Content uploaded by Daniel Remick
Author content
All content in this area was uploaded by Daniel Remick on May 07, 2023
Content may be subject to copyright.
Applied Occupational and Environmental Hygiene, 18: 953–960, 2003
Copyright c
Applied Industrial Hygiene
ISSN: 1047-322X print / 1521-0898 online
DOI: 10.1080/10473220390237683
Evaluation of Hypersensitivity Pneumonitis Among
Workers Exposed to Metal Removal Fluids
Douglas Trout,1David N. Weissman,2Daniel Lewis,2Rodney A. Brundage,2
Alfred Franzblau,3,4and Daniel Remick4,∗
1Division of Surveillance, Hazard Evaluations, and Field Studies, National Institute for Occupational
Safety and Health, Cincinnati, Ohio; 2Health Effects Laboratory Division, National Institute
for Occupational Safety and Health, Morgantown, West Virginia; 3University of Michigan School of Public
Health, Ann Arbor, Michigan; 4University of Michigan Medical School, Ann Arbor, Michigan
Hypersensitivity pneumonitis (HP) was identified among
employees in an automobile parts manufacturing facility.
Mycobacteria immunogenum (MI) was identified as a metal
removal fluid (MRF) contaminant at this facility and had
been identified as a contaminant in other facilities where HP
had occurred. We therefore questioned whether measure-
ment of MI-specific cell-mediated immunity would be asso-
ciated with HP in this facility. We also questioned whether
measures of cell-mediated immunity would be more infor-
mative about the presence of HP than evaluation of serum
anti-MI antibody levels. Workers were categorized for expo-
sure and disease status by questionnaire and review of med-
ical records. Cell-mediated immunity to MI was assessed
by measuring in vitro secretion of cytokines (interleukin
8, tumor necrosis factor alpha, and interferon-γ)from pe-
ripheral blood mononuclear cells or anticoagulated whole
blood induced by culture with MI antigen. Serum antibod-
ies against MI were also measured. Six study participants
met our survey definition for HP and 48 did not. As has
been reported for various agents causing HP, serum an-
tibody levels against MI were increased in both exposed
workers and workers with HP. Serum antibodies did not
distinguish between the two. When expressed as a percent-
age of secretion induced by lipopolysaccharide, MI induced
a significant increase in interleukin-8 secretion in exposed
participants’ whole blood cultures. There were trends for
increased MI-induced secretion of interferon-γby periph-
eral blood mononuclear cells from both exposed workers and
workers with HP. However, these trends did not attain sta-
tistical significance. Thus, several measures of immunity to
MI distinguished between exposed and unexposed workers
∗Supported in part by National Institutes for Health grant GM 50401.
but not between workers with and without HP. These eval-
uations of cell-mediated immunity were not more informa-
tive than measurement of serum antibodies. As was done at
this facility, institution of a comprehensive safety and health
plan for MRF is necessary to eliminate (or minimize) health
effects related to occupational exposures in the machining
environment.
Keywords Mycobacterium Immunogenum, Hypersensitivity Pneu-
monitis, Metalworking Fluids, Metal Removal Fluids,
Coolant, Machining
In 2002, the National Institute for Occupational Safety and
Health (NIOSH) received a request for a health hazard evalu-
ation (HHE) from an automobile parts manufacturing facility.
The request concerned respiratory problems and hypersensitiv-
ity pneumonitis (HP) thought to be associated with occupational
exposures to metal removal fluids (MRFs). In response to the
HHE request, NIOSH investigators performed a record review
and a questionnaire and serologic survey among employees in
selected departments at the facility. The goal of this evaluation
was to provide further understanding of potential causes of HP,
which had been diagnosed among some employees at the facil-
ity. This report summarizes the results of this evaluation.
BACKGROUND
At the time of the survey, the facility was three years old
and employed approximately 2000 workers in the manufacture
of automatic transmissions and transmission components. The
facility has a medical department with on-site physician and
nursing staff. Multiple central MRF systems supplied a semi-
synthetic MRF to machines performing a variety of types of
machining; the metals being machined were primarily steel and
aluminum. Machine tools at this facility are generally enclosed
953
954 D. TROUT ET AL.
and exhausted to the outdoors, and 55 percent (28 of 51) of
recent industrial hygiene samples of MRF aerosol (total partic-
ulate) were below 0.2 milligrams per cubic meter (mg/m3). Bulk
sampling of central MRF systems around the time the illnesses
were recognized revealed contamination of several systems with
Mycobacteria immunogenum, a microorganism that has been
found as a contaminant in other machining plants where there
have been outbreaks of HP.(1)
CASE REPORT
In September 2001, a 35-year-old male toolmaker at the facil-
ity developed a non-productive cough, low-grade fever, dyspnea
with minimal exertion, and myalgias. He smoked 1
/
2to 1 pack of
cigarettes per day, but stopped around the time these symptoms
began. He was initially treated with several courses of antibiotics
without improvement. Pulmonary consultation two months after
symptom onset was notable for continued symptoms, a normal
physical examination, and oxygen saturation with brief exertion
(walking) on room air of 81 percent. Spirometry revealed a mod-
erate reduction in forced vital capacity with a normal ratio of
forced expiratory volume in one second to forced vital capac-
ity. Diffusing capacity for carbon monoxide was within normal
limits. A chest radiograph was interpreted as “no acute process.”
