Validation and Use of a Multiple-Antigen ELISA for Detection of Tuberculosis Infection in Elephants
IAAAM 2000
R. Scott Larsen1, DVM, MS; M. D. Salman1, BVMS, PhD; Susan K. Mikota2, DVM; Ramiro Isaza3, DVM, MS, DACZM; Joni Triantis1, MS
1Center of Veterinary Epidemiology and Animal Disease Surveillance Systems, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, CO, USA; 2Audubon Center for Research of Endangered Species, New Orleans, LA, USA; 3Department of Clinical Sciences, College of Veterinary Medicine, Kansas State University, Manhattan, KS, USA; Present address: Environmental Medicine Consortium, Department of Clinical Sciences, College of Veterinary Medicine, North Carolina State University, Raleigh, NC, USA


Mycobacterium tuberculosis has become an important disease agent in the captive elephant populations of the United States.4 This organism not only presents a threat to the health and welfare of elephants, but is an important zoonotic agent as well. Methods for diagnosing M. tuberculosis in elephants have been problematic, with many tests having inadequate sensitivity and/or specificity.2,3 For example, mycobacterial culture is highly specific for diagnosis of M. tuberculosis; however, many factors may lead to false negative culture results, including inadequate numbers of bacteria, inadequate collection procedures, sample contamination, and/or improper sample handling. Intradermal and serologic tests have not shown good correlation with infection status and most indirect testing methods have not been validated in elephants.2,4 The investigation reported here attempted to determine the validity of a multiple-antigen enzyme-linked immusorbent assay (ELISA) for detection of M. tuberculosis infection in elephants.

Serum samples from 32 Asian elephants (Elaphus maximus) and 15 African elephants (Loxodonta africana) were analyzed using a panel of six antigens. This multiple-antigen ELISA was patterned after the assay used by VanTiem and by Gaborick et al. for detecting M. bovis infection in cattle and cervids.1,5 Antigens used were: M. bovis strain AN5 culture filtrate (CF); purified protein derivative from the standard USDA bovine tuberculin (PPD); modified protein 70, purified from M. bovis strain AN5 (MPB); lipoarabinomannan antigen from the virulent Erdman strain of M. tuberculosis (ERD); lipoarabinomannan antigen from the virulent H37Ra strain of M. tuberculosis (RA); and purified protein derivative from M. avium (AVPPD). Samples were diluted 1:100 and conjugated using non-species-specific conjugates (Proteins A and G horseradish peroxidase). Seroreactivity was determined by measuring optic density. Duplicate trials were performed for each sample for each antigen. The mean value of the duplicate trials was subtracted from the mean value of blank controls, and an optic density ratio value (OD) was determined using a sample of M. bovis-positive bovine serum.

Elephants were considered M. tuberculosis-positive if they had a M. tuberculosis-positive trunk culture. Elephants were considered non-infected if they had no M. tuberculosis-positive trunk cultures, had no contact with infected elephants within the last 5 yr, and had no travel history within the last 5 yr. In order to avoid false-positive results, elephants were excluded from analysis if intradermal tuberculin testing had been performed within the 6 mo prior to serum sampling. Of the 47 elephants, seven Asian elephants were culture-positive for M. tuberculosis; 25 Asian elephants and all 15 African elephants were considered non-infected.

Two-sample t-tests were used for detecting significant differences in OD values between infected and non-infected groups, as well as differences between Asian elephants and African elephants. Discriminant analysis was used to evaluate the six antigens simultaneously and thereby determine the linear combination of antigens that accurately predicted the true infection status of the most animals. The resulting classification functions were then used to calculate the percentage of animals that were correctly classified (i.e., specificity and sensitivity). Level of significance was P '0.05; values are reported with 95% confidence intervals.

Mean OD values were significantly higher in the infected elephants than in the non-infected elephants for all antigens except MPB. Discriminant analysis revealed that the best combination of antigens for differentiating infected and non-infected elephants were CF and RA. The specificity of the discriminant model was 100% (91.9-100%); sensitivity was also 100% (54.4-100%). Of the six antigens used in the panel, CF showed the highest individual specificity of 95% (83.0-100%), and sensitivity of 100% (54.4-100%). Mean OD values for all antigens except MPB and RA were significantly higher in non-infected Asian elephants than in non-infected African elephants; however these differences were substantially smaller than the differences between infected and non-infected elephants.

