Abstract
Background: Metal-on-metal bearing surfaces have been reintroduced
for use during total hip replacement. To assess tissue reactions to various
types of articulations, we studied the histological appearance of
periprosthetic tissues retrieved from around metal-on-metal and
metal-on-polyethylene total hip replacements and compared these findings with
the appearance of control tissues retrieved at the time of primary
arthroplasty.
Methods: Periprosthetic tissues were obtained at the time of
revision of twenty-five cobalt chromium-on-cobalt chromium, nine cobalt
chromium-on-polyethylene, and ten titanium-on-polyethylene total hip
arthroplasties. Control tissues were obtained from nine osteoarthritic hips at
the time of primary total hip arthroplasty. Each tissue sample was processed
for routine histological analysis, and sections were stained with hematoxylin
and eosin. Quantitative stereological analysis was performed with use of light
microscopy.
Results: Tissue samples obtained from hips with metal-on-metal
implants displayed a pattern of well-demarcated tissue layers. A prominent
feature, seen in seventeen of twenty-five tissue samples, was a pattern of
perivascular infiltration of lymphocytes. In ten of the tissue samples
obtained from hips with metal-on-metal prostheses, there was also an
accumulation of plasma cells in association with macrophages that contained
metallic wear-debris particles. The surfaces of tissues obtained from hips
with metal-on-metal prostheses were more ulcerated than those obtained from
hips with other types of implants, particularly in the region immediately
superficial to areas of perivascular lymphocytic infiltration. The lymphocytic
infiltration was more pronounced in samples obtained at the time of revision
because of aseptic failure than in samples retrieved at the time of autopsy or
during arthrotomy for reasons other than aseptic failure.
Total-joint-replacement and surface-replacement designs of metal-on-metal
prostheses were associated with similar results. Tissue samples obtained from
hips with metal-on-polyethylene implants showed far less surface ulceration,
much less distinction between tissue layers, no pattern of lymphocytic
infiltration, and no plasma cells. The inflammation was predominantly
histiocytic. Tissues retrieved from hips undergoing primary joint replacement
showed dense scar tissue and minimal inflammation.
Conclusions and Clinical Relevance: The pattern and type of
inflammation seen in periprosthetic tissues obtained from hips with
metal-on-metal and metal-on-polyethylene implants are very different. At the
present time, we do not know the prevalence or clinical implications of these
histologic findings, but we suggest that they may represent a novel mode of
failure for some metal-on-metal joint replacements.
The recognition that osteolysis can result from the accumulation of
polyethylene wear debris has led to a resurgence of interest in alternative
bearing surfaces. Improvements in engineering tolerances and a better
understanding of tribology have resulted in a new generation of metal-on-metal
prostheses that are made from cobalt-chromium
alloys1-3.
The Metasul bearing was introduced in 1988, and more than 60,000 such bearings
have now been implanted
worldwide4. In 2000,
Willert et al.5
analyzed tissues that had been retrieved from fourteen hips in which a
contemporary metal-on-metal joint replacement had failed after an average of
thirty months in vivo. The investigators found an unusual perivascular
lymphocytic infiltrate, which they suggested was similar to that found in
association with type-IV hypersensitivity. A review of tissues retrieved from
the area around a number of McKee-Farrar first-generation metal-on-metal
bearings revealed a similar pattern, but in a smaller percentage of
cases6. Several
authors have reported on the histological appearance of periprosthetic tissues
obtained from the area around early
metal-on-metal7-9
and
metal-on-polyethylene10-12
joint replacements. Those investigators used semiquantitative scales to assess
the degree of cellular infiltration, but they did not identify any lymphocytic
or plasma-cell infiltration of the periprosthetic tissues.
The purpose of the present study was to undertake a quantitative
stereological analysis of periprosthetic tissues retrieved from patients with
contemporary metal-on-metal and metal-on-polyethylene joint replacements. The
appearance of these tissues was compared with that of control tissues obtained
from osteoarthritic hips at the time of primary joint arthroplasty.
