- Email: [email protected]
The effectiveness of parallel gamma-interferon testing in New Zealand’s bovine tuberculosis eradication programme
Preventive Veterinary Medicine 127 (2016) 94–99
Contents lists available at ScienceDirect
Preventive Veterinary Medicine journal homepage: www.elsevier.com/locate/prevetmed
The effectiveness of parallel gamma-interferon testing in New Zealand’s bovine tuberculosis eradication programme J.A. Sinclair a,∗ , K.L. Dawson a , B.M. Buddle b a b
OSPRI New Zealand, 50 Church Rd., Hamilton 3200, New Zealand AgResearch, Hopkirk Research Institute, Palmerston North 4442, New Zealand
a r t i c l e
i n f o
Article history: Received 29 November 2015 Received in revised form 25 February 2016 Accepted 31 March 2016 Keywords: Bovine tuberculosis Eradication Sensitivity Gamma-interferon
a b s t r a c t In bovine tuberculosis (bTB) eradication programmes, especially where prevalence is low, sensitivity of testing in infected herds must be maximised to reduce the possibility of recrudescence of prior infection and the risk to other herds via animal movement. The gamma-interferon (␥-IFN) assay applied in parallel with intradermal tuberculin testing has been shown to increase test sensitivity. The aim of this work was to substantiate this effect in the field. A retrospective observational study was conducted on 239 New Zealand cattle breeding and dairy herds with bTB infection between 1 July 2011 and 1 September 2015 to evaluate the outcomes of new policy introduced in 2011. The investigation defined the number and proportion of reactors (animals testing positive and slaughtered) found with lesions of bTB in intradermal caudal fold testing (CFT) and parallel ␥-IFN testing, at the breakdown test or first whole herd test after breakdown, WHT(1), and at the final or projected final whole herd test, WHT(F). Parallel ␥-IFN testing was used in 26.8% of the 239 herds at WHT(1), and 430 animals in 49 herds were deemed reactors. One hundred and sixty (37.2%) of these reactors from 32 herds were found to have bTB lesions, despite having been negative to caudal fold testing. These 160 infected animals accounted for 29.6% of all infection found at WHT(1). At WHT(F), parallel ␥-IFN testing was conducted on 93 herds and detected a total of 122 reactors in 49 herds, in addition to those found by CFT. Twenty-one of these reactors, from 13 herds, had bTB lesions at slaughter, accounting for 67.7% of all reactors found with bTB at WHT(F). Eleven of these 13 herds would have had their movement restrictions revoked based on a negative herd CFT alone, and could potentially have caused outward transmission of bTB to other herds, as well as experiencing recrudescent breakdowns. We conclude that ␥-IFN testing in infected herds, in parallel with intradermal tuberculin testing, is a valuable tool in a bTB eradication programme, as it enables higher test sensitivity at both herd and animal level. The use of the ␥-IFN test over a risk cohort early in a breakdown assists in removal of early infection and some cases of anergy to intradermal tuberculin testing. Parallel ␥-IFN with compulsory slaughter of reactors should be considered in breeding and dairy herds in conjunction with tuberculin testing before movement control is revoked, and will assist in achieving TB freedom on a herd level and nationally. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Bovine tuberculosis was introduced to New Zealand by cattle imports with settlers in the mid-1800s. By the 1950s, the disease had become widespread in both the North and South Islands. Voluntary test-and-slaughter schemes failed to eradicate bTB, and in 1967 infection was identified in Australian brushtail possums,
∗ Corresponding author. E-mail address: [email protected] (J.A. Sinclair). http://dx.doi.org/10.1016/j.prevetmed.2016.03.020 0167-5877/© 2016 Elsevier B.V. All rights reserved.
Trichosurus vulpecula (Ekdahl et al., 1970). Possums were subsequently found to be the principal wildlife reservoir for M. bovis in New Zealand. Control of possums was undertaken in the 1970s but funding was interrupted in the 1980s, resulting in a dramatic geographical spread of wild animal infection. Annual infected domestic herd numbers peaked at 1684 infected herds at 30 June 1995 (data held by OSPRI New Zealand) and by 2004, bTB had become endemic in wild animal populations over an estimated 39% of New Zealand’s land area (Livingstone et al., 2015). Recognition by the farming industries and Government of financial and trade impacts of bTB led to the establishment of the Animal
J.A. Sinclair et al. / Preventive Veterinary Medicine 127 (2016) 94–99
Health Board (AHB) and the deployment of a National Pest Management Strategy (later called Plan, NPMP) under the Biosecurity Act 1993. The AHB employed a three-pronged approach to control bovine TB. This consisted of (i) a national surveillance system of TB test and slaughter in domestic cattle and deer herds and inspection of all carcases at slaughter premises, (ii) movement restrictions out of areas with high infected wildlife risk and from infected herds, and (iii) possum control to manage TB in infected wildlife populations nationally. As a result of these control strategies, the period prevalence of TB herds has decreased from a peak of 3.87% as at 30th June 1995 to 0.16% at 30th June 2015 (Anon., 2015a). The NPMP was reviewed in 2011, and officially stated for the first time that eradication of bTB from New Zealand was an objective. Translocation of cattle now accounts for a higher proportion of bTB herd infection in New Zealand than it has in the past, due to both a reduction in the extent of infected wildlife populations over 20 years of possum control, and an increasing volume and geographical extent of stock movements (Mark Stevenson, unpublished data). The national surveillance of cattle herds in New Zealand has been based on the use of the intradermal caudal fold tuberculin test (CFT). When bTB herd prevalence was high, skin test positive animals were routinely slaughtered as reactors. The ␥-IFN test became commercially available as BovigamTM in 2001 (Wood and Jones, 2001) and has been used in New Zealand since the early 2000s. As the disease prevalence dropped, serial ␥-IFN testing began to be routinely applied in skin test positive animals in low-risk herds. This decreased the wastage from false positive skin tests and maintained farmer acceptance of the scheme, although by using two tests in series, some loss of sensitivity was to be expected, to achieve an increase in test specificity. The imperfect accuracy of the currently available tests for bTB is universally acknowledged (de la Rua-Domenech et al., 2006). As the number of infected herds decreased in New Zealand, the parallel ␥-IFN test has been increasingly used in high disease prevalence situations, such as in infected herds, to improve the overall testing sensitivity. The parallel ␥-IFN test is applied only to animals that are negative to the CFT (Anon., 2011). Anecdotal evidence from New Zealand in the field since 2008 has shown that parallel ␥-IFN testing has found residual infection in herds that had no infection detected by CFT. Based on the CFT alone, these herds would have been given a clear status and had their movement restrictions revoked, allowing freedom to trade. The increase in sensitivity offered by the parallel CFT and ␥-IFN test combination allows the removal of some infected animals that are not responsive to skin testing. The ␥-IFN test is capable of detecting bTB infection earlier than the CFT test (Wood and Jones, 2001). There are major industry-wide benefits to achieving the highest possible sensitivity of detection before a herd is free to trade. In the later stages of an eradication programme, maximum sensitivity of detection is required in infected herds, even though it comes at the cost of reduction in specificity (Buddle et al., 2001). In order to have movement restrictions revoked, a herd must complete two clear whole herd tests (WHT) using the CFT on all animals over six weeks of age, with at least six months between the two tests. A clear test is defined as one where no confirmed TB infected animals are identified. The aim is not to clear herd status as fast as possible, but to ensure that residual infection poses less risk to the herd in the future, and to other herds through animal trading. Following the 2011 NPMP review, dairy and breeding beef herds in New Zealand are now additionally required to complete a parallel ␥-IFN test of all animals in the infected cohort at the final whole herd test before having their infected herd status revoked and movement restrictions lifted. This policy change would have resulted in a dramatic rise in cost if applied to all of the infected herds at that time, so was rolled out in herds based on risk of
95
recrudescence; therefore not all eligible herds actually completed the parallel testing. Herds were considered to be at highest risk of recrudescence if they had multiple prior episodes of bTB or had more than one confirmed infected animal found at the latest episode (Dawson et al., 2014), and herds meeting these criteria were given preference for parallel testing at the final whole herd test. Use of parallel ␥-IFN testing occurs in three main ways in New Zealand. Firstly, the parallel ␥-IFN assay has been used in skin test negative animals in infected herds at the first WHT after the detection of an infected episode, WHT(1). The objective is to increase the probability of early detection of diseased animals within the herd and thus shorten the time the herd is deemed to be infected. Herds are eligible to receive parallel ␥-IFN testing when more than one confirmed infected animal has been found at breakdown, whether by routine slaughter inspection or by tuberculin skin testing. Secondly, parallel ␥-IFN is also used to improve confidence that the herd is truly clear of bTB at the final WHT, WHT(F), before movement restrictions are lifted. Finally, parallel ␥-IFN is used to confirm the absence of infection in skin test negative animals from lowrisk cohorts (e.g. dairy heifers) before movement out of an infected herd to another property, a permitted activity under stringent conditions. As the new policy has been in place for four years, the aim of this retrospective observational study was to evaluate the effectiveness of the use of parallel ␥-IFN in New Zealand’s bTB eradication scheme.
