m

. 3.

 

 

 

 

"alum.“

v.32

 

 

 

‘ -

.1

 

 

 

 

 

. u
. Kit
I 33%.:

1.51!

If

I ..x . I...
x 3.. a is.

‘ in
“flag... _

I .. huh...) I:
rfiiurirnhn...
3:: . I

 

v

.:

.._ ghfifiwafimé

.. as. . 333, . _ ”$3 3%.
:zu_.;...4...::._._. , x V “M... . ‘ . .. . ‘ . , V . . t I

it. 11.57 I.(

 

w 4 LIBRARY
5 c c 3 50 90 Michigan State
University

This is to certify that the
thesis entitled

THE EFFECT OF MYCOBACTERIUM AVIUM SUBSP.
PARA TUBERCULOSIS ON THE CAUDAL FOLD
TUBERCULIN (CFT) AND GAMMA INTERFERON (y-IFN)
TESTS FOR BOVINE TUBERCULOSIS

presented by

JOHN RICHARD DUNN

has been accepted towards fulfillment
of the requirements for the

MS. degree in LARGE ANIMAL CLINICAL
SCIENCES (EPIDEMIOLOGY)

 

 

WWI

Major Professor’s Signature

December 5, 2003

 

Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

Wslliiamfl

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRC/DateDuepGS—p. 1 5

 

 

THE EFFECT OF M YCOBA C TERI UM A VI UM SUBSP. PARA TUBERCULOSIS ON
THE CAUDAL FOLD TUBERCULIN (CFT) AND GAMMA INTERFERON (y-IFN)
TESTS FOR BOVINE TUBERCULOSIS

By

John Richard Dunn

A THESIS

Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of

MASTER OF SCIENCE

Department of Large Animal Clinical Sciences
(Epidemiology)

2003

ABSTRACT
THE EFFECT OF M YCOBA C TERI UM A VI UM SUBSP. PARA TUBERCULOSIS ON

THE CAUDAL FOLD TUBERCULIN (CFT) AND GAMMA INTERFERON (y—IFN)
TESTS FOR BOVINE TUBERCULOSIS

By

John Richard Dunn

The studies described were performed to determine whether cattle testing positive
for Mycobacterium avium subsp. paratuberculosis (M. paratuberculosis) were more
likely to have false positive results on the caudal fold tuberculin (CF T) test and the y-IFN
assay for bovine tuberculosis (TB) than negative cattle. M. paratuberculosis fecal culture
and antibody ELISA were performed for 1,043 lestein cattle from 10 Michigan herds
on the day the CFT test was read. Blood samples were also used to test for y-IFN
response following overnight stimulation with phosphate buffered saline (control) and
purified protein derivative (PPD) from M. bovis and M. avium.

The total number of CFT suspects from all herds was 180 (17.3%), and 8 cattle
(0.8%) were positive for y-IFN stimulated by M. bovis. Cattle testing positive for M.
paratuberculosis, as measured by a positive M. paratuberculosis fecal culture or antibody
ELISA test, appeared to have an increased likelihood of false positive results on the CF T
test, although the association was not statistically significant. Further studies involving a
larger sample size need to be conducted to confirm these findings. No significant
association was found between cattle testing positive by M. paratuberculosis fecal culture
or antibody ELISA, and positive results of the y-IFN assay for TB. Cattle positive for y-
IFN stimulated by M. avium were more likely to be CFT suspects than negative cattle,

which may be related to early stages of J ohne’s disease.

In loving memory of my father,
Richard J. “Dick” Dunn
1930-2003

iii

ACKNOWLEDGMENTS

I would like to thank the USDA-National Research Initiative, Michigan State
University-Animal Initiative Grant, and the Population Medicine Center for their
financial support of this project.

A special thank you also goes out to the herd owners and veterinarians that gave
us access to their herds and were patient to let us collect our samples during TB testing,
and RoseAnn Miller in the Population Medicine Center for data analysis assistance.

My committee was an enormous support from the beginning to the end of this
project. Thank you, Dan Grooms for your time spent assisting with sampling and
introducing new ideas and perspectives as I was sorting through the results. Thank you,
Steve Bolin for your assistance running the antibody ELISA and y-IFN tests, and time
spent working with me in the lab. Thank you also to Carole Bolin and Colleen Bruning-
Fann for your expertise and assistance with running fecal cultures and organizing the
field portion of this study.

Finally, I would like to especially thank my main advisor, John Kaneene. I have
truly enjoyed the time spent working with you and take pride knowing I’ve learned from
one of the best. You have been an important role model to me as a veterinary

epidemiologist and as a human being.

iv

TABLE OF CONTENTS

LIST OF TABLES ............................................................................ vii
LIST OF FIGURES ........................................................................... ix
KEY TO ABBREVIATIONS ............................................................... x
INTRODUCTION ............................................................................ 1
Purpose ................................................................................. 1
Objectives ............................................................................. 1
Hypotheses Tested ................................................................... 2
Overview .............................................................................. 2
CHAPTER 1
LITERATURE REVIEW .................................................................... 3
Johne’s disease ....................................................................... 3
Diagnostic tests for J ohne’s disease ............................................... 6
Prevalence of Johne’s disease ...................................................... 8
Bovine tuberculosis .................................................................. 9
Screening tests for bovine tuberculosis ........................................... 10
Prevalence of TB ..................................................................... 14
Conclusion ............................................................................ 15
CHAPTER 2

THE EFFECT OF INFECTION WITH M Y COBA C TERI UM A VI UM SUBSP.
PARA TUBERCULOSIS ON THE CAUDAL FOLD TUBERCULIN (CFT) TEST FOR

BOVINE TUBERCULOSIS IN CATTLE ................................................ 16
Abstract ................................................................................ 16
Introduction ........................................................................... 17
Methods ............................................................................... 17
Results ................................................................................. 20
Discussion ............................................................................. 22

CHAPTER 3

THE EFFECT OF INFECTION WITH MY COBA C T ERIUM A VI UM SUBSP.
PARA TUBERCULOSIS ON THE GAMMA INTERFERON (y-IFN) ASSAY FOR

BOVINE TUBERCULOSIS IN CATTLE ................................................. 32
Abstract ................................................................................ 32
Introduction ........................................................................... 33
Methods ............................................................................... 33
Results ................................................................................. 36
Discussion ............................................................................. 38

CHAPTER 4

EVALUATION OF THE ASSOCIATION BETWEEN THE GAMMA
INTERFERON (y-IFN) ASSAY, STIMULATED BY M Y COBA C TERI UM
A VI UM, AND THE CAUDAL FOLD TUBERCULIN (CFT) TEST AND

JOHNE’S DISEASE IN CATTLE ......................................................... 45
Abstract ................................................................................ 45
Introduction ........................................................................... 46
Methods ............................................................................... 47
Results ................................................................................. 51
Discussion ............................................................................. 52

SUMMARY AND CONCLUSIONS ...................................................... 63

APPENDIX .................................................................................... 65
Detailed Laboratory Protocol ...................................................... 66

REFERENCES ................................................................................ 70

vi

LIST OF TABLES

Table 1-1: Sensitivity and specificity for M. paratuberculosis fecal culture 7

Table 1-2: Sensitivity and specificity for Parachek M. paratuberculosis

antibody ELISA ............................................................................... 8

Table 1-3: Sensitivity and specificity for the CFT test ................................... 13
Table 1-4: Sensitivity and specificity for Bovigam y-IFN assay ........................ 14
Table 2-1: Description and descriptive statistics for risk factors evaluated ........... 25

Table 2-2: Descriptive analysis of CFT test results based on positive Johne’s
disease tests ..................................................................................... 26

Table 2-3: Results of a multivariable logistic regression analysis of the effect of
individual and herd level M. paratuberculosis test results on the CFT test as the
outcome (classified as suspect or negative) ................................................ 27
Table 2-4: Sample size required to detect a true association based on the actual
percentage of CF T positive cattle and M. paratuberculosis test results, and power
calculations based on the actual sample size ............................................... 28

Table 3-1: Description and descriptive statistics for risk factors evaluated .......... 41

Table 3-2: Descriptive analysis of TB test results based on positive J ohne’s
disease tests .................................................................................... 42

Table 3-3: Results of a multivariable logistic regression analysis of the effect of
individual J ohne’s Disease test results and herd levels on the y-IFN test as the
outcome (classified as positive or negative) ............................................... 43

Table 4-1: Description and descriptive statistics for risk factors evaluated ........... 57

Table 4-2: Descriptive analysis of M. avium y-IFN results based on positive
Johne’s disease test results ................................................................... 58

Table 4-3: Descriptive analysis of CFT test results based on positive Johne’s
disease test results ............................................................................. 59

Table 4-4: Results of a multivariable logistic regression analysis of the effect of
Johne’s disease test results on the M. avium y-IFN test as the outcome. . . . . . . . . . ..6O

vii

Table 4-5: Results of a multivariable logistic regression analysis of the effect of
individual Johne’s Disease test results and herd levels on the CFT test as the
outcome (classified as suspect or negative) ................................................ 6]

viii

LIST OF FIGURES

Figure 1-1: Johne’s disease progression based on clinical presentation and
shedding of M. paratuberculosis, including an approximation of the number
of cattle in stages I-III based on one animal in stage IV (Whitlock and

Buergelt, 1996) ................................................................................ 5
Figure 1-2: Agricultural districts of Michigan showing the counties with numbers

of herds used for our‘ study ................................................................... 9
Figure 2-1: Graph of the percentage of positive test results by herd ................... 29

Figure 2-2: Percentage of cattle positive for M. paratuberculosis fecal culture and

antibody ELISA tests separated by CF T status ........................................... 30
Figure 2-3: Percentage of cattle with positive CFT results based on

positive M. paratuberculosis test results ................................................... 31
Figure 3-1: Graph of the percentage of positive test results by herd ................... 44
Figure 4-1: Graph of the percentage of positive test results by herd ................... 62

ix

KEY TO ABBREVIATIONS

 

 

 

 

 

CCT test Comparative Cervical Tuberculin test
CF T test Caudal Fold Tuberculin test

ELISA Enzyme-Linked Irnmunosorbent Assay
M. avium Mycobacterium avium

M. bovis Mycobacterium bovis

 

M. paratuberculosis

Mycobacterium avium subsp. paratuberculosis

 

 

 

 

 

 

 

 

MDA Michigan Department of Agriculture
OD Optical Density

PPD Purified Protein Derivative

TB Bovine tuberculosis

Th1 cells Type 1 Helper T lymphocyte

USDA United States Department of Agriculture
y-IFN Gamma-interferon

 

 

 

INTRODUCTION
Purpose
J ohne’s disease and bovine tuberculosis (TB) are both important diseases in cattle
caused by Mycobacterium avium subsp. paratuberculosis (M. paratuberculosis) and
Mycobacterium bovis, respectively. The caudal fold tuberculin (CFT) test and gamma-
interferon ('y-IFN) assay are screening tests for detection of TB in cattle. Since Johne’s
disease and TB are both caused by mycobacteria, it is speculated that cross-reactivity
may occur on the CFT test and y—IFN assay causing false positives in animals infected
with Johne’s disease. The effect of infection with M. paratuberculosis on the CFT test
and y-IFN assay is important for determining confidence of positive results among herds
infected with J ohne’s disease.
Objectives
1. Compare CFT test results for TB between cattle testing positive and negative
for M. paratuberculosis.
2. Compare y-IF N assay results for TB between cattle testing positive and
negative for M. paratuberculosis.
3. Evaluate the association between y-IFN assay results stimulated by M. avium
PPD with test results for M. paratuberculosis.
4. Compare CFT test results for TB between cattle with a positive response for

y-IFN stimulated by M. avium PPD and cattle with a negative response.

Hypotheses Tested

1.

Overview

Cattle testing positive for M. paratuberculosis will have a higher likelihood of
false positive results on the CF T test compared to cattle testing negative for
M. paratuberculosis.

Cattle testing positive for M. paratuberculosis will have a higher likelihood of
false positive results on the y-IFN assay for TB compared to cattle testing

negative for M. paratuberculosis.

. Cattle testing positive for M. paratuberculosis will have a higher likelihood of

positive results for y-IFN stimulated by M. avium compared to cattle testing
negative for M. paratuberculosis.

Cattle with a positive response of y-IFN stimulated with M. avium will have a
higher likelihood of false positive results on the CFT test compared to cattle

with a negative response.

Chapter 1 is a literature review of J ohne’s disease and TB with primary emphasis

on ante-mortem diagnostic tests for each disease. Chapter 2 details a cross-sectional

study performed in Michigan evaluating the effect of infection with M. paratuberculosis

on the results of the CFT test in cattle. Chapter 3 is from the same study performed in

Michigan, but this chapter focuses on the effect of infection with M. paratuberculosis on

the results of the y-IFN assay for TB in cattle. Chapter 4 evaluates the y-IF N assay,

specifically the response from M. avium PPD stimulation and the association with M.

paratuberculosis infection. The conclusion summarizes the findings as a whole and

presents the relevance of these findings and potential recommendations.

CHAPTER 1
LITERATURE REVIEW
Johne’s disease
Mycobacterium avium subsp. paratuberculosis (M. paratuberculosis) is the
causative agent for Johne’s disease in cattle. Johne’s disease is a chronic disease
characterized by gradual weight loss despite a normal appetite. Decrease in production is
common among infected cattle, as well as a decrease in reproductive performance. Cattle
positive for M. paratuberculosis antibody ELISA have been to shown to have a 28-day
increase in days open compared to negative cows (J ohnson-Ifearulundu et al., 2000).
Calves are the most susceptible to infection during their first year of life, but clinical
signs generally do not develop until the animal is greater than two years old.
The terminal part of the ileum is the most frequent site of lesions associated with
J ohne’s disease. Once ingested, M. paratuberculosis is taken up in the ileum by
microfold cells (M cells), which are specialized epithelial cells associated with Peyer’s
patches and lymphoid follicles. M cells are efficient antigen-presenting cells that present
the bacteria directly to lymphocytes within the Peyer’s patch. The immune system
responds by recruiting macrophages and developing giant cells (Buergelt et al., 1978).
The intestinal wall develops a corrugated appearance from increasingly thickened villi as
the organism continues to multiply, and gradually the intestinal wall loses its ability to
absorb nutrients. Eventually the intestinal wall begins to leak protein from blood into the
intestine and further lowers the animals absorbed nutrients, despite having a normal
appetite. The malabsorption and protein-losing enteropathy result in the

hypoproteinemia, which in advanced disease commonly results in cachexia and

pipestream diarrhea. Interrnandibular edema also develops in rare cases. In addition to
the intestine, lesions have been observed in the liver, spleen, lungs, kidneys, uterus,
placenta, and nonmesenteric lymph nodes.

