- Email: [email protected]
Accelerated atherogenesis in autoimmune rheumatic diseases
Autoimmunity Reviews 1 (2002) 338–347
Accelerated atherogenesis in autoimmune rheumatic diseases P.A. Bacon*, R.J. Stevens, D.M. Carruthers1, S.P. Young, G.D. Kitas2 Department of Rheumatology, Division of Immunity and Infection, University of Birmingham, Birmingham B22 2TT, UK Accepted 28 August 2002
Abstract The observation that systemic inflammatory rheumatic diseases such as rheumatoid arthritis (RA) are associated with a significantly increased rate of cardiovascular disease, which often occurs at a younger age than in the normal population, is particularly important given the increasing interest in the role of inflammation in atherogenesis in the general population. This review examines the accumulating evidence for accelerated atherogenesis of RA and updates the hypothesis that vasculitis plays a major role in this. Endothelial dysfunction (ECD), widely regarded as initial lesion in atherogenesis, has been shown to occur commonly in primary vasculitis. This ECD is a diffuse event, demonstrable in more than one vascular bed. It is not simply due to scarring in the vessel wall, related to the focal inflammation of the underlying vasculitis, since it may be reversed by suppression of the immune inflammation. However, the mechanisms for this ECD differ from that of the primary vasculitis. Preliminary evidence suggests that inflammatory mediators such as CRP, TNF, or sphingolipids may be involved. The diffuse ECD of vasculitis may have important consequences for both the progression of the primary disease and for cardiovascular events. A model for the role of vasculitis-induced ECD in the accelerated atherogenesis of rheumatic diseases is presented. These concepts are discussed together with the messages they suggest for ‘idiopathic’ atherosclerosis in the general population. 䊚 2002 Elsevier Science B.V. All rights reserved. Keywords: Cardiovascular disease; Atherosclerosis; Vasculitis; Rheumatoid arthritis; Inflammation
1. Introduction The worlds of chronic rheumatic disease and acute cardiology are coming closer together. On the one hand, coronary artery disease (CAD) is *Corresponding author. Department of Rheumatology, Division of Immunity and Infection, School of Medicine, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK. Tel.: q44-121-414-6776; fax: q44-121-414-6792. E-mail address: [email protected] (P.A. Bacon). 1 City Hospital, Birmingham, UK. 2 Dudley Hospital, Dudley, UK.
increasingly viewed as an inflammatory disease which shares mechanisms in common with those of rheumatoid synovitis w1x. At the same time the evidence for inflammatory load as an independent risk factor for both future CAD and outcome in an acute coronary episode is now overwhelming w2x. On the other hand, there is increasing evidence for accelerated atherosclerosis in inflammatory rheumatic diseases such as rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE) w3x. The purpose of this article is to review the
1568-9972/02/$ - see front matter 䊚 2002 Elsevier Science B.V. All rights reserved. PII: S 1 5 6 8 - 9 9 7 2 ( 0 2 ) 0 0 1 0 0 - 3
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
evidence for the latter; examine the factors such as endothelial injury which may initiate such enhanced CAD, using primary systemic vasculitis (18SNV) as a model; and look at the implications of this for both rheumatology and cardiology. 1.1. Heart disease in RAySLE A number of studies have shown that both RA and SLE associate with an increase in overall mortality w4x. In RA, there is a decrease of 10–15 years in life expectancy compared with the general population, particularly in severe disease w5,6x, and this does not appear to have changed over the last 3–4 decades despite changes in treatment approaches w6x. Almost half of all deaths in RA are due to cardiovascular causes, with cardiovascular standardised mortality ratios reported to be between 1.2 and 5 w4x. This is predominantly due to myocardial infarction or congestive heart failure w7,8x, suggesting that the main cause is ischaemic heart disease rather than rheumatoid heart disease w3x. The latter, though common on echocardiography or necropsy w3,9x, rarely causes significant haemodynamic consequences to account for early death. A major role for cardiac ischaemia is supported by several other lines of indirect and direct evidence including the finding of accelerated peripheral (carotid) atherosclerosis in RA w10x and a high incidence of predominantly ischaemic events in an RA cohort, which interestingly could not be explained by traditional cardiac risk factors w11x. Probably the most direct evidence for ischaemic heart disease as a major common pathology in RA arises from our studies, based on objective assessments with stress myocardial perfusion imaging with SPECT. These showed that almost 50% of unselected RA patients aged 40– 70 have definitive evidence of cardiac ischaemia, compared with 27% of controls matched for all classical cardiac risk factors. This occurs approximately a decade earlier in the RA cohort suggesting that like diabetes, RA itself (or pathogenic processes associated with it), are independent risk factors for earlier ischaemic heart disease, probably due to accelerated atherosclerosis w12x. This is further supported by the finding that patients with
339
inflammatory polyarthritis do not appear to be at increased risk for the development of cardiovascular disease before the onset of their arthritis w13x. Similar findings have been observed in SLE. A high incidence of coronary heart disease in this condition is also well-established, particularly in young (-45 years) women, who are reported to be almost 50 times more likely to sustain a myocardial infarction than population controls w14–16x. 1.2. Atheroma as an inflammatory immune disease The increased incidence of CAD in RA raises questions about the mechanisms of atherosclerosis, which itself is now widely viewed as a chronic inflammatory disease w1x. The aetiology is clearly multifactorial but the association with an inflammatory immune disease places a focus on this area and the accelerated nature of the disease in RA suggests that RA itself may be important in the initiation of the process. The early phases of primary atheroma are sub-clinical, occur at undefined times, and are thus difficult to study. However, there is increasing evidence for a role for inflammation. Several studies have convincingly established that minor elevations in proteins of the acute phase response, such as CRP, are an independent risk factor for the later development of CAD in apparently healthy men w17x. The possible direct role of infection has also caused considerable interest—but no single agent has been consistently associated with CAD and the current consensus is that the total inflammatory load is the major factor w18x. In the CAD of RA, we have previously suggested inflammatory vasculitis to be a key factor w19x. There are two models which might provide the mechanisms for this. The first is the direct one whereby inflammation in the vessel wall becomes a persistent low-grade process that attracts monocytes into a cytokine rich local environment to phagocytose the cellular debris and lipid deposits, forming the basis for progressive changes in the vessel wall. This would be analogous to the accelerated atherosclerosis that complicates a very high percentage of organ grafts in both animal models
340
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
Fig. 1. 18 Vascular as a model for CVS events in RA.