2. Materials and methods A list of all herds that had an infected status between 1 July 2011 and 1 September 2015 was obtained from the OSPRI1 disease management databases (DMIS—Disease Management Information System and DMS—Disease Management System), maintained by TBfree New Zealand (previously the Animal Health Board). There were 289 herd numbers eligible for consideration. Fifty herds were excluded from the study. These included deer and beef meat production herds where parallel ␥-IFN testing was not used. The parallel ␥-IFN testing is only used on beef breeding and dairy herds. In addition herds were not included where bTB was not confirmed by culture. Ninety-nine beef breeding and 140 dairy herds were eligible for the study, making a total of 239 study herds. Data were extracted from DMS to gather the following information: 1. Whether the herd had one or more prior bTB breakdowns in its history; 2. For the first WHT at or after detection of bTB, WHT(1)* : a Number of skin test positive animals slaughtered b Number of confirmed bTB cases in the skin test positive animals slaughtered c Whether a parallel ␥-IFN of the infected cohort was undertaken d Number of parallel ␥-IFN reactors slaughtered e Number of confirmed bTB cases in the parallel ␥-IFN positive animals slaughtered; 3. For the final WHT (a minimum of six months after the herd had completed a WHT where there had been no confirmed bTB), WHT(F)** :
1 OSPRI, Operational Solutions for Primary Industries, the entity that manages the TBfree and NAIT (National Animal Identification and Tracing) schemes in New Zealand.
96
J.A. Sinclair et al. / Preventive Veterinary Medicine 127 (2016) 94–99
Table 1 Classification of the 239 study herds by risk factors for undetected infection.
bTB history in herd No bTB history in herd Total
Multiple bTB cases
Single bTB case
Total
24 34 58
119 62 181
143 96 239
a Number of skin test positive animals slaughtered b Number of confirmed bTB cases in the skin test positive animals slaughtered c Whether a parallel ␥-IFN of the infected cohort was undertaken d Number of parallel ␥-IFN reactors slaughtered e Number of confirmed bTB cases in the parallel ␥-IFN positive animals slaughtered.
Table 2 Number of reactors slaughtered at WHT(1), and number and proportion of reactors found to have lesions of bTB at slaughter for the 239 dairy and beef breeding herds in the study. Number of reactors with bTB lesions at slaughter
Percentage
653 430
380 160
58.2% 37.2%
1083
540
49.9%
Reactors slaughtered
CFT testing Parallel ␥-IFN (CFT negative) testing Total
Table 3 Number and percentage of 239 study herds where parallel ␥-IFN was used at WHT(1), reactors slaughtered and bTB lesions found.
* If
the breakdown was detected at a WHT, this became WHT(1). If the breakdown was detected by meat inspection, the WHT(1) was completed subsequently. ** Thirty-one herds were still on infected movement control as at 1 September 2015, and had not been eligible for a WHT(F). The definition of WHT(F) is a test that occurs at least six months after a clear whole herd CFT in which there are no confirmed cases of bTB, where a second clear test would qualify the herd for revocation of movement restrictions. If bTB was found at this test, it was still termed WHT(F) for the purposes of this study, although the herd remained on an infected status. 2.1. Risk categories of study herds Ninety-three out of the 239 study herds actually had a parallel test applied at WHT(F). Thirty-one herds were not yet eligible for a WHT(F) by the end date of the study because they had not yet completed a clear CFT of the whole herd. The remaining 208 herds were eligible under the 2011 policy, but the test was preferentially rolled out in the highest risk herds due to the prohibitive cost of parallel testing all herds at WHT(F). The decision on whether to apply the parallel test was discretionary and also involved a consideration of cost and logistics. Ten per cent (24) of the 239 study herds had both a history of bTB in the herd and multiple bTB animals found during their current episode (Table 1). These were considered the highest risk for undetected infection and therefore were given highest priority for parallel testing. Twenty-six per cent (62) of the herds had no bTB history and only a single confirmed infected animal found before the final whole herd testing. These herds were considered low risk. The remaining 64% (153) of herds had only one risk factor: either a bTB history or multiple infected animals in the episode. Sixty-seven of the 93 herds parallel tested had at least one of the risk factors for undetected infection, and of these, 16 herds (80% of the 20 test eligible herds with both risk factors) were parallel tested at WHT(F). 2.2. The caudal fold test (CFT) The CFT was undertaken by intradermally injecting a 0.1 ml volume of bovine PPD (3000 IU/0.1 ml dose; Prionics, Lelystad, The Netherlands) in the caudal fold of the tail. The caudal fold was inspected 72 h later and any detectable swelling was regarded as a positive response. 2.3. The -IFN test Heparinised blood samples were collected from the cattle at 0–30 days following the CFT test and the blood cultures were set up the day following collection, but within 30 h of collection. Blood
Parallel ␥-IFN test used Reactors slaughtered bTB lesions found in reactors
Number of herds
Percentage
64 49 32
26.8% of 239 herds 76.6% of 64 herds 65.3% of 49 herds
samples (1.0 ml) were dispersed into wells of a 24-well plate and preservative-free bovine PPD prepared from M. bovis or PPD prepared from M. avium, avian PPD (24 g/ml final concentration; Prionics, Schlieren-Zurich, Switzerland), or phosphate-buffered saline (negative control) was added. After incubation at 37 ◦ C for 16–24 h, the plasma supernatants were harvested and their ␥-IFN levels measured using a sandwich ELISA kit (Prionics). For parallel interpretation, a Bovine PPD optical density (OD) minus Avian PPD OD of >0.07 was deemed “high risk”; “medium risk” between OD of 0.04–0.07, and “low risk” OD <0.04. The decision to designate animals as reactors depended on the judgement of the veterinarian in charge of the case, and the variable cut-off (Buddle et al., 2015) allowed the slaughter of ‘cuts’ of animals from the highest OD down, with subsequent slaughter decisions dependent on the previous post mortem results. Therefore the same reactor cut-off was not applied in every herd, and the retrospective nature of this study did not allow us to determine the actual cut-off value applied, or to compare the effects of different cut-off values on the lesion findings. 3. Results 3.1. First whole herd test CFT testing at WHT(1) detected 653 reactors over the 239 herds. At slaughter, 380 of these reactors (58.2%) had lesions of bTB detected (Table 2). Parallel ␥-IFN testing was only undertaken on animals which were negative in the CFT and was used in 64 (26.8%) of the 239 eligible herds at WHT(1). Of these herds, a total of 430 reactors were found over 49 (76.5%) herds at ␥-IFN testing. When these reactors were slaughtered, 160 animals from 32 herds (37.2% of reactors and 50% of herds) were found to have lesions of bTB (Tables 2 and 3). Of dairy herds, 46 of 140 had parallel ␥-IFN testing applied and reactors were found in 33 of these 46 herds (71.2%), whereas in beef breeding herds, 18 out of 99 herds were parallel ␥-IFN tested, with reactors found in 16 of 18 herds (88.9%). 3.2. Final whole herd test Herds were eligible for a WHT(F) if they had completed a whole herd CFT with no confirmed bTB animals found at least six months previously, and 208 of the 239 study herds were eligible for a WHT(F). At WHT(F), CFT testing found 74 reactors in the 208 eligible herds, and 10 of these reactors had confirmed bTB lesions at
J.A. Sinclair et al. / Preventive Veterinary Medicine 127 (2016) 94–99 Table 4 Number of reactors slaughtered at WHT(F), and number and proportion of reactors found to have lesions of bTB at slaughter for the 239 dairy and beef breeding herds in the study. Reactors Number of slaughtered reactors with bTB lesions at slaughter CFT testing 74 Parallel ␥-IFN (CFT negative) testing 122 Total 196
10 21 31
13.5% 17.2% 15.8%
Number of herds
Percentage
93 49 13
44.7% of 208 herds 53.9% of 93 herds 26.5% of 49 herds
Table 6 Number of herds with reactors slaughtered and bTB cases, number of reactors and bTB cases per herd for the 38 study herds that had both WHT(1) and WHT(F) applied. WHT(1) Herds with reactors slaughtered Herds with bTB cases Average number of reactors slaughtered per herd Average number of bTB cases per herd
31 22 6.8 2.5
119 (83%) and 24 (17%) respectively (Table 1). A statistically significant association was found between bTB history or no prior history, and the finding of one or multiple confirmed cases (X2 = 9.86, df = 1, p = 0.0017).
Percentage
Table 5 Number and percentage of 208 eligible study herds where parallel ␥-IFN used at WHT(F), reactors slaughtered and bTB lesions found.
