Alumina–alumina artificial hip joints. Part I: a histological analysis and characterisation of wear debris by laser capture microdissection of tissues retrieved at revision

Alumina–alumina artificial hip joints. Part I: a histological analysis and characterisation of wear debris by laser capture microdissection of tissues retrieved at revision

Biomaterials 23 (2002) 3429–3440 Alumina–alumina artificial hip joints. Part I: a histological analysis and characterisation of wear debris by laser c...

430KB Sizes 0 Downloads 0 Views

Biomaterials 23 (2002) 3429–3440

Alumina–alumina artificial hip joints. Part I: a histological analysis and characterisation of wear debris by laser capture microdissection of tissues retrieved at revision A. Hattona, J.E. Nevelosb, A.A. Nevelosc, R.E. Banksd, J. Fisherb, E. Inghama,* a

School of Biochemistry & Molecular Biology, Division of Microbiology, University of Leeds, Leeds LS2 9JT, UK b Department of Mechanical Engineering, University of Leeds, Leeds LS2 9JT, UK c Department of Orthopaedic Surgery, Bradford Royal Infirmary, Bradford, UK d Department of Clinical Medicine, University of Leeds, Leeds LS2 9JT, UK Received 17 May 2001; accepted 31 January 2002

Abstract The aims of this study were to investigate the tissues from uncemented Mittelmeier alumina ceramic-on-ceramic total hip replacements using histological methods and to isolate and characterise the ceramic wear debris using laser capture microdissection and electron microscopy. Tissues from around 10 non-cemented Mittelmeier alumina ceramic on ceramic THRs were obtained from patients undergoing revision surgery. Tissues were also obtained from six patients who were undergoing revisions for aseptic loosening of Charnley, metal-on-polyethylene prostheses. Tissue sections were analysed using light microscopy to determine histological reactions and also the location and content of alumina ceramic wear debris. Tissue samples were extracted from sections using laser capture microdissection and the characteristics of the particles subsequently analysed by TEM and SEM. The tissues from around the ceramic-on-ceramic prostheses all demonstrated the presence of particles, which could be seen as agglomerates inside cells or in distinct channels in the tissues. The tissues from the ceramic-on-ceramic retrievals had a mixed pathology with areas that had no obvious pathology, areas that were relatively rich in macrophages and over half of the tissues had in the region of 60% necrosis/necrobiosis. In comparison, the Charnley tissues showed a granulomatous cellular reaction involving a dense macrophage infiltrate and the presence of giant cells and o30% necrosis/necrobiosis. The tissues from the ceramic prostheses also showed the presence of neutrophils and lymphocytes, which were not evident in the tissues from the Charnley retrievals. There were significantly more macrophages ðpo0:05Þ; and giant cells ðpo0:01Þ in the Charnley tissues and significantly more neutrophils ðpo0:01Þ in the ceramic-on-ceramic tissues. TEM of the laser captured tissue revealed the presence of very small alumina wear debris in the size range 5–90 nm, mean size7SD of 24719 nm whereas SEM (lower resolution) revealed particles in the 0.05–3.2 mm size range. This is the first description of nanometre sized ceramic wear particles in retrieval tissues. The bi-modal size range of alumina ceramic wear debris overlapped with the size ranges commonly observed with metal particles (10–30 nm) and particles of ultra-high molecular weight polyethylene (0.1–1000 mm). It is possible that the two size ranges of contributed to the mixed tissue pathology observed. It is speculated that the two types of ceramic wear debris are generated by two different wear mechanisms in vivo; under normal articulating conditions, relief polishing wear and very small wear debris is produced, while under conditions of microseparation of the head and cup and rim contact, intergranular and intragranular fracture and larger wear particles are generated. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Alumina ceramic; Total hip arthroplasty; Wear debris; Histology

1. Introduction The biological response to ultra-high molecular weight polyethylene (UHMWPE) wear debris, either *Corresponding author. Tel.: +44-113-233-5691; fax: +44-113-2335638. E-mail address: [email protected] (E. Ingham).

alone or in combination with other factors is believed to play an important role in the late aseptic loosening of metal-on-UHMWPE total hip replacements [1–4]. It is currently believed that UHMWPE wear debris, generated at the articulating surfaces enters the periprosthetic tissue where it is phagocytosed by macrophages. The macrophages then release pro-inflammatory cytokines and other mediators of inflammation that stimulate

0142-9612/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 2 - 9 6 1 2 ( 0 2 ) 0 0 0 4 7 - 9

