Role of infrapatellar fat pad in knee osteoarthritis

Transcription

1 Role of infrapatellar fat pad in knee osteoarthritis Lipid lowering drugs as potential therapeutic strategy Proefschrift voorgelegd tot het behalen van de graad van doctor in de Wetenschappen aan de Universiteit Antwerpen te verdedigen door Stefan Clockaerts PROMOTORS prof. dr. Gerjo Van Osch prof. dr. Johan Somville CO-PROMOTOR prof. dr. Luc De Clerck Faculteit Geneeskunde en gezondheidswetenschappen Departement Orthopedische Chirurgie en Traumatologie Antwerpen

2 Role of infrapatellar fat pad in knee osteoarthritis Lipid lowering drugs as potential therapeutic strategy Stefan Clockaerts 1

3 The research for this thesis was performed in collaboration with the department of Orthopaedics of the Erasmus MC, Rotterdam The research for this thesis was performed within the framework of the Erasmus Postgraduate School Molecular Medicine and the TI Pharma consortium. Printing of this thesis and organization of the defense was financially supported by: All rights reserved University of Antwerp Universiteitsplein 1, 2610 Antwerp X Layout: Nathalie Van Dijck 2

4 Role of infrapatellar fat pad in knee osteoarthritis Lipid lowering drugs as potential therapeutic strategy Rol van infrapatellair vet in artrose van de knie Lipidenverlagende middelen als potentiële therapie Proefschrift voorgelegd tot het behalen van de graad van doctor in de Wetenschappen aan de Universiteit Antwerpen te verdedigen door Stefan Clockaerts Promotors prof. dr. Gerjo Van Osch prof. dr. Johan Somville Co-Promotor prof. dr. Luc De Clerck Voorzitter jury prof. dr. G. Hubens Juryleden prof. dr. P. Parizel prof. dr. G. Stassijns Externe juryleden prof. dr. F. Luyten prof. dr. P. Verdonk Faculteit Geneeskunde en gezondheidswetenschappen Departement Orthopedische Chirurgie en Traumatologie Antwerpen

5 4

6 To my wife Katia and to my children Lisa and Mathis, Without their support and encouragement this thesis would not have been possible. In fond memory of Marleen Goossens, For her love and support, and her belief in my succeeding. 5

7 6

8 Table of contents List of abbreviations p. 9 Chapter 1: Introduction and aims of this thesis p. 13 Chapter 2: The infrapatellar fat pad should be considered as an active p. 27 osteoarthritic joint tissue: a narrative review Chapter 3: Infrapatellar fat pad of patients with end-stage osteoarthritis p. 41 inhibits catabolic mediators in cartilage Chapter 4: Cytokine production by infrapatellar fat pad can be stimulated p. 59 by interleukin 1β and inhibited by PPARα agonist Chapter 5: Peroxisome proliferator activated receptor alpha agonist p. 81 decreases inflammatory and destructive responses in osteoarthritic cartilage Chapter 6: Statin use is associated with reduced incidence and progression p. 97 of knee osteoarthritis in the Rotterdam study Chapter 7: General discussion, conclusion and clinical perspectives for p. 115 the future Biografie en curriculum vitae p. 127 Nederlandse samenvatting p. 135 Referenties p. 149 Lekensamenvatting p. 171 Dankwoord p

