*

m

RHEUMATOID ARTHRITIS

1

Systemic NonarticularManifestations of Rheumatoid

Arthritis: Focus on Inflammatory Mechanisms

Larry W. Moreland, MD,* and Jeffrey R. Curtis, MD, MPH†

Objective: Extra-articular (“nonarticular”) manifestations of rheumatoid arthritis (RA) are com-mon and greatly affect physical and emotional health, as well as prognosis, including survival.Several plausible mechanisms have been advanced for many nonarticular manifestations but thereis increasing evidence that pro-inflammatory cytokines (eg, tumor necrosis factor-alpha [TNF-�],interleukin [IL]-1, and IL-6) are also involved. The purpose of this review is to provide a conciseappraisal of recent studies investigating the involvement of inflammatory cytokines in the patho-genesis of nonarticular RA manifestations.Methods: A Medline search for articles published between January 1995 and October 2007 wasconducted using the following keywords: rheumatoid arthritis, anemia, cardiovascular, atheroscle-rosis, bone loss, osteopenia, osteoporosis, pulmonary, thrombocytopenia, lymphadenopathy, ker-atoconjunctivitis sicca, uveitis, scleritis, keratitis. The review focused on articles describing apotential role of inflammatory mediators in these conditions.Results: Studies of many nonarticular manifestations strongly implicate pro-inflammatory cyto-kines and specific mechanisms by which these mediators are likely to act have even been elucidated.The inflammatory cytokines implicated are numerous but particularly include members of theTNF family and the interleukins, particularly IL-1 and IL-6. In bone loss, activated T-cells havebeen shown to express pro-inflammatory cytokines (eg, TNF, IL-1, IL-7, and IL-17) that differ-entially upregulate and downregulate mechanisms that mediate the balance between bone resorp-tion and formation. Cytokine-mediated inflammation has also been implicated, for example, inthe early stages of atherogenesis and this may explain the observed increase in cardiovasculardisease among patients with RA. However, for some nonarticular manifestations, the associationwith pro-inflammatory cytokines has been less firmly established and potential mechanisms aremore speculative.Conclusions: Overall, further research in this area will add to our understanding of the mechanismsof extra-articular manifestations in RA patients. These insights should allow clinicians to selecttherapies to better match the spectrum of joint disease and nonarticular manifestations in indi-vidual patients. This may be particularly relevant for newer biologic agents with specific inhibitoryeffects on cytokines such as TNF-� and IL-6.© 2008 Elsevier Inc. All rights reserved. Semin Arthritis Rheum 39:132-143Keywords: rheumatoid arthritis, extra-articular, nonarticular, systemic, manifestations, complications,cytokines, pro-inflammatory, RA, TNF-�, IL-6

Margaret Jane Miller Endowed Professor for Arthritis Research, University of Pittsburgh, Pittsburgh, PA.†Assistant Professor of Medicine, University of Alabama at Birmingham, Birmingham, AL.

Address reprint requests to: Larry Moreland, MD, University of Pittsburgh, S711 Biomedical Science Tower, 3500 Terrace Street, Pittsburgh, PA 15261. E-mail:orelandl@dom.pitt.edu.

32 0049-0172/09/$-see front matter © 2008 Elsevier Inc. All rights reserved.doi:10.1016/j.semarthrit.2008.08.003

Er1tFfl(tnprcdrssreootets

fit

ts

ewIwctacmaa

tbmfl

L.W. Moreland and J.R. Curtis 133

xtra-articular (or “nonarticular”) manifestations ofrheumatoid arthritis (RA) are common in Westerncountries (1,2). A retrospective review of medical

ecords of 609 cases of RA at the Mayo Clinic (1955-994) found at least 1 extra-articular disease manifesta-ion (eg, pericarditis, pleuritis, cutaneous vasculitis,elty’s syndrome, neuropathy, ophthalmological mani-estations, glomerulonephritis, and other types of vascu-itis) in 247 patients (41%) over a mean of 11.8 yearsrange, 0.1 to 42.8 years) (3). This study also found thathe incidence of severe extra-articular manifestations hadot changed significantly over the decades. However, therevalence of extra-articular disease in this study may noteflect current prevalence rates, since clinical practice hashanged over the last decade and the study period pre-ated the use of more aggressive disease-modifying anti-heumatic drug (DMARD) therapy. The introduction ofuch treatments as well as the results of experimental re-earch has led to a greater appreciation of the importantole that inflammatory mediators play in RA and in thextra-articular manifestations of this disease. While previ-usly it was thought that extra-articular manifestationsccur late in the disease process (4), the recognition thathere is a common mechanism between RA itself andxtra-articular manifestations has led to a growing accep-ance that extra-articular manifestations are no longereen only in the most severe cases of RA.

Table 1 summarizes published prevalence rates for dif-erent extra-articular manifestations of RA observed byndividual studies (2,3,5-14). These suggest that consti-utional symptoms such as fatigue and cardiac and hema-

Abbreviations

BFU-E Erythroid burst-forming unitsBMD Bone mineral densityCRP C-reactive proteinCTLA4 Cytotoxic T-lymphocyte antigen 4CV CardiovascularCVD Cardiovascular diseaseDMARD Disease-modifying antirheumatic drugESR Erythrocyte sedimentation rateFDA Food and Drug AdministrationG-CSF Granulocyte colony stimulating factorHDL High-density lipoproteinHR Hazard ratioIL InterleukinILD Interstitial lung diseaseMI Myocardial infarctionNFAT Nuclear factor of activated T-cellsNSAID Nonsteroidal anti-inflammatory drugRA Rheumatoid arthritisRANKL Receptor activator for nuclear factor-êB

ligandTNF Tumor necrosis factorTNFRSF1A Tumor necrosis factor receptor

isuperfamily, member 1A

ological diseases appear to be among the most commonystemic associations with RA.