3430

A. Hatton et al. / Biomaterials 23 (2002) 3429–3440

osteoclastic bone resorption leading to osteolysis and eventual loosening of the prosthesis [5–8]. Evidence to support this is has been gathered through the analysis of explanted tissues and in vitro and in vivo studies of the biological effects of UHMWPE wear debris. Histological studies have indicated a chronic granulomatous reaction associated with UHMWPE debris local to the implant with the presence of numerous particle-laden macrophages and giant cells [9]. UHMWPE wear debris has been shown to stimulate osteolytic cytokine release by macrophages in vitro and factors such as the size and volume of the debris have been shown to be critical in the response [10,11]. The problems associated with metal-on-polyethylene hip replacements have led to an increased interest in the development and use of alternative bearing surfaces, such as metal-on-metal and ceramic-on-ceramic, for artificial hip joints. Boutin was the first to introduce alumina ceramic materials for use as bearing surfaces in artificial hips in the early 1970s [12]. Alumina ceramic total hip arthroplasties have now been in clinical use, mainly in Europe, for more than 25 years. Recent improvements in design and the quality of the alumina ceramic used in production and manufacture have eliminated the initial complications such as femoral head fractures and catastrophic breakage of the ceramic sockets. Alumina ceramic has very attractive mechanical properties for use as a bearing surface. It has a high Young’s modulus and a hardness second only to that of diamond. These properties have made the alumina-on-alumina combination, a coupling of tremendous potential in the orthopaedic field. Several studies have reported good performance of the alumina ceramic-on-ceramic bearing in terms of low annual wear, in the order of o5 mm, [13–18]. However, careful operative technique with correct positioning of the prosthesis is necessary to avoid excessive wear [19–21]. Impingement between the femoral neck and the rim of the acetabular cup has been associated with massive wear and osteolysis [22] and high wear rates have been associated with ‘‘stripe’’ and severe wear of Mittelmeier prostheses [23]. Nevertheless, given good surgical technique it is possible to achieve very low volumes of wear with ceramic-onceramic prostheses. In comparison to UHMWPE and indeed, metal wear debris, little is known about the accumulation of alumina wear debris in periprosthetic tissues and its in vivo effects. Histological studies of tissues, from around revised alumina ceramic total hip replacements have produced varied results depending on whether the prostheses were cemented or uncemented. Kummer et al. [24] investigated the pseudosynovial tissue obtained from revisions due to mechanical failure of uncemented Autophor prostheses, 8–44 months following implantation and found evidence of surface roughening of the femoral head and acetabular cup associated

with wear between 10 mm and 3 mm. There were numerous particles with a mean size of 5 mm in the tissue. Henssge et al. [25] reported on the ceramic debris in pseudosynovial and soft tissue membranes from cemented alumina-on-alumina prostheses that had been revised for loosening and autopsy specimens that had been stable for more then 5 years. They observed sharpedged polygonal yellow-brown particles of up to 5 mm in diameter and smaller, granular debris inside macrophages by light microscopy in all tissue specimens. They quantitated the particles and showed that the stable prostheses had tissues with fewer particles than the loosened prostheses. They reported evidence of thickening of the membrane, extended necrosis, scar-like fibrosis, fibrin exudates and microhemorrhages in all of the tissues. These alterations were, however, less marked in stable autopsy specimens compared to tissues obtained at revision for loosening which were in turn quantitatively less than in UHMWPE-on-ceramic or metal prosthesis revisions. Lerouge et al. [26] isolated the particles from the soft tissue from the bone/cement interface of the acetabular components of Cerevar– Osteal prostheses revised for loosening after a mean implant life of 8.7 years. Analysis of the particles by scanning electron microscopy revealed a mean size of 0.4470.25 mm. They compared the histology of tissues from around the ceramic-on-ceramic prostheses with tissues from metal-on-UHMWPE total hip replacements and concluded that there was no significant difference between the two groups in relation to the cellular reaction (presence of macrophages, neutrophils, giant cells). They concluded that the cellular response observed in the ceramic-on-ceramic group was attributable to zirconia particles used as opacifying agents in the cement. Osteolysis in association with ceramic-onceramic articulations has been rarely reported, and generally osteolysis has been limited to cases of massive wear [26]. Huo et al. [27] found no evidence of osteolysis after a mean implantation time of 9 years for uncemented components. In marked contrast, a report by Yoon et al. [28] suggested that of 103 uncemented Mittelmeier Biolox ceramic-on-ceramic prostheses, 23 femoral components and 49 acetabular components had radiological signs of osteolysis at a mean implantation time of 92 months. The osteolysis associated with the femoral component was reported to be linear in 12 and scalloping in 11. The acetabular osteolysis was linear not focal. The interfacial tissue was reported to be a highly vascularised, fibrous connective tissue rich in macrophages containing electron-dense material within phagosomes. Abundant ceramic particles were observed with a mean size of 0.71 mm (range 0.13–7.2 mm.). They concluded that ceramic wear particles could stimulate a foreign body response leading to periprosthetic osteolysis. The ceramic bearing surfaces of these failed prostheses showed wear in the form of cracking and

A. Hatton et al. / Biomaterials 23 (2002) 3429–3440

3431

pain—no revision (3), stem loosening and proximal migration of acetabulum (4), proximal migration of acetabulum (5), shifting of acetabulum followed by dislocation (6), loosening of acetabulum (7), unexplained pain (8) stem loosening (9) and unexplained pain (10). The explants were taken from the successful clinical series previously described by our group [30]. The average implantation time was 6 years 4 months, range 1–8 years. The explanted Mittelmeier heads and cups had the characteristic stripe wear described by Nevelos et al. [29]. Tissues were also obtained from six patients who were undergoing revisions for aseptic loosening of Charnley, cemented metal-on-polyethylene prostheses (Table 1). The average implantation time for these prostheses was 11years and 2 months, range 6–19 years.

wear scars on the weight-bearing surface. The extent of the wear was not however, quantified and therefore, it is not known if these prostheses were subject to abnormally high wear rates. This series had a high rate of failure compared to another study of uncemented Mittelmeier prostheses in which there were no reported incidences of osteolysis [29,30]. It is possible that osteolysis may be a problem in ceramic-on-ceramic bearings once a threshold level of debris in the phagocytosible size range has been generated. This study was undertaken to investigate further the true relationship between the tissue reaction and the nature of ceramic wear debris in uncemented aluminaon-alumina prostheses. The aims of this study were to investigate the tissues from uncemented Mittelmeier alumina ceramic-on-ceramic total hip replacements using histological methods and to isolate and characterise the ceramic wear debris using laser capture microdissection and electron microscopy.

