanother decade.

Latin America and the CaribbeanThe prevalence of bovine TB and themeasures applied to control it differ quite a lot across Latin American andCaribbean countries (Table 7). Argentinaand Brazil (the two major cattle breedingand beef exporting countries in theregion) launched national compulsorytuberculin test and slaughter programmesin 1999 and 2001, respectively. Despitethese efforts, the TB situation in bothcountries is still far from being undercontrol, as evidenced by a high incidenceof reactor herds (and animals) and arather worrying rate of cattle, water



7

Incidence of reactors Countries Approximate totalcattle population(2003)

Very low incidence All the Caribbean countries except 61.6 million(<0.1%) or virtually TB Haiti and the Dominican Republicfree countries Panama

HondurasBelizeColombiaVenezuelaSurinamUruguay

0.1 – 1% Mexico 47.4 millionDominican Republic (of which 29.5 millionNicaragua head in Mexico)Costa RicaEl SalvadorParaguay

>1% or unknown Haiti 265.1 millionGuatemala (of which 53.5 millionArgentina head in Argentina and Bolivia 182 million in Brazil)BrazilChileEcuadorGuyanaPeru

Table 7Classification ofLatin American andCaribbean countriesaccording to theincidence oftuberculin testreactors (source:Pan-AmericanHealth Organisation,OIE and de Kantorand Ritacco 2006)

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 42

43

buffalo and pig carcase condemnationsresulting from TB at slaughterhouses.Furthermore, the occurrence of TB inlarge tracks of Argentina and Brazil is still largely unknown. Mexico has beenramping up its TB control programmeover the last decade and madesubstantial progress in the beef herds ofthe northern states. However, significantpockets of infection remain in dairy herdsand, as a result, the export of Holsteincattle to the USA is still banned. Bycontrast, Cuba, Jamaica, Panama,Uruguay and Venezuela have all buteradicated TB from their national herdsafter long-running campaigns. The mainhurdles to theeradication of TB inLatin Americaappear to be aninadequateveterinaryinfrastructure,incompleteepidemiologicaldata and a lack ofstandardised diagnostic methods andreagents across the region. On thepositive side, no reservoirs of M. bovisinfection have been reported in SouthAmerican wildlife or in native species of domesticated mammals (e.g. New World camelids).

AfricaBovine TB was considered a diseaseexotic to Africa, although it is nowadayspresent in cattle in almost all Africancountries. Information on the occurrenceof bovine TB across this continent is very scarce, although there is sufficientevidence to indicate that M. bovisinfection is widespread and found at highprevalences in some animal populations.It is believed that domestic cattle are themain host to M. bovis in Africa, but theorganism has also been isolated fromsmall ruminants, dromedaries and a largenumber of wild mammals. The public

health threat of M. bovis in Africa is alsoconsiderable, given that nearly 85% of itscattle and 82% of its human populationare estimated to live in areas where theinfection is prevalent or only partiallycontrolled. However, it is difficult toassess the precise contribution of M. bovis infections from animal sourcesto the human TB epidemic in Africa.Limited laboratory resources preclude thedifferential diagnosis of human M. bovisand M. tuberculosis infections.

It is in South Africa where perhaps moreinformation has been published aboutbovine TB in the continent. A national TBeradication scheme was introduced in

that country in1969. Althoughcompleteeradication of thedisease has neverbeen achieved, theoverall incidence incommercial cattleremains relativelylow. High morbidity

breakdowns in cattle herds have occurredsporadically since the first report of thedisease in 1880. Those were believed tobe the source of the first incident of TB inAfrican wildlife, described in 1928 whenM. bovis was isolated from lesions infree-ranging greater kudu (Tragelaphusstrepsiceros) and lesser ungulates in theEastern Cape province. There have beennumerous reports of suspected andbacteriologically confirmed M. bovisinfections in African wildlife since then(Table 8). Nowadays, the African (Cape)buffalo (Syncerus caffer) in Uganda andSouth Africa, the Kafue lechwe antelope(Kobus leche kafuensis) in Zambia and,possibly, the greater kudu in South Africaare all regarded as true maintenancehosts of M. bovis and a source ofinfection for cattle and other Africanwildlife. The best studied of these wildlifereservoirs (and probably the most


kCONTINUED ON p44


The public health threat of M. bovis inAfrica is also considerable, given thatnearly 85% of its cattle and 82% ofits human population are estimated tolive in areas where the infection isprevalent or only partially controlled.

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 43


important host for M. bovis in Africa) isthe buffalo in the Kruger National Park,South Africa’s largest wildlife reserve.There is strong circumstantial evidencethat bovine TB was introduced between1950 and 1960 by infected domesticcattle grazing in close proximity withbuffalo on the southern border of thepark. TB was eventually discovered in theKruger’s buffalo population in 1990 andhas now reached epidemic proportions,

with prevalences of infection as high as90% in some herds. Infection ismaintained in buffalo herds mainlythrough aerosol transmission. Carcases of tuberculous buffalo are a source ofinfection to predators and scavengingspecies via the oral route. Other wildanimals contract bovine TB throughenvironmental contamination. Theexpanding geographical range andprevalence of infected buffalo herds pose


8

Table 8Reports of M. bovisisolations from free-living African wildlife

*presumed true maintenance hosts of M. bovis (in regions where infected individuals have been found). The role of the lion as a maintenance or spillover host in South Africa is debatable.

Country Host species Location Year of report

Uganda African buffalo (Syncerus caffer)* Ruwenzori Nat’l Park 1963, 1970s, 1997Warthog (Phacochoerus aethiopicus) Ruwenzori Nat’l Park 1970s, 1997

Kenya Olive baboon (Papio cynocephalus) Maasai Mara Game 1980sReserve

Zambia Kafue lechwe (Kobus leche Kafue river basin 1956, 1970s-90skafuensis)* Private game ranch 1998

near LusakaBushbuck (Tragelaphus scriptus) Private game ranch 1998

near LusakaEland (Taurotragus oryx) 1984

Tanzania Wildebeest (Connochaetes taurinus) Northern Tanzania 2005Topi (Damaliscus lunatus) Northern Tanzania 2005Kudu (Tragelaphus spp.) Northern Tanzania 2005

South Africa African buffalo* Several private game 1998reservesHluhluwe-iMfolozi Park 1986(HiP)Kruger National Park Several reports(KNP) from 1990 to date

Greater kudu (Tragelaphus Eastern Cape Province 1928, 1940, 1992Strepsiceros)* KNP and HiP Several reports

from 1996 to dateDuiker (Sylvicapra grimmia) Eastern Cape Province 1928Bushbuck Eastern Cape Province 1928Chacma baboon (Papio ursinus) KNP and HiP 1995Lion* (Panthera leo) KNP and HiP 1995Cheetah (Acinonyx jubatus) KNP 1995Leopard (Panthera Pardus) KNP 1999Warthog KNP 2005Bushpig (Potamochoerus porcus) HiP After 1986

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 44


a threat to rare wildlife species in thepark and in neighbouring countries. Thisalso represents a big challenge on long-term management strategies to prevent re-infection of domestic livestock. Where co-grazing of wild ungulates and domestic cattle takes place, recyclingof infection is likely to continue withsubsequent spillover to predators andscavengers at the top of the food chain.In the absence of an effective vaccineagainst TB and validated diagnostic toolsfor most species, M. bovis infection islikely to persist in free-ranging Africanwildlife for the foreseeable future.

Further readingAvailable on request to the author.

Ricardo de la RuaDomenech is aVeterinaryAdvisor withDefra specialisingin bovine TB

l Information


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 45


IntroductionIn April 2003 the Veterinary LaboratoriesAgency (VLA) received funding fromDefra to carry out a nine month jointproject with the Environmental ResearchGroup Oxford (ERGO), to examine theCattle Tracing System (CTS) data andassess the importance of cattle movementin the spread of bovine tuberculosis (BTB).The work followed on from an earlierVLA/ERGO collaboration assessing theenvironmental correlates of BTB distribution.

Geo-referencing and datatransformations carried out on CTS dataBefore any modelling of CTS data couldbe undertaken, two main processingoperations had to be carried out. These were firstly the geo-referencing of the CTS data – to enable its spatialrepresentation in a GeographicalInformation System (GIS). The secondstage required the ‘pairing’ of movementrecords. Because of the structure ofmovement records in CTS, where a singlemovement is stored in two elements –leaving a location (‘off’) and arriving at adestination (‘on’) – these two elementsneeded to be linked as a single paired recordbefore further analysis could proceed.

Various methods were used to geo-

reference CTS locations. These methodseither involved obtaining an Easting andNorthing from CTS spatial data (thelocation address or its OS reference) orobtaining it by linking to other Defra datasets (e.g. VetNet or the Livestock Census).98% of the locations associated withcattle movement were successfully geo-referenced.

The ‘pairing’ of movement records in CTS allows the creation of a logicalmovement history for a given animal. To enable the creation of logical movementhistories the VLA developed a set of rules,with an underpinning rationale that if anypart of an animal’s movement historycould not be resolved, or was clearlyincorrect (e.g. movements after death),then all movements associated with thatanimal were excluded from the pairedmovements dataset used in subsequentanalysis. This approach resulted in 17.4million animals being included in theanalysis out of a total of 25.8 million,giving an animal rejection rate of 32.4%.Figure 1 shows the proportion of animals‘included’ by birth cohort. This figureclearly shows the improvement over timeof animals having a logical movementhistory which, since 1999 has stabilised at around 80%.

Bovine TB: Modelling and predicting itsdistribution in GB using CTS data

Andrew MitchellA u t h o r

100

80

60

40

20

01996 1997 1998 1999 2000 2001 2002 2003

Figure 1Proportion ofAnimals ‘included’ by Birth Cohort

1

Perc

enta

ge

Bovine TB: Modelling and predicting its distribution in GB using CTS data

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 46

Characterisation of the national herdusing CTSOnce the preparatory cleaning and dataprocessing transformations had beencarried out, it was possible to examinesome of the basic characteristics of cattlemovement and cattle populations in GB.

These characteristics included the age (atdeath) of the national herd (Figure 2); themovement of animals between Defraregions (Figure 3), and the details ofmovements out of particular regions(Figure 4).


150,000

100,000

50,000

0

10 20 30 40 50 60

Months

0

Figure 2Age at deathfrequencydistributions

2

13%

35%

25%

10%

17%

14%

27%

16%

17%

26%

3a

EastNorthScotlandWalesWest


Figure 3The distribution oforiginating region ofmovement in 2002

(a) Originatinatingregion of movement

(b) Distribution ofholdings on VetNet

3b

The regional origin of 1.84 million farm-to-farm cattle movements in 2002, shown in (a), was broadly similar to the distribution of cattle holdings in the VetNet database (b).

kCONTINUED ON p48


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 47


It was also possible, by combining CTS data with existing TB data sets, toinvestigate specific TB related questionssuch as the movement of cattle betweenthe different parish testing intervalregimes (Figure 5).

Extensive use has since been made of this type of data. The pattern ofmovements between different areas,holding types, herd types and parishtesting regimes have been examined toattempt to gauge the effects of differentpre-movement testing scenarios.



4a

4b

Figure 4Movements out ofthe Western regionin 2002

(a) Receiving farmlocation

(b) Farm-to-farmcattle movementsfrom the WesternRegion

120,000

100,000

80,000

60,000

40,000

20,000

01 2 3 4

In 2002, there were 453,505 farm-to-farm cattle movementsoriginating from the Western Region.

Figure 5Movements betweenparish testingregimes in 2002

Year 2002

(a) 1 Year

5a

5%3%

0%5%

87%


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 48


kCONTINUED ON p50

(b) 2 Year

(c) 3 Year

(d) 4 Year

60,000

50,000

40,000

30,000

20,000

10,000

01 2 3 4

150,000

100,000

5,000

01 2 3 4

150,000

100,000

500,000

01 2 3 4

5b

5c

5d



Year 2002

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 49


Results of modelling the spread of TB The main purpose of the data cleaningand transforming of the CTS data was to allow the CTS data to be put into GridData format (Figure 6), so that it could be put into the GIS models and used as a potential predictor of BTB.

Two different modelling approacheswere undertaken on the cleaned andtransformed data. The first beingdistribution modelling, which seeks toestablish what parameters are importantin the spread of BTB and which can thenbe used for short-term BTB predictions.The second method used was stochasticsimulation (a simulation governed by thelaws of probability); this provides a dynamicdisease spread model, which can be usedfor giving medium term predictions.

Distribution modelling uses logisticregression to establish a relationshipbetween the presence of BTB and a series of significant predictor variables.The final equation is then applied to thecomplete predictor grid coverages togenerate a predicted BTB distribution.

The predictor variables used are shownin Table 1. These included those found to be significant from the previousmodelling work (including indicators ofseasonality, climatic variation, vegetation,land use, demography, cattle density and an index of badger presence) plusnew predictors obtained from the CTS.All the CTS variables were found to besignificant predictors of disease presence,with the proportion of movements intoan area from infected areas consistentlyoutperforming all other variables in the model.

Stochastic simulation modelling againuses logistic regression to model the spreadin BTB distribution. The model is appliedto a known starting distribution andproduces predictions over a number of years that can be projected into the medium-term future.

For the simulation method, predictorswere identified separately for ‘core’ and‘remote’ TB areas. Core areas weredefined as those cells in which diseasehad been found in at least two of theprevious three years, with remote areasbeing all other areas. For core areas onlytwo variables were significant (thesewere the BTB status of both the cell and the surrounding area in previousyears) whilst for remote areas theproportion of movements from infectedareas was again the best predictor ofBTB. This suggests that the variation inthe levels of movement of infectedanimals has a greater impact on diseaselevels in remote areas where the diseasehas yet to become established.

Figure 7 shows an example of thepredictions given by the simulationmodelling. These show good agreementwith those obtained by distribution


KeyTesting frequency / years

1

2

3

4

(e) Parish testingfrequencies asrecorded on VETNET @ 17/12/03

5e



Year 2002

10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 50

51

modelling and they highlight the potentialfor the local spread of the disease inCumbria and the Scottish borders.

The futureThe work carried out and the experiencegained in using the CTS data will enableits future use for the epidemiologicalanalysis of cattle movements and disease spread.