Nonarticular RA manifestations have a wide range offfects on physical, cognitive, and emotional well-being asell as the prognosis and survival for RA patients (1,15).

n 1 retrospective medical record review of 424 patientsith RA, extra-articular manifestations of RA were asso-

iated with a markedly decreased survival compared withhe general population and RA patients without extra-rticular manifestations (P � 0.001), prompting the con-lusion that extra-articular disease is a major predictor ofortality in patients with RA (1). Reduced survival has

lso been positively correlated with the severity of extra-rticular manifestations.

In recent years, the role of inflammatory mediators inhe development of nonarticular complications of RA haseen increasingly recognized. Indeed, these systemicanifestations appear to be driven by the same pro-in-

ammatory cytokines (eg, tumor necrosis factor [TNF],

Table 1 Selected Extraarticular Manifestations Reportedin Patients with RA and Prevalence (2,3,5,10,26,27,32,74,80-83)

Manifestations Examples

PublishedPrevalenceEstimates

Hematologic Hypochromic-microcyticanemia with low serumferritin and low ornormal iron-bindingcapacityFelty syndrome

33 to 60%a

Constitutionalsymptoms

Fatigue 80 to 93%

Bone Osteoporosis 22%b

OsteopeniaCardiac Pericarditis 18%c, 30%d

Heart failure/cardiomyopathy/leftventricular failure

Myocardial infarctionStroke

Ocular Uveitis RareScleritisPeripheral ulcerative

keratitisPulmonary Pleurisy �5%e

Parenchymal nodulesInterstitial involvementAirway diseasePulmonary vasculitis

(rare)

Notes: a, mild anemia; b, prevalence of osteoporosis (T-score ��2.5) at either the femoral neck or the lumbar spine or at bothsites in patients with mean disease duration of 15.4 years; c,angina; d, cardiac chest pain in patients with mean diseaseduration of 12.5 years; e, prevalence of pleural clinical disease(40 to 75% demonstrate pleural involvement postmortem).

nterleukin [IL]-1, and IL-6) responsible for joint damage

(tmmim((acuiit((aCc

tmpipa

M

TvR(ltolsbssorit

R

Tasf

A

Asc

tR(acm

baswei(iTcsit

hmhlfcla(wiaomcwccrbt

etfsttlwcoaas

134 Systemic nonarticular manifestations of RA

9,16,17). This understanding is critical to the applica-ion of newer therapeutic agents that inhibit the inflam-atory pathways associated with RA and the develop-ent of agents under investigation. Currently, 3 TNF

nhibitors are approved by the U.S. Food and Drug Ad-inistration (FDA) for the treatment of RA: etanercept

Enbrel®), infliximab (Remicade®), and adalimumabHumira®). Anakinra (Kineret®), an IL-1 inhibitor, islso approved by the FDA for the treatment of RA. To-ilizumab (ActemraTM), an IL-6 receptor inhibitor, isnder investigation for the treatment of RA; the agent

nhibits the IL-6 receptor complex and thus blocks IL-6-nitiated signaling, which may mediate some manifesta-ions of extra-articular disease. In addition, rituximabRituxan®), which depletes B-cells, and abataceptOrencia®), a cytotoxic T-lymphocyte antigen 4 (CTLA4)ntibody that inhibits T-cell activation by binding toD80 and CD86, can also decrease inflammation and

ytokine activity (18).The major objective of the present review is to survey

he recent literature with regard to systemic nonarticularanifestations of RA with a focus on inflammatory

athological mechanisms. A secondary aim of this reviews to place the above-mentioned biological therapies inerspective with regard to their potential to treat the non-rticular manifestations of RA.

ETHODS

o outline the role of inflammatory mediators in the de-elopment of common extra-articular manifestations ofA, the present study searched Medline for recent articles

published between January 1995 and October 2007) re-ated to this topic. These keywords were rheumatoid ar-hritis, anemia, cardiovascular, atherosclerosis, bone loss,steopenia, osteoporosis, pulmonary, thrombocytopenia,ymphadenopathy, keratoconjunctivitis sicca, uveitis,cleritis, and keratitis. Articles were selected if they hadeen published as full journal articles in English (ab-tracts, poster presentations, conference proceedings, orimilar reports were excluded). Articles could be eitherriginal research papers or review articles but needed toelate to either disease mechanisms of nonarticular man-festations in RA or treatment studies that shed light onhese potential mechanisms.

ESULTS

he literature search as described in Methods highlightednumber of published studies. These are presented in this

ection according to the individual extra-articular mani-estations presented in the literature.

nemia

nemia in RA patients is well described and several “clas-ical” pathological mechanisms may explain this compli-

ation. First, iron-deficiency anemia can occur secondary m

o occult or chronic bleeding caused by therapies used forA, particularly nonsteroidal anti-inflammatory drugs

NSAIDs). The risk for gastrointestinal bleeding associ-ted with NSAIDs is worsened with concomitant glu-ocorticoids. An anemia of chronic disease is also com-on among RA patients.Several studies identified by this review point to possi-

le inflammatory-mediated mechanisms that may medi-te the anemia of chronic disease. First, a large-scale pro-pective database review showed that in 2120 patientsith RA, anemia was closely associated with elevated lev-

ls of acute-phase reactants (particularly erythrocyte sed-mentation rate [ESR] and C-reactive protein [CRP])14). Several studies have also provided more specific ev-dence that implicates pro-inflammatory cytokines (eg,NF-�, IL-1, and IL-6) in the pathogenesis of anemia of

hronic disease among patients with RA (9,16). In 1tudy, serum levels of TNF-�, IL-1, and IL-6 were signif-cantly higher among anemic RA patients, compared withheir nonanemic counterparts (16).

