The pathology of joint replacement and tissue engineering

Anthony Freemont

Anthony Freemont BSc MD FRCP FRCPath is Professor of Osteoarticular

Pathology in the Faculty of Biology, Medicine and Health, University of

Manchester,

UK. Conflicts of interest: none declared.

Abstract

Joint replacement is very common and undertaken in most hospitals in one

form or another. Tissue engineering of connective tissues generally, and joints

in particular, is becoming more common. With increased usage, these

techniques generate iatrogenic morbidity. The diagnosis and exclusion of

iatrogenic disease is an increasingly important area of pathologists’ working

lives. This article discusses the disorders that can arise in association with joint

replacement and tissue engineering of joints and describes a relatively new

disease (Adverse Reaction to Metal Debris [ARMD]) the first of what may

become many new disorders associated with the new therapeutics covered in

this article.

Keywords infection; joint replacement; osteoarticular; pathology; tissue

engineering

Background

The pioneering research of John Charnley is generally considered to have

triggered the revolution that led to modern joint replacement surgery.1

1

Charnley’s work laid the practical foundations for the use of inorganic

engineered materials, such as metals and plastics, to restore joint function by

replacing malfunctioning articular surfaces. These new techniques were not

without their problems. More recently two new branches of connective tissue

medicine, tissue engineering and regenerative medicine, have focused minds

on the possibility of replacing damaged tissue with new.

As with all medical innovations, problems are recognized that with time

change the practice of pathology. The problems associated with the pathology

of joint tissue replacement and regeneration can be considered under four

headings:

Replacement of articular surfaces

Replacement of non-articulating intra-articular structures

Intra-articular injectable agents

Tissue engineering/regeneration

Replacement of articular surfaces

Background

The purpose of joint replacement (arthroplasty) is to improve the limited

movement and reduce pain caused by damage to articular surfaces. This is

achieved by either replacing diseased articulating surfaces with artificial ones

(e.g. hip replacements) or by replacing the entire joint with a prosthesis (e.g.

finger joints).

Artificial articulating surfaces are usually constructed of metals, plastics or

ceramics, or combinations of these materials. Some are used solely for

resurfacing the joint, whilst others replace the articulating surfaces and

2

adjacent bone. Often they are held in place using cements consisting most

commonly of acrylic resins. Recently there has been a trend towards coating

the stems of implants with materials such as hydroxyapatite that are said to

promote bone growth into the prosthesis, with the hope of preventing

loosening, a problem that requires revision and a new prosthesis.

Typically, both sides of a joint wear out together and arthroplasty surgery

replaces both articular surfaces. If this is the case, the two opposing prosthetic

surfaces may consist of the same or different materials. Traditionally the

articulations were metal on plastic. They are highly successful but in a

proportion of patients the differences in physical properties of the two

surfaces leads to wear and formation of significant numbers of wear particles,

which initiate a macrophage and giant cell reaction leading to bone loss and

implant loosening. Advances in materials design has resulted in manufacture of

articulations with reduced wear particle production. Of these the combination

of metal articulating on crosslinked polyethylene (PE), metal on metal, and

ceramic-on-ceramic articulations have all demonstrated lower rates of in vivo

wear particle generation than the original metal-on-plastic ones.

The properties of wear particles derived from inorganic materials used in

joint replacement

Cross-linked polyethylene: ultrahigh molecular weight polyethylene

(UHMWPE) is the most commonly used material for acetabular and tibial

prostheses.2 It has good biocompatibility but it does undergo wear. The wear

particles vary in size from a few to several hundred micrometres long. The

particles are intensely birefringent and remain in tissue after conventional

processing (Figure 1a).

3

Metals: the most common metals used for articulations are low carbon

stainless steels and cobalt-chrome. Metals may be scratched, particularly

during implantation, which initiates wear.

Many stems of knee and hip implants are made of alloys of titanium, vanadium

and aluminium. This material can also wear. There is evidence that titanium

can preferentially elute from the alloy. Although inert, titanium can

disseminate widely throughout the body.3

Metal wear particles show up as tiny dark flecks, usually within macrophages in

tissue sections of synovium (Figure 1b) or the soft tissue membranes that

develop between implants and surrounding bone.

Ceramics: these (e.g. alumina [Al2O3]) are harder than the equivalent metals

and can produce very smooth surfaces that are extremely unlikely to wear.