9 8

10 List of abbreviations ACL anterior cruciate ligament ACTB β-actin (used as house keeping gene) ADAMTS a disintegrin and metalloproteinase with thrombospondin motifs AIA adjuvant induced arthritis B2M β2-microglobulin (used as house keeping gene) BMD bone mineral density BMI body mass index CRP C reactive protein CD cluster of differentiation CI confidence interval CTXII cross-linked C-telopeptides of type II collagen DMB dimethylmethylene blue DMEM Dulbecco s Modified Eagle Medium DMOAD disease modifying drugs for osteoarthritis DMSO dimethylsulfoxide ELISA enzyme-linked immunosorbent assay ESR erythrocyte sedimentation rate FCM fat conditioned medium FGF fibroblast growth factor GAG glycosaminoglycan GAPDH glyceraldehyde-3-phosphate dehydrogenase (used as house keeping gene) GEE generalized estimated equations GM-CSF granulocyte monocyte colony stimulating factor HPRT1 hypoxanthine phosphoribosyltransferase 1 (used as house keeping gene) IκBα inhibitor κbα IL interleukin inos inducible nitric oxide synthase IPFP infrapatellar fat pad ITS insuline-transferrin-selenium K&L Kellgren and Lawrence LIF leukemia inhibitory factor LDH lactate dehydrogenase LDL low density lipoprotein MCP monocyte chemoattractant protein MEC medical ethical committee MMP matrix metalloproteinase mpges microsomal prostaglandin synthase NAMPT nicotinamide phosphoribosyltransferase, pre-b-cell colony-enhancing factor 1 (PBEF1) or visfatin NFκB nuclear factor κb 9

11 NO nitric oxide NSAID non steroidal antiinflammatory drug OA osteoarthritis OR odd ratio PG prostaglandin PPAR peroxisome proliferator activated receptor PTGS2 prostaglandin endoperoxide synthase 2, also known as cyclooxygenase 2 RA rheumatoid arthritis RT-PCR real time polymerase chain reaction SD standard deviation spla2 soluble phospholipase A2 TGF transforming growth factor TIMP tissue inhibitor of metalloproteinases TNF tumor necrosis factor VEGF vascular endothelial growth factor 10

12 11

13 12

14 Chapter 1 Introduction and aims of this thesis 13

15 14

16 Chapter 1: Introduction and aims of this thesis Osteoarthritis (OA) is the most common joint disease worldwide, affecting 9.6% of men and 18% of women aged >60 years 1. It can occur in any joint, but is most prevalent in the hip, knee, joints of the hand, foot, ankle and spine 2. OA accounts for 3% of global years of living with disability, making it one of the leading causes of disability worldwide 3. OA is the most important cause of disability in people aged >65 years 4. In addition, OA is the second most common cause of chronic pain, after migraine 5. The occurrence of OA results in high direct and indirect costs. Direct costs such as contacts with health professionals, medical examinations, drugs, and hospital stays accounted for a mean total of 44.5 euro per OA patient per month in a cohort of city employees of Liege city 6. In addition, indirect costs such as sick-leave days were 66.3 euro per OA patient per month 6. Age is the strongest predicator for OA in general. Other risk factors for OA are more joint specific (Table 1). The United Nations estimate that the world population will increase at least by 2.5 billion between 2005 en 2050 and half of the increase will involve people over 60 years of age. The rising prevalence of obesity (Figure 1), developing nation progress and lifestyle changes also make that OA prevalence and the disease burden of OA will increase even further the following years 3. Considerable evidence indicates that OA has a multifactorial etiology with a combination of biomechanical, genetic, inflammatory and hormonal factors Abnormal biomechanical loading can be caused by joint malalignment, muscle weakness, partial or total meniscectomy (in the knee joint), obesity or peripheral neuropathy 8, 11. The increased repetitive overloading of joints leads to tissue damage or wear, or activates mechanoreceptors in chondrocytes and osteoblasts, which in turn induces chondrocyte apoptosis and activates inflammatory pathways Inflammation is known to play a role in the development of OA. Clinical features of inflammation, such as joint pain, swelling and stiffness, are present during clinical examination 16. Inflammatory mediators, such as interleukin (IL)1β and tumor necrosis factor (TNF)α, are released by cartilage, bone, synovium and other joint tissues and are capable of inducing matrix metalloproteinases (MMP), a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) 4,5 and other catabolic factors. These cartilage breakdown enzymes cause remodelling of the extracellular cartilage matrix. Inflammatory mediators also induce the production of prostaglandin E2 by stimulating the expression or activity of cyclooxygenase (COX)2, microsomal prostaglandin E2 synthase 1 (mpges1), and soluble phospholipase A2 (spla2). They also upregulate the production of nitric oxide via inducible nitric oxide synthase (inos) and induce proinflammatory cytokines such as IL1β, TNFα, IL6, leukemia inhibitory factor (LIF), IL17, IL18, and chemokines, including IL8. The expression of a number of proteins associated with a differentiated chondrocyte phenotype, including collagen type II, are suppressed by inflammatory mechanisms 8. The suppression of productions of cartilage matrix molecules (mainly aggrecan and collagen type II) and the upregulation of cartilage break down enzymes (MMPs and ADAMTS) result in an imbalance in favour of degradation. 15