The patient was begun on prednisone (60 milligrams per day)
due to concerns that he was possibly experiencing pulmonary
manifestations of an autoimmune disease. Approximately three
weeks after initial pulmonary consultation the patient was taken
off work. A ventilation/perfusion scan was performed and was
interpreted as “low probability to indeterminate”for pulmonary
embolism and the patient was admitted to the hospital for anti-
coagulation (treatment for presumed thromboembolic disease).
During that hospitalization, physical examination was notable
for a few coarse interstitial crackles bilaterally. Computed to-
mography of the chest revealed diffuse interstitial infiltrates
(“ground glass”changes). Anticoagulation treatment was dis-
continued. Open lung biopsy approximately six weeks after
initial pulmonary consultation revealed chronic interstitial pneu-
monia with rare giant cells and vague non-necrotizing gran-
ulomas most consistent with HP. Follow-up examination 21
/
2
months after initial pulmonary consultation revealed resolution
of symptoms, oxygen saturation with brief exertion on room air
of 94 percent, and normal spirometry. The patient was main-
tained on a tapering dose of prednisone until it was discontin-
ued approximately three months after initial pulmonary consul-
tation. He returned to work, approximately five months after
initial pulmonary consultation, to an area of the facility outside
the machining areas where he had formerly worked.
STUDY RATIONALE
Occupational exposure to MRF aerosols causes a variety of
pulmonary symptoms and illnesses (including HP); these health
effects have been summarized previously.(1−4) HP (also called
extrinsic allergic alveolitis) is a granulomatous, interstitial lung
disease that occurs after repeated inhalation and sensitization to
a wide variety of microbial agents (bacteria, fungi, amoebae),
animal proteins, and low-molecular weight chemical antigens.
It is marked by a pneumonitis, which is reversible if exposure to
the antigen is stopped; continued exposure can lead to a chronic
interstitial fibrosis or scarring of the lungs. In general, HP is
marked by nonspecific symptoms. Suggestions for uniform cri-
teria for the diagnosis of HP have been published.(5,6)
HP is thought to involve both humoral and cell-mediated
immune responses.(7,8) Detection of antibody is associated with
exposure to a suspect causative antigen, but does not by itself
confirm the presence of an active disease.(9) With recognition
of the difficulty of performing specific inhalation challenges
to evaluate the potential causative role of a specific antigen,
measurements of the secretion of substances playing roles in
cell-mediated immunity (as a surrogate for clinical illness) have
been described.(8)
One study found that peripheral blood cells from subjects
with pigeon breeder’s disease (a form of hypersensitivity pneu-
monitis) produced increased amounts of a cytokine (migration
inhibitory factor) after in vitro challenge compared with those
from asymptomatic persons.(10) Interleukin 8 (IL-8), tumor
necrosis factor alpha (TNF-α), and interferon gamma (IFN-γ)
have been suggested as playing a role in the pathogenesis of sev-
eral types of HP.(11) We therefore questioned whether measure-
ment of MI-specific cell-mediated immunity as MI-stimulated
secretion of IL-8, TNF-α, or IFN-γby peripheral blood cells
would be associated with HP in this facility. We also questioned
whether such measures would be more informative about the
presence of HP than evaluation of serum anti-MI antibody levels.
METHODS
Medical Record Review and Questionnaire
Facility medical records concerning employees from machin-
ing areas who had presented to the facility medical department
within the previous several months with respiratory symptoms
and/or illnesses were reviewed. Subsequently, a brief question-
naire, which included questions concerning job duties, current
symptoms, and medical history was administered in June 2002
to employees in selected manufacturing departments. Included
among those departments were three machining departments
and one assembly department. Current cigarette smoking sta-
tus was determined by response to the question “Do you now
smoke cigarettes?”The goal of the questionnaire was to iden-
tify groups of employees to take part in the serological survey
described below. The protocol for this HHE, including ques-
tionnaire administration and serologic survey, was approved by
the NIOSH and University of Michigan human subjects review
boards. All participants provided informed consent.