The modelling work undertaken willhave three main uses for future TBcontrol and policy. It will allow theidentification of new TB risk areas and high-risk movement types. Also, by adjusting movement data within themodel, the effect of future movementrestrictions can be modelled.

kCONTINUED ON p52


Type Variable

Anthropogenic Distance to roads (DROADS) and city lights(LIGHT)

Demographic Human population density (HPOP)

Land cover Percentage cultivation and managed grassland(PCU); Proportion urban and suburban landcover (PUR); Woodland (PWO), Open Woodland(POW); Grassland (PGR); Water (PWA);Normalised Deviation Vegetation Index (NDVIa);Length of Growing Period (LGP).

Geographic Longitude (LONG), Latitude (LAT)

Topographic Elevation (ALT)

Temperature Air Temperature (ATa); Land SurfaceTemperature (LSTa); Middle Infrared Reflectance(MIRa);

Water and moisture Vapour Pressure Deficit (VPDa); Distance toRivers (DRIV); Potential Evapo-transpiration(PEV)

Zoological Cattle density (CAD); Holding density (HOD);Proportion of dairy cattle (PDC); Herd size (HES);Badger record density (BAD); Distance tonearest recorded badger presence (DBR).

Cattle movement Total movements in (TM); Total movements infrom infected areas (TMI); Proportion ofmovements from infected areas (PMI).

BTB persistence and local spread Number of years of past BTB infection in the5km cell (NYBTB); number of infected cells inthe previous year in a 5km radius doughnutwindow (DNT); Distance to the nearest cell withBTB present (DBTB)

1

Table 1Predictor data usedin the analysis of BTB occurrence


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 51


Further reading Available on request to the author.

AcknowledgementsI would like to acknowledge the fellowcontributors at the VLA to this project –Dr Richard Clifton-Hadley and JoshuaMawdsley. I also wish to acknowledgethe large contribution made to thisproject by Dr David Bourn, Dr Willy Wintfrom the Environmental Research GroupOxford (ERGO) and Dr Marius Gilbertfrom the Free University of Brussels. The time given by Mel Reader (CTS consultant working for Defra) in explaining the complexities of the CTS data was very much appreciated.


Figure 6Example of CTS inGrid data format(grid cell resolution –5km)

Figure 7The probability ofBTB presence in2005 as predicted bysimulation modelling

Distribution of cattle ‘on’ movements that came from grid cells classified as infected in 2000.

6 7

AndrewMitchell is theTB databasemanager, in theCentre forEpidemiologyand RiskAnalysis at theVeterinaryLaboratoriesAgency inWeybridge.

l Information


10493 AW Defra GVJ_03-52 29/8/06 14:20 Page 52

53

Bovine tuberculosis is a chronic bacterialdisease of animals and humans caused byMycobacterium bovis. This bacterium isan obligate pathogen, but it can survivefor several months in the environment in moist soil and in darkness. It is highlysusceptible to sunlight (UV radiation) and remains viable for long periods (>6 months) in frozen tissues.

Cattle are major hosts but some wildlifepopulations act as reservoirs for theagent. M. bovis has been isolated fromnonhuman primates and also in buffalo,bison, zebu cattle,sheep, goat,equine, camel, pig,wild boar, cervidae,antelope, dog, cat,fox, mink, possum,ferret, rat, llama, kudu, eland, tapir, elk, elephant, situtunga, oryx, addax,rhinoceros, feline (exotic and domestic),squirrel, otter, seal, hare, racoon andother mammals. There is evidence tosuggest that M. bovis has becomeestablished in the environment in Great Britain where badgers are

considered to be an important reservoirof infection.

Infection in cattle is usually detected in the live animal on the basis of thetuberculin skin test. After death, it isdiagnosed by post-mortem examinationand then confirmed in the laboratory byhistopathological and bacteriologicaltechniques.

The Foot and Mouth Disease outbreakin 2001 brought a temporary halt to thecattle skin testing programme and whenthis resumed there proved to be a

dramatic increase inthe number of tissuesamples submittedfor laboratoryexamination.Laboratory tests are

lengthy and expensive and consequently,the drive for improvement hasintensified.

BacteriologyDue to its hazardous nature, all work isconducted in a containment level threelaboratory. Field samples and cultures are


The laboratory diagnosis of Bovine Tuberculosis

Keith Jahans andDanny Worth

A u t h o r s

1 2

Figure 17H11, Stonebrinksand Lowenstein-Jensen mediashowing colonies of M. bovis

Figure 2Ziehl-Neelsenstained slide of M. bovisshowing ‘chording’

Infection in cattle is usually detectedin the live animal on the basis of thetuberculin skin test.

kCONTINUED ON p54


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 53


handled in a Class one microbiologicalsafety cabinet (MSC) with a Defraapproved disinfectant to hand. Staff wearprotective clothing and surgical gloves.

Tissues for cultural examination arehomogenized with a solution of 5%oxalic acid, in order to destroy anycontaminating organisms, centrifuged at1100g for ten minutes and the resultingdeposit washed and re-suspended in10ml of 0.85% sterile saline. Thesuspension are sown on to the followingmedia slopes: three Stonebrink’s, twoLowenstein Jensenbase (LJ), twoLowenstein Jensen +Glycerol (LJG), twoLowenstein Jensen +Pyruvate (LJP) andthree Middlebrook7H11. The slopes are incubated at 37°Cfor up to six weeks and then examinedfor growth.

M. bovis will grow on all these mediabut most strains are inhibited by glyceroland enhanced by pyruvate. The organismhas a distinctive morphology on themedia (Figure 1). On 7H11 in particular, it produces flat, matt colonies that floatoff the media when the condensationliquid is run up the slope. The media will

also support the growth of a wide range of other mycobacteria. Ziehl-Neelsen (ZN) staining may also prove useful indiagnosis. Mycobacteria are acid-fast andstain red. M. bovis sometimes produces a typical ‘chording’ effect (Figure 2).Diagnosis is based on these culturalobservations with the help of histology.

Molecular typing Individual strains of M. bovis can bedistinguished according to their geneticfinger-print. The principal typing method

is known asspoligotyping. It is based on DNApolymorphismpresent at oneparticularchromosomal locus,

the ‘Direct Repeat’ (DR) region, which is unique to the Mycobacteriumtuberculosis complex bacteria of whichM. bovis is a member. This methoddistinguishes between strains byidentifying the presence or absence ofknown spacer oligonucleotides (Figure 3)in the DR region, hence the termspoligotyping.

Strains are grown, heat killed andsubjected to Polymerase Chain Reaction

Figure 3Representationillustratingdifferences in spaceroligonucleotides

3


Diagnosis is based on these culturalobservations with the help ofhistology.

Spoligotyping

Genome Direct Repeat Region

Strain A

Strain B

Strain C


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 54

55

(PCR) using primers based on the DRregion allowing amplification of thespacers in between. Forty-three uniquespacer oligonucleotides are hybridisedpermanently to a nylon membrane. The PCR products from the killed strainsare hybridised to the membrane at rightangles to the spacers. After Hybridisationthe membrane is developed by anenhanced chemiluminescence system. An image showing the presence orabsence of spacers, giving the impressionof individual bar codes representing each sample, is developed byautoradiography (Figures 4 and 5).

The different spoligotypes, combinedwith spatial clustering, has provideduseful information on the spread of the disease (Figures 6 and 7). Data from spoligotyping is now availableto the SVS to help trace the source ofdisease breakdowns. Types nine and 17 have been found to compriseapproximately 70% of M. bovis positivecultures. Because of this, an alternativetyping method is being introduced toprovide further differentiation.Genetic loci containing variable

numbers of tandem repeats (VNTR) form the basis for human gene mappingand identification, forensic analysis andpaternity testing. The genome sequenceof M. tuberculosis revealed a largenumber of targets of tandem repeats and so this was the method chosen.Since applying this technique, VLA hasshown that the same spoligotype may be split into a number of VNTR types. In addition, VNTR has been shown todiscriminate between M. bovis isolateswith the same spoligotype but fromdifferent geographical areas of GreatBritain. Used in conjunction withspoligotyping it has provided valuableinformation about the evolution of the M. tuberculosis complex strains.

Acid-fast organisms which are notrecognised as M. bovis by spoligotypingor do not show growth characteristicstypical of M. bovis are tested by multiplex PCR. The amplification processis moderated by several oligonucleotideprimers that, typically, are 20-30nucleotides long. This is where themultiplex PCR differs from a conventionalPCR which generally only uses one set ofprimers. The primers are single-strandedDNA that have sequences complementaryto the flanking regions of the targetDNA. Following PCR, the amplificationproduct can be detected using gelelectrophoresis. Visible bands aremeasured against reference DNA bandsadded at either end of the gel preelectrophoresis. When the gel is run thereference DNA produces a ‘ladder’ ofbands at intervals of 100bp. All membersof the Mycobacterium genus produce a band at 1030bp. However, the TBComplex (which includes – M. bovis, M. microti, M. tuberculosis and M. africanum) show a second band at372bp, M. avium a second band at180bp and M. Intracellulare a secondband at 850bp.

kCONTINUED ON p56

4


Figure 4Autorad showingdifferent ‘bar code’spoligotype patterns


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 55


Molecular typing direct from tissuesThe traditional cultural techniques cantake up to six weeks for growth to becharacterised and speciated. For thisreason, investigations are being carriedout using PCR to detect M. bovis directlyfrom tissues. DNA was extracted frombovine lymph glands by means of acommercial kit (NucliSens, Organon-Teknika). The PCRassays targetedcytochrome b inorder to detectbovine DNA, the RD4region specific to theM. bovis genomeand the insertionsequence IS1081which is specific for organisms of theMycobacterium TB complex. Initialfindings show that this method is not assensitive as culture. However, it could be used as an initial rapid screening testuntil the results of cultural tests becomeavailable. Studies are underway toinvestigate the use of an automatedsystem for animals showing TB lesions-likelesions at slaughter.

Further developmentsCurrent laboratory procedures arecontinually being reviewed, developedand evaluated. There is much to be doneand a great deal has been achieved.

Twenty years ago guinea pigs wereused routinely in the diagnosis of bovinetuberculosis. Improved selective mediahas now made this unnecessary. The

prescribed wayto distinguishbetween bovineand human typetuberculosis wasto inoculate a rabbit, as thelaboratory rearedanimal is highly

susceptible to M. bovis whereas M. tuberculosis fails to produce disease.Nowadays not only can we distinguisheasily between these two species bymolecular typing but we can alsodistinguish between strains.

Liquid culture systems, such as thoseproduced by Becton Dickenson, havebeen developed for the isolation of M. tuberculosis from human clinical


Acid-fast organisms which are not recognised as M. bovis byspoligotyping or do not show growthcharacteristics typical of M. bovisare tested by multiplex PCR.

Spoligotypes 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43

H37Rv � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

DT � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

BCG � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

AN5 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

9 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

17 � � � � � � � � � � � � � � � � � � � � � � � �

10 � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

11 � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

12 � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

13 � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

15 � � � � � � � � � � � � � � � � � � � � � � � � � �

20 � � � � � � � � � � � � � � � � � � � � � � � � � �

21 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

22 � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

25 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

35 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �

Spoligotypes

5


Figure 5Representation of spoligotype ‘bar codes’

10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 56

57

samples and are being used by someveterinary laboratories for isolating M. bovis. These systems are able toidentify M. tuberculosis after only 15days. Work at Weybridge by the TBresearch group has shown that M. bovisdiffers from M. tuberculosis in the way it metabolises certain key substrates.Recent studies at VLA Starcross and VLA Weybridge have shown that theBecton Dickenson MGIT 960 systemcould be used for M. bovis isolation and further work is being carried out to evaluate its used with farm animal and wild-life samples.

The sequencing of the M. tuberculosisand M. bovis genomes has meant thatwe are beginning to understand moreabout the biology of the organisms and the nature of the disease. This hasalready provided us with valuableinformation to fine tune and furtherdevelop our laboratory techniques. As our understanding increases so will the efficiency of diagnostic tests.

Thus the quality of the information that the VLA provides to the SVS can only be improved.


Figure 6Cattle Spoligotypes2000

Figure 7Cattle Spoligotypes2005

6 7

39101112131517192021222528323435515263657374758187

Newspol.txt

Gbab.shp

25

43457879

Newspol.txt

Gbab.shp

910111213151719202122

323435

8889

Keith Jahans isthe TB andAnthraxDiagnosisWorkgroupLeader in theStatutory andExotic BacteriaUnit at theVeterinaryLaboratoriesAgency.

Danny Worthmanages the TBDiagnosticLaboratory withinthe Statutory andExotic BacteriaUnit at theVeterinaryLaboratoriesAgency.

l Information


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 57


The introduction of pre and post-movement TB testing in Scotland forcattle from high incidence TB areas

The incidence of bovinetuberculosis in ScotlandScotland has, for many years, enjoyed a low and relatively stable incidence ofbovine tuberculosis (TB) in its cattle herdsFigure 1. From 1986-2000, the numberof new confirmed incidents each yearwas generally in single figures. The threeexceptions were in 1988 and 1990 when12 new confirmed incidents wererecorded and in 1996 when ten newconfirmed incidents were disclosed. The annual number of reactors over this period ranged from nine (1998) to167 (1992) and the larger numbers ofreactors reflect the years when a wholeherd slaughter was required. Whole herd slaughters are occasionally requiredin four yearly testing areas (all parishes in Scotland are on four-yearly testing)because the disease has longer toestablish itself between routine periodic tests.

In 2001, routine TB testing wassuspended as all available veterinaryresources were redirected to theeradication of the national foot andmouth disease outbreak. The two TBcases detected in Scotland in 2001 wereboth slaughterhouse cases. When testingresumed in 2002, 28 new confirmedcases were disclosed and it has takenuntil 2005 to see the numbers fallingback to 13 new confirmed cases, with a reasonable prospect, in most areas, of now returning to pre-FMD levels of TB.The increase in the number of TB cases in Scotland post-FMD can be put downto a combination of:

(a) The suspension of TB testing in 2001 leading to the detection of ‘twoyears worth’ of cases in 2002;

(b) A reflection of the worsening TBsituation in parts of England and Waleswhere Scottish farmers could trade cattlefreely and;

Dr MartynJ Blissitt

A u t h o r


250

200

150

100

50

0

1986 1989 1992 1995 1998 2001 2004

Year

Num

bers

1

Figure 1Incidence of BovineTuberculosis

2005

The introduction of pre and post-movement TB testing in Scotland for cattle from high incidence TB areas

10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 58

59

(c) Post-FMD restocking movements. In fact, only six new confirmed incidentsof TB in Scotland were on farms restockingpost-FMD (three in 2002 and three morein 2003). In each case disease was quicklyeradicated under the existing policies oftest and slaughter.

