Knee injuries and their consequences : Knee injuries and their consequences – the impact of impact
Swärd, Per
2014
Link to publication
Orthopaedics (Lund)]. Department of Orthopaedics, Lund University.
Total number of authors:
1
Citation for published version (APA):
Swärd, P. (2014). Knee injuries and their consequences – the impact of impact. [Doctoral Thesis (compilation),
General rights
Unless other specific re-use rights are stated the following general rights apply:
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors
and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the
legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study
or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal
Read more about Creative commons licenses: https://creativecommons.org/licenses/
Take down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove
access to the work immediately and investigate your claim.
LUND UNIVERSITY
PO Box 117
221 00 Lund
+46 46-222 00 00
Knee injuries and their consequences – the impact of impact
Per Swärd
DOCTORAL DISSERTATION
by due permission of the Faculty of Medicine, Lund University, Sweden
To be defended at Segerfalksalen, BMC, Sölvegatan 17, Lund, on 23 May 2014 at
1 pm
Faculty opponent
Professor Lars Engebretsen
1
Knee injuries and their consequences – the impact of impact
Per Swärd
3
Cover illustration: Mouse knee joint. Bone, hyaline cartilage and meniscus
visualised by toluidine blue staining (40 x magnification). Published with the kind
permission of Dr. André Struglics.
© Per Swärd
Lund University, Faculty of Medicine
Doctoral Dissertation Series 2014:60
ISBN 978-91-87651-86-1
ISSN 1652-8220
Printed in Sweden by Media-Tryck, Lund University
Lund 2014
En del av Förpacknings- och
Tidningsinsamlingen (FTI)
4
5
To Catrin
6
7
Contents
Abstract 9
Populärvetenskaplig sammanfattning 11
Introduction 13
Preface 13
List of studies 14
Abbreviations 15
Background 17
The knee joint 17
The anterior cruciate ligament 17
Anterior cruciate ligament tears 19
Anterior cruciate ligament injury mechanism 20
Knee injury panorama 22
Osteoarthritis 25
Knee osteoarthritis 25
Post-traumatic knee osteoarthritis 26
Cartilage 27
Synovium 31
The meniscus 32
Bone 32
Inflammation 34
Objectives 39
Specific objectives 39
Subjects 41
Methods 47
Radiography of the knee (Studies I & II)
47
MRI of the knee (Studies I & IV)
Alcian blue precipitation (Study III)
Electrochemiluminescence (Study III)
ELISA (Study IV)
Western blot (Study V)
Statistical methods
Methodological considerations
49
49
50
50
50
51
53
Subjects 53
Imaging 54
Results
55
Radiological characteristics of post- and non-traumatic knee
osteoarthritis (Study 1)
Alignment and the risk of post-traumatic osteoarthritis (Study II)
Cartilage and bone markers and inflammatory cytokines in the acute
phase of knee injury (Studies III & IV)
Osteochondral fractures and joint inflammation (Study IV)
Protease activity in bovine cartilage explants co-incubated with joint
capsule and/or mechanically injured (Study V)
55
56
57
58
60
Conclusions 63
Discussion and future perspectives
65
Post-traumatic and non-traumatic knee osteoarthritis
The knee injury and the impact
Mechanical impact and its role in osteoarthritis initiation after ACL
injury
Immunity vs biomechanics in driving the OA process after anterior
cruciate ligament rupture
Concluding remarks
65
66
67
68
73
Tack 75
References 77
8
Abstract
Anterior cruciate ligament (ACL) injuries are common, severe knee injuries that
result in a high risk of developing knee osteoarthritis (OA) in the affected
individuals. As proof of high impact forces applied to cartilage and bone at the
time of injury, traumatic bone marrow lesions and osteochondral fractures, located
predominantly in the lateral tibiofemoral compartment, are commonly associated
with an ACL injury. The subsequent risk of OA may be closely associated with the
knee injury mechanism and the panorama of injuries in the knee sustained at the
onset of injury. The purpose of this work was to acquire a better understanding of
how the initial impact, related to the trauma mechanism of acute knee injuries,
may influence acute and chronic knee pathology.
In this work it was found that subjects with post-traumatic OA secondary to an
ACL injury have more joint space narrowing and more osteophytes in the lateral
compartment than in the medial compartment, compared with subjects with non
traumatic OA. Furthermore, it was found that an acute knee injury is associated
with instant and sustained synovial fluid biochemical alterations within the first
month of knee injury, suggestive of increased cartilage turnover and severe joint
inflammation. Those subjects who sustained an osteochondral fracture with
disrupted cortical bone in association with the soft tissue knee injury had increased
joint inflammation. In an in vitro bovine cartilage study, mechanical injury to
cartilage increased the matrix metalloproteinase-induced cleavage of cartilage
aggrecan. Moreover, findings from this model suggest that the aggrecan
degradation may differ between cytokine-stimulated cartilage explants compared
with cartilage explants mechanically injured and (or) co-incubated with joint
capsule.
Conclusively, the findings in this work underline the fact that the initial impact
associated with an ACL appears to be important in terms of the risk of developing
post-traumatic OA. In addition, this work emphasizes how the acute biological
response to injury could be involved in cartilage degradation. A greater
understanding of these processes could lead to the improved management of knee
injured patients and possibly delay, or even prevent, OA development.
9
10
Populärvetenskaplig sammanfattning
Främre korsbandsskador är vanliga, allvarliga och traumatiska knäskador. I
kombination med associerade meniskskador och ändrad ledbelastning leder skadan
på sikt till att en stor andel individer utvecklar artros i ung ålder. Detta benämns
post-traumatisk artros. På kort sikt kan skadan leda till oförmåga att fortsätta sin
aktivetet på samma nivå som tidigare. De kraftiga kompressionskrafter över brosk
och ben då skenben och lårben kolliderar i skadeögonblicket leder till skador och
celldöd i vävnaderna Detta bidrar sannolikt till risken för att utveckla post
traumatisk artros. Hos fler än hälften av de skadade individerna kan man med MR
visualisera en fraktur i kortikalt ben på skenben och/eller lårben. Skadorna som
uppstår i samband med traumat är framförallt lokaliserade lateralt i knäleden och
kan tillsammans med blödning i knäleden initiera ett inflammationssvar. Dessutom
initieras en läkningsprocess som åtföljs av en ökad omsättning (nysyntes och
nedbrytning) av olika broskproteiner. Det broskprotein som framförallt har
undersökts under utförandet av detta avhandlingsarbete är aggrekan. Molekylen
aggrekan består av ett protein på vilket det sitter ett stort antal sockerkedjor. Dessa
är kraftigt negativt laddade och bidrar till broskets funktion genom att attrahera
positiva motjoner, vilka genom diffusion drar till sig vatten. Detta leder till ett
svullnadstryck och medför att brosket kan motstå de krafter som uppstår när man
belastar knäleden. Under artrosprocessen bryts aggrekanet ned av olika enzym
som finns i broskmiljön. Framförallt antas så kallade aggrekanaser och matrix
metalloproteaser (MMP) ha stor betydelse. Dessa enzym klyver
aggrekanmolekylen på olika ställen och genererar olika långa aggrekanfragment
som kan mätas i ledväska och i broskmedium. Ett fragment som bildas efter
klyvning av aggrekanaser och som analyseras i både arbete III och V är ARGS
aggrekan. Ett fragment som bildas efter klyvning av MMP är FFGV-aggrecan
(arbete V).