and human disease w20x. The extreme rarity of vasculitis affecting coronary arteries in RA argues against this model and we hypothesized an indirect model whereby local vascular inflammation would produce a distant effect in vessels not involved in the inflammatory vasculitis. There are problems in pursuing that concept in RA itself, where the potential effects of synovial inflammation could obscure those due to vasculitis. We therefore elected to use primary vasculitis as a model to study the initiation of atherogenesis. There were also several reasons from our studies in vasculitis to suggest that unexplored damage to vascular endothelium could provide an explanation for the widespread development of organ damage and the persistent disability noted by patients on selfassessment w21x. We therefore set up a series of studies of endothelial function in 18SNV, using
well established endothelial-dependent flow studies w22x (Fig. 1). 2. EC dysfunction in 18SNV 2.1. Occurrence in medium-sized arteries and microcirculation The first attempt to ask whether endothelial dysfunction (ECD) does occur in vasculitis used brachial artery ultrasound measurements in a crosssectional study of an unselected group of 18SNV. This showed for the first time that ECD is a frequent finding in vasculitis w23x. The mechanism could relate to local arterial wall damage in the subgroup with classical polyarteritis nodosa
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
(PAN), a disease of medium-sized vessels, even though no patients had any clinical evidence of brachial artery involvement. This was unlikely to be the case in those with ANCA-associated small vessel vasculitis such as Wegeners. Such sub-group analysis required a larger study which both confirmed that EC function is depressed in 18SNV compared to matched controls and showed this to be the case in small vessel vasculitis (Wegeners and Churg-Strauss) as well as PAN w24x. This strongly supported the concept that the ECD was a diffuse response to distant vascular inflammation, not a local effect. To confirm this, EC function was also assessed in an alternative vascular bed, using laser-Doppler fluxmetry to test microcirculation responses in clinically normal skin. This again showed significantly depressed EC-dependent responses in 18SNV compared to controls, which was apparent in both small and medium vessel vasculitis. This established that diffuse ECD is occurring in vasculitis, distant to the primary inflammatory lesions of vasculitis. 2.2. Reversibility with immunosuppression These data suggested that ECD occurring in 18SNV does not necessarily represent fixed damage to the tissue. We have performed limited serial studies to examine the reversibility of ECD with treatment of active flares of 18SNV. The first looked at a small number of patients treated with our standard cyclophosphamideysteroid pulse regime to induce remission in flares and looked at EC function before and after 3–4 months therapy w23x. This showed significant normalisation after remission had been achieved. The second ongoing study at an earlier stage in therapy, looks at the alterations in ECD seen acutely with TNF blockade using Infliximab. This agent appears to induce rapid improvement in the depressed EC function, apparent within 24 h. However, this was not persistent but reverted to the pre-treatment levels by 14 days w25x. The important point is that both these therapies achieved significant improvement in EC dependent flow, showing that the ECD observed in vasculitis patients is not necessarily
341
due to fixed damage but represents a dynamic state. The conclusion we draw from these studies is that ECD, of a type that has been widely associated with CAD and frequently used as a surrogate for coronary angiography in primary atheroma w22x, is common in primary vasculitis. This ECD appears to be diffuse, since it can be detected in two vascular beds not related to the sites of the primary inflammation of 18SNV. This suggests that the mechanisms of ECD are not the same as those of the vasculitis itself. 3. Mechanisms of induction of ECD 3.1. 18SNV The mechanisms of induction of ECD in 18SNV have not yet been unravelled. The trivial explanation would be that it reflected the renal involvement which is common in 18SNV. Renal impairment, even of quite mild degree, is well documented to depress EC function w26x. Patients with renal failure had been specifically excluded from our EC function studies but previous renal involvement could have been the cause. We therefore compared Wegener’s patients with and without evidence of renal involvement at any stage— and found no difference in their levels of ECD. This led us to investigate the more interesting concept that the inflammatory mechanisms involved in induction of EC injury at the sites of active vasculitis were also involved in the diffuse EC stunning. The obvious factor to look at was ANCA, since these antibodies have a well-documented role in the pathogenesis of several major forms of 18SNV w27x. However, our data showed that ECD, assessed at either vascular bed, was not restricted to the ANCA-associated vasculitides and within the latter the level of EC response did not correlate with either ANCA status or current ANCA titre w24x. There are no other well defined aetiopathogenic inflammatory mechanisms which can be simply tested for in vasculitis, and so the relationship with a well-validated clinical activity score was investigated. In these cross-sectional
342
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
studies no statistically significant correlation was found between ECD and clinical activity. The data from the therapy studies discussed above suggests that serial data may reveal a different picture if pursued over a sufficient time interval. 3.2. Tumour necrosis factor alpha One factor in immune inflammation that is an attractive candidate for a role in the induction of EC dysfunction is tumour necrosis factor alpha (TNFa). This is a major cytokine involved with the innate immune system as well as with acute and chronic inflammatory diseases. TNFa has pleiotropic functions, including proinflammatory effects such as regulating downstream production of cytokines including interleukin IL-1, IL-6, IL-8 and GM-CSF. In RA it is crucial to driving joint inflammation and destruction. It is found in both synovial fluid and serum, and promotes resorption of cartilage and bone; activation of granulocytes; and stimulation of fibroblasts with synovial release of prostaglandins and collagenases. Recent studies showing the clinical success of the new TNFa blocking therapies have emphasised the importance of this molecule at the apex of the aetiopathogenic cytokine cascade in RA. Thus it is an obvious candidate as an initiator of the accelerated atherosclerosis of RA, particularly since TNFa has been linked to vascular injury in both acute and chronic inflammatory states. It has also been shown in vitro to specifically increase monolayer permeability through EC injury. It can directly induce endothelial cell apoptosis w28x and is involved in EC injury via neutrophil-dependent mechanisms w27x. Both TNFa and IL-1 transcriptionally activate EC genes via NF?B, leading to up-regulation of both cellular adhesion molecules, such as ICAM-1, and neutrophil attracting chemokines. These changes promote neutrophil adhesion at sites of inflammation and injury following release of superoxide and proteases from the neutrophil. 3.3. CRP—marker or mediator? TNFa is thus viewed as an important molecule
at the site of the primary inflammatory process in vasulitis. The manner in which it might be involved in the diffuse ECD distant from that site is not so clear. One interesting possibility is through the cytokine cascade initiated by TNFa. This includes IL-1 and IL-6, which are potent inducers of the acute phase response. CRP, the most frequently measured acute phase protein, not only predicts future coronary events but in one study has been shown to correlate with EC dysfunction w29x. That was true both at a single point in time and more strikingly in a serial study. Here, improvement in EC function followed within 3 months in the subgroup which had a fall in CRP from the initial higher values. This indicates that elevated CRP is a very close marker of ECD and raises the possibility that it could even be a mediator of the latter, although there is little direct evidence to support that idea. 3.4. Sphingolipids as mediator of TNF effects on EC An alternative mechanism whereby TNFa can induce effects on distant EC involves sphingolipids, which are present in all cell membranes. These have a role in TNFa-induced cell differentiation and intra-cellular signalling w30x. Thus, TNFa induces sphingomyelin hydrolysis with generation of ceramide and sphingosine-1-phosphate in EC w31x, which are released in quantity from TNFa treated EC in vitro w32x and are established as active intra-cellular signalling molecules w33x. Relevant to atherosclerosis, ceramide promotes LDL aggregation and accumulation in arterial lesions. At the endothelial level, TNFa promotes increased expression of the cell adhesion molecules E-selectin and VCAM—an effect which is imitated by ceramide w34x. Thus, sphingolipid release appears to be a key event in EC activation and inflammatory cell adhesion w30x. This appears likely to have crucial relevance to induction of the diffuse ECD at sites distant to local inflammation. At the latter, TNF may be released at sufficiently high concentration to directly activate EC. However, in the circulation, this would be rapidly
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
diluted. In contrast, both sphingomyelinase and sphingolipid breakdown products induced by the TNFa can be detected in the circulation and thus may be important to the induction of distant ECD. There are some data to show that, in vitro at least, sphingomyelinase or sphingosine can significantly depress EC dependent vasodilatation induced by thrombin or ionophore—without affecting the control nitroprusside response w35x. However, other experiments appear to contradict this and the situation requires clarification. In vivo experiments have shown that soluble sphingomyelinase can be up-regulated by LPS or cytokines, providing further support for the concept of sphingolipids as the link between inflammation and atherosclerosis. 4. Consequences of ECD 4.1. Model for initiation of atheroma in rheumatic diseases The observation that ECD is a frequent finding in 18SNV leads to speculation about the consequences of diffuse dysfunction of the blood vessels both for the field of autoimmune inflammatory rheumatic disease as well as the messages this may have for primary atheroma. A schematic of our current view of the central role of ECD in the initiation of the accelerated atherosclerosis of RA is shown in the figure. The dysfunctional endothelium is viewed as being at the centre of the early stages of this process. It may also have a role in later stages, as well as being exacerbated by some of the factors related to the primary inflammatory disease. This self-reinforcing pattern could contribute to the acceleration of atherosclerosis in RA while the frequently observed diffuse change related to vasculitis could promote the increased incidence of CAD. The central concept is that dysfunctional EC are more sensitive to other insults. Thus, they will respond more strongly to all the other diseaserelated factors, as well as classic CVS risk factors. For example, RA is associated with alterations in lipids, which occur as a result of multiple factors including inactivity, obesity, and steroid therapy
343
w36x. Some other aspects of therapy may also be important, such as elevations of homocysteine associated with the frequent use of methotrexate. In addition there is evidence that important alterations in lipid sub-fractions, together with accumulation of oxidised lipids, relate directly to the synovial inflammation w37x. The dysfunctional endothelium is predicted to be more sensitive to all these insults compared to normal endothelium. Some factors, such as oxidised lipids, have also been shown to induce endothelial injury, thus perpetuating the cycle. Additionally, endothelium in RA is exposed to the same classic CVS risk factors as the general population and at times their lifestyle can increase those risks. Thus, arthritic patients are inherently unlikely to exercise sufficiently, which also contributes to their increased body mass index that again correlates with evidence of sub-clinical CAD w12x. They also tend to be more frequent smokers than the local population. This has been associated with the incidence of vasculitis in RA, another potential vicious circle. Finally, the late stages of RA have been associated with an enhanced rate of renal impairment, itself probably multifactorial but able to contribute to EC dysfunction w26x. Thus, therapy will have to be multi-pronged as well, unless new ‘endothelial soothing drugs’ can reverse the EC dysfunction to break the loop. 4.2. Relevance of ECD to vasculitis The development of diffuse alterations in EC function is predicted to have unfortunate consequences for the activity of the underlying vasculitis itself. It offers the possibility of enhancing vasculitic inflammation by priming the endothelium for neutrophil mediated damage. In vitro systems used to investigate the mechanisms of ANCA-induced EC injury use TNFa to prime the EC w27x. This is an essential first step to up-regulate selectins and adhesion molecules, which in turn allow inflammatory cell adhesion and activation. The release of damaging oxygen radicals and proteases by inflammatory cells then occurs directly at the endothelial surface, producing local EC injury. Diffuse ECD may be acting in an analogous
344
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
manner in vivo. It is not clear at present whether ECD is accompanied by actual expression of cell adhesion molecules or whether it represents a state where the EC are primed for enhanced responses to standard EC activating agents such as cytokines. The latter are active locally at concentrations which are rapidly diluted in the circulation to levels at which they are normally inactive—but primed dysfunctional EC might still respond. At the clinical level, the existence of widespread alterations in EC function offers an explanation for the spread of inflammatory activity in systemic vasculitis from one organ site to another, which occurs in an unpredictable manner. It may also relate to the high relapse rate, which is still a major feature of systemic vasculitis despite improvements in achieving remissions. One prediction would be that ‘endothelial soothing drugs’ would diminish the relapse rate. Thus, understanding these interactions is of practical relevance to therapy. Establishing the presence of diffuse ECD as a frequent finding in SNV has other implications for the late stages of that condition, since it suggests that 18SNV patients will also have an enhanced risk of developing CAD. The hypothesis underlying our studies was that vasculitis was a major factor in initiating the accelerated CAD established to occur in RA. The data supports that concept so far—and the argument must apply equally when referred back to the 18SNV used as the model. Uncontrolled clinical observation supports the idea that atherosclerotic complications do occur in vasculitis w38x. There is as yet no epidemiological data to corroborate this—but long-term outcome figures are beginning to be collected as therapy improves the overall prognosis in 18SNV.
