Fibrates as therapy for osteoarthritis and rheumatoid arthritis? A systematic review

Abstract

Fibrates are used as lipid-lowering drugs to prevent cardiovascular pathology. Fibrates are ligands of peroxisome proliferator-activated receptor α (PPARα). Besides altering lipid metabolism, PPARα ligands exert anti-inflammatory effects on various cell types. In this study, we hypothesized that PPARα agonists exert beneficial effects on osteoarthritis (OA) and rheumatoid arthritis (RA) by their local anti-inflammatory effects, but also by their systemic influences.

A systematic literature search of Medline and EMBASE databases was performed up to August 2011. The main search items were osteoarthritis, rheumatoid arthritis, peroxisome proliferator-activated receptor alpha and fibrates. Inclusion criteria were in vivo or in vitro studies regarding humans or animals in which the effects of PPARα ligands were studied.

Six in vivo human studies, four in vivo animal studies and seven in vitro studies were included. The in vivo human studies showed all beneficial clinical effects of PPARα ligands, but studies were small and only four were randomized. Ligands for PPARα significantly reduced pain, swelling of the joints and decreased systemic inflammatory markers. In vitro and in vivo animal studies indicate that PPARα agonists inhibit bone resorption, and reduce inflammatory and destructive responses in cartilage and synovium.

PPARα agonists such as fibrates should be considered as potential therapeutic strategy for RA. There is no clinical evidence for their use in OA, although in vitro studies indicate that PPARα agonists demonstrate different joint-protective effects locally, and systemic effects on inflammation, serum lipid levels and vascular pathology. Animal studies should be performed and after confirmation of the protective effects of PPARα, large randomized controlled trials could investigate fibrates in OA and RA.

Introduction

Osteoarthritis (OA) is primarily described as a disease of cartilage damage, although there is growing evidence that synovium (hyperplasia and infiltration of monocytes), subchondral bone (remodeling) and adipose tissue (secretion of chemokines, cytokines and growth factors) are also involved [Aspden, 2008; Hunter, 2009; Clockaerts et al. 2010]. Recently, OA has been linked with atheromatous vascular disease, systemic inflammation caused by obesity, and dyslipidemia [Conaghan et al. 2005; Wang et al. 2008, 2009; Davies-Tuck et al. 2009]. Atheromatous vascular disease in the subchondral bone might enhance the OA disease process by decreasing cartilage nutrition or through direct ischemic effects on the subchondral bone [Conaghan et al. 2005]. In addition, visceral adipose tissue secretes cytokines and adipokines that are able to stimulate inflammatory and catabolic processes in the cartilage. Fatty acids have been shown to influence cartilage metabolism [Henrotin et al. 2003]. Based on these studies, OA might partly be considered as a systemic and metabolic disease [Katz et al. 2010].

Rheumatoid arthritis (RA) is a systemic immune disease, in which activated monocytes/macrophages produce proinflammatory cytokines and chemokines, and stimulate CD4+ T-cell-mediated immune reactions which induce inflammatory and destructive processes in synovium, cartilage, bone and bone marrow [Jenkins et al. 2002; Pratt et al. 2009; Schett and Firestein, 2010]. The destructive processes in cartilage might be stimulated by lipoproteins, such as low-density lipoprotein (LDL). Lipoproteins are involved in cartilage degradation in patients with RA by enhancing production of matrix metalloproteinase (MMP)-3 and reactive oxygen species (ROS) [Henrotin et al. 2003; Kakinuma et al. 2004]. Similar to OA, RA is associated with cardiovascular disease, type II diabetes mellitus and dyslipidemia, indicating that metabolic diseases are a part of the RA disease process [Shahin et al. 2010].

The search for disease-modifying drugs for OA has mainly focused on cartilage. However, the disease process in both OA and RA is probably influenced by systemic and metabolic disturbances and inflammatory processes. In this regard, peroxisome proliferator-activated receptor α (PPARα) ligands such as fibrates might be of interest. PPARα agonists reduce serum triglycerides, increase serum high-density lipoproteins and have a moderate lowering effect on serum LDLs. PPARα ligands exert anti-inflammatory effects on various cell types, including macrophages, smooth muscle cells, endothelial cells, lymphocytes and other cell types. The activation of PPARα inhibits the nuclear translocation of nuclear factor κB (NFκB) by the upregulation of inhibitor κB (IκB) [Chung et al. 2008; Lalloyer and Staels, 2010]. Other inflammatory pathways such as mitogen-activated protein kinase phosphorylation may also be influenced by fibrates [Crisafulli and Cuzzocrea, 2009]. As shown in Figure 1, fibrates may have a pleiotropic effect on OA and RA through their effect on several pathways. By exerting anti-inflammatory effects, they can decrease cartilage and bone destruction and have less synovial and systemic inflammation. In addition, fibrates decrease the influence of the dislipidemiae and vascular pathology [Findlay, 2007; Wang et al. 2008; Alagona, 2010; Clockaerts et al. 2010; Rubenfire et al. 2010; Krysiak et al. 2011].

