Dissertation Highlights

DEPOSITION OF MMMF (CERAMIC FIBERS)
IN THE LUNGS OF GUINEA PIGS

INTRODUCTION AND STATEMENT OF THE PROBLEM

INTRODUCTION

Asbestos is one substance which has been clearly demonstrated to
exert adverse health effects on humans. In addition to studies using
animal models in which the resultant diseases have been produced
experimentally, epidemiological studies have shown a causal
relationship between these diseases and asbestos exposure in workers.

The diseases associated with asbestos exposure are namely asbestosis,
lung cancer, and mesothelioma (cancer of the mesothelial lining of the
pleura and peritoneum). The probability of occurrence of these
diseases increases as exposure is extended implying that a certain
amount of fiber burden remains in the lung for life after exposure
(NIOSH, 1976; McConnell et al., 1982).

Asbestos fibers have hemolytic activity and cytotoxic potency to
macrophages in in vitro cultures. In vivo effects also clearly
indicate that all types of asbestos are active in animals and man
(Schepers, 1955; Wagner and Berry, 1969).

Because of its serious health effects, many different materials are
now used as substitutes for asbestos. Man-made mineral fibers (MMMF)
also known as man-made vitreous fibers (MMVF), such as glass fibers,
mineral wool (slag and rock wool), and ceramic fibers are examples.

Occupational exposure to fibrous glass alone has been estimated to
be about 200,000 workers (NIOSH, 1977) and is projected to rise as
production of MMMF is expected to increase sharply in the future
owing to the energy crisis and because of the need for substitute for
asbestos (Hammad and Esmen, 1982).

In the past, when MMMF gained entry through inhalation, they were
regarded as inert dust. Animal experiments supported this.
Extensive literature has been written on this subject and the authors
agreed that MMMF, regardless of chemical, composition did not produce
neoplasia and/or malignant abnormalities. However, occasional patches
of pneumonia and other complications of endemic chronic bronchitis
were present (Wright, 1968; Pott et al., 1976; Gross, 1976). In
contrasts to the above Davis (1982), in his work on the effects of
ceramic fibers on rats, reported both pulmonary fibrosis and neoplasia
following inhalation of the dust. In addition, very recent animal and
epidemiologic studies suggest a low level of fibrogenic and
carcinogenic potential for MMMF.

Because of their physical similarities in size and shape to asbestos
fibers and dust, and since the health effects of asbestos depend on
physical properties, it is important to prove MMMF safe rather than
accept them as safe until proven otherwise.

STATEMENT OF THE PROBLEM

The purpose of this investigation was to study the pulmonary
deposition of ceramic fibers in guinea pig lungs.

SIGNIFICANCE OF THE STUDY

1. Effects of fibers are related to their size.

2. Investigating pulmonary deposition of fibers provides information
    about dimensions of the fibers retained in lung tissue.

3. Although MMMF were regarded as inert dust in the past, recent
    animal and epidemiologic studies suggest a low level of fibrogenic
    and carcinogenic potential for MMMF.

4. Fiber glass alone has been used in more than 30,000 product
    applications (Powell, 1982). Thus occupational and environmental
    exposure to man-made mineral fibers is widespread and is projected
    to rise as production of MMMF increases in the future owing to the
    energy crisis and because of the need for a substitute for asbestos.

5. Studying the deposition of ceramic fibers in the guinea pig lung
    may furnish information about potential deposition, clearance and
    retention of these fibers in the human lung.

6. The results may contribute significantly to the literature and
    future research in this area.

OBJECTIVES

The objectives of this study were to:

1. Generate a ceramic fiber dust cloud of a wide range of fiber
    diameters and lengths suitable for inhalation.

2. Expose guinea pigs to this dust cloud by nose inhalation only.

3. Determine the deposition of various fiber diameter and length
categories in the lobes of the guinea pig lung.

4. Determine the deposition of various fiber size categories in the
    upper respiratory tract of the guinea pig.

5. Determine pulmonary and upper respiratory tract clearance rates for
    various fiber size categories.

6. Compare results with other previously published research.

7. Project potential tissue and health effects of such deposition.

MATERIALS AND METHODS

RESEARCH PLAN AND SCHEDULE

This study consisted of three phases extending over a period of
approximately one and a half years.

Phase one - 5 months period: During this period the literature was
reviewed and the experimental procedures evaluated in terms of
proposed materials and methods.

Phase two - 6 months period: This period covered the actual
experiments. This included preparation of fiber, setting up the
exposure chamber, exposing the animals to the generated dust cloud,
sacrificing the animals and preparing their lungs for fiber recovery,
preparation of the slides, and finally counting the fiber contents.

Phase three - 4 months period: In this, the data was analyzed, tested
for significance, compared with previously published data, and
conclusions were made.