There remain two essential differencesbetween the patterns of TB outbreaksseen in Scotland and those in the higherincidence areas of England and Wales.Firstly, there is no evidence of recurrence

of disease on farms in Scotland andsecondly, there is as yet no evidence of‘hot spots’ of disease involving similarspoligo/VNTR types on neighbouringfarms. Both factors are taken as evidencethat there are no significant wildlifereservoirs of infection for cattle inScotland at the moment. This should also be true in the lower incidence areasof England and Wales. Indeed, there are

kCONTINUED ON p60


2

Figure 2TB in Scottish Wild Animals

2003

1989

1991

1993

1995

Legend

Red deer hind. Shot 12/12/03. Moy Parish. Pleurisy, peritonitis and pathology indicative of mycobacterial infection.

Sika stag. Submitted to Lasswade 17/11/95. Kintyre forest district. No pathology recorded.

Sika hind. Shot 08/12/93. West Lock Tarbert. Sick with diarrhoea, not associated with cattle disease.

Badger submitted by the public in 1991. Tested positive for M. bovis.

Roe buck. Shot 19/10/89. Near Loch Morar. Invernesshire. Lesions in the spleen, liver and lungs – not associatedwith cattle disease.


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 59


only five recent isolations ofMycobacterium bovis in Scottish wildlife– in four deer and a badger from widelydifferent locations (Figure 2). None ofthese cases had any association with a cattle herd breakdown. Officialsurveillance programmes have, however,been limited to the 48 badgers submittedby the public under the Road TrafficAccident Surveillance (1972-1991 where one was positive) and the routinesurveillance of deer at post-mortemexamination. Without more significantevidence of wildlife reservoirs in Scotlandit has not been considered a priority toconduct further surveillance more widely.In any case, there are limitations with the interpretation of data from surveyssuch as those reliant on road trafficaccident data.

It can be reasonably concluded thattraditional TB control policies still workquite efficiently in Scotland and other low incidence areas. The new confirmedannual incidence figures that have beenachieved in Scotland are probably theclosest that current policies can get toeradication – given the free tradingposition with high incidence TB areas in GB and Ireland.

Origins of infectionTable 1 shows the number of newconfirmed breakdowns where the StateVeterinary Service is satisfied that theorigin of a new Scottish breakdown hasbeen established. It is not possible to

definitely identify origins in all cases,however, where cattle have moved fromknown or previously unidentified infectedfarms and where spoligo/VNTR straintypes match, then the origin is consideredto have been established for tracingpurposes. The data shows that theproportion of origins identified variesfrom year to year, however, whereorigins are identified, almost all cases area result of ‘importing’ cattle from highincidence areas of England, Wales andIreland. These account for 48 out of the52 cases where an origin has been identified during this period.

Policy developmentIn 2003 the Scottish ExecutiveEnvironment and Rural AffairsDepartment (SEERAD) and the StateVeterinary Service in Scotland, began to assess the possibility of introducing a statutory requirement for pre and post-movement TB testing of cattle coming toScotland from high incidence TB areas.

Relative effectiveness of pre and post-movement testingA model for pre and post-movementtesting already existed for cattle comingto Scotland from Ireland, where cattle aremoving from a high to low incidence areaand are TB tested before and after themovement. Pre-export TB testing datafrom 1997-2002 was helpfully providedby veterinary colleagues in Belfast andDublin and an analysis was undertaken


Origin of Scottish Cases

2002 2003 2004 2005

Breakdowns 28 22 23 13

Breakdowns where 19 17 11 5Origin identified

Origin associated with 19 13 11 5movements from EW&I

Table 1Origins of Scottish cases

1


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 60

61

to determine the pre-movement testfailure rate. However, it was apparentthat pre export tests may have beencoded as ‘private tests’ or ‘part herd’ tests and it was difficult to establish afailure rate specifically for the populationof cattle tested prior to movement toScotland. However, from the availabledata, and acknowledging the constraintson the pre-export testing analysis, theexercise indicated an approximate pre-movementfailure rate (in terms of reactors) ofbetween one in1,000 animals (ROIdata 1999-2002)and one in 250animals (NI data for 2002).

A more specificpost-import analysis for Irish cattle (1992-2002) revealed that, at 60 day postimport testing, a further one in 847 weresubsequently disclosed as reactors inScotland. For Scotland this has meantrevealing between one and ten reactorsper year in an imported populationaveraging 3,380 cattle/per year. Thesefailure rates are therefore similar to pre-movement failure rates.

The previously accepted view was thatpre-movement testing would disclose the majority of reactors and that post-movement testing would disclose a much

smaller number of reactors that had been incubating disease at the first test.However, it appeared from the analysisabove that the numbers of reactorsdetected at pre and post-movementtests, when cattle went from high to low incidence areas, are not dissimilar.

Cattle movements from highincidence areas of England and Wales to ScotlandWork was undertaken with British CattleMovement Service (BCMS) and IBMcolleagues to quantify the number ofcattle in the ‘at risk’ population, comingto Scotland from high incidence areas ofEngland and Wales. Preliminary analysisof Cattle Tracing Scheme data suggeststhat between 15,000 and 20,000 cattlemake up the population that would beeligible for pre/post-movement testingeach year.

ConsultationIn February 2004, GB AgricultureDepartments jointly published a

consultationdocument‘Preparing for a New GB Strategy onBovine Tuberculosis’.The consultationemphasised theneed to reduce therisk of spreading TBfrom high to low

incidence areas and several options wereconsidered including:

(a) Zoning – banning all cattlemovements from high incidence areas;

(b) Pre-movement testing;(c) Post-movement testing;(d) Pre and post-movement testing.Veterinary advisers in Scotland

indicated that only a complete ban onmovements would eliminate the risk ofdisease spread from cattle movements.

kCONTINUED ON p62


3

Figure 3Herd of highland cows

Preliminary analysis of CTS datasuggests that between 15,000 and20,000 cattle make up the populationthat would be eligible for pre/postmovement testing each year.


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 61


However, the risk could be reduced,and pre and post-movement testing of all cattle moving from high incidence TB areas to low incidence areas (exceptcalves under 42 days and those goingdirect to slaughter) was likely to yieldthe greatest risk reduction. Furthermore,cattle should be permitted to movefreely for a period of 60 days after the pre-movement test. This wasdescribed as SEERAD’s preferred option in theconsultationdocument. A further advantageof employing preand post-movementTB testing is thatthe 80% sensitivityof the TB skin test (a generallyaccepted figure) can be increased to 95% by testing twice, i.e. alsoemploying post-movement testing.In other words, instead of identifying an average of 16 out of 20 previouslyundetected infected cattle, testing twice provides the potential to identify19 out of 20 previously undetectedinfected cattle before they enterthe Scottish herd.

StakeholdersA ‘pre-movement testing stakeholdergroup’ was set up on a GB basis towardsthe end of 2004 and chaired by BillMadders. In addition, a Scottish post-movement testing stakeholder group wasestablished in Edinburgh and includedrepresentatives of the National FarmersUnion of Scotland, The Institute ofAuctioneers and Appraisers Scotland,Local Authority representatives, NationalBeef Association and Scottish Associationof Meat Wholesalers. Stakeholders wereunanimously supportive of a pre andpost-movement testing policy andSEERAD were able to refine the proposedpolicy to address the remaining concernsthat stakeholders had. One such concernwas the need to ensure that Scottishfarmers could identify eligible cattle, i.e. those that should have been pre-movement tested before coming to Scotland and that would need post-movement testing after arrival.

The ‘TB Strategic Framework for the sustainable control of bovinetuberculosis in GB’In February 2005, the GB AgricultureDepartments published the ‘TB Strategic

Framework’document andrecognised that aregional approachwas required to‘slow down andstop the geographicspread of the disease’and to ‘keep cleanareas clean’.

The Tuberculosis (Scotland) Order 2005The new legislation came into force in Scotland on 23 September 2005. It requires Scottish cattle keepers to:

(a) Ensure that cattle from high-riskparishes are TB tested no more than 60 days before the date ofmovement from the holding of origin;


4

Figure 4Highland cow beingprepared for testing

A further advantage of employing preand post-movement TB testing is thatthe 80% sensitivity of the TB skin test(a generally accepted figure) can beincreased to 95% by testing twice.


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 62

63

(b) Undertake a post-movement test of eligible cattle, 60 to 120 days after arrival.

The only exceptions are for calves under42 days old and cattle moving direct toslaughter. TB tests are conducted at the

owner’s expense. Government providesthe tuberculin and administrativeassistance required to support the new policy.

kCONTINUED ON p64


Destination Destination Destination Section 1 Section 2 Section 3Country AHDO County Name

Scotland Ayr AHDO Ayr 4 5 0

Scotland Dumfries 37 162 5

Scotland Kirkcudbright 34 42 0

Scotland Renfrew 0 84 0

Scotland Wigtown 1 20 0

Sub-Total 76 313 5

Scotland Galashiels AHDO Berwick 70 117 2

Scotland East Lothian 0 1 0

Scotland Lanark 3 8 0

Scotland Mid Lothian 21 6 0

Scotland Roxburgh 45 39 18

Sub-Total 139 171 20

Scotland Inverness AHDO Nairn 40 0 0

Sub-Total 40 0 0

Scotland Inverurie AHDO Aberdeen 149 333 20

Scotland Banff 57 13 19

Scotland Kincardine 7 0 1

Scotland Moray 2 8 0

Sub-Total 215 354 40

Scotland Perth AHDO Angus 0 22 0

Scotland Argyll 0 6 0

Scotland Fife 2 8 0

Scotland Perth 50 0 0

Scotland Stirling 2 1 0

Sub-Total 54 37 0

Total 524 875 65

Table 2Eligible movementsto Scotland

2


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 63


Monthly Cattle Movement ReportBCMS and IBM colleagues havecollaborated closely to produce a newmonthly report of cattle movements toScotland from high incidence parishes in England and Wales. These monthlyreports now distinguish between cattlemoved directly from holdings in Englandand Wales to a holding in Scotland, aswell as movements from high incidenceparishes to Scotland via a market inEngland and Wales, or via a market inScotland. Table 2 is the summary sheetdescribing eligible movements toScotland in January 2006. Section 1describes the direct movements, Section2 describes movements via English andWelsh markets and Section 3 describesmovements via Scottish markets. Themovement report is used in two ways.

Firstly, a monthly ‘mailmerge’ exercisesends advisory letters to recipient farmersto confirm that they have acquiredeligible cattle that should have been pre-movement tested before arrival. The letter also provides advice aboutwhat to do if the cattle have not beentested and addresses the concern ofstakeholders that farmers might notknow the provenance of the cattle theyhad purchased.

Secondly, the report is sent to AnimalHealth Offices in Scotland. The StateVeterinary Service then conduct 10%checks on consignments to assesscompliance with pre-movement testingrequirements. The second workstreamwithin Animal Health Offices involvesentering cattle data on to the VETNET ITsystem to ‘mark forward’ post-movementTB tests for 60 to 120 days after thearrival date on the movement report.Where post-movement tests have notbeen conducted by 120 days (or wherepre-movement testing has not been done as required) whole herd movementrestrictions are imposed until the overdue testing has been completed with negative results.

Initial compliance From 23 September 2005 to 31 January2006, 1,410 cattle in 326 consignmentscame to Scotland from high incidenceparishes in England and Wales. 58consignments (31%) of 387 cattle werechecked by SVS staff and around 80%were fully compliant with pre-movementtesting requirements. Most of the otherconsignments had only minor non-compliances – for example, premovement testing outside of the 60-daywindow. At the time of writing, ‘overduepre-movement test codes’ have beenrecorded as being required in only 15consignments of cattle.

The compliance rates for pre and post-movement testing are, so far, very encouraging and the next milestonewill be the disclosure of the first reactor.The post-movement test failure rate forIrish cattle between 1997 and 2002suggested that, on average, one in 847cattle was subsequently disclosed as areactor in Scotland.

With 1,466 cattle having been post-movement tested (to 19 June 2006) wemight have expected to find reactors inthis population already – if test failurerates are similar for cattle from highincidence areas of England, Wales andIreland. It remains to be seen whether thisis the case. The overwhelming support forthese measures at grass-roots level inScotland is extremely encouraging andthe extent to which farmers and otherstakeholders in Scotland have adoptedthe new policy should make a significantcontribution to an important objective ofthe TB Strategic Framework, namely theobjective of keeping a ‘clean area clean’.


Dr Martyn Blissittis a VeterinaryAdvisor forNotifiableDiseases at theState VeterinaryService (SVS),Scotland.

l Information


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 64

65

“…the substance [tuberculin] will become an indispensable diagnosticmeasure in the future. It will enable us to diagnose questionable cases of early phthisis [tuberculosis] even when we fail to detect bacilli…”

Robert Koch (1891)

“These results and those of others show that the tuberculin test, like manyother things in this world, is not perfect… [However, it would be a] greatmistake to question this method because it does not do everything we want.”

B. Bang (1892)


Ante mortem diagnosis of BovineTuberculosis: the significance ofunconfirmed test reactors

Ricardo de laRua-Domenech,Tony Goodchild,

MartinVordermeier,

Richard Clifton-Hadley

A u t h o r s

Introduction‘Tuberculosis’ (TB) is a clinical orpathological diagnosis that, by convention,refers to the clinical signs (or lesions)arising from infections by organisms ofthe Mycobacterium tuberculosis (MTB)complex, a closely related group of bacteriathat includes M. bovis, the causative agentof bovine TB. ‘Skin’ tests measuring adelayed-type hypersensitivity response to the intradermal injection of purifiedmycobacterial proteins (tuberculins) areused throughout the world, in both humanand veterinary medicine, for the diagnosisof the pre-clinical stages of TB. In GreatBritain (GB), the majority of new TBbreakdowns (incidents) in cattle aredisclosed by routine screening or targeted testing of herds with the single intradermal comparative cervicaltuberculin (SICCT) test. This and othertypes of skin tests are the internationallyaccepted standard for the detection of M. bovis-infected cattle and are consideredthe best tests currently available fordiagnosing TB in live animals. Manydeveloped countries have eradicatedbovine TB through systematic applicationof the skin test alone, followed byslaughter of all test reactors.

