Content on this page requires a newer version of Adobe Flash Player.

Get Adobe Flash player

Section 3. Cattle TB Eradication Part 2

Part 2. Identifying infectious cases .. The hidden reservoir of cattle TB

It is critically important to correctly identify and remove cattle TB cases before they can spread TB within and then between herds . Traditionally this has been achieved by using the skin test which in both cattle and humans entails inoculation of a tiny amount of TB protein/antigen intradermally, then checking after 72 hours for any immune response in inflamation /swelling at the site if TB presence has stimulated this cell mediated response. As shown in Figures 3 and 4 , "Becoming infectious" there is a steady increase in the bacillary load in the lungs and infectiousness ; and the skin test works for the middle part of the disease progression. Confirmation of diagnosis of TB in reactors then divides cases traditionally into NVL / Non Visibly Lesioned or unconfirmed cases and VL / Visibly Lesioned or Confirmed cases .. in which in addition it was shown samples did not / did contain M.bovis bacilli. This diagnostic process hence Completely misses early and late TB cases; which together with the NVL/Unconfirmed cases constitute a huge hidden reservoir of cattle TB, in Figure 4, the Iceberg right hand pyramid, below "C"= Confirmed= Below sea level ".

TB is like a slowly metastasizing cancer, the immune system is unable to arrest this progressive (broncho-)pneumonia with steady multiplication of bacilli, outwardly seen in an increase in the number and size of microscopic, then visible lesions , eventually with cavitation as lesions join , and then spread to other organs , in particular udder and uterus with risk of spread to calves in milk or before birth in the womb. Effectively stage 1 cases with NVL unconfirmed TB ; STAGE 2 VL Confirmed cases; stage 3 advanced clinical cases with emaciation, tubercular cough. these intergrading stages , in the progressive increase in the bacillary load hence also shows :-

  • An increase in infectiousness in bacilli shed externally;

  • Tests :- the skin test too goes from non-reactor, to IR /Inconclusive Reactor, then weak to strong reactors (Collins 1997, 2000).. until the immune system becomes swamped with bacilli, so the skin test ceases to work, and the immune system switches from a cell mediated immune (CMI) response to a humoral one ie. production of blood antibodies which can be measured with new tests, eg Ireland routinely use the ENFER Chemiluminescent Multiplex blood ELISA; the OIE have recently approved an IDEXX M.bovis Ab antibody test. A multiplex test detecting early and late TB in deer is used in New Zealand (Griffin 1994). Because the CMI response early on, in the mobilisation of white blood cells to become macrophage/killer cells which can destroy bacilli, is modulated by the chemical (cytochine) gamma Interferon /IFN ; the IFN test is a valuable additional test to skin testing to find cases earlier on, many skin test Inconclusive Reactors were IFN positive (Neill 1994; Vordermeier 2006). The LTT/Lymphocyte Transformation Test likewise detects this , but is impractical for a field test.

Pragmatically , owing to these limitations of the available tests, and since it takes about a year to reach the "good" reactor / VL / More infectious stage In Figures 3 & 4; annual testing is critical because it detects cases before they become "active spreaders". And as described in the previous Historical Section, in essence TB Schemes simply halt and reverse this progression .. Stage 3 advanced cases skimmed off first; then gradually reducing the Stage 2 cases, so that as eradication approaches , the proportion of Stage 2 VL to Stage 1 NVL cases decreases; so that the the majority of cases have only attained Stage 1 NVL /Unconfirmed status.... see Figure 6. Any premature relaxation of widespread annual testing when TB gets to low levels within an area eg. by fewer tests , which often means going to 2 yearly testing simply allows TB spread to resume both within and then to new herds as seen in above "Historical section".

Toil And Trouble : A Seething Cauldron Of Red Herrings

The seven main problem areas in achieving a correct diagnosis of infectious cases within the "hidden cattle TB reservoir "; as shown in the "Iceberg" in Figure 4 .. is rather complex and very confusing, but can be summarised as follows (Palmer 2006, rua Domenech 2006) : --