Neocapsules that had been retrieved from twenty-five hips with
metal-on-metal articulations were donated by the Joint Replacement Institute,
Orthopaedic Hospital, Los Angeles, California, and from the collection of one
of us (H.G.W.) (see Appendix). Tissues were chosen for study if the specimens
were of sufficient length and thickness to allow for an adequate histological
evaluation. No tissues from infected hips were included in the study. Tissue
samples were obtained at the time of revision of the original components
because of aseptic failure (nineteen hips), at the time of exploration of the
joint for impingement (two hips) or for removal of heterotopic bone (one hip),
or at the time of autopsy (three hips). Of the twenty-five metal-on-metal
implants, fourteen were total-hip-replacement designs and eleven were
surface-replacement designs.
Additional tissues that had been retrieved from nineteen hips with a
metal-on-polyethylene surface replacement were donated by the Joint
Replacement Institute, Orthopaedic Hospital, Los Angeles, California. All
nineteen implants were of the porous surface replacement (PSR) design (see
Appendix). Of these, ten were titanium-on-polyethylene implants (Zimmer,
Warsaw, Indiana) and nine were cobalt chromium-on-polyethylene implants
(DePuy, Warsaw, Indiana). All tissues were obtained at the time of revision of
the original components because of aseptic failure, and all hips demonstrated
osteolysis and polyethylene wear. Control capsular tissues were harvested,
after informed consent had been obtained, from nine patients undergoing
primary total joint replacement for the treatment of osteoarthritis of the
hip.
Histologic Analysis
Tissues that were obtained intraoperatively were immediately preserved in
formalin. Several samples from each tissue site were embedded in paraffin,
sectioned, and stained with hematoxylin and eosin.
All histological analyses were performed with use of a Leitz Dialux 22
Laboratory and Research Light Microscope at final magnifications of 25×,
100×, and 400×. A calibrated graticule was used to measure lengths
and areas, and cell counts were made within this known area. The extent of
inflammation was expressed in cells/mm2. The entire tissue section
was first studied at 25× magnification to determine the nature and size
of the specimen. The site and approximate length of any surfaces were
recorded. Any tissue damage or processing artefact was noted. If no surface
layer was present, the section was discarded. The initial anatomical layer of
interest was the surface. The length of the surface was measured with use of
the graticule, and the extent of surface ulceration was classified with use of
a fourcategory scale (Figs.1-A,
1-B,
1-C,
1-D). The surface was
classified as type-1 if the cellular lining was intact, as type-2 if the
cellular lining had been lost but no fibrin exudation was noted, as type-3 if
the cellular lining had been lost and fibrin deposition was noted, and as
type-4 if gross disruption, surface fissuring, and fibrin deposition were
noted. Type-2, 3, and 4 surfaces were considered to represent varying degrees
of ulceration.
Various layers of the tissue specimens were studied
(Fig. 2). Layer 1 comprised the
surface membrane, Layer 2 comprised the subsurface layer lying superficial to
vascular structures, Layer 3 comprised the vascular layer of the neocapsule,
and Layer 4 comprised the tissues deep to the neocapsule, including fibrous
tissue, fat, and muscle.
Five representative sites, equally spaced across the surface of the
specimen, were chosen for analysis. If deeper tissue layers were missing at
the chosen site, the nearest site with intact deeper layers was studied. The
classification of the surface at each site was recorded. Next, the thicknesses
of Layers 2 and 3 were measured in turn with use of the graticule. After the
thickness of Layer 2 had been measured, the graticule was sited over the layer
at 400× magnification, and quantitative cell counts were made with use
of a multiple block of click-counters. In this way, lymphocytes, macrophages,
giant cells, plasma cells, and eosinophils were counted. Macrophages
containing visible wear-debris particles were scored with use of the 4-point
scale described by Doorn et
al.9.
Layer 3 was then assessed. The graticule was placed over the layer at
100× magnification, and the number of vessels falling within the
graticule area was recorded. The number of these vessels that had a cuff of
lymphocytes was recorded, and the percentage of vessels with a lymphocytic
cuff was calculated. The thickness of each cuff was recorded in terms of
numbers of cells as well as in micrometers. The thickness of each cuff was
considered to be the radius of the cuff from the vessel wall to the outer edge
of the cuff. When a cuff had been sectioned obliquely, the mean thickness of
the cuff was recorded. The cells within these cuffs were then identified and
quantified. Individual tissue lymphocytes were found to measure 6 to 8 µm
in diameter, in keeping with the findings of previous
studies13. A
separate cell count was performed in order to record all of the cells that
fell within the original low-power fields but that were not part of a
perivascular cuff. This cell count was used to determine the so-called
background level of inflammation for comparison with sections in which
perivascular cuffing was not present. Macrophages with phagocytosed debris
were recorded on the 4-point scale described by Doorn et
al.9.