superseded by a developing research interest in the earlier phases of the process, fuelled by the recognition that prevention can only come from study of the early lesions. The problem is that the initiation of primary atheroma starts years before clinical symptoms, probably in late youth but at an undefined time. The advantage that vasculitis offers is the potential to study a process with an acute, clinically overt onset, which can affect patients in early adult life. This can be used to test the widely accepted hypothesis that endothelial injury associated with an inflammatory load can initiate atherogenesis. Perhaps even more importantly, it can be used to identify therapies which can ameliorate ECD and observe whether this is truly associated with a reversal of the process of accelerated atherogenesis. For example, large scale trials of a new ‘endothelial soothing drug’ to avert the onset of serious CAD could produce a definitive answer in RA within a few years. The same study might have to continue for decades in a normal population. Serial study of EC function in patients involved in such a study may well show that alteration of CAD risk can be predicted at a far earlier stage. One study has already shown that a fall in elevated CRP from the highest to the lowest quartile is associated with an improvement in EC function within 3 months w29x. The hypothesis that this will be associated with reduction in risk of real life events has yet to be formally tested—but seems intrinsically probable. The need to abandon the idea that atherogenesis is a oneway street to progression is surely inherent in the concept of atheroma as an inflammatory immune disease.
4.3. Messages for 18 atheroma from the vasculitis studies
The induction of EC stunning is likely to have a pathophysiological role—probably as a host defence mechanism. The close association of ECD with elevations of acute phase reactants, particularly CRP supports the concept that both occur in response to cytokines released by inflammatory insults. One such cytokine, TNFa, may well be involved and certainly is relevant to EC activation but this must differ in several ways from ECD.
The concept of ECD as the trigger factor for accelerated atherogenesis has important messages outside the autoimmune rheumatic disease field, especially in cardiology. The long-term focus on the late clinical features of atheroma is being
5. The concept of ‘EC dysfunction’
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
EC activation is viewed as a circumscribed event occurring in vessels exposed to local injurious events. EC stunning, by contrast, is defined as a diffuse EC response to an injurious event which may occur at a distance from the site at which ECD is assayed. This was illustrated by the experiments of the Vallance group, who first showed that ECD could be induced in a hand vein by local infusion of TNFa or IL-1 w39x—but subsequently showed that immunisation using intramuscular typhoid vaccine induced ECD distant from the site of the skin test w40x. The mediators of the latter event remain undefined at present. It is worth considering the connotations of the term ‘ECD’ in the context of immune inflammatory events. The ‘dysfunction’ is defined in terms of flow in response to a physiological stimulus and perhaps may be more appropriately thought of as altered endothelial responses. The association of ECD is with CRP elevations into the upper limit of the normal range, well below the marked elevations seen in inflammatory autoimmune diseases. Such CRP levels are seen in response to almost any infectious or injurious insult that the organism is exposed to. The CRP response is also remarkably conserved across species, indicating it has powerful survival value. This suggests that the epithet ‘dysfunction’ is misleading and that the diffuse ECD response also has evolutionary survival value, as has local EC activation. The latter is a mechanism for promoting the local adherence of circulating inflammatory cells and their subsequent migration into the tissue. This is clearly advantageous at the site of a noxious stimulus but would not be appropriate systemically. Thus downregulation of EC responses—or ‘stunning’—could act to prevent diffuse EC activation and help focus the latter at the site where it is required. It is also possible that ECD has further relevance, for example by priming EC so that they respond faster to subsequent local stimuli, or respond to a lower dose of TNFa. Examining the relevance of ECD in immuneyinflammatory rather than physiological terms may make a real contribution to understanding the development of atheroma. This is a rapidly developing field which poses at least as many questions as it answers.
345
Take-home messages ● Systemic inflammatory rheumatic diseases, such as RA and SLE, are associated with an enhanced cardiovascular mortality ● Atherogenesis in the general population is being increasingly recognised as having major inflammatory components ● Diffuse ECD is found in primary vasculitis, supporting the hypothesis that vasculitis underlies the accelerated atherogenesis in RA ● A variety of inflammatory immune mediators may be involved in the generation of such diffuse ECD, such as CRP, TNFa, and sphingolipids ● The ECD seen in primary vasculitis does not represent fixed vessel wall damage—but is reversible with effective suppression of the primary inflammatory disease ● Study of the enhanced atherogenesis of rheumatic disease appears likely to yield important information for the aetiopathogenesis and therapy of ‘idiopathic’ atheroma in the general population.
References w1x Pasceri V, Yeh ET. A tale of two diseases: atherosclerosis and rheumatoid arthritis. Circulation 1999;100:2124 –6. w2x Ridker PM. On evolutionary biology, inflammation, infection, and the causes of atherosclerosis. Circulation 2002;105:2 –4. w3x Kitas GD, Banks MJ, Bacon PA. Accelerated atherosclerosis as a cause of cardiovascular death in RA. Pathogenesis 1998;1:78 –83. w4x Goodson N. Coronary artery disease and rheumatoid arthritis. Curr. Opin. Rheumatol. 2002;14:115 –20. w5x Pincus T, Callahan LF. Taking mortality in rheumatoid arthritis seriously—predictive markers, socioeconomic status and comorbidity. J. Rheumatol. 1986;13:841 –5. w6x Wallberg-Jonsson S, Johansson H, Ohman ML, Rantapaa-Dahlqvist S. Extent of inflammation predicts cardiovascular disease and overall mortality in seropositive rheumatoid arthritis. A retrospective cohort study from disease onset. J. Rheumatol. 1999;26:2562 –71. w7x Wallberg-Jonsson S, Ohman ML, Dahlqvist SR. Cardiovascular morbidity and mortality in patients with seropositive rheumatoid arthritis in Northern Sweden. J. Rheumatol. 1997;24:445 –51.