Based on laboratory capabilities for the serological stud-
ies, we planned to recruit approximately 15 participants from
each of four groups (see below). “Symptomatic”employees in-
cluded those employees who reported on the questionnaire in
EVALUATION OF HYPERSENSITIVITY PNEUMONITIS 955
the preceding year: a) chest flu (fever, shivering, cough, tired,
weak, ache all over) or pneumonia, with symptoms related to
work; and/or b) episode of breathing difficulty (such as wheez-
ing, shortness of breath, and/or cough) with symptoms related
to work. “Asymptomatic”employees were defined as employ-
ees who reported neither “a”nor “b”above. If more than 15
potential participants were available from any of the groups, 15
participants were chosen randomly from the total number.
Participants were categorized into one of four groups: Group 1
employees working in machining areas (“exposed”to MRF)
who were suspected or confirmed to have respiratory illness
(possibly including HP) by their physicians; Group 2 employ-
ees from the machining areas who were symptomatic but who
had not presented to the facility medical department; Group 3
employees from the machining areas who were asymptomatic;
and Group 4 employees from the assembly area (“unexposed”
to MRF) who were asymptomatic. HP was defined using a pub-
lished case definition previously used in the evaluation of res-
piratory illness among workers in a large machining plant in
1995–1996 (Table I).(6) We considered all persons with “possi-
ble,”“probable,”or “definite”HP to meet our case definition for
HP.
Cytokine and Antibody Analyses
To evaluate whether M. immunogenum, which was cultured
from the MRF at the facility, may be related to the HP diag-
nosed among some workers at the plant, we performed cytokine
and antibody analyses. A finding of increased cytokine secretion
from workers with HP compared to those from workers without
HP would provide some evidence that the affected workers were
specifically sensitized to M. immunogenum. Our cytokine anal-
yses measured the response of workers’whole blood (Assay 1)
and peripheral blood mononuclear cells (PBMC) (Assays 2 and
3) after in vitro exposure to a M. immunogenum antigen prepa-
ration. Blood was obtained by venipuncture and collected into
TABLE I
Diagnostic criteria for hypersensitivity pneumonitis(6)
Physician diagnosis of hypersensitivity pneumonitis
Onset of at least two pulmonary symptoms (cough, wheeze,
chest tightness, shortness of breath) and one systemic
symptom (fever, weight loss) after July 2001
Recurrence of pulmonary or systemic symptoms after three or
more day avoidance
Restrictive pattern on spirometry not due to obesity
Impairment of pulmonary diffusing capacity less than 80%
predicted
Interstitial or reticulonodular pattern on chest x-ray or
computed tomography
Biopsy evidence of non-caseating granulomas
Definite case of HP: Meeting six or seven criteria.
Probable case of HP: Meeting five criteria.
Possible case of HP: Meeting four criteria.
heparinized tubes. Samples were stored on ice and taken di-
rectly for evaluation at the University of Michigan (Assay 1),
or shipped overnight for evaluation at NIOSH-Morgantown
(Assays2&3).
M. immunogenum (MI) was obtained from the American
Type Culture Collection (ATCC), Manassas, VA (#700506).
This isolate has previously been implicated as being associated
with MRF-related HP.(12) Purified protein derivative from MI
(MI-PPD) was prepared in a fashion similar to that described by
Landi.(13) Briefly, MI was cultured in Middlebrook 7H9 broth
supplemented with OADC for 3 weeks at 37◦C. At 3 weeks,
growth was heavy with a large amount of precipitated material
present in cultures. Cultures were sterilized by autoclaving for
15 minutes. Culture supernatants were obtained by centrifuga-
tion (20,000 ×g, 4◦C, 1 hour). Protein was precipitated from
culture supernatants using saturated ammonium sulfate. Precip-
itate was obtained by centrifugation (20,000 ×g, 4◦C, 1 hr) and
re-dissolved in a minimum volume of phosphate-buffered saline
(PBS). This material was dialyzed against four 20-fold volumes
of PBS at 4◦C over the next 48 hours. The dialysate was re-
moved, clarified by centrifugation (20,000 ×g, 4◦C, 4 hours),
passed through a 0.45 µmfilter, aliquoted, and stored frozen
at −70◦C. The resulting MI-PPD preparation was sterile, had
a protein concentration of 31 mg/ml by Lowry assay, and an
endotoxin concentration of 1.3 ng/ml by limulus amoebocyte
lysate (LAL) assay. The optimal dilution of the MI-PPD prepa-
ration for stimulating IFN-γsecretion by a tuberculin-positive
volunteer’s PBMC using Assay 2 (described below) was 1:20.
Assay 1
The PPD prepared from M. immunogenum (MI-PPD; see
above) antigen preparation was prepared at ten times concen-
tration, and 110 microliter (µl) were added to one ml of whole
blood. Enough stimulant was prepared for at least 40 partici-
pants. The blood and the MI-PPD were combined in the fol-
lowing order. The MI-PPD stock was 31 milligram (mg)/ml,
and stimulation was performed at 0.248 µg/ml. The stock was
diluted 1/100 (990 µl control vehicle [RPMI] +10 µl MI-PPD).