17 Knee Hip Hand Age Age Age High body mass index Genetics Intense sport activities Female sex Physical activity Bone density Previous injury Hormone replacement therapy (protective) Vitamin D Quadriceps strength (protective) Malalignment High body mass index Genetics (including congenital deformities) Intense sport activities Physical activity Previous injury High body mass index Genetics Intense sport activities Grip strength Occupation 2, 44, 45 Table 1: Risk factors for incidence/progression of osteoarthritis 16

18 Chapter 1: Introduction and aims of this thesis Figure 1: Changes in prevalence of overweight and obesity in adults in selected countries (International Obesity Task Force EU Platform Briefing Paper prepared in collaboration with the European Association for the Study of Obesity March Brussels; EU Platform on Diet, Physical Activity and Health) 17

19 18 At this moment, the relative importance of biomechanical versus inflammatory pathways leading to OA is unclear. It is not unlikely that this is joint specific, but also, it is plausible that the initiating factor or underlying risk factor determines the pathogenesis of OA. In addition to biomechanical and inflammatory mechanisms, genetic abnormalities can result in an early initiation of OA. For knee and hip OA, the influence of genetic factors on the onset of OA is believed to be between 30 and 70 percent. Many of these genes affect extracellular cartilage matrix and cartilage signalling molecules, and increase the susceptibility of cartilage to loss of structure 10, 17. Cartilage damage is the hallmark of OA, however, other joint tissues such as synovium and subchondral bone are also involved (Figure 2). Synovial inflammation is in part correlated with clinical symptoms 18. Synovitis may be induced by cartilage derived factors such as IL1β or hyaluronan fragments that activate synovial macrophages 19. These macrophages respond by producing catabolic mediators that break down cartilage extracellular matrix, which in turn enhances synovial inflammation and creates a vicious cycle 20. Subchondral bone changes are a main characteristic of OA. Osteophyte formation, bone remodelling, subchondral sclerosis and attrition are visible on radiographs and are caused by biomechanical and inflammatory triggers. Several of these features are not only present in end stage OA, but also before loss of cartilage structure. It is likely that subchondral bone changes can also contribute to progression of cartilage damage and that a mechanical and biochemical (through perforations in the subchondral plate or calcified cartilage) interaction exists between cartilage and subchondral bone 21, 22.

20 Chapter 1: Introduction and aims of this thesis Figure 2: Schematic representation of the healthy versus osteoarthritic knee joint. Osteoarthritis is a disease of the whole joint, including the cartilage, synovium, subchondral bone and blood vessels