When present, the deepest layer (Layer 4) was then studied. This layer
represented the tissues beyond the neocapsule of the prosthesis and was not
universally present because of loss and disruption during the processing of
the tissue specimens.
Statistical Analysis
All data were entered into a computerized database for statistical
analysis. The data were found to be nonparametric, and therefore nonparametric
analysis was performed with use of the chi-square statistic and the
Kruskal-Wallis one-way analysis of variance. The level of significance was set
at p < 0.05.
Table I shows the median thickness of each layer and the mean number
of vessels per low-power field in the vascular layer (Layer 3) for each group
of specimens obtained from hips with different types of implants.
The surfaces of the tissues that had been retrieved from hips with
metal-on-metal implants showed a greater extent and severity of ulceration
when compared with the surfaces of tissues that had been retrieved from hips
with other types of prostheses. The proportion of the surface of each specimen
that demonstrated either type-3 changes (fibrin deposition) or type-4 changes
(gross disruption and fissuring) was expressed as a percentage. Analysis of
the tissues that had been retrieved from the nineteen hips in which a
metal-on-metal prosthesis had been revised because of aseptic loosening
revealed that a mean of 55% (range, 25% to 100%) of the total surface length
showed type-3 or 4 changes. In contrast, the tissues that had been retrieved
from the area around six metal-on-metal prostheses for reasons other than
aseptic loosening did not show the same extent or severity of surface
ulceration, with a mean of 30% (range, 15% to 60%) of the total surface length
demonstrating type-3 or 4 changes. Analysis of the tissues that had been
retrieved from the nine hips with a cobalt chromium-on-polyethylene implant
revealed that a mean of 30% (range, 0% to 65%) of the total surface length had
type-3 or 4 changes. Analysis of the tissues that had been retrieved from the
ten hips with a titanium-on-polyethylene implant revealed that 15% (range, 0%
to 50%) of the surface length had type-3 or 4 changes. The differences between
the conditions of the surfaces of specimens from hips with the various types
of implants are shown in Table
II. These differences were highly significant (p <
0.001).
The tissues that had been retrieved from all twenty-five hips that had a
metal-on-metal implant showed a clear boundary between the subsurface layer
and the vascular layer (Fig.
2). The cellular infiltrate that was seen in Layer 2 of the
tissues that had been obtained from hips with a metal-on-metal implant was
typically diffuse and lymphocytic in nature. There was no relationship between
the quality of the tissue surface and the cellular infiltration of Layer
2.
The cellular infiltrate in the deeper vascular layer (Layer 3) was quite
different from that seen in Layer 2. Analysis of the tissues revealed no
diffuse infiltration but rather a focal arrangement of perivascular
lymphocytic cuffs that surrounded the small nonmuscular vessels of this layer
(Fig. 3). These cuffs typically
were large (radius, >100 µm) and dense (>200 cells/mm2).
Among the specimens that had been obtained from hips with metal-on-metal
articulations, those with the greatest extent and severity of surface
ulceration also showed the greatest extent of perivascular lymphocytic
infiltration of Layer 3. The median density of cells in perivascular
lymphocytic infiltrates was 32 cells/mm2 in areas deep to a type-1
(intact) surface. This density of inflammation increased to 132
cells/mm2 in areas deep to a surface with type-2 changes (loss of
cellular lining). In areas where the surface demonstrated type-3 changes
(fibrin deposition) or type-4 changes (gross disruption), the median density
of inflammation in the perivascular lymphocytic infiltrates increased further
to 234 cells/mm2 and 741 cells/mm2, respectively. The
differences between these densities of inflammation were highly significant (p
< 0.001).
Layer 4 was intact in sixteen of the twenty-five specimens that had been
obtained from hips with a metal-on-metal prosthesis. Some very small
perivascular lymphocytic cuffs were identified in twelve of these sixteen
specimens, but these had a very different appearance from those seen in the
vascular layer of the periprosthetic tissues. These cuffs comprised no more
than a two-cell-thickness layer surrounding thin-walled nonmuscular vessels in
regions of fat or muscle. In each case, these cuffs had a radius of =15
µm. The cellular density of these cuffs was low (<20
cells/mm2).