The primary mode of transmission for M. paratuberculosis is fecal-oral, most
typically from manure or fecal contaminated feed or water (Sweeney, 1996). Neonates
may ingest contaminated manure by suckling from a manure-contaminated teat or by
other manure-contaminated sources in their birth environment. Contaminated colostrum
or milk from the infected darn or originating from other infected cattle are other potential
sources of ingestion (Streeter et al., 1995). In utero infection may also occur in 25% of
calves born to cows with clinical signs, and 18% from asymptomatic infected cows
(Sweeney et al., 1992). Difficulty in detecting subclinical Johne’s disease leads to the
propagation of the disease within a herd, despite common disease preventative
procedures.

Johne’s disease is commonly divided into four stages of infection (Figure 1-1).
Stage I is described as the silent infection stage (Whitlock and Buergelt, 1996). This
stage primarily involves young calves, as calves less than 30 days of age are most
susceptible with the majority of cattle becoming infected before four months of age
(Hagan, 1938). This first stage generally lasts greater than two years and may surpass ten
years. Laboratory tests rarely detect cattle in stage I of infection. Stage II is referred to
as the inapparent carrier adult stage. Cattle with subclinical disease can have an
increased y-IFN response by sensitized T cells or an increased antibody response to M.
paratuberculosis (Stabel, 1996). Infected animals may also potentially be positive on

fecal culture in Stage 11. Animals in Stage II do not have diarrhea or weight loss but are

now shedding organisms in their manure and infecting other animals on the farm. Stage
IH is defined as clinical disease. Animals in this stage have a normal appetite and vital
signs but are beginning chronic diarrhea and gradual weight loss. Infected animals may
test positive forM paratuberculosis on fecal culture or antibody ELISA and agar gel
immunodiffusion (AGID). The final progression of infection, Stage IV, is advanced
clinical disease. Infected animals are normally lethargic, emaciated, and have pipestream
diarrhea and intermandibular edema. The animal’s condition typically deteriorates within

one or two days. Death can result from extreme dehydration and cachexia.

Figure 1-1: Johne’s disease progression based on clinical presentation and shedding
of M. paratuberculosis, including an approximation of the number of cattle in stages
[-111 based on one animal in stage IV (W hitlock and Buergelt, 1996).

 
    
   

Stage
Adv ced clinica isease
1 animal

 

Stage III
Clinic disease, shedding high It mbers
1-2 animals

 

Stage II
In parent carrier adults, shedding low nu ers
6-8 animals

 

Stage I
Silent infection of calves, non-shedding
10-15 animals

 

Uninfected

 

Early stages of M. paratuberculosis infection are generally characterized by a
cell-mediated immune response followed by a predominately humoral response in later
stages of the disease (Bendixen, 1978; Stabel, 2000). In the final stages of infection,
anergy may develop in which there will be no immune response of any type to specific
antigens (Stabel, 1996). The easiest stage in which to diagnose infection with M.
paratuberculosis is during the clinical phase when overt clinical signs exist and
antibodies can be detected using a variety of methods. The greatest challenge exists with
detecting M. paratuberculosis during the subclinical stage, before a humoral response has

begun.

Diagnostic tests for Johne’s disease

Ante-mortem diagnostic tests for Johne’s include those for direct bacteria
detection, serological tests for the detection of antibody, and assays for cellular
immunity. Methods for direct bacteria detection consist of standard fecal culture,
radiometric fecal culture (BACTEC), and DNA probe (PCR). Direct detection of the
bacteria is the definitive method for detecting M. paratuberculosis (specificity is close to
100%). Unfortunately, M. paratuberculosis bacteria are shed intermittently with
shedding increasing as the animal progresses towards the clinical phase of disease. In
addition, the standard fecal culture takes 12-16 weeks, costs more than the rapidly run
ELISA, and has a relatively low sensitivity (see Table l-l). Radiometric fecal culture is
faster than standard culture, however the special technique is expensive, requires special

equipment, and involves the handling of radioisotopes. The DNA probe is the fastest of

the three methods for directly detecting bacteria, however the test is considered less

sensitive and more expensive compared to fecal culture.

Table 1-1: Sensitivity and specificity for standard M. paratuberculosis fecal culture

 

 

Author Sample size Sensitivity
Sockett et al., 1992 182 infected* 45.1%

111 fecal sheddersT 73.8%
Zimmer et al., 1999 75 clinically affected: 84.0%

57 subclinically infected§ 87.7%
Collins et al., 1990 75 positive1l 60%

 

*Isolation of M. paratuberculosis from any fecal or tissue sample
T Positive on either radiometric fecal culture, conventional fecal culture, or commercial
PCR/DNA probe
ICattle showing typical signs in herds known to be affected by J ohne’s disease
§Cattle positive on either Ziehl-Neelsen staining, fecal culture, or DNA-Probe test
{I Positive on either radiometric fecal culture or conventional fecal culture

There are three serological tests used to detect antibodies for M. paratuberculosis.
Enzyme-linked immunosorbent assay (ELISA) is the most sensitive, followed by agar-gel
immunodiffusion (AGID), and complement fixation (CF). The sensitivity of these tests,
however, has several problems related to the course of M. paratuberculosis infections.
One limitation is the relative lateness of the humoral immune response during the course
of Johne’s disease infection (Kreeger, 1991). The introduction of absorbed ELISA’s
improved the second limitation, which was cross-reactivity with other environmental
bacterial species that can potentially reduce specificity of serological tests (Ridge et al.,

1991). The antibody ELISA test is fast and low-cost, however, reported sensitivities are

even lower than those reported for M. paratuberculosis fecal culture (see Table 1-2).

Table 1-2: Sensitivity and specificity for Parachek* M. paratuberculosis antibody
ELISA

 

 

Author Sample size Sensitivity Specificity
Reichel et al., 1999 106 positive, 341 negative 31.1% 97.9%
Sockett etal., 1992 177 positive, 196 negative 43.4% 99.0%
Collins et al., 1991 150 positive, 196 negative 47.3% 99.0%
Ridge et al., 1991 150 positive, 1,000 negative 47.3% 99.8%
Ellis et al., 1998 1000 negative 99.7%

 

*Parachekm, BioCor Animal Health, Omaha NE

The third category of tests used to detect M. paratuberculosis infection includes
those that measure a cell-mediated immune response. These tests include the Johnin skin
test and y-IFN assay. The y-IFN assay is a measurement of y-IFN released by sensitized
T lymphocytes stimulated by johnin PPD. The 'y-IFN assay has a high number of false
positives with specificity in a recent study ranging from 66.1-93.6% depending on the

algorithm used for the calculation (Kalis et al., 2003).

Prevalence of Johne’s disease

M. paratuberculosis is prevalent in the United States with estimates of individual
prevalence ranging from 16-20% in different parts of the country (Braun et al., 1990;
Whipple et al., 1991; Thoen and Baum, 1988; Kreeger, 1991). A recent study in
Michigan found that 55% of the state’s herds had 22 cows that were positive for M.
paratuberculosis with a total prevalence of 6.9%, using antibody ELISA (Johnson-
Ifearulundu and Kaneene, 1999). The highest prevalence of positive herds coincided
With the eastern portion of the Lower Peninsula in agricultural district 6 (94.1%), district

9 (87.5%), and district 3 (80%) (see Figure 1-2).

Figure 1-2: Agricultural districts of Michigan showing the counties with numbers of
herds used for our study

 

 

Alpena Comty
8 Herd: tested

 

 

 

Dgemaw County

I/ 1 Herd tested

 

N
N

 

 

 

Upper Peninsula
Northwest ml.—
Northeast
West Central

Central 3
East Central
Southwest
South Central 1
Southeast

 

 

 

 

 

 

PPFF’P‘PPE"?

 

 

 

 

 

 

 

L i
/ J 1 LL] 38331533”

 

 

 

 

Bovine tuberculosis

Tuberculosis in cattle, most commonly caused by Mycobacterium bovis, is the
other most recognized mycobacterial disease in cattle. Bovine tuberculosis (TB) is an
important health risk due to the potential spread of the disease to other livestock and
humans. Infected cattle often do not show clinical signs, however, some cattle have
chronic weight loss, anorexia, weakness, lethargy, low-grade fever, or a sofi, moist,
chronic cough. Inhalation is the primary mode of natural transmission for M. bovis in 80-
90% of cattle (Morris et al., 1994). Ingestion is considered secondary to inhalation as a
mode of transmission, which can occur directly from infected cattle or from contaminated

pastures, water, or fomites (Menzies and Neill, 2000).

As a result of inhalation exposure, primary lesions for TB are typically found in
the lungs. The majority of lung lesions are found in the caudal lobes of the left or right
lungs, and are capable of being single, multiple, unilateral or bilateral (Mcllroy et al.,
1986). The initial lesion is normally comprised of the invading bacilli surrounded by a
cluster of neutrophils. Macrophages soon replace the neutrophils and ingest the
mycobacteria. During the cell—mediated immune response, sensitized T lymphocytes
produce cytokines that activate the macrophages, increasing their ability to inhibit the
growth of the phagocytosed bacteria or helping to kill them effectively. The activated
macrophages can also change in shape or fuse together to form multinucleated giant cells.
As more macrophages are recruited to the site of antigen insult, classic granulomas form
which also contain multinucleated giant cells, epithelioid macrophages/histiocytes,
lymphocytes, fibroblasts, and necrosis. Tuberculous granulomas are generally caseous,
meaning that the cells and tissue in the center of the lesion are destroyed during the
process caseous necrosis. Collagenous connective tissue eventually begins to surround
the lesion, which may increase in size large enough to penetrate blood or airway vessels
and allow dissemination to other organs. The most frequent gross and histological
lesions have been found in the thoracic lymph nodes of cattle (Whipple et al., 1996).
Only about 1% of infected cattle usually have lesions in the udder, and rarely infect other

cattle by milk (Collins et al., 1987).

Screening tests for bovine tuberculosis
The type IV delayed hypersensitivity reaction responsible for the formation of

granulomas is the basis for ante-mortem screening tests used for TB. Type IV

10

hypersensitivity is a reaction involving sensitized T lymphocytes that react with cell
bound antigen causing the release of cytokines. These cytokines cause mononuclear cell
accumulation, tissue damage, and inflammation usually observed at least 24 hours after
exposure to the antigen. Following initial infection of mycobacteria in an animal, the
bacteria are phagocytosed by macrophages and presented to cells involved in innate and
acquired immune responses. Antigen-specific CD4+, CD8+, and 7/6 T cells become
activated after exposure to the bacteria (Flyrm and Chan, 2001; Hope et al., 2000;
Liebana et al., 1999; Mustafa et al., 1986; Pollock et al., 1996). These sensitized T cells
are recruited to the site of subsequent infections and produce cytokines such as 'y-IFN and
tumor necrosis factor alpha (Feng et al., 1999). The release of y-IFN increases the ability
of macrophages to kill the bacilli that they have phagocytosed (F lesch and Kaufmann,
1991). In addition to releasing y-[FN, CD4+, CD8+, and 7/5 T cells have also been shown
to be directly involved in cytotoxic activity (Tan et al., 1997; Skinner et al., 2003).

Purified protein derivative (PPD) tuberculin is a crude preparation of
mycobacterial antigens used for the detection of TB in cattle. Bovine PPD tuberculin is
produced from M. bovis (AN 5 or Vallee strains), and each batch must be tested in
animals to be compared to a reference standard (Monaghan et al., 1994). A recent study
evaluating the genome of M. bovis ANS, for possible gene deletions during in vitro
passage, determined that the PPD strains do not possess any dramatic differences
between other strains of M. bovis and thus are suitable for detection of M. bovis infection
(Inwald et al., 2003). When PPD tuberculin is injected intradermally, dendritic cells take
up some of the antigen and migrate to the regional lymph node. In a previously

sensitized animal, the PPD will attract sensitized T cells to the injection site, in addition

11

to neutrophils and macrophages. Circulating Th1 cells, such as CD4+, CD8+, and y/S T
cells are activated by the antigen and secrete y-IF N and other cytokines within a few
hours of being stimulated by the antigen. These cytokines activate macrophages to
increase their ability for cell killing, and also lead to the palpable swelling at the site of
PPD injection due to inflammatory edema, fibrin deposition, and local thrombosis.
Intradermal tuberculin skin tests are the most widely used screening tests for TB
in cattle. The two most common of these skin tests in cattle in the US. are the caudal
fold tuberculin (CF T) test and the comparative cervical tuberculin (CCT) test. The CFT
test is administered by injecting bovine PPD into the caudal fold at the base of the tail,
whereas the CCT test compares the response from separate injections of bovine PPD and
avian PPD in the cervical region. Diagnostic tests using PPD, however, are limited
because most of the proteins in PPD are shared between different mycobacteria species
(Aagaard et a1, 2003). The reported sensitivity and specificity of the CFT test has varied
with past studies, but the test is frequently used because it is simplest to perform and
relatively inexpensive (see Table 1-3). A high sensitivity is desired to detect
asymptomatic disease within the herd, whereas a high specificity minimizes the number
of false positives. Sensitivity and specificity are important for determining the presence
of disease in a herd, but predictive value is generally more useful at the level of the
individual animal. Predictive value is an estimate of the likelihood that a test result is
true. Positive predictive value estimates the likelihood that an animal testing positive is
truly positive, a measure of the specificity in the context of disease prevalence. The
predictive value is affected by prevalence, such that the lower the prevalence of disease

in a population, the lower the positive predictive value. The CFT test has a low positive

12

predictive value due to a low prevalence of TB in the cattle population, resulting in the
disadvantage of a higher likelihood of false positives for the test.

