Tissue Reaction to Ceramic Implant Material

J. HARMS and E. MAUSLE, Orthopedic University Clinic,

HomburglSaar, Germany and Pathological Institute of the Saarland
University, Germany

Summary
The biocompatibility of alumina ceramic was tested by means of macrophage
cultures, intraperitoneal (i.p.) and intramuscular (i.m.) application of powdered
particles in rats, and by implantation of solid samples in the paravertebral muscles and
the condylus of femur. No acute cytotoxicity was found in macrophage cultures. The
i.p. and i.m. application of powdered particles in the beginning showed a granulocytic
reaction, later followed by a histiocytic reaction. Also, the morphological changes in
the organs of reticulo histiocytic system (RHS) are shown. In the solid sample
implantation, the fibrogenetic stimulus is measured by morphological analysis of the
connective tissue membrane around the sample. The importance of the individual cell
observation by transmission-electron microscope (TEM) examination is evidenced. The
experimental results are compared to the environmental reaction of smaller animal-
adapted prostheses and prostheses having been implanted in human patients. Good
hiocompatibility is confirmed by these investigations; also regarding central position of
the macrophages, the environmental reaction due to implant material is shown.

INTRODUCTION
Up until now no standardized guidelines for the testing of bio-
compatibility of an implant material were available. For the first time
reference points were drawn up in the ASTM Standard, F 361-372
(19731.l In 1975/76 the working group of the DGOT2 worked out a
recommendation to assess the biocompatibility; the Americans also
attempted to elaborate a standardization (Autian, ref. 3).
In order to determine the biocompatibility of a material, on
principle, we regard three biological reactions to be important. They
are examined in different forms of presentation in different biological
systems. These concern the assessment of the cytotoxicity, the his-
tocompatibility, and the wear behavior i n vivo. Prior to clinical tests of
the combination, the prosthesis ceramic-metal (Mittelmeier, ref.
4) utilized by us was tested in relation to metal and ceramic parts to

Journal of Biomedical Materials Research, Vol. 13,67-87 (1979)

0 1979 John Wiley & Sons, Inc. 0021-9304/79/0013-0067$01.00
68 HARMS AND MAUSLE

determine the biocompatibility. The results of the ceramic tests are

presented here.

MATERIAL AND METHOD

Using alumina ceramic as the implant material, the production of
powdered particles is as follows: The ceramic powdered particles
consist of pure A1303 (alumina). It is rendered soluble from bauxite in
a humid chemical process and then subjected to various glow processes
(bauxite-aluminum hydroxide-aluminum oxide). The aluminum oxide
thus produced is ground down to the desired particle size. The grinding
rollers also consist of alumina ceramic. The production of the solid
bodies is as follows: The solid bodies of alumina ceramic are made by
means of pressing and trickling; their sizes correspond to the ASTM
standard (F 361-372).

Cell Culture

Morphological Examination of the Peritoneal Macrophages

With reference to the method indicated by Beck5 and Unkeless,6 we
obtain peritoneal macrophages. In each case ten female rats with a
weight of 120-150 g are injected intraperitoneally (i.p.) with 5 ml
Thioglycolate 2.5% four days before they are killed. The cells are
explanted in a density of 2.4 X 106 cells/cm2 in square containers and
Leighton tubes. After 24 hr of explantation the cultures are rinsed with
Eagles medium in order to detach the cells not sticking to the surface.
The cell cultures are then contaminated for 2 and 24 hr with a
suspension of ceramic-powdered particles contained in the medium
(size of particles: 0-5 pm and 5-10 pm) in a concentration of 50 pg/ml
adding 1 ml to the square containers and 0.1 ml to the Leighton tubes.
Fixation in 2% Glutaraldehyd, coloring with methylene blue (UNNA),
or according to Pappenheim.

Measuring of the DNS Synthesis in the Lymphocyte

Transformation Test
We obtained peritoneal macrophages as described above. We
contaminated 2 X 106 cells/cm2 with ceramic particles of a size between 0
and 10 pm for 24 hr. We scraped off cells from the containers by means of
repeated freezing and defreezing destruction of the
 TISSUE REACTION TO CERAMIC IMPLANT MATERIAL 69

macrophages; the ceramic particles were centrifuged together with

the cell fragments a t 2500 g/15 min (Hepple~ton~)A. human
lymphocyte culture cultivated with Eagles medium was
contaminated with the supernatant macrophage culture for 66 hr.
After 48 hr, i.e., 18 hr before the end of the experiment, we added
H3thymidin. The cell cultures were absorbed on fiber glass filter
strips; the incorporation of H3-thymidin in comparison with a
nontreated control culture was measured in cpm with the fluid
scintillation counter Aquasol (Intertechnik).

Powdered Particle Implantation i.m. and i.p.