Genetic polymorphisms in the TNF-� gene (majoristocompatibility complex on chromosome 6p21.3)ay also play a role, as shown in a study of longitudinal

emoglobin data collected from 262 patients with estab-ished or early RA (19). In this study, the tumor necrosisactor receptor I (TNFRSF1A) GG genotype was signifi-antly associated with iron deficiency anemia in estab-ished RA (OR, 4.3; P � 0.01), while there was a weakssociation of the G allele with anemia of chronic diseaseOR, 2.2; P � 0.04). This suggests that polymorphismsithin the TNFRSF1A gene are indeed associated with

ron deficiency anemia and, possibly to a lesser extent,nemia of chronic disease in patients with RA (19). An-ther study has examined a TNF-� gene promoter poly-orphism in RA patients (20). In patients with RA, cir-

ulating levels of TNF-� and soluble TNF receptor type Iere significantly increased in RA patients with anemia

ompared with those without anemia (20). In addition,irculating TNF-� and soluble tumor necrosis factoreceptor type I levels correlated negatively with hemoglo-in concentrations, further confirming a role of TNF inhe pathology of anemia of chronic disease in RA.

Another study found that TNF-� production affectsrythropoiesis in patients with active RA, and this inhibi-ion may be ameliorated with treatment (21). This studyound that erythroid burst-forming units (BFU-Es) wereignificantly lower in RA patients (and particularly inhose with anemia of chronic disease) compared with con-rols. Increased TNF-� levels were also inversely corre-ated with BFU-Es and hemoglobin levels and positivelyith the percentage of cells apoptotic to surface cell gly-

oproteins (such as CD34�/CD71�) normally foundn hematopoietic cells. Furthermore, administration ofnti-TNF-� antibody (cA2) increased BFU-E numbersnd apparently normalized hematopoiesis, thus furtherubstantiating the inhibitory role of TNF-� (21). In sum-

ary, TNF-�-mediated apoptotic depletion of bone mar-

rd

H

HclbaetiFup(

C

Scr(cdaolfl(cmplc

tvlTeeam

T

FRDcaiTIuabnt

F

Isomeab

C

Faaaiyetsdtsatfdadtc

Fprdir2o

L.W. Moreland and J.R. Curtis 135

ow erythroid cells may contribute to anemia of chronicisease in RA.

epcidin

epcidin is a liver-derived acute phase reactant thatlosely regulates iron homeostasis by controlling the re-ease of iron from cells. Hepcidin production is increasedy inflammation via pro-inflammatory cytokines and hasn inverse relationship with physiologic hemoglobin lev-ls. Stimulated hepcidin production may be a direct con-ributor to anemia of chronic disease in RA by interrupt-ng the delivery of iron to erythroid precursor cells (22).urthermore, IL-6 acts directly on hepatocytes to stim-late hepcidin production during inflammation (23), arocess that results in hypoferremia of inflammationFig. 1) (24).

linical Aspects

tudies have shown that, in RA patients, anemia ofhronic disease can be safely and effectively treated usingecombinant human erythropoietin and intravenous iron25,26). On the other hand, it has been argued that spe-ifically targeting mild-to-moderate anemia of chronicisease may have little clinical relevance. Support for thisrgument comes from different sources, including resultsf a study that found that improvement in anemia fol-owed effective RA treatment (combinations of either in-iximab � methotrexate or methotrexate � placebo)27). This specifically demonstrates the difficulties whichan be involved in determining causality for nonarticularanifestations of RA. Indeed, the demonstrated role of

ro-inflammatory cytokines in anemia suggests that bio-ogic agents that target these cytokines may have benefi-ial effects on hemoglobin levels in RA patients. To date,

ig. 1 The role of inflammation in the regulation of hepcidinroduction. In inflammatory states, IL-6 is released from mac-ophages, and acts on hepatocytes to increase hepcidin pro-uction. High circulating levels of hepcidin inhibit the release of

ron from macrophages and the intestinal absorption of iron,esulting in hypoferremia. Nat Clin Pract Gastroenterol Hepatol004;1:39-45; doi:10.1038/ncpgasthep0019). (Color version

sf figure is available online.)

his has been confirmed in at least 1 treatment study in-olving anti-TNF-� antibody therapy (21). This particu-ar study also provided some circumstantial evidence forNF-�-mediated apoptotic depletion of bone marrow

rythroid cells to account for the anemia of chronic dis-ase in RA, and that anti-TNF-� antibody therapy maymeliorate this anemia by down-regulating the apoptoticechanisms involved in erythropoiesis (21).

hrombocytopenia

ew studies concern thrombocytopenia in patients withA. Traditionally observed as an adverse reaction toMARDs, particularly methotrexate (28,29), thrombo-

ytopenia has also been associated with TNF-� blockinggents, possibly representing an idiosyncratic reaction tondividual agents rather than a class effect (30,31). AsNF-� regulates pro-inflammatory cytokines such as

L-1, IL-6, and IL-8, as well as macrophage colony stim-lating factor, it is also possible that TNF-�-blockinggents may block stem-cell differentiation with resultantone marrow failure. Overall, further studies are clearlyeeded to determine causes of thrombocytopenia relatedo pro-inflammatory cytokines and similar mechanisms.

