Epidemiologic Investigation

The exotic animal farm was visited on numerous occasions to evaluate the type and degree of contact between elephants and employees. The farm, located in a rural area and surrounded by barbed wire and trees, originally housed 18 Asian and 2 African elephants. Thirteen elephants were tethered on a chain in one large barn, four were housed in a separate large room (two in a common stall), and a baby elephant was in a third room with 5-6 tigers. A separate barn housed approximately 80 tigers.

TB Screening of Employees

The animal handlers (trainers and caretakers) who had direct contact with the elephants were administered purified protein derivative (PPD) skin tests. Initial screening was performed in August 1996, with subsequent screenings in December 1996 and March, June, and September of 1997. Testing was performed by the McHenry County Department of Health, except in two handlers who had subsequent skin tests performed elsewhere. As part of the screening process, handlers were questioned about their risk factors for TB, including previous bacillus Calmette-Guérin (BCG) vaccination.

Handlers were tested by the two-step method using 5 tuberculin units of PPD (0.1 ml) by intradermal injection in the flexor surface of the forearm. A positive result was defined as an induration of >5 mm. Handlers with positive skin tests were evaluated by a TB health-care worker and had chest radiographs taken. Sputum samples from any handler with a chest radiograph consistent with TB were submitted to the Illinois Department of Public Health Laboratory. Samples were examined by direct microscopy for acid-fast organisms, stained with fluorochrome, and processed for culture by standard methods.

Examination of Isolates

The human isolate and the four elephant isolates were sent to the National Tuberculosis Genotyping and Surveillance Network at the Michigan Community Public Health Agency for restriction fragment length polymorphism (RFLP) analysis. Southern blots of Pvu II restricted whole chromosomal DNA, resolved in 1% agarose gels, were probed with a DNA fragment corresponding to the right side of IS6110 and detected by chemiluminescence (11). The number and size of the hybridizing fragments for each isolate were compared in the same gel. Isolates with identical RFLP patterns or with < 2 band differences were considered to represent the same strain. Additionally, Pvu II digested DNA was similarly typed after probing with the repetitive element TBN12.

Epidemiologic Findings

Elephant handlers worked in very close proximity with the elephants around the clock, whereas tiger handlers had little direct contact with the elephants. Most of the elephant handlers lived on the farm in a separate section of the barn; four lived in trailers on the grounds. The handlers' living quarters had a separate ventilation system from the elephants'; however, the doors between the two quarters were open for unknown periods. Handlers indicated that they held social events in a building connected to the elephant barn.

Necropsies of elephants were performed on the farm and were attended by a number of elephant and tiger handlers (including the handler with the active case). The necropsy of the elephant that died in 1994, also performed on the farm, showed caseous necrosis of the lungs and pleural exudates whose culture yielded M. tuberculosis.

In addition to the three elephants that died of M. tuberculosis infection, a fourth living elephant was also infected with the mycobacterium; this infection was diagnosed in late December 1996 from a trunk culture obtained in October 1996. Subsequent cultures from this and the other animals have been negative for mycobacteria. Another elephant from this farm died of M. tuberculosis infection in 1981 (5), but contact between this elephant and the present herd or any of the handlers could not be established.

Twenty-two handlers at the exotic animal farm had moderate to frequent contact with the infected animals; 12 were elephant handlers and 10 were tiger handlers. Initial PPD testing was performed for 14 handlers in August 1996, 2 in October 1996, and 5 in December 1996. One who was PPD-positive in November 1995 reported receiving BCG more than 10 years before.

Eleven (50%) of 22 handlers were found PPD-positive as part of this investigation. Eight of the 11 had positive PPD skin test results upon initial testing, with a median induration of 12 mm (range, 10 to 19 mm). Four of the eight were elephant handlers and four were tiger handlers. The skin test reaction of three handlers converted from negative to positive with a median induration of 12 mm (range, 8 to 15 mm). The three PPD converters were initially tested in August 1996; one was positive upon retesting in January 1997, and two tested positive in April 1997 (Table).

Eight of the 11 handlers reported that they had negative skin tests in the past and had not received BCG. The other three reported some type of reaction from a previous skin test in the past but did not know the results. All three also reported receiving BCG more than 10 years before. Eight of the 10 handlers with negative PPD skin tests had at least one negative follow-up test at 3 months; two left the farm and did not receive follow-up testing.

The attack rates were approximately equal for the elephant and tiger handlers. Of the 12 elephant handlers tested, 6 (50%) were PPD-positive with two conversions documented in April 1997; of the 9 tiger handlers, 5 (56%) were PPD-positive, with one conversion documented in January 1997. Overall, a very high rate (52%) of handlers tested positive.

All 12 handlers with positive PPDs (including the one with the known positive PPD) received an evaluation and chest radiograph; one had irregular nodules and interstitial changes in the right apex without retraction of the lungs, consistent with active TB, but no cough, chest pain, fever, night sweats, weight loss, or fatigue.

Three sputum samples were smear-negative for acid-fast bacilli, although one yielded M. tuberculosis upon culture. Isoniazid (INH), rifampin, pyrazinamide, and ethambutol treatment was initiated in September 1996, and after 2 months, was reduced to INH and rifampin when the isolate showed no resistance to antituberculous medications. Subsequent chest radiographs revealed improvement or clearing of the initial lesions. Nine of the remaining 11 PPD-positive handlers were prescribed INH prophylaxis; two declined because of the risk for adverse reactions.