They are, however, more brittle and more likely to fracture (in practice this

occurs in <1:400 cases). The fragments produced by wear and fracture have no

specific features, being a granular material that resembles hydroxyapatite.4

Cement: most commonly cements are acrylic resins, such as

polymethylmethacrylate (PMMA). PMMA forms by hyperthermic

polymerization of a liquid monomer. During polymerization the cement sticks

to the implant’s stem and, because it is introduced into the marrow as a liquid

that can permeate between bone trabeculae, when it polymerizes, it binds

itself into the trabecular network, fixing the prosthesis tightly to the adjacent

bone. The cement is doped with particles of metal such as barium to render

the cement radiodense. Under the microscope cement appears as a large,

clear, rounded nodule surrounded by macrophages and multinucleate giant

cells and containing dark refractile rounded granules of the doping metal

(Figure 1c).

4

Hinge materials: some small joints (eg. Finger joints) are replaced by plastic

hinges. The hinge materials are usually made from the silicone-based polymer,

silastic. With frequent use silastic may break, releasing particulate material

which can spread to other areas of the body via the lymphatics. This material is

refractile, granular and usually intracellular within macrophages or

synoviocytes (Figure 1d).

Complications of joint replacement surgery

Arthroplasty is a very successful procedure, however, there is a 10 year failure

rate of about 2%, the major causes of which are:

Dislocation. This is a complication that is related to the success of the

surgical procedure itself and the laxity of supporting tissues.

Damage to the prosthesis. With modern materials these events are

rare.5

Aseptic loosening

Infection. This is relatively uncommon, important and dealt with later.

Aseptic loosening

Background: aseptic loosening is the most common complication of joint

replacement surgery accounting for between half and three quarters of all

revisions. It is often associated with periprosthetic bone loss/osteolysis, which

may be rapid and make revision surgery difficult.

The majority of prostheses are attached to a stem that is pushed into the shaft

of the bone. Around the stem is a potential space in continuity with the joint

cavity. This space is often filled by cement and/or host fibrous tissue and bone.

Under load an imperfectly fixed stem may work loose opening a gap between

the implant into which wear particles generated inside the joint can be forced.

5

Here they initiate a macrophage and osteoclast reaction that leads to bone loss

and loosening.

Of the potential wear particles polyethylene is the most bioactive. The particles

vary in size but the submicron particles induce a greater inflammatory

response in vitro than do larger particles.6 The cellular response is also

dependent on the number and shape of the particles; elongated particles

generating a more severe inflammatory reaction than globular ones.7

Metal wear particles are smaller than those from polyethylene. Despite their

size, they may be so numerous that they stain the capsule black (“metallosis”).

In histological sections metal particles are usually seen within macrophages of

synovium and pseudomembranes around prosthetic stems. They may be

associated with necrosis. Particles of metal may also migrate through the bone

marrow and be found in regional lymph nodes. In the synovial fluid they can

cause direct (third body) wear of the surface of implants.

Ceramic materials tend to have better biocompatibility than metal alloys8 but

the size, shape, number, distribution and reactivity of the respective wear

particles has not been fully determined. This said they seem to be less

inflammogenic than either metal or plastic particles. As a generalization,

ceramic wear particles are granular and non-birefringent measuring 0.5-20 μm

in diameter. Like metal particles they can form aggregates. They tend to be

found in macrophages but can elicit a foreign body giant cell response. They

are also found extracellularly, where they appear brown in colour, but lack the

refractility of haemosiderin deposits.

In addition to wear particles from articular surfaces, particles of cement may

enter the synovial fluid and joint tissues. Polymethylmethacrylate (PMMA) is

brittle and particles and fragments of diameter 30-100 μm break off under load

6

forming debris with properties similar to those of surface wear debris. PMMA

itself is dissolved during tissue processing and under the microscope holes

formed from ghosts of dissolved PMMA fragments can be seen containing dark

metal particles. Because of their relatively large size the PMMA particles

initiate a foreign body-type giant cell reaction.