Subsequent to this investigation, several elephants were followed serologically over time. Seroreactivity in these elephants was based on a classification function determined by the previously described discriminant analysis.

 Group 1. Four elephants from a large herd had been exposed to two M. tuberculosis-positive elephants. These four animals were serologically sampled over the course of 2 yr. Three of the four were initially culture negative and seropositive; one of the four was culture negative and seronegative. Two of the three seropositive elephants had positive trunk cultures a few months after serum sampling. These results suggest that, in some cases, the ELISA may be able to detect infection, earlier than trunk culture.

 Group 2. Five Asian elephants from a large herd were serologically sampled over the course of 1 yr. These elephants had been exposed to a M. tuberculosis-positive elephant. All of these elephants remained culture negative and seronegative.

 Group 3. Four elephants from a non-infected herd were sampled over several months. All four elephants were seronegative initially; three of the four received intradermal tuberculin testing. One of the three that received intradermal tuberculin became seropositive within 2 wk, while the other two elephants remained seronegative. The seropositive animal slowly became seronegative over several months. These results, along with previous reports, suggest that intradermal tuberculin testing may alter serologic tests such as the ELISA.2

Six of the seven culture-positive elephants were treated with several-month-long multi-drug treatment regimens. One of these elephants was not followed over time. Three of the treated elephants became seronegative during the course of treatment, but were seropositive a few months after treatment had been stopped. Although they once again became seropositive, these three elephants have been culture-negative since treatment was instituted. In contrast, two of the six elephants showed even greater seroreactivity during the course of treatment. One of these two was culture-positive shortly after the treatment regimen was completed and a second course of treatment was instituted for this animal. The other elephant has not been culture-positive subsequent to treatment. The clinical significance of seroreactivity after treatment is unknown; the validity of the multiple antigen ELISA during and after anti-tuberculous treatment has not been fully evaluated and warrants further investigation.

Limitations such as sample size, compromised ability to ascertain each animal's true infection status, and absence of known-infected African elephants suggest that much additional research needs to be conducted regarding the use of this ELISA. Furthermore, the significant differences in seroreactivity between non-infected Asian and non-infected African elephants suggests that separate evaluation of these two species may be more appropriate in future investigations. However, the results indicate that this multiple-antigen ELISA may be a valuable screening test for M. tuberculosis infection in elephant herds.


We thank the many veterinarians, owners, caretakers, and managers of elephant-owning institutions that participated in this investigation. We also thank Dr. Richard Montali for his invaluable contributions.


1.  Gaborick CM, MD Salman, RP Ellis, J Triantis. 1996. Evaluation of a five-antigen ELISA for diagnosis of tuberculosis in cattle and Cervidae. J. Am. Vet. Med. Assoc. 209: 962-966.

2.  Montali RJ, LH Spelman, RC Cambre, D Chatterjee, S Mikota. 1998. Factors influencing interpretation of indirect testing methods for tuberculosis in elephants. Proc. Amer. Assoc. Zoo Vets. / Amer. Assoc. Wildl. Vets. Pp 109-112.

3.  Thoen CO, K Mills, MP Hopkins. 1980. Enzyme-linked protein A: an enzyme-linked immunosorbent assay reagent for detecting antibodies in tuberculous exotic animals. Am. J. Vet. Res. 41: 833-835.

4.  U.S. Department of Agriculture. 2000. Guidelines for the control of tuberculosis in elephants. Animal and Plant Health Inspection Service; Animal Care. Washington, D.C.

5.  VanTiem JS. 1992. The serologic diagnosis of Mycobacterium bovis. MS thesis. Department of Environmental Health, Colorado State University, Fort Collins, Colorado.

Speaker Information
(click the speaker's name to view other papers and abstracts submitted by this speaker)

R. Scott Larsen, DVM, MS

MAIN : All : Tuberculosis in Elephants
Powered By VIN


777 W. Covell Blvd., Davis, CA 95616


  • Toll Free: 800-700-4636
  • From UK: 01-45-222-6154
  • From anywhere: (1)-530-756-4881
  • From Australia: 02-6145-2357