2.2. Histology Tissue samples were taken from around the femoral component of each patient ðn ¼ 10Þ and the acetabular tissues of five of the patients from the ceramic-onceramic group. The tissues were fixed in 10% (v/v) phosphate buffered formaldehyde for a minimum of 2 weeks. Tissues were embedded in paraffin and 8 mm sections were taken and stained using acidified Harris’ haematoxylin and eosin. Tissue sections were analysed using light microscopy to determine histological reactions and also the location and content of alumina ceramic wear debris. The tissues were semi-quantitatively graded for several parameters: necrosis, necrobiosis, macrophages, giant cells, neutrophils, lymphocytes and wear debris. The different cell types were identified by looking at cell size, nuclear morphology and cytoplasmic colour. For necrosis, necrobiosis,

2. Materials and methods 2.1. Patients and tissues Tissues from around 10 non-cemented Mittelmeier alumina ceramic on ceramic THRs were obtained from patients undergoing revision surgery. All of these prostheses had been implanted by the same surgeon (AAN) operating at the same institution during the period 1985–1998. The patient details and reasons for revision are given in Table 1. None of the prostheses showed evidence of radiographic osteolysis. Reasons for revision were aseptic loosening in both components (Patient 1), stem loosening (2), a re-exploration after

Table 1 Details of patients, prostheses and reasons for revision Patient number

Sex

Age at surgery

Prosthesis type

Prosthesis survival (years)

Reason for revision

1 2 3 4 5 6 7 8 9 10

f m f m m f f f m f

42 51 47 (left) 49 (right) 19 68 16 29 40 54

Mittelmieir Mittelmieir Mittelmieir Mittelmieir Mittelmieir Mittelmieir Mittelmieir Mittelmieir Mittelmieir Mittelmieir

8 5 still in 6 (fem) 8 (acet) 1 2 6 7 7 7 (fem) 3 (acet)

Aseptic loosening stem and cup Aseptic loosening stem and cup Re-exploration for pain—no revision Aseptic loosening stem, aseptic loosening cup Aseptic loosening cup Aseptic loosening and dislocation cup Aseptic loosening cup Pain Aseptic loosening stem Aseptic loosening cup, pain

1 2 3 4 5 6

m m f m m f

70 60 76 69 78 71

Charnley Charnley Charnley Charnley Charnley Charnley

7 19 6 6 10 19

Aseptic Aseptic Aseptic Aseptic Aseptic Aseptic

loosening loosening loosening loosening loosening loosening

A. Hatton et al. / Biomaterials 23 (2002) 3429–3440

3432

Table 2 Grading system used in histological analysis of retrieved tissues Grade

Necrosisa

Necrobiosisa

Particlesb

Giant cellsc

Neutrophilsc

Lymphocytesc

Macrophagesc

1 2 3 4 5 6 7

o15% 15–30% 30–45% 45–60% above 60%

o15% 15–30% 30–45% 45–60% above 60%

o285b 286–570 571–855 856–1140 above1140

1–8 9–21 22–33 34–46 above 46

1–8 9–21 22–33 34–46 above 46

1–8 9–21 22–33 34–46 above 46

1–21 22–126 127–189 190–252 253–315 316–526 above 526

a

% of high power (  400) field of view. Number observed per high power field of view. c Numbers per mm2 of tissue. b

Table 3 Results of histological analysis of retrieved tissues Patient and tissue

Necrosis

Necrobiosis

Particles of ceramic

Giant cells

Neutrophils

Lymphocytes

Macrophages

1-fem 1-acet 2-fem 3-fem 3-acet 4-fem 4-acet 5-fem 6-fem 7-fem 8-fem 9-fem 10-fem 10-acet

3 2 1 1 2 5 3 1 1 1 2 1 2 2

2 1 2 1 2 2 1 3 1 2 2 2 2 2

5 4 2 5 4 1 2 1 4 2 4 2 2 2

1 1 1 1 1 1 1 1 1 1 1 1 1 1

2 1 2 2 2 3 1 1 2 2 2 2 2 2

2 1 1 1 3 1 4 2 2 1 2 2 2 2

6 5 6 7 6 2 5 6 5 5 5 5 6 5

C1-fem C2-fem C3-fem C4-fem C5-fem C6-fem

1 1 1 1 2 1

1 2 1 2 1 2

NA NA NA NA NA NA

2 2 1 3 3 2

1 1 1 1 1 1

1 1 1 1 1 1

7 7 5 7 7 7

Psm-1 Psm-2

1 1

1 1

NA NA

1 1

1 1

1 1

2 5

Tissues were graded using the scales illustrated in Table 2. fem, femoral tissue; acet, acetabular tissue; C, Charnley; Psm, primary synovial membrane; NA, not applicable.

particles, giant cells, neutrophils and lymphocytes, the tissues were graded on a scale of 1–5, 1 being the lowest and 5 representing the most severe (Table 2). For the grading of macrophage content the scale was increased from 1 to 7. At least 6 high power fields (HPFs) were graded from three sections per tissue sample. A HPF represented the total field of view at a magnification of 400 using the light microscope. Results were averaged and expressed as the mean grade per mm2 tissue (Table 3). Comparisons were made for each parameter in the femoral tissues with femoral tissues retrieved from 6 Charnley total hips that failed from aseptic loosening using the Mann–Whitney U-test [31].

