Assessment of an ELISA method to support surveillance of bovine tuberculosis in Albania

Bovine tuberculosis (bTB) is an important bacterial disease caused by Mycobacterium bovis. It occurs worldwide [1, 2] and affects a wide range of wild and domestic animals [2]. Bovine tuberculosis is a zoonotic disease: man can become infected via milk, aerosols, consumption of infected meat and accidental laboratory exposure [1, 3]. M. bovis is responsible for 5–10 % of all human tuberculosis cases [4] but the infection rate varies widely from country to country [1]. After exposure to infection, animals respond by activating both cellular and humoral immunity. M. bovis moves via the lymphatic system within cells to regional lymph nodes similar to other mycobacterial infections, and delayed hypersensitivity reactions typically commence between 30 and 50 days after the establishment of infection [2, 3, 5]. Tuberculosis is typically a chronic infection. Many infected animals in the herd remain undetected for a long time but they can shed the bacteria by aerosol, via milk, urine and faeces. Consequently, the infection spreads from chronically infected carrier animals to susceptible animals [3, 5]. Programmes to control bTB are based on an identification system for all animals, and the detection of infected individuals at post-mortem examination, especially during meat inspection. Intensive, systematic surveillance, commonly by means of skin tests, is usually linked to the slaughter of positive, reactor animals. Animal movement control and the cleansing and disinfection of contaminated environments [2, 5–9] are important adjuncts to effective control of bTB.

Despite the long history of attempted control and eradication, bTB remains an important disease and its prevalence - in global terms—is almost unchanged; in some countries its prevalence is increasing [1]. This picture is reported in developing countries and more developed countries, including those with highly organized and functional veterinary services that implement eradication programmes correctly [8, 9].

In Albania, the prevalence of bTB is poorly documented. After 1990, the cattle management system changed dramatically: the large, collective, state-owned farms were replaced by many thousands of privately owned small holdings, each with very few cattle. It is estimated that in 2012 there were 480,000 cattle in 375,217 holdings, 73 % with 1–4 animals, and only 5 % with more than 50 cattle. In epidemiological terms, these circumstances complicate the control of bTB. Conversely, the national cattle herd structure might reduce disease transmission since herds are small and are kept separated. Over the last 25 years, the bovine tuberculosis control programme has been based on a skin test, by use of a Purified Protein Derivative (PPD) bovine tuberculin produced by the national reference laboratory, the quality of which was not independently certified. On farms where reactor animals were found, no sanitary measures were implemented. During the last three years, no systematic control measures for bTB have been applied. Moreover, during recent years, there has reportedly been an increase in the prevalence of extra-pulmonary tuberculosis in humans from which M. bovis and M. caprae have been isolated (personal communication, Dr. Silva Tafaj)¹, thus suggesting a bovine origin of the infection. Diagnostic tests that are available today have low sensitivity [2, 6–8, 10–12]. Active surveillance based on the skin test is expensive, and requires well-trained people along with field mobility, adequate supervision; specific, expensive equipment (syringes, needles, etc. not available in Albania.); certified tuberculin, etc.

Advances in the diagnosis of tuberculosis aim to introduce standardized, alternative tests with greater sensitivity and ease of use, and reduced costs, e.g. the M. bovis ELISA test (IDEXX®) [12, 13]. Available data suggest that the sensitivity of this ELISA test is increased by the previous stimulation of the immune system by the intradermal skin test [13], which is currently not done routinely in Albania. However, the higher age of Albanian cattle (frequently >10 years) in many holdings could give more time for them to develop a chronic humoral immune response detectable in the ELISA. For this reason, we conducted this survey to estimate the sero-prevalence of bovine tuberculosis using the IDEXX M.bovis ELISA test and to assess the suitability of this test for bTB surveillance in Albania, which could reduce costs of surveillance.

This observational study was conducted according to the international guidelines stated in the strobe checklist (www.strobe-statement.org), this study was approved on 12 July 2013 by Department of Veterinary Public Health.

Of the 2661 cattle tested, 1.4 % (95 % CI, 0.8–2.1 %) were ELISA positive. Of the 154 herds tested, 18.1 (95 % CI, 1.9–25.8 %) contained at least one ELISA positive animal. The ELISA results are presented in Table 1. Herd and animal prevalence did not vary significantly with district (chi-square for proportion of herds was 8.59 (p = 0.980), chi-square for proportion of cattle tested was 9.51 (p = 0.964)). Although no positive animals were detected in 7 of the 20 districts, neither herd or animal-level prevalence varied significantly between districts (p = 0.980 for herds, p = 0.964 for animals).

In the follow on survey, two animals reacted positively to the skin test (Table 2); these were in the same farm in the village of Katundi i Ri (Fig. 1). Two animals produced inconclusive reactions (Table 2), one in Sohodoll and one in Katundi i Ri (Fig. 1). One of the two positive animals (a cow approximately 15 years old) had skin lesions and a fluctuating oedema, whereas the second reactor (a cow approximately 7 years old) showed only skin thickening at the injection site. The third animal present in the same farm (cow approximately 3 years old) was negative.