Forty µl of the above dilution was added to 5 ml of RPMI for a
final concentration of 2.48 µg/ml. The lipopolysaccharide (LPS)
antigen was prepared as follows. The stock was 100 µg/ml and
we stimulated at 100 nanogram (ng)/ml (50 µl of stock was
added to 5 ml RPMI for a final concentration of 1000 ng/ml.
After adding blood, the tubes were placed on the rotator in the
incubator (37◦C) for 20 hours. At the end of 20 hours, the tubes
were spun at 2000 ×g for three minutes, and the plasma was
drawn off and frozen at −70◦C. Enzyme-linked immunosor-
bent assay (ELISA) was used to analyze for TNF-α, IFN-γ,
and IL-8.
Assays 2 and 3
PBMC were prepared by centrifugation over sodium dia-
trizoate/Ficoll gradients (Isolymph, Gallard Schlesinger, Carle
Place, NY).
956 D. TROUT ET AL.
Cultures for Assay 2 (MI-PPD stimulated IFN-γsecretion by
PBMC) were performed in 96-well round-bottomed microtiter
plates using RPMI-1640 tissue culture medium supplemented
with 25 mM HEPES, 2 mM L-glutamine, 100 units/ml peni-
cillin, 0.1 mg/ml streptomycin, 5 percent fetal calf serum, and
5 percent human AB serum (CRPMI). 2×105PBMC in a volume
of 200 µl culture medium per well were cultured with CRPMI
alone (negative control), phytohemagglutinin-P (PHA) 5 µg/ml
(positive control), purified protein derivative (PPD) 20 µg/ml
(Accurate Chemical & Scientific Co., Westbury, NY), or MI-
PPD (see above) diluted to a final concentration of 1:10, 1:20,
or 1:100. Cultures were performed at 37◦C and 5 percent carbon
dioxide (CO2) for 5 days, an interval previously documented to
be optimal for determination of PPD-induced IFN-γsecretion
by PBMC from tuberculin positive individuals (data not shown).
After 5 days, conditioned culture supernatants were obtained
and saved frozen at −70◦C until assayed for IFN-γcontent by
ELISA (Interferon Gamma Duoset,R&DSystems, Minneapo-
lis, MN). The data for the 1:20 dilution are presented here; these
data correlated with data for the 1:10 dilution (r =0.84) and less
well with data from the 1:100 dilution (r =0.37).
To perform ELISPOT assays (Assay 3), 96-well PVDF
membrane-bottomed plates (MAIPS4510, Millipore, Bedford,
MA) were coated overnight with 50 µl/well of 10 µg/ml solu-
tion of anti-human IFN-γmonoclonal antibody (clone 1-D1K;
Mabtech Inc., Cincinnati, OH) in carbonate buffer, pH 9.6. After
washing and blocking with CRPMI, cultures were performed
in CRPMI alone (negative control), PHA-P 5 µg/ml (positive
control), purified protein derivative (PPD) 20 µg/ml (Accurate
Chemical & Scientific Co., Westbury, NY), or MI-PPD diluted
toafinal concentration of 1:20. 5 ×104cells were added to cul-
tures performed with PHA-P; all other cultures were performed
with 2 ×105cells. Plates were incubated at 37◦C, 5 percent
CO2overnight. To visualize spots, plates were incubated with
the following reagents with appropriate washing steps between
incubations: biotinylated anti-human IFN-γmonoclonal anti-
body 1 µg/ml (clone 7-B6-1, Mabtech Inc., Cincinnati, OH),
Vectastain AB reagent (Vectastain ABC Elite kit, Vector Labo-
ratories, Burlingame, CA), and Vector VIP peroxidase substrate
(Vector Laboratories). Plates were washed with water, dried, and
spots counted using a dissecting microscope. Data are expressed
as spots per 105PBMC.
Antibody Studies
An antigen extract was prepared from the same isolate of M.
immunogenum used for the MI-PPD preparations. The antigen
preparation, which contains both cell-associated and secreted
products, was made by collecting the bacterial cells by cen-
trifugation and re-suspending in saline as a 10 percent volume/
volume suspension. The bacterial suspensions were sonicated
for one minute using a Branson Model 350 sonifier set at 40 per-
cent output, 50 percent duty cycle pulse. The sonicates were
clarified by centrifugation at 3500 rpm for 20 minutes, and the
supernatant fluid was recovered, stored at −20◦C, and used as
the source of antigen. The protein content of bacterial sonicate
was determined using a modified Lowry method (BioRad) ac-
cording to the manufacturer’s recommendations. The bacterial
sonicates were diluted with carbonate coating buffer (pH 9.6)
for use in the ELISA assay.