346
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
w8x Gabriel SE, Crowson CS, O’Fallon WM. Comorbidity in arthritis. J. Rheumatol. 1999;26:2475 –9. w9x Guedes C, Bianchi-Fior P, Cormier B, Barthelemy B, Rat AC, Boissier MC. Cardiac manifestations of rheumatoid arthritis: a case–control transesophageal echocardiography study in 30 patients. Arthritis Rheum. 2001;45:129 –35. w10x Jonsson SW, Backman C, Johnson O, et al. Increased prevalence of atherosclerosis in patients with medium term rheumatoid arthritis. J. Rheumatol. 2001;28:2597 – 602. w11x del Rincon ID, Williams K, Stern MP, Freeman GL, Escalante A. High incidence of cardiovascular events in a rheumatoid arthritis cohort not explained by traditional cardiac risk factors. Arthritis Rheum. 2001;44:2737 –45. w12x Banks MJ, Flint J, Bacon PA, Kitas GD. Expression and prevalence of ischaemic heart disease in rheumatoid arthritis. Arthritis Rheum. 1998;41:S209. w13x Goodson NJ, Wiles NJ, Lunt M, et al. Increased mortality in seropositive patients during the early years of inflammatory polyarthritis. Rheumatology 2001;40(S1):71. w14x Manzi S, Meilahn EN, Rairie JE, et al. Age-specific incidence rates of myocardial infarction and angina in women with systemic lupus erythematosus: comparison with the Framingham Study. Am. J. Epidemiol. 1997;145:408 –15. w15x Urowitz MB, Bookman AA, Koehler BE, Gordon DA, Smythe HA, Ogryzlo MA. The bimodal mortality pattern of systemic lupus erythematosus. Am. J. Med. 1976;60:221 –5. w16x Ward MM. Premature morbidity from cardiovascular and cerebrovascular diseases in women with systemic lupus erythematosus. Arthritis Rheum. 1999;42:338 –46. w17x Lagrand WK, Visser CA, Hermens WT, et al. C-reactive protein as a cardiovascular risk factor: more than an epiphenomenon? Circulation 1999;100:96 –102. w18x Zhu J, Quyyumi AA, Norman JE, et al. Effects of total pathogen burden on coronary artery disease risk and Creactive protein levels. Am. J. Cardiol. 2000;85:140 –6. w19x Bacon PA, Kitas GD. The significance of vascular inflammation in rheumatoid arthritis. Ann. Rheum. Dis. 1994;53:621 –3. w20x Aranda JM, Hill J. Cardiac transplant vasculopathy. Chest 2000;118:1792 –800. w21x Bacon PA, Luqmani R. Assessment of disease, activity and damage. In: Ball E, Bridges SL, editors. Vasculitis. Oxford: Oxford University Press, 2002. p. 246 –54. w22x Raitakari OT, Celermajer DS. Flow-mediated dilatation. Br. J. Clin. Pharmacol. 2000;50:397 –404. w23x Raza K, Thambyrajah J, Townend JN, et al. Suppression of inflammation in primary systemic vasculitis restores vascular endothelial function: lessons for atherosclerotic disease? Circulation 2000;102:1470 –2. w24x Filer AD, Gardner-Medwin J, Thambyrajah J, et al. Diffuse endothelial dysfunction is common to ANCA-
w25x
w26x
w27x
w28x
w29x
w30x
w31x
w32x
w33x
w34x
w35x
w36x w37x
associated systemic vasculitis and polyarteritis nodosa. Ann Rheum Dis, in press. Stevens RJ, Gardner-Medwin J, Carruthers DM, Cheliah G, Townend JN, Bacon PA. Endothelial cell dysfunction improves acutely after Infliximab. Rheumatology 2002;41(S1):24. van Guldener C, Lambert J, Janssen MJ, Donker AJ, Stehouwer CD. Endothelium-dependent vasodilatation and distensibility of large arteries in chronic haemodialysis patients. Neph. Dial. Transplant. 1997;12(S2):14. Savage CO, Harper L, Holland M. New findings in pathogenesis of antineutrophil cytoplasm antibody-associated vasculitis. Curr. Opin. Rheumatol. 2002;14:15 – 22. Schuger L, Varani J, Marks RM, Kunkel SL, Johnson KJ, Ward PA. Cytotoxicity of tumor necrosis factoralpha for human umbilical vein endothelial cells. Lab. Invest. 1989;61:62 –8. Fichtlscherer S, Rosenberger G, Walter DH, Breuer S, Dimmeler S, Zeiher AM. Elevated C-reactive protein levels and impaired endothelial vasoreactivity in patients with coronary artery disease. Circulation 2000;102:1000 –6. Auge N, Negre-Salvayre A, Salvayre R, Levade T. Sphingomyelin metabolites in vascular cell signaling and atherogenesis. Prog. Lipid Res. 2000;39:207 –29. Xia P, Wang L, Gamble JR, Vadas MA. Activation of sphingosine kinase by tumor necrosis factor-alpha inhibits apoptosis in human endothelial cells. J. Biol. Chem. 1999;274:34499 –505. Wong ML, Xie B, Beatini N, et al. Acute systemic inflammation up-regulates secretory sphingomyelinase in vivo: a possible link between inflammatory cytokines and atherogenesis. Proc. Natl. Acad. Sci. USA 2000;97:8681 –6. Bornfeldt KE, Graves LM, Raines EW, et al. Sphingosine-1-phosphate inhibits PDGF-induced chemotaxis of human arterial smooth muscle cells: spatial and temporal modulation of PDGF chemotactic signal transduction. J. Cell Biol. 1995;130:193 –206. Modur V, Zimmerman GA, Prescott SM, McIntyre TM. Endothelial cell inflammatory responses to tumor necrosis factor alpha. Ceramide-dependent and -independent mitogen-activated protein kinase cascades. J. Biol. Chem. 1996;271:13094 –102. Murohara T, Kugiyama K, Ohgushi M, Sugiyama S, Ohta Y, Yasue H. Effects of sphingomyelinase and sphingosine on arterial vasomotor regulation. J. Lipid Res. 1996;37:1601 –8. Situnayake RD, Kitas GD. Dyslipidemia and rheumatoid arthritis. Ann. Rheum. Dis. 1997;56:341 –2. Hurt-Camejo E, Paredes S, Masana L, et al. Elevated levels of small, low-density lipoprotein with high affinity for arterial matrix components in patients with rheumatoid arthritis: possible contribution of phospholipase A2 to this atherogenic profile. Arthritis Rheum. 2001;44:2761 –7.
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347 w38x Filer AD, Nicholls D, Corston R, Carey P, Bacon PA. Takayasu arteritis and atherosclerosis: illustrating the consequences of endothelial damage. J. Rheumatol. 2001;28:2752 –3. w39x Bhagat K, Moss R, Collier J, Vallance P. Endothelial ‘stunning’ following a brief exposure to endotoxin: a
347
mechanism to link infection and infarction? Cardiovasc. Res. 1996;32:822 –9. w40x Hingorani AD, Cross J, Kharbanda RK, et al. Acute systemic inflammation impairs endothelium-dependent dilatation in humans. Circulation 2000;102:994 –9.
The World of Autoimmunity; Literature Synopsis Anti-ThyTo in Systemic Sclerosis Anti-ThyTo antibody is an antinucleolar antibody which immunoprecipitated ribonucleoproteins associated with H1y8-2 and Thy7-2 RNAs, components of the RNA processing enzymes RNAse P and RNAse MRP. Kuwana et al. (Ann Rheum Dis 2002;61:842) measured serum samples from 1048 Japanese patients with various autoimmune diseases for these antibodies. Anti-ThyTo antibodies were found in 14 of 303 patients with systemic sclerosis (4.6%) and in 7 of 745 patients without systemic sclerosis (0.9%). More patients having this disease showed immunoreactivity to hPop1 (RNAse P subunit) than those without systemic sclerosis (93% versus 14%). In addition, certain HLA molecules were found associated with anti-ThyT0 antibodies: in patients having those antibodies DRB1*1502 or 0802 were detected more often, but DRB10405DQB1*0401 haplotype were found less often than in patients without these autoantibodies. These cumulative results aid to differentiate between patients with and without systemic sclerosis, based on anti-ThyT0 response, HLA class II alleles and anti-hPop1.
IL-10 signaling in SLE Features of abnormal death signaling in lymphocytes from SLE patients have been studied by Gergely et al. (J Immunol 2002;169:1092). CD3yCD28 costimulation of peripheral blood lymphocytes elicit transient mitochondrial hyperpolarization and intracellular pH elevation, followed by increased reactive oxygen intermediates production. T cell activation-induced changes were blunted in 15 SLE patients compared with healthy donors or rheumatoid arthritis patients. Whereas interleukin-10 induced transient mitochondrial transmembrane hyperpolarization in control lymphocytes, it had opposing effects on mitochondrial signaling in lupus patients. Interleukin-10 induced reactive oxygen intermediates production and cell death in lupus peripheral blood lymphocytes, but it did not affect similarly control lymphocytes. Blockade of interleukin10 or stimulation with interleukin-12 normalized most of the induced changes. Therefore, altered interleukin10 responsiveness in SLE is associated with mitochondrial hyperpolarization, increased reactive oxygen intermediates and cytoplasmic alkalization.