A smaller percentage of TB incidents inGB (10% to 40%, depending on the localfrequency of skin testing) are initiated bysuspect lesions of TB found by meatinspectors during normal meat production(‘slaughterhouse cases’). Additionally, thegamma-interferon (γ− IFN) blood test(BOVIGAM®) has been used in GB since2002 to maximise the detection ofinfected animals in herds with acute orpersistent infection that cannot be clearedby short-interval skin testing alone.

Somewhat arbitrarily, a TB incident isnot considered ‘confirmed’ for statisticaland other purposes until at least one ofthe slaughtered animals presents itselfwith characteristic TB lesions on postmortem examination or, in the absenceof lesions, M. bovis can be cultured froma pool of selected lymph nodes. Accordingto Annex B of European Directive64/432/EEC (as amended), all skin testreactors are considered to be affectedwith TB and must be slaughtered. UnderEU legislation there is also a requirement toconduct post mortem and bacteriologicalexaminations on those animals. Thereason for this, however, is not so muchthe validation of skin test results, but

kCONTINUED ON p66

Ante mortem diagnosis of Bovine Tuberculosis: the significance of unconfirmed test reactors

10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 65


rather to determine (1), the severity of disease (and thus the infectiousness of reactors to other cattle and human in-contacts) and (2) the number andinterpretation of subsequent skin testsrequired to restore the official TB-freestatus of the affected herd. Moreover, the SVS uses the results of these post-mortem examinations to inform decisionson the reporting of bovine TB incidents to the medical public health authorities,and for the instigation of tracings andtesting of contiguous herds. Finally, tissuesamples from reactors are cultured toprovide sufficient material for DNAextraction and molecular typing(spoligotyping) of M. bovis isolates. Thisinformation is important for epidemiologicaland geographical monitoring of thespread of M. bovis in GB and to supportbreakdown investigations by the SVS.

Throughout GB, the overall proportionof individual test reactors that areconfirmed by post mortem and/orbacteriological examination is relativelylow: approximately 40% of all skin testreactors (Figure 1) and 18% of cattleslaughtered due to a positive γ− IFN test result. Similarly low confirmationpercentages have been observed for skintest reactors in the Republic of Ireland,Northern Ireland, New Zealand and other

countries. This may give the impression of a test accuracy far lower than thatreported in the scientific literature and a wasteful policy that leads to theunnecessary destruction of 60% of all cattle reacting to the skin test. The presence of these so-called ‘falsepositives’ undermines the confidence ofmany veterinarians and farmers in theskin and γ− IFN tests. In this paper we willattempt to explain that this is a simplisticview and there are several reasons for thefailure to ‘confirm’ the presence of TB in tuberculin reactors (or γ− IFN positiveanimals) by post mortem examinationand bacterial culture. We will also showthat the perception that too many falsepositive reactors are being slaughteredarises from two misunderstandings, namely(1), that post mortem examination andculture of the causative organism of TBare the ‘gold standards’ against which one should assess the validity of an antemortem test result, and, (2) a failure toappreciate the difference between thefalse positive rate (‘given that the animalis not infected, what is the probabilitythat the test will be positive?’) and thepredictive value of a positive test result(‘given that the animal has testedpositive, what is the probability that the animal is infected?’).


70%

60%

50%

40%

30%

20%

10%

0%

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Percentage of reactors confirmed per monthPercentage of reactors confirmed (12-month rolling mean)

1

Figure 1Overall monthly‘confirmation rate’for all tuberculin testreactors disclosed inthe period January1986 to December2005 (Courtesy ofRaquel Gopal, VLAWeybridge)


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 66

67

Epidemiological andImmunological principlesrelevant to TB screening tests

Predictive value of a positive testThe diagnostic accuracy of a screeningtest is primarily defined in terms of itssensitivity and specificity, which arecalculated respectively, from theproportions of infected and uninfectedanimals that are correctly diagnosed. Test sensitivity is the proportion ofdiseased (infected) animals detected as positive in the diagnostic assay.Diseased animals that test negative arethus false negatives. Test specificity is theproportion of disease-free (uninfected)animals that are correctly identified asnegative by a diagnostic test. Disease-free animals that test positive are then,by definition, false positives. Nobiological test for any disease is perfect(i.e. both 100% sensitive and 100%specific) and the tuberculin and γ− IFNtests are no exception. Furthermore, in allquantitative diagnostic tests a cut-offpoint for a positive result has to bedefined on the continuous scale ofpossible test results. As a consequence,there is an inverse relationship betweentest sensitivity and specificity and acompromise between the two must beselected. Field studies evaluating thesensitivity of the SICCT test in variouscountries suggest that its sensitivity liesbetween 52.0% and 100%, with medianvalues of 80.0% and 93.5% for standardand severe interpretations, respectively.Similarly, other studies conducted in TB-free cattle populations have found the specificity of the SICCT test to bebetween 88.8% and 100% (median of99.9%). Looking at the prevalence offield reactors in locations of very low TBprevalence (as judged by post mortemexamination), the specificity of the SICCTtest used in GB can be estimated at99.99%. This means that fewer than one

in 1,000 non-infected animals tested willbe wrongly classified as skin test reactors.Imperfect test specificity leads to falsepositives, also known within the contextof TB testing programmes as ‘non-specific reactors’ (NSRs). The causes ofcross-reactions to the bovine tuberculinused in the skin and γ− IFN tests arecomplex and fall outside the scope of this paper, but they are summarised inthe box below. The relevance of this toour discussion is that only a very smallproportion of all Non Visible Lesion (NVL),culture-negative reactors disclosed inGreat Britain should be attributed to non-specific (false positive) reactions.

Potential causes of false positiveresults to the tuberculin and otherdiagnostic tests for TB in cattle• Infection with M. tuberculosis

(human TB bacillus) or mycobacteriaof the M. tuberculosis complex otherthan M. bovis

• Infection with M. avium subsp. avium (avian TB)

• Infection with M. avium subsp.paratuberculosis (Johne’s disease)

• Immunisation of calves with Johne’sdisease vaccine

• Unidentified acid-fast bacilliassociated with the presence ofsubcutaneous granulomas commonlyknown as ‘skin TB’

• Experimental vaccination with M. bovis BCG strain

• Sensitisation to tuberculin byingestion of various non-pathogenicenvironmental mycobacteria found insoil, stagnant water and vegetation.Bryophytes (mosses), in particular,appear to be a rich source of theseorganisms

• ‘Atypical’ mycobacteria recoveredfrom mammals

• Non-mycobacterial organisms such as Nocardia spp.

kCONTINUED ON p68



10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 67

68EXOTIC DISEASES IN DOGS AND CATS – THE DACTARI SCHEME

A key point to remember is that theability of a diagnostic test to predictaccurately the true status of a positiveanimal is not constant. Rather, thisdepends on the sensitivity and specificityof the test, as well as the prevalence ofthe disease in the population tested. The relationship between these threeparameters is given by a probabilityformula, which defines whatepidemiologists call the ‘predictive value of a positive test result’.

For any test with less than 100%sensitivity and 100% specificity (i.e.virtually every diagnostic test), the higherthe disease prevalence, the test sensitivityand test specificity, the more likely it isthat a positive test result will be trulypredictive of the disease (or infection). To illustrate this point, let us consider twoherds. In herd A, the proportion of trulyinfected animals is 20%, while in herd B,the same proportion is 1%. The sametest, with sensitivity and specificity bothof 90%, would have a positive predictivevalue of 69% in herd A, but only 8% in

herd B. It is, therefore, misleading to usethe proportions of NVL reactors (or herds)to judge the merits of the skin and γ− IFNtests. This is because the herd or reactorNVL proportion will vary according to theprevalence of M. bovis infection in thescreened population, regardless of testaccuracy. It also follows, therefore, that in an area of endemic high incidence ofTB (e.g. annual and, to a degree, biennialtesting parishes) or in a herd breakdownin which Visible Lesion (VL) cattle havebeen previously disclosed, one canassume with a high degree of confidencethat the majority of cattle that fail theskin test (which is highly sensitive andextremely specific) are infected with M. bovis or have been exposed to thebacterium and developed an immuneresponse against it.

Gold standards for the diagnosis of TBCulture of M. bovis provides thedefinitive proof of infection. Disclosure oftypical tuberculous lesions in itself doesnot prove infection, as such lesions can


Increased bacterial load

Increasing pathology

Time

Gamma interferonTest (IFN−γ)

Tuberculin Skin Test

Antibody Response

Tuberculin Skin Test

Antibody Response

Gamma Interferon Test (IFN−γ)

Ane

rgy

Imm

une

Resp

onse

2

Figure 2Schematicrepresentation of the spectrum ofpathological andimmune responses incattle infected withMycobacterium bovis


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 68


occasionally be caused by othermycobacteria. However, in combinationwith a positive skin or gamma-interferontest result, the presence of TB lesions isconsidered sufficient to confirm infectionwith M. bovis. In other words, bacteriologyand post mortem examination of suspectanimals are extremely specific methodsfor the diagnosis of M. bovis infections(the infection is almost always present inthe animal if those tests yield positiveresults). The downside is that postmortem examination (particularly whenconducted in a commercial slaughterhouse)is a less sensitive technique than theimmunodiagnostic tests (e.g. skin and γ−IFN tests)conducted in liveanimals. Thesetests detect a cell-mediated immuneresponse of thehost againstinfection with M. bovis, which isoften subclinicaland may or maynot have progressed to the disease stage(TB) when the infected animal is identified.Depending on animal, test and (possibly)M. bovis strain characteristics, it usuallytakes between one and nine weeks forcattle to become reactors to the skin testfollowing infection with M. bovis. Mostanimals will be identified as skin testreactors from three to six weeks post-infection (this could be slightly less forthe γ−IFN test). This lag period betweeninfection and reactivity to tuberculin (pre-allergic phase) appears to be largelyindependent of the M. bovis infectivedose. By contrast, it may take severalmonths to reach the visible lesion (VL)stage in a majority of cattle infectedunder natural conditions. Due to thisdelay from infection to the developmentof detectable signs of bovine TB, cellularimmune responses against the invadingorganism will be evident at an earlier

stage than the pathological changescaused by the infection, and probably wellbefore bacterial loads are sufficientlylarge to be found by standard culturemethods (Figure 2). This is one of thereasons why the VL percentage inreactors taken in tests that disclose a newbreakdown is significantly greater than inreactors taken in short-interval tests.Furthermore, M. bovis is an intracellularpathogen with a slow generation rateand is not particularly easy to culture inthe laboratory. Culture of the organismfrom a pooled sample of NVL lymphnodes becomes even more complicatedthan isolation from a VL animal, which is

generally attemptedfrom the lesion itself.Even when VLs arepresent, these do not always contain a sufficient number of viable bacteria to grow in vitro. For these reasons, the culture of M. bovis in the

laboratory and the examinations forTB lesions at slaughter should not beconsidered the gold standards for theinfection status of cattle. At post mortemexaminations, a proportion of these testpositive animals (reactors) will not haveVLs, nor will M. bovis be subsequentlycultured from tissue samples. To implythat such animals were ‘false positives’ is inappropriate, as they may have been at an early stage of disease when thebacillus could not yet be detected. Such animals could have been harbouringthe bacilli, and when the disease hadprogressed further, became infectious toother cattle. From a TB control perspective,the early removal of skin test reactors andgamma-interferon test positive animals is a desirable outcome of testing,particularly in areas of endemic TB.

The diagnostic accuracy of ascreening test is primarily defined interms of its sensitivity and specificity,which are calculated respectively,from the proportions of infected and uninfected animals that arecorrectly diagnosed.

kCONTINUED ON p70


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 69

The Empirical EvidenceAssuming 99.99% specificity for thesingle intradermal comparative tuberculintest, we have estimated that theprobability that any animal gives a false positive test result tends towards0.00016 (i.e. 0.016% or 1.6 in 10,000animals tested) in counties with lowprevalence of TB. For herds in which 50cattle were tested, one in 200 herdswould have one false positive reactor. For a herd in which 200 cattle weretested, approximately one in 50 herdswould have one false positive reactor. In other words, the probability of a falsepositive result increases with herd size,

although it is extremely rare to find more than one false positive in a herd(Figure 3). However, in areas of GreatBritain with a high incidence ofconfirmed TB breakdowns, there are far more ‘unconfirmed’ breakdowns than could be accounted for simply bythe estimated rate of false tuberculinreactors. Based on data from the period1996-1999 only one in between two

and five unconfirmed incidents in thoseareas were due to non-specific reactors. The excess was the fraction ofunconfirmed TB breakdowns attributableto M. bovis infections.

In areas where TB is endemic,unconfirmed breakdowns tend to cluster around foci of confirmed TB. The historical TB data for Great Britainshow that, following an unconfirmed TB breakdown, the risk that a confirmedbreakdown will occur in the next 12months is five to ten times greater thanfor herds in the same county without ahistory of breakdowns (data for 1996-1999). In other words, disclosure of

unconfirmed reactors is often the firstindication of a developing TB problem ina herd, which will manifest itself in thepresence of VL animals in future herdtests. In relation to the γ− IFN test, studiesin Northern Ireland have demonstratedthat gamma-IFN test positive, skin testnegative cattle that are not immediatelyslaughtered are far more likely to react to subsequent skin tests.