Metallic particulate debris was identified within macrophages in twelve of
the twenty-five tissue specimens that had been obtained from hips with
metal-on-metal implants. In ten of the specimens in which debris-filled
macrophages were prominent, the macrophages were closely associated with a
localized infiltration of plasma cells
(Fig. 4). In the material
investigated for the present study, plasma cells were identified only in
specimens with macrophages that were laden with metallic debris and only in
the immediate vicinity of those macrophages. Plasma cells were not identified
in any of the available tissue specimens that had been obtained from hips with
metal-on-polyethylene prostheses.
The tissue surface was typically less ulcerated in specimens from hips with
metal-on-polyethylene prostheses than it was in specimens from hips with
metal-on-metal prostheses (Table
II). While the anatomical boundary between the subsurface layer
(Layer 2) and the vascular layer (Layer 3) could be discerned in most cases,
the distinction was far less obvious than it was in specimens from hips with
metal-on-metal implants. The cellular component of Layer 2 was not lymphocytic
but comprised a foreign-body reaction to the presence of extensive
polyethylene debris. Macrophages and fibroblasts were the predominant cell
types identified, with giant cells also being present. Metallic debris was not
seen in any of the specimens from hips with cobalt chromium-on-polyethylene
implants. Metal-stained macrophages were present in eight of the ten specimens
from hips with titanium-on-polyethylene implants. Notably, perivascular
lymphocytic cuffs were not seen in the vascular layer of the neocapsule of any
of the specimens from hips with metal-on-polyethylene implants. The number of
specimens in each group that demonstrated a focal perivascular infiltration of
lymphocytes within the vascular layer is shown in
Table III. The differences
between these groups were significant (p < 0.001). Once again, very small
perivascular cuffs of lymphocytes were identified in areas of fat within Layer
4 of two of the six specimens from hips with a cobalt chromium-on-polyethylene
implant and in one of the five specimens from hips with a
titanium-on-polyethylene implant in which this layer was intact. These cuffs
were very different from those seen in the specimens from hips with
metal-on-metal implants and in each case comprised no more than a
two-cell-thickness layer around the vascular lumen and had a radius of no more
than 15 µm. These cuffs were of low cellular density (<15
cells/mm2).
Plasma cells were not identified in any of the high-power images of
specimens from hips with metal-on-polyethylene implants.
Table IV shows the number of
specimens from each group in which plasma cells were identified. The
differences between the various groups were significant (p < 0.01).
Tissues harvested from patients undergoing primary joint replacement for
the treatment of osteoarthritis were predominantly fibrous, with a very sparse
cellular component. The surfaces of the joint capsule were classified as type
1 or 2 in all cases. Lymphocytes were not seen either diffusely in Layer 2 or
focally in Layer 3 of the joint capsule in any of these tissues. Layer-4
tissues were present in five of the nine control specimens. In two specimens,
very small perivascular cuffs of lymphocytes were identified within areas of
fat in Layer 4, beyond the neocapsule. Again, these cuffs were very different
from those seen in the vascular layer of the tissues from hips with
metal-on-metal prostheses and in each case comprised a one-cell-thickness
layer around the vascular lumen with a radius of no more than 8 µm and a
low density of cells (<10 cells/mm2). Plasma cells were not seen
in any of the specimens from hips undergoing primary joint replacement. The
anatomical boundaries between the subsurface and vascular layers were
difficult to identify because of the hypovascularity of the tissues.
We found that the anatomical layers of the neocapsule that forms
around different types of articulations exhibit very different patterns of
inflammation. Perhaps this finding is not surprising given the differing types
of wear debris generated by the various articulations. Metal-on-metal
prostheses generate less volumetric wear than comparable metal-on-polyethylene
prostheses do4. The
particles produced are, however, orders of magnitude
smaller14 and
therefore as many as 100 times more particles are produced by metal-on-metal
as compared with metal-on-polyethylene
bearings14.
Such very small particles may provoke biological reactions that have not
been previously recognized.