Accelerated atherogenesis in autoimmune rheumatic diseases P.A. Bacon*, R.J. Stevens, D.M. Carruthers1, S.P. Young, G.D. Kitas2 Department of Rheumatology, Division of Immunity and Infection, University of Birmingham, Birmingham B22 2TT, UK Accepted 28 August 2002
Abstract The observation that systemic inflammatory rheumatic diseases such as rheumatoid arthritis (RA) are associated with a significantly increased rate of cardiovascular disease, which often occurs at a younger age than in the normal population, is particularly important given the increasing interest in the role of inflammation in atherogenesis in the general population. This review examines the accumulating evidence for accelerated atherogenesis of RA and updates the hypothesis that vasculitis plays a major role in this. Endothelial dysfunction (ECD), widely regarded as initial lesion in atherogenesis, has been shown to occur commonly in primary vasculitis. This ECD is a diffuse event, demonstrable in more than one vascular bed. It is not simply due to scarring in the vessel wall, related to the focal inflammation of the underlying vasculitis, since it may be reversed by suppression of the immune inflammation. However, the mechanisms for this ECD differ from that of the primary vasculitis. Preliminary evidence suggests that inflammatory mediators such as CRP, TNF, or sphingolipids may be involved. The diffuse ECD of vasculitis may have important consequences for both the progression of the primary disease and for cardiovascular events. A model for the role of vasculitis-induced ECD in the accelerated atherogenesis of rheumatic diseases is presented. These concepts are discussed together with the messages they suggest for ‘idiopathic’ atherosclerosis in the general population. 䊚 2002 Elsevier Science B.V. All rights reserved. Keywords: Cardiovascular disease; Atherosclerosis; Vasculitis; Rheumatoid arthritis; Inflammation
1. Introduction The worlds of chronic rheumatic disease and acute cardiology are coming closer together. On the one hand, coronary artery disease (CAD) is *Corresponding author. Department of Rheumatology, Division of Immunity and Infection, School of Medicine, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK. Tel.: q44-121-414-6776; fax: q44-121-414-6792. E-mail address: [email protected] (P.A. Bacon). 1 City Hospital, Birmingham, UK. 2 Dudley Hospital, Dudley, UK.
increasingly viewed as an inflammatory disease which shares mechanisms in common with those of rheumatoid synovitis w1x. At the same time the evidence for inflammatory load as an independent risk factor for both future CAD and outcome in an acute coronary episode is now overwhelming w2x. On the other hand, there is increasing evidence for accelerated atherosclerosis in inflammatory rheumatic diseases such as rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE) w3x. The purpose of this article is to review the
1568-9972/02/$ - see front matter 䊚 2002 Elsevier Science B.V. All rights reserved. PII: S 1 5 6 8 - 9 9 7 2 ( 0 2 ) 0 0 1 0 0 - 3
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
evidence for the latter; examine the factors such as endothelial injury which may initiate such enhanced CAD, using primary systemic vasculitis (18SNV) as a model; and look at the implications of this for both rheumatology and cardiology. 1.1. Heart disease in RAySLE A number of studies have shown that both RA and SLE associate with an increase in overall mortality w4x. In RA, there is a decrease of 10–15 years in life expectancy compared with the general population, particularly in severe disease w5,6x, and this does not appear to have changed over the last 3–4 decades despite changes in treatment approaches w6x. Almost half of all deaths in RA are due to cardiovascular causes, with cardiovascular standardised mortality ratios reported to be between 1.2 and 5 w4x. This is predominantly due to myocardial infarction or congestive heart failure w7,8x, suggesting that the main cause is ischaemic heart disease rather than rheumatoid heart disease w3x. The latter, though common on echocardiography or necropsy w3,9x, rarely causes significant haemodynamic consequences to account for early death. A major role for cardiac ischaemia is supported by several other lines of indirect and direct evidence including the finding of accelerated peripheral (carotid) atherosclerosis in RA w10x and a high incidence of predominantly ischaemic events in an RA cohort, which interestingly could not be explained by traditional cardiac risk factors w11x. Probably the most direct evidence for ischaemic heart disease as a major common pathology in RA arises from our studies, based on objective assessments with stress myocardial perfusion imaging with SPECT. These showed that almost 50% of unselected RA patients aged 40– 70 have definitive evidence of cardiac ischaemia, compared with 27% of controls matched for all classical cardiac risk factors. This occurs approximately a decade earlier in the RA cohort suggesting that like diabetes, RA itself (or pathogenic processes associated with it), are independent risk factors for earlier ischaemic heart disease, probably due to accelerated atherosclerosis w12x. This is further supported by the finding that patients with
339
inflammatory polyarthritis do not appear to be at increased risk for the development of cardiovascular disease before the onset of their arthritis w13x. Similar findings have been observed in SLE. A high incidence of coronary heart disease in this condition is also well-established, particularly in young (-45 years) women, who are reported to be almost 50 times more likely to sustain a myocardial infarction than population controls w14–16x. 1.2. Atheroma as an inflammatory immune disease The increased incidence of CAD in RA raises questions about the mechanisms of atherosclerosis, which itself is now widely viewed as a chronic inflammatory disease w1x. The aetiology is clearly multifactorial but the association with an inflammatory immune disease places a focus on this area and the accelerated nature of the disease in RA suggests that RA itself may be important in the initiation of the process. The early phases of primary atheroma are sub-clinical, occur at undefined times, and are thus difficult to study. However, there is increasing evidence for a role for inflammation. Several studies have convincingly established that minor elevations in proteins of the acute phase response, such as CRP, are an independent risk factor for the later development of CAD in apparently healthy men w17x. The possible direct role of infection has also caused considerable interest—but no single agent has been consistently associated with CAD and the current consensus is that the total inflammatory load is the major factor w18x. In the CAD of RA, we have previously suggested inflammatory vasculitis to be a key factor w19x. There are two models which might provide the mechanisms for this. The first is the direct one whereby inflammation in the vessel wall becomes a persistent low-grade process that attracts monocytes into a cytokine rich local environment to phagocytose the cellular debris and lipid deposits, forming the basis for progressive changes in the vessel wall. This would be analogous to the accelerated atherosclerosis that complicates a very high percentage of organ grafts in both animal models
340
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
Fig. 1. 18 Vascular as a model for CVS events in RA.