21 The most important symptom of OA is pain, experienced intermittent and worst during or after activities. It is also typically present after a period of inactivity or in the morning and resolves in less than 30 minutes. Another feature is disability with limitations in mobility and day-to-day activities. OA is associated with depression and disturbed sleep, which influences pain and disability 38. During clinical examination, joint enlargement due to joint effusion and/or bony swelling, restrictions in passive movement, crepitus, locking and joint deformities may be visible. Plain radiograph remains the gold standard in imaging of OA since it is inexpensive, fast and easily available. Radiographs can only visualize calcified bone and therefore the cartilage can only be evaluated indirectly by observing joint space. Other features of OA on radiographs are osteophytes, joint deformities and subchondral bone sclerosis. Other imaging techniques such as computed tomography (CT) scan, MRI and ultrasound have the major disadvantage of being time consuming, expensive and/or require higher radiation exposure, while adding insufficient amount of information to be used routinely for OA. Another potential technique to assess and diagnose OA is the use of biomarkers. Biomarkers are molecules generated by the joint metabolism or OA disease process and are present in synovial fluid, blood or urine. More than 20 biomarkers have already been described in literature. These are markers of cartilage, bone and synovium synthesis or degradation. Cross linked C-telopeptide of type II collagen (CTXII) in urine and cartilage oligomeric matrix protein (COMP) in serum have been studied most extensively, but similar to all biomarkers, they do not seem effective to aid diagnosis or prognosis in individual OA patients at this moment 39. An effective diagnostic and prognostic technique to evaluate OA may identify patients at risk for OA, which may benefit from (future) preventive strategies. A more precise diagnosis of OA could differentiate the choice of therapy for patients and would improve clinical studies for OA by defining the study population more precisely. Treatment of OA should aim at reducing pain and improving joint function. In addition, therapeutic strategies should prevent disease progression or restore joint tissues. Possible non-pharmacological interventions are: informing the patient about OA, improvement of social and mental wellbeing to decrease the influence of psychological co-morbidity, exercise to improve muscles and aerobic condition, and weight reduction of at least 10%. Braces, canes or other forms of joint protection, insoles, lasers, transcutaneous electrical nerve stimulation, ultrasound, electrotherapy, or acupuncture have only limited effect sizes 40. Oral analgetics such as paracetamol or non-steroidal anti-inflammatory drugs (NSAIDs) can be used to decrease pain, and if ineffective, weak opioids and narcotic analgesics can be used during a short period if no side effects subside. Topical NSAIDs have been reported to be as effective as oral NSAIDs 40. Today, no effective disease modifying drugs exist for OA. Nutraceuticals such as glucosamines and chondroitin sulphate, joint lubrificans such as hyaluronic acid, or the IL1β inhibitor diacerein have still not 20

22 Chapter 1: Introduction and aims of this thesis fully proven their efficacy in well designed clinical trials or meta-analysis. Other potential disease modifying drugs targeting cartilage catabolism (inhibitors of MMPs, ADAMTS, inos or cell signalling pathways), cartilage anabolism (e.g. transforming growth factor β, bone morphogenetic protein, fibroblast growth factor), synovial layer or subchondral bone (e.g. biphosphonates, calcitonin) are still being investigated 41. At this moment, total joint replacements are the most effective treatment available with significant decreases in out-of-pocket expenses for patients and use of drugs. In addition, studies report patient satisfaction rates for hip and knee replacement up to 100% and 75% respectively, even after 8 years 42. Taken together, although the physical and economical burden of OA is high and will increase in the future, effective disease modifying drugs are still unavailable. More knowledge is required concerning the etiological mechanisms of the OA disease process to elucidate new targets for potential disease modifying drugs for OA. 21

23 The link between obesity and OA Outcome of multiple potential underlying disease processes Obesity and high body mass index are associated with a higher risk of developing osteoarthritis 12. Changed kinetics of weight-bearing joints can lead to the initiation and progression of OA, and obesity could enhance this mechanism by increasing loading forces 23, 24. However, the risk of developing OA in non-weight-bearing joints such as joints of the hand, is also higher in obese than in healthy people This correlation indicates that alternative mechanisms might be responsible for the link between obesity and OA. There is growing evidence from epidemiological studies that the OA disease process might be influenced by metabolic diseases associated with obesity. Observational studies have described a correlation between atheromatous vascular disease and OA 32, but the causal relationship is still subject of investigation. The main hypothesis is that venous outlet obstruction and intraosseous hypertension in the subchondral bone might impair the supply of nutrients to the overlying cartilage plate and lead to an alteration in the mechanical properties of the bone. The impaired mechanical properties of the bone may in turn result in a reduced ability of the bone to absorb forces, which enhances the susceptibility of cartilage to damage during repetitive loadings 30. Hypercholesterolemia and hypertriglyceridemia have also been related to incidence and progression of OA in epidemiological studies. The correlation between dyslipidemia and OA might be the result of the immunomodulatory effects of fatty acids on cartilage. n-3 polyunsaturated fatty acids lower GAG release by cartilage under inflammatory conditions whereas n-6 polyunsaturated fatty acids increase GAG release under inflammatory conditions Alternatively, the correlation between BMI and OA might also be explained by the increased secretion of cytokines (e.g. IL6, monocyte chemoattractant protein 1) and adipokines (e.g. leptin, resistin, nicotinamide phosphoribosyltransferase) by adipose tissue in people with obesity 29. This low grade systemic inflammation may enhance inflammatory processes in the OA joint. Interestingly, the increased release of cytokines and adipokines by adipose tissue is not only due to the presence of a larger amount of adipose tissue in obese people. There is also an increased cytokine/adipokine production by visceral adipose tissue 70 compared to subcutaneous adipose tissue. This difference in endocrine behavior of adipose tissue located at different anatomical sites was confirmed by epidemiological studies investigating cardiovascular risk factors, that indicate that the ratio between visceral/subcutaneous tissue shows higher associations with cardiovascular disease than body mass index (BMI) 208. Although literature is inconclusive regarding the mechanisms leading to the difference in adipose tissue phenotype, the anatomic location of adipose tissue may play a significant role in its behavior as endocrine organ. Different responses of adipose tissues to stress related interleukin 1β (IL1β) may in part explain adipose tissue distribution and variation in production 57. In this regard, it is interesting to notice 22