atigue

n patients with RA, fatigue is a commonly reportedymptom that is often severe and greatly impacts qualityf life (32). A wide variety of assessment scales are used toeasure fatigue in RA, although most do not have good

vidence for validation (33). The heterogeneity amongssessment tools makes it difficult to compare fatigue ratesetween different studies.

linical and Psychosocial Factors

atigue in patients with RA may be secondary to or associ-ted with other symptoms, such as pain, anxiety, depression,nemia, inability to exercise, or comorbidities (eg, fibromy-lgia). Several of these empirical correlations were confirmedn a study of 229 patients with a duration of RA of at least 5ears (11). In particular, it was found that fatigue is largelyxplained by pain, but psychosocial factors were also impor-ant, especially individual coping abilities and tendency toeek help and social support. On the other hand, laboratoryata available for 75 patients did suggest a correlation be-ween fatigue and hemoglobin level (11). A more recenttudy suggested that fatigue in RA is linked mainly to painnd depression and is secondarily associated with disease ac-ivity (10). Evidence for this association was suggested by theact that treatment with DMARDs or TNF inhibitors re-uced fatigue in patients whose RA was associated with painnd depression, independent of somewhat a direct effect onisease activity (10). The argument that fatigue among pa-ients with RA is predominantly mediated centrally is furtheronfirmed by a longitudinal study that examined relation-

hips between psychosocial characteristics and fatigue in 122

Rfwcsrd

I

DdstevzcwnimRpselihgg(tfoTmHt

rstetsd

dpaatAati

sm

G

Efe(aitemad

I

T(mRoRsisvmdtbdaRAbR

ketpimIahmtcma(

s

136 Systemic nonarticular manifestations of RA

A patients and 122 matched controls (34). This studyound that, although common in healthy controls, fatigueas closely associated with psychosocial factors such as in-

reased anxiety, more disability, less social support, and moreocial stress in the RA group. Social stress was identified as aelatively unexplored but potentially modifiable variable in-ependently related to fatigue in RA (34).

nflammatory Factors

espite these data indicating that fatigue is centrally me-iated and related to psychosocial associations with RA,everal studies suggested a potential role of pro-inflamma-ory cytokines in the etiology of fatigue. The first line ofvidence has emerged from several treatment studies in-olving biologic agents, including etanercept, tocili-umab, and adalimumab (35-37). In 1 of these multi-enter, randomized, double-blind clinical trials, patientsith recent-onset RA who received etanercept had a sig-ificantly faster improvement in fatigue than those receiv-

ng methotrexate in the first 2 months and the improve-ent in fatigue was sustained for up to 46 months (36).eduction in fatigue was strongly correlated with reducedain and Health Assessment Questionnaire disabilitycores. Other evidence has emerged from a study thatxamined whether chronic interpersonal stress (closelyinked with fatigue) is associated with cellular markers ofnflammation in RA patients (18). The study found thatigher chronic interpersonal stress was associated withreater stimulated IL-6 production (P � 0.05) as well asreater resistance to hydrocortisone inhibition of IL-6P � 0.05). This study is also significant in that it con-rolled for pain as a confounding variable (covariate) andound that stimulation of IL-6 contributing to fatigueccurred over and above any contribution by pain (18).herefore, it appears that there is heightened pro-inflam-atory cytokine activity in RA patients at risk for fatigue.owever, fatigue symptoms were not found to be related

o plasma levels of CRP (18).To further understand whether fatigue is a modifiable

isk factor that is specific to a particular class of drugs, atudy of 21,016 RA patients (13) concluded that, al-hough anti-TNF therapy may lower fatigue levels, thisffect could not be assumed to be greater for anti-TNFherapy than for other RA treatments. Hence, these re-ults appear to argue against a specific role of TNF in theevelopment of fatigue in RA patients.Overall, the literature suggests that inflammatory me-

iators may at least partially mediate fatigue among RAatients. The effects of pain and biopsychosocial factorsre also important in determining how impaired patientsre by this symptom and may suggest complementaryherapies to those specific for controlling disease activity.s such, there is a definite need for further research in thisrea. In the meantime, it is clear that fatigue is a symptomhat significantly impacts the quality of life of affected

ndividuals and that clinicians should be vigilant in a

creening for and measuring this nonarticular diseaseanifestation (32,33).

eneralized Bone Loss

specially when used long term, corticosteroid treatmentor RA comprises a well-established explanation for gen-ralized bone loss as a nonarticular manifestation of RA38). Another established mechanism is reduced mobilitynd functional capacity in RA, which influences the abil-ty to perform weight-bearing activity, an important de-erminant of bone mass. Outside of these causes for gen-ralized bone loss in RA patients, there is evidence fromore recent literature that inflammatory mediators oper-

nt in several mechanistic pathways are involved in theevelopment of bone loss.