CHOICE OF FIBER

With the exception of very few studies, all the literature reviewed
on MMMF has clearly showed a worldwide interest in glass fiber and
mineral wool owing to long-term existence and use. Ceramic fiber,
on the other hand, was developed in the mid-1940's, and today it
constitutes a significant portion of the world's output of man-made
inorganic fibers. The ceramic material has unusual combinations of
properties, thus making it highly desirable for a wide range of
applications. The fiber has an extraordinary durability; it is not
subject to oxidation up to about 1200 °C and decomposes at 1900 °C
(Hlavac, 1983; Birchall et al., 1985). This property of ceramic fiber
makes it resistant to changes that may occur due to digestion or
ashing when extracted from the lung tissue, thus size and count
distributions would not be affected (Hammad, personal communication).
Inhalation studies on ceramic fibers are relatively few when compared
to other MMMF. Therefore, owing to the above reasons, the fibrous
material chosen for this experiment was ceramic fibers.

Because the aerodynamic properties of fibers depend on the density

of the parent material, the density of the fibers used in the study was
determined by a 10 ml pycnometer. The average density of six
determinations was found to be 2.34 ± 0.113 gm/cm³.

CHOICE OF EXPERIMENTAL ANIMAL

The consideration of the animal model for experimental studies is a
function of several parameters. If an animal model is known to
produce results similar to those produced in man, then this model can
be used with more confidence to examine similar or other possible
effects. Although the rat was reported to be the best model to use
for studies on the bioeffects of asbestos fibers (Davis, 1979), the
guinea pig has an advantage over the rat. Guinea pigs have been found
to coat asbestos fibers and to produce asbestos bodies, a reaction
found in man when exposed to asbestos and to other mineral fibers
(Davis, 1979; Hammad, personal communication). In addition, the
guinea pig's respiratory system differs from that of the rat's, which
makes deposition patterns quite different, and therefore, clearance
rates and the bioeffects would be expected to be different. Thus
being able to pinpoint these differences among various species may
help in clarifying some unanswered questions such as tissue reaction
and health effects. In light of the above facts, the guinea pig was
the animal model of choice in this experiment. Another reason for
selecting the guinea pig was the availability of a recent study on the
deposition of ceramic fibers in the rat's lung (Rowhani, 1982) which
could furnish a base line for comparisons.

DUST GENERATION

A fibrous dust generator similar to that described by Timbrell
(Timbrell et al., 1968), with some modification, was used to produce
the fibrous dust cloud of ceramic fibers.

The dust generator consisted of a grinder made of two separable main
parts; the lower part, a shell used as a feeder, and the upper part,
specifically the dispersing chamber. The feeder gradually delivered
the dust into the dispersing chamber, which contained an electrically
driven blade rotor. Whenever the rotating blade hit the fed dust
plug, it shaved off small amounts in the form of single fibers or
minute flocks. By introducing a compressed air stream, the generated
single fibers were moved into a vertical elutriator, installed above
the generator and before the inhalation chamber. The elutriator
removed very large dust particles and agglomerated fibers, and only
allowed the fine part of the dust cloud to enter the inhalation
chamber. From the elutriator, the dust cloud passed to the outlet,
where it was mixed with dilution air to enter the exposure chamber
(Hammad, 1982; Hammad et al., 1982).

The use of the dust depends on the physical characteristics at the
time of the experiment. Because the material consisted of very long
fibers and aggregates, some pretreatment was necessary. Fibers were
broken, ground and milled to satisfactory physical dimensions, that
is, length and diameter with an aspect ratio of at least 3:1.
Pretreatment would only change length dimensions and had no effect on
diameter. During the process of refining the fibers, water was added
to wet the material to prevent dispersion of fibers into the
environment. Following this preliminary preparation, the fibers were
separated from any possible shot content by liquid sedimentation.
The remaining slurry was then dried in a low-temperature oven until it
was ready to be used. The resulting ceramic dust was then mixed with
acetone to make a slurry from which dust plugs were packed and
prepared for the generator.

The desired fiber dimensions were in the range of submicrometer to
about 3 μm in diameter and from 10 μm to approximately 100 μm in
length. This decision was based on the fact that the most serious
health effects occur with fibers of these dimensions (Stanton and
Wrench, 1972; Timbrell, 1982). Another reason for choosing this size
was the nature of this study, which was the deposition of the fibers
in the alveolar tissue. Fibers larger than the stated dimensions did
not contribute to the study due to the probability of them being
retained in the upper respiratory tract (Timbrell, 1965).

Reasons for choosing this type of dust generators were several: The
generator has been in use since 1968 and still attracts researchers
due to its good performance. It was simple to use and reliable.
Another reason for selecting it was that this generator was available
at the institution where this experiment was carried out.