24 Chapter 1: Introduction and aims of this thesis that the knee joint contains a large infrapatellar fat pad (IPFP) that might act as a third type of adipose tissue, next to visceral and subcutaneaous adipose tissue. Considering its anatomical location in the joint, the IPFP is prone to be exposed to high concentrations of IL1β present in arthritic joints. Fain et al 70 demonstrated that not just the adipocytes, but the macrophages are mainly responsible for the secretion of inflammatory mediators by adipose tissue. Macrophages are present in high amount in visceral adipose tissue. Studies indicate that macrophages also drive inflammatory and destructive responses in the OA joint 113. These macrophages were assumed to be present in the synovium, but are even likely to be present in the IPFP, making the IPFP a potential source of inflammation in the joint. This is confirmed by studies reporting different concentrations of adipokines in the synovial fluid and serum, which cannot be explained by permeability of the inflamed synovium alone; resistin for example, is present in lower amount in synovial fluid than leptin, although its molecular weight is similar 79,80,95,109. This was the first indication that production by local, intra-articular adipose tissue determines cytokine/adipokine concentration in the synovial fluid. Although numerious studies have demonstrated the influence of BMI on visceral and subcutaneous cytokine production 70, no data are available yet conconcerning the influence of BMI on cytokine or adipokine production by IPFP. 23

25 24

26 Chapter 1: Introduction and aims of this thesis Aims and scope of this thesis The association between obesity and OA exists due to different etiological mechanisms. Next to increased joint loading, metabolic and inflammatory pathways related to visceral adipose tissue, data indicate that the IPFP may act as a local source of inflammatory mediators influencing the cartilage. The aim of this thesis is to investigate: 1) the role of intra-articular adipose tissue, such as the infrapatellar fat pad (IPFP) in knee OA. 2) whether lipid lowering drugs, such as statins and fibrates, can be considered as disease modifying drugs for OA (DMOADs) because they target inflammatory and destructive processes in IPFP, but also cartilage and synovium. In Chapter 2, we performed a review of literature to examine the potential role of intraarticular adipose tissue in the OA disease process of the knee joint. Based on these results, we described the hypothesis that the infrapatellar fat pad may influence the OA disease process by the excretion of cytokines, adipokines and growth factors directly into the knee joint. We tested this hypothesis by examining the effect of the infrapatellar fat pad on inflammatory and destructive processes in cartilage (Chapter 3). We analyzed cytokines secreted by the infrapatellar fat pad and examined whether their production is influenced by body mass index or by inflammatory stimuli that are also present in osteoarthritic synovial fluid (Chapter 4). Then we analyzed whether we could counteract the production of inflammatory mediators by adding a ligand for Peroxisome Proliferator Activated Receptor (PPAR)α. PPARα ligands are lipid lowering drugs such as fibrates, that also exert anti-inflammatory effects on blood vessels, kidneys and the liver. To investigate their effect, we co-cultured infrapatellar fat pad explants with this drug (Chapter 4). In addition, we performed culture experiments adding PPARα ligands to synovial membrane (Chapter 4) and cartilage (Chapter 5) explants to examine their anti-inflammatory and anti-destructive effects on both joint tissues. To investigate the hypothesis that lipid lowering drugs might inhibit the OA disease process, we analyzed the association between statin use and OA progression of the knee and the hip in a large population based cohort study (Chapter 6). Finally, Chapter 7 presents the general discussion in which I discuss and summarize the results of the work performed and elaborate on potential applications or future research. 25

