Identification of tuberculosis in cattle slaughtered
in Mexico
Feliciano Milian-Suazo, MVZ, PhD; M. D. Salman, DVM, PhD; Carolina Ramirez, MVZ, MC;
Janet B. Payeur, DVM, PhD; Jack C. Rhyan, DVM, MS; Marco Santillan, MVZ
Objectives—To determine epidemiologic factors
associated with tuberculosis (TB) in dairy cattle
slaughtered in 6 important regions for milk production
in Mexico.
Animals—2,500 cattle.
Procedure—Tissue specimens with lesions typical of
TB were obtained during routine inspection of car-
casses at abbatoirs between July 1996 and January
1997. Infection with Mycobacterium organisms was
confirmed by histologic examination and bacteriolog-
ic culture. Species identification was made by use of
selective growth medium, conventional biochemical
tests, and radiometric procedures. Epidemiologic
information for affected cattle was obtained by per-
sonal interviews with cattle dealers and owners.
Results—400 (16%) of 2,500 cattle carcasses had
gross lesions typical of TB. Of the 400 infected cattle,
336 (84%) had lesions in ≥ 1 lymph node. Infection
was confirmed in 87% of cattle with gross lesions by
histologic examination, in 77% by bacteriologic culture
at a laboratory in the United States, and in 59% by
bacteriologic culture at a laboratory in Mexico. Most
cattle were adult females in fair to good body condi-
tion that came from large herds (> 500 cattle) and
were not included in the Mexican TB control program.
Conclusions and Clinical Relevance—Mean preva-
lence of lesions typical of TB in dairy cattle at 6 loca-
tions in Mexico was 16%. Mycobacterium infection
was confirmed by various techniques in most lesions.
Recognition of typical gross lesions at slaughter may
expedite TB control procedures. (Am J Vet Res 2000;
61:86–89)
T uberculosis (TB) is a zoonotic disease with high
prevalence among humans and animals in develop-
ing countries. According to the World Health
Organization, 80% of cases in humans are detected in
Third World countries.1 Tuberculosis in humans is pri-
marily caused by Mycobacterium tuberculosis; however,
M bovis, the cause of TB in cattle, is also responsible for
disease in humans.2,3 Most species of the genus
Mycobacterium are capable of infecting several animal
Received Aug 10, 1998.
Accepted Feb 4, 1999.
From the CENID-Microbiología, Instituto Nacional de
Investigaciones Forestales, Agrícolas y Pecuarias, Palo Alto, D. F.
Delegación Cuajimalpa, 05110 México (Milian-Suazo); the
Facultad de Ciencias Naturales, UAQ, 16 de Septiembre 63 Ote,
C. P. 76000, Querétaro, México (Milian-Suazo, Ramírez, Santillan);
the Department of Environmental Health, College of Veterinary
Medicine and Biological Sciences, Colorado State University, Ft
Collins, CO 80523 (Salman); and the National Veterinary Services
Laboratories, APHIS:USDA, 1800 Dayton Ave, PO Box 844, Ames,
IA 50010 (Payeur, Rhyan).
Address correspondence to Dr. Salman.
86
species. Four possible relationships between TB in
humans and cattle have been documented; these
include TB in humans caused by M tuberculosis, TB in
humans caused by M bovis, TB in cattle caused by
M bovis, and TB in cattle caused by M tuberculosis.a
Therefore, eradicating M bovis from livestock is of sub-
stantial importance to human health and welfare.
In Mexico, the prevalence of TB is reported to be
much higher in dairy cattle than in beef cattle.b The
threat to human health by dairy cattle with TB is sub-
stantial. In 1992, it was estimated that 40% of milk
produced in Mexico was consumed unpasteurized, and
5 to 8% of TB cases in humans may have been caused
by infection with M bovis.b It is not known, however,
whether infection was caused by ingestion of milk.