In each case standardized powdered particles of 2-5 pm and 5-10
pm size in a NaCl suspension are injected intramuscularly (i.m.) and
intraperitonealy (i.p.) into four animals. In i.m. application the
powdered particle concentration is 150 mg/ml and in i.p. application
it is 500 mg/ml. Tissue was taken off 1 day, 4 weeks, and 26 weeks
after implantation under chloral hydrate anesthesia. In i.m.
application the surrounding muscle shell and the lymphatic nodes in
the area of thea. iliaca externa and the bifurcation of the aorta are
taken off; in i.p. application parietal peritoneum, mesenteric root,
spleen liver, and vertebral bodies are taken out. Fixation in
10%formalin, and in 3.5% phosphate-buffered (pH = 7.4)
glutaraldehyde. Embedding in paraffin, the tissue fixed in
glutaraldehyde is washed in phosphate buffer, subsequent fixation
with phosphate-buffered 1% Os04 for 2 hr, embedding in araldit.

Solid Body Implantation i.m. and i.0.

The samples made according to the ASTM-standard (F 361-372)
(4,1.6 mm and length, 6.3 mm) are implanted i.m. into the paraver-
tebral muscles and intraosseously (i.0.) transversally into the femure
condyle. In each case samples were taken of surrounding muscle and
bone shell, respectively, after 2,4,26, and 52 weeks from four animals
under chloral hydrate anesthesia. Fixation took place in phosphate-
buffered (pH = 7.4) 3.5%glutaraldehyde. In i.m. implantation the
muscle shell is divided parallel to the longitudinal axle of the implant
body after fixation, the implant body is taken out, and fixed in
glutaraldehyde. In i.0. implantation decalcification of the femure
condyle in 20%EDTA (PH = 7.2, Trilon B@)at a temperature of 56OC
up to 36 hr. Here, too, the femure condyle parallel to the longitudinal
I0 HARMS AND MAUSLE

axis was divided, when the implanted solid body was taken off, and
when muscle and bone disks were cut, respectively, in a vertical
direction to the longitudinal axis of the implant. There was
subsequent fixation with Os04 for 2 hr and flat embedding in
araldit. The semithin cuts of a thickness of 0.5 pm are colored with
methylene blue (UNNA). Part of the preparation is processed for the
examination with the electron microscope: ultrathin sections of 500-
800 A are made (Reichert Om42), contrasting with uranyl acetate and
lead citrate, by examination in the electron microscope (Zeiss-EM 9 2).

Morphornetry

In i.m. implantation the membrane thickness is determined with

the measuring ocular for the four animals at each point in time for
each of the ten cases in which points at a magnification of 250X
were measured. Subsequent variance analysis after prior
determination of homogenous variances with the Bartlett test. The
differences are verified by the multiple T test according to Duncan.

Implantation of Smaller Adapted Prostheses in Dogs

Under sterile conditions smaller animals-adapted prostheses with

a metal alloy stem, but a head and socket of ceramics, were
implanted in ten dogs without using bone cement. After 3,6, and 12
months the animals were killed, taking out the newly formed
capsula. The process used consisted of fixation in formalin,
embedding in paraffin, and HE staining.

Comparison of These Results with the Newly Formed Capsulas

in Human Implantations

From 15 patients, who had been supplied -with a ceramic-metal

combination prosthesis for a period of 3 to 18 months, newly
formed capsulas could be taken for a histological examination. The
process used consisted of fixation in formalin, embedding in
paraffin, and HE staining.
 TISSUE REACTION TO CERAMIC IMPLANT MATERIAL 71

RESULTS

Cell Culture
Within a reaction of 2 and 24 hr of ceramic-powdered particles of
a size of up to 10 pm to cell cultures of glycolate stimulated
peritoneal macrophages, no cell damage specific to the material is
to be discerned. Within the cytoplasm of the control peritoneal
macrophages vacuoles can be seen, which seem to be concerned
glycolate-induced vacuoles. After contamination the ceramic
particles are partly absorbed by macrophages (Fig. 1): they partly
are adsorbed a t the cell membrane, and partly free in the medium.
No quantifyable differences between reaction periods of 2 and 24 hr
can be stated. A comparison with the control macrophages on the
Leighton tubes shows that after a 24-hr implantation, remarkably
less particle-loaded peritoneal macrophages stick to the Leighton
tubes than is the case after a 2-hr exposition of powdered particles.
Within the few scattered fibroblasts we could only find a minimal
absorption of powdered particles.
With the lymphocyt transformation test no decrease of the
synthesis rate under the influence of the supernatant macrophages
treated with powdered particles could be stated. In the case of the
nontreated control sample, 16 000 cpm were measured as compared

Fig. 1. Cell culture of peritoneal macrophages: powdered particles of

alumina ceramic (5 pm) stored within macrophages. Pappenheim, X806.
72 HARMS AND MAUSLE

to 15 200 in the treated macrophages. The decrease is statistically

insignificant.

Powdered Particle Implantation i.m. and i.p.