False positives on the CFT test have been attributed to the subjectivity of reading
the CFT test, as well as exposure of cattle to M. paratuberculosis, M. avium, M.
tuberculosis, environmental Mycobacterz'a spp., or certain non-mycobacteria agents such
as Nocardia spp (Karlson, 1962). A study in 1981 showed that cattle infected with M.
avium complex serotypes 6, 14, and 18 produced positive reactions to the CFT test
(Ketterer et al., 1981). The reclassification of M. paratuberculosis as a subspecies of M.
avium would suggest the animal might respond similarly to M. avium on the CFT and

CCT tests (McIntyre and Stanford, 1986).

Table 1-3: Sensitivity and specificity for the CFT test

 

 

Author Sample size Sensitivity Specificity
Whipple et al., 1995 204 positive 80.4%

Wood et al., 1991 125 positive; 6177 negative 65.6% 99.5%
Wood et al., 1992 22 positive; 1340 negative 68.2% 97.3%
Francis et al., 1978 135 positive; 3820 negative 72.0% 98.8%

 

The y-IFN assay is also approved as a screening test for TB in cattle. This in vitro
test measures y-IFN that is released by lymphocytes during cell-mediated immune
response to antigen stimulation. The assay compares the animal’s y-IF N response to
bovine PPD and avian PPD. The y-IFN assay has the benefits that animals do not have to
be contained for 72 hours to wait and read the test, subjective differences in interpreting
the results between veterinarians is eliminated, and the cattle don’t have to be injected

with PPD. Sensitivity and specificity estimates of the test have varied (see Table 1-4).

13

Table 1-4: Sensitivity and specificity for Bovigam* y—IF N assay

 

 

Author Sample size Sensitivity Specificity
Ryan et al., 2000 163 positive, 213 negative 85% 93%
Scacchia et al., 2000 36 positive, 70 negative 91.7% 84.3%
Lauzi et al., 2000 1557 negative 88.8%
Wood et al., 1992 22 positive, 1340 negative 81.8% 99.1%

 

*BovigamTM, BioCor Animal Health, Omaha NE

Prevalence of TB

M. bovis is currently endemic in free-ranging white-tailed deer (Odocoileus
virginianus) in northeastern Michigan, and since 1998 it has been diagnosed within the
same region in 25 beef cattle herds, 5 dairy cattle herds (Schmitt et al., 2002; Michigan
Bovine TB Web site, 2003), and in one captive cervid farm (Kaneene et al., 2002). M.
bovis has also been detected in a cat in the same endemic area (Kaneene et al., 2003), as
well as other free-ranging carnivores including coyotes, raccoons, a red fox, black bear,
and bobcat (Bruning-Fann et al., 2001). Since 2001, TB has also reemerged in areas of
Texas, California, and New Mexico (TAHC, 2002; CDFA, 2002; APHIS, 2003). In deer,
the disease is efficiently transmitted through nasal secretions, saliva, or contaminated
feed (Palmer et al., 2001 ). Infected deer are considered to be a reservoir for the spread of
TB to cattle in Michigan, with increased spread among deer likely as a result of
unnaturally dense congregations around contaminated feed from supplemental feeding

and baiting (Kaneene et al., 2002; Miller et al., 2003; Payeur et al., 2002).

14

Conclusion

J ohne’s disease and TB are important diseases in cattle, both caused by
mycobacteria species. Limitations exist for ante-mortem tests used for each disease,
especially with regards to sensitivity and specificity. Cross-reactivity between varying
mycobacterial antigens is one of the' limitations for ante-mortem testing and has led to
speculation that M. paratuberculosis infection may significantly influence an animal’s
reaction to the CF T test (Morrison et al., 2000). Due to a high prevalence of M.
paratuberculosis among cattle in Michigan and the United States, the studies described in
the following chapters were designed to determine whether cattle testing positive for M.
paratuberculosis are more likely to have false positive results on the CFT test and y-IFN
assay than cattle testing negative for M. paratuberculosis in Michigan. Evaluating the
effect of M. paratuberculosis on the CFT test and y-IFN assay under field and laboratory
conditions is an important step for determining the amount of confidence given to

positive results among herds infected with Johne’s disease.

15

CHAPTER 2
THE EFFECT OF INFECTION WITH MYCOBAC T ERIUM A VIUM SUBSP.
PARA TUBERCULOSIS ON THE CAUDAL FOLD TUBERCULIN (CFT) TEST FOR
BOVINE TUBERCULOSIS IN CATTLE.

Abstract
Objective - To determine whether cattle testing positive for Mycobacterium avium subsp.
paratuberculosis (M. paratuberculosis) were more likely to have false positive results on
the caudal fold tuberculin (CF T) test than cattle testing negative.
Animals — 1,043 Holstein cattle from 10 herds in Michigan
Procedure —Fecal and whole blood samples were collected from all cattle 224 months of
age on the day the CF T test was read. Fecal samples were submitted for mycobacterial
culture, and samples of plasma were tested for antibody against M. paratuberculosis.
Results — The total number of CFT suspects from all herds was 180 (17.3%). Of the 1043
cattle tested in this study, 45 (4.3%) were positive for M. paratuberculosis by fecal
culture and 115 (11.0%) were positive by antibody ELISA. Percentage of cattle positive
for the CFT test ranged from 16.6%, in cattle negative for M. paratuberculosis fecal
culture and antibody ELISA, to 25.0% in cattle positive for both tests. No variables were
significantly associated with a positive CFT test.
Conclusions —Cattle testing positive for M. paratuberculosis, as measured by a positive
M. paratuberculosis fecal culture or antibody ELISA test, appear to have an increased
likelihood of false positive results on the C FT test, although the association is not

statistically significant.

16

Introduction

M. avium subsp. paratuberculosis (M. paratuberculosis) has been speculated to
influence an animal’s reaction to the CF T test due to cross reactivity between varying
mycobacterial antigens (Morrison et al., 2000). This study was designed to determine
whether cattle testing positive for M. paratuberculosis are more likely to have false
positive results on the CF T test when compared to cattle testing negative for M.
paratuberculosis in Michigan. Evaluating the effect of M. paratuberculosis on the CF T
test under field conditions is an important step for determining the amount of confidence

given to positive results among herds infected with Johne’s disease.

Methods
Animals

Samples of feces and whole blood were collected from 1043 Holstein cattle from
10 separate herds from three counties in Michigan. Eight of the herds sampled were
within a TB high-risk area (Alpena county), one in a TB-free area (lngham county), and
one near the high-risk area (Ogemaw county) (Figure l-l). Criteria for inclusion were a
willingness to participate, having a total number of cattle less than 250 (for cost reasons),
and being able to sample the herd on the day their required whole-herd TB tests were
being read. All cattle greater than or equal to 24 months of age were sampled from each
herd. None of the herds had been previously vaccinated for M. paratuberculosis or M.

bovis.

l7

Tuberculosis skin testing

The CFT tests were performed by either accredited private practice veterinarians
or by veterinarians employed by the Michigan Department of Agriculture (MDA) using
the methods described in the USDA Uniform Methods and Rules (USDA, 1999). The
same MDA veterinarian tested seven of the 10 herds. All comparative cervical tuberculin
(CCT) tests were performed by USDA or MDA veterinarians. Results of both skin tests

were interpreted 72 :t 6 hours after injection of tuberculin.

Sampling

Blood and fecal samples were collected on the day the CFT test was read. A new
plastic sleeve was used for each animal to collect fecal samples, which were placed in
separate plastic whirl-pack bags afier collection. Fecal samples were transported in
coolers at ambient temperature to the Diagnostic Center for Population and Animal
Health at Michigan State University and stored at -80°C until cultured. Blood was
collected on the day the CFT test was read via the middle coccygeal vein using a 20-
gauge, l-in. needle. The blood was collected into a 10 ml Vacutainer tube containing
sodium heparin (Coming Glass Works, Corning, NY). Blood samples were transported
to the lab in plastic coolers, chilled with ice packs, and were processed within 24 hours of

when the samples were drawn.

Laboratory methods -— A detailed description of laboratory methods is located in the

appendix.

18

M. paratuberculosis testing

M. paratuberculosis laboratory diagnostic tests consisted of fecal culture and
testing for antibodies using a plasma ELISA. Fecal samples were cultured to detect the
presence of M. paratuberculosis using standard procedures recommended by USDA-
National Veterinary Service Laboratory and based on the procedures used by Whitlock et
al. at the University of Pennsylvania (Fyock and Whitlock, 1999).

Plasma samples were tested for antibodies to M. paratuberculosis with a
commercial antibody ELISA test kit.1 The samples were tested in duplicate wells and the
average optical density (OD) was calculated. The corrected OD was calculated by
subtracting the average OD of two negative serum control wells from the average OD of

duplicate sample wells. A corrected OD greater than 0.1 was considered positive.

Statistical methods

Prevalence of J ohne’s disease within each herd was computed for each type of
Johne’s test (fecal culture or antibody ELISA) by the number of positive cattle for each
test divided by the total number of cattle tested within each herd.

Statistical analyses were conducted using a standard software packagez, testing
associations between CF T test results and CFT testing veterinarian, animal age, and
J ohne’s disease status. In the analyses, multivariable logistic regression models with
random effects were developed to assess associations between animal CFT status and
animal factors and Johne’s disease status. Because the analysis was done on an

individual animal level, the random effect function was used in all the models to adjust

 

' Parachekm, BioCor Animal Health, Omaha NE
2 SAS V8. SAS lnstitue, Cary, NC

19

for the fact that animals from the same herd are more alike in terms of the exposure than
animals in other herds. Model outcome was CF T test status (positive or negative), and
risk factors including testing veterinarian, animal age, results of individual Johne’s tests,
and herd prevalence of Johne’s disease tests (Table 2-1).

A backwards stepwise model development approach was used to create final
models with risk factbrs significant at p_<_0.05. In brief, a full model was generated, and
possible interactions and confounding were assessed and corrected during the model
development process. Odds ratios (OR) with 95% confidence intervals were computed
for parameter estimates. With the exception of the potential confounders forced into the
model, each risk factor was tested by examining the effects of removal of that factor from
the model. If removal of the risk factor resulted in a change in the OR of the remaining
variables of more than 10%, the risk factor and its interaction terms were retained in the

model.

Results
Johne ’3 disease results

Of the 1043 cattle tested in this study, 45 (4.3%) were fecal culture positive and
115 (11.0%) were positive by antibody ELISA (Table 2-2). The herd prevalence for each
test ranged from 0 to 15.4% (mean 4.5%, median 2.8%) for the M. paratuberculosis fecal
culture, and 1.2 to 30.8% (mean 9.7%, median 6.9%) for the M. paratuberculosis

antibody ELISA.

20

TB results

All cattle that responded to the CF T test (any visible or palpable response) were
classified as suspect and retested with the CCT test. Of the 1043 total cattle, 180 were
CFT suspects (17.3%). Herd response ranged from 8.0% to 56.1% (mean 21.2%, median
20.1%) for the CFT test (Figure 2-1). Five of the 180 CFT suspects were also CCT
suspects and a sixth animal was a CCT reactor. The CCT suspects and the CCT reactor
were classified as negative for TB based on necropsy, histological examination of tissues
and mycobacterial culture. Johne’s disease test results were separated displayed,
comparing those animals positive versus those negative to the CFT test (Figure 2-2).
CFT test results were also displayed based on Johne’s disease test results (Figure 2-3).
The percentage of cattle positive for the CFT test ranged from 16.6%, in cattle negative
for M. paratuberculosis fecal culture and antibody ELISA, to 25.0% in cattle positive for

both tests (Figure 2-3).

Statistical results

In the multivariable regression model for the CFT test, the individual animal and
herd level test results of M. paratuberculosis fecal culture and antibody ELISA did not
significantly affect the CFT test results (Table 2-3). The veterinarian performing the CFT
test and age of the animal also lacked a significant association with a positive CFT test.
M. paratuberculosis fecal culture and antibody ELISA test results were poorly correlated

with a spearman correlation coefficient of 0. l 663 (p<0.0001).

21

Discussion

Cross-reactivity between different mycobacterial antigens have led to the
suggestion that infection with M. paratuberculosis may lead to false positive responses to
the CFT test (Morrison et al., 2000). It is important to determine the cause of false
positives responses as they may lead to unnecessary culling, increased time and cost of
animal handling, increased cost of follow-up testing, as well as psychological stress to
producers and veterinarians.

False positive responses to the CFT test have been reported in a study following
vaccination of calves against M. paratuberculosis (thler et al., 2001). Four calves were
vaccinated at 28 days of age with a commercially available modified-live M.
paratuberculosis vaccine, resulting in positive or suspicious skin reactions on the CFT
test for as long as 2 years after vaccination (Kohler et al., 2001). Vaccination in their
study was shown to create a cell-mediated immune reaction detectable by tuberculin skin
tests over a long period of time. Our study included adult cattle that were not vaccinated
for M. paratuberculosis, but rather attempted to evaluate the relationship between natural
M. paratuberculosis infection and the CFT test.

When comparing cattle that were positive to the CF T test with cattle negative to
the CFT test, a higher percentage of CFT positive cattle were positive for M.
paratuberculosis fecal culture, antibody ELISA, and a combination of the two tests
(Figure 2-2). Similarly, cattle testing positive for M. paratuberculosis fecal culture,
antibody ELISA, and a combination of the two tests had a higher percentage of cattle
testing positive for the CFT test compared to cattle negative by fecal culture and ELISA.