Parallel ␥-IFN test used Reactors slaughtered bTB lesions found in reactors
97
WHT(F) 17 4 1.2 0.2
slaughter, thereby indicating a failure to eliminate bTB and a continuation of the infected status and movement restrictions. For the 208 herds, a parallel ␥-IFN was undertaken in 93 (44.7%) herds and 49 of these (52.7%) had reactors found. Of these 49 herds, 13 (26.5%) had a total of 21 reactors with lesions of bTB detected at slaughter (Tables 4 and 5). These 13 herds remained on movement restriction and an infected status. The percentage of dairy and beef breeding herds with reactors found at parallel testing was similar (53.1% of 64 dairy herds and 51.7% of 29 beef herds). Eleven of the 13 herds in which parallel ␥-IFN reactors with confirmed bTB were found at WHT(F) had no infection identified by CFT, and therefore would have had movement restrictions revoked if no parallel testing had been done. Table 4 shows the number of CFT test and parallel reactors found at WHT(F) and the percentage of bTB confirmed lesions in these groups. 3.3. Herds receiving parallel testing at both WHT(1) and WHT(F) Thirty-eight herds had received a parallel test at both WHT(1) and WHT(F) (Table 6). In WHT(1), a mean of 6.8 reactors were found by the parallel, dropping to 1.2 at the WHT(F). The mean number of reactors found with bTB lesions at slaughter correspondingly fell between WHT(1) and WHT(F) from 2.5 to 0.2 (Table 6). 3.4. Risk factors for undetected infection Sixty per cent of the 239 study herds had a history of one or more bTB breakdowns in the herd prior to the study period. The test interval before detection of bTB and length of time since prior bTB breakdown was not determined from the data available. Twentyfour per cent of the study herds had more than one bTB confirmed case found in the current breakdown. Ten per cent of the herds had both multiple bTB cases and at least one prior breakdown (Table 1). Of herds that had no bTB history, 62 (65%) had only a single bTB animal found in the current breakdown, and 34 (35%) had multiple confirmed cases. For herds with a bTB history, these figures were
4. Discussion New Zealand’s TB control since the mid 1990’s has been very effective and based mainly on the CFT, with extra testing and movement restrictions from the known areas of high risk for infected possums (Vector Risk Areas, VRAs). After twenty years of control of possums there has been a change in the nature of breakdowns with a greater proportion occurring in the Vector Free Areas (VFAs) through the trading of stock. As the infected herd prevalence has dropped, more effort is now needed to increase the sensitivity of detection of disease in infected herds. Sixty per cent of study herds had a history of at least one prior bTB episode and period of revocation of movement restrictions prior to the study breakdown. In many cases the most recent bTB breakdown may represent residual infection from an historic breakdown. The length of time since the most recent breakdown, and testing interval before detection was not determined from the data but is worthy of investigation in its own right. Singleton breakdowns were more likely in herds that had experienced a prior breakdown, and reasons may include the earlier detection of disease as a result of more intensive testing post-clearance, or the detection of a single residual case post clearance. Study herds that had a bTB history were significantly more likely to have only a single bTB case in the study breakdown. The results of this study show that many of these repeat breakdowns could have been prevented by maximising test sensitivity before clearance. Parallel ␥-IFN testing helps achieve this sensitivity and is now more affordable than in the past, due to the lower number of herds involved. The test is capable of detecting infection at an earlier stage than is possible with the skin tuberculin test (Buddle et al., 1995; Wood et al., 1991; Wood and Jones, 2001) and therefore may detect early infected animals that would otherwise remain as a source of ongoing transmission within the herd (Houlihan et al., 2008). Eleven of the 93 herds in our study in which a parallel ␥IFN test was used failed their WHT(F) because confirmed infection was found at the parallel test, despite the herd being clear to CFT. This study has demonstrated the efficacy of the parallel ␥-IFN assay under New Zealand field conditions for detecting many additional cases of bTB infection in infected herds that could not be found by intradermal tuberculin testing alone. The policy of parallel ␥-IFN testing in breeding or dairy herds as part of final clearance testing applied since 2011 has therefore been successful in preventing some potential movement-related or recrudescent breakdowns. The benefit to New Zealand of parallel testing as part of its eradication plan is likely to significantly outweigh the cost of the supplementary testing. Each additional herd breakdown prevented is estimated to save a minimum of NZD 20,000 for a single animal breakdown. Two extensive multiple animal breakdowns caused by the inadvertent purchase of infected animals in the North Island of New Zealand in 2013 cost an estimated NZD400,000 each in testing costs and reactor compensation before the herds were returned to a clear status, not including costs to the farmer from lost revenue and additional labour (J. Sinclair, unpublished data). The number of additional breakdowns prevented by increased testing sensitivity before clearance cannot be quantified, but undetected infection at clearance puts at least the newly cleared herd at risk, and potentially many herds through trading. In the past six years, two cases of milk-borne spread to young calves from cows with undetected udder bTB infection resulted in infection being
98
J.A. Sinclair et al. / Preventive Veterinary Medicine 127 (2016) 94–99
later found in an additional seven and 10 herds respectively (J. Sinclair, unpublished data). Reasons for the parallel test being selectively applied during the roll-out period even in high-risk herds in this study were herd size and logistics. Dairy herds in New Zealand have an average of 419 cows, with 12% of herds comprised of over 750 cows (Anon., 2015b). A whole herd parallel test is much more financially and logistically viable in a small to medium sized herd. Four eligible herds in the study had both a recent bTB history and multiple bTB animals found, yet did not receive parallel testing at WHT(F). Two of these herds were very large (5122 and 4145 animals). One had bTB found at CFT testing in their WHT(F) and therefore had already failed, and the last had no identifiable reasons for not completing a parallel at WHT(F). Where the parallel ␥-IFN test was applied at WHT(F), the lesion rate in the reactors to this test was higher than the lesion rate in CFT positives. This observation gives rise to concern that herds that did not receive parallel testing may have had residual infection found, if they had been tested. As New Zealand’s bTB eradication scheme enters a more rigorously targeted phase, a team approach to case management will replace the individual judgement-based approach of the past. Flexibility in test application and interpretation is still important, as this enables experiential learning. A team focus will help achieve transparency and the long-term management of risk from undetected infection to the country as a whole. Additionally, with infected herd numbers now below 40 nationally, parallel testing of even very large herds at clearance is financially viable, and the aim is to apply these tests in every infected herd. Where herds had received parallel testing at both the WHT(1) and a WHT(F), markedly fewer reactors were slaughtered and less infection found at WHT(F) than at WHT(1). This gives confidence that the disease control measures being applied in infected herds are effective at identifying disease and preventing its spread. The sensitivity of the ␥-IFN assay has been estimated to be between 73 and 100%, with a median sensitivity of 87.6% (de la Rua-Domenech et al., 2006). Under New Zealand conditions, the sensitivity was estimated to be 85% with a specificity of 93% for animals that had reacted positively in the CFT (Ryan et al., 2000), with the sensitivity considered equivalent or greater to that of the CFT. The sensitivity and specificity of the ␥-IFN test is dependent on the cut-off value used for positive interpretation (Goodchild and Clifton-Hadley, 2001; Wood and Jones, 2001; de la Rua-Domenech et al., 2006). Lowering the cut-off from OD > 0.1 to OD > 0.04 increased the sensitivity from 94% to 98% in one trial (Buddle et al., 2001). In the past, policy in New Zealand has allowed discretionary application of testing and cut-off values within guidelines. This has been necessary due to herd differences in length of time infected, presence of nonspecific reactions and within-herd transmission, and has enabled empirical learning. The discretionary nature of the OD cut-off used for disease management in infected herds is problematic in the interpretation of the results of this study. No conclusions can be drawn about the effect on the lesion finding rate of reactors of using a lower cut-off. Future research could examine the correlation between OD reading and post-mortem or culture positive rate. The ␥-IFN assay has been used in other countries with parallel interpretation and a lower cut-off to improve sensitivity of detection in high risk herds (Buddle et al., 2001; Wood et al., 1991; Wood and Jones, 2001; de la Rua-Domenech et al., 2006), especially for clearance testing, where a decision must be made on whether to release a herd from movement restrictions. In infected herds, the use of two tests capable of detecting different subpopulations of infected animals will improve the rate of resolution of a bTB episode and should also reduce the chance of recurrent infection in the absence of movement or vector-related reinfection (de la Rua-