Figure 1.

Pleiotropic effect of fibrates on osteoarthritis and rheumatoid arthritis.

We performed a systematic search for studies investigating the effects of fibrates or other PPARα ligands in in vitro and in vivo studies on OA and RA. In this study, we hypothesized that PPARα agonists exert beneficial effects on OA and RA by their anti-inflammatory effects, but also by their systemic influences.

Methods

Data sources and searches

A systematic literature search of Medline and EMBASE databases was performed for studies in English, Dutch and German up to August 2011. The main search items were osteoarthritis, rheumatoid arthritis, fibrates and PPARα. The following keywords were used: (arthritis OR osteoarthrit*) OR (cartilage OR synovium OR chondrocytes OR synoviocytes OR subchondral bone OR osteoclasts OR osteoblasts) AND ((‘peroxisome proliferator activated’ NEAR/2 receptor*) OR PPAR* OR (fibric NEAR/2 acid*) OR clofibr* OR fibrate* OR atromid OR beclobrate OR beclobrinic acid OR bezafibrate OR binifibrate OR ciprofibrate OR clinofibrate OR dulofibrate OR eniclobrate OR etofibrate OR etofylline clofibrate OR fenofibrate OR fenofibric acid OR halofenate OR metaglidasen OR methylclofenapate OR nicofibrate OR simfibrate OR tazasubrate).

Additionally, reference lists of the selected papers were screened for further publications.

Study selection

Articles were screened by their title and abstract by one observer (ICMvE) and checked by a second observer (SC). After inclusion, full text articles were read for further screening. Inclusion criteria were in vitro studies using human or animal material, in vivo animal studies, randomized controlled trials (RCTs) or clinical trials in which the effects of PPARα ligands on OA or RA were studied. Case reports, editorials, reviews and letters were excluded.

Assessment of quality and data extraction

Methodological quality of the clinical studies was assessed by one observer (ICMvE) and checked by a second observer (SC) using the quality scoring table of the Cochrane reviews with a maximum score of five points and a minimum score of 0 points representing sequence generation, allocation concealment, blinding of participants, personnel and outcome assessors, data outcome and selective outcome reporting. Quality items of in vivo animal studies were analyzed with a list of quality criteria compiled by Sniekers and colleagues [Sniekers et al. 2008]. For in vitro studies no quality items were scored. The scores of publications were noted and data were pooled independently of outcome quality.

Data of the in vivo human and animal studies were extracted by one observer (ICMvE) and in vitro data were summarized by another observer (SC). All general features, such as author, journal, year of publication, aim, type of drug used and patient information were extracted. The specific variables for in vivo human studies were level of pain [visual analog score (VAS) of the patient], scores for clinical assessment based on morning stiffness, grip strength and articular index [Ritchie et al. 1969], inflammatory markers [C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR)], and blood pressure as a marker for cardiovascular disease.

The variables extracted from the animal studies consisted of joint swelling, assessment of histology (synovial hyperplasia and infiltration of immune cells) serum parameters (such as CRP, ESR), radiographic changes and bone volume.

Data synthesis and analysis

Clinical data were standardized by calculating effect sizes [with mean ± standard deviation (SD); difference in outcomes for both groups divided by the SD of the outcome]. If the data were not expressed as mean ± SD, we entered an increase (↑) or reduction (↓) of the symptoms, or no effect (o) depending on the statistical significance of the results (p < 0.05).

The results were summarized separately for clinical effects (clinical severity and occurrence of RA or OA) and mechanistic effects (systemic inflammatory parameters, inflammation and destruction in cartilage, synovial immune cell infiltration, bone changes, improvement in systemic and local lipid profile and prevention of vascular pathology).