12%

10%

8%

6%

4%

2%

0%

0 200 400 600 800 1,000 1,200

Herd size

Prop

ortio

n of

her

ds g

ettin

g fa

lse +

ve re

acto

r(s)

at a

rout

ine

test

Where specificity = 99.99%

Probability of finding two ormore false positive reactor

Probability of finding one ormore false positive reactors

3

Figure 3The percentage ofherds with falsepositives (skin testreactors) increaseswith herd size. It is,however, very rare to find more thanone false positive in a herd


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 70

71

ConclusionIn summary, saying that a TB breakdownor a reactor is ‘not confirmed’ meansdifferent things in different circumstances,depending on local TB prevalence,thoroughness of post-mortem examination,number of reactors detected and herdsize. The chronic and latent nature of TBmeans that NVL and culture-negativereactors in herds with confirmed TB orherds in endemic TB areas should not beautomatically regarded as false positives.Such animals may well have been exposedto M. bovis and infected even if thepresence of the disease cannot beestablished. This is because the ‘goldstandard’ tests (post mortem examinationand culture) are significantly less sensitivethan the ante mortem diagnostic tests forTB (the skin test and gamma-interferontests) that measure indicators of infectionrather than disease. Animals that arepositive to a skin or gamma-interferontest, but NVL and culture negative can,therefore, be indicative of:

• non-specific cross reactions to theantigens used in the ante mortemtests (true false positives);

• early detection of M. bovis infection,when tuberculous granulomas are still too small to be detected byroutine post mortem examination and there are very low numbers of bacilli in the lymph node poolcollected for culture in the absence of lesions;

• infection with lesions in a locationthat is not normally checked atroutine post mortem examination;

• animals with arrested infection, i.e.temporarily able to contain thebacterium in a condition of latency, a phenomenon well characterised inhuman TB.

In regions where the prevalence of TB islow or the disease is believed to be absent,the predictive value of a positive skin testresult will be less than in the endemic TBareas. In other words, NVL, culture-negativereactors in low-TB prevalence regions aremore likely to arise from non-specificsensitisation to tuberculin. For that reason,where NVL, culture-negative reactors arepersistently found in herds in 4-, 3- or two−yearly testing areas and non-specificsensitization is suspected, it is possible touse the γ− IFN test for sequential testingof those animals prior to slaughter. The γ− IFN test results, the evidence from adetailed necropsy, laboratory culture, theherd’s testing history and the TB historyof the area, are all taken into accountwhen assessing the likely infection statusof those animals and before considering if such herds should have their TBrestrictions lifted under the so-called ‘nonspecific reactor’ procedures (described inVIPER 23). For the reasons explainedabove, such protocols are never appliedon reactors in annually tested herds orherds in which infection with M. bovishas already been confirmed.

In the end, from a disease controlperspective, test performance comes down,not to arguments about specificity andgold standards, but whether a given testis useful in the field. This was the pragmaticapproach that underpinned the γ− IFN fieldtrial of 2002-2005 and the SVS instructionsfor ad hoc application of the γ− IFN test.Namely, that the γ− IFN test is most likelyto be useful to the SVS as a supplementto the skin test in identifying early infectedanimals in severe or chronic TB breakdowns,where removal of small numbers of falsepositives is only a minor concern.


Ricardo de laRua-Domenech is a VeterinaryAdvisor,TB Division,Defra.

Tony Goodchild is a bovine TBepidemiologist,VeterinaryLaboratoriesAgency Centre for Epidemiologyand Risk Analysis,Weybridge.

MartinVordermeier isLeader of theImmunologyGroup, TBResearch Unit,VeterinaryLaboratoriesAgency,Weybridge.

Richard Clifton-Hadley is Head of Statutory andExotic BacteriaUnit, VeterinaryLaboratoriesAgency,Weybridge.

l Information


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 71

AbstractThe BOVIGAM® IFN-γ test constitutes anadditional ante-mortem test to identifycattle infected with Mycobacteriumbovis, the causative agent of bovinetuberculosis. We will explain why we think that at present its mostadvantageous use will be as an ancillarytest to the tuberculin skin test with a view to maximising the numbers ofinfected animals detected. This paper willalso describe the principle of the test andinterpretation of the results in the contextof the patho-biology of tuberculousinfections. Furthermore, the results ofstudies carried out to determine thesensitivity and specificity of this test willbe addressed, including the presentationof unpublished results from recent GBtrials. In addition, we will discuss thepotential use of defined antigens todifferentiate M. bovis infected cattle from those vaccinated with futurevaccines currently under development.

1. Infection biology of bovinetuberculosis in cattle in relation to diagnostic testsIt is generally accepted that cell-mediatedimmune (CMI) responses are the principaland earliest immune responses to developafter infection with Mycobacterium bovis,the causative agent of bovine tuberculosis(Figure 1, (1-3)). Antibody responsesdevelop considerably later than CMIresponses (Figure 1) and it is thereforenot surprising that assays based onmeasuring CMI, like the tuberculin skintest and the BOVIGAM® IFN−γ assay, havebeen used extensively for the diagnosis of bovine tuberculosis (BTB) in cattle.Currently there is no recognisedserelogical test available for TB diagnosis.

To better understand the underlyingprinciples of the tuberculin skin and theBOVIGAM® IFN-γ tests, it is important tounderstand the kinetics of the immuneresponse. BTB in cattle is presentedprimarily as visible lesions in the lungsand associated lymph nodes and/or in the lymph nodes of the head. M. boviscan also be cultured from tissues with or without visible lesions. As M. bovis, isclosely related to the pathogen causinghuman tuberculosis, (M. tuberculosis) ithas been proposed that the principles ofM. bovis pathogenesis in cattle followsclosely that of M. tuberculosis in humans,which is described below.

Humans in contact with TB patients willbe exposed to M. tuberculosis (as will becattle in a herd with confirmed bovineTB). A certain proportion (10-30% in thecase of humans) of exposed individualswill actually become infected (as defined by an acquired delayed typehypersensitivity and/or IFN−γ response).About 5% of infected humans developdisease within one year of infection(primary tuberculosis), whilst 95% ofinfected individuals do not. Suchindividuals – i.e. those that are tuberculinskin test positive or show IFN−γ responsesin vitro to antigens such as ESAT-6 but donot present with clinical or radiologicalsigns of disease – are classified as latentlyinfected. However, 5-10% of latentlyinfected humans develop clinicaltuberculosis during their lifetime through re-activation of the latentinfection (re-activation tuberculosis).Thus, the population of latently infectedhumans presents a huge reservoir of M. tuberculosis. Although exposed,latent and diseased states are likely to beequally applicable to bovine tuberculosis


The BOVIGAM® assay as ancillary test to the Tuberculin skin test

MartinVordermeier,Adam Whelan,Katie Ewer, TonyGoodchild,Richard Clifton-Hadley, JohnWilliams, R. GlynHewinsonVeterinaryLaboratoriesAgency (VLA) –Weybridge, TBResearch Group

A u t h o r s


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 72


in cattle (Figure 2), the proportions thatthese disease states occur in cattle arenot known. It is likely that the proportionof latent infection will be considerablylower in bovine tuberculosis than inhuman tuberculosis. However, theargument that latently infectedindividuals (culture-negative NVL, skintest reactors forexample) constitutea continuous andunpredictablesource of re-infection, is equallyvalid for cattle as it is for human TB.

The concept ofinfection without signs of disease(latency) as defined in the previousparagraph, as opposed to detectabledisease, has important implications forthe interpretation of tests like the IFN−γtest or the skin test (see also Figures 1and 2). At the early stages of infection, or in latently infected cattle, a period will occur when M. bovis appears to beabsent because the bacillary load is not

large enough to be detected by culture. In addition, the pathological changescaused by the bacilli are not yet profoundenough to be detected during routineabattoir inspection. However, cellularimmune responses will be detectable in these animals at an earlier stage of

infection than thepathologicalchanges caused bythe disease (e.g.visible lesions), orbefore the bacterialloads exceed thenumbers necessaryto be able to cultureM. bovis from tissuesamples (Figure 1).

This has two consequences: firstly,immuno-diagnostic assays of cellularimmunity (which includes the IFN−γtest as well as the tuberculin skin test),because they detect infection rather than disease, can be very sensitive atidentifying M. bovis infected cattle.Secondly, at abattoir inspections, aproportion of these identified animals willnot have visible lesions, nor can M. bovis

Anergy

Increasing bacillary load

Pathology

Cell-Mediated Immunity

Antibody Response

Imm

une

resp

onse

Cell-Mediated Immunity

Antibody Response

Figure 1Development ofimmune responses incattle following M. bovis infection:Bovine tuberculosisas spectral disease

kCONTINUED ON p741

It is generally accepted that cell-mediated immune (CMI) responsesare the principal and earliest immuneresponses to develop after infectionwith Mycobacterium bovis, thecausative agent of bovine tuberculosis.


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 73

be subsequently cultured from tissuesamples. To designate these animals apriori as ‘false-positive’ is inappropriate as they can harbour bacilli, and becomeinfectious to other cattle, specificallywhen the disease has progressed further.

2. Tuberculin skin testsThe most commonly applied diagnostictests for BTB in cattle are tuberculin skintests. Two types of tuberculin skin testsare in use, namely, the single intradermaltest (SIT), and the single intradermalcomparative cervical tuberculin test(SICCT). The SIT, using bovine tuberculinpurified protein derivative (PPD) alone,can be carried out as the caudal fold test,(CFT) (which is used in countries such asUSA, Australia, New Zealand) or as aninjection into the skin of the neck (cervicalSIT). The SICCT, used in the UK and Ireland,compares responses to avian and bovinetuberculin PPD. The following sensitivityand specificity ranges have been cited in recent reviews. SIT: sensitivity, 68-96.8% (CFT), 80-91% (cervical SIT); and specificity, 96-98.8 (CFT), 75.5-96.8(cervical SIT). For the SICCT the following

values have been reported for sensitivityand specificity, respectively (standardinterpretation): 55.1-93.5% andapproximately 88.8-100%.

Importantly, BCG vaccination will resultin positive tuberculin responses to theskin test (and the BOVIGAM® assay, seebelow) in a large proportion of vaccinatedanimals, thus compromising the use oftuberculin as a diagnostic antigen inconjunction with vaccination. As theprinciple of the tuberculin skin test hasbeen amply described previously, we will


CMI diagnosis(TT or IFN–γ)

Negative

Positive

Positive(or negative, e.g.anergic cattle)

Visible lesions/M. bovis +ve

Negative

Negative orPositive

Positive

No Infection, OR:Infection, immediateclearance

Infection, diseasecontrolled (latency)

Infection, progressivedisease

Figure 2Biology of bovinetuberculosis and the consequences on the outcome ofdiagnostic tests.Please note that thedegree of latency incattle is unknown

2

M. bovisexposure

Diseaseprogression

2


10493 AW Defra GVJ_53-90 29/8/06 16:02 Page 74


not expand on it further in this review butwill concentrate on discussing the principleof the IFN−γ test (BOVIGAM® assay).

3. BOVIGAM® IFN−γ blood testThe BOVIGAM® IFN−γ test is an in vitrolaboratory-based immuno-diagnostic test which detects gamma interferon(IFN−γ) that is produced after thestimulation of blood cells with antigenssuch as bovine tuberculin. The test isperformed in two stages; firstly, bloodsamples are transported to the laboratory

within 24 hours of sampling, andcultured at 37°C in the presence ofbovine or avian tuberculin.

Practical advantages of the testinclude: (i) flexibility in its interpretation by

setting appropriate cut-offs;(ii) only one farm visit is required;(iii) the test does not interfere with the

host’s immune status, so thatassays can be repeated morefrequently than tuberculin skin tests;

Country

Australia

Australia

USA

Brazil

Spain

Spain

Northern Ireland

New Zealand

New Zealand

Italy

Italy

Italy

Ethiopia

Romania

Eire

Eire

GB

GB

Median

Range

IFN−γ Sensitivity (%)

84.4

81.8

80.9

100

84.9

87.6

89.3

85

94.6

ND

96.6

NT

95.5

92.5

84.2

88.5

88.2

89.7

88.4

80.9-100

Specificity (%)

98

99.1

ND

ND

ND

ND

99.2

93

(73.6a)

88.4

98

97.3-98.6

87.7

ND

96.2

NT

92

96.6

96.6

87.7-99.2

1

Table 1Previous studies of sensitivity andspecificity of theBOVIGAM® assay

A number of different cut-off values and interpretation protocols were applied in the different trials. However, results in all studiesare based on the comparison of responses to avian and bovine PPD. A study concentrated on false-positive skin test positiveanimals, i.e. specificity of skin test is 0% in this cohort, data not used to determine median specificity. ND, not determined.

kCONTINUED ON p76


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 75

(iv) it is highly amenable to theinclusion of defined antigens thatallow the differentiation betweeninfection and vaccination(differential diagnosis, see below).

After 24 hours incubation, plasmasupernatants are collected, and theamount of IFN−γ produced quantified byenzyme-immuno-assay (EIA) using thecommercially available BOVIGAM® kit. A test result is interpreted as positive forbovine tuberculosis when the blood cellsfrom an infected cow produce more IFN−γ after stimulation with bovinetuberculin than after stimulation withavian tuberculin. It is also accepted thatthe BOVIGAM® test detects animals thatescape skin testing, probably because itdetects animals earlier after infectionthan tuberculin skin testing.

The BOVIGAM® test is generallyconsidered to be more sensitive thantuberculin skin testing. Whilst specificitieshave been reported that, although notwidely dissimilar to those seen withtuberculin skin testing, were lower inmost studies than those reported for theSICCT. However, direct comparisons ofthe two tests in the same study haverarely been made.

A number of trials assessing theBOVIGAM® assay have been conductedsince 1991, and Table 1 summarises 18 of these trials conducted in 11 countrieswith respect to sensitivity and specificity.The assay was applied in these trials in its basic form using bovine and aviantuberculin. However, a range of cut-offswere used, and in some studies cattlewere also tested before skin testapplication, whereas in others they


Study

Wood, 1991

Whipple, 1995

Gonzalez, 1999

Collins, 2002

Goodchild, 2004

IFN−γ + (%)

94

73

85

88

70b

ST+ (%)

66

80

80

74a

65b

ST+/IFN−γ +(%)

95

91

93

93a

88b

Goodchild, 2006

IFN−γ + ST-

142

ST+d

387

IFN−γ + and ST+

529

B. Maximising detection of infected animals (project SB4008)c

Numbers of VL and /or Cu+ cattle detected


2

Table 2Increase in overallsensitivity by usingSICCT andBOVIGAM® test in parallel

a) Severe interpretation of skin test. b) Whole herd removal operation.c) Data based on results from 78 herds. d) Skin test reactorswere not tested with the BOVIGAM® IFN−γ assay. NA, not applicable. ST, Skin test, IFN−γ , BOVIGAM® test.