Domenech et al., 2006; Ameni et al., 2010). Skin and ␥-IFN testing in parallel is thought to be 20% more sensitive than skin testing alone (Wood et al., 1991; Whipple et al., 1995; González Llamazares et al., 1999) and is estimated to have a sensitivity of 93% (Gormley et al., 2006). In this study, half of all ␥-IFN reactors at WHT(1) had no gross evidence of infection. The absence of lesions at slaughter is not proof of the absence of infection, as up to 10% of reactors with no visible lesions were shown to be culture positive for M.bovis in one study (Corner, 1994); therefore infection may be missed despite critical examination of prescribed body sites. Reactors slaughtered with no visible lesions may be accounted for by (i) early infection, where the immune response has been activated but lesions are too small to detect at post-mortem examination (Menzies and Neill, 2000); (ii) latent infection, where the organism is present in a dormant state without disease (Cassidy et al., 2001); (iii) cross-reactivity with other mycobacterial organisms (de la RuaDomenech et al., 2006) or (iv) the lesions were not detected in reactor post-mortem (Corner, 1994). Several studies have shown that parallel ␥-IFN reactors not slaughtered were significantly more likely to be later identified as infected with bTB. A longitudinal study from Ireland followed ␥IFN reactors remaining in their herds for a further 155 days after the initial assay. At the next two skin tests these reactors were 7–9 times more likely than their ␥-IFN negative herd-mates to be skin test positive (Gormley et al., 2006). The odds ratio of ␥-IFN reactors subsequently becoming skin test positive compared with non-reactors was 7–20 in another study (Collins, 2000). When 100 animals were followed after an initial ␥-IFN positive result, they were significantly more likely to be identified as having bTB at subsequent testing at a rate ratio of 17.18 compared to the animals negative to both the skin and ␥-IFN test (Welsh et al., 2008). A Northern Ireland study of 1107 ␥-IFN positive reactors that were not slaughtered showed that these animals were 2.31 times more likely than ␥-IFN negative animals to become reactors to the intradermal tuberculin test in the following five years (Lahuerta-Marin et al., 2015). In New Zealand, the presence of reactor animals without confirmed bTB at a final clearance test increased the hazard of a future breakdown by 2.8 times in the first two years after clearance (Dawson et al., 2014). The removal of these potentially early infected animals negates the possibility that they may pose a future transmission risk to their herd-mates or to other herds postmovement. The slaughter of some false positive reactors in infected herds due to the lower specificity of the test combination of CFT and ␥-IFN is a necessary price to pay for the eradication of the disease, and helps to prevent future losses. In confirmed infected herds and in areas with endemic wildlife disease, reactors are highly predictive of infection (de la RuaDomenech et al., 2006). Parallel reactors may indicate ongoing within-herd transmission and may be the only indicator of the presence of an anergic shedder within the herd. New Zealand’s bTB surveillance scheme has relied upon slaughterhouse surveillance and annual to triennial caudal fold tuberculin testing in eligible stock, depending on their proximity to vector risk. Both of these elements have imperfect sensitivity. This study has shown that in bTB infected herds, CFT failed to identify a significant proportion of the infected animals in some herds. Selection pressure may have resulted in the emergence of strains that are less detectable by CFT. The sensitivity of tuberculin testing is dose and site dependent (de Kantor et al., 1984). To improve sensitivity, mid cervical testing and/or increasing the tuberculin dose could be considered in infected herds in New Zealand to support the aim of eradication. If the sensitivity of the CFT is as low as these findings suggest, it is difficult to explain how the surveillance scheme been so successful in reducing the burden of infected herds. However, it is easier to
J.A. Sinclair et al. / Preventive Veterinary Medicine 127 (2016) 94–99
find infection with the CFT at a herd level when infection is more prevalent, and the CFT has therefore been useful in reducing the bulk of infection when applied over large numbers of herds. This study has confirmed that the CFT has poor sensitivity at the animal level. Increased use of parallel testing in the future, perhaps with adjustments to the use of tuberculin, will undoubtedly contribute to ongoing success of New Zealand’s bTB eradication scheme by improving animal-level sensitivity. Our study has confirmed under field conditions in New Zealand that the greater test sensitivity offered by the combination of CFT and parallel ␥-IFN enables the detection of many more cases of within-herd infection at the first test following detection of bTB in a herd than with skin tuberculin testing alone. Thus the combination of tests enables more complete removal of infected animals and therefore prevention of ongoing transmission, resulting in a more rapid resolution of within-herd infection. Parallel ␥-IFN before revocation of movement restrictions gives greater confidence that the herd will remain clear of disease in the long term (Gormley et al., 2006) providing economic benefits to farmers and cattle industries. Acknowledgements The authors thank the AsureQuality and VetEnt technicians for undertaking the CFT and collection of the blood samples for the ␥IFN, and the AgResearch bTB diagnostic team for the ␥-IFN testing and culture of M. bovis from slaughtered animals. References Ameni, G., Aseffa, A., Hewinson, G., Vordermeier, M., 2010. Comparison of different testing schemes to increase the detection Mycobacterium bovis infection in Ethiopian cattle. Trop. Anim. Health Prod. 42, 375–383. Anon, 2011. National bovine tuberculosis pest management strategy. In: National Operational Plan Part B: Operational Policies. Animal Health Board (now known as OSPRI), Wellington, New Zealand. Anon., Annual Report. OSPRI, Wellington, New Zealand, 2015. Anon, 2015b. New Zealand Dairy Statistics 2014–15. Livestock Improvement Corporation and DairyNZ Limited, 52 pp. Buddle, B.M., de Lisle, G.W., Pferrer, A., Aldwell, F.E., 1995. Immunological responses and protection against Mycobacterium bovis in calves vaccinated with a low dose of BCG. Vaccine 13, 1123–1130. Buddle, B.M., Ryan, T.J., Pollock, J.M., Andersen, P., De Lisle, G.W., 2001. Use of ESAT-6 in the interferon-test for diagnosis of bovine tuberculosis following skin testing. Vet. Microbiol. 80, 37–46. Buddle, B.M., de Lisle, G.W., Griffin, J.F.T., Hutchings, S.A., 2015. Epidemiology, diagnostics, and management of bovine tuberculosis in domestic cattle and deer in New Zealand in the face of a wildlife reservoir. N. Z. Vet. J. 63, 19–67. Cassidy, J.P., Bryson, D.G., Gutierrez Cancela, M.M., Forster, F., Pollock, J.M., Neill, S.D., 2001. Lymphocyte subtypes in experimentally induced early-stage bovine tuberculous lesions. J. Comp. Pathol. 124, 46–51.
99
Collins, J.D., 2000. A longitudinal study of cattle found positive to the interferon-gamma assay for Mycobacterium bovis infection. Soc. Vet. Epidem. Econ. Proc., 1263–1265. Corner, L.A., 1994. Post mortem diagnosis of Mycobacterium bovis infection in cattle. Vet. Microbiol. 40, 53–63. Dawson, K.L., Stevenson, M.A., Sinclair, J.A., Bosson, M.A., 2014. Recurrent bovine tuberculosis in New Zealand cattle and deer herds, 2006–2010. Epidemiol. Infect. 142, 2065–2074. de la Rua-Domenech, R., Goodchild, A.T., Vordermeier, H.M., Hewinson, R.G., Christiansen, K.H., Clifton-Hadley, R.S., 2006. Ante mortem diagnosis of tuberculosis in cattle: a review of the tuberculin tests, -interferon assay and other ancillary diagnostic techniques. Res. Vet. Sci. 81, 190–210. de Kantor, I.N., Odeon, A.C., Steffan, P.E., Auza, M.J., Madrid, C.R., Marchevsky, N., 1984. Sensitivity of the cervical and the caudal fold tuberculin tests with Mycobacterium bovis in infected cattle of Argentina. Rev. Sci. Tech. OIE 3, 137–150. Ekdahl, M.O., Smith, B.L., Money, D.F., 1970. Tuberculosis in some wild and feral animals in New Zealand. N. Z. Vet. J. 18, 44–45. González Llamazares, O.R., Gutiérrez Martín, C.B., Aranaz Martín, A., Liébana Criado, E., Domínguez Rodríguez, L., Rodríguez Ferri, E.F., 1999. Comparison of different methods for diagnosis of bovine tuberculosis from tuberculin-or interferon-reacting cattle in Spain. J. Appl. Microbiol. 87, 465–471. Goodchild, A.V., Clifton-Hadley, R.S., 2001. Cattle-to-cattle transmission of Mycobacterium bovis. Tuberculosis 81, 23–41. Gormley, E., Doyle, M.B., Fitzsimons, T., McGill, K., Collins, J.D., 2006. Diagnosis of Mycobacterium bovis infection in cattle by use of the gamma-interferon (Bovigam) assay. Vet. Microbiol. 112, 171–179. Houlihan, M.G., Dixon, F.W., Page, N.A., 2008. Outbreak of bovine tuberculosis featuring anergy to the skin test, udder lesions and milkborne disease in young calves. Vet. Rec. 163, 357–361. Lahuerta-Marin, A., Gallagher, M., McBride, S., Skuce, R., Menzies, F., McNair, J., McDowell, S.W.J., Byrne, A.W., 2015. Should they stay, or should they go? Relative future risk of bovine tuberculosis for interferon-gamma test-positive cattle left on farms. Vet. Res. 46. Livingstone, P.G., Hancox, N., Nugent, G., de Lisle, G.W., 2015. Toward eradication: the effect of Mycobacterium bovis infection in wildlife on the evolution and future direction of bovine tuberculosis management in New Zealand. N. Z. Vet. J. 63, 4–18. Menzies, F.D., Neill, S.D., 2000. Cattle-to-cattle transmission of bovine tuberculosis. Vet. J. 160, 92–106. Ryan, T.J., Buddle, B.M., De Lisle, G.W., 2000. An evaluation of the gamma interferon test for detecting bovine tuberculosis in cattle 8–28 days after tuberculin skin testing. Res. Vet. Sci. 69, 57–61. Welsh, M.D., McNair, J., McDowell, S.W.J., Buchanan, J., Hill, R., McBride, S., 2008. The Northern Ireland interferon-gamma (IFN-g) testing programme. Cattle Practice 16, 136–139. Whipple, D.L., Bolin, C.A., Davis, A.J., Jarnagin, J.L., Johnson, D.C., Nabors, R.S., Payeur, J.B., Saari, D.A., Wilson, A.J., Wolf, M.M., 1995. Comparison of the sensitivity of the caudal fold skin test and a commercial gamma-interferon assay for diagnosis of bovine tuberculosis. Am. J. Vet. Res. 56, 415–419. Wood, P.R., Jones, S.L., 2001. BOVIGAMTM : an in vitro cellular diagnostic test for bovine tuberculosis. Tuberculosis 81, 147–155. Wood, P.R., Corner, L.A., Rothel, J.S., Baldock, C., Jones, S.L., Cousins, D.B., McCormick, B.S., Francis, B.R., Creeper, J., Tweddle, N.E., 1991. Field comparison of the interferon-gamma assay and the intradermal tuberculin test for the diagnosis of bovine tuberculosis. Aust. Vet. J. 68, 286–290.