Silastic wear particles are 10-100 μm in size. They are crenulated and

birefringent, usually eliciting a brisk macrophage and multinucleated giant cell

reaction. They can cause osteolysis, painful synovitis, and lymphadenopathy in

regional lymph nodes.9

Wear particles, either following phagocytosis or by activation of cell

surfaces, can change the function of different cell types (particularly

macrophages, fibroblasts, osteoblasts and osteoclasts) within the

bone/marrow around the stem of the prosthesis. 10 Macrophages activated by

phagocytosing wear particles in the synovium and pseudomembrane produce

numerous cytokines, growth factors, chemokines and other mediators (notably

IL-1ß, IL-6, M-CSF, nitric oxide, metalloproteinases)11 that stimulate changes in

local cell and matrix biology. In particular they can give rise to a painful

inflammatory synovitis and osteoclast-mediated periprosthetic osteolysis and

loosening. The osteolysis may be worsened by direct inhibition of osteoblastic

activity by the wear particles themselves, particularly metal particles.

In addition to osteolysis being mediated by a foreign bodytype response to

wear particles, in some patients the presence of a granulomatous response

suggests a type IV hypersensitivity reaction.

Pathological findings: in patients with a painful joint replacement the clinician

is usually seeking guidance on the causes of pain and/or swelling. In particular

7

(s)he is asking the pathologist to exclude infection or ARMD (see below). One

or more of three tissues is/are commonly sampled:

Synovium

Pseudomembrane

Synovial fluid

Synovium - After the joint capsule is closed around an implant, a crude

synovium forms to line the new joint space. The new synovium has the same

functions and is prone to the same disease processes as synovium anywhere.

Wear particle debris enters the synovium eliciting a macrophage/giant cell

response. The macrophages are often numerous with diffuse brown coloured

cytoplasm, or contain distinct metal/plastic/cement particles. The synovial

subintima is often fibrotic, the synovial surface may become replaced by

granulation tissue, and necrosis may be present. Only rarely is there a

significant lymphocytic infiltrate and polymorphs should not feature unless

infection is present. Necrosis associated with macrophages and giant cells may

be confused with necrotizing granulomatous inflammation especially in small

biopsies. In this setting the recognition of particulate material using direct

vision, polarizing microscopy or special stains (e.g. Oil Red O for polythene)

then becomes paramount.

Although usually carried out for osteoarthritis, joint replacement is sometimes

undertaken as part of the management of inflammatory arthropathies,

particularly rheumatoid arthritis. In this setting the “new” synovium can take

on an identical appearance to that seen in the primary disease.

Pseudomembrane -the pseudomembrane between a loosening implant and

native bone shows the same fibrosis, focal necrosis and macrophage and giant

cell response to that seen in the synovium. The surface of the

8

pseudomembrane abutting the implant may even develop a synoviocyte-like

layer of palisaded “synoviocytes”.11

Synovial fluid - the synovial fluid in joints with implants is usually non-

inflammatory (i.e viscid and with a cell count of <500 cells/mm3). The presence

of >1700 cells/mm3 should alert one to the possibility of infection (see below).

Wear debris, haemosiderin and macrophages are present in the fluid. The

presence of wear debris indicates wear and not necessarily loosening. Indeed

there is no cytological test for diagnosing loosening.

Special pathology of metal-on-metal articulations Adverse Reaction to Metal

Debris (ARMD)

Background: As a major problem with joint prostheses is loosening of the

stem, attempts have been made to design “stemless” prostheses. As the

earliest and still most frequent joint replacement, loosening of femoral

stemmed prostheses has become a common problem, not because the

technique is flawed but simply because of the huge numbers of procedures

being performed. The problem is particularly serious in the elderly in whom

bone loss through osteoporosis makes revision difficult.

Resurfacing of the damaged articular surface of the femoral head seemed a

perfect solution as it leaves the femoral shaft unaffected should a revision to a

stemmed prosthesis be necessary. Over the past 15 years cobalt-chrome has

become the material of choice both for the femoral head and also the

acetabulum, creating a metal-on-metal articulation. These work extremely

well, however in the last 10 years a complication (ARMD) has become apparent

in a proportion of patients that has been shown to manifest in a number of

different ways.

Pathology: The pathology of ARMD is characterized by (Figure 2a):

9

The formation of pseudotumours; para-articular, often cystic, swellings

that involve fat, fibrous tissue ad muscle

Very extensive tissue necrosis of the synovium and subintima, which

may be very extensive and extend into and through the joint capsule

into surrounding tissues.