Results

A total of 19 publications met all the inclusion criteria: six in vivo human studies, four in vivo animal studies and nine in vitro studies (Figure 2). The six in vivo studies were scored for methodological quality resulting in a maximum score of 5. As shown in Table 1, all in vivo human studies were of poor methodological quality, resulting in a high risk of bias. The study of Goto scored best with 3 (missing clear allocation concealment and free of other bias) and the noncontrolled trial of Forster and colleagues scored worst with 0 (Table 1) [Forster and McConkey, 1986; Goto, 2010]. The animal studies were screened based on the list of quality compiled by Sniekers and colleagues [Sniekers et al. 2008]. The most prevalent shortcomings were the mentioning of dropouts, sample size number, age and body weight of the animals.

Figure 2.

Flow chart.

Table 1.

Overview of in vivo human trials.

Study	Type		Score	Patients (n)		Drug		Follow up	Outcome
				Control	Intervention	Type	Dosage		Painmarker			BPD		Function
									VAS	CRP	ESR	Sys	Dias	Grip strength	Morning stiffness	Articular index
Goto$ [2010]	RCT	RA	3	21	23	Fenofibrate	200 mg/day	6 months	0.67*	0.1	–0.03	0.06	–0.07	-	-	–
Bird et al.$ [1983]	RCT	RA	2	31‡	25	Clobuzarit	100 and 300 mg/day	24 weeks	↓*	↓*	↓*/↓*	–	–	↑*/↑*	o/↓*	o/↑*
Dixon et al.$ [1983]	RCT	RA	2	60‡	30	Clobuzarit	100 and 300 mg/day	24 weeks	–	0.10/0.85*	–	–	–	0.60/0.53	0.82/0.86*	0.70/0.75*
Reynolds et al.$ [1981]	RCT	RA	1	20	20	Clobuzarit	50 mg/day	6 months	–	↓	↓	–	–	↑	↓	↑
McConkey et al.$ [1980]	T	RA	1	–	34	Clobuzarit	50–200 mg/day §	1 year	–	↓*	↓*	–	–	–	-	–
Forster and McConkey [1986]	T	RA	0	-	3	Clobuzarit	200 mg/day	2 years	–	↓	–	–	–	–	–	–

Outcomes given as effect size (calculated with difference in outcomes for both groups divided by the standard deviation of the outcome).

^* p < 0.05; –, missing value; ↑, increase, effect size cannot be determined; ↓, reduction, effect size cannot be determined.

^$Compares with baseline only.

^‡Four groups taken together into patients given clofibrate (100 mg and 300 mg a day) and not given clofibrate.

^§Started with 50 mg/day and increased in stages to 200 mg/day.

BPD, blood pressure; co, crossover; CRP, C-reactive protein; DAS, disease activity score; Dias, diastolic; ESR, erythrocyte sedimentation rate; RA, rheumatoid arthritis; RCT, randomized controlled trial, SJC, swollen joint count; Sys, systolic; T, trial one group only; TJC, tender joint count; VAS, visual analog score.

Clinical effects

Human studies

Four randomized controlled trials and two noncontrolled trials have been reported for the use of fibrates in RA (Table 1). Goto and Bird and colleagues reported a significant reduction in VAS score in patients with RA and dislipidemia after treatment with fibrates [Bird et al. 1983, Goto, 2010]. Reynolds and colleagues mentioned a nonsignificant reduction in pain [Reynolds et al. 1981]. McConkey and colleagues demonstrated a significant reduction in clinical score (subjective clinical state score) for RA based on questionnaires in patients feeling better or worse after treatment with fibrates than before treatment [McConkey et al. 1980]. Dixon and colleagues and Bird and colleagues reported a significant improvement in the articular index [Ritchie et al. 1969] and early morning stiffness after treatment with 300 mg clobuzarit. This reduction was not seen with 100 mg clobuzarit [Bird et al. 1983; Dixon et al. 1983]. Bird and colleagues also showed a significant improvement in grip strength when 100 mg or 300 mg clobuzarit was given, which was not seen in the study by Dixon and colleagues [Bird et al. 1983; Dixon et al. 1983]. Reynolds and colleagues found no improvement in all assessments of function after the use of 50 mg clobuzarit [Reynolds et al. 1981]. No clinical studies for OA have been performed.