We also found that the surface ulceration was more extensive and more
severe in periprosthetic tissues from hips with aseptic loosening of
metal-on-metal prostheses when compared with those from hips with well-fixed
metal-on-metal or metal-on-polyethylene prostheses. The extent and severity of
ulceration of the surface tissue was associated with the presence of
perivascular infiltration with lymphocytes in the deeper vascular layer of the
neocapsule. The association between perivascular lymphocytic cuffing and
ulceration of the adjacent tissue surface is not yet understood. In recent
unpublished experiments, we found that synovial fluid retrieved from hips with
metal-on-metal implants may be toxic to human fibroblasts in tissue culture.
It is therefore possible that the cells lining the pseudosynovial neocapsule
around these prostheses may not have the secretory and regulatory capabilities
of normal synovial cells and thus may expose the deeper layers of the
neocapsule to higher concentrations of these substances. Whether these
substances are capable of eliciting the distinct lymphocytic reaction seen in
tissues from hips with metal-on-metal implants is not known.
Hartman et al.15
described the histologic appearance of biopsy specimens of skin that had been
obtained from the sites of sensitization and challenge tests following surface
application of 2,4-dinitrochlorobenzene (DNCB). Their findings were remarkably
similar to ours. Howie and
Vernon-Roberts16
demonstrated that intra-articular injection of cobalt-chromium wear particles
in rats produced synovial surface ulceration and a dense infiltrate of small
lymphocytes in the subsurface tissue layer. These findings lend weight to the
hypothesis that the primary insult is to the surface of the tissue and that
the perivascular lymphocytic infiltrate is the secondary phenomenon. The
findings of Howie and
Vernon-Roberts16
also indicate that exposure of articular synovial tissue to cobalt-chromium
debris is sufficient to provoke surface ulceration and a lymphocytic
infiltrate in the absence of a loose prosthesis.
Perivascular lymphocytic vasculitis has been previously described in a
variety of clinical settings. One variant of giant-cell arteritis features a
nonspecific panarteritis with a mixed inflammatory infiltrate composed largely
of lymphocytes and macrophages but no giant
cells17. Takayasu
arteritis and polyarteritis nodosa also can have an identical histological
picture with perivascular cuffing of the vasa vasorum with lymphocytes and
other mononuclear
cells17. The
histological changes in some forms of dermatitis also feature perivascular
cuffing with a variety of cell types, including
lymphocytes18. The
earliest stages of contact dermatitis feature a superficial perivascular
lymphocytic infiltrate. Such changes also are seen in association with
erythema multiforme and discoid lupus
erythematosus18.
Perivascular lymphocytic infiltration therefore has been identified in a
variety of systemic vasculitides affecting small, medium, and large vessels
and also is a characteristic feature of both acute and chronic forms of
dermatitis. At the present time, it is unclear to what extent the perivascular
cuffs of lymphocytes that we observed represent a true active vasculitis in
the periprosthetic tissues around metal-on-metal prostheses or whether this
finding is a novel form of immunological reaction with unknown, perhaps
benign, consequences.
Lymphocytic infiltration also has been observed in periprosthetic tissues
obtained during revision of metal-on-metal joint-replacement implants. Willert
and Semlitsch19
reported infiltrates of lymphocytes and plasma cells in the absence of
evidence of bacterial infection in some tissues in their series of
first-generation metal-on-metal prostheses. Similar features were then found
in tissues retrieved from all fourteen hips with contemporary metal-on-metal
prostheses that had failed at an average of thirty months after
implantation5,6.
Other authors have reported individual cases of lymphocytic infiltration
around failed modern metal-on-metal
implants20,21.
In contrast, infiltrates of lymphocytes or plasma cells have not been reported
in studies of tissues from around
metal-on-polyethylene10-12,19,22,23
or
ceramic-on-ceramic5,6
prostheses. Boss et
al.24 described
"inflammatory infiltrates" including white blood cells around
failed short-term cemented metal-on-polyethylene hip replacements, but they
concluded that the predominant cell species present in periprosthetic tissues
were macrophages and giant cells. The histological appearance of the tissues
from metal-on-polyethylene articulations in the present study was similar to
that in other
series10-12,19,22-24.
The predominant cell types were macrophages and fibroblasts, with occasional
giant cells.