A. Sensitivity increase

10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 76


were tested after skin test application.Interestingly, despite the application ofdifferent cut-offs across different studypopulations, the results are often verysimilar, which also highlights the easewith which it is possible to adjustBOVIGAM® cut-offs to local conditionswithout overtly affecting testperformance. As Table 1 indicates,median sensitivity and specificity levelsacross these trials were 88.4% and96.6%, respectively. During 2005, aDefra-funded study was conducted todetermine the specificity of the test in GBwith tight confidence intervals by using alarger number of animals from TB freeherds than had been tested previously in GB. Results based on 874 cattledetermined test specificity to be 96.7%(95% CI: 95.6-97.8, unpublished results).This underscores the relatively highspecificity of this test and is in line withprevious published results. Furthermore,it confirms the specificity valuesdetermined previously (96.6%, Table 1).

Although themajority of infectedanimals will bepositive in bothtuberculin skin testand BOVIGAM®

assay, thepopulations ofanimals reacting toeither of the assaysare not fullyoverlapping and aproportion of animals will be IFN−γ-positive/skin test negative, IFN−γ-negative/skin test positive or negative inboth tests (Figure 3). Therefore, the mostbeneficial application of this test isalongside the tuberculin skin test as anancillary test in herds with high diseaseprevalence or with persistent infectionthat cannot be cleared with thetuberculin skin test (also called paralleltesting). This increases the overalldiagnostic sensitivity and will thus have a major impact on disease control.

The advantage of applying the two teststogether has also been acknowledged bythe European Union who allow the use of the test to ‘maximise the detection ofinfected animals’. Table 2 presentspublished and unpublished data

illustrating this point. The recentlycompleted ‘NationalIFN−γ Trial’ (projectSB4008) hasunderscored the factthat the BOVIGAM®

test will identify asignificant portion of skin test-negativeanimals with visiblelesions (VL) and/or

which are culture positive for M. bovis. In the national trial, 27% more of suchanimals were identified by the IFN−γ testthan had been found at the disclosingskin test (Table 2).

What would be the possibleconsequence of leaving IFN−γ+/ST-animals in affected herds? It was recentlyreported that animals that were IFN−γpositive/skin test negative animals wereseven to nine times more likely to become

3

ST+/IFN-

ST-/IFN-

ST+/IFN+

The BOVIGAM® test is generallyconsidered to be more sensitive thattuberculin skin testing, whilstspecificities have been reported that,although not widely dissimilar tothose seen with tuberculin skintesting, were lower in most studiesthan those reported for the SICCT.

ST-/IFN+

Figure 3Outcome of paralleltesting usingtuberculin skintesting and IFN−γassay in combinationPie chart representsVL and/or M. bovisculture positiveanimals. ST,tuberculin skin test;IFN, BOVIGAM®

IFN−γ test results

kCONTINUED ON p78


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 77

subsequently skin test positive than IFN−γnegative animals. The authors of this studytherefore concluded that not removingIFN−γ+ animals could leave a significantsource of re-infection in these herds.

4. Influence of blood storage onBOVIGAM® test performanceIn practical terms this can be condensedto the question as to whether bloodshould be tested on the day of sampling(within eight hours), or could be leftovernight (within 24 hours) without lossof sensitivity. Studies in New Zealand andthe USA have concluded that about equalproportions of animals could be definedas positive when the BOVIGAM® test wasperformed on the day of sampling orwithin 24 hours of sampling, despite the observation that the OD values afterovernight storage of the blood sampleswere substantially lower than when theblood culture was started on the day ofsampling. These results formed the basisof New Zealand and USA policy ofstarting the assay on the day after thesample was taken but within 24 hours.

Our own results confirmed thesefindings. Despite decreases in signalstrength (OD450), we did not observe a reduction in the numbers ofexperimentally infected cattle detected.In field reactors we detected a modestdrop in relative sensitivity from 96% toabout 90%, when we assessed SICCT

(comparative tuberculin skin test) fieldreactors with confirmed TB (Figure 4)using a cut-off value (PPD-B minus PPD-A > 0.1 OD450). More dramaticallyhowever, we observed an increase inspecificity when the blood was keptovernight rather than cultured freshlyfrom about 85-97% (Figure 4). Anincrease in specificity following overnightstorage has also been observed by others.

In contrast to these findings thatsuggest that overnight storage of bloodhas no significant impact on sensitivity, a recent paper suggested that the delay inblood culture reduced test sensitivitysignificantly. Applying the BOVIGAM® testto visibly lesioned animals within eight or24 hours did not decrease sensitivitysignificantly in line with what the otherstudies described above had shown.However, they observed a decrease in thepercentage of animals testing positive inthe BOVIGAM® test after 24 hour storagecompared to eight hour for SICCT-negative animals. However, it isimpossible to assess the disease status ofthis skin-test negative group as no post-mortems were conducted. It is thereforenot possible to assess if this decrease inthe number of animals testingBOVIGAM®-positive is caused by adecrease in sensitivity due to overnightstorage, or indeed an increase inspecificity due to the detection of lessfalse-positive animals.


Cut-off: 0.1

100%

95%

90%

85%

80%

75%Sensitivity Specificity

8 h

24 h

Figure 4 Overnight storageimproved specificityof BOVIGAM® assay.Antigens: PPD-B,PPD-A responses,cut-off: OD450 PPD-B minus PPD-A > 0.1

4


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 78


5. Blood sampling in relation to skintest applicationEarly data suggested that IFN−γ responseswere affected by the application oftuberculin skin tests. In these studies,increases in the OD450 values after invitro stimulation with avian and bovinePPD were observed between seven and59 days post-skin test. However, the testapplied in these studies was the caudalfold test (CFT).Since thennumerous studieshave looked at thedevelopment ofIFN−γ responsespost-skin test, andreached conflictingconclusions.However, by comparing these studies, a general conclusion can be drawn thatCFT can boost IFN−γ responses whilst the application of the SICCT does not.Furthermore, our data, as well as that ofothers, demonstrated that blood could be taken three days post-SICCT applicationwithout affecting test sensitivity.

6. Specific antigens and differentialdiagnosis of vaccinated and infected animalsOne of the main advantages of theBOVIGAM® assay is the ease with which defined antigens can be added.The defined antigens used commonly are antigens like ESAT-6, CFP-10, whosegenes are absent from the genome of thevaccine strain M. bovis BCG or proteins

that are not highlyexpressed by somestrains of BCG likeMPB70. Suchdefined antigensmay also increasetest specificitycompared to the useof avian and bovine

tuberculin in animals sensitised byenvironmental mycobacteria. This is in contrast to the use of defined proteins for skin testing in cattle, where positiveresponses could only be elicited byapplying large protein doses, or when the antigens were applied with immuno-modulating agents like lipo-peptides.

3

2

1

0

Med ESAT-6 Med ESAT-6PPD-B CFP-10 PPD-B CFP-10

BCG M. bovis

Although the majority of infectedanimals will be positive in bothtuberculin skin test and BOVIGAM®

assay, the populations of animals are not fully overlapping.

5

Figure 5Differential diagnosisof M. bovis infectedfrom BCG vaccinatedcattle by the use ofESAT-6 and CFP-10in the BOVIGAM®

IFN−γ test.Heparinized bloodwas collected fromBCG vaccinatedcattle (n=5) and fromcattle experimentallyinfected with M. bovis (n=5) and incubated withbovine tuberculin(PPD-B), orrecombinant ESAT-6or CFP-10 protein.IFN−γ production wasdetermined by ELISAand data is expressedas mean OD450values + SE

OD

450

kCONTINUED ON p80


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 79

Interestingly, in the recent GB specificitystudy, the use of a peptide cocktail ofESAT-6 and CFP-10 increased thespecificity compared to PPD only marginally(from 96.7-97%, unpublished data),whereas the combined use of both PPDand the peptide cocktail (i.e. only animalsgiving a positive result to both antigenpreparations were deemed test-positive)lead to a specificity of around 99%(unpublished data). However, these latterresults should be viewed with caution asthis approach would inevitably also resultin reduced sensitivity.

The selection of antigens absent from the BCG genome highlights theinterest in antigens applicable for thedifferentiation of (BCG) vaccinated fromM. bovis infected animals (differentialdiagnosis), which may constitute themajor application for such definedantigens. BCG vaccination will result inpositive tuberculin responses both in theskin test and the BOVIGAM® assay, thuscompromising the use of tuberculin as a diagnostic antigen for vaccinatedanimals. However, several studies haveshown that by using ESAT-6 and CFP-10,which are absent in BCG, as diagnosticantigens in the BOVIGAM® assay, M. bovis infected animals could bedistinguished from BCG vaccinated cattle(Figure 5). For example, in one studyapproximately 70% of BCG vaccinatedanimals tested positive after stimulationwith tuberculin, whereas none respondedto stimulation with a cocktail of ESAT-6and CFP-10. However, the test sensitivitycompared to tuberculin was still lower,and studies are currently underway toidentify further antigens that, when usedin combination with ESAT-6 and CFP-10,will increase test sensitivity to levels akinto tuberculin. This endeavour has beengreatly facilitated by the elucidation of the genome sequences of M. tuberculosis, M. bovis, and BCG. This has allowed the application of insilico methods of comparative genomicsto predict ‘specific antigens’, which can

then be tested in cattle, either in the formof synthetic peptides, or recombinantantigens, for their suitability fordifferential diagnosis. Several antigenshave been identified in this way that areable to increase test sensitivity when usedin combination with ESAT-6 and CFP-10.This approach of identifying specificantigens is an on-going and successfulresearch effort and steady progress isbeing made to close the sensitivity gapbetween these specific antigens andtuberculin. However, this gap, whilstnarrowing, still exists, and further specificantigens will have to be identified toclose it completely.

7. Concluding remarksBoth IFN−γ and tuberculin skin tests are based on the detection of cellularimmune responses and therefore detectinfection rather than disease. A corollaryof this statement is that animals will beidentified by these tests that are in earlyor latent disease states where it may beimpossible to detect visible pathology orto culture M. bovis from tissues. Suchanimals should not be dismissed as false-positives; by allowing the disease toprogress they could become criticalsources of re-infection. As the populationsof animals detected by the BOVIGAM®

test and tuberculin skin test do notcompletely overlap, their combinedapplication in parallel, with theBOVIGAM® test used as a strategicancillary test, will maximise the numbersof infected cattle that can be detected.The BOVIGAM® test is a flexible testformat that can readily be tailor-made toparticular applications by refining cut-offsfor positivity or by the inclusion of more defined antigens than tuberculin.The most promising application of suchdefined and specific antigens could be in the differentiation of vaccinated frominfected cattle, and genome-basedapproaches have been successful atidentifying additional specific antigens.


l Information

MartinVordermeier isthe ImmunologyTeam Leader at the VLA.

Tony Goodchildis a bovine TBepidemiologistat the VLA.

Adam Whelanis a researchscientistspecialising in bovinetuberculosis.

Katie Ewer is animmunologiststudying bovineimmunity totuberculosis at the VLA.

Tony Goodchildis a bovine TBepidemiologistat the VLACentre forEpidemiologyand RiskAnalysis.

Richard Clifton-Hadley is Headof Statutoryand ExoticBacteria Unit atthe VeterinaryLaboratoryAgency.

John Williamsis Deputy WorkGroup leaderwithin theLaboratoryTesting Unit atthe VeterinaryLaboratoriesAgency.

Professor GlynHewinson isHead of the TBResearch Groupwithin theStatutory andExotic BacteriaUnit at theVeterinaryLaboratoriesAgency.


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 80


Tuberculous pneumonia and BVD in housed calves

Veterinary Laboratories Agency (VLA) – TruroA three-month-old calf showing signs ofpneumonia died within 48 hours of thefirst clinical signs becoming apparentdespite veterinary treatment. The calf wassubmitted to VLA Truro for post mortemexamination which revealed extensivelesions of tuberculous pneumonia fromwhich Mycobacterium bovis (M. bovis)was subsequently isolated. This casereport describes the clinical, post mortemand laboratory findings in a group of 22‘in-contact’ calves.

HistoryThe farm carried 100 Holstein milkingcows, together with 150 dairy and beeffollowers. Cows and youngstock werefed on silage, straw and bought inconcentrates alone. Calves were fed onhay, straw and bought in concentratetogether with raw milk, in buckets,including mastitic milk withheld from the bulk tank.

In a five month period betweenNovember and the following March eightyoung cows had aborted two to three

months before full term. In June and Julytwo calves had been born apparentlyblind and one of these was ataxic. The area surrounding the farm had ahistory of TB in cattle and badgers but thisfarm had no experience of TB during thelast 20 years. In July of the same year,however, the eventual slaughter of aninconclusive reactor milking cow revealedtuberculous lesions in the retropharyngeallymph nodes from which M. bovis wasisolated. A herd TB test followed on 24 August and this identified 11 reactors,eight of which showed visible lesions ofTB. Two of these were adults in which thelesions were confined to the lymphaticglands of the head and chest. The othersix were calves, between six and twomonths of age, and in four of these thetuberculous lesions were again confinedto the lymphatic glands of the head andchest. However, two also showed miliarycaseous lesions in the apical lobes of thelungs. M. bovis was isolated from thetuberculous lesions of the two cows andfive of the six calves. M. bovis was alsoisolated from the lungs of one calf whichwas considered to have shown lesions

2

R J Monies,BVMS, MRCVS,

VeterinaryInvestigation

Officer

A u t h o r

Figure 1Lung of calf no. 400– granulomatousmycobacterialpneumonia

Figure 2Bronchial lymphnodes of calf no. 400 caseouslymphadenitis

1

kCONTINUED ON p82


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 81

typical of acute pneumonia rather than tuberculosis.