Contents lists available at ScienceDirect
Preventive Veterinary Medicine journal homepage: www.elsevier.com/locate/prevetmed
The effectiveness of parallel gamma-interferon testing in New Zealand’s bovine tuberculosis eradication programme J.A. Sinclair a,∗ , K.L. Dawson a , B.M. Buddle b a b
OSPRI New Zealand, 50 Church Rd., Hamilton 3200, New Zealand AgResearch, Hopkirk Research Institute, Palmerston North 4442, New Zealand
a r t i c l e
i n f o
Article history: Received 29 November 2015 Received in revised form 25 February 2016 Accepted 31 March 2016 Keywords: Bovine tuberculosis Eradication Sensitivity Gamma-interferon
a b s t r a c t In bovine tuberculosis (bTB) eradication programmes, especially where prevalence is low, sensitivity of testing in infected herds must be maximised to reduce the possibility of recrudescence of prior infection and the risk to other herds via animal movement. The gamma-interferon (␥-IFN) assay applied in parallel with intradermal tuberculin testing has been shown to increase test sensitivity. The aim of this work was to substantiate this effect in the field. A retrospective observational study was conducted on 239 New Zealand cattle breeding and dairy herds with bTB infection between 1 July 2011 and 1 September 2015 to evaluate the outcomes of new policy introduced in 2011. The investigation defined the number and proportion of reactors (animals testing positive and slaughtered) found with lesions of bTB in intradermal caudal fold testing (CFT) and parallel ␥-IFN testing, at the breakdown test or first whole herd test after breakdown, WHT(1), and at the final or projected final whole herd test, WHT(F). Parallel ␥-IFN testing was used in 26.8% of the 239 herds at WHT(1), and 430 animals in 49 herds were deemed reactors. One hundred and sixty (37.2%) of these reactors from 32 herds were found to have bTB lesions, despite having been negative to caudal fold testing. These 160 infected animals accounted for 29.6% of all infection found at WHT(1). At WHT(F), parallel ␥-IFN testing was conducted on 93 herds and detected a total of 122 reactors in 49 herds, in addition to those found by CFT. Twenty-one of these reactors, from 13 herds, had bTB lesions at slaughter, accounting for 67.7% of all reactors found with bTB at WHT(F). Eleven of these 13 herds would have had their movement restrictions revoked based on a negative herd CFT alone, and could potentially have caused outward transmission of bTB to other herds, as well as experiencing recrudescent breakdowns. We conclude that ␥-IFN testing in infected herds, in parallel with intradermal tuberculin testing, is a valuable tool in a bTB eradication programme, as it enables higher test sensitivity at both herd and animal level. The use of the ␥-IFN test over a risk cohort early in a breakdown assists in removal of early infection and some cases of anergy to intradermal tuberculin testing. Parallel ␥-IFN with compulsory slaughter of reactors should be considered in breeding and dairy herds in conjunction with tuberculin testing before movement control is revoked, and will assist in achieving TB freedom on a herd level and nationally. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Bovine tuberculosis was introduced to New Zealand by cattle imports with settlers in the mid-1800s. By the 1950s, the disease had become widespread in both the North and South Islands. Voluntary test-and-slaughter schemes failed to eradicate bTB, and in 1967 infection was identified in Australian brushtail possums,
∗ Corresponding author. E-mail address: [email protected] (J.A. Sinclair). http://dx.doi.org/10.1016/j.prevetmed.2016.03.020 0167-5877/© 2016 Elsevier B.V. All rights reserved.
Trichosurus vulpecula (Ekdahl et al., 1970). Possums were subsequently found to be the principal wildlife reservoir for M. bovis in New Zealand. Control of possums was undertaken in the 1970s but funding was interrupted in the 1980s, resulting in a dramatic geographical spread of wild animal infection. Annual infected domestic herd numbers peaked at 1684 infected herds at 30 June 1995 (data held by OSPRI New Zealand) and by 2004, bTB had become endemic in wild animal populations over an estimated 39% of New Zealand’s land area (Livingstone et al., 2015). Recognition by the farming industries and Government of financial and trade impacts of bTB led to the establishment of the Animal
J.A. Sinclair et al. / Preventive Veterinary Medicine 127 (2016) 94–99
Health Board (AHB) and the deployment of a National Pest Management Strategy (later called Plan, NPMP) under the Biosecurity Act 1993. The AHB employed a three-pronged approach to control bovine TB. This consisted of (i) a national surveillance system of TB test and slaughter in domestic cattle and deer herds and inspection of all carcases at slaughter premises, (ii) movement restrictions out of areas with high infected wildlife risk and from infected herds, and (iii) possum control to manage TB in infected wildlife populations nationally. As a result of these control strategies, the period prevalence of TB herds has decreased from a peak of 3.87% as at 30th June 1995 to 0.16% at 30th June 2015 (Anon., 2015a). The NPMP was reviewed in 2011, and officially stated for the first time that eradication of bTB from New Zealand was an objective. Translocation of cattle now accounts for a higher proportion of bTB herd infection in New Zealand than it has in the past, due to both a reduction in the extent of infected wildlife populations over 20 years of possum control, and an increasing volume and geographical extent of stock movements (Mark Stevenson, unpublished data). The national surveillance of cattle herds in New Zealand has been based on the use of the intradermal caudal fold tuberculin test (CFT). When bTB herd prevalence was high, skin test positive animals were routinely slaughtered as reactors. The ␥-IFN test became commercially available as BovigamTM in 2001 (Wood and Jones, 2001) and has been used in New Zealand since the early 2000s. As the disease prevalence dropped, serial ␥-IFN testing began to be routinely applied in skin test positive animals in low-risk herds. This decreased the wastage from false positive skin tests and maintained farmer acceptance of the scheme, although by using two tests in series, some loss of sensitivity was to be expected, to achieve an increase in test specificity. The imperfect accuracy of the currently available tests for bTB is universally acknowledged (de la Rua-Domenech et al., 2006). As the number of infected herds decreased in New Zealand, the parallel ␥-IFN test has been increasingly used in high disease prevalence situations, such as in infected herds, to improve the overall testing sensitivity. The parallel ␥-IFN test is applied only to animals that are negative to the CFT (Anon., 2011). Anecdotal evidence from New Zealand in the field since 2008 has shown that parallel ␥-IFN testing has found residual infection in herds that had no infection detected by CFT. Based on the CFT alone, these herds would have been given a clear status and had their movement restrictions revoked, allowing freedom to trade. The increase in sensitivity offered by the parallel CFT and ␥-IFN test combination allows the removal of some infected animals that are not responsive to skin testing. The ␥-IFN test is capable of detecting bTB infection earlier than the CFT test (Wood and Jones, 2001). There are major industry-wide benefits to achieving the highest possible sensitivity of detection before a herd is free to trade. In the later stages of an eradication programme, maximum sensitivity of detection is required in infected herds, even though it comes at the cost of reduction in specificity (Buddle et al., 2001). In order to have movement restrictions revoked, a herd must complete two clear whole herd tests (WHT) using the CFT on all animals over six weeks of age, with at least six months between the two tests. A clear test is defined as one where no confirmed TB infected animals are identified. The aim is not to clear herd status as fast as possible, but to ensure that residual infection poses less risk to the herd in the future, and to other herds through animal trading. Following the 2011 NPMP review, dairy and breeding beef herds in New Zealand are now additionally required to complete a parallel ␥-IFN test of all animals in the infected cohort at the final whole herd test before having their infected herd status revoked and movement restrictions lifted. This policy change would have resulted in a dramatic rise in cost if applied to all of the infected herds at that time, so was rolled out in herds based on risk of
95
recrudescence; therefore not all eligible herds actually completed the parallel testing. Herds were considered to be at highest risk of recrudescence if they had multiple prior episodes of bTB or had more than one confirmed infected animal found at the latest episode (Dawson et al., 2014), and herds meeting these criteria were given preference for parallel testing at the final whole herd test. Use of parallel ␥-IFN testing occurs in three main ways in New Zealand. Firstly, the parallel ␥-IFN assay has been used in skin test negative animals in infected herds at the first WHT after the detection of an infected episode, WHT(1). The objective is to increase the probability of early detection of diseased animals within the herd and thus shorten the time the herd is deemed to be infected. Herds are eligible to receive parallel ␥-IFN testing when more than one confirmed infected animal has been found at breakdown, whether by routine slaughter inspection or by tuberculin skin testing. Secondly, parallel ␥-IFN is also used to improve confidence that the herd is truly clear of bTB at the final WHT, WHT(F), before movement restrictions are lifted. Finally, parallel ␥-IFN is used to confirm the absence of infection in skin test negative animals from lowrisk cohorts (e.g. dairy heifers) before movement out of an infected herd to another property, a permitted activity under stringent conditions. As the new policy has been in place for four years, the aim of this retrospective observational study was to evaluate the effectiveness of the use of parallel ␥-IFN in New Zealand’s bTB eradication scheme.