1. One day after i.m. application the powdered particle depots are
interspersed by neutrophile, and in a few cases we also find eosinophile
granulocytes and few scattered lymphocyts. At the border of the depots
we find that muscle fibers are disintegrating, which are interspersed by
granulocytes as well. After 4 weeks of implantation only very few
granulocytes are found between the powdered particle depots; the
powder is now mixed with histiocyts. The powder is very densely
packed within the macrophages 26 weeks later; the nuclei themselves
are very small and chromatin dense. Between powder loaded cells we
find a large number of capillaries and very fine collagenous fibers. A
connective tissue capsula, however, was not formed. Between the
macrophages there are also a few lymphocyts. In a few places we also
found that small foci can be observed, in which lymphocyts are lying
closely together in a folliclelike manner. In the regional lymphatic
nodes several ceramic-loaded macrophages are found. With a particle
size of 10 pm there are essentially the same reactions; the histiocyts are,
however, not as loaded as in the 5 pm powder, and thus the powder
depots seem to have more cells.
2. In i.p. application of a ceramic powder of a size of 5 pm can be
found one day thereafter in the serosa and in the mesenterium. The
powder particle depots are again interspersed by granulocytes; in
addition, however, there are more lymphocyts than seen in i.m.
application and a few macrophages, of which some have already
absorbed particles. The liver and spleen do not contain any ceramic
particles. The 10 pm particles show an identical reaction.
Four weeks later in the mesenterium we find, in addition to small
groups, several separate large focuslike concentrations of ceramic-
powdered particles interspersed with mononuclear cells. The leucocyts
have nearly completely disappeared. For a part we observe a beginning
formation of fibers. In the spleen and the liver there are only a few
powder-loaded macrophages to be found. The foci at 10 pm particles
show a wall of lymphocyts and many large macrophages around the
particles. At this time no powdered particles can be evidenced in the
liver and spleen.
Twenty-six weeks later there is essentially the same picture as in
i.m. application. There are a large number of capillaries between the
 TISSUE REACTION TO CERAMIC IMPLANT MATERIAL 73

powder-loaded histiocyts. Only rarely fine collagenous fibers have

formed. In the spleen we find a few macrophages loaded with
ceramic-powdered particles. In the liver we observe scattered cells
with ceramic particles in the portal area. In the case of 10 pm
particles there is a similar picture. There are, however, more
powdered par-ti,cles to be seen lying freely between the
macrophages. Compared to the 5 pm particles there seems to be a
slightly increased fiber formation (Fig. 2). Less particles lie in the
liver and spleen than with the application of the 5 p-size. None of
the animals had ceramic-powdered particles in the bone marrow.

Solid Body Implantation i.m. and i.0.

I.m. Implantation
Light microscopic evidence: 1. In the case of i.m. implantation of
solid samples after 2 weeks there is a connective tissue membrane rich
in cells having many histiocyts, sparse lymphocyts, and granulocytes
around the implant. At this time a closed layer of flattened histiocyts
and multinuclear giant cells lie adjacent to the implant. There are
sparse collagenous fibers only in the outer rim of the membrane. At
the border we find a few scattered atrophic muscle cells. Four weeks
later the connective tissue membrane is thinner

Fig. 2. Powdered psticles of AlzO3-ceramic, i.p. application, 26 weeks post-op.:

particles (5 pm) densly packed within macrophages HE, X640.
74 HARMS AND MAUSLE

and contains fewer cells. In the direction towards the implant there
is a uninterrupted cell layer (Fig. 3). Here 1-3 layers having
relatively high cell content are concerned, which all lie on the same
level. In addition to these mononuclear histiocytic cells we also
have multi-nuclear giant cells. In the area of the inmost parts of the
membrane, already matured and aligned collagenous fibers can be
evidenced. Twenty-six weeks later the picture is quite similar. The
cover cells superposed to the membrane, however, mostly consist
of one layer. While the membrane seems to be smaller, it also
seems to have more fibers. Within the membrane there are narrow
drawn cells (fibro-blasts and histiocyts) to be found. Within the
histiocyts, in the covering layer as well as in the membrane,
ceramic particles can be found. Fifty-two weeks later there is a
similar picture. At the border of the implant-host tissue as well as
within the membrane more often ceramic particles stored within the
cells are observed. A few individual ceramic-loaded cells lie in the
sectors at the outer parts of the membrane, preferably perivasal. In the
light microscope the cells of the connective tissue membrane and the
cover cells do not show any damage. After 52 weeks an uninterrupted
coat of adipose tissue has formed around the connective tissue
membrane, which separates the membrane from the muscle (Fig. 4). A
further cellular reaction cannot be observed in the area and also no
changes of the surrounding muscle shell are observed.

Fig. 3. Solid samples of AlzOs-ceramic, i.m. application, 4 weeks post-op.:

connective tissue membrane with an uninterrupted cell layer of macrophages, semithin
section, methylene blue, X640.
 TISSUE REACTION TO CERAMIC IMPLANT MATERIAL 75

Fig. 4. Solid samples of AlzOa-ceramic, i.m. application, 52 weeks post-op.: fatty

tissue ( t) separates connective tissue membrane from the muscle; semithin section,
methylene blue, X40.