A direct ELISA for antibodies to M. immunogenum was
developed using the procedures described by Voller and Bid-
well (Manual of Clinical Immunology, 4th edition, ASM Press,
Washington, DC). ELISA plates were coated with the M. im-
mungenum extract (3 to 5 µg protein/ml) in carbonate coat-
ing buffer overnight at 4◦C, blocked with 1 percent human
serum albumin. The subject sera were initially tested in dupli-
cate at a 1:100 dilution, and antibody binding was detected using
peroxidase-labeled anti-human immunoglobulins (Sigma) and
developed with TMB substrate. Appropriate positive and neg-
ative human serum controls were performed with each plate,
including an antigen blank. Results were reported as optical
density at 450 nm.
Data Analysis
To control for the number of viable cytokine-producing cells,
for Assays 1 and 2 the secretion of cytokine was expressed as a
percentage of cytokine secretion after stimulation from a strong
stimulant (lipopolysaccharide [LPS] [Assay 1] or phytohemag-
glutinin [PHA] [Assay 2]). The equation(14) used to calculate the
“% LPS”(or %PHA) was:
[(cant −ccon)/(cLPS −ccon )] ×100 [1]
where cant =cytokine stimulated by antigen, ccon =cytokine
stimulated by control, and cLPS =cytokine stimulated by LPS (or
in Assay 2, by PHA). For the ELISPOT (Assay 3), the index used
in the evaluation was the difference between antigen-stimulated
result and the result from the negative control. Comparisons
of cytokine concentrations (Assays 1 and 2), ELISPOT results
(Assay 3), and antibody levels were made by case (HP case ver-
sus non-case) and exposure (exposed versus unexposed) status
using the Wilcoxon two-sample test.Apvalue of ≤.05 was used
to indicate statistical significance.
RESULTS
Medical Record Review
As of the start of the NIOSH HHE, medical records for 16
workers were available and reviewed by a physician; all work-
ers were from machining departments. Job titles included oper-
ator (8 persons), job setter (4 persons), toolmaker (2 persons),
and machine repair and janitor (1 each). Using the case defini-
tion outlined above,(6) 7 workers met the definition for possible
(2 workers), probable (2 workers), or definite (3 workers) HP.
Among the 7 meeting a case definition for HP were 2 operators,
2 job setters, 2 toolmakers, and 1 janitor. Among the remaining
9 workers for whom records were reviewed, 7 of the 9 appeared,
based on the physician’s review, to have work-related respiratory
EVALUATION OF HYPERSENSITIVITY PNEUMONITIS 957
or upper respiratory symptoms, and 2 of the 9 appeared to have
respiratory symptoms that were not work-related. Among the
14 workers with work-related respiratory conditions (7 with HP
and 7 with other work-related respiratory conditions), 11 (73%)
participated in our laboratory testing, including 6 of the 7 with
HP and 5 of the 7 categorized as having other work-related res-
piratory disorders.
Questionnaire Administration and Selection of
Participants for Laboratory Testing
The 14 workers mentioned above made up the Group 1 par-
ticipants. The primary goal of the questionnaire survey was to
identify participants in Groups 2–4. Although not used for group
categorization, questionnaires were also completed for all those
available in Group 1. Of the 150 questionnaires distributed to
workers in the departments of interest, 140 (93%) were com-
pleted. Of the 140 persons who completed the questionnaire, 36
(26%) declined to participate any further in the HHE. Among the
remaining 104 workers, 22 from the machining areas were symp-
tomatic (and therefore made up potential members of Group 2),
and 13 were asymptomatic (and therefore made up potential
members of Group 3). Fourteen of the 104 were asymptomatic
workers from the assembly department (and therefore made up
potential members of Group 4). The remainder of the partici-
pants who completed the questionnaire reported some respira-
tory symptoms but did not report work-relatedness and were not
included in any of the four groups. Fifteen of the 22 symptomatic
workers from the machining areas were randomly selected to be
offered the laboratory testing as part of Group 2.
Therefore, 56 workers were initially identified as potential
participants for the laboratory portion of the survey from Groups
1–4. Seven (13%) of the workers on the original list of 56 were
not available (mostly due to vacation) to participate in the in-
formed consent procedures and were therefore dropped from the
survey. During that time, 9 workers who worked in the depart-
ments of concern, but who had not been at work during the ques-
tionnaire administration, presented requesting to be included in
the HHE.
All completed the questionnaire, were grouped according to
their exposure status and symptoms as reported in the question-
naire, and participated in the laboratory testing. Therefore, 58
workers were scheduled to have blood drawn. Fifty-one (88%)
of the 58 completed the blood sampling. Due to timing of ship-
ment of blood samples, 5 of the 51 blood samples could not be
transported to the University of Michigan laboratory. Three of
the 7 workers unavailable at the time of the initial blood draw
provided blood samples when NIOSH representatives returned
to the facility 2 weeks after the primary testing date; those 3
samples were analyzed only by the NIOSH laboratory because
the samples could not be transported to the University of Michi-
gan laboratory. This yielded the following: 46 samples analyzed
by the Michigan laboratory (Assay 1) and 54 samples analyzed
by the NIOSH laboratory (Assays 2 and 3).