Avhandlingen utgår från frågeställningen av vad ett vridvåld i knäleden
(knäledsdistorsion) innebär i det akuta skedet och på längre sikt. I första delarbetet
gjordes en jämförelse av lokalisationen av artrosförändringar mellan knäleder som
har ett definierat trauma (främre korsbandsskada) mot de som inte har detta. Vi
fann mer lateralt lokaliserad ledspringesänkning och osteofyter i knäleder där det
funnits ett definierat trauma vilket indikerar att det initiala traumat har en viktig
roll i den post-traumatiska artrosprocessen. Dessutom kan det vara en viktig
11
klinisk implikation, då det verkar som att efter en främre korsbandsskada
utvecklas artros både medialt och lateralt i knäleden. Artros hos individer utan
tidigare trauma är oftast lokaliserad medialt i knäleden.
I andra delarbetet belyses vilken inverkan knäledens ställning har i samband med
artrosutveckling efter en korsbandsskada. Vi fann att individer med varusställning
(hjulbenthet) hade mer artros i sitt skadade knä jämfört med de med normal eller
valgusställning (kobenthet). Även om skillnaderna mellan grupperna var stora, var
de inte statistiskt säkerställda. Mer forskning på området för att utreda om
varusställning ökar risken att utveckla artros efter en främre korsbandsskada är
indicerad.
Arbete III och IV utgår från mätning av ledvätskekoncentrationen av olika brosk-,
ben- och inflammationsmarkörer. Vid jämförelse mellan individer med akut
knäskada och knäfriska kontroller, visade sig ledvätskekoncentrationer i knäleden
av total aggrekan och ARGS-aggrekan vara förhöjda från 1-3 dagar efter skadan
upp till 23 dagar efter skadan. Ett annat broskprotein, COMP var förhöjt från 2-3
dagar efter skadan upp till 23 dagar efter skadan. Den nya kunskap som dessa fynd
indikerar är att omsättningen av både aggrekan och COMP ökar nästan omedelbart
efter skadan. Även flera pro-inflammatoriska proteiner och benassocierade
proteiner återfanns i högre koncentrationer i knäleden hos de knäskadade
individerna från 0-23 dagar efter knäskadan. I arbete IV undersöktes hur ovan
nämnda biomarkörer relaterade till förekomst av MR-visualiserade osteokondrala
frakturer där skenben och lårben kolliderade i samband med den främre
korsbandsskadan. De inflammatoriska cytokinerna interleukin-8 och tumour
necrosis factor-α var förhöjda i ledvätskan hos de individer som ådragit sig en
osteokondral fraktur med åtföljande avbrott i det kortikala benet. Sådana frakturer
skulle således kunna vara viktiga för det initiala inflammationssvaret på skadan,
men även för risken att utveckla artros på lång sikt. Om så är fallet får utvärderas i
framtida studier.
I arbete V studerades effekterna på aggrekannedbrytning av ett trubbigt våld mot
ungt kalvbrosk i laboratoriemiljö samt effekterna av att odla ungt kalvbrosk
tillsammans med ledkapsel. Mekanisk skada av brosket ledde till ökad MMP
aktivitet, medan en antydan till ökad aggrekanasaktivitet sågs i mekaniskt skadat
brosk som odlades tillsammans med ledkapsel. I mekaniskt skadat brosk som
behandlades med cytokin sågs mycket kraftig aggrekanasaktivitet, men ingen
MMP-aktivitet.
Denna avhandling belyser en del processer som skulle kunna ha betydelse för
utvecklingen av artros efter en allvarlig knäskada och speciellt efter en främre
korsbandsskada. Sammantaget stärker den antagandet att traumat över ben och
brosk som sker i samband med att en främre korsbandsskada uppkommer kan ha
stor betydelse för risken att utveckla post-traumatisk artros.
12
Introduction
Preface
Like many diseases, the manifest disease and pathogenesis of osteoarthritis (OA)
are multidimensional. The dimensions of OA range from clinical symptoms and
macroscopic features of disease through biomechanics to microscopic events at
cell, extracellular matrix and molecular level. An increasing amount of knowledge
is starting to bridge these dimensions, but there are still many large holes to fill, in
order better to understand the disease.
The main advantage of post-traumatic OA from the scientist’s perspective is that
the time of disease onset is known and the progression of the disease is relatively
rapid compared with non-traumatic OA. One of the major drawbacks from the
patients’ perspective is that they are affected by the disease at a young age.
During my time as a PhD student, I have tried better to understand why and
through which processes a severe knee injury can lead to post-traumatic OA. This
knowledge is crucial in order to understand the disease and to be able to construct
viable treatment options and advice for subjects with severe knee injuries. The
main hypothesis underlying the present thesis was that the impact on cartilage,
bone and soft tissues of the knee, inflicted at the time of injury, plays an essential
role when it comes to the future risk of developing post-traumatic OA. The
strength of the present thesis is that, via the papers in the study, it links different
clinical parameters and molecular mechanisms that may be important for post
traumatic OA development.
13
List of studies
I.
II.
III.
IV.
V.
Swärd P, Kostogiannis I, Neuman P, Von Porat A, Boegard T, Roos H.
Differences in the radiological characteristics between post-traumatic and
non-traumatic knee osteoarthritis. Scand J Med Sci Sports.
2010;20(5):731-739.
Swärd P, Fridén T, Boegard T, Kostogiannis I, Neuman P, Roos H.
Association between varus alignment and post-traumatic osteoarthritis
after anterior cruciate ligament injury. Knee Surg Sports Traumatol
Arthrosc. 2013;21(9):2040-2047.
Swärd P, Frobell R, Englund M, Roos H, Struglics A. Cartilage and bone
markers and inflammatory cytokines are increased in synovial fluid in the
acute phase of knee injury (hemarthrosis)--a cross-sectional analysis.
Osteoarthritis Cartilage. 2012;20(11):1302-1308.
Swärd P, Struglics A, Englund M, Roos H, Frobell R. Soft tissue knee
injury with concomitant osteochondral fracture is associated with higher
degree of acute joint inflammation. Am J Sports Med. Published online
March 24, 2014. DOI: 10.1177/0363546514524924.
Swärd P, Hansson M, Lohmander SL, Wang Y, Grodzinsky A, Struglics
A. Evidence of increased protease activity in mechanically injured
cartilage co-cultured with joint capsule. Manuscript.