2. Morphometry of the connective tissue membrane: Two weeks

later the implant has a surrounding connective tissue membrane of an
average thickness of 77.5 pm. Thereafter we observe a continual
decrease in the thickness: after 4 weeks to 46 pm, after 8 weeks to 36
pm, and after 26 weeks to 24.4 pm. Later there is no further decrease
of the thickness noticeable. The graphic chart shows an asymptotic
behavior of the membrane thickness.
3. Findings with the electron microscope: Two weeks later closely
adjacent histiocyts form an interior layer containing only scattered
lymphocyts and granulocytes. Towards the exterior means: With regard to
the connective tissue membrane the histiocyts lay interior
(= adjacent to the implant) and the fibroblasts are situated beneath the
histiocyts, i.e. exterior fibroblasts with well developed RER with
wide tubules follow. The histiocyts touch each other with numerous
long pseudopodia. Already at this point the inmost cell layer forms an
epitheliumlike layer. The mononuclear and multinuclear cells are
cytoplasm rich and stick together by fingerlike interlocking
protrusions. The cell membrane of this inmost cell layer is extremely
dense in contrast. In the slightly osmiophilic hyaloplasm there are
small plump mitochondria with lamellar cristae, numerous small
vesicles, which carry mostly ribosomes, relatively few free ribosomes,
numerous small primary, and several larger secondary lysosomes. The
lysosomes partly lie preferably around the several Golgi fields
76 HARMS AND MAUSLE

near the nucleus. In most cases the nucleolus is situated peripherally.

In a few cases there are decaying cells in the inner half of the
connective tissue membrane. The outer part of the connective tissue
membrane shows long histiocyts with a varying content of secondary
lysosomes as well as activated fibroblasts.
In the interstice we find a few collagenous fibrils only in the exterior
parts of the membrane; in the more interior areas, we observe only a
small fiber felt of protofibrils.
After 4 weeks the inmost histiocyt layer consists only of 1-3 rows.
The granulocytes and lymphocyts have nearly disappeared. The inmost
cell layer is hardly unchanged. Only in the secondary lyso-somes in
some cases there are small crystal-like ceramic inclusions which are
dense in contrast. Under the cell membrane, close to the implant, there
is a narrow zone of hyaloplasm without organelles. The inclusions of
ceramic particles can also be found in the histiocyts of the exterior
layers of the connective tissue membrane. The fi-broblasts are still
activated. Transversely striated collagenous fibrils are, however,
already lying under the histiocyt layer in the area of the inmost
connective tissue membrane.
After 8 weeks we find only a single cell layer inside the connective
tissue membrane. The epitheliumlike connected macrophages are
narrower. Their hyaloplasm is denser in contrast and the mito-
chondria are larger, rounder, and have an electron-dense matrix. Up to
26 and 52 weeks the connective tissue membranes become richer in
fibers; the collagenous fibrils are located in vertical, and possibly in
spiral-formed rows. The innermost layer of macrophages is long and
narrow. In the dense hyaloplasm we find large round mito-chondria
(Fig. 5), which are only sparse ribosomes-carrying vesicles.
Sometimes under the cell membrane there are scattered small
concentrical layers of lamelles. In the secondary lysosomes we more
often observe crystal deposits of ceramic. The fibrocyts are narrow and
have narrow RER tubules. Between the rows of collagenous fibers lie
narrow long histiocyts, which increasingly contain particles of
ceramics in secondary lysosomes. These lysosomes show-insofar as no
artefacts were produced by the knife of the ultramicrotome-an
undamaged membrane with, in most cases, several crystals in the
interior at the same time (Fig. 6). The mitochondria are normal and
degenerative changes cannot be discerned.
As compared to the different metal alloys (Ti-Al-Va alloy and Co-
 TISSUE REACTION TO CERAMIC IMPLANT MATERIAL 77

Fig. 5. Solid samples of A120a-ceramic,.m. application, 52 weeks post-op.: inmost

layer of macrophages, within the hyaloplasm large round mitochondria. X17,760.

Ni-Cr-Mo alloy) lipoid deposits within the macrophages cannot be

evidenced.

Intraosseous Implantation
After 2 weeks the wide space around the hole of the prosthesis shows
entensive new bone formations in the form of narrow and branched-out
trabecules, partly in a radial and partly in a shell-like order. There is very
little connective tissue around the prosthesis. After 4 weeks we find, to a
large extent, uninterrupted bone cover with sparse connective tissue, and
at the exterior part, wide medullary spaces.
Up to the 26th week no fundamental change occurs. The trabecules
show, however, a functional alignment and are connected with the
original bone. In the meantime the latter has adapted itself also to the
new functional unit.
After 52 weeks the uninterrupted bone cover around the prosthesis
is slightly thicker than in the 26th week. The trabecules aligned
functionally are a little plumper and show signs of a repeated
reconstruction. Also after this period of time there are still narrow rests
78 HARMS AND MAUSLE

Fig. 6. Solid samples of AlzOa-ceramic, i.m. application, 52 weeks post-op.: crystal

of ceramic in sec. lysosomes ( f ) membrane undamaged (A),collagenous fibrils (CF)
located in vertebral rows. X28,500.

of a connective tissue membrane or a separate cell layer between

the implant and the bone cover.