TABLE II
Description of casesAand non-cases participating in health
hazard evaluation
# (%) # (%)
N Mean age Male Current smoker
Non-case 48 40 30 (63) 11 (23)
HP case 6 44 6 (100) 0 (0)
AMeeting case definition for hypersensitivity pneumonitis.(6)
Cytokine/Antibody Analyses
Cytokine and antibody results are presented by case (HP)
and exposure status; descriptive summaries of those groups are
presented in Tables II and III.
Assay 1 (MI-PPD Stimulated Secretion of Cytokines
in Whole Blood Cultures)
The limit of detection (LOD) for each of the three cytokines
after stimulation with M. imunogenum,M. tuberculosis, and
RPMI was 60 picograms per ml (pg/ml); the LOD for the cy-
tokines after stimulation with LPS was 300 mg/ml. IFN-γwas
detected above the LOD (63 and 77 pg/ml) in only 2 of 46
samples after stimulation with M. immunogenum—one sample
from a worker with HP, one from a worker in Group 4 (unex-
posed, asymptomatic). TNF-αwas detected above the LOD (89
and 93 pg/ml) in only two samples after stimulation with M.
immunogenum—one sample from a worker with HP, one from
a worker who had respiratory symptoms that were not work-
related (Group 1).
Summary data for IL-8 secretion are presented in Tables IV
(HP cases versus non-cases) and V (MRF-exposed versus those
unexposed to MRF). There were no statistically significant dif-
ferences in the concentrations of IL-8 among the persons with
HP (compared to those without HP), or among the persons ex-
posed to MRF (compared to those unexposed); the median val-
ues were greater among the persons without HP and among the
persons unexposed to MRF. There was no statistically significant
difference in the index measure for IL-8 secretion (IL-8 secre-
tion expressed as a percent of LPS secretion) between persons
with HP and without. However, this index of IL-8 secretion was
statistically significantly greater among those exposed to MRF
(median 0.33) compared to the unexposed (median 0).
TABLE III
Description of MRF-exposed and MRF-unexposed
participating in health hazard evaluation
# (%) # (%)
N Age (mean) Male Current smoker
Unexposed 11 41 5 (45) 2 (18)
Exposed 43 40 31 (72) 9 (21)
958 D. TROUT ET AL.
TABLE IV
Cytokine/antibody results by case status
Cytokine/antibody Group Mean Median p valueA
Assay 1 Non-case 4607 2923 0.3
IL-8BHP case 4654 1398
Assay 1 Non-case 2.8 0 0.9
IL-8 %LPSCHP case 3.6 0
Assay 2 Non-case 102.0 0 0.2
IFN-γDHP case 143.3 141
Assay 2 Non-case 6.1 0 0.3
IFN-γ%PHAEHP case 7.1 4.3
Assay 3 Non-case 3.8 1.0 0.4
IFN-γELISPOTFHP case 2.8 1.6
Assay 3 Non-case 1.0 0.5 0.7
IFN-γELISPOT HP case 2.4 1.0
(index)G
Antibody to Non-case 0.098 0.074 0.001
M. immunogenumHHP case 0.29 0.21
AWilcoxon two-sample test; two-sided.
BInterleukin-8 (pg/ml).
CInterleukin-8 index (% of lipopolysaccharide).
DInterferon gamma (pg/ml).
EInterferon gamma index (% of phytohemagglutinin).
FInterferon gamma ELISPOT-producing cells per 105PBMC.
GELISPOT index is differencebetween antigen-stimulated result and
that from control.
HOptical density at 450 nm.
Assays 2 (MI-PPD Stimulated Secretion of IFN-γ
into Culture Supernatants by PBMC) and 3 (MI-PPD
Stimulated IFN-γELISPOT)
Summary data for IFN-γsecretion are also presented in
Tables IV and V. There were no statistically significant differ-
ences in the concentrations of IFN-γ(both absolute secretion
and index value [IFN-γ%PHA]) as determined by Assay 2
among the persons with HP (compared to those without HP),
or among the persons exposed to MRF (compared to those un-
exposed). The median values for both measures were greater
among the persons with HP and among the persons exposed to
MRF.
There were no statistically significant differences in the ab-
solute and index values for IFN-γsecretion as determined by
Assay 3 among the persons with HP (compared to those without
HP) and among exposed persons (compared to those unexposed
to MRF). The median values for both absolute and index mea-
sures were higher among HP cases (compared to non-cases),
while the median values for the absolute and index measures
among exposed persons (compared to those unexposed to MRF)
did not show the same consistent trend.