and human disease w20x. The extreme rarity of vasculitis affecting coronary arteries in RA argues against this model and we hypothesized an indirect model whereby local vascular inflammation would produce a distant effect in vessels not involved in the inflammatory vasculitis. There are problems in pursuing that concept in RA itself, where the potential effects of synovial inflammation could obscure those due to vasculitis. We therefore elected to use primary vasculitis as a model to study the initiation of atherogenesis. There were also several reasons from our studies in vasculitis to suggest that unexplored damage to vascular endothelium could provide an explanation for the widespread development of organ damage and the persistent disability noted by patients on selfassessment w21x. We therefore set up a series of studies of endothelial function in 18SNV, using
well established endothelial-dependent flow studies w22x (Fig. 1). 2. EC dysfunction in 18SNV 2.1. Occurrence in medium-sized arteries and microcirculation The first attempt to ask whether endothelial dysfunction (ECD) does occur in vasculitis used brachial artery ultrasound measurements in a crosssectional study of an unselected group of 18SNV. This showed for the first time that ECD is a frequent finding in vasculitis w23x. The mechanism could relate to local arterial wall damage in the subgroup with classical polyarteritis nodosa
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
(PAN), a disease of medium-sized vessels, even though no patients had any clinical evidence of brachial artery involvement. This was unlikely to be the case in those with ANCA-associated small vessel vasculitis such as Wegeners. Such sub-group analysis required a larger study which both confirmed that EC function is depressed in 18SNV compared to matched controls and showed this to be the case in small vessel vasculitis (Wegeners and Churg-Strauss) as well as PAN w24x. This strongly supported the concept that the ECD was a diffuse response to distant vascular inflammation, not a local effect. To confirm this, EC function was also assessed in an alternative vascular bed, using laser-Doppler fluxmetry to test microcirculation responses in clinically normal skin. This again showed significantly depressed EC-dependent responses in 18SNV compared to controls, which was apparent in both small and medium vessel vasculitis. This established that diffuse ECD is occurring in vasculitis, distant to the primary inflammatory lesions of vasculitis. 2.2. Reversibility with immunosuppression These data suggested that ECD occurring in 18SNV does not necessarily represent fixed damage to the tissue. We have performed limited serial studies to examine the reversibility of ECD with treatment of active flares of 18SNV. The first looked at a small number of patients treated with our standard cyclophosphamideysteroid pulse regime to induce remission in flares and looked at EC function before and after 3–4 months therapy w23x. This showed significant normalisation after remission had been achieved. The second ongoing study at an earlier stage in therapy, looks at the alterations in ECD seen acutely with TNF blockade using Infliximab. This agent appears to induce rapid improvement in the depressed EC function, apparent within 24 h. However, this was not persistent but reverted to the pre-treatment levels by 14 days w25x. The important point is that both these therapies achieved significant improvement in EC dependent flow, showing that the ECD observed in vasculitis patients is not necessarily
341
due to fixed damage but represents a dynamic state. The conclusion we draw from these studies is that ECD, of a type that has been widely associated with CAD and frequently used as a surrogate for coronary angiography in primary atheroma w22x, is common in primary vasculitis. This ECD appears to be diffuse, since it can be detected in two vascular beds not related to the sites of the primary inflammation of 18SNV. This suggests that the mechanisms of ECD are not the same as those of the vasculitis itself. 3. Mechanisms of induction of ECD 3.1. 18SNV The mechanisms of induction of ECD in 18SNV have not yet been unravelled. The trivial explanation would be that it reflected the renal involvement which is common in 18SNV. Renal impairment, even of quite mild degree, is well documented to depress EC function w26x. Patients with renal failure had been specifically excluded from our EC function studies but previous renal involvement could have been the cause. We therefore compared Wegener’s patients with and without evidence of renal involvement at any stage— and found no difference in their levels of ECD. This led us to investigate the more interesting concept that the inflammatory mechanisms involved in induction of EC injury at the sites of active vasculitis were also involved in the diffuse EC stunning. The obvious factor to look at was ANCA, since these antibodies have a well-documented role in the pathogenesis of several major forms of 18SNV w27x. However, our data showed that ECD, assessed at either vascular bed, was not restricted to the ANCA-associated vasculitides and within the latter the level of EC response did not correlate with either ANCA status or current ANCA titre w24x. There are no other well defined aetiopathogenic inflammatory mechanisms which can be simply tested for in vasculitis, and so the relationship with a well-validated clinical activity score was investigated. In these cross-sectional
342
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
studies no statistically significant correlation was found between ECD and clinical activity. The data from the therapy studies discussed above suggests that serial data may reveal a different picture if pursued over a sufficient time interval. 3.2. Tumour necrosis factor alpha One factor in immune inflammation that is an attractive candidate for a role in the induction of EC dysfunction is tumour necrosis factor alpha (TNFa). This is a major cytokine involved with the innate immune system as well as with acute and chronic inflammatory diseases. TNFa has pleiotropic functions, including proinflammatory effects such as regulating downstream production of cytokines including interleukin IL-1, IL-6, IL-8 and GM-CSF. In RA it is crucial to driving joint inflammation and destruction. It is found in both synovial fluid and serum, and promotes resorption of cartilage and bone; activation of granulocytes; and stimulation of fibroblasts with synovial release of prostaglandins and collagenases. Recent studies showing the clinical success of the new TNFa blocking therapies have emphasised the importance of this molecule at the apex of the aetiopathogenic cytokine cascade in RA. Thus it is an obvious candidate as an initiator of the accelerated atherosclerosis of RA, particularly since TNFa has been linked to vascular injury in both acute and chronic inflammatory states. It has also been shown in vitro to specifically increase monolayer permeability through EC injury. It can directly induce endothelial cell apoptosis w28x and is involved in EC injury via neutrophil-dependent mechanisms w27x. Both TNFa and IL-1 transcriptionally activate EC genes via NF?B, leading to up-regulation of both cellular adhesion molecules, such as ICAM-1, and neutrophil attracting chemokines. These changes promote neutrophil adhesion at sites of inflammation and injury following release of superoxide and proteases from the neutrophil. 3.3. CRP—marker or mediator? TNFa is thus viewed as an important molecule
at the site of the primary inflammatory process in vasulitis. The manner in which it might be involved in the diffuse ECD distant from that site is not so clear. One interesting possibility is through the cytokine cascade initiated by TNFa. This includes IL-1 and IL-6, which are potent inducers of the acute phase response. CRP, the most frequently measured acute phase protein, not only predicts future coronary events but in one study has been shown to correlate with EC dysfunction w29x. That was true both at a single point in time and more strikingly in a serial study. Here, improvement in EC function followed within 3 months in the subgroup which had a fall in CRP from the initial higher values. This indicates that elevated CRP is a very close marker of ECD and raises the possibility that it could even be a mediator of the latter, although there is little direct evidence to support that idea. 3.4. Sphingolipids as mediator of TNF effects on EC An alternative mechanism whereby TNFa can induce effects on distant EC involves sphingolipids, which are present in all cell membranes. These have a role in TNFa-induced cell differentiation and intra-cellular signalling w30x. Thus, TNFa induces sphingomyelin hydrolysis with generation of ceramide and sphingosine-1-phosphate in EC w31x, which are released in quantity from TNFa treated EC in vitro w32x and are established as active intra-cellular signalling molecules w33x. Relevant to atherosclerosis, ceramide promotes LDL aggregation and accumulation in arterial lesions. At the endothelial level, TNFa promotes increased expression of the cell adhesion molecules E-selectin and VCAM—an effect which is imitated by ceramide w34x. Thus, sphingolipid release appears to be a key event in EC activation and inflammatory cell adhesion w30x. This appears likely to have crucial relevance to induction of the diffuse ECD at sites distant to local inflammation. At the latter, TNF may be released at sufficiently high concentration to directly activate EC. However, in the circulation, this would be rapidly
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
diluted. In contrast, both sphingomyelinase and sphingolipid breakdown products induced by the TNFa can be detected in the circulation and thus may be important to the induction of distant ECD. There are some data to show that, in vitro at least, sphingomyelinase or sphingosine can significantly depress EC dependent vasodilatation induced by thrombin or ionophore—without affecting the control nitroprusside response w35x. However, other experiments appear to contradict this and the situation requires clarification. In vivo experiments have shown that soluble sphingomyelinase can be up-regulated by LPS or cytokines, providing further support for the concept of sphingolipids as the link between inflammation and atherosclerosis. 4. Consequences of ECD 4.1. Model for initiation of atheroma in rheumatic diseases The observation that ECD is a frequent finding in 18SNV leads to speculation about the consequences of diffuse dysfunction of the blood vessels both for the field of autoimmune inflammatory rheumatic disease as well as the messages this may have for primary atheroma. A schematic of our current view of the central role of ECD in the initiation of the accelerated atherosclerosis of RA is shown in the figure. The dysfunctional endothelium is viewed as being at the centre of the early stages of this process. It may also have a role in later stages, as well as being exacerbated by some of the factors related to the primary inflammatory disease. This self-reinforcing pattern could contribute to the acceleration of atherosclerosis in RA while the frequently observed diffuse change related to vasculitis could promote the increased incidence of CAD. The central concept is that dysfunctional EC are more sensitive to other insults. Thus, they will respond more strongly to all the other diseaserelated factors, as well as classic CVS risk factors. For example, RA is associated with alterations in lipids, which occur as a result of multiple factors including inactivity, obesity, and steroid therapy
343
w36x. Some other aspects of therapy may also be important, such as elevations of homocysteine associated with the frequent use of methotrexate. In addition there is evidence that important alterations in lipid sub-fractions, together with accumulation of oxidised lipids, relate directly to the synovial inflammation w37x. The dysfunctional endothelium is predicted to be more sensitive to all these insults compared to normal endothelium. Some factors, such as oxidised lipids, have also been shown to induce endothelial injury, thus perpetuating the cycle. Additionally, endothelium in RA is exposed to the same classic CVS risk factors as the general population and at times their lifestyle can increase those risks. Thus, arthritic patients are inherently unlikely to exercise sufficiently, which also contributes to their increased body mass index that again correlates with evidence of sub-clinical CAD w12x. They also tend to be more frequent smokers than the local population. This has been associated with the incidence of vasculitis in RA, another potential vicious circle. Finally, the late stages of RA have been associated with an enhanced rate of renal impairment, itself probably multifactorial but able to contribute to EC dysfunction w26x. Thus, therapy will have to be multi-pronged as well, unless new ‘endothelial soothing drugs’ can reverse the EC dysfunction to break the loop. 4.2. Relevance of ECD to vasculitis The development of diffuse alterations in EC function is predicted to have unfortunate consequences for the activity of the underlying vasculitis itself. It offers the possibility of enhancing vasculitic inflammation by priming the endothelium for neutrophil mediated damage. In vitro systems used to investigate the mechanisms of ANCA-induced EC injury use TNFa to prime the EC w27x. This is an essential first step to up-regulate selectins and adhesion molecules, which in turn allow inflammatory cell adhesion and activation. The release of damaging oxygen radicals and proteases by inflammatory cells then occurs directly at the endothelial surface, producing local EC injury. Diffuse ECD may be acting in an analogous
344
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
manner in vivo. It is not clear at present whether ECD is accompanied by actual expression of cell adhesion molecules or whether it represents a state where the EC are primed for enhanced responses to standard EC activating agents such as cytokines. The latter are active locally at concentrations which are rapidly diluted in the circulation to levels at which they are normally inactive—but primed dysfunctional EC might still respond. At the clinical level, the existence of widespread alterations in EC function offers an explanation for the spread of inflammatory activity in systemic vasculitis from one organ site to another, which occurs in an unpredictable manner. It may also relate to the high relapse rate, which is still a major feature of systemic vasculitis despite improvements in achieving remissions. One prediction would be that ‘endothelial soothing drugs’ would diminish the relapse rate. Thus, understanding these interactions is of practical relevance to therapy. Establishing the presence of diffuse ECD as a frequent finding in SNV has other implications for the late stages of that condition, since it suggests that 18SNV patients will also have an enhanced risk of developing CAD. The hypothesis underlying our studies was that vasculitis was a major factor in initiating the accelerated CAD established to occur in RA. The data supports that concept so far—and the argument must apply equally when referred back to the 18SNV used as the model. Uncontrolled clinical observation supports the idea that atherosclerotic complications do occur in vasculitis w38x. There is as yet no epidemiological data to corroborate this—but long-term outcome figures are beginning to be collected as therapy improves the overall prognosis in 18SNV.