Prominent lymphoid cell (usually lymphocyte but occasionally plasma

cell) aggregates in fibrous tissue often, but not exclusively, deep to areas

of necrosis 12. This was originally called “aseptic lymphocytic vasculitis-

associated lesion” or ALVAL. There is no vasculitis.

A band of macrophages within the necrotic tissue. The macrophages

sometimes contain obvious metal particles but in a significant

proportion of cases they do not, instead having a brown tinge to the

cytoplasm or appearing “foamy”. In places macrophages show evidence

of cytotoxicity with histologically detectable cell membrane damage.

Pathogenesis: These patients often have raised levels of cobalt and chromium

in their blood and synoval fluid and these are believed to be the clue to the

pathogenesis of this disorder. It is now accepted that the probable mechanism

of necrosis and ALVAL is stimulation of the adaptive immune response by the

presence of nano-sized wear particles of cobalt and/or chrome. Metallosis as

one of the manifestation of ARMD, is caused by accumulation of larger metal

particles but may also be accompanied by necrosis and ALVAL.

In a recent review 13 Athanasou discusses all the elements of aseptic loosening

and in particular discusses the singular pathology of metal on metal hip

prostheses.

Similar features may be seen in non-metal on metal articulating prostheses

with modular cobalt chrome stems where it has been postulated that the

10

driver of the pathological changes is believed to be wear at the interface

between the modular elements.

Infection

Background: As in all surgery there is a risk of introducing infection at

arthroplasty. The infection rate varies from 1% to 5%. The incidence rises in

revision arthroplasty and in patients with compromized immune systems such

as those with rheumatoid arthritis or diabetes. Factors such as the

characteristics of the operating theatre, the quality of the host bone and soft

tissue, and the complexity and length of the operation all contribute to the

infection risk. Prosthetic material/particulate debris contributes to infection

because bacteria, particularly Staphylococcus epidermidis, readily bind to most

arthroplasty materials. The ability of many bacteria to remain attached to a

surface is enhanced by their production of a biofilm consisting of saccharides,

proteins and nucleic acids. Once established, biofilms attract other bacteria

with less specific adhesion properties to the prosthetic surface, thus forming a

colony of mixed bacterial species.

Frank septic arthritis is rare following arthroplasty, but the presence of the

biofilm encourages low grade, insidious infection which nonetheless leads to

implant loosening indistinguishable from that caused by aseptic loosening. This

condition of periprosthetic infection is also known as septic loosening. It is

particularly poorly responsive to systemic and local antibiotic treatment.

Failure to recognize that osteolysis is infection-related can have disastrous

results if the patient undergoes revision arthroplasty into an infected bone, the

new implant rapidly failing.

If infection is suspected or proven concerted action can eradicate the infection.

High dose long-term antibiotics can be successful, but once loosening has

11

started antibiotic treatment is inneffective and a two stage revision is

undertaken in which the infected implant is removed and replaced with a

cement spacer doped with antibiotic. Usually this is coupled with prolonged

oral antibiotic therapy and only once biomarkers of infection have returned to

normal is the revision completed.

Diagnosis of septic loosening has therefore become a diagnostic priority.

Diagnosing the infected prosthesis: When there is acute septic arthritis or

acute osteomyelitis associated with an arthroplasty, the diagnosis is relatively

straightforward, as there is necrosis and large numbers of polymorphs within

the tissue; together with a high nucleated cell count and high proportion of

polymorphs in the synovial fluid. The causal agent is usually a Gram positive

coccus which is identified by Gram staining or culture.

Low grade infection of the type that leads to periprosthetic infection, is much

more difficult to diagnose, the key feature being the presence of polymorphs

which are never a significant component of inflammatory infiltrates in aseptic

loosening or ARMD.

In most cases, the diagnosis of an infected total joint replacement depends on

a combination of clinical features, radiographic findings, and laboratory test

results, but in the low grade infections that characterize periprosthetic

infection, laboratory techniques are key. A peripheral blood leukocytosis,

raised ESR and CRP might suggest infection, but these tests have

poorsensitivity and specificity, and imaging adds little, nor does culture of

synovial fluid or joint tissue. In this setting tissue biopsy and synovial fluid

analysis prove to be pivotal.