Implantation of Smaller Adapted Prostheses in Dogs

The examination of the newly formed capsula after implantation
of smaller animal-adapted ceramic- metal combination prosthesis
shows only little ceramic wear debris within the period under
observation (Fig. 7). The loaded histiocyts are stored without any
further cell reaction, preferably pericapillar. At a few of the
prostheses, increased metal debris can be evidenced (insufficiently
marked conic clamping between metal stem and ceramic head). We
cannot prove an increase of the wear debris within the capsula due
to the limited time of our experiment.

Human Implantation
In the newly formed capsula examined by us after total endopros-
thesis implantation (Fig. 8) with ceramic articulating parts, we found
very little wear debris. Only in one case we stated a high A1203wear
debris; postoperatively there occurred a relapsing subluxation of the
prosthesis, which led to a friction of the edge with a high quantity of
A1203wear debris. The debris particles had a size of 0.5-2 pm, and
 TISSUE REACTION TO CERAMIC IMPLANT MATERIAL 79

Fig. 7. New formed capsula in dog, 52 weeks after implantation of a ceramic pros-
thesis: small amount of wear debris, particles, 1pm, mostly stored within macrophages.
HE, X320.

also form a conglomerate. The wear debris particles are mostly stored
within macrophages; partly they lie separately in the tissue [Fig. 8(b)].
The ceramic-loaded macrophages accumulate perivasal in the capsula
parts, mostly at the exterior part. In the cases examined by us we did
not find any foreign-body giant cells.

DISCUSSION
Peritoneal Macrophages
The usual foreign body implants are detached from the host tissue
by means of a connective tissue coat on which, towards the implant,
we find a more or less closed layer of macro phage^.^-^^ Thus the
macrophage represents the cell reacting with the implant. Therefore it
is also useful to test the cytotoxicity of a potential implant material
against a macrophage culture, we use the models known from the
silicose research.'5J6 Cultures of fibroblasts do not seem to be suitable
any longer for the envisaged examination, because the fibroblast is
never in direct contact with the implant. In contrast to Beck et al.16 we
could only evidence minor storage capability of the fibroblasts in the
cultures used in our tests. With examination in the light microscope we
do not find any cell damage for a reaction period of 2 and 24 hr. This
corresponds to the results of the phagocytosis of Korund
80 HARMS AND MAUSLE

(a)
Fig. 8. (a) Self-locking prostheses (Mittelmeier). (b) New formed capsula in human
patient. 52 weeks after implantation a ceramic prosthesis, only small amount of wear
debris, stored within macrophages, HE, X640.

(Al.oxyde) titanium dioxyde and asbestos, whereas in a quartz-

phagocytosis we can evidence a cell damage already after 5-8 hr,
which in the light microscope expresses itself in a growing
vacuolization and in a ball -shape of the cells.16 These findings are in
line with the findings for the electron microscope, which we observed
with the macrophages in the surrounding membrane in the case of
solid body implantation after 52 weeks. As in the phagocytosis of
nontoxic substances-Korund-titanium dioxyde and asbestos16-our
ceramic particles lie covered by a membrane in secondary lysosomes
without stating a damage of the membrane as this is the case with the sili-
cates.lG2O Also the lymphocyt transformation test applied by us for the
first time to determine the cytotoxicity did not prove any inhibition of the
DNA synthesis by macrophages possibly damaged by ceramic. We
applied the lymphocyt transformation test in the method modified by
Greaves21and Pees.22 In it the DNS synthesis
 TISSUE REACTION T O CERAMIC IMPLANT MATERIAL 81

(b)
Fig. 8. (Continued fromprevious page.)
of the human lymphocyts stimulated by PHA is determined under
the presence of the examined biomaterial. In uitro, it serves for the
recording of the immunological cell reaction against the added
implant material.
We, however, do not arrive at a statement concerning a possibly
existing material -specific toxic effect on the mitotic capacity of the
macrophages by this testing. Also a material-specific fibrogenetic
stimulation7cannot be determined by this LT test. However, we have
the impression that the informative character of the in vitro test can be
improved by applying the lymphocyt transformation test together with
the morphological assessment of the macrophage cultures. Here
a possibly existing material-specific cytotoxicity is, corresponding to
the silicose formation, effective only after passing through
macrophages, which are always the primarily reacting cells. We do not
regard the direct contamination of an implant material with cell
cultures and their morphological a s s e ~ s m e n t ~as~ ,suitable~~-~
~to determine the cytotoxicity of an implant material.
The informative content of the cytotoxic damage missing in the cell
82 HARMS AND MAUSLE

culture is, however, limited by the fact that in the in vitro

examination the interference existing in vivo between lymphocyts
and macrophages cannot be sufficiently taken into account.