DESCRIPTION OF THE EXPOSURE CHAMBER

The animals were exposed to dust in an inhalation chamber designed
by Hammad (1982). The exposure chamber was made of a plexiglass
rectangular chamber (10.5x10.5x30 in). Aerosol from the dust
generator, after passing through the vertical elutriator, together
with filtered dilution air, was fed into the top of the chamber.
The chamber was designed to hold the animals in plexiglass cylinders.
Each cylinder was fitted at one end with a machined aluminum
nosepiece. The guinea pigs were held in the tubes (cylinders) by an
adjustable plexiglass circular piece fitted with a rubber "O" ring.
The chamber had 24 holes, of which 21 ports were used for exposing the
animals. The chamber also had 4 separate ports for air-dust sampling.
Air from the exposure chamber was exhausted to a clean-up system
consisting of two in line high-efficiency particulate filters. The
exposure system was completely inclosed within a 1.5 m³ chamber.

MEASUREMENT OF THE DUST CONCENTRATION
IN THE CHAMBER

Dust concentrations in the chamber was determined by sampling the air
with Millipore membrane filters. A cellulose ester membrane filter
37 mm in diameter with pore size 0.8 μm was used. The pump was
adjusted to sample 1.00 liters per minute. The pump was calibrated by
a soap bubble meter and the flow rate of the sample was maintained by
critical orifices. The volume of the air sampled was adjusted so that
an optimum concentration of dust was collected on the filter. The
sample should not be too dense, since samples in which particles
overlap must be rejected as uncountable. However, a large enough
volume of air was sampled in order to obtain sufficient numbers of
fibers in each field. It was suggested that there be no more than
10-20 fibers per field. Experience indicated that more than 150
fibers per field including non-fibrous background would interfere with
counting the sample. Four samples were collected during the entire
exposure period. The results were averaged to obtain the fiber
concentration. All sampling filters were maintained in an open-faced
fashion to prevent dust accumulation in the center of the filter and
clogging caused by inertial effects of large fibers. Collected
samples were ashed and then suspended in isopropyl alcohol to produce
a homogeneous mixture. An aliquot of this mixture was then filtered
on a 0.22 μm pore membrane filter and counted by phase-contrast microscopy.

ANIMAL SAMPLE SIZE

Three groups (7 guinea pigs each) were determined to be used for this
study. All three groups were exposed at the same time. The reason
for this selection was based on a study by Hammad et al. (1982). It
was reported that the highest observed standard deviation (S.D.),
median coefficient of variation (C.V.) and mean (X) for fiber
retention in 5 rats' lungs were 4.76, 0.76 and 10 respectively. The
observed data came from 5 different categories based on diameters and
lengths of the fibers.

Assuming that the mean (μ) and the median coefficient of variation
(C.V.) in the overall study were 10 and 0.76 respectively, then the
standard deviation (s) would become:

σ/μ = C.V.
σ = (μ)(C.V.)
hence, σ = (10)(0.76) = 7.60

At a 90% level of confidence, Z (unit normal distribution) = 1.645.
The next step was to select the width of the interval (w). Assuming
an interval width (w) of 5 units, (1/2 the mean), the experiment was
designed so that we would be 90% sure that the observed mean (X) and
the actual mean (μ) would differ by 5 units or less, i.e. | μ-X | ≤ 5.

Therefore, for a standard deviation of 7.6, confidence level of 0.90
and w = 5, the minimum size of the animal sample (n) was calculated by
the following equation (Daniel, 1983).

w = Zσ/(n)^1/2which, when solved for n, gave
n = Z²σ²/w²and therefore,
n = (1.645)²(7.60)²/(5)² = 6.25

Rounded up to the next largest whole number, n became 7. Because the
calculated sample size was 7 animals per each group, 21 guinea pigs
were exposed in this experiment. However, two of the twenty-one
guinea pigs died during the exposure, leaving 19 animals available for
the experiment. Because of their odd number the animals were divided
into 6, 6, and 7 for groups I, II, and III respectively.

Further more, in the digestion process of the lungs, two lobes, the
accessory right lobe and the middle left lobe, of animals #11 and #12
of group II were inadvertently destroyed.

ESTIMATION OF THE ANIMAL'S MINUTE VOLUME

The minute volume for guinea pigs have been reported differently by
different authors and researchers. Crosfill and Widdicombe (1961)
reported the minute volume of the guinea pig to be 0.13 l/min, with
values ranging from 0.08 to 0.19 l/min. These values were measured
while the animal was anaesthetized. Guyton (1947), on the other hand,
reported values ranging between 0.10 and 0.38 l/min with an average
observation of about 0.16 l/min. Applying the formula described by
Guyton, the minute volume, for guinea pigs of body weight ranging
from 0.274 to 0.941 kg with an average of 0.466 kg, would be about
210 ml/min (Guyton, 1947). Mauderly et al. (1979) and

Silbaugh et al. (1981) also reported similar values of 0.171

and 0.25 l/minute respectively.