2.3. Laser capture microdissection (LCM) Five to eight micrometer thick histological sections of paraffin embedded tissue were heat fixed to alcoholcleaned glass microscope slides and stained using haematoxylin and eosin. A thermoplastic polymer film was placed onto the tissue section to be analysed in the LCM device (Arcturus PixCell IIs). A near IR-gallium arsenide laser was pulsed onto the region of tissue to be extracted (beam intensity 75 mA, beam spot size 30 mm). Local heating of the thermoplastic polymer caused a rapid decrease in viscosity at the precise region targeted by the laser. The melted polymer entered the interstitial

A. Hatton et al. / Biomaterials 23 (2002) 3429–3440

gaps between the targeted cells and wear debris where it quickly hardened to form a strong bond between the targeted tissue and the polymer. Subsequent removal of the polymer layer allowed extraction of the targeted tissue due to a stronger bond formed between the tissue and the polymer than the tissue and the glass slide. This process did not damage the targeted tissue in any way. The extracted tissue was then processed and analysed using scanning electron microscopy (SEM) and transmission electron microscopy (TEM) for visualisation and characterisation, and energy dispersive X-ray analysis (EDX) for elemental identification. 2.4. Scanning electron microscopy The thermoplastic polymer film containing the extracted tissue was mounted on an aluminium SEM stub and carbon coated. SEM was then used to identify regions in the tissue that resembled alumina wear debris (CamScan series 4). Suspected regions of wear debris were analysed using EDX to determine if the particles were alumina (Link ISIS, Oxford Instruments). The images were digitised and the alumina wear debris was characterised using image analysis software (Image Pro Plus 4.0, Media Cybernetics). At least 150 particles were characterised from each patient tissue and particle length was determined.

3433

showed a granulomatous cellular reaction involving a dense macrophage infiltrate and the presence of giant cells and o30% necrosis/necrobiosis (Fig. 1a). The tissues from around the ceramic-on-ceramic prostheses all demonstrated the presence of particles which could be seen as agglomerates inside cells or in distinct channels in the tissues as illustrated in Figs. 1d and e. The tissues from the ceramic prostheses also showed the presence of neutrophils and lymphocytes (Fig. 1f) which were not evident in the tissues from the Charnley retrievals. The mean grades for each parameter in all the ceramic-on-ceramic tissues, the ceramic-on-ceramic femoral tissues, the ceramic-on-ceramic acetabular tissues and the Charnley femoral tissues were calculated and are presented in Fig. 2. The data for the ceramic-onceramic femoral tissues were compared to the data for the Charnley femoral tissues using the Mann–Whitney U-test. This revealed no significant difference between the tissues for the levels of necrosis, necrobiosis or lymphocytes. There were significantly more macrophages ðpo0:05Þ; and giant cells ðpo0:01Þ in the Charnley tissues and significantly more neutrophils ðpo0:01Þ in the ceramic-on-ceramic tissues. The femoral and acetabular tissues from the ceramic-on-ceramic retrievals were also compared but there were no statistical differences between them for any of the parameters measured.

2.5. Transmission electron microscopy The thermoplastic polymer film containing the extracted tissue was stained with osmium tetroxide, which stained the tissue black allowing easy sectioning. The film was embedded in Spurrs resin and 90 nm sections were cut using an ultra-microtome, mounted on copper TEM grids and carbon coated. The sections were examined using TEM (Philips CM20) and EDX (Link ISIS, Oxford Instruments). At least 150 particles were characterised for each patient tissue and the particles were measured by length.

3. Results 3.1. Histological analysis of retrieved tissues The results for the histological analysis of the retrieved tissues from the ceramic-on-ceramic prostheses together with the results from the Charnley tissues are presented in Table 3. The tissues from the femoral and acetabular regions of the ceramic-on-ceramic retrievals showed a mixed pathology with areas that had no obvious pathology, areas that were relatively rich in macrophages and over half of the tissues had in the region of 60% necrosis/necrobiosis (Fig. 1b and c). In comparison, the Charnley tissues

3.2. Laser capture microdissection and characterisation of the ceramic debris by electron microscopy Laser capture microdissection (LCM) was successfully carried out on all of the ten femoral and five acetabular tissues from the ceramic-on-ceramic retrievals. TEM images of the LCM extracted tissue were obtained for all the tissue samples, TEM was carried out at a magnification of  50–60,000, which revealed the presence of very small shards of alumina wear debris in the size range 5–90 nm, mean size 7SD of 24719 nm (Figs. 3a and 3b). Debris in this size range was present in all 15 tissue samples and the data for each tissue is presented in Table 4. The debris was identified as alumina by EDX analysis. These results indicated that the larger (micrometre) sized particles that were readily visible by light microscopy were few in number in comparison to the nanometre sized particles and were not present in the sections analysed. In order to capture images of the larger debris for analysis, LCM extracted tissues were further analysed by SEM at a lower magnification of  5–15,000. Images of only five of the tissues were successfully captured using this method. At this lower resolution, alumina wear debris in a larger size range was visible and polygonal shaped particles in the size range 0.046–3.2 mm were clearly visible (Figs. 3c and d). The mean size of the particles determined by