  1. Only "Open" Visible Lesion Cases Infectious

    This is one of the most damaging and persistent of Darwin's "false facts" in the entire badgers and TB debate; impairing a true interpretation of TB schemes eg. Hewson 1987, the Ostertag method (Francis 1947). In human TB many tubercles do become encapsulated and " closed "so non-infectious lesions; and recent studies suggest badgers too may develop harmless lung tubercle encapsulation (Gallagher 1998). But this is a major difference as regards cattle since apparently "closed" tubercles continue to seep exudate peripherally (Francis 1947, Jubb ). And tubercles are "open" and infectious, shedding bacilli at any stage of the disease progression, so that all reactors should be regarded as infectious ( see quotes in Introduction :- Francis 1947, 1958; Blood). Well understood by M'Fadyean 1888 (Pritchard 1988); who also warned against the artificial distinction between NVL/ VL cases; Bang 1892; Stamp 1948; Richards 1972; O'Connor 1986). Recent studies have re-discovered the fact that cattle may be infectious before they develop visible lesions or react to the skin test :- 20 % of NVL cases sputum positive ( McIlroy 1986, Neill 1988, 1992, 1994 )... although one too brief in-contact experiment found no transmission to 32 contacts, despite most having lung lesions (O'Reilly 1988). Early on in the GB TB scheme Steward 1941 found that 399 out of 400 "sputum positive" cows had visible lesions. Becoming an "efficient disseminator" has always been problematic .. only a minority of human lung TB patients are seriously infectious (Griffin 1995); but cattle superexcretors do occur in the field and in one infamous cattle show with rapid spread to contacts over a 4 day period (Steger 1970)... TB also returned to Guernsey after a show (Francis 1947).

    One problem has always been knowing the time it takes to reach the active spreader infectious stage.... about a year , 6- 18 months. Fischer 2005 notes 6 mnths. Griffin 2002 suggested 9 % by 6 m, 17 % by 12 m, 26 % by 18 m, 34 % by 24 months. VL status achieved some 6 months after previous clear test (3-20 mnths ; Neill 1988, McIlroy 1986).

  2. False Tubercles.. lesions may be due to other bacteria Actinobacillus, Corynebacterium

    Nocardia; or be due to lung worms or pneumoniocosis / silicosis as in miners .. latter two also in badgers which in addition suffer from fungal adiaspiromycosis (Higgins 1985). However, recent studies found no evidence of a silica susceptibility in badgers (Gallagher 2000, Corner 2012). (Wilesmith 1987, Liebana 2008).

  3. No Tubercles..there may be no tubercles because truly false positive reactors do not have TB

    Early unconfirmed cases may have yet to develop visible lesions (false negatives) ; abattoir inspection aims at finding gross lesions to protect the foodchain, but may only be 50 % accurate, missing less obvious lesions (Corner 1990, 1994). Late on as eradication schemes near completion, it becomes critically important to find early NVL cases, so for example, Australia relied heavily on a NGSP National Granuloma Submissions Programme (More 2006).

  4. False M.bovis negative.

    In many countries in Africa and Asia without easy access to hospital clininal facilities, identifying and treating infectious patients is via sputum smears , but it has has long been recognised this is difficult when low numbers of bacilli are shed intermittently :- needs c. 100, 000 bacilli/cc ( Steward 1941, Liebana 2006) Similarly in identifying AFBs (Acid Fast Bacilli) in stained histology sections.

  5. False Positives

    Specificity of tests is the percentage of reactors which truly have TB ; 99.9% means only 0.1 % wrongly identified. Cross-reactors may be due to other Mycobacteria, eg. M. avium, avium- paratuberculosis or Johnes disease including vaccinated cattle, human, skin, TB. Historically, cows infected with M. tuberculosis "muddied" the testing progress in Finland, but was of little significance in GB when fewer human TB cases around (Lesslie 1960, 8 examples, Francis 1947 ). Cross reactions also as regards brucellosis, liver fluke or fascioliasis (Richards 1972, Wilesmith 1987 .. recently emphasised by D. Williams of Liverpool university with AFBINI, Ulster). With up to 14 % of cattle reacting to avian TB, both GB and Ireland have long had EU derogation to use the comparative skin test with avian/bovine antigens to avoid false positives.

    Alas, based on this truly false positive reaction to avian TB, it has long been assumed wrongly that NVL/UNConfirmed cattle apparently without gross lesions or detectable M.bovis are also false positive. This is perhaps the single most important mistake in the whole "Badgers and TB Debate" . Up to 85 % of new breakdowns at the peak of Mad Cow replacements in 1994 were Unconfirmeds ; and since Foot & Mouth 2001, some 200, 000 out of 340, 000 cattle culled were also Unconfirmeds. But skin test specificity 99.99 % , so only 1 in a thousand are truly false positive. Several detailed studies of unconfirmed reactors found that many DID have TB :-

    McIlroy 1986, serial lung sections showed 73 % with micro-lesions;

    Wilesmith 1987, Dunnet 1986 para. 32 .. 70- 80 % exposed to M.bovis;

    Nassel 1956 out of 1000 NVLs, 50 % confirmed .. 87 % with micro-lesions, 13 % by culture (Myers 1969);

    Lesslie 1960 :- 1952-1960, 1990 NVL reactors, 660 were Mycobacterium bovis, 30 M.avium, and just 12 M. tuberculosis.