Tissue biopsy: there are a number of studies that have shown the diagnostic

and prognostic significance of identification of polymorphs in the superficial

12

synovium. Arguably the most significant of these was in 1995 when Athanasou

and colleagues in Oxford14 showed that an average of one or more neutrophil

polymorphs per high powered microscope field across a frozen section of

synovium taken intraoperatively gave a diagnostic sensitivity for low grade

infection of 90% and specificity of 96%.

Using the criterion of five polymorphs per high powered field in at least five

fields as evidence of active infection within periprosthetic tissue from the

bone-cement interface or the pseudocapsule, Feldman et al15 achieved a

sensitivity of 100% and a specificity of 96%. As a rule of thumb, >10

polymorphs per high powered field (Figure 2b) over five fields is predictive of

infection, whilst 5-9 is suspicious and <5 has no diagnostic significance. Even

though the organism causing infection is most commonly a Staphylococcus,

Gram staining intra-operative smears or later on tissue sections has a very low

diagnostic yield and adds very little, if anything, to the overall diagnostic

process.

Although lacking the immediacy of intra-operative frozen sections the same

criteria can used for diagnosing infection in conventionally processed tissue,

thus pre-operative open, arthroscopic or blind needle biopsy of synovium can

be used for planning definitive surgery.

Synovial fluid: Synovial fluid analysis has recently been shown to be a useful

investigation in the diagnosis of low grade, periprosthetic infection, and one

that can be used before surgery, allowing planned revision.16 In a prospective

study a synovial fluid leukocyte count >1700/mm3 had a sensitivity of 94% and

specificity of 88% for diagnosing prosthetic joint infection; and if neutrophils

accounted for >65% of the nucleated cells the sensitivity and specificity rose to

97% and 98% respectively.

13

In terms of assessing infection prior to revision for knee arthroplasty this

positions synovial fluid analysis at the forefront of diagnostic tests.

Replacement of intra-articular structures other than articulating surfaces

Attempts have been made to replace intra-articular ligaments such as the

cruciate ligaments and fibrocartilagenous structures such as menisci with

inorganic materials.

Cruciate ligament replacement are usually constructed of loosely twisted or

braided fibrillary materials including plastics, carbon fibre and even

polymerized naturally occurring molecules such as lactic acid. The articifical

ligament has innate strength but the hoped for ingrowth of fibroblasts and

replacement by organized fibroelastic tissue is rarely achieved and the

presence of a macrophage and giant cell response to degrading biomaterial is

the norm.

The synovial fluid in such cases is invariably of low cell count, but fragments of

fibrillar prosthetic material might be seen.

Intra-articular injectable agents

Intra-articular injection is frequently used for delivering therapy into diseased

joints. Examples include steroids and hyaluronans therapeutically and local

anaesthetics prior to arthroscopic surgery.

Steroids

Steroids used to treat inflammation and pain, are usually injected as relatively

insoluble, crystalline preparations which dissolve slowly to give a longer lasting

effect. The carrier medium can very rarely induce a hypersensitivity reaction

with eosinophils and mast cells in the synovial fluid. The steroid crystaloids

14

which can remain in the joint for several weeks, can be mistaken for

pathogenic crystals on subsequent synovial fluid examination. Steroids reduce

inflammation within the synovium and the number of polymorphs within the

synovial fluid.

Hyaluronans

Hyaluronans are being used increasingly for treating noninflammatory

arthropathies, such as trauma and osteoarthritis. They have some beneficial

effects through: decreasing pain nerve sensitivity; enhancing chondrocytic

proteoglycan synthesis; and reducing effects of proinflammatory mediators

and matrix metalloproteinases.

Rarely, patients develop an acute or subacute reaction with pain and swelling

of the joint, which usually settles spontaneously within a day or two. Synovial

fluid samples aspirated at the time have cell counts of 2000-100,000

cells/mm3. This acute reaction can mimic septic arthritis.

A second uncommon reaction has been noted recently in which aggregates of

hyaluronans are absorbed into the synovium where they induce a

granulomatous reaction. In H&E stained sections they appear as naked

granulomata with a central grey amorphous centre surrounded by

mononuclear and multinucleate macrophages 17 (Figure 3a).