Cattle positive for M. paratuberculosis were estimated to be four times as likely to be

22

positive on the CFT test than negative cattle, although the estimation was not statistically
significant (Table 2-3).

Despite the minimal trend of a higher likelihood of a positive CF T test among
cattle testing positive for M. paratuberculosis, the most surprising finding of our study
was the lack of a statistically significant association. There was no significant association
at either the individual animal or the herd level between test status for Johne’s disease
and risk of CF T response. The lack of significant association between positive Johne’s
disease tests and the CFT test could have the following possible explanations. First, both
tests used to detect M. paratuberculosis in this study have relatively poor sensitivities
(Table 1-1; Table 1-2). A low sensitivity would result in a high number of false
negatives, resulting in an underestimation of the true number of cattle infected with M.
paratuberculosis. This misclassification would reduce the power to detect a true positive
association between cattle testing positive for M. paratuberculosis and a positive CFT
test result. Second, the sample size may have been too small to detect a true association.
The sample size used for this study was based on the ability to detect a 25% difference in
the CFT test results between cattle testing positive and negative for M. paratuberculosis.
In this study, the difference in CF T test results were only 7% between cattle testing
positive and negative by M. paratuberculosis fecal culture, and 4% between cattle testing
positive and negative by ELISA. With 80% power (the probability of detecting a true
association), the true sample size needed to detect a significant association between M.
paratuberculosis tests and the CFT test would have to be 6,304 based on culture and
7,736 based on ELISA, compared to the total sample size of 1,043 used for this study

(Table 2-4). Third, there was a possible lack of the true prevalence of Johne’s disease

23

between herds. All ten herds in this study had a Johne’s disease prevalence of 15% or
less, as measured by M. paratuberculosis fecal culture. There were no high prevalence
herds that would have increased the number of total positive animals, helping to increase
the test positive sample size, as well as increasing the power of detecting an association
between herd level M. paratuberculosis results and positive CFT test results.

The results of this study show a minimal trend between positive M.
paratuberculosis test results and a positive CFT test result for individual animals. A
combination of factors may have led to the lack of statistical significance of this
association. The most likely combination is a misclassification of true positives as
Johne’s disease test negative due to the low sensitivity of the current tests. This
misclassification may have led to a difference of the percent of CFT positive cattle
between cattle testing positive and negative for M. paratuberculosis that was smaller than
the true difference. The smaller difference decreased the statistical power, which would '
make a larger sample size necessary to detect a significant association.

The CF T test is prone to false positives from cross—reactivity due to the fact that
many of the proteins in PPD are shared between different mycobacteria species. Current
research is underway to identify proteins that are specific to M. bovis, able to be used in a
whole-blood assay. A recently published study concluded that a mix of the proteins
ESAT-6, CFP-10, T8274, and TB10.4 may have a sensitivity as good as the CFT test
with an improved specificity (Aagaard et al., 2003). It is important that further research
continue with either larger sample sizes or more specific diagnostic tests in order to
determine whether an association truly exists between cattle testing positive for M.

paratuberculosis and positive CFT test results.

24

Table 2-1: Description and descriptive statistics for risk factors evaluated

 

 

 

 

 

 

 

 

 

 

 

Outcome Values Description It %*
1 Positive 1 80 1 7
CFT Test Result .
0 Negative 863 83
Risk Factor Values Description n %
Veterinarian 1 (CFT tested
1 195 19
I herd)
Veterinarian 2 (CFT tested
. 2 94 9
CFT Testing I herd)
Veterinarian Veterinarian 3 (CFT tested
3 647 62
7 herds)
Veterinarian 4 (CFT tested
4 107 10
I herd)
1 Positive 115 11
M. paratuberculosis ELISA .
0 Negative 928 89
M. paratuberculosis 1 Positive 45 4
Culture 0 Negative 998 96
M. paratuberculosis ELISA 1 Positive for both 16 2
and Culture 0 Negative for one or both 1027 98
Continuous variables
Age (months) 24-144 Age of animal in months 1043 100
M. paratuberculosis % cattle positive for M.
Culture 0—15.4 paratuberculosis fecal 1043 100
Herd Prevalence culture for each herd
% cattle positive for M.
M. paratuberculosis ELISA .
1.2-3 0.8 paratuberculosis antibody 1043 100

Herd Prevalence

ELISA for each herd

 

*Percentage of cattle described by the respective value (= n/1043)

25

Table 2-2: Descriptive analysis of CFT test results based on positive Johne’s disease

tests

 

 

 

 

 

 

 

, . All CFT CFI‘
Johne 5 disease test .
Cattle suspect negatrve
(positive results) (n=1043) (n=180) (n=863)
# % # % # %
Fecal Culture 45 4.3 11 6.1 34 3.9
Antibody ELISA 115 11.0 24 13.3 91 10.5
Fecal Culture 9_r
. 144 13.8 31 17.2 113 13.1
Antibody ELISA
Fecal Culture em
16 1.5 4 2.2 12 1.4
Antibody ELISA
Negative to both Johne’s
899 86.2 149 82.8 750 86.9

disease tests

 

 

 

 

26

 

Table 2-3: Results of a multivariable logistic regression analysis of the effect of
individual and herd level M. paratuberculosis test results on the CFT test as the
outcome (classified as suspect or negative).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Full Model
Risk Factor Estimate S.E. x2 p OR 95% CI.
1 .34 1.61 0.83 1.41 .06 - 33.1
2 2.37 3.33 0.48 10.67 .02 — 7288
Testing vet:
3 3.92 3.38 0.25 50.26 .07 — 3.8E4
4 0 - - - -
Age (months) .01 .01 0.10 1.01 .998 - 1.02
Day 3 Para Result -.37 .89 0.67 0.69 .12 — 3.93
Culture 1.46 1.01 0.15 4.29 .59 — 31.0
CultureANDELISA .01 .84 0.99 1.01 .20 — 5.22
A CultureHerdPrev -1.23 7.04 0.86 0.29 3.0E-7 — 2.9E5
ParaHerdPrev 14.39 12.2 0.24 1.8E6 7.3E-5 — 4.3E16
Culture*Day3ParaResult 0 - - l -
Agem*Day3ParaResult .01 .02 0.41 1.01 .98-1.04
Agem*Culture -.02 .02 0.29 0.98 .95-1.02
Model —2 Res Log Likelihood = 5139.0
Reduced Mod_el
Risk Factor Estimate 8.153. X2 p OR 95% CI.
Age (months) 0.01 .004 0.06 1.01 .999-1.02
Model —2 Res Log Likelihood = 5088.9

 

 

27

 

Table 2-4: Sample size required to detect a true association based on the actual
percentage of CFT positive cattle and M. paratuberculosis test results, and power
calculations based on the actual sample size

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.. t... ,H o
5’ (<2 53 32 8 a) a
'5 :1 5 N 35 N E o
g a; 1.1.1 2 LL] 2 '5 8 '7) 8 a
assesses
‘— ‘F‘ .—
CFT resultsvs. g “at g 8. .3: 3 2 8 g ‘5 Eng .3
ELISA “é o :< Eta-5.2 wao e
o D“ LL} W o 53 o a E t: E Q‘ '5
U “-1 :i R“ 5’,” 9- o :3 <1: 5 <1 g
0 LL] I—‘ C: I— Q E {’2 E a s.-
.9 “‘" L“ 'E ._1 'E -1 "‘
*5 U U --' I-I-I --- LL] 8
at .\° .\° 2 2 13
Sam le size re uir dt
1’ q .3. ° 95% 80% 9282115 17% 21% 6,883 853 7,736
detect true assocratron
Power based '
onactual 95% 17% 9282115 17% 21% 928 115 1,043
sample size
-- D Q) Q)
.°>’ .53 5 “5 2 “5 o g
g) cu E 3 .g g E. E .2
o a) ..>. 53 g 33 g g; g g o e-
g ,_ S 'a 0 § 0 § 3.2. a}: g
CFT results vs. g “3’ § 8. 2 a, I}: o g 13) E95 1:
Culture ‘15 9‘3 ”:5 95’ '5 g '3 E "’ 8 "’ 8, .022)
o 0-
o E g 8. go 8. 8 E e E 2: §-
2L) ,1: = I: c" .5. .3 .E _3 *-
s U 0 § 5 63$ 8
ed ..\°‘ C.\" [—
Sam 1e size re uired to
P ‘1. . 95% 80% 998:45 17% 24% 6.032 272 6,304
detect true assocratron
Power based
on actual 95% 20% 998:45 17% 24% 998 45 1,043
sample size

 

28

 

553.1553 Bum

8315 3315 @515 £15 Gene 3.1.5 9215 A815 A315 @315

 

 

 

 

 

 

 

 

3 a a b w m w m N g
1 LI _ . 1 A c _ 1 _ 1 c
a m H -
fl -
I H 1 e—
1 3
1 an 02.7.93
«swam .x.
Bongo SENSSQSQBR .3 I
1 av
2330
Room SEESQSSSR S m
1 cm
mfioamam .50 I
1 3.

PS: 3 838.. :8 02:8.— ..e anew—89.2— 2: .3 .3an "TN 9.33

29

Figure 2-2: Percentage of cattle positive for M. paratuberculosis fecal culture and
antibody ELISA tests separated by CFT status

 

I M. paratuberculosis ELISA
Only

III M. paratuberculosis Culture
and ELISA

% Positive

I M. paratuberculosis Culture
Only

 

 

 

 

 

CFT suspect CFT negative
cattle (n=180) cattle (n=863)

30

Figure 2-3: Percentage of cattle with positive CFI‘ results based on positive M.

paratuberculosis test results

Negative for Culture and ELISA
(n=899)

ELISA only (n=99)

Culture only (n=29)

Culture and ELISA (n=16)

 

 

 

 

 

31

10 15 20
“/11 CFT positive

25

30

CHAPTER 3
THE EFFECT OF INFECTION WITH M Y C OBA C T ERIUM A VI UM SUBSP.
PARA TUBERCULOSIS ON THE GAMMA INTERFERON (y-IFN) ASSAY FOR
BOVINE TUBERCULOSIS IN CATTLE.
Abstract
Objective — To determine whether cattle testing positive for Mycobacterium avium subsp.
paratuberculosis (M. paratuberculosis) are more likely to have false positive results on
the y-IFN assay for bovine tuberculosis than cattle testing negative.
Animals — 1,043 Holstein cattle from 10 herds in Michigan
Procedure — Fecal and whole blood samples were collected from all cattle 224 months of
age on the day each herd’s annual caudal fold tuberculin (CFT) test was read. Fecal
samples were submitted for mycobacterial culture. Samples of plasma were tested for
antibody against M. paratuberculosis, as well as for y-IFN after whole blood samples
were stimulated with phosphate buffered saline solution or purified protein derivative
(PPD) from either M. bovis or M. avium.
Results — Of the 1043 cattle tested in this study, 45 (4.3%) were positive for M.
paratuberculosis by fecal culture and 115 (11.0%) were positive by antibody ELISA. A
total of 8 cattle (0.8%) were positive in the y-IFN assay after stimulation with PPD from
M. bovis. None of the eight cattle were also positive for either M. paratuberculosis fecal
culture or antibody ELISA.
Conclusions — No significant association was found between cattle testing positive by M.
paratuberculosis fecal culture or antibody ELISA. and positive results of the y-IFN assay

for bovine tuberculosis.

32

Introduction

The y-IFN assay for TB compares an animal’s y-IFN response to whole blood
stimulated by bovine purified protein derivative (PPD) tuberculin with the y-IFN
response stimulated by avian PPD. Cross-reactivity is speculated to cause false positives
on diagnostic tests that utilize PPD due to shared proteins among mycobacterial species
(Aagaard et al., 2003). As a result, the high prevalence of J ohne’s disease in Michigan
cattle may lead to cross-reactivity on the y-IF N assay used to detect bovine tuberculosis
(TB). A recent study in Denmark demonstrated a lower specificity of the y—IFN assay for
TB among M. paratuberculosis-infected herds compared to uninfected herds (Jungersen
et al., 2002).

This study was designed to determine whether individual cattle testing positive
for M. paratuberculosis are more likely to have false positive results on the y-IFN assay
for TB than cattle testing negative for M. paratuberculosis. Evaluating the effect of M.
paratuberculosis on the y-IFN assay is an important step for determining the amount of

confidence given to positive results among herds infected with Johne’s disease.

Methods
Animals

The study included 1,043 Holstein cattle from 10 separate herds located in three
counties in Michigan. Eight of the herds sampled were within a TB high-risk area
(Alpena county), one in a TB-free area (lngham county), and one near the high-risk area
(Ogemaw county) (Figure 1-1). Criteria for inclusion were a willingness to participate,

having a total number of cattle in the herd less than 250 (for cost reasons), and being able

33

to sample the herd on the day their required whole-herd TB test was being read. All
cattle greater than or equal to 24 months of age were sampled from each herd. None of

the herds had been previously vaccinated for M. paratuberculosis or M. bovis.

Sampling

Blood and fecal samples were collected on the day the CFT test was read. A new
plastic sleeve was used for each animal to collect fecal samples, which were placed in
separate plastic whirl-pack bags after collection. Fecal samples were transported in
coolers at ambient temperature to the Diagnostic Center for Population and Animal
Health at Michigan State University and stored at -80°C within 24 hours of sampling until
cultured. Blood was collected via the middle coccygeal vein using a 20-gauge, l-in.
needle. The blood was collected into a 10 ml Vacutainer tube containing sodium heparin
(Corning Glass Works, Corning, NY). Blood samples were transported to the lab in
plastic coolers, chilled with ice packs, and were processed within 24 hours of when the

samples were drawn.

Laboratory methods — A detailed description of laboratory methods is located in the

appendix.