There were notable differences between the histological changes seen in the
specimens obtained from the nineteen hips with aseptic loosening of
metal-on-metal implants and those obtained from the six hips with well-fixed
metal-on-metal implants. The latter group showed less-extensive and
less-severe surface ulceration. Thus, the histologic changes of surface
ulceration and lymphocytic perivascular cuffs appear to be associated with the
aseptic loosening of some of these contemporary metal-on-metal implants. These
findings are in agreement with those of previous studies that have analyzed
tissues from patients with failed metal-on-metal
prostheses5,6,19-21.
It is important to note that these changes were seen in association with both
total-hip-replacement and surface-replacement designs of metal-on-metal
prostheses. Devices from a number of different manufacturers showed similar
histological changes. Surface ulceration and prominent perivascular
lymphocytic cuffs therefore appear to be associated with metal-on-metal
prostheses per se and not with any particular design of such implants.
The finding of a thin layer of lymphocytes around small vessels within
areas of fat and muscle in Layer 4 of all of the tissue types studied suggests
that a small number of such features may be a normal histological component of
deeper, nonarticular tissues before and after joint replacement. Such features
are known to be a histological feature of a variety of normal tissues,
including skin. The perivascular cuffs found in Layer 4 universally had less
than a two-cell thickness around the lumen of small vessels and had an
entirely different appearance compared with the large, prominent cuffs seen in
Layer 3 of the tissues obtained from hips with metal-on-metal prostheses.
Plasma cells were seen in ten of the tissue specimens from hips with
metal-on-metal articulations. Plasma cells are mature B-lymphocytes that are
primed to manufacture a particular antibody subtype upon repeated exposure to
a specific antigen. In the hips available for study, these cells were closely
associated with macrophages laden with metallic debris
(Fig. 4). This close
association suggests a form of local immunological response to the presence of
that metal. In the present study, no such reaction was seen in specimens from
hips with cobalt chromium-on-polyethylene implants, suggesting either that the
metallic debris generated by metal-on-metal articulations may differ in some
way from that generated by metal-on-polyethylene articulations or that the
plasma-cell reaction is not related to the presence of metallic debris. The
prominent perivascular lymphocytic cuffing seen in Layer 3 of the tissues from
hips with metal-on-metal implants also suggests the occurrence of some form of
immunological response that was not seen in the specimens from hips with
metal-on-polyethylene implants. These findings, taken together with the early
clinical failure of some of the metal-on-metal implants, suggest that a novel
biological mechanism may lead to the failure of some of these devices. The
frequency with which such an immunological response might cause the premature
failure of a metal-on-metal joint replacement is probably low but is presently
unknown.
Tables presenting details of all specimens studied are available with the
electronic versions of this article, on our web site at
(go to
the article citation and click on "Supplementary Material") and on
our quarterly CD-ROM (call our subscription department, at 781-449-9780, to
order the CD-ROM). ?
Note: The authors are grateful for the support of the Wishbone
Trust of the British Orthopaedic Association and to The Royal College of
Surgeons of England for their assistance in funding the work for the current
study.
Weber BG. [Metal-metal total prosthesis
of the hip joint: back to the future]. Z Orthop Ihre Grenzgeb.1992;130: 306-9.
German.130306
1992
[PubMed][CrossRef]
Weber BG. Experience with the Metasul
total hip bearing system. Clin Orthop.1996;329 Suppl:
S69-77.329S69
1996
[PubMed][CrossRef]
Streicher RM, Semlitsch M, Schon R,
Weber H, Rieker C. Metal-on-metal articulation for artificial hip joints:
laboratory study and clinical results. Proc Inst Mech Eng [H].1996;210:
223-32.210223
1996
[PubMed]
Sieber HP, Rieker CB, Kottig P. Analysis
of 118 second-generation metal-on-metal retrieved hip implants. J Bone
Joint Surg Br.1999;81:
46-50.8146
1999
[CrossRef]
Willert HG, Buchhorn GH, Fayyazi A,
Lohmann Ch. Vascular changes and lymphocyte aggregations in aseptic loosening
of all-metal hip joint replacements. Presented as a poster at the Sixth
World Biomaterials Congress; 2000 May 15-20; Kamuela,
HI.
2000
Willert H, Buchhorn G, Fayyazi A,
Lohmann C. Histopathological changes around metal/metal joints indicate
delayed type hypersensitivity. Primary results of 14 cases.