Based on the above reported values, estimating a minute volume for
the guinea pig of 250 ml/min seemed reasonable. However, the minute
volume for each animal was determined and used to calculate deposition
in that particular animal. Minute volumes were estimated using
Guyton's equation (Guyton, 1947).

Minute volume for any individual lobe of a lung was also estimated as
percentage of the animal's M.V. in the same ratio of lobe weight to
total body weight (Hammad and Rowhani, 1984).

ESTIMATION OF THE EXPOSURE DUST CONCENTRATION

Hammad (1982), in an inhalation study on rats exposed to ceramic
fibers, reported that the mean airborne fiber concentration was 303
with a standard deviation of 17 fibers/cc of air. Rowhani (1982)
reported that the dust concentration in rats was 709 fibers/cc of air.
For this study, the exposure dust concentration was estimated to be
about 300 fibers/cc of air. This number was reached by the following
assumptions and calculations.

Assumptions:

1. about 5 fibers would be counted per field;
2. the area of the field was 1.12225x10^-3 mm²;
3. the actual collection diameter of the 47 mm diameter membrane
    filter, used for the recovered fibers from the animal lung, was
    35 mm, thus making its effective collection area 962.11 mm²;
4. the actual collection diameter of the 37 mm diameter membrane
    filter, used for the dust sampling in the chamber, was 35 mm,
    thus making its effective collection area 962.11 mm².
    Therefore, the number of fields per filter would be 8.57x10⁵.
    Hence, the number of fibers estimated per filter would be
    approximately 4.285x10⁶.

Fibers per filter were determined by multiplying the concentration
(fiber/cc) by duration (total minutes of exposure) by minute volume
(ml/min) by percentage deposition in the alveolar tissue. That is,

fibers/filter = (C)(duration)(minute volume)(% deposition)

For an average minute volume of 250 ml/minute for guinea pigs and a
percentage deposition in the alveolar tissue of about 10%-15%
(estimated from other studies; Raab et al., 1977 and Rowhani, 1982), the
concentration in the exposure chamber, for 6-hour exposure, would be
then expected to be approximately 317± fibers/ml of air.

Considering the air flow rates entering the exposure chamber as follows:

1. generation air: 10 liters/min, that took the fibers through
    the elutriator and into the chamber;
2. dilution air: 40 liters/min, that diluted the dust cloud
    before entering the chamber;

The dust concentration generated by the dispenser and carried by the
generation air would be then approximately 1500 fiber/ml of air.

Sampling time per filter was then determined by equally dividing the
entire time of exposure by 4 (the number of air samples), thus
yielding 90 minutes for each.

DESCRIPTION OF THE EXPOSURE TECHNIQUE

All animals were exposed to the dust cloud by nose only. Exposure was
as described in section "D" above. The animals were exposed for a
total of six continuous hours. Following exposure, the animals were
divided and scheduled to be sacrificed.

SURVIVAL TIME

Morgan et al. (1977) reported that there was a rapid fall/clearance
(t_1/2< 1 day) in the lung fiber contents, assuming that this decline
represented mucociliary clearance of the airways. Hammad et al.(1982),
on the other hand, showed that the alveolar clearance was a slow process
(t_1/2 = 40 days). This clearly showed that clearance in the lungs occurs
in two compartments; the ciliated and the non-ciliated airways.

To be able to account for these two types of clearances in this
experiment, group (I) was sacrificed immediately after exposure; group
(II) 24 hours later; and group (III) 5 days after exposure.

ANIMAL SACRIFICE

Animals were anaesthetized by sodium pentobarbital (50 mg/Kg Nembutal).
In order to reduce organic material and mass that could interfere
with microscopic examination of fibers recovered from the lungs, the
blood content of the lungs was removed by perfusion. The abdomen was
opened, a catheter inserted into the right ventricle and the
descending aorta transacted. A saline-heparin solution was infused
for three minutes at 15 cm water pressure to clear the lungs. The
lungs were further infused with fixative solution (Karnovesky's) for
3 minutes. The trachea and the lobes of the lungs were then carefully
separated and placed in preservative tubes. The nose was then washed
3 times with 10 ml of saline and the collected wash fluid was added to
the trachea's tube. All samples were then stored until digestion and
preparation for microscopic counting.

PREPARATION OF THE LUNGS FOR FIBER COUNTING

Tracheae and lung lobes were all digested in the same manner. The
lobes of each animal were weighted prior to digestion. All digestion
procedures for any sample were carried out in one tube, so that the
possibility of losing fibers was minimized. After placing each lung
tissue in a separate tube, Clorox bleach (sodium hypochlorite) was
added. All tubes were tightly closed and left overnight at room
temperature. If some of the tissue was not digested completely, more
Clorox was added and the tube was agitated gently on a shaker until
all the tissue was completely digested.

