Comparison of 4 agar gel immunodiffusion kits for serologic detection of equine infectious anemia virus antibodies

Abstract

Using 85 sera collected from horses that had been experimentally infected with equine infectious anemia virus (EIAV) and 200 field sera collected from racehorses in Japan, we compared 4 agar gel immunodiffusion (AGID) kits for serologic detection of EIAV antibodies from Idexx, VMRD, IDvet, and the National Engineering Research Center of Veterinary Biologics, China (NECVB). The positive control lines were sufficiently clear in all kits for evaluation to be made, with slight differences in sharpness: NECVB was the sharpest, followed by VMRD, IDvet, and Idexx. The test results for all 285 samples agreed among the 4 kits, with 62 positives and 223 negatives. The sensitivities and specificities of VMRD, IDvet, and NECVB compared with the Idexx kit were 100%, and the kappa coefficient values between the kits were 1.0 for all combinations. We concluded that the testing capacity of these 4 kits was virtually identical.

Equine infectious anemia (EIA) is caused by the equine infectious anemia virus (EIAV; Retroviridae, Lentivirus). ¹ In nature, the virus is transmitted among horses by the bites of blood-feeding insects. ¹ However, EIAV transmission can also occur via the reuse of injection needles, syringes, and surgical equipment, as well as by direct administration of infected blood and plasma products.^1,5,9 Major clinical signs of EIA are recurring fever, anemia, depression, edema, hemorrhage, and wasting syndrome, although many infected horses do have obvious signs in the chronic phase and even on first exposure.^1,8 Horses infected with EIAV carry the virus throughout their lifetime and are potential reservoirs for infecting other horses, regardless of whether they are clinically ill. Elimination of infected horses is therefore essential for the control of EIA.

The diagnosis of EIA is supported by virus isolation, molecular methods for detection of viral nucleic acid, and serologic methods that include agar gel immunodiffusion (AGID), ELISA, and immunoblotting. ¹⁰ Of these, AGID is simple to perform in the laboratory and has the best sensitivity and specificity, which makes it the preferred confirmatory test. Although ELISAs are rapid and suitable for testing many samples, a substantially higher rate of false-positives relegates them to use as a screening test only. ¹⁰ Additionally, AGID can be used for confirmation of freedom from infection for populations, individual horses, and eradication programs ¹⁰ ; hence, AGID is used widely as a standard method.

Commercial kits for EIA AGID are available in many countries, but some companies have stopped providing these products. To maintain appropriate quarantine and surveillance systems for EIA, it is essential to evaluate multiple kits for their testing performance, allowing alternative products to be secured as backups. Few reports show direct comparisons of these kits^4,6,7; however, there is one report of agreement between AGID kits from Idexx and VMRD (kappa of 0.933; 600 sera [285 positive and 315 negative samples using the Idexx kit]). ⁷ Another study reported complete agreement of test results of AGID kits from Idexx, VMRD, and IDvet. ⁴ However, the 316 serum samples tested in the latter study included only 5 positives (1.6%). ⁴ Therefore, further evaluations are needed using sera that provide more positive samples. An AGID kit from the National Engineering Research Center of Veterinary Biologics, China (NECVB, http://necvb.cn/) was developed by the World Organisation for Animal Health (WOAH) Reference Laboratory for EIA (The Chinese Academy of Agricultural Sciences, Harbin Veterinary Research Institute), but there are hitherto no reports that compare its performance with those of other kits, to our knowledge.

We compared 4 EIA AGID kits using 85 sera collected from 10 horses that had been experimentally infected with EIAV (strain P337), and 200 field sera collected from racehorses in Japan. The EIAV infection experiments were performed from 1982 to 1996, and had been approved by the Animal Care Committee of the Japan Racing Association, Equine Research Institute. The sera were collected sequentially from the day of viral inoculation to 199 d later, and were stored at −20°C. Another series of sera (n = 200) were collected in 2022 from racehorses at Japan Racing Association training centers.

In 2017, the Japanese Ministry of Agriculture, Forestry and Fisheries had concluded, based on the results of a survey of 70,133 horses, including racehorses, riding horses, and wild horses, that EIA had been eradicated in Japan. The Japanese Council of Equine Health has reported no EIA-positive horses since 2020 in their annual survey, which includes ~1,000 horses. ² Our series of 200 field-collected sera in 2022 was therefore anticipated to be negative for EIAV antibodies.