2. Materials and methods A list of all herds that had an infected status between 1 July 2011 and 1 September 2015 was obtained from the OSPRI1 disease management databases (DMIS—Disease Management Information System and DMS—Disease Management System), maintained by TBfree New Zealand (previously the Animal Health Board). There were 289 herd numbers eligible for consideration. Fifty herds were excluded from the study. These included deer and beef meat production herds where parallel ␥-IFN testing was not used. The parallel ␥-IFN testing is only used on beef breeding and dairy herds. In addition herds were not included where bTB was not confirmed by culture. Ninety-nine beef breeding and 140 dairy herds were eligible for the study, making a total of 239 study herds. Data were extracted from DMS to gather the following information: 1. Whether the herd had one or more prior bTB breakdowns in its history; 2. For the first WHT at or after detection of bTB, WHT(1)* : a Number of skin test positive animals slaughtered b Number of confirmed bTB cases in the skin test positive animals slaughtered c Whether a parallel ␥-IFN of the infected cohort was undertaken d Number of parallel ␥-IFN reactors slaughtered e Number of confirmed bTB cases in the parallel ␥-IFN positive animals slaughtered; 3. For the final WHT (a minimum of six months after the herd had completed a WHT where there had been no confirmed bTB), WHT(F)** :
1 OSPRI, Operational Solutions for Primary Industries, the entity that manages the TBfree and NAIT (National Animal Identification and Tracing) schemes in New Zealand.
96
J.A. Sinclair et al. / Preventive Veterinary Medicine 127 (2016) 94–99
Table 1 Classification of the 239 study herds by risk factors for undetected infection.
bTB history in herd No bTB history in herd Total
Multiple bTB cases
Single bTB case
Total
24 34 58
119 62 181
143 96 239
a Number of skin test positive animals slaughtered b Number of confirmed bTB cases in the skin test positive animals slaughtered c Whether a parallel ␥-IFN of the infected cohort was undertaken d Number of parallel ␥-IFN reactors slaughtered e Number of confirmed bTB cases in the parallel ␥-IFN positive animals slaughtered.
Table 2 Number of reactors slaughtered at WHT(1), and number and proportion of reactors found to have lesions of bTB at slaughter for the 239 dairy and beef breeding herds in the study. Number of reactors with bTB lesions at slaughter
Percentage
653 430
380 160
58.2% 37.2%
1083
540
49.9%
Reactors slaughtered
CFT testing Parallel ␥-IFN (CFT negative) testing Total
Table 3 Number and percentage of 239 study herds where parallel ␥-IFN was used at WHT(1), reactors slaughtered and bTB lesions found.
* If
the breakdown was detected at a WHT, this became WHT(1). If the breakdown was detected by meat inspection, the WHT(1) was completed subsequently. ** Thirty-one herds were still on infected movement control as at 1 September 2015, and had not been eligible for a WHT(F). The definition of WHT(F) is a test that occurs at least six months after a clear whole herd CFT in which there are no confirmed cases of bTB, where a second clear test would qualify the herd for revocation of movement restrictions. If bTB was found at this test, it was still termed WHT(F) for the purposes of this study, although the herd remained on an infected status. 2.1. Risk categories of study herds Ninety-three out of the 239 study herds actually had a parallel test applied at WHT(F). Thirty-one herds were not yet eligible for a WHT(F) by the end date of the study because they had not yet completed a clear CFT of the whole herd. The remaining 208 herds were eligible under the 2011 policy, but the test was preferentially rolled out in the highest risk herds due to the prohibitive cost of parallel testing all herds at WHT(F). The decision on whether to apply the parallel test was discretionary and also involved a consideration of cost and logistics. Ten per cent (24) of the 239 study herds had both a history of bTB in the herd and multiple bTB animals found during their current episode (Table 1). These were considered the highest risk for undetected infection and therefore were given highest priority for parallel testing. Twenty-six per cent (62) of the herds had no bTB history and only a single confirmed infected animal found before the final whole herd testing. These herds were considered low risk. The remaining 64% (153) of herds had only one risk factor: either a bTB history or multiple infected animals in the episode. Sixty-seven of the 93 herds parallel tested had at least one of the risk factors for undetected infection, and of these, 16 herds (80% of the 20 test eligible herds with both risk factors) were parallel tested at WHT(F). 2.2. The caudal fold test (CFT) The CFT was undertaken by intradermally injecting a 0.1 ml volume of bovine PPD (3000 IU/0.1 ml dose; Prionics, Lelystad, The Netherlands) in the caudal fold of the tail. The caudal fold was inspected 72 h later and any detectable swelling was regarded as a positive response. 2.3. The -IFN test Heparinised blood samples were collected from the cattle at 0–30 days following the CFT test and the blood cultures were set up the day following collection, but within 30 h of collection. Blood
Parallel ␥-IFN test used Reactors slaughtered bTB lesions found in reactors
Number of herds
Percentage
64 49 32
26.8% of 239 herds 76.6% of 64 herds 65.3% of 49 herds
samples (1.0 ml) were dispersed into wells of a 24-well plate and preservative-free bovine PPD prepared from M. bovis or PPD prepared from M. avium, avian PPD (24 g/ml final concentration; Prionics, Schlieren-Zurich, Switzerland), or phosphate-buffered saline (negative control) was added. After incubation at 37 ◦ C for 16–24 h, the plasma supernatants were harvested and their ␥-IFN levels measured using a sandwich ELISA kit (Prionics). For parallel interpretation, a Bovine PPD optical density (OD) minus Avian PPD OD of >0.07 was deemed “high risk”; “medium risk” between OD of 0.04–0.07, and “low risk” OD <0.04. The decision to designate animals as reactors depended on the judgement of the veterinarian in charge of the case, and the variable cut-off (Buddle et al., 2015) allowed the slaughter of ‘cuts’ of animals from the highest OD down, with subsequent slaughter decisions dependent on the previous post mortem results. Therefore the same reactor cut-off was not applied in every herd, and the retrospective nature of this study did not allow us to determine the actual cut-off value applied, or to compare the effects of different cut-off values on the lesion findings. 3. Results 3.1. First whole herd test CFT testing at WHT(1) detected 653 reactors over the 239 herds. At slaughter, 380 of these reactors (58.2%) had lesions of bTB detected (Table 2). Parallel ␥-IFN testing was only undertaken on animals which were negative in the CFT and was used in 64 (26.8%) of the 239 eligible herds at WHT(1). Of these herds, a total of 430 reactors were found over 49 (76.5%) herds at ␥-IFN testing. When these reactors were slaughtered, 160 animals from 32 herds (37.2% of reactors and 50% of herds) were found to have lesions of bTB (Tables 2 and 3). Of dairy herds, 46 of 140 had parallel ␥-IFN testing applied and reactors were found in 33 of these 46 herds (71.2%), whereas in beef breeding herds, 18 out of 99 herds were parallel ␥-IFN tested, with reactors found in 16 of 18 herds (88.9%). 3.2. Final whole herd test Herds were eligible for a WHT(F) if they had completed a whole herd CFT with no confirmed bTB animals found at least six months previously, and 208 of the 239 study herds were eligible for a WHT(F). At WHT(F), CFT testing found 74 reactors in the 208 eligible herds, and 10 of these reactors had confirmed bTB lesions at
J.A. Sinclair et al. / Preventive Veterinary Medicine 127 (2016) 94–99 Table 4 Number of reactors slaughtered at WHT(F), and number and proportion of reactors found to have lesions of bTB at slaughter for the 239 dairy and beef breeding herds in the study. Reactors Number of slaughtered reactors with bTB lesions at slaughter CFT testing 74 Parallel ␥-IFN (CFT negative) testing 122 Total 196
10 21 31
13.5% 17.2% 15.8%
Number of herds
Percentage
93 49 13
44.7% of 208 herds 53.9% of 93 herds 26.5% of 49 herds
Table 6 Number of herds with reactors slaughtered and bTB cases, number of reactors and bTB cases per herd for the 38 study herds that had both WHT(1) and WHT(F) applied. WHT(1) Herds with reactors slaughtered Herds with bTB cases Average number of reactors slaughtered per herd Average number of bTB cases per herd
31 22 6.8 2.5
119 (83%) and 24 (17%) respectively (Table 1). A statistically significant association was found between bTB history or no prior history, and the finding of one or multiple confirmed cases (X2 = 9.86, df = 1, p = 0.0017).
Percentage
Table 5 Number and percentage of 208 eligible study herds where parallel ␥-IFN used at WHT(F), reactors slaughtered and bTB lesions found.