superseded by a developing research interest in the earlier phases of the process, fuelled by the recognition that prevention can only come from study of the early lesions. The problem is that the initiation of primary atheroma starts years before clinical symptoms, probably in late youth but at an undefined time. The advantage that vasculitis offers is the potential to study a process with an acute, clinically overt onset, which can affect patients in early adult life. This can be used to test the widely accepted hypothesis that endothelial injury associated with an inflammatory load can initiate atherogenesis. Perhaps even more importantly, it can be used to identify therapies which can ameliorate ECD and observe whether this is truly associated with a reversal of the process of accelerated atherogenesis. For example, large scale trials of a new ‘endothelial soothing drug’ to avert the onset of serious CAD could produce a definitive answer in RA within a few years. The same study might have to continue for decades in a normal population. Serial study of EC function in patients involved in such a study may well show that alteration of CAD risk can be predicted at a far earlier stage. One study has already shown that a fall in elevated CRP from the highest to the lowest quartile is associated with an improvement in EC function within 3 months w29x. The hypothesis that this will be associated with reduction in risk of real life events has yet to be formally tested—but seems intrinsically probable. The need to abandon the idea that atherogenesis is a oneway street to progression is surely inherent in the concept of atheroma as an inflammatory immune disease.
4.3. Messages for 18 atheroma from the vasculitis studies
The induction of EC stunning is likely to have a pathophysiological role—probably as a host defence mechanism. The close association of ECD with elevations of acute phase reactants, particularly CRP supports the concept that both occur in response to cytokines released by inflammatory insults. One such cytokine, TNFa, may well be involved and certainly is relevant to EC activation but this must differ in several ways from ECD.
The concept of ECD as the trigger factor for accelerated atherogenesis has important messages outside the autoimmune rheumatic disease field, especially in cardiology. The long-term focus on the late clinical features of atheroma is being
5. The concept of ‘EC dysfunction’
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
EC activation is viewed as a circumscribed event occurring in vessels exposed to local injurious events. EC stunning, by contrast, is defined as a diffuse EC response to an injurious event which may occur at a distance from the site at which ECD is assayed. This was illustrated by the experiments of the Vallance group, who first showed that ECD could be induced in a hand vein by local infusion of TNFa or IL-1 w39x—but subsequently showed that immunisation using intramuscular typhoid vaccine induced ECD distant from the site of the skin test w40x. The mediators of the latter event remain undefined at present. It is worth considering the connotations of the term ‘ECD’ in the context of immune inflammatory events. The ‘dysfunction’ is defined in terms of flow in response to a physiological stimulus and perhaps may be more appropriately thought of as altered endothelial responses. The association of ECD is with CRP elevations into the upper limit of the normal range, well below the marked elevations seen in inflammatory autoimmune diseases. Such CRP levels are seen in response to almost any infectious or injurious insult that the organism is exposed to. The CRP response is also remarkably conserved across species, indicating it has powerful survival value. This suggests that the epithet ‘dysfunction’ is misleading and that the diffuse ECD response also has evolutionary survival value, as has local EC activation. The latter is a mechanism for promoting the local adherence of circulating inflammatory cells and their subsequent migration into the tissue. This is clearly advantageous at the site of a noxious stimulus but would not be appropriate systemically. Thus downregulation of EC responses—or ‘stunning’—could act to prevent diffuse EC activation and help focus the latter at the site where it is required. It is also possible that ECD has further relevance, for example by priming EC so that they respond faster to subsequent local stimuli, or respond to a lower dose of TNFa. Examining the relevance of ECD in immuneyinflammatory rather than physiological terms may make a real contribution to understanding the development of atheroma. This is a rapidly developing field which poses at least as many questions as it answers.
345
Take-home messages ● Systemic inflammatory rheumatic diseases, such as RA and SLE, are associated with an enhanced cardiovascular mortality ● Atherogenesis in the general population is being increasingly recognised as having major inflammatory components ● Diffuse ECD is found in primary vasculitis, supporting the hypothesis that vasculitis underlies the accelerated atherogenesis in RA ● A variety of inflammatory immune mediators may be involved in the generation of such diffuse ECD, such as CRP, TNFa, and sphingolipids ● The ECD seen in primary vasculitis does not represent fixed vessel wall damage—but is reversible with effective suppression of the primary inflammatory disease ● Study of the enhanced atherogenesis of rheumatic disease appears likely to yield important information for the aetiopathogenesis and therapy of ‘idiopathic’ atheroma in the general population.
References w1x Pasceri V, Yeh ET. A tale of two diseases: atherosclerosis and rheumatoid arthritis. Circulation 1999;100:2124 –6. w2x Ridker PM. On evolutionary biology, inflammation, infection, and the causes of atherosclerosis. Circulation 2002;105:2 –4. w3x Kitas GD, Banks MJ, Bacon PA. Accelerated atherosclerosis as a cause of cardiovascular death in RA. Pathogenesis 1998;1:78 –83. w4x Goodson N. Coronary artery disease and rheumatoid arthritis. Curr. Opin. Rheumatol. 2002;14:115 –20. w5x Pincus T, Callahan LF. Taking mortality in rheumatoid arthritis seriously—predictive markers, socioeconomic status and comorbidity. J. Rheumatol. 1986;13:841 –5. w6x Wallberg-Jonsson S, Johansson H, Ohman ML, Rantapaa-Dahlqvist S. Extent of inflammation predicts cardiovascular disease and overall mortality in seropositive rheumatoid arthritis. A retrospective cohort study from disease onset. J. Rheumatol. 1999;26:2562 –71. w7x Wallberg-Jonsson S, Ohman ML, Dahlqvist SR. Cardiovascular morbidity and mortality in patients with seropositive rheumatoid arthritis in Northern Sweden. J. Rheumatol. 1997;24:445 –51.