14
Abbreviations
ACL
Anterior cruciate ligament
ADAMTS A disintegrin and metalloproteinase with thrombospondin motifs
BMI
Body mass index
BSP Bone sialoprotein
C2C
Newly formed epitope after cleavage of collagen at the type II
collagen primary cleavage site
COMP
CRP
Cartilage oligomeric matrix protein
C-reactive protein
CS Chondrotin sulphate
DAMP
Damage-associated molecular patterns
ECM Extracellular matrix
HA
Hyaluronan
KS Keratan sulphate
MMP Matrix metalloproteinase
MRI
Magnetic resonance imaging
OA Osteoarthritis
OCL
Osteocalcin
OPN Osteopontin
PF Patellofemoral
sGAG
Sulphated glycosaminoglycan
SPARC
TGF-β
TIMP
TNF-α
secreted protein acidic and rich in cysteine
Tumour growth factor-β
Tissue inhibitor of matrix metalloprotease
Tumour necrosis factor-α
TF Tibiofemoral
15
16
Background
The knee joint
The knee joint is located between the two other joints of the lower limb; the hip
and the ankle. The proximal end of the tibia and the distal end of the femur form
the medial and lateral tibiofemoral compartments. The patella and the anterior part
of the distal femur form the patellofemoral joint. Together, these joints form the
knee joint. High demands are imposed on the knee joint and it has several
functions which are essential for human beings to walk, run and jump. Primarily, it
enables flexion-extension of the lower limb in the sagittal plane. In extension, full
or close to full, the knee must be able to withstand the strong forces imposed on
the knee by gravity. In flexion, rotation at the knee enables the leg to position the
foot before placement. Small movements in the varus/valgus direction are also
possible at the knee joint but only when the knee is flexed. Joint stability during
movement is attained by the shape of the articular surfaces, the collateral and
cruciate ligaments, the menisci and tendons and muscles crossing the knee joint
(Figure 1) [1].
The anterior cruciate ligament
The anterior cruciate ligament (ACL) has two functional bundles which connect
the femur and the tibia; the anteromedial and the posterolateral bundles, named
after their insertion sites on the tibia (Figure 2). These bundles bridge the
posteromedial aspect of the lateral condyle and the medial tibial plateau where
they insert next to and anterior to the tibial spines. The ACL has an intra-articular
location, but it is separated from the synovial fluid by a synovial lining. The main
blood supply originates from the femur and specifically from the central geniculate
artery [2, 3]. Pacinian corpuscles, Golgi tendon organs and Ruffini endings are
mechanoreceptors present in the ACL which contribute to the proprioceptive sense
[4]. The extracellular matrix (ECM) of the ACL contains collagen types I, II, III
and V, elastin and proteoglycans. The tensile properties of the ACL are mainly
related to bundles of collagen type I and cross-linking of these [5, 6]. In the normal
17
ACL, fibroblasts reside along collagen bundles and are important for normal
ligament turnover. Moreover, cells with progenitor potential are present [6]. The
function of the ACL is to provide tibiofemoral joint stability in anterior-posterior
translation and in internal-external rotation [7]. The ACL also restrains movements
in the varus-valgus direction. The anteromedial and posterolateral bundles of the
ACL act in synergy to stabilise the knee joint through its entire range of motion. In
flexion, the anteriomedial bundle is tauter; in extension, the posterolateral bundle
is tauter [3]. The combination of valgus and internal rotation of the tibia has been
shown to increase ACL strain more than either motion alone [8].
Figure 1. Posterior view of the left knee showing the anterior and posterior cruciate ligaments and
the menisci. The image which is from the 20th US edition of Gray's Anatomy of the Human Body
was originally published in 1918 and has been transferred into the public domain.
18
Figure 2. Schematic drawing of the double-bundle ACL anatomy. AM, anteromedial bundle; PL,
posterolateral bundle. Published with kind permission. © C. Kartus.
Anterior cruciate ligament tears
In Sweden, football (soccer) is the most common activity associated with ACL
injury for both men and women. The second most common activity associated
with ACL injury is downhill skiing for women and floorball for men
(www.aclregister.nu). The incidence of ACL tears was shown to be 81 per
100,000 subjects aged between 10 and 64 years [9]. Based on these numbers, some
6,000 ACL injuries occur in the Swedish population every year, of which ~3,000
ACLs are surgically reconstructed. The indications for ACL reconstruction in
Sweden are symptoms of instability and the failure of conservative treatment
(www.aclregister.nu). In the USA, approximately 200,000 ACL reconstructions
19
are performed every year [10]. Women are known to be more susceptible to ACL
rupture compared with men, and are injured at a younger age [11, 12]. Anatomic,
neuromuscular and hormonal variations between men and women have been
proposed to explain this observed difference [13-16]. From the individual’s
perspective, the injury may, in the short term, lead to knee dysfunction and an
inability to continue sports participation at the same level as pre-injury. In a recent
review, it was summarised that, although normal to nearly normal knee function
was regained in most ACL-reconstructed individuals, a relatively small number of
individuals returned to their pre-injury activity level and competitive sports [17].
Psychological factors, including fear, lifestyle changes and personality, have an
impact on why subjects do not return to their pre-injury sports activity after ACL
reconstruction [18]. In professional sports, the return to play after ACL injury is
much greater and Waldén et al. [12] reported that 94% of elite level football
players returned to training within 10 months and that 89% participated in a match
within 12 months after ACL reconstruction.
An ACL tear also leads to a high risk of developing OA of the injured knee at a
young age [19-22]. ACL re-injury/graft rupture and the high risk of contralateral
ACL injury are other important issues that require consideration when advising an
individual to return to sport [23]. In a reasonably sized study, in terms of the
queries tested, it was reported that 4.5% sustained a graft rupture and 7.5%
sustained a contralateral ACL injury during the five-year follow-up after ACL
reconstruction. Importantly, 29% of individuals younger than 20 at the time of the
first ACL injury sustained an ACL injury to either knee during follow-up [24].
Taken as a whole, the high risk of ACL injury, specifically in young women, and
future complications related to the injury are regarded as one of the major
problems in sports medicine [25].
Anterior cruciate ligament injury mechanism
In the literature, the ACL injury mechanism is most often described as contact or
non-contact. This definition may be misleading, as most ACL injuries that occur in
contact sports are associated with a “non-contact” mechanism [26, 27]. A contact
ACL injury mechanism typically occurs in American football, when the subject
plants his foot and, at the same time, is tackled at the knee from the lateral side,
resulting in a valgus collapse of the knee joint. The injury panorama of this injury
mechanism often results in lesions of the ACL, medial collateral ligament and
medial meniscus, referred to as the O’Donoghue triad [28]. Several mechanisms
that may result in a non-contact ACL injury have been proposed. Most typically,
the injury occurs as the athlete plants the foot with the knee in slight flexion and
20
during landing, side-cutting or deceleration. In this position, knee valgus motion,
internal rotation of the tibia towards the femur or external rotation of the femur
towards the tibia, combined with anterior translation of the tibia, ensue. This leads
to high strain in and the rupture of the ACL [26, 29, 30]. This ACL injury
mechanism, also known as the pivot shift injury, induces simultaneous lateral
tibiofemoral compartment subluxation and combined compressive force,
explaining the typical location of bone marrow lesions, on the posterolateral tibial
plateau and the midportion of the lateral femoral condyle [2]. The impact site on
the lateral femur is related to the degree of knee flexion at the time of injury. With
the knee in a high degree of flexion or extension at the time of injury, the lateral
tibia impacts the posterior or the anterior part of the lateral femur respectively [2].