The resulting solution was then washed with distilled water from the
tube onto a 47 mm membrane filter with a pore size of 0.22 μm. The
tube was rinsed with isopropanol and poured over the filter. The
isopropanol was used to dissolve colloidal particles in the solution
and thus yielded a better distribution of fibers on the filter.
After this, the tube was washed with xylene and transferred to the
filter. This was necessary to remove any fats and other organic
soluble tissue materials from the filter without affecting the fibers.
Afterwards the filters were dried for about 1 day at room temperature
and made ready for counting.

COUNTING OF RECOVERED FIBERS

A fiber is defined as a particle whose length is at least three times
greater than its diameter. A wedge of the dried prepared filter
(section "L") was cut with an arc of about 1 cm and placed on the edge
of a clean glass slide. A drop of mounting medium was placed on the
center of the slide with an applicator and smeared with a tooth pick
into the shape and size similar to that of the wedge. With forceps,
the cut wedge was then repositioned to the center of the slide and
placed on the mounting medium with the dust-side up. Care was taken
not to wash off the fibers by excess amount of the medium. A cover
slip No. 1½ was then placed on top of the wedge. A drop of immersion
oil was applied. All slides were counted with phase-contrast, 10X
eyepiece (fitted with a calibrated Porton graticule) and 100X
objective (1.30 N.A.) for a total magnification of 1000X.

The Porton graticule was calibrated against a stage micrometer. The
diameter (D) of the circles located above the large rectangle of the
graticule was determined by applying the following equation:

D = L (2^N)^1/2where L = the length of the rectangle (measured by the stage
micrometer)/200, and N is the number of that circle.

Microscopic fields were then selected at random and each examined from
left to right, right to left and top to bottom. Slides with uneven
distribution containing more than 25 fibers per field were disregarded
and new prepared slides were counted instead.

Fiber diameters were measured between circles number 1 and 8, while
fiber lengths were measured between circle number 4 up to and
including circle number 16. Mid size of the interval between any two
circles was used to calculate the median length and the median
diameter of the fiber distribution. Fibers which were inside the
reticle or entered it from any side were counted and sized. Fibers
with any portion extending outside the reticle were counted as half fibers.

Number of fields and fibers counted per each sample (air or animals)
was a function of lobe size and size of filter used. As a result, an
average of 800 fibers were counted in 200 fields for each air sample
and 800-1000 fibers in 250-450 fields per each animal sample.

DEFINITION OF TERMS

Nineteen male Hartly albino guinea pigs were utilized in this study.
Animals were divided into groups I, II, and III with 6, 6, and 7
animals respectively. Animals were assigned numbers from 1 to 19.
The respiratory tract of the animal was divided into two compartments;
the upper respiratory tract (URT) for trachea and nose, and the lower
respiratory tract (LRT) for the lungs. Both (URT) and (LRT) were
referred to as the entire respiratory tract (ERT).

Further division of the lungs was as the following.

(UR) upper right lobe, (right cranial & middle lobes);
(AR) accessory right lobe, (right accessory lobe);
(LR) lower right lobe, (right caudal lobe);
(UL) upper left lobe, (left cranial lobe);
(ML) middle left lobe, (left middle lobe);
(LL) lower left lobe, (left caudal lobe).

RESULTS

BODY WEIGHT EFFECT

All animals showed little if any effects of the exposure on their

body weight. Mean body weight for group I did not change at all

during the six hours exposure. However, groups II and III

showed a slight reduction. The mean body weight for both groups,

at time of sacrifice, was respectively 1.25% & 2.21% less than

that at time of exposure.

LUNG WEIGHT

Correlation between body weights and lung weights indicated that

the two parameters have a strong relationship and that the weight

of the lungs is highly dependent on the animals' body weight.

MINUTE VOLUME (M.V.)

Minute volume for each guinea pig was estimated by Guyton's

equation which relates the M.V. to the body weight.

Minute volume for each animal was estimated at time of exposure.

Minute volumes for lobes were also estimated.

Minute volumes used to calculate retention of fibers in the lobes of
an animal were those M.V.s estimated for that animal. Mean minute
volumes calculated for the entire group were not used in calculating
retention.

CONCENTRATION AND SIZE DISTRIBUTION OF

FIBERS IN THE EXPOSURE CHAMBER

Concentration:
The overall time weighted concentration of airborne fibers in the
exposure chamber was 297 ± 25.5 fiber/ml air for the entire period of
six hours. Concentrations determined by the four air sampling filters
(90 minutes each) were 327, 295, 265, and 302 fiber/ml.

Size Distribution:
The means of the count median length (MCML) and the count median
diameter (MCMD) of airborne fibers in the exposure chamber were
13.3 μm and 0.92 μm with geometric standard deviations (GSD) of 2.64
and 1.96 respectively.