    Largely unnoticed , but the TBEG 2009 report para. 54 belatedly recognised the significance of unconfirmed cases . Under Directive EC 64/432 both confimed AND UNconfirmed breakdowns must have 2 clear tests to de-restrict the herd, and IR Inconclusive Reactors removed at first re-test. Rather ironically, with up to 80 % of TB badgers NVL , recent studies have replicated cattle scenarios .. far more confirmed cases via more detailed autopsies with serial sections of tissues, and culturing of samples and DNA Spoligotype absolute confirmation (Crawshaw 2008, Murphy 2010).

  6. False Negatives

    sensitivity of tests is the percentage of actual TB cases missed by the test ; 80 % means 20 % missed . Scotlands use of pre- and post-import testing means 16 out of 20 found / then 19 out of 20 , so few slip through the net. 80 % accuracy means 2 in 10, or 200 in 1000 residual cases so herd never does go clear. Also 20 in 100 singleton herds missed (Morrison 2000). There are at least five different sorts of false negatives :-

    1. Very early cases not found even by the IFN test .. an often quoted figure of 8 - 51 or 65 days to become skin test reactors, but that probably was a high dose result (Francis 1947). Cattle may be reactors at birth if infected in utero, or with high dose nasal inoculation react within 2-3 weeks (Cassidy 1998, 1999); but in natural infection with just 1-2 airborne droplet nuclei it may take much longer to reach reactor status. A classic but forgotten study found 13 "in -contact" calves caught TB from the main herd, by aerosol with a 6 m gap , 5 became reactors at 6 months, the other 8 at 12 months (Svensson 1904, in M'Fadyean 1910) ; the ISG study found half in-contact cattle were reactors by 7 months (p. 235); and in the only proven badger to cattle transmission yard experiment 8 calves out of 12 caught TB and were reactors by 6- 10 months (Little 1982). So, Goodchild 2001 suggest a "latent period" of unresponsive + reactor but "uninfectious" of c. 6- 12 months;

    2. Poorly understood and under-reported is the role of immunosuppression on the test; the skin test antigen inoculation means cattle become desensitized, repeat tests not before 60 days (Coad in ICMB 5);

    3. IBR , and both BVD and BIV , the bovine form of human HIV/AIDS and liver fluke may suppress the immune system (Monies 2006, Wilesmith 1987);

    4. Lastly, the complex interplay of nutritional status/stress/ and interplay of reproductive hormonal --immune system factors affects tests....a significant minority of cattle may become transient inconclusive or weak reactors, then suppress the progression of TB until often puberty/pregnancy alters this balance so re-activating disease progression. Latency. A majority of cattle show a steady progression of disease, become rectors within a year, so the majority of breakdowns test "clear" within a year. MAFF traditionally regarded a breakdown as part of a single "episode" if cleared within 15 months, beyond that they were assumed to be "new" introductions. However, arrested or slow development of latent TB has been recognised for over a century (Bang & M'Fadyean 1890s ; Francis 1947). Cases can remain latent for at least 7 - 11 years (Goodchild 2001). Good 1994, noted a batch of calves infected in 1987, 1 reactor ; 20 reactors in 1989; then 1 each 1990-2, so 24 reactors/14 were VL,over 6 years .. all stayed in the original herd, but if they had moved to new herds, that would have been X more "due to badgers" ! A dozen IR Inconclusive Reactors took up to 4 years to surface later on as full reactors (Hewson 1987 , Zuckerman p.73, L.Negus Cornwall 2 examples, Costello 1998, TBF 63 & 72). One IR accompanied by 4 companion VL cases re-surfaced 11 years later, having passed intervening tests, when it developed udder TB affecting 16 calves (Good M pers.com). Goodchild 2001 ... since risk of repeat breakdowns rises with severity of previous breakdown and herd size , using 1999 data ; suggested that breakdowns of 6 to 40 reactors have a 4 or 9 % chance of 1 case being missed after the 2 clear tests // so at 6 mnth check test a 13.5 % chance of a repeat breakdown // and for herd sizes of 100/ 101-200 / 200+; further reactors in 6.9 / 10.4 / 14.2 % of cases. Clegg too found 10 % of herds likely to have a missed case.