Differences in the ACL injury mechanism between the sexes have been described.
Investigating basketball players, Krosshaug et al. [31] reported that a valgus
collapse in association with the injury was five times more common in women. A
lower incidence of meniscal tears associated with the ACL injury has been
indicated in women [32], which could be related to differences in injury
mechanism between men and women. Furthermore, besides from showing that
young age was associated with more traumatic bone marrow lesions on the lateral
femoral condyle, it was in a study by Bisson et al. [33] demonstrated that male
gender associated with mild traumatic bone marrow lesions on the lateral femoral
condyle and tibial plateau. Male gender also associated with moderate and severe
traumatic bone marrow lesions on the lateral femoral condyle. These findings may
further highlight gender-specific differences in ACL injury mechanism and as a
result, factors other than valgus collapse may be of greater importance for
compressive injuries to soft tissues and bone associated with ACL injury. Even if
they were not investigating bone-bruise patterns, Fridén et al. [27] proposed that
the degree of compression between the tibia and femur was related to weight
bearing or non-weight-bearing at the time of injury. A non-weight-bearing ACL
injury mechanism, which typically occurs when skiing, was associated with larger
numbers of intact menisci, indicating a lower degree of joint compression in this
type of injury [27]. In a large-scale investigation of 525 subjects, younger age and
not jumping at the time of ACL injury were associated with a bone bruise [34].
Jumping at the time of injury may indicate a non-weight-bearing ACL injury and
these findings are thus in line with the hypothesis put forward by Fridén et al. [27].
21
Knee injury panorama
An acute knee injury with joint effusion and intra-articular bleeding (hemarthrosis)
suggests significant intra-articular pathology. Anterior cruciate ligament tears,
meniscal lesions and lesions of the medial collateral ligament are common after
rotational knee injury [9]. In subjects (n=1,145) with an acute knee injury in whom
MRI (magnetic resonance imaging) was performed a median of eight days after
the trauma, 52% had sustained an ACL injury, 17% had transient patellar
dislocation and 28% a medial collateral ligament tear. Among the ACL-injured
subjects, 55% also sustained an associated meniscal tear [35]. These findings are
in line with previous investigations of smaller study samples, regarding both the
panorama of knee injuries [9] and the high prevalence of meniscal tears
concomitant to the ACL tear [36, 37]. The majority of studies and the historical
view indicate that lateral meniscal tears are more common than medial tears after
ACL injury [36, 38-40]. However, recent studies have described a similar
prevalence or even more medial than lateral meniscal tears in association with
ACL injury [9, 37]. In a recent MRI-based study, the most common meniscal tear
associated with the ACL injury was a longitudinal tear of the posterior horn of the
medial meniscus [9]. The authors speculate that these tears may be difficult to
detect during routine arthroscopy and could progress and produce symptoms in an
unstable knee [9]. Isolated or multiple articular cartilage lesions are also a frequent
finding after ACL injury [32, 41]. Posterolateral knee injury, which is relatively
uncommon, is an important diagnosis which, if left untreated, can lead to severe
knee disability [42, 43].
In almost all individuals, an acute knee injury also leads to a collision between the
tibial plateau and the femoral condyle, as visualised using MRI by traumatic bone
marrow lesions at the site of impact. These have been described as “fingerprints of
injury mechanism”, or “kissing lesions” and have been detected in almost all
subjects who have suffered an ACL tear [2, 44, 45]. Typically, the collision occurs
in the lateral tibiofemoral compartment between the non-articular posterolateral
tibial plateau and the articular midportion of the lateral femoral condyle [44, 46,
47]. Depending on the injury mechanism, other locations of traumatic bone
marrow lesions may be present (see above). Traumatic bone marrow lesions in the
medial tibiofemoral compartment have been ascribed to the contre-coup
mechanism, but they are less common and are mostly associated with bone
marrow lesions of the lateral compartment [48]. It should be noted that medial
bone marrow lesions are commonly found in subjects with a combined ACL and
posterolateral knee injury [42].
At the site of impact between tibia and femur, more than half of ACL-injured
subjects also sustain a traumatic osteochondral fracture [39, 44]. These fractures
22
are prevalent in the lateral tibiofemoral compartment overlying the traumatic bone
marrow lesions and present with or without the disruption of cortical bone (Figure
3). They indicate strong impact forces between the tibia and femur at the time of
injury and correlate with the size of bone marrow lesions [39, 44] and the
likelihood of associated meniscal tears [39]. The traumatic bone marrow lesions
represent trabecular fractures, bleeding or oedema. In one study, human biopsy
samples of cartilage and subchondral bone overlying MRI-detected bone bruises
were obtained at a median of 4.5 weeks after ACL injury. Glycosaminoglycan loss
and chondrocyte necrosis were observed in the overlying cartilage. In the
subchondral bone, osteocyte necrosis was indicated by empty lacunae [49].
Figure 3. Magnetic resonance images of knees used in Study IV. Upper row: proton density T2
weighted sequence and bottom row: short-tau inversion recovery (STIR) sequence of knees with
osteochondral fractures with disrupted cortical bone in the femur (yellow arrow) and without
disrupted cortical bone in the femur (blue arrow). The STIR sequence images (bottom row) clearly
show the surrounding post-traumatic bone marrow lesion of the osteochondral fracture above,
indicating traumatic impact forces (white arrows).
23
24
Osteoarthritis
Knee osteoarthritis
Knee OA is more common in women and knee OA incidence and prevalence
increase with increasing age. Knee OA may be isolated, or may be part of
generalised OA, defined as affecting the hands and at least one large joint, or
affecting three or more joints [50]. Obesity has been shown to be a highly
important and modifiable risk factor for OA development and progression [51].
Adding to the complexity of the disease, the metabolic syndrome and two of its
components, central obesity and hypertension, are also associated with the
incidence of severe knee OA requiring total knee replacement, independent of
body mass index (BMI) [52]. Systemically increased factors derived from visceral
adipose tissue have been shown to increase the risk of hand OA and may also be
involved in the pathogenesis of knee OA [53]. Today, knee OA is a leading cause
of global disability. The ageing population and increasing rates of obesity
worldwide forecast an increasing need for health care related to knee OA in the
future [54]. Symptomatic knee OA affects almost 7% of the US population 50-84
years of age, of which approximately 50% are obese [55]. Petersson et al. [56]
reported that 1.5% of individuals 35-54 years of age, living in a district in the
southwest of Sweden, had symptomatic non-traumatic knee OA. The rates of
radiographic knee OA are much higher. In the elderly (over the age of 60 to 75),
the prevalence of radiographic OA was shown to be between 31-45% [57-59].