27 26

28 Chapter 2 The infrapatellar fat pad should be considered as an active osteoarthritic joint tissue: a narrative review S. Clockaerts, Y.M. Bastiaansen Jenniskens, J. Runhaar, G.J.V.M. Van Osch, J.F. Van Offel, J.A.N. Verhaar, L.S. De Clerck, J. Somville Osteoarthritis Cartilage Jul;18(7): Epub 2010 Apr 22. Review. 27

29 Abstract Introduction Osteoarthritis (OA) of the knee joint is caused by genetic and hormonal factors, and by inflammation in combination with biomechanical alterations. It is characterized by loss of articular cartilage, synovial inflammation and subchondral bone sclerosis. Considerable evidence indicates that the menisci, ligaments, periarticular muscles and the joint capsule are also involved in the OA process. This paper will outline the theoretical framework for investigating the infrapatellar fat pad as an additional joint tissue involved in the development and progression of knee-oa. Methods A literature search was performed in Pubmed from 1948 untill October 2009 with keywords intrapatellar fat pad, Hoffa fat pad, intraarticular adipose tissue, knee, cartilage, bone, cytokine, adipokine, inflammation, growth factor, arthritis, osteoarthritis. Results The infrapatellar fat pad is situated intracapsularly and extrasynovially in the knee joint. Besides adipocytes, the infrapatellar fat pad from patients with knee-oa contains macrophages, lymphocytes and granulocytes, which are able to contribute to the disease process of knee-oa. Furthermore, the infrapatellar fat pad contains nociceptive nerve fibers that could in part be responsible for anterior pain in knee-oa. These nerve fibers secrete substance P, which is able to induce inflammatory responses and cause vasodilation, which may lead to infrapatellar fat pad edema and extravasation of the immune cells. The infrapatellar fat pad secretes cytokines, interleukins, growth factors and adipokines that influence cartilage by upregulating the production of matrix metalloproteinases, stimulating the expression of pro-inflammatory cytokines and inhibiting the production of cartilage matrix proteins. They may also stimulate the production of pro-inflammatory mediators, growth factors and matrix metalloproteinases in synovium. Conclusion These data are consistent with the hypothesis that the infrapatellar fat pad is an osteoarthritic joint tissue capable of modulating inflammatory and destructive responses in knee- OA. 28

30 Chapter 2: The infrapatellar fat pad should be considered as an active osteoarthritic joint tissue: a narrative review Introduction In the recent past, knee osteoarthritis (OA) was considered as a pathologic condition that affects cartilage and bone. We now appreciate that all joint tissues, including the synovium, menisci, ligaments, periarticular muscles and the joint capsule are involved 43. In addition to these structures, the knee joint also contains adipose tissue: the infrapatellar fat pad (IPFP) 46. Although it has been known that adipose tissue secretes inflammatory mediators that are able to influence cartilage and synovium, a theoretical framework for a role of the IPFP in knee-oa has not been described 47. Considering the intraarticular location of the IPFP, the metabolic properties of adipose tissue and the fact that the OA disease process involves all the joint tissues, it is likely that the IPFP could also be involved in knee-oa. Our hypothesis is that immune cells could infiltrate IPFP as a result of knee-oa and that the combination of immune cells, nerve fibers and adipocytes could then contribute to the disease process by producing and releasing inflammatory mediators, capable of modifying inflammatory and destructive responses in cartilage and synovium (Figure 1). In this article, an overview is given of literature that supports this hypothesis. 29