Local anaesthetic agents

It is common practice to perform arthroscopies under a form of local

anaesthesia in which a local anaesthetic agent is instilled into the joint. There is

no pathology associated with this as such but some recent evidence has

emerged implicating some local anaesthetic agents in chondrotoxicity. The

significance, if any, of this in Man is unknown.18

15

Tissue engineering/regeneration

Background

Connective tissues have been the focus of much basic and translational tissue

engineering and regenerative medicine research. The major area that has

impacted on the pathologist relates to cartilage replacement/regeneration.

Cartilage resurfacing

The ability of articular cartilage to heal spontaneously is limited, and injury

predisposes to osteoarthritis. Whilst joint replacement is very successful in

relieving symptoms, the life of an implant may be short and revision has a

relatively high morbidity, with the revised implant having, on average, a

shorter life than the original. This is particularly the case in young implant

recipients which increases the reluctance of surgeons to perform joint

replacement surgery in the young and middle aged.

One answer is joint resurfacing with inorganic materials such as metals,

another is to restore the cartilage. In young people the initial, often trauma

induced, damage to cartilage is a relatively small, focal defect. A variety of

techniques have been tried to promote repair by stimulating natural stem cells

most of which involve damaging the subchondral bone and exposing marrow

in the floor of the defect. The mesenchymal stem cells within the marrow are

stimulated to form new bone and cartilaginous tissues within the defect. This

results in the formation of a repair tissue that whilst it may consist in large part

of fibrocartilage which, whilst lacking many of the physical properties of

articular cartilage, successfully achieves the primary goals of pain relief and

delay in the onset of osteoarthritis.

Attempts have also been made to implant tissue engineered cartilage grown

outside the body, but biointegration of such a construct is a real problem.

16

In vivo tissue engineering has been attempted with rather more success. About

30 years ago successful in vivo engineering of cartilage from autologous

chondrocytes (autologous chondrocyte implantation - ACI) was reported.19

The initial approach to ACI involved expanding a population of chondrocytes

harvested from relatively normal but non-weight bearing cartilage in the

affected joint in a laboratory. After population expansion, the cells are placed

in a debrided defect in the cartilage and held in place by sowing a sheet of

material (often periosteum) across the top of the defect.20 At subsequent

arthroscopy, the regenerating tissue has a rigid, elastic consistency and

sometimes grows over the surrounding cartilage. Histologically it usually has

the appearance of fibrocartilage, but sometimes true hyaline cartilage forms.

This has gained relatively wide usage and more recently similar effects have

been documented using autologous mesenchymal stem cells.

The technique cannot be used on large or convex defects. In these settings

cylinders of autologous bone and cartilage have been transplanted into sites of

cartilage loss either to initiate repair or to act as a scaffold for ACI.

Most recently hybrid technologies involving combining damaging subchondral

bone with the use of autologous chondrocytes or mesenchymal stem cells in

biomaterials (mainly gels such as alginate and chitosan), and transplanted

osteochondral cylinders have been trialled. Both autologous cell and tissue

transplantation in all these settings have proved successful in forming cartilage.

The histopathologist is occasionally asked to assess the quality of

biointegration (Figure 3b), particularly at the interface with native cartilage.

Metachromatic stains often delineate engineered cartilage from host cartilage

and the junction can also be identified using polarizing microscopy, the

organization and birefringence of the newly formed cartilage/fibrocartilage

being different from that of the host cartilage. Assessment of the extent of

17

biointegration has clinical value as poor biointegration increases the chances of

failure of the engineered cartilage.

REFERENCES

1. Wroblewski BM. Professor Sir John Charnley (1911-1982). Rheumatology (Oxford) 2002; 41: 824-5.

2. Kurtz SM, Muratoglu OK, Evans M, Edidin AA. Advances in the processing, sterilization, and crosslinking of ultra-high molecular weight polyethylene or total joint arthroplasty. Biomaterials 1999; 20: 1659-88.

3. Urban RM, Jacobs JJ, Tomlinson MJ, Gavrilovic J, Black J, Peoc’h M. Dissemination of wear particles to the liver, spleen, and abdominal lymph nodes of patients with hip or knee replacement. J Bone Joint Surg Am 2000; 82: 457-76.

4. Hannouche D, Hamadouche M, Nizard R, Bizot P, Meunier A, Sedel L. Ceramics in total hip replacement. Clin Orthop Relat Res 2005; 430: 62-71.