There may, furthermore, be ethnic differences regarding knee OA location. For
example, in the Chinese, the lateral tibiofemoral compartment is affected more
often than in Caucasians [60]. Importantly, OA has a large impact on quality of
life and was estimated to reduce the remaining quality-adjusted life expectancy in
persons with knee OA by 10-13%, with the higher rates applying to younger
individuals with knee OA [55]. As indication of the increasing incidence of
symptomatic knee OA, the numbers of total knee replacements performed in the
USA more than doubled from 1999-2008. In younger individuals (45- to 64-year
olds), the numbers more than tripled. In addition to increasing obesity rates, this
increase may be related to wider indications of surgery and increasing numbers of
severe knee injuries in young individuals [61]. An increased understanding of knee
25
OA risk factors, how best to treat symptoms and find treatments that can prevent
or stop OA progression is needed.
Apart from age, gender and obesity, also genetics, knee injuries and abnormal joint
loading influence the risk of developing knee OA. Seven genetic variants
associated with knee OA or total knee replacement have been identified [62] and
39% of the risk of developing knee OA has been attributed to genetic variation
[63]. The possibility that abnormal, or overly high loads can lead to progressive
cartilage degradation was indicated by the increased risk of knee OA in
occupations where there is frequent heavy lifting, kneeling or squatting [64].
Moreover, malalignment has been shown to increase the risk of OA [65, 66] and
elderly male former elite athletes engaged in non-impact sports have an increased
knee OA prevalence after adjustment for previous knee injury, age, gender, BMI
and occupational load [67].
Post-traumatic knee osteoarthritis
As things stand, some 12% of the total OA burden has been ascribed to post
traumatic OA [68]. It has been proposed that the incidence of post-traumatic OA is
increasing in relation to increased numbers of individuals engaged in sports and
over the last few decades, the increase in sports participation has been substantial.
In the USA, the number of women participating in high-school sports has roughly
doubled every decade [46]. Instability is as a major cause of disability after ACL
injury [69]. This has influenced the treatment of ACL injuries and ACL
reconstruction has been a preferred treatment, with the aim of re-establishing knee
stability [70]. However, no differences in the long-term risk of OA development
have been shown between ACL-injured subjects treated with or without primary
ACL reconstruction in recent systematic literature analyses and from the early
results of a randomised controlled trial [19, 21, 71, 72]. Associated injuries may be
of greater significance for the long-term prognosis. In a review by Øiestad et al.
[21], it was concluded that the prevalence of knee OA after an isolated ACL injury
was 0-13%. The prevalence of knee OA after an ACL tear with an associated
meniscal tear was 21-48% [21].
Differences in the classifications relating to the radiographic grading of OA
between different studies have led to difficulties juxtaposing the current
knowledge [21, 71]. In a recent meta-analysis only including studies using the
Kellgren & Lawrence classification and with a minimum follow-up time of 10
years, it was shown that non-ACL-reconstructed knees had an increased relative
risk of developing any grade of OA. However, the relative risk of progression to
moderate or severe OA tended to be higher in ACL-reconstructed knees [71].
26
Cartilage
The proximal tibia, the distal femur and the patella are covered by a thin layer of
hyaline cartilage, forming the articular surfaces of the knee joint. Normal cartilage
is avascular, aneural, with no lymphatic vessels and the only cell type found in
cartilage, the chondrocyte, obtains nutrition mainly from passive diffusion from
synovial capillaries (see below). In normal articular knee cartilage, only 1-3% of
the wet weight tissue consists of chondrocytes [51]. Some 70% of the tissue
consists of water, whereas 20% of the wet weight is collagen (mainly type II; the
fibril-forming collagen of cartilage) and approximately 5% is aggrecan [73]. The
main function of cartilage is to distribute the load applied to the underlying bone
and to allow movements of low friction at the knee joint [74]. These functional
properties are preserved by the chondrocytes which maintain cartilage hemostasis
by inducing proteolysis and the production of non-fibrillar collagens,
proteoglycans and other non-collagenous molecules, as a reaction to
biomechanical and biochemical stimuli [75].
The capacity of cartilage to absorb and distribute high loads at a specific site is
related to the composition of cartilage ECM and specifically to the integrity of the
fibrillar collagen and aggrecan networks that are present. Aggrecan forms large
aggregates by binding to hyluronan and is substituted by negatively charged
sulphated glycosaminoglycans (sGAGs) [76]. The high negative charge attracts
counter-ions and, by diffusion, water is attracted to the aggrecan molecule. This
results in a swelling pressure in cartilage which is retained by the tensile strength
of the collagen fibril network and gives cartilage the ability to withstand
compressive loads [76]. The fibrillar collagen and aggrecan networks, on the other
hand, are dependent on other molecules present in the cartilage ECM, such as
cartilage oligomeric matrix protein (COMP), non-fibrillar type IX collagen and
members of the small leucin-rich repeat protein family. Important functions of
these molecules include the regulation of collagen fibril formation and other
collagen networks in the cartilage ECM. They also enable interactions between
different ECM molecules and interactions with the chondrocytes [76-78].
Type II collagen has a long half-life (about 100 years) and its degradation is
believed to be practically irreversible [79]. Several collagenases are able to cleave
fibrillar type II collagen, of which the matrix metalloproteinase (MMP)-13 is
believed to be the most important [51, 77]. Cleavage at the primary cleavage site
in type II collagen by these collagenases generates two fragments, of which one
can be detected by a neo-epitope antibody (against the newly formed epitope;
named C2C). Interestingly, increased synovial fluid concentrations of C2C are
associated with pre-radiographic cartilage lesions in the knee joint of ACL-injured
subjects [80]. Aggrecan has a more rapid turnover. The cleavage of aggrecan can
27
be mediated by several proteolytic enzymes (Figure 4). The aggrecanases-1 and -2
(ADAMTS-4 and -5; a disintegrin and metalloproteinase with thrombospondin
motifs), and, secondly, the MMPs are believed to be most important [81-84].
Another somewhat abundant protein in cartilage is COMP, which is believed to be
important for the fibrillation of type II collagen in cartilage and for stabilising the
collagen network in the adult cartilage. Its turnover is increased during the early
progression of OA [77]. The turnover of COMP is also increased after an acute
knee injury, as observed by increased synovial fluid concentrations from within
one week of injury up to several years later [85].
HA
G1
ADAMTS-4/5
G2
IGD
G3
L
MMP
KS
CS
Figure 4. Aggrecan is bound to hyaluronan (HA); an interaction stabilised by the link protein (L).
Aggrecan is composed of three globular domains; G1 (N-terminal side), G2 and G3, an inter
globular domain (IGD), one region rich in keratan sulphate (KS) and two regions rich in chondroitin
sulphate (CS1 and CS2). Several enzymes have been shown in vitro to cleave aggrecan in the IGD
domain [81, 83, 84, 86]. Aggrecanase cleavage at the TEGE373↓374ARGS site and MMP cleavage
at the IPEN341↓342FFGV site have been demonstrated in humans [81, 83, 84, 87-89]. Suggestive of
rapid aggrecan turnover by increased aggrecanase activity, increased concentrations of ARGS-SELE
and ARGS-CS1 fragments have been shown in the synovial fluid of subjects with acute knee injury.