nflammatory Factors

he receptor activator for nuclear factor � B ligandRANKL, otherwise known as osteoprotegerin ligand) is aember of the TNF (ligand) superfamily. Functionally,ANKL is a surface-bound molecule that activates oste-clasts leading to bone resorption. Not surprisingly,ANKL overproduction is implicated in a variety of ero-

ive or destructive bone diseases, including RA. Indeed,ncreased RANKL production has been observed in theynovial tissue of patients with RA (39). The present re-iew identified different studies suggesting that T-cellsay influence bone metabolism through a RANKL-me-

iated pathway. Activated T-cells express pro-inflamma-ory cytokines (eg, TNF, IL-1, IL-7, and IL-17) that haveeen shown to differentially upregulate RANKL andownregulate osteoprotegerin (which inhibits differenti-tion of osteoclast precursors), thereby altering theANKL/osteoprotegerin ratio to favor bone resorption.recent study, for example, found that IL-17 produced

y helper T-cells stimulated osteoclastogenesis via aANKL-mediated pathway (Fig. 2) (40).Other studies point to effects of inflammatory cyto-

ines on bone independent of RANKL pathways. Forxample, in 1 study, IL-6 and IL-11 were added to cul-ures of human peripheral blood mononuclear cells in theresence of macrophage colony stimulating factor, whichnduced the formation of multinucleated cells positive for

olecules expressed by osteoclasts (41). Hence, IL-6 andL-11 appear capable of inducing osteoclast formation byRANKL-independent mechanism (41). Other studiesave suggested that increased IL-1 levels, which occurore frequently in RA patients with active disease may,

ogether with increased IL-1 receptor agonist activity, in-rease bone resorption by stimulating the release of matrixetalloproteinases and other degradative products as well

s by promoting osteoclast differentiation and activation42,43).

In support of a potential role of members of the TNFuperfamily in the development of generalized bone loss

re studies using anti-TNF antibodies. In 1 study of mice

w(iAtsclimoopomaOibi

clptstasisBte

ispoat

F

FamsemmdkptChsapaipsbcoctd

pcdaau(tam

C

CiscCatb

FcsmMcrzpNtb2

L.W. Moreland and J.R. Curtis 137

ith collagen-induced arthritis, the bone mineral densityBMD) change from baseline to sacrifice was significantlyncreased by anti-TNF-� antibodies (versus saline) (44).nti-TNF-� antibodies also increased trabecular bone

hickness (P � 0.02 versus saline) to a value close to thateen in control mice. Another recent study found that aombination of antirheumatic drugs (methotrexate, buci-lamine, and sulfasalazine) with effects on T-cells and pro-nflammatory cytokines could interfere with a RANKL-

ediated pathway leading to an inhibitory effect onsteoclastogenesis by acting on osteoclast precursor andsteoclastogenesis-supporting cells (45). The specificathways involved the downregulation of nuclear factorf activated T-cells (NFAT) c1 induction, which is theaster transcription factor for osteoclastogenesis, as well

s inhibitory effects on T-cells and RANKL itself (45).ne caveat involves the finding that ablation of NFATc1

n mature osteoclasts does not prevent RANKL-mediatedone resorption activity, suggesting NFATc1 is involvedn cell fusion events and not bone resorption per se (46).

Hence, inhibiting TNF and other pro-inflammatoryytokines in RA patients may provide clinical benefits forocal articular and generalized systemic bone loss. Theositive results observed even with agents that have lessargeted effects on cytokine-mediated pathways opens uptrong possibilities for agents that have more specific ac-ions on TNF and cytokines such as IL-1, IL-6, IL-11,nd IL-17. However, several studies have failed to showignificant benefit in terms of bone loss for agents thatnhibit pro-inflammatory cytokines. For example, in 1tudy (the BeSt trial), there was no observed difference inMD after 1 year between infliximab- and corticosteroid-

reated RA patients (47). Another study found only mod-

ig. 2 Inflammatory pathways potentially involved in in-reased bone destruction observed in RA patients and theites at which these may be blocked by DMARD antirheu-atic therapy (45). Key: Buc, bucillamine; IL-1, interleukin-1;TX, methotrexate; NFATc1, nuclear factor of activated T-

ells; RANK, receptor activator for nuclear factor �B; RANKL,eceptor activator for nuclear factor �B ligand; SFZ, sulfasala-ine (salazosulphapyridine); TNF-�, tumor necrosis factor (al-ha). Reproduced with permission from Suematsu A, Tajiri Y,akashima T, Taka J, Ochi S, Oda H, et al. Scientific basis for

he efficacy of combined use of antirheumatic drugs againstone destruction in rheumatoid arthritis. Mod Rheumatol007;17:17-23.

st benefits for long-term infliximab treatment in guard- w

ng against generalized bone loss with BMD remainingtable while levels of serum type I procollagen N-terminalropeptide decreased and returned to pretreatment levelsnly after 1 year (48). Further studies, including thosessessing the effects of new treatments, are warranted inhis area.

elty’s Syndrome

elty’s syndrome is the constellation of RA, neutropenia,nd splenomegaly. Literature related to Felty’s syndromeainly concerns the considerable clinical experience that

urrounds the use of the recombinant human hemopoi-tic growth factors granulocyte (G-CSF) and granulocyte-acrophage colony-stimulating factors (49). An im-une-mediated neutropenia associated with a terminal

efect in neutrophil maturation is likely to underlie thisey clinical characteristic of Felty’s syndrome and a highroportion of patients with this nonarticular manifesta-ion have peripheral blood expansions of CD3� andD8� large granular lymphocytes (50). At least 1 studyas analyzed whether the clinical manifestations of Felty’syndrome, especially neutropenia, might be explained bybnormalities in cytokine production by examining theroduction of 5 cytokines (IL-1�, TNF-�, IL-8, G-CSF,nd granulocyte-macrophage colony stimulating factor)nvolved in the maturation and activation of polymor-honuclear cells (51). This study found that anti-CD3-timulated IL-8 production was significantly increased inoth RA and Felty’s syndrome, compared with normalontrols. However, it was concluded that the neutropeniaf Felty’s syndrome cannot be definitely explained byhanges in peripheral blood cytokine production, despitehe suggestion of a possible role for G-CSF and IL-8 in theevelopment of certain Felty’s syndrome complications.Treatment studies using agents that target cytokine

roduction reflect, to some extent, this apparent lack of alear role for cytokines in the pathogenesis of Felty’s syn-rome. For example, in 2 patients with refractory RAssociated with severe Felty’s syndrome, rituximab ther-py was ineffective perhaps owing to the inability of rit-ximab to bind to plasma cells, which are CD20 negative52). Etanercept has been shown to severely worsen neu-ropenia in a patient with longstanding seropositive RAnd Felty’s syndrome, despite improving articular inflam-ation (53).