Tuberculosis is also directly or indirectly responsible
for large economic losses in the livestock industry. For
example, in Mexico, it is estimated that the dairy
industry loses approximately $40 million/y as a result
of culling because of TB.4 Indirectly, TB affects com-
mercial trade between Mexico and the United States. It
has been estimated that > 1 million Mexican steers
enter the United States each year; approximately 0.01%
of these steers are found to be infected at slaughter. In
the past 10 years, approximately 66% of tuberculous
cattle detected in the United States have been traced
back to Mexico.3
Official intervention to eliminate TB in Mexico is
based on identification of mycobacterial species by
culture in the laboratory. This method, however, is
slow, and intervention may be delayed from 4 to 8
weeks. A tentative diagnosis of TB in cattle can be
made when typical lesions are detected in cattle car-
casses at abattoirs. It has been suggested that in early
stages of a TB control program, when disease preva-
lence is high, great reliance can be placed on diagno-
sis by detection of macroscopic lesions during meat
examination5; however, more research is needed to
support this statement.
Most TB infections in cattle are acquired by inhala-
tion; 70 to 90% of lesions are found in the lymph nodes
of the head or thoracic cavity.5 Therefore, inspection of
these organs during slaughter may provide a more
rapid and reliable diagnosis of TB than bacteriologic
culture, depending on the diligence of the inspector.
Results of certain studies estimate that approximately
47% of infected carcasses are detected by routine
inspection,6 compared with 70 to 90% detected when
examination is performed more carefully. However, it
has also been estimated that despite the low sensitivity
of abattoir inspection, it is a cost-effective method for
monitoring large numbers of cattle for lesions of TB.
Corner et al6 found that careful examination of ≥ 6
AJVR, Vol 61, No. 1, January 2000
Unauthenticated | Downloaded 10/17/23 08:41 AM UTC
pairs of lymph nodes and the lungs can identify 95% of
cattle with macroscopic lesions of TB. The objective of
the study reported here was to describe epidemiologic
characteristics of TB diagnosed in cattle in 6 locations
in Mexico, including geographic location, organ affect-
ed, and lesion severity.
Materials and Methods
Specimens—Tissues with macroscopic lesions typical of
infection with M bovis were obtained after careful examina-
tion of carcasses of dairy cattle killed in selected abattoirs.
Specimens were collected by 1 investigator (FMS) between
July 1996 and January 1997; for each abattoir, day of speci-
men collection was chosen randomly, and number of visits
ranged from 2 to 60 because of geographic location and lim-
ited resources. On each visit, all tissues with macroscopic
lesions typical of TB were collected. Inspection included the
entire carcass with emphasis on the retropharyngeal,
mandibular, parotid, mediastinal, tracheobronchial, hepatic,
and mesenteric lymph nodes; lungs; and liver. Lymph nodes
were carefully incised with a knife, whereas lungs, liver,
spleen, and other organs were only incised if lesions were
detected on their surface. Specimens with lesions typical of
TB were carefully removed from the carcass to prevent cont-
amination, placed in plastic bags in a cooler with ice, and
frozen 3 to 4 hours after collection.
Geographic locations—Six geographic locations
(Aguascalientes, Queretaro, Tizayuca, Tlaquepaque, Torreon,
and Tulancingo) that correspond to 6 main Mexican dairy
regions were included in the study to represent different
parts of the country and to ensure diversity in origin of
Mycobacterium isolates. Abattoirs were chosen in areas with
high concentrations of dairy cattle to increase chances of
finding infected cattle, because it is known that in Mexico the
prevalence of the disease is higher in dairy cattle.b
Confirmation of infection—Specimen infection was
confirmed by histologic examination and bacteriologic cul-
ture of Mycobacterium spp. Bacteriologic culture of unpre-
served specimens was performed at the National Institute
for Forestry, Agriculture and Animal Research (INIFAP) in
Mexico. After specimens were thawed, bacteriologic culture
was performed by use of a standard procedurec; after decont-
amination by placement in 200 ml of sodium hypochlorite
(1:1000 dilution contained in a 600 ml plastic or glass
beaker) and removal of fat and other peripheral tissues, spec-
imens were blended for 2 to 3 minutes in 50 ml of a 4% solu-
tion of phenol red. Seven milliliters of the macerated tissue
suspension were placed in a test tube with 5 ml of NaOH
(0.5 N) for 10 minutes. Using a Pasteur pipette, 10 to 15
drops of HCI (6 N) were added to the tube with the macer-
ated tissue suspension until the mixture turned yellow; the
color of the mixture was returned to pale pink by adding
NaOH (0.1 N) drops. The suspension was centrifuged at
2,000 X g for 20 minutes and 90% of the overlying fluid dis-
carded. Two drops of residual fluid were placed in a culture
tube containing Lowenstein-Jensen medium and a culture
tube containing Stonebrink medium incubated at 37 C; these
tubes were observed for bacterial growth every 3 days during
the first week and every week for 8 weeks.