In Vivo Implantation of Powdered Particles

The in vivo implantation of powdered particles is to simulate model-
like the local (i.m. application) and the general (i.p. application)
reactions within the framework of a possible material wear debris. The
intraperitoneal test was introduced in the silicose research to determine
the fibrogenetic effectivity of powdered particles.26 In order to arrive
at comparable values that can be transferred to the in vivo
implantation, the powdered particles must be characterised with regard
to their chemical-physical state and must be applied in model-like
sizes.20,25,27-29For the ceramic-powdered particles we chose a size
between 0 and 5 pm and between 5 and 10 pm, that is approximately
the actual particle size with in vivo metal and ceramic wear debris.30-
35 A t the beginning a mostly granulocytic reaction prevails with more
or less lymphocytic infiltration. From the second week on, the acute
inflammation phase receded and the histiocytic reaction prevails. We
could not find a well-marked fiber formation
as it was observed in the case of stainless In accordance with E s c a
l a ~there~~ was no immunologically significant reaction discerned
by us.
We limited ourselves to a descriptive assessment of the
environmental reaction, since the semiquantitative evaluations
carried out so far26,29,32,36-39do not allow an exact classification
and we regard also in accordance with Willert40a quantitative
assessment of the cell reaction as impossible due to the high quantity of
interfering factors in the case of the powder application.
The ceramic particles are deported into the regional lymphatic
nodes, spleen, and liver. The particle-loaded macrophages
evidenced in these organs do not show any significant cell damage
in the light microscope. Within the examination period we could not
prove any deposits in the bone marrow. Whether this reaction-free
behavior continues on with a longer implantation period and
increased powdered particle quantity, as it is, on principle,
conceivable in case of a prosthesis fracture, is questionable. With
pathological high wear debris in animal tests already after eight
weeks clumps of Kupffer cells were formed in the liver.41
As is the case with all other implant materials we do not know up
 TISSUE REACTION TO CERAMIC IMPLANT MATERIAL 83

until now, in the case of ceramics, where the particles are finally
deposited and what biological reactions are there generated by the
debris particles. So far examinations after long-term application of
debris particles have only been carried out at the place of the
primary reaction, the newly formed capsula; a tissue metaplasia
cannot be evidenced there,10*35@though Mittelmeier and
Singer43pointed out this possibility already.

Solid Body Implantation i.m. and i.0.

The solid body implantation makes a morphometric analysis of
the membrane and an observation of the individual cell on the
border area implant-host tissue possible. From the fourth week on
signs of inflammation are not any more evidenced; under the
histiocyt layer there is a connective tissue membrane, which with
proceeding maturation of the collagenous fibers becomes narrower.
In contrast to silicates the ceramic obviously does not induce any
fibrogenetic stimulus. Up to the 52nd week a shifting layer of fatty
tissue has formed between membrane and muscle for a functional
exclusion of the implant.
By the morphometric analysis of the connective tissue membrane we
observed up to the 26th week a continual decrease of the membrane
thickness. Our membrane thickness at that time (24.4 pm) corresponds to the
Escalas et al. values.32 When using a CaOA1203-ceramic, however, the
membrane is considerably thicker (100-200 pm).44 The time comparison
with the histiology shows that from the membrane thickness we can draw
conclusions as to the duration of the stimulus generated fibrogenetically by
the foreign body, what also corresponds to the Laings i d e a ~ . ~Thus5 the
morphometry enables us to assess this foreign body stimulus also
quantitatively, what is considerably difficult when we only regard the cell r e
a ~ t i o n . ~In~ , ~ ~ , ~ ~ comparison to the observation with the electron
microscope here, however, only the fibrogenetic stimulus is determined.
With the electron microscope the cells superposed to the connective
tissue membrane towards the implant can be identified as macro-
phages,14 which do not suffer any damage during the experiment,
although with increasing duration a higher number of particles broken
out of the solid body are absorbed in the cells. In contrast hereto li-poid
inclusions into the macrophages, sometimes also formation of
granulomas can be observed with different metal alloys under similar
experimental conditions.& In the case of these macrophages the same
84 HARMS AND MAUSLE