CONCENTRATION AND SIZE DISTRIBUTION OF

FIBERS IN ANIMALS

Concentration:
Fibers counted in animals were recovered from two compartments;

LRT and URT. The sum of fibers counted in the two compartments

was reported as total number of fibers in the animal or ERT.

The average total number of fibers recovered from the entire
respiratory tract (ERT) of groups I, II and III were
8.5x10⁶ ± 7.28x10⁵, 7.6x10⁶ ± 5.62x10⁵ and 6.2x10⁶ ± 6.9x10⁵fibers/animal respectively.

Size Distribution:
The MCML and MCMD of fibers recovered from the lungs LRT slightly
decreased from 10.6 μm and 0.76 μm in the first group to 9.95 μm and
0.75 μm in the second group to 9.25 μm and 0.71 μm in the third group.
In the URT, the MCML slightly increased from groups I and II to group
III; i.e. from 14.1 μm in groups I and II to 15.5 μm in group III.
The MCMD increased from 0.88 μm in the first group to 0.91 μm in the
second group, and sharply decreased to 0.69 μm in the third group.
The MCML and MCMD for the ERT showed trends similar to those

of the LRT.

FIBER RETENTION IN ANIMALS

The overall mean retention with S.D. for the LRT was 28.5% ± 1.44 for
group I, 25.8% ± 1.05 for group II and 21.9% ± 1.62 for group III.
For URT, these values were 2.25% ± 0.484, 1.30% ± 0.425 and
0.078% ± 0.008, and for ERT, retention was 30.8% ± 1.61, 27.1% ± 1.33,
and 21.9% ± 1.66 respectively.

Assuming that the average number of fibers of all sizes recovered from
group I represents 100% retention (zero clearance), then the overall
retention percent in the LRT of groups II and III are 91.1 and 78.5
respectively. In the URT, relative retention in groups II and III are
59.0 and 3.58, and that in the ERT are 89.4 and 72.9.

CONCENTRATION AND SIZE DISTRIBUTION OF

FIBERS IN LOBES AND TRACHEAE

Concentration:
The average number of fibers recovered from the lobes and tracheae of
animals in group I ranged from 1.40x10⁵ ± 3.42x10⁴ in the ML lobe to
1.88x10⁶ ± 2.28x10⁵ in the LR lobe. Lobes of groups II and III showed
similar trends with lowest number of fibers counted for ML lobes and
highest number of fibers counted for LR lobes.

In groups I and II, the number of fibers counted in the trachea was
considerably higher than the number of fibers counted in the trachea
of the third group. The average number of fibers ranged from 6.14x10⁵in the first group to 2.2x10⁴ in the third group.

Size Distribution:
Results indicated that the CML and CMD were not dependent on the
weight of the lobe.

FIBER RETENTION IN LOBES AND TRACHEAE

Results clearly showed that the mean retention of fibers in the upper
left lobe (UL) in all three groups was the highest. Fiber retention
in the lobes indicated that retention was not dependent on the weight
and size of the lobe but on the position of the lobe within the
respiratory tract as indicated later in section "I" below.

The average number of fibers counted per lobe, however, indicated

that retention consistently decreased from group I to group III implying
some clearance over time. The same is true for fiber retention in the
trachea. However, clearance in this compartment of the respiratory
tract was much more rapid. The average number of fibers dropped

from 6.14x10⁵ in the first group to 3.62x10⁵ in the second group and

to 2.20x10⁴ in the third group.

EVENNESS INDEX (E.I.) FOR FIBERS

RETAINED IN LOBES

The evenness index was determined for all lobes of the animals in the
three groups to eliminate the effect of lobar weights on fiber
retention. The evenness index for any lobe was calculated by dividing
the fiber concentration (number of fibers per gram weight of that
lobe) by the fiber concentration per gram weight of the entire lung,
and multiplying the result by 100. If fibers were proportionally
retained in all lobes then the evenness index would be approximately
100% for all lobes.

Results showed that, in groups I and II, the mean evenness index was
identical in the order of the lobes share of all fibers retained in
the lungs. In a descending order, these lobes were UL, UR, AR, LL,
LR, and ML with an evenness index of >100% for the first three and
<100% for the last three lobes. This trend was slightly different in
group III. The order of the lobes in this group was UL, UR, LR, AR,
LL, and ML with E.I. of >100% for the first two and <100% for the last
four lobes.

In all groups, however, the highest evenness index was for the UL lobe
and the lowest for the ML lobe. The results also showed that for any
lobe the evenness index was not significantly different from one group
to another. This was clearly indicated by the overlapping of the
confidence limits.

FIBER CLEARANCE

The average number of fibers recovered from the LRT of the three
groups were 7.9 ± 0.76, 7.2 ± 0.55, and 6.2 ± 0.69 million
fibers/guinea pig. For the URT, these averages were 0.61 ± 0.10,
0.36 ± 0.10, and 0.02 ± 0.002 million fibers. The sum of these two
compartments represents the total average number of fibers recovered
from the entire respiratory tract (ERT) of the three groups.