Bacteriologic culture was attempted from unpreserved
specimens, because preservation of infected tissue in sodium
borate reduces viability of mycobacteria6; however, randomly
chosen specimens were also submitted in a 6% solution of
sodium borate to the National Veterinary Services
Laboratory (NVSL) in Ames, Iowa, for bacteriologic culture
as described. Specimens were also submitted to NVSL in neu-
tral-buffered 10% formalin for histologic examination,
AJVR, Vol 61, No. 1, January 2000
processed in a routine manner, sectioned at 5 to 8 µm,
stained with H&E and fresh fuchsin acid-fast stain, and
examined under a light microscope.
Species identification was made by use of selective
growth medium, conventional biochemical tests (susceptibil-
ity to izoniacid, pyrazinamide, nitrase activity, and oxygen
preference),7 and radiometric procedures.
Collection of epidemiologic data—Epidemiologic data
was obtained by use of a data-collection formd that included
sex, age, breed, and body condition. Cattle owners were
interviewed to obtain information related to herd size, herd
of origin (if applicable), source of milk fed to calves, popula-
tion density, and status in the Mexican TB control program.
Information regarding origin of infected cattle was also
obtained by examination of certificates documenting move-
ment of the cattle and by interviews with cattle dealers.
Pathologic findings, such as organ affected, specific lymph
nodes affected, severity of lesions, and localized versus sys-
temic infection, were obtained during carcass inspection.
Localized infection was defined as infection that involved 1
organ or tissue located in any region of the carcass. Systemic
infection was defined as infection that involved several
organs or tissues in > 1 region of the carcass. Lesion severity
was determined on the basis of the proportion of the affected
organ. Body condition was classified by visually scoring mus-
cle mass of the carcass after the skin had been removed.
Data analyses—Descriptive statistics were determined
for categoric data. Mean number of specimens per day for
each abattoir was calculated by dividing the number of affect-
ed tissues by the number of days needed to obtain the speci-
mens. Expected percentage of infected cattle for each abattoir
was estimated by dividing the number of infected cattle by
the product of the mean number of cattle inspected per day
and the mean number of specimen collection days and mul-
tiplying by 100. Agreement among results for the 3 diagnos-
tic procedures (bacteriologic culture performed in Mexico,
bacteriologic culture performed at NVSL, and histologic
examination) was evaluated by the kappa test; 95% confi-
dence intervals (CI) for kappa values were also estimated.8
A P value < 0.05 was considered significant.
Results
Of the 2,500 cattle carcasses inspected in 6 abattoirs,
400 (16%) had lesions typical of TB infection. Most affect-
ed cattle were Holsteins (97%), females (92%), > 2 years
old (93%), and in fair or good body condition (75%).
Ninety percent of affected cattle had localized lesions
affecting a single organ or tissue. Thirty-six of 78 (46%)
cattle owners were interviewed. Most affected cattle
(74%) were from herds that contained > 500 cattle, used
replacement cattle from the United States and Canada
(63%), used milk from the same herd to feed calves
(71%), and had moderate or low population density (> 25
m2/animal; 70%). Most cattle owners (67%) did not par-
ticipate in the Mexican program for control of TB.