M. paratuberculosis testing
M. paratuberculosis laboratory diagnostic tests consisted of fecal culture and
testing for antibodies using a plasma ELISA. Fecal samples were cultured to detect the

presence of M. paratuberculosis using standard procedures recommended by USDA-

34

National Veterinary Service Laboratory and based on the procedures used by Whitlock et
al. at the University of Pennsylvania (Fyock and Whitlock, 1999).

Plasma samples were tested for antibodies to M. paratuberculosis with a
commercial antibody ELISA test kit.3 The samples were tested in duplicate wells and
the average absorbance values, or optical density (OD) was calculated. The corrected
OD was calculated by subtracting the average OD of two negative serum control wells
from the average OD of duplicate sample wells. A corrected OD greater than 0.1 was

considered positive.

y-IFN testing

Whole blood samples were tested for y-IFN using a commercially available
antigen capture ELISA, following the manufacturers recommended protocol.4 Using the
OD values, an animal was classified as positive for M. bovis y-IFN if the difference
between the mean OD value from the bovine PPD samples and the nil antigen samples
were 2 0.1 and the difference between the mean OD value from the bovine PPD samples

and avian PPD samples were also 2 0.1.

Statistical methods
Prevalence of J ohne’s disease within each herd was computed for each type of
J ohne’s disease test (fecal culture or antibody ELISA) by the number of positive cattle for

each test divided by the total number of cattle tested within each herd.

 

3 Parachekm, BioCor Animal Health. Omaha NE
4 Bovigam”, BioCor Animal Health. Omaha NE

35

Statistical analyses were conducted using a standard software packages, testing
associations between 'y-IFN assay results and animal age and Johne’s disease status. In
the analyses, multiple logistic regression with random effects was performed with
positive y-IFN result for TB as the model outcome and risk factors including Johne’s test
results, J ohne’s herd prevalence for each test, and animal age (Table 3-1). Because the
analysis was done on an individual animal level, the random effect function was used in
all the models to adjust for the fact that animals from the same herd are more alike in
terms of the exposure than animals in other herds.

A backwards stepwise model development was used to create final models with
risk factors significant at p50.05. In brief, a full model was generated, and possible
interactions and confounding were assessed and corrected during the model development
process. Odds ratios (OR) with 95% confidence intervals were computed for parameter
estimates. With the exception of the potential confounders forced into the model, each
risk factor was tested by examining the effects of removal of that factor from the model.
If removal of the risk factor resulted in a change in the OR of the remaining variables of

more than 10%, the risk factor and its interaction terms were retained in the model.

Results
Johne ’s disease results

Of the 1043 cattle tested in this study, 45 (4.3%) were positive for M.
paratuberculosis fecal culture and 115 (11.0%) were positive for antibody ELISA (Table

3-2). The herd prevalence for each test ranged from 0 to 15.4% (mean 4.5%, median

 

5 SAS V8, SAS 1118111116, Cary. NC

36

2.8%) for M. paratuberculosis fecal culture, and 1.2 to 30.8% (mean 9.7%, median 6.9%)

for M. paratuberculosis antibody ELISA (Figure 3-1).

TB results

Eight cattle (0.8%) were positive for y-IF N after stimulation of blood samples
with PPD from M. bovis. Herd response ranged from 0.0% to 3.5% (mean 0.9%, median
0.3%) (Figure 3-1). Two of the eight cattle positive for y—IFN were CFT suspects,
however, both were negative on the CCT test. This study was done before the y-IFN
assay became an official screening test for TB, so none of the eight cattle that were y-IF N
positive were removed from the herd for post-mortem examination. In addition, none of
the eight cattle that were positive for y-IFN were also positive for M. paratuberculosis
fecal culture or antibody ELISA (Table 3-2). To date, M. bovis has not been diagnosed

on any of the farms in this study.

Statistical results

The y-IFN test results following stimulation with M. bovis PPD for each animal
were analyzed with individual Johne’s test results, Johne’s herd prevalence for each test,
and age (Table 3-3). None of the animals positive for y-IFN were also positive for M.
paratuberculosis antibody ELISA or fecal culture. No variables were significantly

associated with positive y-IFN test results.

37

Discussion

The y-IF N assay has several advantages over tuberculin skin tests for diagnosis
of bovine tuberculosis. First, blood samples are taken only once, so animals are not
required to be held for 72 hours and handled twice, which is required for administration
and reading of the CFT and CCT skin tests. Second, cattle can be retested at any time,
unlike the skin tests, which require retesting within 10 days or after 60 days from the
previous skin test (USDA, 1999). Third, unlike the skin tests, the immune response of
the animal is not potentially compromised from the intradermal injection of tuberculin
(Radunz and Lepper, 1985). Fourth, the assay eliminates the subjective variability
present between veterinarians reading skin tests. The y-IF N assay also has several
disadvantages. Laboratory facilities are required with the ability to cultivate the blood
cells and detect the y-IF N production. The blood must also be processed quickly since
cells must be viable in order to produce y-IFN during stimulation with PPD. A study in
1992 determined that blood must be processed within 8 hours for optimal production of
y-IFN (Rothel et al., 1992).

The USDA’s Center for Veterinary Biologics licensed the y-IFN assay on
December 10, 2001 as an official supplemental test for the bovine TB eradication
program (USDA, 2002). The manufacturer recommends that the test be used as an
ancillary test for serial (confirmation of positives) or parallel (confirmation of negatives)
testing, three to thirty days after skin testing (Biocor, 2002). False positive responses
could cause a problem if used as a confirmation of cattle with positive responses from

tuberculin skin testing. These false positive responses could lead to unnecessary culling,

38

unnecessary time and cost of follow-up testing, as well as possible psychological stress to
the producer and veterinarian.

There were only eight cattle positive for the y-IFN assay for TB from 1,043 cattle
tested in this study. Two of the eight cattle were positive on the CFT test, but were
negative on the follow-up CCT test. None of the eight cattle positive for M. bovis y-IFN
were also positive for M. paratuberculosis antibody ELISA or fecal culture. Because
there was no overlap between positive M. paratuberculosis test results and positive M.
bovis y-IFN results, the 95% confidence intervals for the odds ratios were extremely
large. The lack of overlapping values resulted in a large standard error for the estimate of
the independent variable, leading to the infinite range of the confidence interval.

Although no significant effect was found between the M. paratuberculosis test
results and positive M. bovis 'y-IFN results, the lack of effect was only based on eight 7-
IF N positive animals. Due to the small number of false positives, a true association due
to M. paratuberculosis would have been impossible to identify. The sensitivity of the y-
IFN assay used in this study has been reported to range from 81 .8-91 .7%, with a
specificity ranging from 84.3-99.1% (Table 1-4). The low number of animals positive for
y-IFN in this study supports a high specificity for the assay. If M. paratuberculosis truly
has no effect on the y-IFN assay, the low number of false positives may make it a useful
supplemental screening test in herds infected with Johne’s disease.

Further studies involving a large sample size, however, need to be conducted to
confirm these findings. In addition, research should focus on using proteins that are
specific to M. bovis, rather than PPD that share proteins with other mycobacterial species.

The recent mapping of the complete genome sequence for M. bovis allows specific

39

proteins to be identified that are unique to M. bovis (Garnier et al., 2003). A recently
published study identified a mixture of the proteins ESAT-6, CFP-lO, TB27.4, and
TB10.4 as a potential novel cocktail for the development of new tests for M. bovis
(Aagaard et al., 2003). It is important that genomic sequence be studied in depth to help
generate more possible proteins that can be tested for potential highly specific diagnostic

assays for TB.

40

Table 3-1: Description and descriptive statistics for risk factors evaluated

 

 

 

 

 

 

 

 

 

 

Outcome Values Description 11 %*
y-IF N Assay Result 1 Positive 8 1
0 Negatlve 1035 99
Risk Factor Values Description 11 %
M. paratuberculosis ELISA 1 Positive 115 ll
' 0 Negative 928 89
M. paratuberculosis Culture 1 Positive 45 4
0 Negative 998 96
M. paratuberculosis ELISA and 1 Positive for both 16 2
Culture 0 Negative for one or both 1027 98
Continuous variables
Age (months) 24-144 Age of animal in months 1043 100
M. paratuberculosis Culture % cattle positive for M.
Herd Prevalence 0-15.4 paratuberculosis fecal 1043 100
culture for each herd
M. paratuberculosis ELISA % cattle positive for M.
Herd Prevalence 1.2-30.8 paratuberculosis antibody 1043 100

ELISA for each herd

 

*Percentage of animals described by the respective value (=n/1043)

Table 3-2: Descriptive analysis of TB test results based on positive Johne’s disease
tests

 

 

 

 

 

 

All y-IFN
Cattle positive
Johne’s disease test (n=1043) (n=8)
(positive results) # % # %
Fecal Culture 45 4.3 0 0
Antibody ELISA 1 15 1 1.0 0 0
Fecal Culture o_r
. 144 13.8 0 0
Antibody ELISA
Fecal Culture fl
16 1.5 0 0
Antibody ELISA
Negative to both Johne’s
. 899 86.2 8 100
disease tests

 

 

 

 

 

42

Table 3-3: Results of a multivariable logistic regression analysis of the effect of
individual Johne’s Disease test results and herd levels on the y-IFN test as the

outcome (classified as positive or negative).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Full Model
Risk Factor Estimate S.E. X2 p OR 95% Cl.
Age (months) 0.01 .01 0.49 1.01 .98 — 1.04
Day 3 Para Result -16.67 10864 0.998 5.71:"8 0-66
Culture -16.95 13297 0.999 4315’8 0-66
CultureANDELISA 17.83 11201 0.999 5.5137 0-66
CultureHerdPrev -3.08 1 1.1 0.78 .05 1.6E‘” — 1.3138
ParaHerdPrev -1.79 6.00 0.77 .17 1.31:"- 2.113“
Culture*Day3ParaResult 0 - - 1 -
Agem*Day3ParaResult -001 232 1.00 .99 331349330397
Agem*Culture -002 209 0.999 .98 1.21~:'”3-7.913T77

 

 

 

Model —2 Res Log Likelihood = 9909.4

 

43

 

€215 sans Gene Gene
2 a a A

 

 

 

25.» see E I

<m5m
xwonscm mafizoameoneQ .3 m

23:3 _mocm ESEEmcEEeA .3. a

 

 

5:35.53 as:

515 5.15 @315 A315 $15 3:15
e

 

 

 

 

 

 

m .. m N _
H _Em u H _ u _ 11” o
m - m m -.
1 1 E
u a
m m“ 96.8.—
w m 1 3 .x.
1 ma
1 - a
1 mm

 

Eu: .3 3.53.. .3. «>53: me «max—3.3.— 2: no Err—U "Tm .9.sz

44

CHAPTER 4
EVALUATION OF THE ASSOCIATION BETWEEN THE GAMMA
INTERFERON (y-IFN) ASSAY, STIMULATED BY M Y COBA C T ER] UM A VI UM,

AND THE CAUDAL FOLD TUBERCULIN (CFT) TEST AND J OHNE’S DISEASE IN
CATTLE

Abstract

Objective - To determine whether cattle testing positive for Mycobacterium avium subsp.
paratuberculosis (M. paratuberculosis) are more likely to be positive on the y-IFN assay
stimulated by Mycobacterium avium PPD than cattle testing negative. The association
between the M. avium y-IFN assay and the CFT test was also evaluated.

Animals — 1,043 Holstein cattle from 10 herds in Michigan

Procedure — Fecal and whole blood samples were collected from all cattle 224 months of
age on the day the CFT test was read. Fecal samples were submitted for M.
paratuberculosis culture. Samples of plasma were tested for antibody against M.
paratuberculosis, as well as for y-IPN after whole blood samples were stimulated
overnight with phosphate buffered saline solution (control) and purified protein
derivative (PPD) from M. bovis or M. avium.

Results -The multivariable model for M. avium y-IFN test results identified positive M.
paratuberculosis fecal culture and herd prevalence of M. paratuberculosis antibody
ELISA as statistically significant risk factors, with odds ratios of 27.1 (p=0.0066) and 8.4
E5 (p=0.0461), respectively. Cattle positive for M. avium y-IFN were 5.4 times more

likely to be CFT suspects than those negative for M. avium y-IFN (p<0.0001).

45

Conclusions and Clinical Relevance — Cattle positive for M. avium y-IFN were more
likely to be CFT suspects than negative cattle, which may be related to early stages of

J ohne’s disease.

Introduction

Determining whether an animal is infected with J ohne’s disease is a challenge,
especially in the earlier stages of the disease, before a humoral response has begun. The
beginning stage of J ohne’s disease is typically characterized by cell-mediated immunity
to antigen stimulation, which progresses to a strong humoral immune response as the
disease progresses from subclinical to clinical stages (Bendixen, 1977). Because y-IFN
relies on a cell-mediated response, we attempted to use the y-IFN assay, comparing
stimulation of blood with M. avium PPD to phosphate buffered saline (control), as an
indicator for early cell-mediated stages of Johne’s disease. M. avium PPD was used
instead because M. paratuberculosis is a subspecies of M. avium, and Johnin PPD is not
commercially available (McIntyre and Stanford, 1986). The ‘y-IFN response to M. avium,
however, is not an approved diagnostic test for M. paratuberculosis.

A recent study measuring 'y-IFN after stimulation with M. avium PPD and Johnin
PPD concluded that the y-IFN assay could be used to identify subclinical J ohne’s disease,
although the test is highly susceptible to cross-reactivity between mycobacteria species
(J ungerson et al., 2002). This first goal of our study was to measure this association
between individual cattle with positive M. paratuberculosis test results and the 'y-IFN
response to M. avium. To measure this association, our study calculated the odds of a

positive y-IFN response based on positive results by M. paratuberculosis fecal culture

46

and antibody ELISA tests. This assumes that a cell-mediated immune response remains
detectable after progression of Johne’s disease to a humoral response, detected by
ELISA.