Animal studies

Three publications reported the clinical effect on RA through collagen-induced arthritis (CIA) or adjuvant-induced arthritis (AIA) (Table 2). Bloxham and colleagues investigated swelling of the joints in rats and found that 100 mg/kg clobuzarit significantly reduced swelling in AIA and 60 mg/kg clobuzarit significantly reduced joint swelling in CIA [Bloxham et al. 1990]. Okamoto and colleagues showed a significant reduction in swelling of joints after treatment with fenofibrate [Okamoto et al. 2005]. This was confirmed by Castillero and colleagues, who found a significant reduction in swelling and erythema in rats treated with 300 mg/kg fenofibrate [Castillero et al. 2011].

Table 2.

Overview of animal studies.

Study	Animal (strain)		Treatment	Administration	Dosage	Frequency	Age	Number	Follow up	Outcome
										Swollen joints	Infiltration immune cells	Bone	Effect
Okamoto et al. [2005]	Rats (Lewis)	AIA	Fenofibrate	Oral	10 mg/kg 100 mg/kg	Daily	7 weeks	–	18 days	↓*	↓*	–	–
Castillero et al. [2011]	Rats (Wistar)	AIA	Fenofibrate	Oral	300 mg/kg	Daily		N = 16/16/16	15 days	↓*		–	–
Bloxham et al. [1990]	Rats (AHH/R) Rats (AP1/R2)	AIA CIA	Clobuzarit	Oral	10 mg/kg 30 mg/kg 100 mg/kg 60 mg/kg	Daily		N = 5/5/5/5/5/5/5 N = 8/8/8/8/8/8/8	5 days 15 days	↓ ↓ ↓* ↓*	–	Radiographic assessment	0.97*
Still et al. [2008]	Rats (Wistar)	-	Bezafibrate	Subcutaneous injection	10 mg/kg	Daily	3 months	N = 15/15/15	12 weeks	–	–	Trabecular bone Cortical bone Bone mass	0.55* 0.96* 0.13

Outcomes given as effect size (calculated with difference in outcomes for both groups divided by the standard deviation of the outcome).

^* p < 0.05; ↓, reduction, effect size cannot be determined.

AIA, adjuvant-induced arthritis; CIA, collagen-induced arthritis.

Mechanisms

We reviewed possible etiological factors of OA and RA, such as systemic inflammation, destruction of cartilage, synovitis, bone changes, vascular pathology and altered lipid profiles, which are potentially altered by fibrates.

Reduction of systemic inflammatory parameters

McConkey and colleagues reported a significant reduction in CRP and ESR serum levels with fibrate treatment in patients with RA, while in the study by Goto, fibrates had no effect [McConkey et al. 1980; Goto, 2010]. Bird and colleagues demonstrated a significant reduction in CRP and ESR after treating patients with RA with 300 mg clofibrate, which was also shown by Dixon and colleagues for CRP [Bird et al. 1983; Dixon et al. 1983]. Reynolds and colleagues reported a nonsignificant reduction in CRP and ESR levels [Reynolds et al. 1981].

Reduction of inflammation and destruction in cartilage

None of the human trials investigated the effects of fibrates on cartilage. In addition, cartilage was never evaluated in the studies examining the effects of fibrates using animal models for OA or RA. However, seven in vitro studies investigated the effect of fibrates or other PPARα ligands on cartilage.

In in vitro studies, PPARα agonists were able to counteract the interleukin (IL)-1β-induced increase in MMP1, -3 and -13 production in cartilage [Francois et al. 2006; Clockaerts et al. 2011]. Similar effects were also observed for production of nitric oxide, prostaglandin E2 and release of glycosaminoglycans, although this was contradicted by Bordji and colleagues and Fahmi and colleagues, who reported that the addition of PPARα ligands did not counteract the effects of IL1β on the production of nitric oxide proteoglycan synthesis and MMP-13 by rat and human chondrocytes [Bordji et al. 2000; Fahmi et al. 2001; Clockaerts et al. 2011]. PPARα ligands inhibited the transforming growth factor (TGF)-β stimulated production of proteoglycans and collagen in chondrocytes [Poleni et al. 2007, 2010]. PPARα ligands also inhibited the stimulatory effect of TGF-β1 on tissue inhibitor of matrix metalloproteinase 1 gene expression and protein synthesis [Poleni et al. 2010].

Reduction in synovial immune cell infiltration

No human clinical studies have examined the effect of fibrates on synovium.