3434

A. Hatton et al. / Biomaterials 23 (2002) 3429–3440

Fig. 1. Histology of retrieved tissues. (a) Section of the femoral tissue from Charnley patient (6) showing the characteristic granulomatous response. Stained with heamatoxylin and eosin  400 magnification. M (macrophages). (b) Section of the femoral tissue from ceramic-on-ceramic Patient (9) showing an area of necrobiosis. Stained with heamatoxylin and eosin  1000 magnification. (c) Section of the femoral tissue from ceramic-onceramic Patient (3) showing the presence of ceramic wear particles in channels in the tissue associated with an infiltrate of macrophages. Stained with heamatoxylin and eosin  400 magnification. CD (ceramic debris) M (macrophages). (d) Section of the femoral tissue from ceramic-on ceramic Patient (6) showing the presence of agglomerated ceramic wear particles inside cells. Stained with heamatoxylin and eosin  1000 magnification CD (ceramic debris). (e) Section of the acetabular tissue from ceramic-on-ceramic Patient (10) showing an area of necrosis and a clump of ceramic particles. Stained with haematoxlyn and eosin  1000 magnification. CD (ceramic debris). (f) Section of the femoral tissue from ceramic-on-ceramic Patient (4) showing the presence of lymphocytes and neutrophils in an area of necrotic tissue. Stained with heamatoxylin and eosin  1000 magnification. N (neutrophils) L (lymphocyte).

SEM was 0.42870.325 mm. The data for each tissue is presented on Table 4. The data from all of the tissue samples analysed by TEM was combined and the frequency of particles

within different sizes is presented in Fig. 4. The mode of the particle size distribution was 15–20 nm and at the high resolution used, o10% of the particles were greater than 50 nm. The data from the five tissues analysed by

A. Hatton et al. / Biomaterials 23 (2002) 3429–3440

(d) Giant Cells

3.5

3.5

3

3

Mean Giant Cells grade

Mean necrosis grade

(a) Necrosis

2.5 2 1.5 1 0.5

2 1.5 1 0.5 0

All

Femoral

Acetabular

All

Charnley

(b) Necrobiosis

Femoral

Acetabular

Charnley

(e) Neutrophils 2.5 Mean Neutrophil grade

2.5 Mean Necrobiosis grade

**

2.5

0

2 1.5 1 0.5

2

**

1.5 1 0.5

0

0 All

Femoral

Acetabular

Charnley

All

(c) Lymphocytes

Femoral

Acetabular

(f) Macrophages

2.5 2 1.5 1 0.5 0

Charnley

*

8 Mean Macrophage grade

Mean Lymphocyte grade

3435

7 6 5 4 3 2 1 0

All

Femoral

Acetabular

Charnley

All

Femoral

Acetabular

Charnley

Fig. 2. Comparison of the histological grades for all ceramic-on-ceramic tissues (All), acetabular ceramic-on-ceramic tissues (acetabular), femoral ceramic-on-ceramic tissues (femoral) and Charnley femoral tissues (Charnley): (a) Necrosis (b) Necrobiosis (c) Lymphocytes (d) Giant cells (e) Neutrophils (f) Macrophages. Tissues were graded as described in Table 1 and the mean grade for each group calculated. The data for the ceramicon-ceramic femoral tissues and the Charnley femoral tissues were compared using the Mann–Whitney U test; * indicates significant difference ðpo0:05Þ; ** indicates significant difference ðpo0:01Þ:

SEM was also combined and the frequency of the particles in different size ranges determined (Fig. 5). By SEM, particles less than approximately 50 nm could not be resolved and the mode of the particle size distribution was 0.25–0.35 mm.

4. Discussion This study was undertaken to determine the nature of the cellular response in the periprosthetic tissues from 10 non-cemented Mittelmeier ceramic-on-ceramic total hip

replacements and to isolate and characterise the alumina wear debris present in the tissues using LCM and electron microscopy. The histological approach used was based upon an analysis and grading system adapted from Doorn et al. [32] in their studies of tissues from around metal-on-metal prostheses. The grading system was adapted by increasing the scale from 1–3, as used by Doorn, to 1–5 and, in the case of macrophage numbers, 1–7; since the numbers varied greatly. Histological analysis of the tissues revealed a mixed pathology throughout the series. A common feature was the presence of a macrophage infiltrate that was devoid of

3436

A. Hatton et al. / Biomaterials 23 (2002) 3429–3440

Fig. 3. Electron micrographs of ceramic wear debris. (a) TEM of ceramic wear debris from the femoral tissue of ceramic-on-ceramic Patient (9) showing the nanometer sized shards of debris: Ceramic debris . (b) TEM of the ceramic wear debris from the acetabular tissue of ceramic-onceramic Patient (4) showing a similar size distribution of nanometer size shards. The debris was confirmed to be alumina by EDX analysis: ceramic debris . (c) SEM of ceramic wear debris from the femoral tissue of ceramic-on-ceramic Patient (3) showing a large (B3 mm) particle associated with ; ceramic debris . (d) SEM of ceramic wear debris from the femoral tissue of ceramic-on-ceramic Patient polygonal granuels B0.5 mm: tissue (5) showing the agglomerated polygonal granules in the 0.5–2 mm size range: tissue ; ceramic debris .

giant cells. The tissues also contained areas of necrosis and necrobiosis and lymphocytes and neutrophils were common. Necrobiosis was determined following the description given by Doorn et al. [32]; the collagenous background remained but the tissues were acellular or

scattered with dissipating cell nuclei. As proposed by Doorn et al., it may be that the necrobiosis would have progressed into necrosis over time. The necrosis observed was of the conventional, caseous type and was observed in all 15 tissue samples. There were,