5. Schmalzried TP, Shepherd EF, Dorey FJ, et al. The John Charnley Award. Wear is a function of use, not time. Clin Orthop Relat Res. 2000; 381: 36-46.

6. Jacobs JJ, Hallab NJ, Urban RM, Wimmer MA. Wear particles. J Bone Joint Surg Am 2006; 88(suppl 2): 99-102.

7. Yang SY, Ren W, Park Y, et al. Diverse cellular and apoptotic responses to variant shapes of UHMWPE particles in a murine model of inflammation. Biomaterials 2002; 23: 3535-43.

8. Campbell P, Shen FW, McKellop H. Biologic and tribologic considerations of alternative bearing surfaces. Clin Orthop Relat Res 2004;418: 98-111.

9. Yamashina M, Moatamed F. Peri-articular reactions to microscopic erosion of silicone-polymer implants. Light and scanning electron microscopic studies with energy-dispersive X-ray analysis. Am J Surg Pathol 1985; 9: 215-9.

18

10. Purdue PE, Koulouvaris P, Potter HG, Nestor BJ, Sculco TP. The cellular and molecular biology of periprosthetic osteolysis. Clin Orthop Relat Res 2007; 454: 251-61.

11. Athanasou NA. The pathology of joint replacement. Curr Diagn Pathol 2002; 8: 26-32.

12. Davies AP, Willert HG, Campbell PA, Learmonth ID, Case CP. An unusual lymphocytic perivascular infiltration in tissues around contemporary metal-on-metal joint replacements. J Bone Joint Surg Am 2005; 87: 18-27.

13. Athanasou NA The pathobiology and pathology of aseptic implant failure.Bone Joint Res. 2016;5:162-8.

14. Athanasou NA, Pandey R, de Steiger R, Crook D, Smith PM. Diagnosis of infection by frozen section during revision arthroplasty. J Bone Joint Surg Br 1995; 77: 28-33.

15. Feldman DS, Lonner JH, Desai P, Zuckerman JD. The role of intraoperative frozen sections in revision total joint arthroplasty. J Bone Joint Surg Am 1995; 77: 1807-13.

16. Trampuz A, Hanssen AD, Osmon DR, Mandrekar J, Steckelberg JM, Patel R. Synovial fluid leukocyte count and differential for the diagnosis of prosthetic knee infection. Am J Med 2004; 117: 556-62.

17. Michou L, Job-Deslandre C, de Pinieux G, Kahan A. Granulomatous synovitis after intraarticular Hylan GF-20. A report of two cases. Joint Bone Spine 2004; 71: 438-40.

18. Piper SL, Kim HT. Comparison of ropivacaine and bupivacaine toxicity in human articular chondrocytes. J Bone Joint Surg Am 2008; 90: 986-91.

19. Grande DA, Singh IJ, Pugh J. Healing of experimentally produced lesions in articular cartilage following chondrocyte transplantation. Anat Rec 1987; 218: 142-8.

19

20. Brittberg M, Lindahl A, Nilsson A, Ohlsson C, Isaksson O, Peterson L. Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med 1994; 331: 889e95.

Practice points

In assessing diseased joints and those with an implant it is an imperative to exclude infection

Infection is best diagnosed by synovial fluid analysis which must include: nucleated cell count; polymorphs expressed as a proportion of all nucleated cells; ragocytes expressed as a proportion of all nucleated cells; and the presence of haematoidin crystals

In metal-on-metal hip implants ARMD must be confirmed or excluded

Figure legends

Figure 1 Examples of different materials seen in biopsies of synovium: (a) HMWPE. (b) Cement. (c) Silastic. (d) Metal particles.

Figure 2 (a) Typical histological appearance of ALVAL with necrosis (N), Band of Macrophages (M), Lymphocyte aggregates (L). (b) Infective arthritis with polymorphs in the necrotic tissue and fibrin at the surface of the synovium.

Figure 3 (a) Naked granulomata consisting of giant cells surrounding hyaluronan in the synovium of a professional footballer treated for knee pain with hyaluronan injections. (b) Needle biopsy of the interface region between ACI derived (A) and native residual articular cartilage (N) showing good lateral biointegration.

20

Figure 1

Figure 2

21

Figure 3

22