The second goal of this study was to evaluate the effect of M. paratuberculosis,
using a positive response of y-IFN to M. avium as well as positive M. paratuberculosis
fecal culture and antibody ELISA, with a positive response on the CFT test for bovine
tuberculosis. This second goal is intended to supplement Chapter 2, by adding positive

M. avium y-IFN results as third diagnostic test for Johne’s disease infection.

Methods
Animals

The study included 1,043 Holstein cattle from 10 separate herds located in three
counties in Michigan. Eight of the herds sampled were within a TB high-risk area
(Alpena county), one in a TB-free area (lngham county), and one near the high-risk area
(Ogemaw county) (Figure 1-1). Criteria for inclusion of a herd in this study were a
willingness to participate, having a total number of cattle in the herd less than 250 (for
cost reasons), and being able to sample the herd on the day when their required whole-
herd TB test was read. All cattle greater than or equal to 24 months of age were sampled
from each herd. None of the herds had been previously vaccinated for M.

paratuberculosis or M. bovis.

47

Tuberculosis skin testing

The CFT tests were performed by either accredited private practice veterinarians
or by veterinarians employed by the Michigan Department of Agriculture (MDA) using
the methods described in the USDA Uniform Methods and Rules (USDA, 1999). Seven
of the 10 herds were CFT tested by the same MDA veterinarian. All comparative
cervical tuberculin (CCT) tests were performed by USDA or MDA veterinarians. Results

of both skin tests were interpreted 72 d: 6 hours after injection of tuberculin.

Sampling

Blood and fecal samples were collected on the day the CF T test was read. A new
plastic sleeve was used for each animal to collect fecal samples from the rectum. The
fecal samples were placed in separate plastic whirl-pack bags afier collection. Fecal
samples were transported in coolers at ambient temperature to the Diagnostic Center for
Population and Animal Health at Michigan State University and stored at -80°C within 24
hours of sampling until cultured. Blood was collected via the middle coccygeal vein
using a 20-gauge, l-in. needle. The blood was collected into a 10 ml Vacutainer tube
containing sodium heparin (Corning Glass Works, Corning, NY). Blood samples were
transported to the lab in plastic coolers, chilled with ice packs, and were processed within

24 hours of when the samples were drawn.

Laboratory methods — A detailed description of laboratory methods is located in the

appendix.

48

M. paratuberculosis testing

M. paratuberculosis laboratory diagnostic tests consisted of fecal culture and
testing for antibodies using a plasma ELISA. Fecal samples were cultured to detect the
presence of M. paratuberculosis using standard procedures recommended by USDA-
National Veterinary Service Laboratory and based on the procedures used by Whitlock et
al. at the University of Pennsylvania (F yock and Whitlock, 1999).

Plasma samples were tested for antibodies to M. paratuberculosis using a
commercial ELISA kit.° The samples were tested in duplicate wells, as recommended by
the manufacturer. The average optical density (OD) was calculated for each pair of
wells. The corrected OD was calculated by subtracting the average OD of two negative
serum control wells from the average OD of duplicate sample wells. A corrected OD

greater than 0.1 was considered positive.

M. avium y—IFN testing

Whole blood samples were tested for y-IFN using a commercially available
antigen capture ELISA, following the manufacturers recommended protocol.7 Using the
OD values, an animal was classified as positive for M. avium y-IFN if the difference
between the mean OD value from the avian PPD samples and the nil antigen samples
were 2 0.1 and the difference between the mean OD value from the avian PPD samples

and bovine PPD samples were also 2 O. 1.

 

b Parachek”, BioCor Animal Health, Omaha NE
7 Bovigam”, BioCor Animal Health, Omaha NE

49

Statistical methods

Prevalence of Johne’s disease within each herd was computed for each type of
J ohne’s disease test (fecal culture, antibody ELISA, or M. avium 'y-IFN) by the number of
positive cattle for each test divided by the total number of cattle tested within each herd.

Statistical analyses were conducted in two parts, using a standard software
packages: 1) testing associations between M. avium y-IFN test results and animal age and
Johne’s disease status; and 2) testing associations between CFT test results and age, CFT
testing veterinarian, and Johne’s disease status. In the first analyses, multivariable
logistic regression models with random effects were developed with the outcome of
positive y-IFN response from M. avium stimulation and risk factors included J ohne’s
disease test results, J ohne’s disease herd prevalence for each test, and animal age. In the
second analyses, model outcome was CFT test status (positive or negative), and risk
factors included testing veterinarian, animal age, results of individual Johne’s disease
tests, and herd prevalence of J ohne’s disease (Table 4-1). Because the analysis was done
on an individual animal level, the random effect function was used in all the models to
adjust for the fact that animals from the same herd are more alike in terms of the
exposure than animals in other herds.

A backwards stepwise model development was used to create final models with
risk factors significant at pS0.05. In brief, a full model was generated, and possible
interactions and confounding were assessed and corrected during the model development
process. Odds ratios (OR) with 95% confidence intervals were computed for parameter
estimates. With the exception of the potential confounders forced into the model, each

risk factor was tested by examining the effects of removal of that factor from the model.

 

8 SAS vs, SAS Institue, Cary. NC

50

If removal of the risk factor resulted in a change in the OR of the remaining variables of

more than 10%, the risk factor and its interaction terms were retained in the model.

Results
Johne ’s disease results

Of the 1043 cattle tested in this study, 45 (4.3 %) were positive for M.
paratuberculosis by fecal culture, 115 (11.0%) were positive by antibody ELISA, and
154 (14.8%) were positive for M. avium y-IFN (Table 4-2). The herd prevalence for each
test ranged from 0 to 15.4% (mean 4.5%, median 2.8%) for M. paratuberculosis fecal
culture, 1.2 to 30.8% (mean 9.7%, median 6.9%) for M. paratuberculosis antibody

ELISA, and 0 to 68.2% (mean 9.3%, median 1.7%) for M. avium y-IFN (Figure 4-1).

TB results

All cattle that responded to the CF T test (any visible or palpable response) were
classified as suspect and retested with the CCT test. Of the 1043 total cattle, 180 were
CFT suspects (17.3%). Herd response ranged from 8.0% to 56.1% (mean 21.2%, median
20.1%) for the CFT test (Figure 4-1). Five of the 180 CFT suspects were also CCT
suspects and a sixth animal was a CCT reactor. The CCT suspects and the CCT reactor
were classified as negative for TB based on necropsy, histological examination of tissues
and mycobacterial culture. Of the cattle positive for M. paratuberculosis by fecal culture,
24.4% were CFT suspects, while 20.9% of the antibody ELISA positive cattle were CFT

suspects, and 19.5% of those positive for M. avium 'y-IF N were CFT suspects.

51

Statistical results

The initial multivariable model for M. avium y-IFN assay results identified
positive individual animal M. paratuberculosis fecal culture and increasing herd
prevalence of M. paratuberculosis antibody ELISA as statistically significant risk factors
for positive result on the 'y-IFN assay, with odds ratios of 27.1 (p=0.0066) and 8.4 E5
(p=0.0461), respectively (Table 4-3).

In the multivariable regression model for the CF T test, y-IFN released from the
stimulation of blood with M. avium PPD was the only statistically significant risk factor
identified (OR=5.4, p<0.0001) (Table 4-4). The individual animal and herd level of M.
paratuberculosis fecal culture and antibody ELISA test results did not significantly affect
the CFT test results. The veterinarian administering the CFT test was identified as a

confounding variable, and therefore controlled for in the modeling process.

Discussion

Cross-reactivity between different mycobacteria have led to the assumption that
infection with M. paratuberculosis may lead to false positive responses to the CFT test
(Morrison et al., 2000). It is important to determine the cause of false positives responses
as they may lead to unnecessary culling, increased time and cost of animal handling,
increased cost of follow-up testing, and possible psychological stress to producers and
veterinarians.

In the current study, Johne’s disease infection was classified according to M.
paratuberculosis fecal culture and antibody ELISA results, and y-IFN response when

stimulated by M. avium. As mentioned earlier, M. avium y-IFN is not an approved

52

diagnostic test for J ohne’s disease but was used for this purpose in our study. We
calculated the odds of a positive response of 'y-IFN stimulated by M. avium, based on
positive results from M. paratuberculosis fecal culture and antibody ELISA. After
learning the relationship between M. avium y-IFN and other approved Johne’s disease
diagnostic tests, we evaluated the effect these diagnostic tests had on the outcome of the
CFT test, including the y-IFN response to M. avium stimulation.

Because J ohne’s disease is thought to progress from cell-mediated immunity to a
strong humoral immune response as the disease progresses from subclinical to clinical
stages, we attempted to use M. avium y~IFN as an indicator for early stages of Johne’s
disease where the immune response is predominately cell-mediated. For later stages of
J ohne’s disease that have progressed to humoral immunity, we attempted to detect
antibodies using antibody ELISA and to detect shedding bacteria with fecal culture.

First, we examined the effect of individual Johne’s disease test results on the
outcome of a positive M. avium 'y-IFN result. Cattle positive for M. paratuberculosis
fecal culture were 27.1 times more likely to be positive for M. avium 'y-IFN than cattle
with a negative culture (Table 4-3). Herds with a higher prevalence of cattle testing
positive for M. paratuberculosis antibody ELISA were also more likely to be positive for
M. avium y-IFN than cattle from herds with lower prevalence. The 95% confidence
interval for the odds ratio was large for ELISA herd prevalence, so it is difficult to
comment on the true effect on a positive M. avium y-IFN response. This relationship
suggests that M. avium y-IFN may be detecting a portion of cattle infected with M

paratuberculosis.

53

Cattle positive for M. avium y-IFN were 5.4 times more likely to be CF T suspects
than cattle negative for M. avium y-IFN (Table 4-4). A positive trend existed between
positive M. paratuberculosis fecal culture and antibody ELISA tests and false positive
reactions to the CFT test, however this association was not statistically significant (See
Chapter 2).

Although positive M. paratuberculosis fecal culture results were associated with
positive M. avium y-IFN results, this study was not designed to determine whether M.
avium y-IFN indicates early or late stages of J ohne’s disease. The association was
significant, however, which may indicate that cattle infected with M. paratuberculosis
can successfully be detected with M. avium y-IFN, or the association may also be due to
an increased false positive cross-reactivity for the y-IF N assay due to the genetic
relationship between M. avium and M. paratuberculosis. Animals with positive M.
avium 'y-IF N results were more likely to be CFT suspects compared to negative animals.
It is difficult to conclude whether this association was due to proteins in common with
PPD since both tests are a measurement of cell~mediated immunity based on PPD, or if
specific infection with Johne’s disease causing positive M. avium y-IFN results resulted
in the false positive CF T results.

Although positive M. paratuberculosis fecal culture results were associated with
positive M. avium y-IF N results, cattle positive for M. paratuberculosis fecal culture were
not more likely to be CFT suspects. Several possible explanations may account for this
lack of significant association. First, the sensitivity of fecal culture is low, due to the fact
that the bacteria are only shed intermittently and in low numbers for most stages of

Johne’s disease. A low sensitivity would result in a large number of false positives,

54

which would decrease the likelihood of detecting an association due to the
misclassification. Second, the sample size may have been too small to detect a true
association between positive fecal culture and a positive CFT test. In this study, the
difference in CF T test results were only 7% between cattle testing positive and negative
by M. paratuberculosis fecal culture. The sample size used for this study was based on
the ability to detect a 25% difference in the CFT test results between cattle testing
positive and negative for fecal culture, thus the power to detect a true association was not
adequate. Third, all ten herds in this study had a prevalence of positive fecal culture of
15% or less, such that we did not have any high prevalence herds which would have
increased the number of positive cattle, increasing the ability to detect individual and
herd level associations with positive CFT results. Fourth, despite the mild trend between
M. paratuberculosis positive cattle and positive CF T results, it is possible that M.
paratuberculosis may not significantly affect the CFT test. Either way, further studies
with a larger sample size are necessary to confirm whether an association truly exists or
not. Determining that a true association exists would help people in TB eradication
programs decide whether the CF T test could be most useful in J ohne’s disease-free herds
or if a high number of false positives are inherent to the test despite Johne’s disease
status.

Future studies are important to determine whether y-IFN is successful at detecting
early stages of Johne’s disease, undetectable by M. paratuberculosis fecal culture or
antibody ELISA. If v-IFN proves to be an accurate diagnostic assay for subclinical
Johne’s disease, the high number of CFT suspects in this study may indicate that the

majority of subclinically infected cattle in these herds are predominately exhibiting cell-

55

mediated immunity and haven’t progressed to humoral immunity. The result could be a
large number of CFT false positives in regions that have a high prevalence of Johne’s
disease.

Finally, it is important that research continues for improving diagnostic tests for
TB that successfully compares M. bovis to M. avium and M. paratuberculosis during the
initial round of testing. The recent mapping of the complete genome sequence for M.
bovis allows specific proteins to be identified that are unique to M. bovis (Garnier et al.,
2003). A recently published study identified a mixture of the proteins ESAT-6, CFP-lO,
TB27.4, and TBlO.4 as a potential novel cocktail for the development of new tests for M.
bovis (Aagaard et al., 2003). Further work identifying new diagnostic tests that detect

antigens specific to M. bovis will dramatically aid in the eradication of TB.