A. Hatton et al. / Biomaterials 23 (2002) 3429–3440

however, no statistically significant differences in levels of necrobiosis and necrosis between femoral tissues from the Mittelmeier and Charnley prostheses. This was probably due to the small sample size examined. Upon

observing the tissue sections, the percentages of the fields of view that were necrotic/necrobiotic in the tissues from the Mittelmeier prostheses tended to be higher. There were no significant differences between the levels of necrosis/necrobiosis in tissues from around the acetabulum and femoral stem of the Mittelmeier prostheses. A necrobiotic and necrotic response has been observed in tissues from retrieved first generation metal-on-metal total hip replacements. In the case of metal-on-metal hips, this particular tissue pathology has been suggested to be associated with the release of corrosion products of metal particles into the periprosthetic tissue, which are believed to be toxic to surrounding cells [32]. Lerouge et al. [26] reported no difference between cemented ceramic-on-ceramic and metal-polyethylene periprosthetic tissues with respect to cellularity This was in contrast to the differences in the cellular infiltrates in the uncemented Mittelmeier tissues compared to the Charnley tissues in the present study. The levels of macrophages were higher in the Charnley tissues and the levels of neutrophils and lymphocytes were higher in the Mittelmeier tissues. Neutrophils are often associated with necrotic tissue and so, this was in keeping with the

Table 4 Alumina ceramic particle sizes (mean length nm7SD) determined by TEM and SEM Patient

TEM Particle length (nm)7SD

1-fem 1-acet 2-fem 3-fem 3-acet 4-fem 4-acet 5-fem 5-acet 6-fem 7-fem 8-fem 9-fem 10-fem 10-acet

68719 23715 30710 22715 33712 31714 26712 21718 1978 25713 17711 22717 24710 17714 37723

SEM Particle length (nm)7SD 123760 7857540

124750 3567110

7567400

fem, femoral tissue; acet, acetabular tissue.

5 35 -4 0 40 -4 5 45 -5 0 50 -5 5 55 -6 0 60 -6 5 65 -7 0 70 -7 5 75 -8 0 80 -8 5 85 -9 0 90 -9 95 5 -1 00 >1 00

0

-3 30

5

-3 25

0

-2 20

5

-2 15

0-

10

10

20 18 16 14 12 10 8 6 4 2 0 -1

% Particles

3437

Particle Size (nm) Fig. 4. The frequency of the size distribution of ceramic wear particles observed by TEM. The data represents the combined data for all 15 tissues.

30 % Particles

25 20 15 10 5 >1.5

1.4-1.45

1.3-1.35

1.2-1.25

1.1-1.15

1.00-1.05

0.9-0.95

0.8-0.85

0.7-0.75

0.6-0.65

0.5-0.55

0.4-0.45

0.3-0.35

0.2-0.25

0.1-0.15

0

Particle Size (um) Fig. 5. The frequency of the size distribution of ceramic wear particles observed by SEM. The data represents the combined data for the five tissues as shown in Table 4.

3438

A. Hatton et al. / Biomaterials 23 (2002) 3429–3440

presence of dying tissue. When present in the Mittelmeier tissues, lymphocytes tended to be found in the perivascular areas and may have migrated out of the vessels in response to the inflammatory process. The numbers of lymphocytes were not so high as to indicate that they were concerned in a specific response. The levels of macrophages in Charnley tissues are known to be associated with the UHMWPE particle load [33] and the lower levels in the Mittelmeier tissues was probably due to a lower ceramic particle load in these tissues [29]. Giant cells, which are characteristic of the granulomatous response in tissues surrounding Charnley prostheses were not observed in tissues from the Mittelmeier retrievals. This was probably because of the small size range of the ceramic wear debris present in the tissues with no large particles present to initiate macrophage fusion. The Mittelmeier hip consists of a non-cemented, screw-in monolithic alumina cup articulating against an alumina head on a titanium (later cobalt chrome alloy) press fit stem (the Autophor prosthesis, CeramTecAG). Since the prostheses were non-cemented in design, the presence of necrosis and necrobiosis could not be attributed to either bone cement particles as postulated by Henssge et al. [25], or radio-opacifying agents such as zirconium dioxide or barium sulphate. In addition, the macrophage response, although not as intense as that observed with metal-polyethylene implants, was not attributable to particles of zirconia. Using light microscopy, it was only possible to characterise wear debris greater than approximately 0.5 mm in size, anything smaller could not be resolved. An abundance of alumina ceramic wear particles were observed in the tissues, as has been reported previously by Henssge et al. [25], Lerouge et al. [26] and Yoon et al. [28]. Ceramic particles were seen as translucent, yellowish polygonal granules frequently present in one of two formations in the tissues: clumps or channels. These could have been formed as a result of macrophage clustering, possibly around lymphatic drainage channels. The debris was isolated from the tissue sections using the technique of LCM. LCM is a method originally designed for use in cancer research in which pure populations of targeted cells are extracted from specific microscopic regions in tissue sections for subsequent analysis. The use of LCM in the present study represents a novel technique for the isolation and characterisation of wear debris from periprosthetic tissues and had the advantage that it preserved the wear debris in its original in vivo condition. Unfortunately, the LCM technique did not preserve the cellular features so it was not possible to determine the location of particles within the cells. Characterisation of the debris revealed a bi-modal size distribution that was unique to the method of visualisation. SEM revealed particles in the 0.05–3.2 mm size range and TEM