Parallel ␥-IFN test used Reactors slaughtered bTB lesions found in reactors
97
WHT(F) 17 4 1.2 0.2
slaughter, thereby indicating a failure to eliminate bTB and a continuation of the infected status and movement restrictions. For the 208 herds, a parallel ␥-IFN was undertaken in 93 (44.7%) herds and 49 of these (52.7%) had reactors found. Of these 49 herds, 13 (26.5%) had a total of 21 reactors with lesions of bTB detected at slaughter (Tables 4 and 5). These 13 herds remained on movement restriction and an infected status. The percentage of dairy and beef breeding herds with reactors found at parallel testing was similar (53.1% of 64 dairy herds and 51.7% of 29 beef herds). Eleven of the 13 herds in which parallel ␥-IFN reactors with confirmed bTB were found at WHT(F) had no infection identified by CFT, and therefore would have had movement restrictions revoked if no parallel testing had been done. Table 4 shows the number of CFT test and parallel reactors found at WHT(F) and the percentage of bTB confirmed lesions in these groups. 3.3. Herds receiving parallel testing at both WHT(1) and WHT(F) Thirty-eight herds had received a parallel test at both WHT(1) and WHT(F) (Table 6). In WHT(1), a mean of 6.8 reactors were found by the parallel, dropping to 1.2 at the WHT(F). The mean number of reactors found with bTB lesions at slaughter correspondingly fell between WHT(1) and WHT(F) from 2.5 to 0.2 (Table 6). 3.4. Risk factors for undetected infection Sixty per cent of the 239 study herds had a history of one or more bTB breakdowns in the herd prior to the study period. The test interval before detection of bTB and length of time since prior bTB breakdown was not determined from the data available. Twentyfour per cent of the study herds had more than one bTB confirmed case found in the current breakdown. Ten per cent of the herds had both multiple bTB cases and at least one prior breakdown (Table 1). Of herds that had no bTB history, 62 (65%) had only a single bTB animal found in the current breakdown, and 34 (35%) had multiple confirmed cases. For herds with a bTB history, these figures were
4. Discussion New Zealand’s TB control since the mid 1990’s has been very effective and based mainly on the CFT, with extra testing and movement restrictions from the known areas of high risk for infected possums (Vector Risk Areas, VRAs). After twenty years of control of possums there has been a change in the nature of breakdowns with a greater proportion occurring in the Vector Free Areas (VFAs) through the trading of stock. As the infected herd prevalence has dropped, more effort is now needed to increase the sensitivity of detection of disease in infected herds. Sixty per cent of study herds had a history of at least one prior bTB episode and period of revocation of movement restrictions prior to the study breakdown. In many cases the most recent bTB breakdown may represent residual infection from an historic breakdown. The length of time since the most recent breakdown, and testing interval before detection was not determined from the data but is worthy of investigation in its own right. Singleton breakdowns were more likely in herds that had experienced a prior breakdown, and reasons may include the earlier detection of disease as a result of more intensive testing post-clearance, or the detection of a single residual case post clearance. Study herds that had a bTB history were significantly more likely to have only a single bTB case in the study breakdown. The results of this study show that many of these repeat breakdowns could have been prevented by maximising test sensitivity before clearance. Parallel ␥-IFN testing helps achieve this sensitivity and is now more affordable than in the past, due to the lower number of herds involved. The test is capable of detecting infection at an earlier stage than is possible with the skin tuberculin test (Buddle et al., 1995; Wood et al., 1991; Wood and Jones, 2001) and therefore may detect early infected animals that would otherwise remain as a source of ongoing transmission within the herd (Houlihan et al., 2008). Eleven of the 93 herds in our study in which a parallel ␥IFN test was used failed their WHT(F) because confirmed infection was found at the parallel test, despite the herd being clear to CFT. This study has demonstrated the efficacy of the parallel ␥-IFN assay under New Zealand field conditions for detecting many additional cases of bTB infection in infected herds that could not be found by intradermal tuberculin testing alone. The policy of parallel ␥-IFN testing in breeding or dairy herds as part of final clearance testing applied since 2011 has therefore been successful in preventing some potential movement-related or recrudescent breakdowns. The benefit to New Zealand of parallel testing as part of its eradication plan is likely to significantly outweigh the cost of the supplementary testing. Each additional herd breakdown prevented is estimated to save a minimum of NZD 20,000 for a single animal breakdown. Two extensive multiple animal breakdowns caused by the inadvertent purchase of infected animals in the North Island of New Zealand in 2013 cost an estimated NZD400,000 each in testing costs and reactor compensation before the herds were returned to a clear status, not including costs to the farmer from lost revenue and additional labour (J. Sinclair, unpublished data). The number of additional breakdowns prevented by increased testing sensitivity before clearance cannot be quantified, but undetected infection at clearance puts at least the newly cleared herd at risk, and potentially many herds through trading. In the past six years, two cases of milk-borne spread to young calves from cows with undetected udder bTB infection resulted in infection being
98
J.A. Sinclair et al. / Preventive Veterinary Medicine 127 (2016) 94–99
later found in an additional seven and 10 herds respectively (J. Sinclair, unpublished data). Reasons for the parallel test being selectively applied during the roll-out period even in high-risk herds in this study were herd size and logistics. Dairy herds in New Zealand have an average of 419 cows, with 12% of herds comprised of over 750 cows (Anon., 2015b). A whole herd parallel test is much more financially and logistically viable in a small to medium sized herd. Four eligible herds in the study had both a recent bTB history and multiple bTB animals found, yet did not receive parallel testing at WHT(F). Two of these herds were very large (5122 and 4145 animals). One had bTB found at CFT testing in their WHT(F) and therefore had already failed, and the last had no identifiable reasons for not completing a parallel at WHT(F). Where the parallel ␥-IFN test was applied at WHT(F), the lesion rate in the reactors to this test was higher than the lesion rate in CFT positives. This observation gives rise to concern that herds that did not receive parallel testing may have had residual infection found, if they had been tested. As New Zealand’s bTB eradication scheme enters a more rigorously targeted phase, a team approach to case management will replace the individual judgement-based approach of the past. Flexibility in test application and interpretation is still important, as this enables experiential learning. A team focus will help achieve transparency and the long-term management of risk from undetected infection to the country as a whole. Additionally, with infected herd numbers now below 40 nationally, parallel testing of even very large herds at clearance is financially viable, and the aim is to apply these tests in every infected herd. Where herds had received parallel testing at both the WHT(1) and a WHT(F), markedly fewer reactors were slaughtered and less infection found at WHT(F) than at WHT(1). This gives confidence that the disease control measures being applied in infected herds are effective at identifying disease and preventing its spread. The sensitivity of the ␥-IFN assay has been estimated to be between 73 and 100%, with a median sensitivity of 87.6% (de la Rua-Domenech et al., 2006). Under New Zealand conditions, the sensitivity was estimated to be 85% with a specificity of 93% for animals that had reacted positively in the CFT (Ryan et al., 2000), with the sensitivity considered equivalent or greater to that of the CFT. The sensitivity and specificity of the ␥-IFN test is dependent on the cut-off value used for positive interpretation (Goodchild and Clifton-Hadley, 2001; Wood and Jones, 2001; de la Rua-Domenech et al., 2006). Lowering the cut-off from OD > 0.1 to OD > 0.04 increased the sensitivity from 94% to 98% in one trial (Buddle et al., 2001). In the past, policy in New Zealand has allowed discretionary application of testing and cut-off values within guidelines. This has been necessary due to herd differences in length of time infected, presence of nonspecific reactions and within-herd transmission, and has enabled empirical learning. The discretionary nature of the OD cut-off used for disease management in infected herds is problematic in the interpretation of the results of this study. No conclusions can be drawn about the effect on the lesion finding rate of reactors of using a lower cut-off. Future research could examine the correlation between OD reading and post-mortem or culture positive rate. The ␥-IFN assay has been used in other countries with parallel interpretation and a lower cut-off to improve sensitivity of detection in high risk herds (Buddle et al., 2001; Wood et al., 1991; Wood and Jones, 2001; de la Rua-Domenech et al., 2006), especially for clearance testing, where a decision must be made on whether to release a herd from movement restrictions. In infected herds, the use of two tests capable of detecting different subpopulations of infected animals will improve the rate of resolution of a bTB episode and should also reduce the chance of recurrent infection in the absence of movement or vector-related reinfection (de la Rua-