cell layer is concerned, that is also described by other authors using b i

o g l a ~ s . l ~ J "Hench~~ and Paschal111see the possibility of a
chemical compound between living and nonliving structures in the
very close connection between these cells and the glass ceramic 45
S5F. We do not support this opinion, because we see a macrophage
covering layer with the various implant material without evidence of
direct connection between implant material and host tissue.
In the case of intraosseous application the sequestrationlike
exclusions of the implant proceeds by an uninterrupted newly formed
bone ring. As Boutin48 and Dorre et al.49we were not able to prove
any direct contact between newly formed bone and ceramic implant.
Thus we cannot confirm a direct bone-ceramic contact50 or a direct
connection bone-implant through collagenous structures51 as well as
the ingrowth of vessels into the ceramic body.52 In our investigations
we always found a histiocytic interim layer between implant and bone.
On account of preparative difficulties it is, however, not always
possible to keep this cell layer uninterrupted.
In examinations of newly formed capsulas of dogs and human
patients there is a wear debris of ceramic particles. These are stored in
macrophages; on account of the perivasal position of the particle-
loaded macrophages we can assume a deportation into the RHS. Here
we can, however, not regard the extent of the scar formation at the
newly formed capsula only as an expression of the fibrogenetic
effectivity of the ceramic. Here the functional stimulus is stronger than
that specific to the material, however, an unphysiological high wear
debris can change this situation.

SUMMARY
To determine the biocompatibility of the implant material A1203-
ceramic the cell and tissue reactions caused by ceramic were examined
in vitro and in uiuo.
In the cell culture from peritoneal macrophages no material specific
damage of the cells is to be evidenced; in correspondence hereto there
is no impairment of the DNS synthesis in the lymphocyt
transformation test.
With i.m. and i.p. application of 0-10 pm size AlZOs-powdered
particles we cannot discern a granuloma formation comparable to
silicose. After the fading of the acute inflammatory phase from the
second week on, there prevails a mostly histiocytic reaction. We did
not observe a significant fiber formation.
 TISSUE REACTION TO CERAMIC IMPLANT MATERIAL 85

With i.m. and i.0. implantation of solid bodies, no direct connection

between implant and host tissue is produced: there is always a his-
tiocyt layer in between. The phagocytosis of small ceramic particles by
these histiocyts leads, in contrast to the quartz powder phagocytosis, to
no damage of the cell organelles, especially of lysosomes. The
decreasing thickness of the newly formed connective tissue mefnbrane
supports the opinion that the foreign body effect generated by the
implant is not passed on to the environment by the intermittent
histiocyts. In i.m. application the functional exclusion of the implant is
done by a fatty tissue shifting layer that is formed between membrane
and muscle, in i.0. application by a newly formed bone ring that
surrounds the implant completely.
The good biocompatibility of ceramic in these experiments is
also confirmed under functional stress in animal tests with smaller
pros-theses and with human patients in newly formed capsulas that
were obtained by prostheses changes. The test method used by us
makes it possible to determine model-like and comparably the cell
and tissue reactions in uitro and in animal tests that are
responsible for the biocompatibility of an implant material. Due to
this test model, the risk for human patients in the transformation to
human in implantation can be kept a t a low level.

References
1. ASTM-Standard: ASTM., designation F 361-372, annual book of ASTM-
standards, 1973.
2. Deutsche Gesellschaft fur Orthopadie und Traumatologie (German Society of
Orthopedic and Traumatological Surgery), Arbeitskreis Biomaterial: Vorschlag fur
eine stoffliche und biologische Prufung von Implantatwerkstoffen und Im-
plantaten fur die orthopadische Chirurgie (1976).
3. J. Autian, Plast. Med. Surg. 27,l (1976).
4. H. Mittelmeier, in Aduances in Artificial Hip and Knee Joint Technology, M. S.
Chaldach und D. Hohmann, Eds., Springer Verlag, Berlin, 1976.
5. E. G. Beck, Koln-Opladen, Westd. Verlag, Berlin, 1970.
6. J. Unkeless, S. Gordon, and E. Reich, J . Exp. Med., 139,834 (1974).
7. A. G. Heppleston, in Inhaled Particles, Vol. I., W. H. Walton, Ed., Unwin
Brothers Ltd., Surrey, England, 1970, pp. 357-369.
8. D. H. Collins, J . Path. Bact., 65, 109 (1953).
9. W .Hoppe, Dt. Zahnarztl. Ztschr., 11,837 (1956).
10. H. G. Willert and P. Puls, Arch. Orthop. Unfall-Chir., 72,33 (1972).
11. L. L. Hench and H. A. Paschal1,J. Biomed. Mater. Res. (Symp .4), 25 (1973).
12. H. U. Cameron and V. L. Fornasier, J . Biomed. Mater. Res., 10,769 (1976).
13. A. E. Clark, L. L. Hench, and H. A. Paschall, J . Biomed. Mater. Res., 10, 161
(1976).
14. J. Harms and E. Maude, Med. Orthopad. Tech., 4,103 (1976).
86 HARMS AND MAUSLE