The above results indicate that clearance of fibers in the respiratory
tract appears to be associated with two clearance half-lives; the
first representing the fast clearance in the ciliated airways, and the
second representing a slow clearance of fibers in the non-ciliated
airways or the pulmonary compartment. The equations for fiber
retention in both compartments as a function of time were determined
by linear regression procedure. Estimates of the constants of these
equations show that the half-life is about 25 hours in the URT and
about 362 hours in the LRT.

For URT; R = 0.633[EXP(-0.671T)]
For LRT; R = 7.75[EXP(-0.046T)]

where R is retention in million fibers and T is time in days.

EQUIVALENT DIAMETER CONCEPT

The question of particle size arises when considering aerosol
particles of extreme shapes such as fibers where the particle
aerodynamic behavior is not well defined as in the case of perfect
spheres. For non-spherical particles several measurements defining
particle behavior can be made using properties such as particle mass,
surface area, mobility in an electric field, and impaction. These
measures are described as "equivalent" diameters. In this study,
fiber retention in the ERT of the three groups was calculated by three
concepts of equivalent diameters.

Retention by D_ie for groups I, II and III peaked at 1.58, 1.49 and
1.45 μm respectively. Retention calculated by D_me for the three
groups peaked at 1.66, 1.56 and 1.52 μm. Retention consistently
decreased as the equivalent diameter of the fiber increased
approaching zero for fibers >15 μm.

RETENTION AND CLEARANCE OF

LONG AND THIN FIBERS

Long and thin fibers (>10 μm & <0.56 μm) constituted about 21% of all
fibers retained in the animal's LRT in all three groups. Retention of
fibers of these sizes at any time was determined by the following equation:

R = 1.651[EXP(-0.047T)]

where R is retention in million fibers and T is time in days.
Estimates of the constant of this equation indicate that the half-life
for these fibers is about 354 hours.

DISCUSSION OF RESULTS

The main objective of this study was to determine deposition of
ceramic fibers in the lungs of guinea pigs. Nineteen animals were
exposed to the dust by inhalation for six hours and sacrificed
immediately, 1 day and 5 days after exposure. Retention of fibers in
any compartment of the respiratory tract was calculated as percent of
all fibers inhaled by that compartment. Overall retentions were then
used to calculate fiber clearance as a function of time. Retention
was also calculated in relation to different equivalent diameters.

The CML and CMD of airborne fibers and fibers recovered from the
respiratory tract of the exposed animals indicated that generating and
delivering of the desired dust cloud properties were efficient and
that the exposure procedure was satisfactory.

The exposure had no apparent negative effects on the well being of the
guinea pigs as shown by the physical and morphological conditions of
the animals after exposure for six continuous hours. The reduction in
body weights exhibited by groups II and III was not significant.

The determination of the minute volume by Guyton's equation yielded
values within the ranges reported in literature. The mean and standard
deviation determined for all 19 animals were 264 ± 20.7 ml/minute.
This value is in agreement with minute volumes reported in other studies
Guyton (1947), Mauderly et al. (1979) and Silbaugh et al. (1981).
Lobar minute volumes estimated by using the equation described by

Hammad and Rowhani (1984), developed for rats, also appeared to

be applicable for guinea pigs.

Characterization of the airborne fibers in the inhalation chamber
showed a concentration remarkably close to the target concentration
required. The CML and CMD and their respective GSDs indicated

that the dust cloud generated covered necessary size ranges and the
characteristics of the dust cloud clearly showed that the components
of the exposure chamber were performing as expected.

The fiber size distribution in the three groups revealed a slight
decrease in the MCML in the LRT and ERT indicating that long fibers
were clearing at a faster rate than short fibers. This rapid
clearance of long fibers suggests that a significant portion were
retained in the ciliated airways and were gradually removed upward by
the movement of the cilia. In the case of the URT, the MCML slightly
increased from groups I and II to group III, further supporting the
conclusion that long ceramic fibers were deposited in the trachea and
ciliated airways.

The MCMD also decreased from group I to Group III in the

LRT and ERT. In the URT, however, the MCMD increased from

group I to group II and then decreased in group III.

The observation that long and thin fibers were able to penetrate
deeper in the lungs than long and thick fibers suggests that thick
fibers were intercepted in the upper respiratory tract.

In the lobes, although the MCML was slightly higher for the upper left
and right lobes in group I, the fiber size distribution showed that
the MCML and MCMD in all three groups were not dependent on

the lobe weight nor on the lobe location. The MCML in all the lobes

slightly decreased from group I to group III. For the MCMD this trend

was not the same, however, the difference was subtle.