56

Table 4-1: Description and descriptive statistics for risk factors evaluated

 

 

 

 

 

 

 

 

 

 

 

 

 

Outcome Values Description 11 %*
1 Positive 180 17
CFT Test Result _
O Negatrve 863 83
Risk Factor Values Description 11 %
Veterinarian l
l 195 19
(CFT tested I herd)
Veterinarian 2
2 94 9
(CFT tested I herd)
CFT Vet
Veterinarian 3
3 647 62
(CFT tested 7 herds)
Veterinarian 4
4 107 10
(CFT tested I herd)
1 Positive 154 15
M. avium y-IFN
O Negative 889 85
1 Positive 115 11
M. paratuberculosis ELISA
0 Negative 928 89
1 Positive 45 4
M. paratuberculosis Culture .
O Negatrve 998 96
M. paratuberculosis ELISA and 1 Positive for both 16 2
Culture 0 Negative for one or both 1027 98
Continuous variables
Age (months) 24-144 Age of animal in months 1043 100
% cattle positive for M.
M. paratuberculosis Culture
0-15.4 paratuberculosis fecal culture 1043 100
Herd Prevalence
for each herd
% cattle positive for M.
M. paratuberculosis ELISA 1.2- .
paratuberculosis antibody 1043 100
Herd Prevalence 30.8
ELISA for each herd
M. avium y-IF N Herd % cattle positive for M. avium
O-68.2 1043 100

Prevalence

y-IFN for each herd

 

*Percentage of animals described by the respective value (=n/1043)

57

Table 4-2: Descriptive analysis of M. avium y-IFN results based on positive Johne’s
disease test results

 

 

 

 

 

All Cattle M. avium y-IFN M. avium 7-IF N
Johne’s test positive negative
(positive results) (n=1043) (n=154) (n=889)
# % # % # %
Fecal Culture 45 4.3 12 7.8 33 3.7
Antibody ELISA 115 11.0 35 22.7 80 9.0
ELISA + Culture 16 1.5 6 3.9 10 1.1
Negative for both 899 86.2 1 13 73.4 786 88.4

 

 

 

 

 

 

58

Table 4—3: Descriptive analysis of CFT test results based on positive Johne’s disease
test results

 

Johne’s test A" Cattle CFT suspect CFT negative

 

 

. . (n=1043) (n=180) (n=863)
(posrtrve results) # 0 A1 # % # %
Fecal Culture 45 4.3 1 1 6.1 34 3.9
Antibody ELISA 115 11.0 24 13.3 91 10.5

 

M. avium 'y-IFN 154 14.8 30 16.7 124 14.4

 

Johne’snegative 749 71.8 121 67.2 628 72.8

 

 

 

 

 

 

59

Table 44: Results of a multivariable logistic regression analysis of the effect of
Johne’s disease test results on the M. avium 7-IF N test as the outcome.

 

 

Final Model
x2
Risk Factor Estimate S.E. OR 95% CI.
P
M. paratuberculosis Culture 3.30 1.21 .01 27.1 2.52-292

 

M. paratuberculosis ELISA Herd 5 11
13.64 6.83 .05 8.4E 1.28-5.5E

 

 

 

Prevalence
M. paratuberculosis Culture Herd 4 2.0E4T
11.09 12.3 .37 6.5E 15
Prevalence 2. 1 E
Age (months) -.003 .01 .66 1.00 .98-1.01
Age (months)*

-.02 .02 .33 .98 .93-1.02
M. paratuberculosis Culture

 

Model —2 Res Log Likelihood = 6796.4

 

60

Table 4-5: Results of a multivariable logistic regression analysis of the effect of
individual Johne’s Disease test results and herd levels on the CFT test as the
outcome (classified as suspect or negative).

 

 

 

 

Final Model
Risk Factor Estimate S.E. X2 p OR 95% C].
M. avium y-IFN 1.69 .40 < .0001 5.45 247-1203
Age (months) .01 .01 .07 1.01 999-102
1 -2.07 1.22 .09 0.13 .01-1.38
CFT 2 -.95 1.21 .43 0.39 .04-4.13
Vet 3 .13 0.89 .89 1.13 .20-6.49
4 0 - - 1 -

 

Model —-2 Res Log Likelihood = 5154.9

 

61

5535.53 the:

$21.5 sans 3.15 £15 515 c.1115 €215 A815 A315 @215

  

 

 

 

 

 

 

 

 

 

 

3 a w h e m w m N _
HIWm _ w
” 1 S
/
/ .1
” cm
/
M 1 an
0 1 an 358.—
Z11:1> 25.33 SD / .x.
W 1 cm
<m3m .
€898 £3323553& .3 I L 1 cc
23:3 Room MioanmaEEom fi 1 E.
1 ca
88913 150a

 

 

 

Eu.— .3 3.39. «>53; .3 «wax—3.5.. 2: he .3qu “TV 01.5w:

62

SUMMARY AND CONCLUSIONS

In the first chapter, an introduction to Johne’s disease and bovine tuberculosis
(TB) was presented along with a review of current ante-mortem diagnostic tests used for
each disease. Johne’s disease and TB both exist in Michigan, with spread of TB in cattle
occurring from infected white-tailed deer. A major problem for TB screening tests is the
relatively high number of false positive responses, thought to occur as a result of cross-
reactivity between mycobacteria and other closely related species. Due to the high
prevalence of J ohne’s disease among Michigan herds, information was needed about the
association between infection with M. paratuberculosis and false positive responses to
the CFT and y-IFN screening tests for TB.

The second chapter described a cross-sectional study performed in Michigan
evaluating the effect of Mycobacterium avium subsp. paratuberculosis (M.
paratuberculosis) on the CFT test for TB in cattle. Cattle testing positive for M.
paratuberculosis, as measured by a positive M. paratuberculosis fecal culture or antibody
ELISA test, appear to have an increased likelihood of false positive results on the CF T
test, although the association is not statistically significant. Further studies involving a
larger sample size need to be conducted to confirm these findings.

The third chapter was an evaluation of the effect of M. paratuberculosis on the y-
IFN assay for TB in cattle, using the same animals and herd visits described in chapter 2.
No significant association was found between cattle testing positive by M.
paratuberculosis fecal culture or antibody ELISA. and positive results of the 'y-IFN assay
for bovine tuberculosis. It may not be valid to conclude whether an association truly

exists because only eight cattle were positive for y-IFN stimulated by M. bovis. None of

63

the eight cattle with a positive y-IFN response were also positive for either M.
paratuberculosis fecal culture or ELISA. Further studies involving a large sample size
with more false positives need to be conducted to confirm these findings.

The fourth chapter looked at the possibility of y-IFN, stimulated by M. avium, as a
test for diagnosis of early stages of Johne’s disease. Cattle positive for M.
paratuberculosis fecal culture were 27.1 times more likely to be positive for M. avium y-
IFN than cattle with a negative culture. Herds with a higher prevalence of cattle testing
positive for M. paratuberculosis antibody ELISA were also more likely to be positive for
M. avium y-IFN than cattle from herds with lower prevalence. Cattle positive for M.
avium y—IF N were 5.4 times more likely to be CF T suspects than cattle negative for M.
avium y-IF N. It is difficult to conclude whether this association was due to proteins in
common with PPD since both tests are a measurement of cell-mediated immunity based
on PPD, or if true infection detected by M. avium y—[FN resulted in the false positive C FT
results. Further work is necessary to determine whether M. avium y-IFN is detecting
early stages of Johne’s disease and whether early stage Johne’s disease is the reason for
an increased likelihood of positive CFT test results, as measured by M. avium y-IFN.
Confirmation of the results in these studies will help to determine necessary
improvements for screening tests currently used for detection or TB. As diagnostic tests
are improved, further work identifying new diagnostic tests that detect antigens specific

to M. bovis will dramatically aid in the eradication of TB.

64

APPENDIX

65

DETAILED LABORATORY PROTOCOL

Fecal Culture

Fecal culture was performed on all samples following a protocol used by
Whitlock et al. at the University of Pennsylvania (F yock and Whitlock, 1999). Samples
were stored at -80°C until processed. Two grams of each fecal sample were diluted with
sterile distilled water in a labeled tube, shaken and allowed to stand for 30 minutes before
transferring 5mL from the center of the supernatant into a tube containing 1/2XBHI-HPC
(Half strength Brain Heart Infusion [BHI] broth with l-Hexadecylpyridinium Chloride).
The tube was then incubated overnight at 35-3 7°C. After incubation, the samples were
centrifuged at 900 x G for 30 minutes. Supernatant was then discarded and lmL of
antibiotic brew was added with a sterile disposable pipet, followed by incubation
overnight at 35-3 7°C. Antibiotic brew consists of 1 Liter of half strength BHI broth
mixed with 5mL Amphotericin B (lOmg/mL), 10mL Vancomycin (lOmg/mL), and 5mL
Naladixic Acid (10mg/mL).

Inoculation of the media took place either the next day or the following two days.
Five tubes of Herrold’s Egg Yolk Agar, four containing Mycobactin J and one without
Mycobactin J, were then inoculated with 0.25mL of well mixed suspension. Tubes were
rolled to coat the entire service of media with inoculum. Tubes were incubated in a 35-
37°C aerobic incubator in a slanted position with the media service horizontal until the
inoculum dried. Afier drying, the tubes were returned to upright position and the caps
tightened. Tubes were evaluated for grth and contamination every 2 weeks for 16

weeks. Colonies appearing after 6 weeks were evaluated for typical acid fastness and

66

morphological appearance of M. paratuberculosis. Each culture was then evaluated for

mycobactin dependency before being reported as positive.

Antibody ELISA

Plasma samples were tested for antibodies to M. paratuberculosis with a
commercial antibody ELISA test kitb, using the manufacturer’s recommended protocol.
Briefly, 25 uL of plasma and control samples were combined with serum diluent buffer,
mixed and incubated for 1 hour at room temperature. Next, 100uL of test and control
samples were added to microtiter plates containing bound M. paratuberculosis antigens
and incubated at room temperature for 30 minutes. The microtiter wells were then
washed 6 times with an appropriately diluted buffer supplied by the manufacturer. Next,
lOOuL of horseradish peroxidase labeled anti-bovine lg was added to each well and
incubated at room temperature for 30 minutes. Plates were washed six times with wash
buffer, followed by the addition of lOOuL enzyme substrate solution to each well.
Finally, 50uL of enzyme stopping solution (0.5M H2804) was added to each well once
positive control wells had absorbance between 0.35 and 0.40 using a 620nm filter. The
absorbance was then read using a 450nm filter between 2 and 5 minutes after stopping

the reaction.

y-IFN testing
Whole blood samples were tested for y-IFN using a commercially available
antigen capture ELISA, following the manufacturers recommended protocol.9

Heparinzed blood was separated into 1.5mL aliquots and placed into four wells of a 24

 

9 Bovigam“, BioCor Animal Health, Omaha NE

67

well tissue culture plate for each animal. To the first well, lOOuL phosphate buffered
saline (0.01M, pH 7.2) was added as a negative control (nil antigen). In the second and
third wells, lOOuL bovine PPD10 or avian PPDll were added to stimulate the release of y-
IFN, respectively. Pokeweed mitogen12 was added to the fourth well to a final
concentration of 10ug/mL of blood (Stabel, 1996). The pokeweed mitogen served as a
control for cell function, as this mitogen stimulates production and release of y-IFN from
mononuclear leukocytes. Samples were incubated for 16-24 hours at 38°C in a
humidified atmosphere. Samples were centrifuged at 1,200 x g for 15 minutes following
incubation and 500uL of plasma were harvested from each well. For the test procedure,
SOuL of test plasma and an appropriate number of positive and negative control samples
provided by the manufacturer were combined with plasma diluent buffer, mixed and
incubated for 60 minutes at room temperature. The microtiter wells were then washed
out and filled with wash buffer six times, removing as much wash buffer as possible after
the last wash. Next, lOOuL of horseradish peroxidase labeled anti—bovine y-IFN was
added to each well and incubated at room temperature for 60 minutes. Plates were
washed six times with wash buffer as earlier, followed by the addition of lOOuL enzyme
substrate solution to each well, and then incubated for 30 minutes. Finally, 50uL enzyme
stopping solution (0.5M H2SO4) was added to each well at the end of the incubation and
the absorbance was read using a 450nm filter within 5 minutes of stopping the reaction.
Using the absorbance values measured by the spectrophotometer, or optical density (OD)
values, an animal was classified as positive for M. avium y-IF N if the difference between

the mean OD value from the avian PPD samples and the nil antigen samples were 2 0.1

 

‘0 CSL Limited, Victoria, Australia, Cat. No. 20901501
" CSL Limited, Victoria, Australia, Cat. No. 20911501
'2 Phytolacca americana, cell culture tested, Sigma. St. Louis. MO

68

and the difference between the mean OD value from the avian PPD samples and bovine
PPD samples were also _>_ 0.1. An animal was classified as positive for M bovis y-IFN if
the difference between the mean OD value from the bovine PPD samples and the nil
antigen samples were 2 0.1 and the difference between the mean OD value from the

bovine PPD samples and avian PPD samples were also 2 0.1.

69

REFERENCES

70

REFERENCES

Aagaard C, Govaerts M, Okkels LM, et al. Genomic approach to identification of
Mycobacterium bovis diagnostic antigens in cattle. J Clin Microbiol 2003;
41(8):3719-3728.

Animal and Plant Health Inspection Service (APHIS). USDA amends bovine
tuberculosis classification of New Mexico. Press Release July 24, 2003.
Washington, DC.

Bendixen PH. Immunologic reactions caused by infection with Mycobacterium
paratuberculosis. A Review. Nord Vet Med 1978; 30:163-168.

Biocor Animal Health. Veterinary Diagnostic Newsletter. A quarterly bulletin, Fall
2002. Omaha, NE.

Braun RK, Buergelt CD, Litell RC, et a1. Use of an enzyme-linked immunosorbent assay
to estimate prevalence of paratuberculosis in cattle of Florida. J Am Vet Med
Assoc 1990;196:1251-1254.

Bruning-Fann CS, Schmitt SM, Fitzgerald SD, et al. Bovine tuberculosis in free-ranging
carnivores from Michigan. J Wildl Dis 2001; 37(1):58-64.

Buergelt CD, Hall C, McEntee K, et al. Pathological evaluation of paratuberculosis in
naturally infected cattle. Vet Pathol 1978; 15:196-207.

California Department of Food and Agriculture (CDFA). Bovine tuberculosis found in
cattle. News Release June 3, 2002. Sacramento, CA.

Collins C, Kelly P, Bryne C, et al. Is bovine, atypical or resistant tuberculosis a problem?
Ir Med J 1987; 80(2):66-67.