Antibody Assays
Antibody levels against the M. immunogenum antigen were
greater among both persons with HP (compared to those without
HP) (Table IV) and persons exposed to MRF compared to those
unexposed (Table V). Because antibody production to inhaled
antigens has been shown to be inhibited by cigarette smoking,(11)
analyses were performed in which current cigarette smokers
were excluded. There was no substantial change in the increased
level of antibody among HP cases and among MRF-exposed
persons with the analyses performed among non-smokers only.
DISCUSSION AND CONCLUSIONS
The primary goal of our evaluation was to learn whether mea-
sures of cell-mediated immunity to M. immunogenum, which had
been identified as an MRF contaminant in the areas of the facil-
ity where workers had been diagnosed with HP, were associated
with the presence of HP among these workers. Except for the
IL-8 percent LPS assay, our cytokine studies revealed no statisti-
cally significant relationships between secretion of the cytokines
tested and either the presence of HP or exposure to MRF. Of the
cytokine assays used in this HHE, Assay 2 (MI-PPD stimulated
IFN-γsecretion into culture supernatants) showed increased
secretion (for both the absolute and index measures of IFN-γ)
among persons with HP, although the differences were not sta-
tistically significant. A study with a larger number of affected
workers would improve our ability to detect small differences in
IFN-γsecretion from PBMCs of persons with HP. However, due
to overlap between groups with and without HP, it is doubtful
TABLE V
Cytokine/antibody results by exposure status
Cytokine/antibody Group Mean Median p valueA
Assay 1 Unexposed 3185 3131 0.9
IL-8BExposed 4601 2364
Assay 1 Unexposed 0 0 0.02
IL-8 %LPSCExposed 3.3 0.33
Assay 2 Unexposed 34 0 0.2
IFN-γDExposed 124 12
Assay 2 Unexposed 2.0 0 0.2
IFN-γ%PHAEExposed 7.4 0.38
Assay 3 Unexposed 15.7 1.1 0.8
IFN-γELISPOTFExposed 1.4 0.88
Assay 3 Unexposed 0.72 0.25 0.3
IFN-γELISPOT Exposed 1.24 0.5
(index)G
Antibody to Unexposed 0.057 0.057 0.03
M. immunogenumHExposed 0.16 0.10
AWilcoxon two-sample test; two-sided.
BInterleukin-8 (pg/ml).
CInterleukin-8 index (% of lipopolysaccharide).
DInterferon gamma (pg/ml).
EInterferon gamma index (% of phytohemagglutinin).
FInterferon gamma ELISPOT-producing cells per 105PBMC.
GELISPOT index is difference between antigen-stimulated result and
that from control.
HOptical density at 450 nm.
EVALUATION OF HYPERSENSITIVITY PNEUMONITIS 959
that this type of testing will be a useful tool in medical screening
for HP at other machining facilities.
The results of our testing for antibodies against Mycobac-
teria were consistent with antibody testing against causative
antigens in the evaluation of other types of HP. For example,
elevated levels of antibody against purified avian antigens have
been associated with increased incidence of pigeon breeders dis-
ease (a form of HP),(15) and increasing exposure to pigeons has
been associated with increase in antibody response to pigeon
antigens.(16)
However, as with other antibody testing, our antibody testing
was non-specific in that we found elevated levels of antibody in
exposed persons without regard to the presence of HP. MRF as
they are used in production facilities are complex mixtures of
multiple potential antigens, both microbial and chemical. There-
fore, based on antibody testing alone, we cannot rule out that
some other antigen(s) may be present along with the M. immuno-
genum and may be involved in the pathogenesis of HP.
There were several limitations to our evaluation. Although
we based our grouping of workers on the most up-to-date clin-
ical information that was available to us, our records may have
been incomplete so it is possible that participants may have been
misclassified in terms of case/non-case status. Likewise, it is
possible that our departmental-level determination of exposure
status may have led to some misclassification of participants by
exposure at the individual level. Finally, the fact that we were
using PBMCs may be an important limitation of our labora-
tory work. Because the target organ of inhaled antigens thought
responsible for HP is the lung, recent work related to the patho-
genesis of HP has involved the cellular immune response of cells
from the lung (obtained via bronchoalveolar lavage [BAL]).(17)
Cells obtained from BAL may be more accurate indicators of
cell-mediated immune responses occurring in the lung; how-
ever, these types of cells are not readily available from workers
involved in evaluations such as this HHE.
A number of actions were taken by the facility in response to
the illnesses among employees. The primary actions taken in-
volved improvements in maintenance, cleaning, and formulation
of the MRF in the machining areas. Based on review of indus-
trial hygiene records at the facility, the engineering controls in
place appeared to be effectively maintaining the workers’ex-
posure to MRF aerosol at concentrations substantially less than
the NIOSH-recommended exposure limit. However, there was
concern that some workers at the facility may be exposed to
greater concentrations of MRF aerosol for short periods of time
due to job tasks which entail entering machine tool enclosures.