demonstrated particles in the 5–90 nm size range. This is the first description of nanometre sized ceramic wear particles in retrieval tissues. This is due to the resolution limitation of SEM in that particles less than approximately 50 nm cannot be visualised using the majority of microscopes. TEM, on the other hand, allows visualisation of much smaller particles down to a nanometre. However, the method of preparation of the sample for TEM involved using 100 nm sections and any particles larger than this could not be observed. The bi-modal size range of alumina ceramic wear debris overlaps the size ranges commonly observed with metal particles (10– 30 nm; [34]) and particles of UHMWPE (0.1–1000 mm; [35]). The size of the alumina wear debris observed by SEM of 0.05–3.2 mm, mean size 0.4470.33 mm was almost identical to that observed by Lerouge et al. [26]; 0.4470.25 mm and similar to that observed by Yoon et al. [28]; 0.13–7.21 mm. Whether the tissues analysed in these studies also contained the nanometre sized particles is not known since they did not use TEM to look for smaller particles. It is possible that the two size ranges of contributed to the mixed tissue pathology observed. The larger sized particles may have been phagocytosed by macrophages and stimulated further macrophage infiltration via the release of pro-inflammatory cytokines. It has been shown the micrometer sized alumina wear particles will stimulate macrophages to produce osteolytic cytokines in vitro, but only at particle volumes to cell number ratios above 100 mm3/cell [36]. Although there is no evidence that micrometer sized alumina ceramic particles are toxic to cells, particles in the nanometre size range have not yet been tested. It is possible that the very small ceramic particles may release ions that are toxic to cells and these might have contributed to the observed tissue necrosis/necrobiosis. On the other hand, the necrotic/ necrobiotic response which was also seen in some of the Charnley tissues may have been unrelated to wear particles generated at the articulating surfaces. There is a possibility that metal particles were liberated by fretting wear of the stem, and these could have led to cell death, however, metal particles were not observed in the majority of the tissues. CoCr particles were recovered using LCM in one tissue sample only. If the femoral stem was not mechanically stable, micromotion could have resulted in the disruption of the passivation layer surrounding the CoCr stem leading to ion release and subsequent cellular damage. Alternatively, physical damage to the tissues surrounding the implants as a result of micromotion may have contributed to the levels of necrosis/necrobiosis. It has recently been reported that in order to obtain clinically relevant wear rates, wear patterns (stripe wear) and wear mechanisms for ceramic-on-ceramic prostheses in vitro in hip joint simulators, it is necessary to introduce microseparation of the centers during the

A. Hatton et al. / Biomaterials 23 (2002) 3429–3440

swing phase and rim contact on heal strike [37]. It is likely that the wear debris generated by the rim contact occurring during heel strike is different (and larger) than that generated during standard kinematic conditions in which the head and cup remain in contact and a relief polishing wear of the surfaces produces smaller debris. This could explain the bi-modal distribution of the size of the wear debris observed in these tissues from the ceramic-on-ceramic retrievals, in which stripe wear of the head was a characteristic wear pattern [29]. The degree of microseparation and hence rim contact found clinically may depend on alignment, position, soft tissue tension and muscle forces. Thus these factors may also contribute to the levels of micrometer sized alumina ceramic debris generated in vivo and the potential for osteolysis in the longer term.

Acknowledgements We would like to thank the Medical Research Council for financial support, and Adrian Hick, Biochemistry & Molecular Biology, University of Leeds for technical assistance with the TEM work.

References [1] Harris WH, Schiller AL, Scholler JM, Freiberg RA, Scott R. Extensive localised bone resorption in the femur following total hip replacement. J Bone Jt Surg 1976;58A:612–8. [2] Willert HG, Semlitsch M. Reactions of the articular capsule to wear products of artificial joint prostheses. J Biomed Mater Res 1977;11:157–64. [3] Mirra JM, Marder RA, Amstutz HC. The pathology of failed total joint arthroplasty. Clin Orthop Relat Res 1982;170:175–83. [4] Jasty MJ, Floyd WE, Schiller AL, Goldring SR, Harris WH. Localised osteolysis in stable, non-septic total hip replacement. J Bone Jt Surg 1986;68A:912–9. [5] Schmalzried TP, Jasty M, Harris WH. Periprosthetic bone loss in total hip arthroplasty: polyethylene wear debris and the concept of the effective joint space. J Bone Jt Surg 1992;74A:849–63. [6] Santavirta S, Holkka V, Eskola A, Kontinen YT, Paavilainen T, Tallroth K, Amstutz HC, Campbell P, Kossovsky N, Clarke IC. Mechanism and clinical significance of wear debris-induced osteolysis. Clin Orthop 1992;276:7–18. [7] Murray DW, Rushton N. Macrophages stimulate bone resorption when they phagocytose particles. J Bone Jt Surg 1990;72B:988–92. [8] Jiranek WA, Machado M, Jasty M. Production of cytokines around loosened cemented acetabular components. Analysis with immunohistochemical techniques and in situ hybridisation. J Bone Jt Surg 1993;75A:863–79. [9] Revell PA, Al-Saffar N. Inflammatory mediators in aseptic loosening of prostheses. In: Downes S, Dabestani M, editors. Failure of joint replacement, a biological, mechanical or surgical problem? Institute of Orthopaedics, Middlesex, UK, 1994. p. 89–96. [10] Voronov I, Santerre JP, Hinek A, Callahan JW, Sandhu J, Boynton EL. Macrophage phagocytosis of polyethylene particles in vitro. J Biomed Mater Relat Res 1998;39:40–51.