Collins MT, Kenefick KB, Sockett DC, et al. Enhanced radiometric detection of
Mycobacterium paratuberculosis by using filter-concentrated bovine fecal
specimens. J Clin Microbiol 1990; 28:2514-2519.

Collins MT, Sockett DC, Ridge S, et al. Evaluation of a commercial enzyme-linked
immunosorbent assay for Johne’s disease. J Clin Microbiol 1991; 29(2):272-276.

Cox JC, Drane D, Johnes SL, et al. Development and evaluation of a rapid absorbed
enzyme immunoassay test for the diagnosis of J ohne’s disease in cattle. Aust Vet J
1991; 68:157-60.

Ellis TM, Carson BA, Kalkhoven MJ, et al. Specificity of two absorbed ELISAs for

surveys of Mycobacterium paratuberculosis in cattle. Aust Vet J 1998; 76(7):497-
499.

71

Feng CG, Bean AG, Hooi H, et al. Increase in gamma interferon-secreting CD8(+), as
well as CD4(+), T cells in lungs following aerosol infection with Mycobacterium
tuberculosis. Infect Immun 1999; 67:3242-3247.

Flesch IE, Kaufrnann SH. Mechanisms involved in mycobacterial growth inhibition by
gamma-interferon-activated bone marrow macrophages: role of reactive nitrogen
intermediates. Infect Immun 1991; 59:3213-3218.

Flynn J L, Chan J. Immunology of tuberculosis. Annu Rev Immunol 2001; 19:93-129.

Francis J, Seiler RJ, Wilkie IW, et al. The sensitivity and specificity of various
tuberculin tests using bovine PPD and other tuberculins. Vet Rec 1978;
103(19):420-425.

F yock TL and Whitlock RH. Protocol for culture of M. paratuberculosis from feces.
University of Pennsylvania School of Veterinary Medicine, New Bolton Center.
Kennett Square. PA. September 29, 1999.

Gamier T, Eiglmeier K, Camus JC, et al. The complete genome sequence of
Mycobacterium bovis. Proc Natl Acad Sci U S A 2003; 100(13):7877-7882.

Hagan WA. Age as a factor in susceptiblility to Johne’s disease. Cornell Vet 1938;
28:34-40.

Hope JC, Kwong LS, Sopp P, et a1. Dendritic cells induce CD4+ and CD8+ T-cell
responses to Mycobacterium bovis and M. avium antigens in Bacille Calmette
Guerin vaccinated and nonvaccinated cattle. Scand J Immunol 2000; 52:285-291.

Inwald J, Hinds J, Palmer S, et al. Genomic analysis of Mycobacterium tuberculosis
complex strains used for production of purified protein derivative. J Clin
Microbial 2003; 41(8):3929-3932.

Johnson-Ifearulundu Y, Kaneene JB. Distribution and environmental risk factors for
paratuberculosis in dairy cattle herds in Michigan. Am J Vet Res 1999; 60(5):589-
596.

Johnson-Ifearulundu YJ, Kaneene J B, Sprecher DJ, et al. The effect of subclinical
Mycobacterium paratuberculosis infection on days open in Michigan, USA, dairy
cows. Prev Vet Med 2000; 46:171-181.

J ungersen G, Huda A, Hansen JJ, et al. Interpretation of the gamma interferon test for
diagnosis of subclinical paratuberculosis in cattle. Clin Diagn Lab Immun 2002;
9(2):453-460.

72

Kalis CH, Collins MT, Hesselink J W, et al. Specificity of two tests for the early
diagnosis of bovine paratuberculosis based on cell-mediated immunity: the Johnin
skin test and the gamma interferon assay. Vet Microbial 2003;97(1-2):73-86.

Kaneene JB, Bruning-Fann CS, Dunn J, et al. Epidemiologic investigation of
Mycobacterium bovis in a population of cats. Am J Vet Res 2002; 63(1 l):1507-
151 1.

Kaneene JB, Bruning- Fann CS, Granger LM, et al. Environmental and farm management
factors associated with tuberculosis on cattle farms in northeastern Michigan. J
Am Vet Med Assoc 2002; 221(6). 837- 842.

Kaneene JB, VanderKlok M, Bruning-Fann CS, et al. Prevalence of Mycobacterium
bovis infection in cervids on privately owned ranches. J Am Vet Med Assoc 2002;
220(5):656-659.

Karlson AG. Nonspecific or cross-sensitivity reactions to tuberculin in cattle. Adv Vet Sci
1962; 7:148-175.

Ketterer PJ, Rogers RJ, Donald B. Pathology and tuberculin sensitivity in cattle
inoculated with Mycobacterium avium complex serotypes 6, 14, and 18. Aust Vet
J 1981; 57: 61-65.

Kohler H, Gyra H, Zimmer K, et al. Immune reactions in cattle after immunization with
a Mycobacterium paratuberculosis vaccine and implications for the diagnosis of
M. paratuberculosis and M. bovis infections. J Vet Med B 2001; 48:185-195.

Kreeger J M. Ruminant paratuberculosis — A century of progress and frustration. J Vet
Diagn Invest 1991; 32373-383.

Lauzi S, Pasotto D, Amadori M, et al. Evaluation of the specificity of the y-interferon
test in Italian bovine tuberculosis-free herds. Vet J 2000; 160: 1 7-24.

Lesslie IW, Hebert CN. Comparison of the specificity of human and bovine tuberculin
PPD for testing cattle. 3. National trial in Great Britain. Vet Rec 1975; 96:332-
334.

Liebana E, Girvin RM, Welsh M, et a]. Generation of CD8+ T-cell responses to
Mycobacterium bovis and mycobacterial antigen in experimental bovine
tuberculosis. Infect Immun 1999; 67: 1034-1044.

McIlroy SG, Neill SD, McCracken RM. Pulmonary lesions and Mycobacterium bovis
excretion from the respiratory tract of tuberculin reacting cattle. Vet Rec 1986;
118:718-721.

73

McIntyre G, Stanford J L. Irnmunodiffusion analysis shows that Mycobacterium
paratuberculosis and other mycobactin-dependent mycobacteria are variants of
Mycobacterium avium. J Appl Bacterial 1986; 61 :295-298.

Menzies RJ, Neill SD. Cattle-to-cattle transmission of bovine tuberculosis. Vet J 2000;
160292-106.

Michigan Bovine TB Web site. Maps: TB Positive Animals. Available at:
http://www.bovinetb.com. Accessed Sep 24, 2003.

Miller R, Kaneene JB, Fitzgerald SD, et al. Evaluation of the influence of supplemental
feeding of white-tailed deer (Odocoileus virginianus) on the prevalence of bovine
tuberculosis in the Michigan wild deer population. J Wildl Dis 2003; 39(1):84-95.

Milner AR, Mack WN, Coates K], et al. The sensitivity and specificity of a modified
ELISA for the diagnosis of J ohne’s disease from a field trial in cattle. Vet
Microbial 1990; 25:193-198.

Monaghan ML, Doherty ML, Collins JD, et al. The tuberculin test. Vet Microbiol 1994;
40:111-124.

Morris RS, Pfeiffer DU, Jackson R. The epidemiology of Mycobacterium bovis
infections. Vet Microbiol 1994; 40:1 53-177.

Morrison WI, Boume FJ, Cox DR, et al. Pathogenesis and diagnosis of infections with
Mycobacterium bovis in cattle. Vet Rec 2000; 146:236-242.

Mustafa AS, Kvalheim G, Degre M, et al. Mycobacterium bovis ECG-induced human T-
cell clones from ECG-vaccinated healthy subjects: antigen specificity and
lymphokine production. Infect Immun 1986: 53:491-497.

Palmer MV, Whipple DL, Waters WR. Experimental deer-to-deer transmission of
Mycobacterium bovis. Am J Vet Res 2001; 62(5):692-696.

Payeur JB, Church S, Mosher L, et al. Bovine tuberculosis in Michigan wildlife. Ann N
YAcad Sci 2002; 969:259-261.

Pollock JM, Pollock DA, Campbell DG, et al. Dynamic changes in circulating and
anti gen-responsive T-cell subpopulations post-Mycobacterium bovis infection in

cattle. Immunology 1996; 87:236-241.

Radunz BL, Lepper AW. Suppression of skin reactivity to bovine tuberculin in repeat
tests. Aust VetJ1985; 62:191-194.

74

Reichel MP, Kittelberger R, Penrose ME, et al. Comparison of serological tests and
faecal culture for the detection of Mycobacterium avium subsp. paratuberculosis
infection in cattle and analysis of the antigens involved. Vet Microbial 1999;
66: 135-150.

Ridge SE, Morgan IR, Sockett DC, et al. Comparison of the Johne’s absorbed EIA and
the complement-fixation test for the diagnosis of J ohne’s disease in cattle. Am Vet
J 1991; 68:253-257.

Rothel J S, Jones SL, Corner LA, et al. The gamma-interferon assay for diagnosis of
bovine tuberculosis in cattle: conditions affecting the production of gamma-
interferon in whole blood culture. Aust Vet J 1992; 69(1): 1-4.

Ryan TJ, Buddle BM, DeLisle GW. An evaluation of the gamma interferon test for
detecting bovine tuberculosis in cattle 8 to 28 days after tuberculin skin testing.
Res Vet Sci 2000; 69:57-61.

Sanftleben P. Quest continues for fast, reliable test for bovine paratuberculosis.
J Am Vet Med Assoc 1990; 197(3):299-305

Scacchia M, Lelli R, Petrini A, et al. Use of innovative methods in the eradication of
bovine tuberculosis. J Vet Med B 2000; 47:321-327.

Schmitt SM, O’Brien DJ, Bruning-Fann CS, et al. Bovine tuberculosis in Michigan
wildlife and livestock. Ann N Y Acad Sci 2002; 969:262-268.

Skinner MA, Parlane N, McCarthy A, et al. Cytotoxic T-cell responses to
Mycobacterium bovis during experimental infection of cattle with bovine
tuberculosis. Immunology 2003; 110:234-241.

Sockett DC, Carr DJ, Collins MT. Evaluation of conventional and radiometric fecal
culture and a commercial DNA probe for diagnosis of Mycobacterium
paratuberculosis infections in cattle. Can J Vet Res 1992; 56:148-153.

Sockett DC, Conrad TA, Thomas CB, et al. Evaluation of four serological tests for
bovine paratuberculosis. J Clin Microbial 1992; 30:1134-1139.

Stabel J R. Production of y-interferon by peripheral blood mononuclear cells: an
important diagnostic tool for detection of subclinical paratuberculosis. J Vet
Diagn Invest 1996; 8:345-350.

Streeter RN, Hoffsis GF, Bech-Nielsen S, et al. Isolation of M. paratuberculosis from
colostrum and milk of subclinically infected cows. Am J Vet Res 1995; 5621322-
1324.

75

Sweeney RW. Transmission of paratuberculosis. Vet C [in North Am F oad Anim Pract
1996; 12(2):305-312.

Sweeney RW, Whitlock RH, Rosenberger AE. Mycobacterium paratuberculosis isolated
from fetuses of infected cows not manifesting signs of the disease, Am J Vet Res
1992; 53(4):477-480

Tan J S, Canaday DH, Boom WH, et al. Human alveolar T lymphocyte responses to
Mycobacterium tuberculosis antigens: role for CD4+ and CD8+ cytotoxic T cells
and relative resistance of alveolar macrophages to lysis. J Immunol 1997;
159:290-297.‘

Texas Animal Health Commission (TAHC). It’s official. Texas loses cattle tuberculosis
“free” status. News Release June 7, 2002. Austin, TX.

Thoen CO, Baum KH. Current knowledge on paratuberculosis. J Am Vet Med Assoc
1988;192:1609-161 l.

USDA. Bovine Tuberculosis Eradication. Uniform Methods and Rules, Effective January
22, 1999. USDA-APHIS 91-45-011.

USDA. Center for Veterinary Biologics Notice No. 02-03. Issuance of Licenses.
Jaunuary 8, 2002.

Whipple DL, Bolin CA, Davis AJ, et al. Comparison of the sensitivity of the caudal fold

skin test and a commercial y-interferon assay for diagnosis of bovine tuberculosis.
Am J Vet Res 1995; 56(4):415-419.

Whipple DL, Bolin CA, Miller J M. Distribution of lesions in cattle infected with
Mycobacterium bovis. J Vet Diagn Invest 1996; 8:351-354.

Whipple DL, Callihan DR, Jamagin J L. Cultivation of Mycobacterium paratuberculosis
from bovine fecal specimens and a suggested standardized procedure. J Vet Diag
Invest 1991; 31368-373.

Whipple DL, Palmer MV, Slaughter RB, et al. Comparison of purified protein
derivatives and effect of skin testing on results of a commercial gamma interferon
assay for diagnosis of tuberculosis in cattle. J Vet Diagn Invest 2001; 13:117-
122.

Whitlock RH, Buergelt C. Preclinical and clinical manifestations of paratuberculosis
(including pathology). Vet Clin North Am F 00d Anim Pract 1996;12(2):345-56.

Wood PR, Comer LA, Rothel J S, et al. A field evaluation of serological and cellular
diagnostic tests for bovine tuberculosis. Vet Microbial 1992; 31:71-79.

76

Wood PR, Comer LA, Rothel J S, et al. Field comparison of the interferon-gamma assay
and the intradermal tuberculin test for the diagnosis of bovine tuberculosis. Aust
Vet J 1991; 68:286-290.

Wood PR, Rothel J S. In vitro immunodiagnostic assays for bovine tuberculosis. Vet
Microbial 1994; 40:125-135.

Zimmer K, Drager KG, Klawonn W, et al. Contribution to the diagnosis of J ohne’s
Disease in cattle. Comparative studies on the validity of Ziehl-Neelsen staining,
faecal culture and a commercially available DNA-Probe test in detecting
Mycobacterium paratuberculosis in faeces from cattle. J Vet Med B 1999;

46: 1 3 7- 140.

77

 

11111111111111111111