ardiovascular Disease

ardiovascular disease (CVD) is a major contributor toncreased mortality among patients with RA, with studieshowing that RA is associated with a higher prevalence ofardiovascular (CV) endpoints such as angina, stroke, andVD-related mortality (54-56). Increased CV risk

mong patients with RA is not fully explained by tradi-ional CV risk factors alone. This is clearly demonstratedy a comparison of CV event rates over 1 year in patients

ith RA and the general population; after adjustments for

at(ad

Ai

TmRfioatpsa

Eo

AanCphmcstw

ici9aamrivititab

mIpgag

emioiicccmtt

ufhfqrrams

uesIwtg1fsur

AT

DEdwflsiadmv

pt(s

138 Systemic nonarticular manifestations of RA

ge, sex, smoking status, diabetes mellitus, hypercholes-erolemia, systolic blood pressure, and body mass indexie, traditional risk factors), RA was associated with anpproximately 4-fold increased risk of CV events (inci-ence rate ratio, 3.96; 95% CI, 1.86-8.43) (54).

venues of Enquirynto Mechanisms of Disease

he present review identified several mechanisms thatay underlie the increased risk of CVD in patients withA. Research has been performed in 2 broad areas. Therst has highlighted the possible effects of cytokines andther pro-inflammatory mediators on the acceleration oftherosclerosis and endothelial dysfunction. Studies inhis area tend to look at this issue from a mechanistic orathophysiological perspective. The second area concernstudies suggesting an influence of RA and its treatment ondversely affecting traditional CV risk factors.

ffects of Pro-Inflammatory Cytokinesn Atherosclerosis and Endothelial Function

nalysis of the third National Health and Nutrition Ex-mination Survey III population concluded that RA andon-RA participants had similar prevalence of traditionalVD risk factors and previous CVD (57). However, RAatients had lower plasma levels of antioxidants andigher levels of CRP. This suggests that inflammationay be at least partially responsible for the increased in-

idence of CVD in individuals with RA. Other evidenceupports the notion that atherosclerosis is an inflamma-ory disorder sharing common pathological pathwaysith the pannus formation and synovial inflammation.One study supportive of this hypothesis is a recent

nvestigation by Roman and coworkers that examinedlinical and biological markers for atherosclerotic diseasen RA patients (58). In this study, 98 RA outpatients and8 matched controls were assessed for CVD risk factorsnd by carotid ultrasonography. RA patients were alsossessed for disease severity, treatment, and inflammatoryarkers. Results showed that, despite a more favorable

isk factor profile, patients with RA had a 3-fold increasen the prevalence of carotid atherosclerotic plaque (44%ersus 15% in controls; P � 0.001). Although levels ofnflammatory mediators did not differ significantly be-ween RA patients and controls, the use of TNF-� inhib-tors (viewed as a proxy for disease severity and inflamma-ion) was significantly correlated with plaque levels. Theuthors concluded that chronic inflammation and, possi-ly, disease severity are atherogenic in this population.Results of more mechanistic studies have suggested aore direct link between inflammatory cytokines such as

L-6, TNF, and IL-1 with atherogenesis (17,59,60). Inarticular, in fatty streaks and the “cap”/“shoulder” re-ions of atheromatous plaques, macrophage foam cellsnd smooth muscle cells express IL-6. This indirectly sug-

ests a role for this cytokine (17) and several lines of m

vidence suggest a central role for IL-6 in the develop-ent of CVD. First, IL-6 has a key role in the systemic

nflammatory response, as it is the only cytokine capablef stimulating synthesis of all the acute-phase proteinsnvolved in the inflammatory response. Second, IL-6 alsonfluences leukocyte recruitment, the coagulation cas-ade, the endothelium, and other components of the vas-ulature. It has also been shown to predict higher levels ofritical endothelial biomarkers (vascular cell adhesionolecule-1, intracellular adhesion molecule-1, and endo-

helial leukocyte adhesion molecule-1), independently ofraditional CV factors (59).

Epidemiologic support for the role of IL-6 in contrib-ting to CV risk is found in a large-scale study whichound that elevated levels of IL-6 in 14,916 apparentlyealthy men were associated with an increased risk ofuture myocardial infarction (MI); men with the highestuartile of baseline IL-6 had a 2.3 times higher relativeisk of MI than those in the lowest quartile (61). Thiselationship persisted after control for other factors suchs CRP (61). Overall, this suggests a role for cytokine-ediated inflammation in the early stages of atherogene-

is (61).In another study conducted among 3489 patients with

nstable coronary artery disease treated with either anarly invasive (n � 1222) or a noninvasive (n � 1235)trategy (3269 of whom had plasma samples analyzed forL-6 levels), plasma levels of IL-6 �5 ng/L (comparedith levels �5 ng/L, the lower detection limit of the sys-

em used to measure IL-6 levels) were associated withreatly increased mortality in patients (RR, 3.47; 95% CI,.94-6.21), after adjustment for most established CV riskactors (60). It was concluded that circulating IL-6 was atrong independent marker for identifying patients withnstable coronary artery disease who were at an increasedisk of death over 6 to 12 months (60).