Moreover, increased aggrecanase cleavage in the aggrecan CS2 domain generates GRGT-G3,
GLGS-G3 and AGEG-G3 fragments [88]. Figure reproduced with the kind permission of Dr. André
Struglics.
Cartilage can be divided into regions of different ECM organisation and molecular
content, as well as chondrocyte organisation, shape and function throughout the
depth of cartilage: the superficial (tangential), central (intermediate) and deep
(radial) layers of cartilage [90]. The transitional zone between the cartilage and
subchondral bone is called the calcified cartilage layer and it creates a barrier and
28
attachment site between cartilage and subchondral bone [91]. Detected differences
in cell appearance and function between the different layers of cartilage include
the following. (1) The number of chondrocytes per cartilage volume is higher in
the superficial zone compared with deeper layers and the chondrocytes appear
flattened and are aligned horizontally parallel to the joint surface [92, 93]. The
superficial zone chondrocytes, furthermore, produce lubricin, as opposed to the
deeper layers [94]. There is also a high content of collagen fibres running parallel
to the cartilage surface, although there may be differences between weight- and
non-weight-bearing regions (see below) [7]. (2) In the central layer, the
chondrocytes are rounder and the proteoglycan content is higher compared with
the superficial or deep zones [92, 95]. (3) In the deep layer, the collagen fibres are
thicker and round chondrocytes are aligned in columns [90, 92]. In addition to
these depth-related differences, the cartilage ECM also differs depending on the
proximity to the chondrocytes. With an increasing distance from the cells, ECM is
classified as pericellular, territorial and interterritorial, with apparent differences in
ECM organisation and function [77]. Mesenchymal progenitor cells, which could
have the potential to regenerate focal cartilage defects, have been found in
cartilage [96].
Different cartilage regions in the knee joint, furthermore, show differences in cell
and ECM organisation attributed to the different mechanical loads between
regions. Rolauffs et al. [93] detected four distinct superficial zone chondrocyte
patterns; strings, clusters, pairs and singles in the knee joint. Different joint
surfaces of the knee were typically dominated by only one of these four patterns.
The predominant pattern of the femoral condyles, meniscus-covered medial tibial
plateau and patellofemoral grove were strings, pairs and clusters respectively.
Central regions of the tibial plateau are associated with a less organised collagen
fibre orientation, which could be a consequence of high compressive loads over
this region. On the other hand, peripheral regions of the tibial plateau (i.e. beneath
the menisci) are associated with a more organised collagen fibre orientation more
parallel to the surface. This could be a consequence of the high tensile stresses to
which these regions are exposed [7, 97].
Cartilage and knee osteoarthritis
One of the main features of knee OA is the loss of articular cartilage. Over the
years, much of the research on OA has therefore focused on the events involved in
cartilage degradation. At cartilage level, OA development is associated with
cartilage fissures, swelling, chondrocyte hypertrophy and phenotypic changes in
the chondrocytes. As the disease progresses, cartilage thinning and the exposure of
subchondral bone occur [51]. Several important clues to the way cartilage is
affected during the different stages of OA development and progression have been
29
identified during the past few decades. Proteases able to degrade cartilage
constituents have been highlighted and MMP-13 and ADAMTS-5 in particular
[79]. In models using genetically modified mice, deleting the catalytic domain of
ADAMTS -5 protects the cartilage from degradation [98]. Furthermore, knock-in
induced resistance to cleavage at the wild type ARGS↓TEGE-aggrecan site
significantly protects the cartilage from degradation [99]. Induced MMP-13
deficiency also inhibits cartilage degradation [100]. Underlining the complexity of
the post-traumatic OA mouse model, at least 29 different genetic modifications
have been shown to be protective of cartilage erosion and 19 have been shown to
increase cartilage erosion in different OA models [101]. Human OA is
characterised by a whole joint disease and inducing MMP-13 deficiency did not
lead to reduced osteophyte formation in a mouse OA model [100]. Many other
molecular mechanisms have been implicated in OA pathogenesis and recent
studies have, for example, indicated important roles for tumour growth factor
(TGF)-β and complement activation in OA pathogenesis [51, 102, 103].
Cartilage and post-traumatic knee osteoarthritis
Acute knee injuries are associated with radiographic OA progression and joint
space narrowing indicative of cartilage erosion (see above). One major limitation
of using radiographic signs to detect cartilage injury is the relatively long time
from injury until these changes occur and the fact that changes indicating cartilage
loss, such as joint space narrowing, may be related to other factors, such as the
integrity of the menisci [104]. Recent advances in MRI have led to an increase in
our understanding of early post-traumatic changes in cartilage which may precede
post-traumatic OA development [105, 106]. Compositional cartilage MRI (delayed
gadolinium-enhanced imaging of cartilage (dGEMRIC), T1-rho and T2 mapping)
can depict information on the composition of cartilage. Whereas dGEMRIC and
T1rho may distinguish changes in cartilage proteoglycan content, T2 mapping
relates to the cartilage water content and indirectly to type II collagen content and
orientation [105]. Before the development of these techniques, cartilage biopsies
revealed important information on changes in cartilage composition after ACL
injury. Suggestive of collagenase activity, biopsy samples acquired from the non
weight-bearing articular cartilage of the intercondylar notch show collagen
denaturation and the cleavage of type II collagen less than and more than one year
post-ACL rupture [107, 108]. Interestingly, an increase in cartilage GAG content
was also observed [107]. At time points after one year, lower type II collagen
content in the cartilage of the intercondylar notch was observed [108]. Biopsy of
the cartilage adjacent to bone marrow lesions on the lateral femur showed a
decrease in GAG in median 4.5 weeks after ACL injury [49]. In line with these
findings, Li et al. [109] showed elevated T1rho values (indicating GAG loss), of
30
the posterolateral tibial cartilage in ACL-injured knees at baseline. Tiderius et al.
[110] showed that the estimated GAG content (assessed by dGEMRIC) was lower
in both the medial and lateral femoral cartilage within mean three weeks after
ACL injury. At two-year follow-up, recovery but not normalization, was observed
in the lateral tibifemoral compartment whereas the estimated GAG decrease in the
medial tibiofemoral compartment seemed to remain at a constant low level [111].
Potter et al. [41] demonstrated MRI-detectable cartilage lesions at the time of ACL
injury in 100% of patients, primarily affecting the lateral tibial plateau and
secondly the lateral femur. A progressive increase in cartilage lesion severity was
observed with time lapsed after injury [41].
Synovium
The non-bony cavity of the knee joint is enclosed by the synovium. It consists of
the intima, a continuous layer of macrophages and specialised synovial fibroblasts
(synoviocytes), and the subintima, which is the underlying tissue. The subintima is
a fibrous ECM where blood and lymph vessels, nerves, stationary fibroblasts and
immune cells reside. The synovial fibroblasts have a distinct phenotype. Of
particular importance to the joint environment, they experience high activity from
an enzyme which converts UDP-glucose to UDP-glucoronate, an essential
component for hyaluronan synthesis [112]. Furthermore, synovial fibroblasts,
along with superficial zone chondrocytes, produce lubricin, which is particularly
essential for the boundary lubrication of cartilage [113].