Okamoto and colleagues significantly suppressed the progression of arthritis by adding fenofibrate in an adjuvant-induced arthritis model in female Lewis rats. Histological assessment demonstrated a reduction in infiltration of immune cells in synovial tissue and less pannus formation [Okamoto et al. 2005].

In vitro data demonstrate that production of IL-6, IL-8 and granulocyte monocyte colony-stimulating factor were inhibited by PPARα ligands in IL-1β-stimulated synovial fibroblasts from patients with RA [Okamoto et al. 2005]. However, Cheng and colleagues showed no effect of Wy14643 (PPARα ligand) on IL-1-induced prostaglandin E2 production in synovial fibroblasts [Cheng et al. 2004].

Inhibition of bone changes in arthritic joints

No clinical studies have been published on the effect of fibrates on subchondral bone.

Using rats with collagen-induced arthritis, Bloxham and colleagues showed a significant reduction of erosive bone changes and new periosteal bone formation in rats treated with fibrates versus nontreated rats [Bloxham et al. 1990]. Still and colleagues demonstrated an elevated serum concentration of osteocalcin that is indicative for an increase in bone formation in rats and a significantly higher metaphyseal bone mineral density in animals treated with fibrates, while there was no significant difference between their body weights [Still et al. 2008]. Still and colleagues also showed a significantly greater trabecular bone volume and trabecular number, and total cross-sectional areas at the tibia–fibula junction were higher due to increases in cortical bone area [Still et al. 2008].

The in vitro differentiation of peripheral blood monocytes to osteoclasts is inhibited by PPARα ligands as well as the resorptive activity of mature osteoclasts [Okamoto et al. 2005; Chan et al. 2007; Still et al. 2008]. PPARα ligands are also able to increase the number of osteoblastic colonies and induce osteoblastic maturation [Jackson and Demer, 2000; Still et al. 2008].

Improvement of systemic and local lipid profile, prevention of vascular pathology

No reports have been published concerning the indirect effect of lipid lowering and prevention of vascular pathology on the OA or RA disease process. No studies have investigated the effect of fibrate therapy on synovial fluid lipid concentrations.

Discussion

This study indicates that fibrates or other PPARα agonists might lead to clinical improvement in RA. For OA, we found no clinical studies concerning these products. This overview of in vitro, animal studies and human clinical studies shows that PPARα ligands target many tissues, including tissues of the joint and its environment. This multiple targeting results in diminished systemic parameters of inflammation in patients with RA, but may also reduce pain, as indicated by decreased VAS scores, improved clinical scores and reduced swelling of the joints. PPARα ligands might also inhibit bone resorption. An effective dose of fibrates which could benefit patients with RA cannot be determined and no dose–response relationship was presented in any study.

OA has primarily been described as a ‘wear and tear’ disease with loss of cartilage structure. However, inflammation is considered an important aspect of the OA pathogenesis and therefore some etiopathogenic mechanisms might be similar for OA and RA. Levels of cytokines in the synovial fluid of patients with OA are in some cases as high as in patients with RA, and inflammatory and destructive responses have been described in OA and RA cartilage [Bondeson et al. 2010; Simopoulou et al. 2010; Sverdrup et al. 2010]. Furthermore, OA and RA are both characterized by infiltration of immune cells in the synovium. Subchondral bone alterations occur in both diseases and both diseases are not limited to the joint, but may have systemic consequences or systemic causes. Low-grade systemic inflammation has been described in both OA and RA [Hulejova et al. 2007]. In general, OA and RA have a multifactorial etiopathogenesis. Therefore, it might be possible to expect some of the observed effects of fibrates in patients with RA, and in patients with OA.