ssociations Between RA andraditional Cardiovascular Risk Factors

espite the results of the National Health and Nutritionxamination Survey III analysis suggesting that the inci-ence of traditional risk factors is similar between patientsith and without RA, it is possible that the systemic in-ammation associated with RA may influence the expres-ion of these well-established risk factors. Lipid abnormal-ties may be a particularly important CV risk factormong RA patients (62), particularly low levels of high-ensity lipoprotein (HDL) cholesterol (63). Althoughost RA treatments affect lipid levels, the clinical rele-

ance of these changes is not clear.This review identified a number of relevant studies in

atients with RA. First, data from a 56-week trial of pa-ients with early RA comparing combination therapyglucocorticoids, methotrexate, and sulfasalazine) versusulfasalazine alone found total cholesterol increased by a

ean of 0.6 mmol/L and HDL cholesterol increased by

mprcdtpttsmc(mapictcai

T

FuCgtshtr

alisrdtwwpiTrcei

ttdep

Fptnpawahtaswtss(esuir

caatdt

Citsse

psripadn

O

IkTatvie

L.W. Moreland and J.R. Curtis 139

ore than 50% in both groups after treatment (64). Thisroduced an improvement of the total cholesterol/HDLatio, which was particularly rapid in the group receivingombination treatment. Similarly, other studies haveemonstrated that TNF inhibitors elevate total choles-erol and HDL levels (65-68). In another study of 359atients with active RA in whom the response to metho-rexate was inadequate in which some patients werereated with the IL-6 receptor inhibitor tocilizumab,ome of the tocilizumab-treated patients experiencedoderate and reversible increases in the nonfasting total

holesterol and triglyceride levels and reductions in HDL35). Taken together, these data suggest that anti-inflam-atory therapy with either TNF or IL-6 inhibitors may

dversely affect lipid profiles; however, since these thera-ies also typically reduce inflammatory markers, the clin-cal relevance of the net effect on CV risk is unclear. Thesehanges may be related to the anti-inflammatory effect ofhese agents, since at least 1 study showed that the lipidhanges were correlated with a reduction in RA diseasectivity (57), but other factors such as favorable changesn insulin sensitivity may also play a role (57).

reatment Perspectives

rom a practical treatment perspective at both an individ-al and a population level, the role of inflammation inVD among RA patients is of considerable importanceiven that CV is the leading cause of excess mortality forhese individuals. Treatment with DMARDs (includingulfasalazine, hydroxychloroquine, and methotrexate)ave also been linked with significant CVD risk reduc-ion, even after additional correction for the presence ofheumatoid factor and erosions (69).

The findings suggesting a role for inflammatory medi-tors as a risk factor for CVD are further strengthened inight of treatment studies evaluating the effect of suppress-ng inflammation on direct markers of vascular functionuch as endothelium-dependent vasodilation. The presenteview identified a study that assessed endothelial-depen-ent vascular function (vasodilatation) of the brachial ar-ery in 11 RA patients with high disease activity treatedith infliximab 3 mg/kg at weeks 0, 2, and 6. Comparedith baseline, endothelial function substantially im-roved following 12 weeks’ TNF inhibition using inflix-mab (from 3.2 � 0.4% to 4.1 � 0.5%, P � 0.018) (70).hese improvements with infliximab occurred concur-

ently with improvements in inflammatory markers andlinical parameters and suggest that, in patients with RA,ndothelial dysfunction is part of the disease process ands mediated at least in part by TNF-�.

To put the potential clinical benefit of anticytokineherapy into clinical context, another study found thatreatment with etanercept or infliximab significantly re-uced the age–sex-adjusted incidence rate of first CVDvents compared with RA patients not treated (14.0/1000

erson-years at risk versus 35.4/1000 person-years) (71). w

urther evidence supporting the role that inflammationlays in CVD and, in particular, MI comes from data ofhe British Society for Rheumatology Biologics Register, aational prospective observational study, which com-ared MI rates in 8670 patients with RA treated withnti-TNF-� and 2170 patients with active RA treatedith traditional DMARDs (72). In the intent-to-treat

nalysis, RA patients treated with anti-TNF-� did notave a lower incidence of MI compared with RA patientsreated with traditional DMARDs; however, a subgroupnalysis of those anti-TNF-�-treated patients who re-ponded to the treatment within 6 months versus thoseho did not found high differential MI rates between

hese groups (3.5 events per 1000 person-years in re-ponders and 9.4 events per 1000 person-years in nonre-ponders). This equated to an adjusted incidence rate ratio95% CI) for responders compared with nonrespondersqual to 0.36 (0.19-0.69) (72). Overall, therefore, there istrong support for the hypothesis that inflammation contrib-tes to the development of CV events, and that suppressingnflammatory mediators such as TNF-� may reduce theisk of CVD among RA patients.

Similar to cardiac disease (56), the incidence of non-ardiac vascular disease (except cerebrovascular diseaselone) in patients with RA has also been positively associ-ted with the presence of severe nonarticular RA manifes-ations (73). This suggests that the excess risk of noncar-iac vascular disease in RA is likely to be related, in part, tohe systemic inflammation associated with RA.

For both traditional and newer biologic therapies,VD risk reductions have been linked to suppression of

nflammation. This is also in keeping with the observationhat strategies to prevent CVD and mortality focusedolely on controlling traditional CV risk factors may beomewhat less beneficial in RA subjects than in the gen-ral population (74).