The main route for nutritional access to the avascular cartilage is passive diffusion
from capillaries of the synovium. They are located in the subintima, just beneath
the intima. The capillary endothelial cells form a size-selective barrier allowing
the flux of water, nutrients and proteins into the joint cavity. The synovial fluid-to
plasma ratio of plasma proteins decreases as the molecular radius of the protein
increases [114]. Importantly, the cells of the intima are loosely connected and
there is no basement membrane. As a result, the interstitial fluid of the synovium
and the synovial fluid form an unbroken continuum [115]. The efflux of molecules
from the knee joint occurs through drainage via the lymph vessels for molecules
~2 to ~10 nm in size. Smaller molecules, for example, some cytokines, can also
access the circulation via diffusion into capillaries. Larger molecules, such as
hyaluronan and large aggrecan fragments, may be restricted from efflux via the
lymph vessels. At high intra-joint pressures, these large molecules could create a
filter cake, increasing the outflow resistance across the synovium and limiting
fluid escape from the joint [115, 116]. In the event of synovitis (see below), the
31
size selectivity of the endothelial barrier is decreased, leading to an increase in the
influx of large proteins and joint effusion [117].
The meniscus
The meniscus is a crescent-shaped fibro-cartilaginous tissue rich in collagen type
I. The molecular constituents of the meniscus are quite different from those of
cartilage. For example, compared with knee-joint hyaline cartilage, the collagen
fibrils are thin, the amount of aggrecan and chondroadherin is low, whereas the
amount of asporin is high [118]. The tensile strength and function of the menisci
are related to the circumferentially oriented collagen fibres woven together with
radial fibres [119]. In principle, the menisci play an important role in reducing
cartilage loads by distributing the load between the femur and tibia over a larger
area. Compromising this function, such as in the case of meniscus tear or
meniscectomy, leads to increased peak and average loads over the tibiofemoral
cartilage [120] and a high risk of OA [19, 121, 122]. The increased loading of
tibial cartilage is closely related to both the amount and type of medial
meniscectomy (anterior, posterior or longitudinal) [120].
The long-term outcome of ACL injury is closely related to the integrity of the
menisci (see above). However, subjects with an ACL tear (isolated or combined)
present with the first radiographic signs (joint space narrowing) of cartilage
degradation approximately ten years prior to subjects with an isolated meniscal
tear [123]. Not only the status of the menisci at index injury but also the risk of
secondary meniscus injury are important for the risk of future tibiofemoral OA
after ACL injury [124]. Meniscus injury also appears to be important when it
comes to developing patellofemoral OA [125, 126]. The menisci and specifically
the medial meniscus may contribute to joint stability and restrain anterior
translation in the ACL-deficient knee but not in knees with an intact ACL [127,
128]. This renders the medial meniscus susceptible to tears in the ACL-deficient
knee and this has been advocated as an important reason to perform ACL
reconstructive surgery [129].
Bone
The bony parts of the knee joint, the femur, tibia and patella, have a dense and
compact outer structure, the cortical bone, and a porous inner structure, the
trabecular bone. In the knee, trabecular bone and cartilage are separated by the
32
osteochondral plate; a thin layer of cortical bone (the subchondral plate) and the
calcified cartilage layer [130].
Bone is a metabolically active tissue. After the peak bone mass has been reached
at the age of 20-30 years, bone remodelling is characterised by a higher bone
resorption rate compared with the corresponding bone formation. This results in a
net loss of bone mass [131].
Bone and knee osteoarthritis
Bone-related alterations associated with OA include bone marrow lesions,
osteophytes, subchondral bone sclerosis, cyst formation, tidemark duplication and
thickening of the calcified cartilage layer [130]. Microcracks through the
osteochondral plate, ingrowth of nerves, vessels and endothelial proliferation
within the non-calcified cartilage are associated with OA and could increase cross
talk between cartilage and bone [132, 133]. These changes, in combination with
changes in bone homeostasis, may induce cartilage erosion [103, 133, 134]. An
important role for increased TGF-β signalling in the subchondral bone has been
suggested [103].
It is unknown whether pathological changes in bone precede or follow
osteoarthritic changes at molecular level in cartilage. The parallel progression of
disease in both tissues is, however, a likely scenario. This was illustrated by the
increase in both serum bone sialoprotein (BSP) and COMP in subjects with
chronic knee pain who went on to develop early radiographic knee OA [135]. The
serum concentrations of these proteins were, furthermore, higher in subjects with
bone scan abnormalities [136]. Even though these studies indicate parallel
alterations in cartilage and bone homeostasis during the course of early OA
development, abnormal alterations in bone typically precede those in cartilage, as
visualised by different imaging techniques. For example, osteophyte formation and
scintigraphic changes indicative of increased bone turnover typically occur before
joint space narrowing [137]. Changes in bone shape, furthermore, predict the onset
of radiographic OA [138]. These differences could, however, be related to the
higher metabolic rate in bone compared with that in cartilage and the more rapid
response of bone to changes in joint load.
Bone and post-traumatic knee osteoarthritis
In line with OA seen in subjects without previous knee joint trauma, early signs of
bone disturbance are also evident after severe knee injury. Bone mineral loss in the
knee of ACL-injured subjects was demonstrated after ACL reconstruction [139].
33
In a study population of 121 subjects followed prospectively after ACL injury, it
was recently shown that bone shape changes of femur, tibia and patella could be
detected by advanced analyses methods applied to sequential MR images already
after two years. Preliminary analyses suggest that these changes could relate to
osteophyte formation. [140]. In the same study population, changes in articular
bone curvature were observed within three months of the injury [141]. Buckland
Wright et al. [142] demonstrated thickening of horizontal trabeculae and a high
prevalence of osteophytes in the medial tibial compartment within four years of
ACL injury. Taken as a whole, these findings indicate changes in bone metabolism
and remodelling at an early stage after ACL injury.
In subjects with an acute knee injury, cartilage and bone involvement occurs
instantly, as the tibia and femur collide, resulting in cartilage injury, bone marrow
lesions and osteochondral fractures [41, 44, 49, 143]. Traumatic bone marrow
lesions gradually decrease in size, but fewer than half resolve within one year after
the injury [144]. In most studies, their presence has not been shown to correlate to
clinical outcome [34, 145]. Anterior cruciate ligament-injured subjects with a bone
marrow lesion of the medial tibia, a trabecular fracture of the lateral femur or a
more severe intra-articular lesion, i.e. osteochondral fracture, may, however, have
poorer clinical outcome scores [39, 146, 147]. Johnson et al. [146] observed that a
trabecular fracture, by the authors termed a geographical bone bruise [47], on the
lateral femur was associated with the increased size and duration of effusion, a
time-lag in time until the normal range of motion was regained and increased pain
in the acute phase of ACL injury.
Inflammation
Inflammation may play an important role in the pathogenesis of OA [148].