    5. Cows suppress their immune response during pregnancy, the calf being half "foreign" protein, formerly allowing cows with uterine TB to undergo a proliferation of bacilli and lesions ..which rapidly regresses after birth with restoration of "normal" immunity . Birth waters may contain up to a kg of virulent puss, but surprisingly as few as 1 % of calves catch TB in utero (Francis 1947). This sequence is why up to a third of cows are temporarily non-reactors for a couple of months after calving , as first recognised by Kerr 1947 (Blood, Goodchild 2001, Liebana 2008)....perhaps why repeat tests only 65 % accurate so 1 in 3 missed (Monaghan 1994) ..and why around a third of bad breakdowns may suffer a repeat within 14 months (ISG p.239); similarly, Kelly 2008.. A.Conlan, of Cambridge recently suggested 20-50 % suffer repeat breakdowns within 2 years. At the low point whilst 85 % of breakdowns were cleared within a year, some 9 % lasted for 2 years ( 1972-8, 1099 confirmeds,Wilesmith 1983) . It seems unfortunately, that after a number of pregnancy/ calving cycles, a minority of cows fail to regain normal immunity, so becoming "ANERGIC" non-reactors . Such VL infectious cases result in chronic breakdown herds and occurred in 36 and 33 % of 47 / 39 depopulated hards (Costello 1997; Good 2011).

  7. Untested / untraced breakdown herds .

    Most TB testing is in the minority of the national herd within "hotspot annual test" areas.. but TB has always been diffusing outwards into supposedly clean longer test interval areas. It may be impossible to trace up to 75 % of breakdowns to a confirmed TB source (Wilesmith 1983, Hewson 1987, Gopal 2006). SO attribution of causes of breakdowns has always been somewhat "informed guesswork" ( Zuckerman p. 79, Dunnet p. 53, Krebs 158-64, ISG p. 121-38 .... TB 99 and other questionnaire forms ; the ER 76 a & b ones in Eire .. hence Vet. Field Manuals for both GB and Southern Ireland ).

In conclusion : - The hidden TB reservoir has been cattle all along !

As noted in the historical section above, at the low point in the 1970s (Figure 1, Map 2 ).. there was an "archipelago of intractable TB hotspot islands" supposedly due to high density badgers .. but they were in reality high density dairying hotspots (SEE also Figure 6 in section 4). Dairy cows having longer working lives than most beef stock , in bigger herds , and are more prone to develop advanced TB which as variously outlined above and as summarised in the hidden reservoir "ICEBERG" of TB in Figure 4; include a variety of non-reactor early or anergic or otherwise "missed" cases. Whilst most herds have minor TB incidents and go clear with 2-3 re-tests within a year, some 10-15 % of problem herds may have more major incidents with over 6 reactors / breakdown, and so may develop endemic residual TB which is much harder to eliminate (Wilesmith 1983, ISG p. 142). These are the "engine" of within-herd spread, and wrongly testing "clear" are free to export a random scatter of new cases , often IRs, or UNConfirmeds far and wide... up to 85 % of the Mad Cow mid-1990s re-stocking scatter of breakdowns (see maps Krebs 1997) .... and since foot and mouth 2001 some 340, 000 TB cattle removed ..including c. 200,000 UNConfirmeds : - a Huge hidden reservoir spreading amongst the cattle population. As shown in Figure 5, the annual test core area may have a few chronic herds with Anergic reactors, as described for the West Penwith study above ..and the scatter of new breakdowns into 2 or 3-4 year test areas may have more confirmed incidents. Even intractable W. Penwith went clear with depopulation of the key chronic herds in 1985 (Richards 1972). These supposed "Islands" of badger TB were in fact, as shown by the clean ring culling 1982-4 NOT of continuous widespread badger TB, but incredibly tiny micro-pockets of a few badger clans which had caught TB from the epicentre of the cattle herd/s cluster (Krebs 1997).. again, see Figure 6 in next section. So, as was noted as long ago by Bang in 1892 .. and again Blood in 1989, early and late TB cases have always been the cause of recrudescance of TB in cattle herd breakdowns.... not poor Old Brock, since despite 42 years research, no-one has realistically shown how badgers Might give cows a respiratory lung infection ! .

Table : TB Badgers; recorded 1972-1996 (culls 1974-96); from Maff Consultative Panel Badger reports 1-20

3542 TB+BADGERS /19098 culled (18.5 %) ; 1066 +/23032 (4.6 %) RTAs = 4608 TB+/42130 (10.9 %) culls +RTAs; MOST , c. 90% from the worst cattle TB OR the 7 "Badger Problem Counties"... Avon-Glos. + Wilts-Somerset (latterly Heref./Worcs in "southwest problem area); NOTE : few from other cull areas Staffs/Sussex....... Total cull, Excludes some 20,000 gassed early on . Seems rather obvious, that there was no widespread self-maintaining badger TB reservoir, badger TB cases merely a spillover from preceding cattle problem !

Figure 3

Figure 4

Figure 5

Figure 6