Number of specimens obtained per day for each
abattoir ranged from 1.3 to 10, whereas expected per-
centage of infected cattle ranged from 11 to 24%
(Table 1). More cattle (336/400; 84%) had lesions with-
in lymph nodes than other organs or tissues; lesions
were often observed in retropharyngeal (183/336;
49.2%), mediastinal (91/336; 24.4%), mandibular
(24/336; 6.4%), tracheobronchial (12/336; 3%), mesen-
teric (9/336;2.4%), or a combination of these lymph
nodes. Lesions were less commonly observed in lung
87
Unauthenticated | Downloaded 10/17/23 08:41 AM UTC
 Table 1—Descriptive statistics for 400 dairy cattle with typical lesions of Mycobacterium bovis infection
detected by inspection of carcasses at slaughterin 7 abattoirs in 6 geographic locations in Mexico
Specimens
Location n (% of total)
Aguascalientes 38 (10)
Gómez Palacio 6 (1)
Querétaro 209 (52)
Tizayuca 3 3 (8)
Tlaquepaque 56 (15)
Torreón 50 (12)
Tulancingo 8 (2)
*No. of specimens
No. of collection days 3 mean No. of cattle inspected/d
(6/400; 1.5%), udder (6/400; 1.5%), liver (3/400; 0.7%),
and sternum cartilage (1/400; 0.2%).
Bacteriologic culture was performed at INIFAP in
Mexico on all 400 specimens and at the NVSL on 228
specimens; 268 specimens were examined microscopi-
cally at the NVSL. Bacteriologic culture results were
positive for 59% of specimens tested at INIFAP and
77% of specimens tested at the NVSL. Eighty-seven
percent of specimens examined histologically had
lesions that were typical of TB and had acid-fast
microorganisms. Agreement among test results for 228
specimens tested by all 3 procedures was poor; signifi-
cant agreement between results of histologic examina-
tion and bacteriologic culture performed at INIFAP
(kappa, 0.06; 95% CI, –0.06 to 0.20) and between
results of the 2 culture procedures (kappa, 0.08; 95%
CI, –0.05 to 0.22) was not detected. Agreement
between results of histologic examination and bacteri-
ologic culture performed at the NVSL was significant
(kappa, 0.29; 95% CI, 0.13 to 0.46; P < 0.05).
Discussion
The 16% prevalence of dairy cattle with lesions
typical of TB in the study reported here is similar to the
16% prevalence for reactors to the tuberculin test
reported by the Mexican Commission for the Control
of Tuberculosisb; however, cattle and abattoirs in our
study were not randomly selected. The high proportion
of positive results for bacteriologic culture and histo-
logic examination indicate that gross inspection of car-
casses at slaughter may be a reliable method for esti-
mation of the prevalence of TB.
Tracing cattle origin as a means to identify sources
of infection was problematic. For the study reported
here, we attempted to obtain this information by exam-
ining certificates that documented movement of the
cattle and by interviews with cattle dealers. Certificates
were not available for most of the cattle, or certificates
listed multiple cattle, which made tracing individual
cattle impossible. Verbal interview was a better but not
always successful approach, because cattle dealers had
a high degree of loyalty to cattle owners and were often
reluctant to reveal economically important information.
A more reliable method to trace cattle to the farm of ori-
gin is needed to successfully identify infected herds.
During interviews with cattle owners, it was
observed that dairy cattle owners were not as con-
88
Expected
Mean percentage
Collection Mean No. No. cattle infected
days specimens/d inspected/d cattle*
12 3.2 20 16
2 3.0 15 20
60 3.5 20 17
8 4.1 17 24
8 6.9 45 16
5 10.0 80 12
6 1.3 12 11
3 100
cerned as beef cattle owners about participating in the
national program for control of TB. In our opinion, this
lack of interest in the TB control program may be the
result of 3 financial considerations. The prevalence of
the disease is high, and the number of reactors to the
skin test would be large, resulting in high costs for
culling cattle. Furthermore, compensation for cattle
killed under the program is not available, and, from the
cattle owners’ point of view, a commercial advantage
for eliminating the disease is not evident, because a
special price for milk that originates from TB-free
farms has not been established as an incentive. In addi-
tion, cattle owners claim that contaminated milk is not
a hazard to public health, because the bacteria are
killed during pasteurization.