Indicative of this, synovitis, which has been shown to be reflected by increased
plasma C-reactive protein (CRP) concentrations, is a prevalent finding in OA
[149]. More importantly and underlining the importance of inflammation in OA,
synovitis has been shown to be prognostic of MRI detected cartilage loss in
subjects without radiographic OA [150]. Increased numbers of several
inflammatory cell populations have been detected in the OA joint capsule, of
which macrophages, T cells and mast cells are most abundant [151]. Of particular
interest, mast cells have been found at higher cell numbers in OA compared to RA
joint capsule and their mediators (histamine and tryptase) at higher concentrations
in the synovial fluid in subjects with OA compared to those with RA [151, 152].
Inflammatory and synovial cells may contribute to joint synovial and cartilage
34
inflammation by producing cytokines, prostaglandins and proteases, such as
neutrophil elastase, aggrecanases and different MMPs [153-157].
Proteases, which are essential for the degradation of cartilage molecules, can be
induced and activated by inflammatory molecules. Numerous in vitro studies have
demonstrated aggrecanase and MMP activity (including collagenase activity) and
injurious effects on cartilage structure and chondrocyte function (survival) in the
presence of pro-inflammatory cytokines [158-161]. In fact, adding tumour necrosis
factor (TNF)-α to a cartilage explant in vitro can lead to the total dissolution of the
explant mediated through the upregulation of proteases in the chondrocytes [74].
In vivo, in a mouse model, Malfait et al. [162] showed that the intra-articular
administration of TNF-α led to the rapid degradation of the cartilage aggrecan by
the induction of aggrecanase activity. The aggrecan degradation was reversible
and these findings are in line with the work of others [163]. Typically, collagen
degradation, which is irreversible in in vitro cartilage explant systems, does not
begin until after several days of inflammatory stimuli and not until most aggrecan
molecules have been degraded [79]. Further emphasising the protective role of
intact aggrecan in the in vitro cartilage explant system, aggrecanase inhibition not
only prevented interleukin (IL)-1-induced aggrecan degradation but also prevented
collagen degradation [160]. However, in more complex systems, increased
collagen degradation may parallel that of aggrecan. In a horse model, increased
synovial fluid concentrations of the C2C epitope were detected within 24 hours of
lipopolysaccharide-induced joint inflammation, suggestive of increased
collagenase activity [164]. Also suggesting increased protease activity, cross
linked peptides of type II collagen were detected in the synovial fluid early in the
acute phase of knee injury [165].
The fact that a “soup” of several pro-inflammatory cytokines or molecules leads to
more rapid cartilage degradation was furthermore observed when human and
bovine cartilage explants were exposed to both interleukin (IL)-1 and plasminogen
compared with when explants were exposed to IL-1 alone [166]. Moreover, the
combined effects of TNF-α, IL-6 and sIL-6r on human and bovine knee cartilage
explants caused more GAG release than the individual cytokines alone [161].
Inflammation may also mediate pain in OA. In a recent meta-analysis, it was
concluded that subjects with OA have modestly higher circulating CRP levels than
controls and that there is great variation in the observed associations between
different studies [167]. Circulating CRP levels appeared to be associated more
with symptoms than radiographic OA changes [167]. Moreover, synovitis is
associated with symptoms including pain [168]. The fact that intra-articular
corticosteroids supplied to osteoarthritic joints lead to a rapid reduction in joint
pain further implicates inflammation as an important mediator of pain in OA
[169]. Interestingly, this effect may be related to a reduction in synovial
35
inflammation [169]. It has, however, not been demonstrated that, by inhibiting
knee joint inflammation, corticosteroids are able to reduce cartilage degradation.
In a randomised controlled trial, investigating intra-articular hyaluronan treatment
and intra-articular hyaluronan + corticosteroid treatment, no significant MRI
progression of OA was observed in either study group during the one-year course
of the study [170].
Prostaglandins may contribute to sustained joint inflammation and pain. As proof
of this, subjects waiting for total knee replacement displayed a dose-responsive
decrease in knee synovial fluid concentrations of TNF-α, vascular endothelial
growth factor and IL-6 after two weeks of non-steroidal anti-inflammatory drug
treatment [171]. Pain and knee function improvements were also observed.
Molecules (intact or cleavage products) of cartilage ECM or yielded from
chondrocytes are able to activate toll-like receptors and complement and thereby
stimulate inflammation. It has been suggested that the inflammation associated
with OA is mediated primarily through this mechanism [74, 102]. In subjects with
OA who undergo total knee replacement, the symptoms are reduced after the
operation, which could be related to the stopped release of cartilage molecules into
the synovial fluid, leading to a reduced degree of synovial inflammation
subsequent to the operation [74]. However, subjects with higher synovial fluid
concentrations of inflammatory markers at the time of total knee replacement
experience fewer improvements in pain outcomes following surgery, underpinning
the importance of pathological hemostasis in joint tissues other than cartilage for
the clinical symptoms associated with OA (for example, synovitis) [172].
Inflammation and post-traumatic knee OA
The trauma inflicted on joint tissues at the time of an acute knee injury initiates an
immune response. This response may be exaggerated by bleeding into the joint
and the degree of acute hemarthrosis may be important for both acute and chronic
cartilage degradation by activating inflammatory pathways [173]. Typically, acute
tissue injury leads to the release of damage-associated molecular patterns
(DAMPs). These molecules, intact or proteolytically processed ECM molecules
like biglycan, tenacin-C and hyaluronic acid fragments, and cell-associated
molecules, such as high-mobility group box 1 (HMGB1) and uric acid are able to
activate immune cells by binding to pattern recognition receptors, including toll
like receptors [174-176]. In response to this activation, the immune cells release a
variety of different cytokines and chemokines that notify other cells, including
macrophages and neutrophils, of ongoing tissue injury [177, 178]. In the acute
phase of knee injury, these events can be monitored by analysing the synovial
fluid. Rapid increases in synovial fluid concentrations of pro-inflammatory
36
cytokines and proteases such as elastase and MMP-3 have been observed [153,
179, 180]. At a later stage, T-cells that have been recruited to the site of injury,
together with macrophages, may induce a state of chronic inflammation,
depending on the present cytokine environment [178]. These courses of events
after acute knee injury have not been extensively studied. However, clinical joint
effusion and investigations of synovial fluid from acutely knee-injured subjects
suggest an initial hyper-inflammatory state [146, 179-181]. Importantly, these
events are controlled by anti-inflammatory cytokines and protease inhibitors which
may maintain homeostasis during the pro-inflammatory stage and thus reduce the
proteolytic activity in cartilage [166]. The fact that both IL-1Ra and the protease
inhibitor, tissue inhibitor of matrix metalloprotease (TIMP)-1 are increased in the
synovial fluid after acute knee injury indicates that this could be the case [179,
180, 182].
Underpinning the importance of inflammation in post-traumatic OA development,
injections of corticosteroids into the joint of ACL-transected dogs during surgery
and at different time points after surgery led to reduced osteophyte size and
reduced histological severity of cartilage lesions [183]. IL-1ra administered intra
articularly to subjects with recent ACL injury, furthermore, led to a reduction in
pain and knee functional limitations, but it is not known whether such treatment
can i