In the available studies we found no evidence for metabolic mechanisms of actions of PPARα agonists in OA and RA, although there is growing evidence that reinforces the hypothesis that PPARα agonists can prevent OA and RA by their systemic effects. First, cytokines and adipokines secreted by intra-articular adipose tissue may enhance inflammatory and destructive processes in cartilage [Clockaerts et al. 2010]. We found no studies reporting the effects of PPARα agonists on adipose tissue, but studies show that PPARα agonists lower CRP serum levels and inhibit monocyte release of IL-6, IL-1β, monocyte chemoattractant protein-1 and tumor necrosis factor α [Krysiak et al. 2011]. A recent meta-analysis concluded that short-term treatment of fenofibrate decreases CRP levels in a heterogeneous group. Although not all PPARα ligands have exactly the same pharmacokinetics, the main mechanisms of action are thought to be comparable [Miller and Spence, 1998; Ye et al. 2011]. In people with obesity, it is known that inflammation is partly caused by visceral and subcutaneous adipose tissue, and even more specific, by adipose tissue-related macrophages [Clockaerts et al. 2010]. Second, studies have reported an association between serum lipid levels and the development of OA and RA [Nurmohamed, 2007; Wang et al. 2008, 2009; Davies-Tuck et al. 2009; Nurmohamed and Dijkmans, 2009]. The concentration of cholesterol and fatty acids in serum might be correlated with synovial fluid concentrations, and these lipids might be capable of altering cartilage metabolism [Findlay, 2007; Wang et al. 2008]. Therefore, it is likely that PPARα agonists may have a preventive effect on the development of RA and OA through their lipid-lowering modalities. Finally, the link between vascular pathology and OA is becoming more and more established, and vascular pathology is also an important consequence of RA. A recent study by Hoeven and colleagues showed an association between the intima media thickness (or carotic plaque) and the prevalence of OA in the knee, distal interphalangeal and metacarpophalangeal OA in women [Hoeven et al. 2012].

Since PPARα agonists such as fibrates are known to prevent atherosclerosis by altering lipid composition and by their anti-inflammatory effects, they have the potential to prevent OA by counteracting this underlying risk factor [Alagona, 2010; Ji et al. 2010; Rubenfire et al. 2010].

Other lipid-lowering drugs, such as statins, are also shown to exert pleiotropic effects [Ridker et al. 2009]. Several studies already described that statins have potential beneficial effects on RA and OA [Paraskevas, 2008; Lazzerini et al. 2010; Baker et al. 2011] and this was also observed by Clockaerts and colleagues in the Rotterdam Study, a large open population study, where there was a 50% decrease in incidence and progression of knee OA in statin users [Clockaerts et al. 2012]. Statins and fibrates might exhibit similar potential beneficial effects on the joint and on systemic metabolic dysregulations. However, fibrates are drugs with a higher bioavailability compared with statins. For example, fenofibrate is rapidly converted into its active metabolite, fenofibric acid, which is protein bound (primarily to albumin) for 99%. Plasma levels of fenofibric acid peak after 6–8 h after oral administration and it has a half life of 20 h and may therefore reach the joint in higher concentrations and over a longer period [Tziomalos and Athyros, 2006]. Statins, however, are largely taken up by the liver and undergo first-pass metabolism [Lazzerini et al. 2010]. The overall tolerability of fibrates compared with statins or the combination shows no difference, and there is no difference in adverse advents [Farnier et al. 2012].

Despite the availability of RCTs investigating the effect of fibrates on RA, we could not perform a meta-analysis due to the differences in outcome parameters used. We found no clinical studies investigating the effect of PPARα ligands on OA. Other limitations of our study include the use of data of different kinds of species. The effects of PPARα ligands are species dependent, which can influence the results. This is however a well-known problem in in vitro research.

Well-designed RCTs are needed to reinforce the potential inhibitory effect of fibrates on OA and RA development and progression, but only after examination of potential underlying mechanisms, such as the role of systemic inflammation caused by obesity, vascular pathology and dyslipidemia. Spontaneous OA and RA animal models with metabolic alterations such as vascular pathology and disturbed lipid metabolism should be performed to investigate the effect of fibrates on OA and RA development and progression. These studies will help to determine the inclusion criteria for a well-designed human randomized clinical trial to investigate the effect of fibrates on OA and RA development.

Conclusion

OA and RA have a complex and multifactorial etiology. PPARα agonists such as fibrates may prevent disease progression by their ability to modify cartilage, synovium and bone responses in both diseases. For RA, fibrates seem to improve pain and function. For OA, however, there is no clinical evidence.

The effect of PPARα agonists on OA or RA development through the lowering of serum or intra-articular lipid levels, the prevention of vascular pathology or the inhibition of cytokine or adipokine production remains to be elucidated. Animal models for OA and RA, with metabolic alterations, should be performed. After confirmation of a protective effect on joints, well-designed RCTs, using well-defined inclusion criteria, can be started to investigate the symptom improvement and disease-modifying properties of fibrates in OA and RA.

Key messages

1. PPARα agonists should be considered as a potential therapeutic strategy for OA and RA.

2. PPARα agonists seem to demonstrate different joint-protective effects locally and systematically.

3. Animal models for OA and RA with metabolic alterations should be performed.