Domenech et al., 2006; Ameni et al., 2010). Skin and ␥-IFN testing in parallel is thought to be 20% more sensitive than skin testing alone (Wood et al., 1991; Whipple et al., 1995; González Llamazares et al., 1999) and is estimated to have a sensitivity of 93% (Gormley et al., 2006). In this study, half of all ␥-IFN reactors at WHT(1) had no gross evidence of infection. The absence of lesions at slaughter is not proof of the absence of infection, as up to 10% of reactors with no visible lesions were shown to be culture positive for M.bovis in one study (Corner, 1994); therefore infection may be missed despite critical examination of prescribed body sites. Reactors slaughtered with no visible lesions may be accounted for by (i) early infection, where the immune response has been activated but lesions are too small to detect at post-mortem examination (Menzies and Neill, 2000); (ii) latent infection, where the organism is present in a dormant state without disease (Cassidy et al., 2001); (iii) cross-reactivity with other mycobacterial organisms (de la RuaDomenech et al., 2006) or (iv) the lesions were not detected in reactor post-mortem (Corner, 1994). Several studies have shown that parallel ␥-IFN reactors not slaughtered were significantly more likely to be later identified as infected with bTB. A longitudinal study from Ireland followed ␥IFN reactors remaining in their herds for a further 155 days after the initial assay. At the next two skin tests these reactors were 7–9 times more likely than their ␥-IFN negative herd-mates to be skin test positive (Gormley et al., 2006). The odds ratio of ␥-IFN reactors subsequently becoming skin test positive compared with non-reactors was 7–20 in another study (Collins, 2000). When 100 animals were followed after an initial ␥-IFN positive result, they were significantly more likely to be identified as having bTB at subsequent testing at a rate ratio of 17.18 compared to the animals negative to both the skin and ␥-IFN test (Welsh et al., 2008). A Northern Ireland study of 1107 ␥-IFN positive reactors that were not slaughtered showed that these animals were 2.31 times more likely than ␥-IFN negative animals to become reactors to the intradermal tuberculin test in the following five years (Lahuerta-Marin et al., 2015). In New Zealand, the presence of reactor animals without confirmed bTB at a final clearance test increased the hazard of a future breakdown by 2.8 times in the first two years after clearance (Dawson et al., 2014). The removal of these potentially early infected animals negates the possibility that they may pose a future transmission risk to their herd-mates or to other herds postmovement. The slaughter of some false positive reactors in infected herds due to the lower specificity of the test combination of CFT and ␥-IFN is a necessary price to pay for the eradication of the disease, and helps to prevent future losses. In confirmed infected herds and in areas with endemic wildlife disease, reactors are highly predictive of infection (de la RuaDomenech et al., 2006). Parallel reactors may indicate ongoing within-herd transmission and may be the only indicator of the presence of an anergic shedder within the herd. New Zealand’s bTB surveillance scheme has relied upon slaughterhouse surveillance and annual to triennial caudal fold tuberculin testing in eligible stock, depending on their proximity to vector risk. Both of these elements have imperfect sensitivity. This study has shown that in bTB infected herds, CFT failed to identify a significant proportion of the infected animals in some herds. Selection pressure may have resulted in the emergence of strains that are less detectable by CFT. The sensitivity of tuberculin testing is dose and site dependent (de Kantor et al., 1984). To improve sensitivity, mid cervical testing and/or increasing the tuberculin dose could be considered in infected herds in New Zealand to support the aim of eradication. If the sensitivity of the CFT is as low as these findings suggest, it is difficult to explain how the surveillance scheme been so successful in reducing the burden of infected herds. However, it is easier to
J.A. Sinclair et al. / Preventive Veterinary Medicine 127 (2016) 94–99
find infection with the CFT at a herd level when infection is more prevalent, and the CFT has therefore been useful in reducing the bulk of infection when applied over large numbers of herds. This study has confirmed that the CFT has poor sensitivity at the animal level. Increased use of parallel testing in the future, perhaps with adjustments to the use of tuberculin, will undoubtedly contribute to ongoing success of New Zealand’s bTB eradication scheme by improving animal-level sensitivity. Our study has confirmed under field conditions in New Zealand that the greater test sensitivity offered by the combination of CFT and parallel ␥-IFN enables the detection of many more cases of within-herd infection at the first test following detection of bTB in a herd than with skin tuberculin testing alone. Thus the combination of tests enables more complete removal of infected animals and therefore prevention of ongoing transmission, resulting in a more rapid resolution of within-herd infection. Parallel ␥-IFN before revocation of movement restrictions gives greater confidence that the herd will remain clear of disease in the long term (Gormley et al., 2006) providing economic benefits to farmers and cattle industries. Acknowledgements The authors thank the AsureQuality and VetEnt technicians for undertaking the CFT and collection of the blood samples for the ␥IFN, and the AgResearch bTB diagnostic team for the ␥-IFN testing and culture of M. bovis from slaughtered animals. References Ameni, G., Aseffa, A., Hewinson, G., Vordermeier, M., 2010. Comparison of different testing schemes to increase the detection Mycobacterium bovis infection in Ethiopian cattle. Trop. Anim. Health Prod. 42, 375–383. Anon, 2011. National bovine tuberculosis pest management strategy. In: National Operational Plan Part B: Operational Policies. Animal Health Board (now known as OSPRI), Wellington, New Zealand. Anon., Annual Report. OSPRI, Wellington, New Zealand, 2015. Anon, 2015b. New Zealand Dairy Statistics 2014–15. Livestock Improvement Corporation and DairyNZ Limited, 52 pp. Buddle, B.M., de Lisle, G.W., Pferrer, A., Aldwell, F.E., 1995. Immunological responses and protection against Mycobacterium bovis in calves vaccinated with a low dose of BCG. Vaccine 13, 1123–1130. Buddle, B.M., Ryan, T.J., Pollock, J.M., Andersen, P., De Lisle, G.W., 2001. Use of ESAT-6 in the interferon-test for diagnosis of bovine tuberculosis following skin testing. Vet. Microbiol. 80, 37–46. Buddle, B.M., de Lisle, G.W., Griffin, J.F.T., Hutchings, S.A., 2015. Epidemiology, diagnostics, and management of bovine tuberculosis in domestic cattle and deer in New Zealand in the face of a wildlife reservoir. N. Z. Vet. J. 63, 19–67. Cassidy, J.P., Bryson, D.G., Gutierrez Cancela, M.M., Forster, F., Pollock, J.M., Neill, S.D., 2001. Lymphocyte subtypes in experimentally induced early-stage bovine tuberculous lesions. J. Comp. Pathol. 124, 46–51.
99
Collins, J.D., 2000. A longitudinal study of cattle found positive to the interferon-gamma assay for Mycobacterium bovis infection. Soc. Vet. Epidem. Econ. Proc., 1263–1265. Corner, L.A., 1994. Post mortem diagnosis of Mycobacterium bovis infection in cattle. Vet. Microbiol. 40, 53–63. Dawson, K.L., Stevenson, M.A., Sinclair, J.A., Bosson, M.A., 2014. Recurrent bovine tuberculosis in New Zealand cattle and deer herds, 2006–2010. Epidemiol. Infect. 142, 2065–2074. de la Rua-Domenech, R., Goodchild, A.T., Vordermeier, H.M., Hewinson, R.G., Christiansen, K.H., Clifton-Hadley, R.S., 2006. Ante mortem diagnosis of tuberculosis in cattle: a review of the tuberculin tests, -interferon assay and other ancillary diagnostic techniques. Res. Vet. Sci. 81, 190–210. de Kantor, I.N., Odeon, A.C., Steffan, P.E., Auza, M.J., Madrid, C.R., Marchevsky, N., 1984. Sensitivity of the cervical and the caudal fold tuberculin tests with Mycobacterium bovis in infected cattle of Argentina. Rev. Sci. Tech. OIE 3, 137–150. Ekdahl, M.O., Smith, B.L., Money, D.F., 1970. Tuberculosis in some wild and feral animals in New Zealand. N. Z. Vet. J. 18, 44–45. González Llamazares, O.R., Gutiérrez Martín, C.B., Aranaz Martín, A., Liébana Criado, E., Domínguez Rodríguez, L., Rodríguez Ferri, E.F., 1999. Comparison of different methods for diagnosis of bovine tuberculosis from tuberculin-or interferon-reacting cattle in Spain. J. Appl. Microbiol. 87, 465–471. Goodchild, A.V., Clifton-Hadley, R.S., 2001. Cattle-to-cattle transmission of Mycobacterium bovis. Tuberculosis 81, 23–41. Gormley, E., Doyle, M.B., Fitzsimons, T., McGill, K., Collins, J.D., 2006. Diagnosis of Mycobacterium bovis infection in cattle by use of the gamma-interferon (Bovigam) assay. Vet. Microbiol. 112, 171–179. Houlihan, M.G., Dixon, F.W., Page, N.A., 2008. Outbreak of bovine tuberculosis featuring anergy to the skin test, udder lesions and milkborne disease in young calves. Vet. Rec. 163, 357–361. Lahuerta-Marin, A., Gallagher, M., McBride, S., Skuce, R., Menzies, F., McNair, J., McDowell, S.W.J., Byrne, A.W., 2015. Should they stay, or should they go? Relative future risk of bovine tuberculosis for interferon-gamma test-positive cattle left on farms. Vet. Res. 46. Livingstone, P.G., Hancox, N., Nugent, G., de Lisle, G.W., 2015. Toward eradication: the effect of Mycobacterium bovis infection in wildlife on the evolution and future direction of bovine tuberculosis management in New Zealand. N. Z. Vet. J. 63, 4–18. Menzies, F.D., Neill, S.D., 2000. Cattle-to-cattle transmission of bovine tuberculosis. Vet. J. 160, 92–106. Ryan, T.J., Buddle, B.M., De Lisle, G.W., 2000. An evaluation of the gamma interferon test for detecting bovine tuberculosis in cattle 8–28 days after tuberculin skin testing. Res. Vet. Sci. 69, 57–61. Welsh, M.D., McNair, J., McDowell, S.W.J., Buchanan, J., Hill, R., McBride, S., 2008. The Northern Ireland interferon-gamma (IFN-g) testing programme. Cattle Practice 16, 136–139. Whipple, D.L., Bolin, C.A., Davis, A.J., Jarnagin, J.L., Johnson, D.C., Nabors, R.S., Payeur, J.B., Saari, D.A., Wilson, A.J., Wolf, M.M., 1995. Comparison of the sensitivity of the caudal fold skin test and a commercial gamma-interferon assay for diagnosis of bovine tuberculosis. Am. J. Vet. Res. 56, 415–419. Wood, P.R., Jones, S.L., 2001. BOVIGAMTM : an in vitro cellular diagnostic test for bovine tuberculosis. Tuberculosis 81, 147–155. Wood, P.R., Corner, L.A., Rothel, J.S., Baldock, C., Jones, S.L., Cousins, D.B., McCormick, B.S., Francis, B.R., Creeper, J., Tweddle, N.E., 1991. Field comparison of the interferon-gamma assay and the intradermal tuberculin test for the diagnosis of bovine tuberculosis. Aust. Vet. J. 68, 286–290.