15. E. G. Beck, Zbl. Bakt. Parasit. I n f e k . KRH . Hyg., 198,92 (1965).

16. E. G. Beck, J. Bruch, and J. Sach, Ergebn. Unters. Geb. Staub und Silikoseerf.
Steinkohlenbergbau, 6,131 (1967).
17. E. G. Beck, J. Bruch, K. H. Friederichs, W. Hilscher, and F. Pott, in Inhaled
Particles, Vol. I , W. H. Walton, Ed., Unwin Brothers Ltd., Surrey, England, 1970, pp.
447-486.
18. J. Bruch, in Inhaled Particles, Vol. I, W. H. Walton, Ed., Unwin Brothers, Ltd.,
Surrey, England, 1970, p. 447.
19. A. C. Allison, J. S. Harington, and M. Birrbeck, J . Expt . Med., 124,141 (1966).
20. A. C. Allison, in Inhaled Particles, Vol. I, W. H. Walton, Ed., Unwin Brothers
Ltd., Surrey, England, 1970, pp. 437-440.
21. M. F. Greaves, F. Janossy, and M. Doenhoff, Nature (London),248,698 (1974).
22. H. Pess, A. Pappas, and P. G . Scheuerlen, Verhdlg. Dtsch. Ges. Inn. Med.,
81,1203 (1975).
23. L. Hulliger, Arch. Orthop. Linfall-Chir., 54,581 (1962);L. Hulliger, 0. Pohler,
and F. Straumann, 2.Ges. Exp . Med., 114,145 (1967).
24. M. Mital and J. Cohen, J . Bone J t . Surg., 50A, 547 (1968).
25. A. M. Pappas and J . Cohen, J . Bone J t . Surg., 50A, 535 (1968).
26. J. R. Ruttner, in Fortschr. der Staublungenforschung Bd. I , H. Rebloch and W.
Klosterkotter, Eds., Niederrhein, Druckerei, 1963, pp. 639-642.
27. R. Z.Schultz and C. R. Williams, J . Indust. Hyg. Toxicol., 24,75 (1942).
28. A. Swensson, d. Glomme, and G. Bloom, A.M.A. Arch. Ind. Health, 14, 482
(1956).
29. J. Cohen, J . Bone Jt . Surg., 41A, 152 (1959).
30. S.A. V. Swanson, M. A. R. Freeman, and J. C. Heath, J . Bone J t . Surg., 55B,
424 (1973).
31. C. B. Charosky, P. G. Bullongh, and P. D. Wilson, J . Bone Jt . Surg., 55A, 49
(1973).
32 . F. Escalas, d. Galante, and W. Rostoker, J . Biomed. Mater. Res., 10,175 (1976).
33. C. N. J., Wagner, A. H. Shbaik, D. J. Schurman, and H. C. Amstutz, J.
Biomed.
Mater. Res., 10,653 (1976).
34. M. Ungethum, personliche mitteilung, 1977.
35. J . Harms and E. Mausle, Orthop. Praxis, I, 33 (1977).
36. C. A. Homsy, J . Biomed. Mater. Res., 4,341 (1970).
37. J. E. Turner, W. H. Lawrence, and J. Autian, J . Biomed. Mater. Res., 7, 39 (1973).

38. B. J. Zweifach, L. Gramt, and R. T. Melluskey, The Inflammatory Process,

Academic, New York, 1973.
39. H. G. Willert and M. Semlitsch, Med. Orthopiid. Tech., 4,94 (1976).
40. H. G. Willert, D. Kiihne, K. D. Eckert, M. Semlitsch, and E. Dorre, J .
Biomed. Mater. Res., to appear.
41. E. Mausle and J. Harms, unpublished.
42. d. Charnley, Acrylic Cement in Orthopedic Surgery, Williams & Wilkins Company,
Baltimore, Md., 1970.
43. H. Mittelmeier und L. Singer, Arch. Orthop. Unfall -Chir., 48,519 (1956).
44. S. F. Hulbert, S. J. Morrison, and J. J. Klawitter, J . Biomed. Mater. Res., 6,347
(1972).
45. P. G. Laing, A. B. Ferguson, and E. Hodge, J . Biomed. Mater. RPS., 1, 135 (1967).

46. J. Harms and E. Mausle, J . Bioengineering, im Druck.

 TISSUE REACTION TO CERAMIC IMPLANT MATERIAL 87

47. B. A. Blencke, P. Allersee, H. Bromer, and E. Pfeil, Arch. Orthop. Unfull-Chir.,

82,135 (1975).
48. P. Boutin, R6u. Chir. Orthop., 58,229 (1972).
49. E. Dorre, D. Geduldig, M. Happel, R. Lade, P . Prussner, H. G. Willert, and L.
Zichner, J . Biorned.Muter. Res., 10,493 (1976).
50. S. F. Hulbert, J. R. Matthews, J. J. Klawitter, B. W. Sauer, and R. B. Leonard, J .
Biomed. Muter. Res. (Syrnp. 5), (1),85 (1974).
51. R. L. Hentrich, G. A. Graves, H. G. Stein, and P. K. Baipai, J . Biomed. Muter. Res.,
5,25 (1971).
52. F, W. Rhinelander, M. Royweyha, and J. C . Milner, J . Biorned. Muter. Res.,
5,81 (1971).

Received November 13,1977

Revised May 3,1978