Osteologie.2000;9:
2-16.92
2000
Willert HG, Buchhorn GH, Gobel D, Koster
G, Schaffner S, Schenk R, Semlitsch M. Wear behavior and histopathology of
classic cemented metal on metal hip endoprostheses. Clin
Orthop.1996;329 Suppl:
S160-86.329S160
1996
[CrossRef]
Evans EM. Metal sensitivity as a cause
of bone necrosis and loosening of the prosthesis in total joint replacement.
J Bone Joint Surg Br.1974;56:
626-42.56626
1974
[PubMed]
Doorn PF, Mirra JM, Campbell PA, Amstutz
HC. Tissue reaction to metal on metal total hip prostheses. Clin
Orthop.1996;329 Suppl:
S187-205.329S187
1996
[CrossRef]
Heilmann K, Diezel PB, Rossner JA,
Brinkman KA. Morphological studies in tissues surrounding alloarthroplastic
joints. Virchows Arch A Pathol Anat Histol.1975;366:
93-106.36693
1975
[PubMed][CrossRef]
Willert HG. Reactions of the articular
capsule to wear products of artificial joint prostheses. J Biomed Mater
Res.1977;11:
157-64.11157
1977
[CrossRef]
Mirra JM, Amstutz HC, Matos M, Gold R.
The pathology of the joint tissues and its clinical relevance in prosthesis
failure. Clin Orthop.1976;117:
221-40.117221
1976
[PubMed]
Junqueira LC, Carneiro J, Kelley RO.
Basic histology. 6th ed. Norwalk, Conn: Appleton and Lange;
1989. p 112.112
1989
Firkins PJ, Tipper JL, Saadatzadeh MR,
Ingham E, Stone MH, Farrar R, Fisher J. Quantitative analysis of wear and wear
debris from metal-on-metal hip prostheses tested in a physiological hip joint
simulator. Biomed Mater Eng.2001;11:
143-57.11143
2001
[PubMed]
Hartman A, Hoedemaeker PJ, Nater JP.
Histological aspects of DNCB sensitization and challenge tests. Br J
Dermatol.1976;94:
407-16.94407
1976
[CrossRef]
Howie DW, Vernon-Roberts B. The synovial
response to intraarticular cobalt-chrome wear particles. Clin
Orthop.1988;232:
244-54.232244
1988
Schoen FJ, Cotran RS. Blood vessels. In:
Cotran RS, Kumar V, Collins T, editors. Robbins pathologic basis of
disease. 6th ed. Philadelphia: Saunders; 1999. p
493-542.493
1999
Murphy GF, Mihm MC Jr. The skin. In:
Cotran RS, Kumar V, Collins T, editors. Robbins pathologic basis of
disease. 6th ed. Philadelphia: Saunders; 1999. p
1170-1214.1170
1999
Willert HG, Semlitsch M. Tissue
reactions to plastic and metallic wear products of joint endoprostheses.
Clin Orthop.1996;333:
4-14.3334
1996
[PubMed]
Bohler M, Kanz F, Schwarz B, Steffan I,
Walter A, Plenk H Jr, Knahr K. Adverse tissue reactions to wear particles from
Co-alloy articulations, increased by alumina-blasting particle contamination
from cementless Ti-based total hip implants. A report of seven revisions with
early failure. J Bone Joint Surg Br.2002;84:
128-36.84128
2002
[PubMed][CrossRef]
Al-Saffar N. Early clinical failure of
total joint replacement in association with follicular proliferation of
B-lymphocytes: a report of two cases. J Bone Joint Surg Am.2002;84:
2270-3.842270
2002
[PubMed]
DiCarlo EF, Bullough PG. The biologic
responses to orthopedic implants and their wear debris. Clin
Mater.1992;9:
235-60.9235
1992
[CrossRef]
Howie DW. Tissue response in relation to
type of wear particles around failed hip arthroplasties. J
Arthroplasty.1990;5:
337-48.5337
1990
[CrossRef]
Boss JH, Misselevich I, Behar J, Mendes
DG. Histologic analysis of the periprosthetic tissues of long-term surviving
cemented total hip arthroplasties. J Long Term Eff Med
Implants.1996;6:
73-90.673
1996