346
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347
w8x Gabriel SE, Crowson CS, O’Fallon WM. Comorbidity in arthritis. J. Rheumatol. 1999;26:2475 –9. w9x Guedes C, Bianchi-Fior P, Cormier B, Barthelemy B, Rat AC, Boissier MC. Cardiac manifestations of rheumatoid arthritis: a case–control transesophageal echocardiography study in 30 patients. Arthritis Rheum. 2001;45:129 –35. w10x Jonsson SW, Backman C, Johnson O, et al. Increased prevalence of atherosclerosis in patients with medium term rheumatoid arthritis. J. Rheumatol. 2001;28:2597 – 602. w11x del Rincon ID, Williams K, Stern MP, Freeman GL, Escalante A. High incidence of cardiovascular events in a rheumatoid arthritis cohort not explained by traditional cardiac risk factors. Arthritis Rheum. 2001;44:2737 –45. w12x Banks MJ, Flint J, Bacon PA, Kitas GD. Expression and prevalence of ischaemic heart disease in rheumatoid arthritis. Arthritis Rheum. 1998;41:S209. w13x Goodson NJ, Wiles NJ, Lunt M, et al. Increased mortality in seropositive patients during the early years of inflammatory polyarthritis. Rheumatology 2001;40(S1):71. w14x Manzi S, Meilahn EN, Rairie JE, et al. Age-specific incidence rates of myocardial infarction and angina in women with systemic lupus erythematosus: comparison with the Framingham Study. Am. J. Epidemiol. 1997;145:408 –15. w15x Urowitz MB, Bookman AA, Koehler BE, Gordon DA, Smythe HA, Ogryzlo MA. The bimodal mortality pattern of systemic lupus erythematosus. Am. J. Med. 1976;60:221 –5. w16x Ward MM. Premature morbidity from cardiovascular and cerebrovascular diseases in women with systemic lupus erythematosus. Arthritis Rheum. 1999;42:338 –46. w17x Lagrand WK, Visser CA, Hermens WT, et al. C-reactive protein as a cardiovascular risk factor: more than an epiphenomenon? Circulation 1999;100:96 –102. w18x Zhu J, Quyyumi AA, Norman JE, et al. Effects of total pathogen burden on coronary artery disease risk and Creactive protein levels. Am. J. Cardiol. 2000;85:140 –6. w19x Bacon PA, Kitas GD. The significance of vascular inflammation in rheumatoid arthritis. Ann. Rheum. Dis. 1994;53:621 –3. w20x Aranda JM, Hill J. Cardiac transplant vasculopathy. Chest 2000;118:1792 –800. w21x Bacon PA, Luqmani R. Assessment of disease, activity and damage. In: Ball E, Bridges SL, editors. Vasculitis. Oxford: Oxford University Press, 2002. p. 246 –54. w22x Raitakari OT, Celermajer DS. Flow-mediated dilatation. Br. J. Clin. Pharmacol. 2000;50:397 –404. w23x Raza K, Thambyrajah J, Townend JN, et al. Suppression of inflammation in primary systemic vasculitis restores vascular endothelial function: lessons for atherosclerotic disease? Circulation 2000;102:1470 –2. w24x Filer AD, Gardner-Medwin J, Thambyrajah J, et al. Diffuse endothelial dysfunction is common to ANCA-
w25x
w26x
w27x
w28x
w29x
w30x
w31x
w32x
w33x
w34x
w35x
w36x w37x
associated systemic vasculitis and polyarteritis nodosa. Ann Rheum Dis, in press. Stevens RJ, Gardner-Medwin J, Carruthers DM, Cheliah G, Townend JN, Bacon PA. Endothelial cell dysfunction improves acutely after Infliximab. Rheumatology 2002;41(S1):24. van Guldener C, Lambert J, Janssen MJ, Donker AJ, Stehouwer CD. Endothelium-dependent vasodilatation and distensibility of large arteries in chronic haemodialysis patients. Neph. Dial. Transplant. 1997;12(S2):14. Savage CO, Harper L, Holland M. New findings in pathogenesis of antineutrophil cytoplasm antibody-associated vasculitis. Curr. Opin. Rheumatol. 2002;14:15 – 22. Schuger L, Varani J, Marks RM, Kunkel SL, Johnson KJ, Ward PA. Cytotoxicity of tumor necrosis factoralpha for human umbilical vein endothelial cells. Lab. Invest. 1989;61:62 –8. Fichtlscherer S, Rosenberger G, Walter DH, Breuer S, Dimmeler S, Zeiher AM. Elevated C-reactive protein levels and impaired endothelial vasoreactivity in patients with coronary artery disease. Circulation 2000;102:1000 –6. Auge N, Negre-Salvayre A, Salvayre R, Levade T. Sphingomyelin metabolites in vascular cell signaling and atherogenesis. Prog. Lipid Res. 2000;39:207 –29. Xia P, Wang L, Gamble JR, Vadas MA. Activation of sphingosine kinase by tumor necrosis factor-alpha inhibits apoptosis in human endothelial cells. J. Biol. Chem. 1999;274:34499 –505. Wong ML, Xie B, Beatini N, et al. Acute systemic inflammation up-regulates secretory sphingomyelinase in vivo: a possible link between inflammatory cytokines and atherogenesis. Proc. Natl. Acad. Sci. USA 2000;97:8681 –6. Bornfeldt KE, Graves LM, Raines EW, et al. Sphingosine-1-phosphate inhibits PDGF-induced chemotaxis of human arterial smooth muscle cells: spatial and temporal modulation of PDGF chemotactic signal transduction. J. Cell Biol. 1995;130:193 –206. Modur V, Zimmerman GA, Prescott SM, McIntyre TM. Endothelial cell inflammatory responses to tumor necrosis factor alpha. Ceramide-dependent and -independent mitogen-activated protein kinase cascades. J. Biol. Chem. 1996;271:13094 –102. Murohara T, Kugiyama K, Ohgushi M, Sugiyama S, Ohta Y, Yasue H. Effects of sphingomyelinase and sphingosine on arterial vasomotor regulation. J. Lipid Res. 1996;37:1601 –8. Situnayake RD, Kitas GD. Dyslipidemia and rheumatoid arthritis. Ann. Rheum. Dis. 1997;56:341 –2. Hurt-Camejo E, Paredes S, Masana L, et al. Elevated levels of small, low-density lipoprotein with high affinity for arterial matrix components in patients with rheumatoid arthritis: possible contribution of phospholipase A2 to this atherogenic profile. Arthritis Rheum. 2001;44:2761 –7.
P.A. Bacon et al. / Autoimmunity Reviews 1 (2002) 338–347 w38x Filer AD, Nicholls D, Corston R, Carey P, Bacon PA. Takayasu arteritis and atherosclerosis: illustrating the consequences of endothelial damage. J. Rheumatol. 2001;28:2752 –3. w39x Bhagat K, Moss R, Collier J, Vallance P. Endothelial ‘stunning’ following a brief exposure to endotoxin: a
347
mechanism to link infection and infarction? Cardiovasc. Res. 1996;32:822 –9. w40x Hingorani AD, Cross J, Kharbanda RK, et al. Acute systemic inflammation impairs endothelium-dependent dilatation in humans. Circulation 2000;102:994 –9.
The World of Autoimmunity; Literature Synopsis Anti-ThyTo in Systemic Sclerosis Anti-ThyTo antibody is an antinucleolar antibody which immunoprecipitated ribonucleoproteins associated with H1y8-2 and Thy7-2 RNAs, components of the RNA processing enzymes RNAse P and RNAse MRP. Kuwana et al. (Ann Rheum Dis 2002;61:842) measured serum samples from 1048 Japanese patients with various autoimmune diseases for these antibodies. Anti-ThyTo antibodies were found in 14 of 303 patients with systemic sclerosis (4.6%) and in 7 of 745 patients without systemic sclerosis (0.9%). More patients having this disease showed immunoreactivity to hPop1 (RNAse P subunit) than those without systemic sclerosis (93% versus 14%). In addition, certain HLA molecules were found associated with anti-ThyT0 antibodies: in patients having those antibodies DRB1*1502 or 0802 were detected more often, but DRB10405DQB1*0401 haplotype were found less often than in patients without these autoantibodies. These cumulative results aid to differentiate between patients with and without systemic sclerosis, based on anti-ThyT0 response, HLA class II alleles and anti-hPop1.
IL-10 signaling in SLE Features of abnormal death signaling in lymphocytes from SLE patients have been studied by Gergely et al. (J Immunol 2002;169:1092). CD3yCD28 costimulation of peripheral blood lymphocytes elicit transient mitochondrial hyperpolarization and intracellular pH elevation, followed by increased reactive oxygen intermediates production. T cell activation-induced changes were blunted in 15 SLE patients compared with healthy donors or rheumatoid arthritis patients. Whereas interleukin-10 induced transient mitochondrial transmembrane hyperpolarization in control lymphocytes, it had opposing effects on mitochondrial signaling in lupus patients. Interleukin-10 induced reactive oxygen intermediates production and cell death in lupus peripheral blood lymphocytes, but it did not affect similarly control lymphocytes. Blockade of interleukin10 or stimulation with interleukin-12 normalized most of the induced changes. Therefore, altered interleukin10 responsiveness in SLE is associated with mitochondrial hyperpolarization, increased reactive oxygen intermediates and cytoplasmic alkalization.