Four of the seven persons meeting the case definition for HP
were job setters or tool makers, workers whose jobs involve
routine (intermittent) entrance into the machine tool enclosures.
Based on the above, and incomplete knowledge related to the
importance of short-term “peak”exposures to MRF aerosol rel-
ative to ambient (or average) exposures, reduction in such peak
exposures may be warranted. Primary methods to accomplish
this involve changes in engineering controls and work practices.
Such changes should be part of a comprehensive safety and
health plan recommended by NIOSH to minimize health effects
related to occupational exposure to MRF.(1)
ACKNOWLEDGMENTS
The authors thank Barbara MacKenzie, Debbie Sammons,
Toni A. Bledsoe, Janet Simpson, and Erica Janotka for their
assistance in conducting this evaluation.
REFERENCES
1. Kreiss, K.; Cox-Ganser, J: Metalworking Fluid-Associated Hyper-
sensitivity Pneumonitis: A Workshop Summary. Am J Indus Med
32:423–432 (1997).
2. NIOSH: Criteria for a Recommended Standard: Occupational Ex-
posure to Metalworking Fluids, DHHS (NIOSH) Publication Num-
ber 98-102. National Institute for Occupational Safety and Health,
Cincinnati, OH (1998).
3. CDC: Biopsy-Confirmed Hypersensitivity Pneumonitis in Auto-
mobile Production Workers Exposed to Metalworking Fluids—
Michigan. MMWR 45(28):606–610 (1996).
4. Zacharisen, M.C.; Kadambi, A.R.; Schlueter, D.P.; et al.: The
Spectrum of Respiratory Disease Associated with Exposure to
Metal Working Fluids. J Occup Environ Med 40(7):640–647
(1998).
5. Terho, E.O.: Diagnostic Criteria for Farmer’s Lung Disease. Amer
J Indus Med 10:329 (1986).
6. Fox, J.; Anderson, H.; Moen, T.; et al: Metal Working Fluid-
Associated Hypersensitivity Pneumonitis: An Outbreak Inves-
tigation and Case-Control Study. Am J Indus Med 35:58–67
(1999).
7. Newman, L.; Storey, E.; Kreiss, K.: Immunologic Evaluation of
Occupational Lung Disease. In: Occupational Medicine: State of
the Art Reviews. L. Rosenstock, ed., 2 (2):345–372. Hanley and
Belfus, Inc., Philadelphia (1987).
8. Schatz, M.; Patterson, R.; Fink, J.: Immunopathogenesis of Hy-
persensitivity Pneumonitis. J Allergy Clin Immunol 60(1):27–37
(1977).
9. Richerson, H.B.; Bernstein, I.L.; Fink, J.N.; et al: Guidelines for
the Clinical Evaluation of Hypersensitivity Pneumonitis. J Allergy
Clin Immunol 84 (5, part 2): 839–844 (1989).
10. Fink, J.L.; Moore, V.L.; Barboriak, J.J.: Cell-Mediated Hypersensi-
tivity in Pigeon Breeders. Int Arch Allergy Appl Immunol 49:831–
836 (1975).
11. McSharry, C.; Anderson, K.; Bourke, S.J.; Boyd, G.: Takes Your
Breath Away—The Immunology of Allergic Alveolitis. Clin Exp
Immunol 128(1):3–9 (2002).
12. Moore, J.S.; Christensen, M.; Wilson, R.W.; et al.: Mycobacterial
Contamination of Metalworking Fluids: Involvement of a Possible
New Taxon of Rapidly Growing Mycobacteria. Am Indus Hyg
Assoc J 61(2):205–213 (2000).
13. Landi, S.: Production and Standardization of Tuberculin. In: The
Mycobacteria—A Sourcebook, G.P. Kubica; L.G. Wayne, eds.,
Part A. pp. 505–535. Marcel Dekker, Inc., New York (1984).
14. Mazurek, G.H.; LoBue, P.A.; Daley, C.L.; et al.: Comparison of
a Whole-Blood Interferon-γAssay with Tuberculin Skin Testing
for Detecting Latent Mycobacterium tuberculosis Infection. JAMA
286(14):1740–1747 (2001).
960 D. TROUT ET AL.
15. McSharry, C.; Banham, S.W.; Lynch, P.P.; Boyd, G.: Antibody
Measurement in Extrinsic Allergic Alveolitis. Eur J Respir Dis
65(4):259–265 (1984).
16. Banham, S.W.; McSharry, C.; Lynch, P.P.; Boyd. G.: Relationships
Between Avian Exposure, Humoral Immune Response, and Pi-
geon Breeders’Disease Among Scottish Pigeon Fanciers. Thorax
41(4):274–278 (1986).
17. Schuyler, M.; Gott, K.; Cherne, A.: Mediators of Hy-
persensitivity Pneumonitis. J Lab Clin Med 136(1):29–38
(2000).