Three weeks after the herd TB test, a three-month-old calf developedpneumonia and died three days later despite veterinary treatment.Unfortunately, post-mortem examinationwas not carried out and the carcase wasremoved to the local hunt kennels. The following week a three month oldcalf (No. 400) developed hyperpnoea,dyspnoea and a temperature of 103˚F. It showed amucopurulent nasaldischarge, developeda cough and again,despite veterinarytreatment, it died 24hours later. The calfwas submitted to VLA Truro for post-mortem examination on the 21 September, and this revealed caseous foci in sub-mandibular, retropharyngeal,bronchial and mediastinal lymph nodes as well as extensive pneumonia. Acidfastbacilli were present in smears preparedfrom the lesions and M. bovis wassubsequently isolated from these.Histological examination of the lungsshowed a granulomatous and necrotisingmycobacterial pneumonia andlymphadenitis. Because of the apparentrapid onset of disease, the age of thecalves and the severity of the changesseen, a visit was made to examine theremainder of the calves on the farm.

Clinical Examination of the In-Contact CalvesTwenty two calves were present on thefarm and these were kept in two houses,both of which were open at one end tothe cows covered yard and separatedonly by a gate. The calves, therefore,shared air space and had nose to nosecontact with the cows. Fourteen, one-three month-old calves were kept in onehouse and eight, three-four month-oldcalves were in the other. There had been

frequent movements of calves betweenthe houses in both directions.

A number of the calves showedrespiratory signs including hyperpnoea,coughing and nasal discharge. In additionfive calves showed either developmentalabnormalities or central nervous signswhich were suggestive of in-utero BovineViral Diarrhoea Virus infection (BVDV)(Animal Numbers 396, 397, 398, 411and 420). Two had short tails, one calfhad bilateral cataracts and two calves

showed ataxia and circling.

The AnimalHealth Officedecided toslaughter all 22 calves as TB

contacts. Heparin and plain bloodsamples were collected from all the calves prior to their slaughter on the 30 September.

Laboratory InvestigationsPostmortem examinations of all 22 calveswere carried out.

Formalised tissue samples of brainsfrom the five calves showing clinical signssuggestive of in utero BVD virus infectionwere collected. Sections of these tissueswere examined histologically and immunestaining for pestivirus antigen was carriedout on sections from all five calves.

Samples of affected lungs from allcalves showing lung lesions were takenaseptically and innoculated onto 5%sheep blood agar and MacConkey agar.In addition, the samples were innoculatedonto modified Middlebrook 7H11 agarsemi-solid slopes and any suspect colonieswere innoculated onto specialisedLowenstein Jensen and Stonebrinksmedia at different temperatures(Gallagher and Horwill, 1977).

Serum samples were examined by thecomplement fixation test for antibody toHaemophilus somnus and by the ELISAtest for antibody to Mycoplasma bovis.


Three weeks after the herd TB test a three-month old calf developedpneumonia and died three days later despite veterinary treatment.


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 82


Serum samples were screened by theELISA tests for antibodies to RespiratorySyncitial Virus (RSV), Infectious BovineRhinotracheitis (IBR), Parainfluenza 3 (PI3) and BVD.

Heparin blood samples were screenedfor BVD antigen and clots from all theblood samples were screened for BVDvirus by virus isolation. Pooled samples of thymus and spleen tissue from the fivecalves showing clinical signs suggestive of in utero BVD virus infection were alsoscreened for BVD virus by virus isolation.

Laboratory FindingsOf the five calves showing central nervoussigns or developmental abnormalities two showed cerebellar abnormalities –hypoplasia in one (397) and aplasia in theother (398). Histopathology confirmedgranuloprival-type cerebellar dysgenesisin these and also revealed similarhistopathology in a further one (396).Immune-staining failed to identifypestivirus in these but did identifypestivirus in the cerebral neurones of the remaining two (411 and 422), ie thoseshowing no gross or histopathologicalevidence of changes to the central nervoussystem. This finding was considered to beindicative of persistent pestivirus infection.

Lesions suggestive of tuberculosis wereidentified in 14 of the 22 calves, 13 ofthese showed lung lesions ranging from

miliary granulomatous abscesses tocaseous abscesses and generalisedbronchopneumonia. In the majority of these caseous nodules were present in enlarged bronchial, mediastinal and retropharyngeal lymph nodes. The remaining calf showed caseousnodules in the bronchial and mediastinallymph nodes only.

Pasteurella multocida was isolated from the lung lesions of one of the calves only – 398.

Mycobacterium bovis was isolated fromthe lesions in nine of the calves – 386,394, 396, 398, 402, 404, 405, 411 and 422.

BVD antigen was detected in a heparinsample taken from one calf (385). BVDvirus (type I) was isolated from wholeblood samples taken from two of thecalves (385 and 411). In addition BVDvirus (type I) was isolated from a pooledsample of thymus and spleen tissues from two calves (411 and 420).

Serum samples showed no evidence ofantibody to either Haemophilus somnusor Mycoplasma bovis. ELISA examinationfor antibody to RSV, IBR and PI3 virusesshowed titres to all three viruses, adecline in antibody titres to all threeviruses with the increasing age of thecalves was apparent. ELISA examinationfor antibody to BVD virus showed that all

3 4

kCONTINUED ON p84

Figure 3Calf no. 397cerebellar hypoplasia

Figure 4Distribution ofantigen (brown)indicates persistentinfection in Cerebralneurones


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 83

except three had titres greater than 0.4 OD units. One (385) had no titre andcalves 411 and 390 had titres between0.2 and 0.3 OD units. Interestingly, calf420 had a titre of 0.75 OD units.

Further InvestigationsWhilst awaiting the 60 day tuberculin test the farmer culled a cow because ofpersistent mastitis. The cell count hadrisen from 500 (x 1,000/ml) to 2,558 (x 1,000/ml) over the previous fourmonths. The affected quarter wasswollen and firm andduring severalperiods of antibiotictreatment the milkhad been withheldfrom the bulk tankand fed to the calves.At post mortemseveral small caseousabscesses werepresent in the liverand a mesenteric lymph node, showingcaseation and calcification, was enlargedto 5cms in diameter. Multipletuberculous miliary abscesses werepresent throughout one quarter of theudder. Examination of Ziehl-Neelsenstained smears from the udder lesionsshowed very large numbers of acidfastbacilli. Histological examination ofmammary tissue showed sheets ofepitheloid macrophages together withextensive areas of necrosis and central

mineralisation. A Ziehl-Neelsen stainedsection showed numerous acidfast bacilliwithin the lesions and also occasionalbacilli within the ducts. M. bovis wassubsequently isolated from the lesions.

Examination of the udders of all thecows in the herd and scrutiny of theirindividual cell count records was thencarried out. Milk samples were collectedfrom any cows showing abnormal udders or persistently high cell counts.The samples were centrifuged andmicroscopic and cultural examinations

of the sedimentsfor acidfast bacilli were carried out.All of these provednegative.

A total of eightcows were culledbefore the next 60 day herd testand each time thecarcases of these

were examined. Two showed lesions of TB which were confined to theretropharyngeal lymph nodes and liver ofone cow, caseous abscesses were presentin the lungs and liver of one more cow.The 60 day herd test identified a furtherfive reactor cattle. Three of these wereolder cattle showing tuberculosis lesionsin the lymphatic glands of the head orliver. Two, however, were calves agedfour and six-weeks-old both of whichshowed extensive pulmonary tuberculosis.


5 6

Figure 5Mycobacterialmastitis, largenumbers of acidfastbacilli withinnecrotic areas ofmammary glandtissue

Figure 6Two acidfast bacilliwithin lumen ofducts of mammarygland

Lesions suggestive of tuberculosiswere identified in 14 of the 22 calves,13 of these showed lung lesionsranging from military granulomatousabscesses to caseous abscesses andgeneralised bronchopneumonia.


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 84


The two calves were part of a group offour which were the last animals toreceive milk from the cow withtuberculous mastitis before it’s eventualslaughter. The remaining two calves of thefour were slaughtered but showed nolesions of TB.

Blood samples were also collected fromall the cattle on the farm at the sametime as the TB herd test was carried out.Samples were screened for both BVDantigen and BVD antibody. Antigen wasdetected in only one animal – an 18-month-old steer which was thenremoved from the herd. Surprisingly,15% of the herd was still sero-negativeto BVD virus. The subsequent 60 daytuberculin tests did not reveal any moretuberculous animals and has sinceremained clear. Since the episode theherd embarked on a trial to eradicateBVD. This has been based upon aprogramme of vaccination but alsoinvolves testing calves at three months ofage and the removal of persistentlyinfected animals. The preliminary findingsof this suggest a correlation betweenimproved calf health and freedom fromBVD virus. However the results of the trialwill be the subject of a future report.

DiscussionThe significant feature of this case is the high morbidity and severity of thetuberculous disease in the calves.Advanced and disseminated lesions werefound in calves which had passed the TB test less than one month previously. The youngest calf, showing lung lesionsfrom which M. bovis was isolated, wasonly 25 days old. This would suggest that bovine TB is not always a slow and chronic disease, rapid progressionresulting in clinical disease can occur,especially in calves. It is also of note that,despite good evidence that infection was introduced orally to one calf at least,none of the calves showed alimentarytract lesions. However, feeding milk in

buckets to calves may well createaerosols resulting in respiratory infection.It cannot be assumed that all of thecalves were infected from the milk. It islikely that airborne spread also occurredbetween the calves, the lung lesions insome of the calves would have been anabundant source of bacilli for respiratoryinfection of other contact calves. Theposition of the calf houses meant thatthe spread of infection from the calvesback to the cows was indeed possible.

Involvement of respiratory viruses in thesyndrome was not demonstrated. It isunfortunate that it was not possible tocollect paired serum samples from thecalves. The decline in antibody titres toRSV, PI3 and RSV viruses was seen with theincreasing age of the calves and this wouldsuggest that the titres were most likely theresult of maternally derived antibody.

A combination of antigen detection and virus isolation techniques on bloodsamples together with virus isolation andimmune staining procedures on tissuesamples identified three calves whichwere persistently infected with BVD virus. A further three calves showed postmortem changes resulting from in-uteroBVD infection. The presence of an 18month old viraemic animal suggests thatBVD virus had been circulating in theherd for approximately two years at least. The titres to BVD virus seen in these group of calves relate well to theaccepted understanding of BVD in cattle.Calves born with cerebellar anomalieswould be expected to have antibody toBVD virus but not to be viraemic (396,397 and 398). Calves infected at an earlier gestation age would be expectedto be viraemic but to have no antibody to BVD virus (385). There are also thosecalves which fall in between the twogroups. In this case these are representedby the two animals from which virus was recovered and in which virus wasdemonstrated by immune staining.

kCONTINUED ON p86


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 85

These calves (411, 420) showed varying levels of antibody to BVDV. The importance of BVD virus and its rolein immunosuppression and intercurrentdisease has been recognised for sometime. It is known that acute BVD virusinfection in calves can influence the onset of bovine respiratory tract diseaseand this is most probably a sequel to theperiod of leucopaenia which accompanies the onset of acute BVD infection. This poses some interesting questions.

Could a calf which is either persistentlyor acutely infected with BVD virusdevelop extensive tuberculous lesionswhen challenged with M. bovis morequickly than a calf not infected with BVD virus? Also, would such a calf once infected with M. bovis present anunusually high level of challenge to in-contact calves which may themselvesbe transiently infected with BVD virus?Could such calves, as a result of theirsuppressed immune response produce a lowered responseto the intradermalskin test for bovinetuberculosis?

Calves 385, 390and 420 weredemonstrated to bepersistently infected with BVD virus butcultural and histological examinationfailed to confirm the lung changes werecaused by M. bovis. Calf 411 was alsodemonstrated to be persistently infectedand in this instance M. bovis was isolatedfrom the lung lesions. However, as one of the younger animals the opportunityof calf 411 to influence the older calves in the group would have been limited.Before the group of 22 calves wereinvestigated in detail, six calves withtuberculous lesions had been removedfrom the farm. A further two calves, one a confirmed case of tuberculouspneumonia, had died. Unfortunately, atthat stage, the involvement of BVD virushad not been suspected and the BVD

status of those calves remains unknown.It is likely to have been similar to the BVDstatus of the 22 calves described here.

Whether or not persistent and acuteBVD infection influenced the course ofthe mycobacterial infection in this herd is open to speculation. What is importantto recognise is that M. bovis is a cause ofclinical disease in cattle and that whetheror not M. bovis bacilli are excreted in themilk of a dairy cow or in aerosols fromthe respiratory tract of a calf there is an opportunity for zoonotic transmissionparticularly if the immune system of the in contact individual is suppressed.

AcknowledgementsThe assistance of Roger Sainsbury(Animal Health Office, Defra), John Headand Stephen Trethewey (Head and Head,Helston) for bringing the case to myattention and assisting in examining and collecting blood samples from calves is gratefully acknowledged.

The author wouldalso like to thank Dr Sandra Scholes,VLA Lasswade, for carrying outhistologicalexaminations of

tissues and the Virology Departmentof VLA Weybridge for virologicalinvestigations. S. Waterhouse, VLALangford, for serological investigationsand the laboratory staff at VLA Truro andStarcross for bacterological examinations.The author would also like to thank Lyn Penrose for typing the manuscript.


l Information

Robert Monies is a VeterinaryInvestigationOfficer VLA, Truro.

The significant feature of this case isthe high morbidity and severity ofthe tuberculous disease in the calves.


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 86


BackgroundThe spectrum of hosts susceptible to M. bovis infection is very broad. Although cattle are the natural hosts ofthe bacterium, all terrestrial mammals are susceptible to infection to a variabledegree, as determined by the exposurelevel, innate resistance, predominantimmunological pathways and type ofhusbandry. However, their opportunityto transmit the disease to other animals(particularly cattle) varies according totheir ability to maintain TB within theirown populations. Coleman and Cookehave classified host species of M. bovisas follows:

• maintenance hosts (e.g. bovines,badgers and occasionally farmed and wild deer), where the infectionpersists by vertical, pseudo-vertical or horizontal transmission within thespecies, without the need for inputfrom other species; and

• spillover hosts, where TB occurswithin the species only as long asthere is input from an externalsource. Spillover hosts may in turn be either ‘dead-end’ hosts (if theincidence and pathology of thedisease indicates they play nosignificant role in its onwardtransmission) or ‘amplifier’ hosts ifthey appear capable of increasing the prevalence of TB in livestock orother species.