The expected number of infected cattle was simi-
lar for the different geographic locations of the abat-
toirs, possibly indicating that the prevalence of TB in
dairy cattle may be similar throughout the country.
This information also indicates that a single national
strategy for controlling the disease may be feasible.
It has been reported that approximately 90% of
infected cattle can be detected by careful examination of
3 pairs 5 or 6 pairs6 of lymph nodes at slaughter. Results
of the study reported here support that finding; 84% of
infected cattle had lymph node involvement. The
slightly higher prevalence reported by Corner5 may be
attributed to optimal conditions for careful examination
of carcasses; in our study, organs were often removed
before a thorough inspection could be completed.
Poor agreement among results obtained by the 3
diagnostic procedures requires careful consideration.
At the NVSL, histologic examination yielded more pos-
itive results than did bacteriologic culture. It is possi-
ble that certain specimens for which positive histolog-
ic results were obtained were not infected by M bovis
but with a different Mycobacterium sp that would not
replicate in the M bovis-selective culture medium.
Conversely, if all specimens reported positive by histo-
logic examination were infected with M bovis, then
many specimens are misdiagnosed by bacteriologic
culture. If that is true, and disease prevalence for a spe-
cific region is based on culture results, disease preva-
lence in that region may be underestimated. The possi-
bility that a certain proportion of mycobacterial organ-
isms may be killed by disinfection during processing
for culture requires further investigation.
AJVR, Vol 61, No. 1, January 2000
Unauthenticated | Downloaded 10/17/23 08:41 AM UTC
 At present, definitive diagnosis of TB requires cul-
ture of the organism. Recognition of macroscopic lesions
in carcasses at slaughter may provide a diagnosis more
promptly and may avoid delays in disease intervention
while awaiting culture results from a laboratory.
a
Committee on Bovine Tuberculosis, Board of Agriculture. National
Academy Press, 2101 Constitution Ave, NW Washington, DC
20418, 1994.
b
Secretaría de Ganaderia, Direccion General de Salud Animal,
Secretaría de Agricultura y Desarrollo Rural, México: Personal
communication, 1994.
c
Payeur JB, Jarnagin JL, Marquardt JJ, et al. Laboratory methods in
veterinary mycobacteriology for isolation and identification of
mycobacteria. NVSL-USDA:APHIS, Ames, Iowa, 1992.
d
Form available from the senior author upon request.
References
1. Styblo K, Rouillon A. Tuberculosis in the world estimated
global incidence of smear-positive pulmonary tuberculosis.
AJVR, Vol 61, No. 1, January 2000
Unreliability of officially reported figures on tuberculosis. Bull Int
Union Tuberc 1981;56:118–126.
2. Grange JM, Yates DM. Zoonotic aspects of Mycobacterium
bovis infection. Vet Microbiol 1994;40:137–151.
3. Essey MA, Koller MA. Status of bovine tuberculosis in
North America. Vet Microbiol 1994;40:15–22.
4. Valdespino OJR. Impacto de la brucellosis y tuberculosis
bovina durante 1992 en México, in Proceedings. Pan Am Cong Vet Sci
1994;102–120.
5. Corner LA. Post mortem diagnosis of Mycobacterium bovis
infection in cattle. Vet Microbiol 1994;40:53–63.
6. Corner LA, Melville L, McCubbin K, et al. Efficiency of
inspection procedure for the detection of tuberculosis lesions in cat-
tle. Aust Vet J 1990;67:389–392.
7. Collins CH, Yates MD, Grange JM. Subdivision of
Mycobacterium tuberculosis into five variants for epidemiological pur-
poses: methods and nomenclature. J Hyg (Lond) 1982; 89:235–242.
8. Norman GR, Streiner DL. Measures of association for cate-
gorical data. In: Biostatistics: the bare essentials. St Louis: Mosby Year
Book Inc, 1994;163–166.
89
Unauthenticated | Downloaded 10/17/23 08:41 AM UTC