31 Infrapatellar fat pad Articular cartilage INFLAMMATION Synovial layer Fat pad edema Extravasation of WBC Lymphocytes Monocytes Granulocytes Synovial fluid Interleukins Growth factors Nitric oxide Leukotrienes Prostaglandines Blood vessel Substance P Leptin Adiponectin NAMPT Resistin Adipocytes Sensory nerve (C-fibre) Articular cartilage Figure 1: A schematic overview of the infrapatellar fat pad as an active osteoarthritic joint tissue. The infrapatellar fat pad shows signs of inflammation secondary to osteoarthritis of the knee joint. The infrapatellar fat pad contains adipocytes and has an increased number of immune cells such as lymphocytes, monocytes and granulocytes that have migrated from the blood circulation. Substance P nerve fibers are also present in the infrapatellar fat pad and contribute to the immune regulation within the infrapatellar fat pad by the secretion of substance P. Substance P is able to induce extravasation of white blood cells and is also known to enhance inflammation in white blood cells. The combination of these cells is able to secrete adipokines such as leptin, adiponectin, NAMPT and resistin, but also interleukins, growth factors, nitric oxide, leukotrienes and prostaglandins which have shown to influence cartilage, synovium and osteophyte formation. Methods A literature search was performed in Pubmed from 1948 untill October 2009 using key words infrapatellar fat pad, Hoffa fat pad, intraarticular adipose tissue, knee, cartilage, bone, cytokine, adipokine, inflammation, growth factor, arthritis, osteoarthritis. We excluded papers that were not written in English. Cadaver knee joints were dissected to make photographs and figures. 30

32 Chapter 2: The infrapatellar fat pad should be considered as an active osteoarthritic joint tissue: a narrative review Results Inflammation in OA Considerable evidence indicates that OA has a multifactorial etiology with a combination of biomechanical, genetic, inflammatory and hormonal factors Abnormal biomechanical loading is an etiological factor in OA that can be caused by partial or total meniscectomy, malalignement, joint instability, muscle weakness or peripheral neuropathy, and obesity 8, 11. In addition to tissue damage or wear, repeated overloading of joints activates mechanoreceptors in chondrocytes and osteoblasts, which in turn activates inflammatory pathways that lead to the production of cartilage degradation mediators Several lines of evidence indicate that genetic abnormalities can result in an early initiation of OA. For knee-oa, the influence of genetics on the onset of OA is believed to be between 39 and 65 percent. Many of these genes affect extracellular cartilage matrix and cartilage signalling molecules 10. Inflammatory pathways are known to play a role in the development of OA. Clinical features of inflammation, such as joint pain, swelling and stiffness, are indeed present during clinical examination 16. Inflammatory mediators, such as interleukin (IL)1β and tumor necrosis factor (TNF)α, are released by cartilage, bone, synovium and other surrounding joint tissues and are capable of inducing matrix metalloproteinases, aggrecanases and other catabolic genes 8, 48. This causes remodeling of the extracellular cartilage matrix. These inflammatory mediators induce the production of prostaglandin E2 by stimulating the expression or activity of cyclooxygenase (COX)2, microsomal prostaglandin E2 synthase 1 (mpges1), and soluble phospholipase A2 (spla2) 48. They also up-regulate the production of nitric oxide via inducible nitric oxide synthase (inos) and induce proinflammatory cytokines such as IL1, TNFα, IL6, leukemia inhibitory factor (LIF), IL17, IL18, and chemokines, including IL8. The expression of a number of genes associated with a differentiated chondrocyte phenotype, including collagen type II are suppressed by inflammatory mechanisms 8, 48, 49. Transforming growth factor β produced by synovial macrophages, is able to induce the formation of osteophytes in animal studies 50, 51. Inflammatory events in the subchondral bone layer may induce hypertropic differentiation of chondrocytes or stimulate chondrocytes in a paracrine manner However, it is not known whether the subchondral bone inflammation is partly caused or stimulated by inflammatory mediators present in osteoarthritic joints. Obesity is associated with OA Obesity and high body mass index are associated with a higher incidence risk of osteoarthritis 12,24,55,56. Changed kinetics of weight-bearing joints can lead to the initiation and progression of OA, and obesity could enhance this mechanism by increasing the loading forces 23,24. However, non-weight-bearing joints such as joints of the hand also have higher incidence risk for OA in obese people compared to healthy people 25-28, 45. Fur- 31