1. SheepTB is rare in sheep, despite the fact thatmany flocks in TB hotspots must beexposed to M. bovis. The first reportedcase of TB in sheep in Britain occurred in

2003 involving a flock in close associationwith infected cattle. Six sheep reacted tothe tuberculin test, four of which hadtuberculous lesions at post mortem.

2. GoatsBovine TB does occur in goats. Milk frominfected goats is a potential human healthrisk. Goats may be considered spillover(amplifier) or maintenance hosts,depending on the type and location ofthe herd. In Mediterranean countries, for example, goats are known tomaintain M. bovis infection and there is some evidence of transmission of goat-adapted clones of M. bovis (M. bovissubsp. caprae) to humans and cattle.

SVS standing instructions state that‘Goats should be placed undermovement restrictions and officiallytuberculin tested at the Departments’expense if located on premises where TB has been confirmed in cattle, or if M. bovis infection has been confirmed in the goat herd itself. Consideration shouldalso be given to testing other goat herdson contiguous premises. This is stronglyrecommended if any dairy goat herdssupplying/retailing unpasteurised milk are at risk of infection’.

3. HorsesHorses are reported to be resistant tohuman, bovine and avian TB. Isolatedcases do however occur. Thus horses are true dead end hosts.

4. PigsThe oral route is considered the mostimportant route of infection in pigs,

TB in domestic species other thancattle and badgers

Wyn Buick,State Veterinary

Service (SVS),Carmarthen

A u t h o r

kCONTINUED ON p88


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 87

most frequently caused by feeding milkor milk products from infected cows(domesticated pigs) or scavengingcarcases of tuberculous animals (feralpigs). The status of the pig as a truedead-end or amplifier host of TB is stillbeing debated. TB is not considered to be particularly contagious amongst pigsor to spread easily from pigs to otheranimals. In most cases, the disease is self-limiting and no control measures arerequired. However, evidence from NewZealand and Spainsuggests that cross-infection betweenferal pigs/wild boarand extensivelyreared domestic pig populations is possible in theabsence of cattle.

M. bovis was acommon cause ofTB in domesticatedpigs in the early to mid 1900s. In recentyears, the vast majority of suspect TBlesions detected by the Meat HygieneService during slaughterhouse inspectionof pig carcases have proven positive forM. avium. Still, in the period 1999-2003four incidents of M. bovis infection wereconfirmed in pigs. The first one took placein 2000 on a wild boar farm in Cornwall.The second one was disclosed in 2002 intwo fattened pigs in an outdoor unit inWiltshire. Both pig premises were locatedin areas with a well established and

documented history of bovine TB in cattleand badgers. The other two incidentswere in annual cattle testing areas.

5. South America Camelids (llamas,alpacas, vicuñas, guanacos)TB is not a major health problem with camelids, but these species dooccasionally develop the disease.Although reports of infection in theirnatural habitat in South America are few,cases of bovine TB have been diagnosed

in llamas andalpacas in New Zealand, the USA and inGreat Britain. M. avium and M. microti infectionshave also beendocumented. Some authors haveconcluded that TBshould be considered

in the differential diagnosis in all cases ofill-thrift in these species, with or withoutobvious respiratory signs. As with othernon-bovine species, there is littlelegislation underpinning the managementof TB incidents in camelids in Great Britain.There is no requirement to identifycamelids or record their movements.Divisional Veterinary Managers (DVMs) or Local Authorities have no legal powersto enforce TB testing in camelids andslaughter any reactors. Similarly, thereare no provisions to compensate owners


Figure 1Tuberculous Liver –Pig

Figure 2M. Microti Liver –Llama

2

Although cattle are the natural hostsof the bacterium, all terrestrialmammals are susceptible to infectionto a variable degree, as determinedby the exposure level, innate resistance,predominant immunologicalpathways and type of husbandry.

1


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 88


for the loss of such animals.TB testing of camelids should be by the

single intradermal comparative cervicaltest (SICCT), as in cattle. Although notperfect, the SICCT is a suitable test forassessing the status of individual animalsin a camelid flock with confirmed TB. This test provides reasonable sensitivityand specificity and it is the official testingprocedure for camelids imported intothe UK. The γ–IFN test (BOVIGAM®) is not a valid test for the diagnosis of TBin camelids.

6. DogsAlthough TB in dogs was a relativelycommon infection in the late 19thcentury and first half of the 20th century,more recent case reports of canine TB in the veterinary literature are scarce. The lower frequency with which TB iscurrently diagnosed in dogs probablyreflects the reduced prevalence of TB in human beings and cattle in thedeveloped countries. EpidemiologyDogs, like any other terrestrial mammals,can be naturally infected with M. bovis,M. tuberculosis, M. avium complex and other mycobacteria. However, theconsensus of opinion is that dogs do notrepresent an epidemiologically significantsource of infection for cattle, other dogsor other animals.

At present the disease appears to beextremely rare in dogs in GB. Typicallyless than five canine samples are processedat VLA annually for mycobacterialculture. Only one submission has provenpositive to M. bovis since 1999. Pathology In the dog the primary complex occursmost frequently in the lungs and associatedlymph nodes. Pleuritis (pleurisy) is acommon complication. Lesions in theliver and mesenteric lymph nodes are also common. TB lesions in the skin andsuperficial nodes are more commonlyobserved in cats than in dogs.

Treatment Because of the risk of transmission to other animals and human contacts,Defra‘s official line is to discourage the treatment of companion animalsinfected with M. bovis or M. tuberculosis.Furthermore, there are significantdifficulties and financial costs associatedwith any anti-tuberculosis treatment,which involves a prolonged course ofantibiotics for at least six months. The prognosis depends on the extent and severity of infection, but it isgenerally guarded.

In general, the antibiotics used to treatmycobacterial infections in dogs andother companion animals are the same as those used in humans. Some of theseare inherently toxic.

There has been only one confirmedcase of TB caused by M. bovis in dogs inthe last five years. This was in a dog fromGloucestershire. M. bovis spoligotype ten (the most prevalent type of M. bovisin cattle in the eastern half of the county)was isolated from pulmonary lesionssubmitted to VLA for culture in February 2002.

7. FerretsIn New Zealand, high rates of infectionhave been found in feral ferrets (up to66% on infected properties). In the UK

3

Figure 3Mesenteric lymphnode – Ferret

kCONTINUED ON p90


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 89

infection has been found in domesticferrets and in one feral ferret. Waleshas a significant population of polecats(closely related to feral ferrets) whichare well known to regularly frequentfarmyards.

8. CatsThree infectiousdiseases caused bybacteria of the genusMycobacterium havebeen documented incats. These includeclassical tuberculosis(TB), feline leprosy (acutaneous disease)and opportunisticmycobacterialinfections (again, typically, cutaneousconditions). Only classical TB in cats is ofrelevance to State veterinarians in thecourse of their duties and, as such, it isdiscussed in detail in this paper.

Classical TB in cats is quite a rare disease,traditionally caused by two members ofthe Mycobacterium tuberculosis complex:Mycobacterium bovis and M. microti (thevole tubercle bacillus). Infection of catswith M. tuberculosis (the human tuberclebacillus) appears to be even more rare,probably because cats are naturally moreresistant to the human than to the bovinebacillus. Sporadic infections of cats withM. avium have also been described in theveterinary literature.

Cats are considered spill-over hosts ofM. bovis, M. microti and M. tuberculosisbecause removal of the source ofinfection in cattle, wildlife or peopleresults in a reduction in the incidence ofTB in cats. However, cats are not true endhosts as the disease presentation makesthem (at least theoretically) capable oftransmitting these infections to other catsand other mammals, including humans.Cats are thus better regarded as ‘spill-over amplifier’ hosts of M. bovis.

Cats become infected by exposure toinfectious material from tuberculouscattle (or other maintenance hosts of M. bovis), tuberculous wildlife or throughclose contact with infected people.Therefore, the incidence of TB in the cat population is often a reflection of the local prevalence of bovine TB in

cattle/badgers/wildrodents and, to alesser degree, M. tuberculosisinfection in people.

Bovine TB in catsused to be quitecommon in GB.Most cases of felineTB were believed tohave resulted from

the ingestion of contaminatedunpasteurised milk from tuberculouscattle. More recently, after theintroduction of regular TB testing ofcattle and milk pasteurisation, othersources of infection for cats havebeen postulated:

• wounding by tuberculous possums inNew Zealand

• hunting prey (voles, mice) infectedwith M. microti

• direct or indirect contact withtuberculous badgers in GB

• scavenging of tuberculous wild deercarcases or offal in Michigan, USA.

PredispositionMost cases of feline TB are probablysubclinical. Infection usually occurs afterprotracted exposure, e.g. repeatedexposure to infectious small mammals,living on a farm housing tuberculouscattle or living for prolonged periods withinfected humans. TB is therefore seenmainly in adult cats and, interestingly, isseen most commonly in males. Infectionwith feline immunodeficiency virus (FIV)could predispose cats to infection with


Because of the risk of transmission toother animals and human contacts,Defra’s official line is to discouragethe treatment of companion animalsinfected with M. bovis or M. tuberculosis.


10493 AW Defra GVJ_53-90 29/8/06 14:25 Page 90


mycobacteria. However, very few catswith mycobacterial infections haveactually tested positive for FIV or feline leukaemia virus (FeLV). Incidence Between 1980 and 2003 M. bovis wasisolated from cats on 41 occasions atVLA. The incidence of tuberculousdisease in cats caused by M. bovis hasbeen sporadic in GB. However in 2005the VLA received 90 supect felinesamples, 13 of which were positive for M. bovis and 23 for M. microti-like or M. avium-complex organisms. This ismore than double the recorded incidencein previous years.

ConclusionAmong domestic animals in Great Britain, pigs, horses, sheep, cats, dogsand camelids are all considered spilloverhosts as is perhaps the case with all othermammalian species (badgers, cattle anddeer excluded). In other words, thesespecies become infected only when thechallenge level is relatively high, and theycannot sustain the infection within theirown populations in the absence ofinfected cattle or a wildlife reservoir. This does not mean that, if infected,these species cannot occasionallytransmit the disease to other animals and humans, i.e. some of them may actas amplifier hosts.

TB is particularly rare (but not unheardof) in horses and sheep, which should be considered true dead-end hosts.

Pigs, farmed wild boar, dogs, cats andcamelids should be treated as potentialamplifier hosts of TB.

AcknowledgementsMuch of this article has been edited fromthe SVS standing instructions on Bovinetuberculosis – Viper Ch 23 section Yprepared by R. de la Rua-Domenech.

4

Figure 4Popliteal LymphNode – Cat

l Information

Wyn Buick is aVeterinary Officerwith the SVS,Carmarthen.


The Government Veterinary Journal (GVJ) is the official journal of theGovernment Veterinary Service and those who support its work or have aninterest in state veterinary medicine. It is compiled and produced by theDepartment for Environment, Food and Rural Affairs (Defra) for, and on behalf of veterinary surgeons and those who support them across all parts of Government.Its key aims are to enhance the contribution of veterinary expertise within andacross government, promote the work of Government veterinary surgeons andprovide a range of technical, factual and interesting articles in the fields of:

• Disease control • Animal welfare • Public health • Consumer protection

It is intended that the GVJ will also highlight progress in relation to the AnimalHealth and Welfare Strategy and focus on issues identified by the VeterinaryHead of Profession in Government.

Editorial Board

Editor: Linda SmithDefraBlock A GovernmentOfficesColey ParkReadingBerkshire RG1 6DT

Board: Jose BisMartyn BlissittWyn BuickAdrienne ConroyPaul GethingsAndrew GreshamAngela JonesSivam Kumar

Alex MorrowRichard RowlandLinda Smith

Production Team: Peter GreenAngela GreenawayEmily Thorne


Correspondence, including distribution inquiries and requests for additionalcopies should be sent to:

GVJ Production Team Internal Communications andStakeholder Management Team Department for Environment Food and Rural AffairsArea 206 1a Page StreetLondon SW1P 4PQ

Tel: 020 7904 6482 Email: [email protected]

The Government Veterinary Journal canbe viewed on the Government VeterinarySurgeons (GVS) website atwww.gvs.gov.uk. If you would like further information onany of the topics addressed in this issueplease contact your local Animal HealthOffice or the Production Team. Articles appearing in the GovernmentVeterinary Journal are Crown Copyrightand permission must be obtained fromthe Production Team before they can bereproduced elsewhere.The Government Veterinary Journal isdistributed free to its readers. It is notavailable for sale.

Website addresses included in this report are correct at time of publication. If you experience any problems with links, please contact [email protected]

Front Cover: State Veterinary Service vet recording details about the tuberculosis inspection.

The GVJ is printed on recycled paper containing 80% post consumer waste and 20% totally chlorine free virgin pulp.

10493 AW Defra GVJ_03-52 29/8/06 14:27 Page 2

www.gvs.gov.uk

VOLUME 16 | NO1 | 2006

GVJGovernment Veterinary Journal(formerly known as the

State Veterinary Journal)

Bovine TB Special Edition

PB 12125 Nobel House17 Smith SqaureLondonSW1P 3JR

About Defrawww.defra.gov.uk

ContentsForeword 4


2 Some lessons from the history of the eradication of Bovine Tuberculosis in Great Britain 11

3 Bovine Tuberculosis in the European Union and othercountries: current status, control programmes andconstraints to eradication 19



6 The introduction of pre and post-movement TB testing in Scotland for cattle from high incidence TB areas 58

7 Ante mortem diagnosis of Bovine Tuberculosis: thesignificance of unconfirmed test reactors 65




Bo

vine TB

Special Ed

ition

Go

vernm

ent V

eterinary Jo

urn

alV

OLU

ME 16 |

NO

1 |2006

10493 AW Defra GVJ_03-52 29/8/06 14:27 Page 1

Documents

VOLUME 16 NO1 2006 10 TB in domestic species other than ... · GVJ Production Team Internal Communications and Stakeholder Management Team Department for Environment Food and Rural