Overall retentions percent in the LRT of the three groups ranged from
28.5 ± 1.44 in group I to 21.9 ± 1.62 in group III. It is interesting
to note that deposition in humans is about 24% for compact particles
(Task Group On Lung Dynamics, 1966).

Results indicate that short-term fiber clearance by length in the LRT
and ERT of groups II and III have similar trends; as fiber clearance
is much faster for fibers with length >10 μm. In the case of URT,
the clearance rate for all fiber lengths is basically the same.
Short-term clearance patterns by diameter in the LRT and ERT have
trends similar to those obtained by length. Clearance in these two
compartments shows much faster rates for relatively thick fibers with
diameter >1.5 μm. In the URT clearance was basically the same for all
fiber diameters. The fact that the overall retention in the LRT and
ERT of the third group is the same , 21.9%, further supports this
hypothesis. It should be noted that the short-term clearance patterns
are completely different from long-term clearance rates reported in
the literature. In the later, clearance is more rapid for short and
thin fibers (Hammad, 1982).

Results also suggest that clearance of fibers in the respiratory tract
is associated with two half-lives; fast clearance in the ciliated
airways and slow clearance in the non-ciliated airways. Since the
ciliated airways are extended and embedded into the lungs it becomes
impossible to separate the two compartments, and therefore the
retention of fibers as a function of time observed for the LRT does
incorporate two types of clearance mechanisms.

The overall retention and mean retention percent by length for all
diameters and by diameter for all lengths show that fiber retention
peaked at fiber length of about 6.5 μm and at fiber diameter of about
0.5 μm in the LRT and ERT of the three groups. Retention of fibers
>6.5 μm in length and >0.5 μm in diameter decreased consistently as
fiber dimensions increased. This is most likely due to retention and
removal of larger fibers in the upper respiratory tract (URT). This
hypothesis is supported by the retention observed in the trachea of
the animal. Retention in this compartment was higher for long and
relatively thick fibers than short and thin fibers.

Comparison between this study and pulmonary deposition for MMMF

in rats (Hammad, 1982), and asbestos fibers in humans (Timbrell, 1982)
shows that for MMMF and asbestos depositions peaked at fiber length

of 7.5 and 6.5 μm respectively. In this study retention was highest

at 6.5 μm.

Deposition by fiber diameter, on the other hand, peaked at 0.5 μm for
MMMF and at 0.85 μm for asbestos. In this study retention by fiber
diameter was highest at 0.40 μm in groups I and II and at 0.56 for
group III. Retention of fibers by length for all diameters and by
diameter for all lengths in this study agrees with that reported by
both Hammad (1982) and Timbrell (1982). The slight difference in
retention by both length and diameter among the three studies could be
related to the anatomical and physiological differences among the
three models, and the shape and size distribution of the fibrous dust.

In individual lobes, the retention exhibited trends similar to those
observed in the whole lung. Generally, retention in the left lobes
were higher than the corresponding right lobes. The highest retention
was observed in the UL; although, by weight, it was only the fourth
largest of all lung lobes, followed by the UR which was the third
largest lobe. This implies that retention was not proportional to the
weight of the lobe. Retention in the rest of the lobes also indicated
that fiber retention was not dependent on lobe weight. Retention in
the lobes seems to be dependent on the location of the lobes within
the lung and the average distance from the principal bronchus and
trachea. Similar trends of retention values and the evenness index
were observed in the various lobes of groups I and II. For group III,
the trend was little different from I and II. It is not clear whether
this difference is due to the longer survival period or the clearance
mechanism in this group.

Retention by length for all diameters and by diameter for all lengths
in lung lobes shows similar results to that in the entire respiratory
tract. Fiber retention in all lobes peaked at fiber length of about
6.5 μm and fiber diameter of ≤0.5 μm. Retention of fibers >6.5 μm

in length and >0.5 μm in diameter decreased as fiber dimensions

increased.

Retention by equivalent diameters indicated that the D_ae is not
adequate to relate fiber retention to fiber dimensions. Calculation
of fiber retention by D_ie and by D_me indicated that these two concepts
of equivalent diameters are more efficient descriptors of fiber
retention in relation to fiber dimensions.

Polynomial regression of fiber retention in the ERT of the three
groups by D_ie and D_me respectively peaked at fiber dimensions
(equivalent diameter) of about 1.5 μm and rapidly decreased as the
equivalent diameter of fiber dimensions increased approaching zero for
fibers >15 μm. Equations for fiber retention in the ERT of the three
groups showed very high correlation coefficients of approximately 0.95.

As indicated before in the results section, fibers associated with the
highest toxic potential, i.e., those >10 μm and <0.5 μm , constituted a
significant portion (21%) of all fibers retained in the LRT for the
three groups. Their slow clearance rate which is comparable to that
calculated for all fibers indicates that they remain in contact with
lung tissue for a significant period of time to manifest their toxic
effects...