We used the AGID EIA test (lot BU662; Idexx), Equine infectious anemia virus antibody test kit AGID (lot P220201-001; VMRD), EIA AGID (lot J93; IDvet), and the Equine infectious anemia AGID kit (lot 20221018; NECVB). Kits were stored as recommended by the manufacturers, and tests were performed before the expiration dates of each product. The gel used, consisting of 1% w/v agar in 0.145 M borate buffer (9 g H₃BO₃ and 2 g NaOH per L) with 0.1% w/v sodium azide, was prepared according to the WOAH Manual of Diagnostic Tests and Vaccines for Terrestrial Animals. ¹⁰ A 5-mm metal punch was used to form 6 wells encircling a center well of the same diameter; the wells were 3 mm apart. Antigen solution (50 µL), positive control serum, and tested serum were added to the wells. The antigen solution was placed in the central well, the positive control serum was placed in the 2 wells to the left and right of the central well, and serum samples were placed in the remaining 4 wells at the top and the bottom. The plates were incubated at room temperature in a closed humid chamber for 24 h, after which they were visually evaluated by 2 trained and experienced Japan Racing Association Equine Research Institute veterinarians. For any samples over which the 2 judges disagreed, a judgment by an additional trained veterinarian was employed to provide the final result. A serum sample was concluded to be negative when the reference positive control line continued into the test sample well without curving or with a slight curve away from the antigen well and toward the positive control serum. A serum sample was judged positive when the control line joined with and formed a continuous line with the line between the test serum and antigen, or when the control line curved slightly toward the antigen well and away from the reference positive control serum well, but did not form a complete line between antigen and test serum. Any lines formed between the antigen and test serum, but which did not form a continuous line with the positive control line, were regarded as nonspecific lines.

The test result with the Idexx kit was designated arbitrarily as the standard, against which the results of the 3 other kits were compared. The kappa coefficient values were evaluated as follows: < 0–0.20 = slight agreement; 0.21–0.40 = fair agreement; 0.41–0.60 = moderate agreement; 0.61–0.80 = substantial agreement; and 0.81–1.0 = almost perfect agreement. ³

The positive control lines were sufficiently clear for evaluations to be made with all of the kits, but they showed a slight difference in sharpness: NECVB was the sharpest, followed progressively by VMRD, IDvet, and Idexx (Fig. 1A). No nonspecific lines were observed for any samples tested with the 4 kits. All of the kits detected strong-positive lines and weak-positive reactions between the antigen wells and the sample wells (Fig. 1B, 1C). The results for all 285 samples coincided among the 4 kits, with 62 positives and 223 negatives (Table 1). Of 85 sera from experimentally infected horses, 23 were negative for EIAV antibodies. Because these sera were collected at 0–21 d after viral inoculation, we reasoned that the detectable antibodies had not been produced yet. The sensitivities and specificities of VMRD, IDvet, and NECVB kits against the Idexx kit were 100%, and the kappa coefficient values between the kits were 1.0 for all combinations, namely Idexx vs. VMRD; Idexx vs. IDvet; and Idexx vs. NECVB. We, therefore, concluded the testing capacity of these 4 kits to be virtually identical.

Figure 1.

Results of testing for equine infectious anemia viral antibody with agar gel immunodiffusion kits from Idexx, VMRD, IDvet, and the National Engineering Research Center of Veterinary Biologics, China (NECVB). A. Positive control lines formed between the antigen well (center) and positive control wells (left and right). B. Strong-positive lines formed between the antigen well (center) and the sample wells (top left and right, and bottom left and right). The serum in the upper left well was evaluated as being weakly positive with the NECVB kit, but strongly positive with the other kits. Similar differences in levels of positive reactions were observed for some of the other sera, but no samples were judged positive with one kit and negative with another. C. Weak-positive reactions between the antigen well (center) and the sample well (bottom left). The control line curved slightly toward the antigen well and away from the reference positive control serum well, but did not form a complete line between antigen and test serum.

Table 1.

Comparison of the results of 4 AGID kits for the serologic detection of equine infectious anemia antibodies using 285 horse serum samples.

Idexx	VMRD			IDvet			NECVB		Total
	+	−	Total	+	−	Total	+	−
+	62	0	62	62	0	62	62	0	62
−	0	223	223	0	223	223	0	223	223
Total	62	223	285	62	223	285	62	223	285

+ = positive; − = negative.

A limitation of our study is that the 62 sera judged to be positive by all 4 tests were sourced from the 10 horses that had been experimentally infected with the same strain of EIAV; there could be a difference in reactivity of antibodies raised after infection with different EIAV strains. A broader range of serum samples should be tested in future studies, which includes those from horses infected naturally with various EIAV strains representing each endemic area.