3439

[11] Green TR, Fisher J, Stone MH, Wroblewski BM, Ingham E. Polyethylene particles of a ‘‘critical size’’ are necessary for the induction of cytokines by macrophages in vitro. Biomaterials 1998;19:2297–302. [12] Boutin P. Arthroplastie totale de la hanche par prothese en alumina fritte: etude experimentale et premieres applications cliniques. Rev Chir Orthop 1972;58:229–46. [13] Bohler M, Knahr K, Plenk Jr H, Walter H, Salzer M, Schreiber V. Long-term results of uncemented alumina acetabular implants. J Bone Jt Surg 1994;76B:53–9. [14] Nizard RS, Sedel L, Christel P, Meunier A, Soudry M, Witvoet J. Ten-year survivorship of cemented ceramic–ceramic total hip prostheses. Clin Orthop 1992;282:53–63. [15] Walter A. On the material and the tribology of alumina–alumina couplings for hip joint prostheses. Clin Orthop 1992;282:31–46. [16] Boutin P, Christel PS, Dorlot JM, Meunier A, deRoquancourt BD, Herman S, Sedel L, Witvoet J. The use of dense alumina– alumina combination in total hip replacement. J Biomed Mater Res 1988;22:1203–9. [17] Mittelmeier H, Heisel J. Sixteen-years experience with ceramichip prostheses. Clin Orthop 1992;282:64–74. [18] Willman G. Oxide ceramics for articulating components of total hip replacements. In: Sedel L, Rey C, editors. Bioceramics, Proceedings of the 10th International Symposium on Ceramics in Medicine, vol. 10. Paris, Elsevier Science Ltd, 1997. p. 123–26. [19] Dorlot JM, Christel P, Meunier A. Wear analysis of retrieved alumina ceramic heads and sockets of hip prostheses. J Biomed Mater Res 1989;23:299–310. [20] Mahoney OM, Dimon JH. Unsatisfactory results with a ceramic total hip prosthesis. J Bone Jt Surg 1990;72A:663–5. [21] Hoffinger SA, Keggi KJ, Zatorski LE. Primary ceramic hip replacement: a prospective study of 119 hips. Orthopaedics 1991;14:523–31. [22] Wirganowicz PZ, Thomas BJ. Massive osteolysis after ceramic on ceramic total hip arthroplasty. Clin Orthop 1997;338:100–4. [23] Nevelos JE, Ingham E, Doyle C, Fisher J, Nevelos AB. Analysis of retrieved alumina ceramic components from Mittelmeier toal hip prostheses. Biomaterials 1999;20:1833–40. [24] Kummer FJ, Stuchin SA, Frankel VH. Analysis of removed Autophor ceramic on ceramic components from Mittelmeier total hip arthroplasties. J Arthroplasty 1990;5:28–33. [25] Henssge EJ, Bos I, Willman G. Al2O3 against Al2O3 combination in hip endoprostheses. Histologic investigations with semiquantitative grading of revision and autopsy cases and abrasion measures. J Mater Sci: Mater Med 1994;5:657–61. [26] Lerouge S, Huk O, Yahia L, Witvoet J, Sedel L. Ceramic–ceramic and metal-polyethylene total hip replacements. J Bone Jt Surg 1996;79B:135–9. [27] Huo MH, Martin RP, Zatorski LE, Keggi KJ. Ceramic total hip hip replacement done without cement. Long-term follow-up study. Orthop Trans 1995;19:400–1. [28] Yoon TR, Rowe SM, Jung ST, Seon KJ, Maloney WJ. Osteolysis in association with a total hip arthroplasty with ceramic bearing surfaces. J Bone Jt Surg 1998;80A:1459–68. [29] Nevelos JE, Ingham E, Doyle C, Fisher J, Nevelos AB. Analysis of retrieved alumina ceramic components from Mittlmeier total hip prostheses. Biomaterials 1999;20:1833–40. [30] Roy BR, Nevelos AB, Ingham E, Shaw DL, Fisher J. Comparison of ceramic on ceramic to ceramic on polyethylene total hip replacement. Proceedings of the 13th International Smposium on Ceramics in Medicine. Bioceramics, vol 13, Bologna, Italy, 2000. [31] Sokal RR, Rohlf FJ. In: Sokal RR, Rohlf FJ, editors. Biometry. Non-parametric methods in lieu of single classification Anova. New York: W.H. Freeman Company, 1981. p. 443–5 [Chapter 13].

3440

A. Hatton et al. / Biomaterials 23 (2002) 3429–3440

[32] Doorn PF, Campbell PA, Worrall J, Benya PD, McKellop HA, Amstutz HC. Metal wear particle characterisation from metal on metal total hip replacements: transmission electron microscopy study of periprosthetic tissues and isolated particles. J Biomed Mater Res 1998;42:103–11. [33] Livermore J, Ilstrup D, Morrey B. Effect of femoral head size on the wear of the polyethylene acetabular component. J Bone Jt Surg 1990;72A:518–28. [34] Tipper JL, Firkins PJ, Ingham E, Stone MH, Farrar R, Fisher J. Quantitative analysis of the wear and wear debris from low and high carbon content cobalt chrome alloys used in metal on metal total hip replacements. J Mater Sci: Mater Med 1999;10:353–62.

[35] Tipper JL, Ingham E, Hailey JL, Besong AA, Wroblewski BM, Stone MH, Fisher J. Quantitative comparison of polyethylene wear debris, wear rate and head damage in retrieved Charnley hip prostheses. J Mater Sci: Mater Med 2000;11:117–24. [36] Hatton A, Ingham E, Matthews JB, Fisher J. Cytokine release by human peripheral blood macrophages after stimulation with alumina ceramic particles in vitro. Sixth World Congress of Biomaterials, vol. 1186, Hawaii, USA, 2000. [37] Nevelos J, Ingham E, Doyle C, Streicher R, Nevelos AB, Walter W, Fisher J. Microseparation of the centers of alumina–alumina artificial hip joints during simulator testing produces clinically relevant wear rates and patterns. J Arthroplasty 2000;15:793–5.