Districts	Number of tested herds	Number of positive herds	Percentage (%) of positive herds	Number of cattle tested	Positive samples number/percentage
Durrës	12	2	16.7	266	3/1.13
Krujë	2	0	0	10	0/0
Korçë	5	0	0	53	0/0
Shkodër	15	3	20	198	3/1.52
Fier	21	4	19	235	4/1.7
Lushnjë	6	2	33.3	123	2/1.62
Tepelenë	2	0	0	59	0/0
Tiranë	12	2	16.7	63	2/3.17
Peshkopia	10	4	40	312	5/1.6
Sarandë	2	0	0	72	0/0
Vlorë	10	2	20	139	4/2.87
Elbasan	19	3	15.8	374	4/1.06
Lezhë	6	1	16.7	93	1/1.08
Pukë	4	0	0	44	0/0
Librazhd	1	0	0	13	0/0
Tropojë	6	1	16.7	38	1/2.63
Kukës	2	0	0	14	0/0
Bulqizë	6	1	16.7	148	1/0.67
Gjirokastër	5	1	20	126	2/1.59
Mat	8	2	25	281	4/1.42
Total	154	28	18.18	2661	36/1.35

Legend: Of the 2661 cattle tested, the percentage of positive animals was 1.4 % (95 % CI, 0.8–2.1 %). Of 154 herds tested, the percentage of positive was 18.2 % (95 % CI, 1.9–25.8 %)

^a In Peshkopia district, follow on tests were performed on 87 cattle in Katundi i Ri, Pejkë and Sohodoll villages

Criteria used for determining status of animals	Status of animals	Number of animals
Skin thickness difference ˂2 mm	Negative	83
Skin thickness difference 2–4 mm	Doubtful	2
Skin thickness difference >4 mm	Positive	2
ELISA test results	Negative	83
ELISA test results	Positive	4

Legend: Two animals produced positive skin reactions; four animals were positive in the ELISA test; no animal was positive in both tests

There was no overlap between positive or inconclusive skin tests and positive serological results. ELISA test results showed that four animals were positive (Table 2): one in Pejke, one in Sohodoll and two in Katundi i Ri (Fig. 1).

Many of the problems associated with bTB are due to the limitations of diagnostic tests [2, 5–7]. In general, sensitivity of official diagnostic tests for M. bovis infection is low, and is affected by several factors [2, 7]. Concurrent parasitic infections (e.g., Fasciola hepatica and M. bovis) in cattle are known to reduce cell mediated immune responses to M. bovis [9]. However, this lack of sensitivity occurs even without concurrent F. hepatica coinfection [10]. These limitations complicate implementation of appropriate control measures.

In addition to skin tests, serological tests have been used to diagnose bTB in many countries. However, antibody detection tests, such as ELISA, are not capable of detecting sub-clinical infection [5, 13]. As the disease progresses and high numbers of Mycobacterium bovis bacteria accumulate in tissues, importantly, antibody levels increase in blood [5, 8, 11, 12]. At this stage, ELISA tests are useful for serological surveys of dairy herds. The sensitivity of ELISA tests is generally high in animals with clinical disease. A comparison of the skin tests and IDEXX M. bovis Test for diagnosis of bTB at the herd level, indicated that ELISA had a sensitivity of 63 % which is similar to the skin test [12]. The Mycobacterium bovis IDEXX ELISA Test has the advantage of being a more rapid test and it is less vulnerable to errors and biases on the farm.

In this survey, 1.4 % of the sera tested reacted positively for antibody to Mycobacterium bovis. The median specificity of the M.bovis IDEXX ELISA Test used has previously been estimated to be 98.2 % [12], which means that in a M. bovis negative population of 2661 cattle, up to 48 false positive results could have been expected. In our survey, 36 positive samples were found. Some of these positive results were from clusters of cattle. Households in Albania tend to keep cows for many lactations - a factor that could facilitate the detection of infected animals, since older animals could develop clinical disease. The lack of regular compulsory bTB testing in Albanian cattle meant that their immune systems would not have received the boosting effect of tuberculin.

When we compared the ELISA test with the classical skin test method some animals were positive on one or the other test, however no animals were found to be positive in both skin test and ELISA test. The ages of tested animals were much higher than in western European countries, i.e., the mean was 7.7 years: in Pejke 7.0 years, in Sohodoll 7.8 years, and in Katundi i Ri 8.8 years old. Under these conditions, the ELISA test should improve the possibility of finding chronically infected animals. We could not conclude, however, that the ELISA test is an effective test for monitoring bovine tuberculosis in the dairy cattle population. It would be useful to follow cattle that are positive in both tests through to slaughter to seek to detect histological changes that are characteristic of tuberculosis and to isolate M. bovis from lesions and tissues. Since the two tests we used measured different immune responses, it may be useful to use both testing strategies, ideally using the Mycobacterium bovis IDEXX ELISA Test 2 weeks after the skin test [5], to screen the bovine population for the prevalence of bovine tuberculosis. With the serological boost resulting from tuberculin, the cut-off could be increased, probably resulting in increased specificity.

The number of positive animals detected in ELISA test was low. This could reflect the poor sensitivity of the ELISA test on sera from animals that were not previously sensitized by tuberculin skin test. To produce more accurate results, it is suggested that in future all cattle in affected holdings (herds) should be sensitized by means of the tuberculin skin test preferably 14 days prior to the collection of blood samples, which would then be tested by the M. bovis ELISA IDEXX Test. An affordable and practical approach is needed to control bTB in Albania. Consequently, we recommend that if a suspicious lesion is found at a slaughterhouse, the herd of origin of the animal should be traced, and a skin test and M.bovis IDEXX ELISA Test should be performed on the herd. Tuberculin skin test should be performed by adequately trained veterinarians. Movement control is an essential sanitary measure.

Ab, antibody; bTB, bovine Tuberculosis; CI, confidence interval; ELISA, enzyme-linked immunosorbent assay; EU, european union; F.Hepatica, Fasciola Hepatica; IU, international unit; M.Bovis, Mycobacterium Bovis; OD, optical density; P, positive control serum; PAZA, protection against zoonotic diseases (Albania); Phd, philosophy doctor; PPD, purified protein derivate; S, sample; SICST, single intradermal cervical skin tuberculin test