Further research is needed to determine optimal ap-roaches to reducing CV morbidity and mortality in per-ons with RA, including the development of refined CVisk stratification algorithms that take into account themportant role of nontraditional CV risk factors in thisopulation (74). Future studies should also help betterscertain the utility of biological therapies on the inci-ence and clinical course of this particularly importantonarticular manifestation of RA.

cular Effects

nflammatory eye diseases associated with RA includeeratoconjunctivitis sicca, uveitis, scleritis, and keratitis.he present review identified only a small number of

rticles dealing with mechanistic aspects of ocular nonar-icular manifestations of RA. In terms of studies that in-estigated a potential role of pro-inflammatory cytokinesn the development of ocular complications, 1 source ofvidence is derived from a study in which corneal samples

ere collected from 7 RA patients with corneal ulcer-

apa7Aauaapdsf(pal

irwRsi(arsrat

P

Atpalpnsomrr

RP

Rso5pnit

cRpio

fip(citafihtBp

phcdstani9CisR322(aliOai

ntdaeaRctei

140 Systemic nonarticular manifestations of RA

tions/perforations (75). Paraffin sections of these sam-les were examined using specific gene probes for TNF-�nd IL-6 and this detected TNF-� and IL-6 expression in1 and 100% of analyzed tissue samples, respectively.lthough there was no control group for comparison, theuthors of this study concluded that, among RA patients,pregulation of the pro-inflammatory cytokines TNF-�nd IL-6 may induce the production of metalloprotein-ses, resulting in collagenolytic corneal damage. In sup-ort of this finding, a case report of an RA patient whoeveloped necrotizing scleritis refractory to local mea-ures, systemic corticosteroids, and cyclophosphamideound that this patient responded rapidly to infliximab76). This report suggests that biologic agents affectingro-inflammatory cytokines may be useful in ocular non-rticular manifestations of RA, although the evidence isargely limited to case reports.

Arguing against a clear role for inflammatory mediatorsn other ocular manifestations of RA are the results of aecent prospective case-control study in 72 RA patients inhom the severity of dry eye was assessed against that ofA activity (evaluated by assessing duration of morning

tiffness, ESR, grip strength, and joint score, the Lansburyndex) (77). This study found that, although dry eyesymptomatic of keratoconjunctivitis sicca) was commonmong enrolled RA patients, there was no apparent cor-elation between disease activity and the severity of ocularymptoms. This indicates, albeit indirectly, that the cor-elation between sicca symptoms and inflammation is nots strong as for other ocular manifestations of RA. Hence,here is a definite need for further studies in this area.

ulmonary Complications

s with ocular manifestations, pulmonary RA manifesta-ions involve a spectrum of disorders. These can includeleurisy, parenchymal nodules, interstitial involvement,irway disease, and pulmonary vasculitis, although theatter is rare (5). Pleural involvement is the most commonulmonary manifestation and often occurs with pulmonaryodules or interstitial disease. Pulmonary involvement in as-ociation with RA may occur as a result of disease factorsr secondary to treatment for RA, such as D-penicilla-ine or gold-induced obliterative bronchiolitis or, more

arely, methotrexate-induced pneumonitis. Treatment-elated causes will not be covered in detail here.

ole of Inflammation inulmonary RA Manifestations

eview of the literature uncovered a small number oftudies that point to a role of inflammation in the devel-pment of pulmonary manifestations of RA. In 1 study of2 patients with RA assessed using high-resolution com-uted tomography and pulmonary function tests, it wasoted that rheumatoid factor titers and ESR were signif-

cantly higher in the presence of lung abnormalities de-

ected by computed tomography scans (78). These results p

oncur with those of other studies, albeit not specific toA, which have shown that the cytokines involved inulmonary inflammatory and immune disease can be pro-nflammatory (eg, TNF-� and IL-1), anti-inflammatory,r growth-stimulatory.

Another study has demonstrated, via staining of paraf-n-embedded lung biopsy specimens from interstitialneumonia patients with RA (n � 15) and without RAn � 16), that patients with RA have increased CD4�ell counts (79). B-lymphocytes, which have emerged asmportant elements in the events leading to joint destruc-ion in RA, may also be significantly involved in RA-ssociated interstitial pneumonia (80). This was con-rmed in a study that found marked follicular B-cellyperplasia among patients with RA-associated intersti-ial pneumonia, which was also associated with higher-cell cellularity (as compared with idiopathic interstitialneumonia) (80).Studies related to RA treatments and pulmonary com-

lications have yielded interesting results. One review ofospital and death records found no clear pattern ofausal association of RA treatment and interstitial lungisease and no clear evidence to support a causal relation-hip between infliximab, azathioprine, and hospitaliza-ion for this pulmonary manifestation (81). However,mong past therapies received by these patients, pred-isone (hazard ratio [HR], 3.0; 95% CI, 1.0-8.9), inflix-

mab (HR, 2.1; 95% CI, 1.1-3.8), etanercept (HR, 1.7;5% CI, 1.0-3.0), and cyclophosphamide (HR, 3.7; 95%I, 0.9-15.5) were all positively associated with hospital-

zation for interstitial lung disease (ILD). In anothertudy, 9847 consecutive patients from the BSR Biologicsegister with RA who were treated with etanercept (n �969), infliximab (n � 3231), or adalimumab (n �647) were compared with a similarly recruited cohort of906 subjects with active RA requiring DMARD therapy82). This review found that patients with baseline RA-ssociated ILD, those in the anti-TNF cohort were moreikely to die than those with baseline RA-associated ILDn the DMARD cohort (HR, 2.11; 95% CI, 0.59, 7.59).

verall, anti-TNF-� was associated with a worse all-causend RA-associated interstitial lung disease specific mortal-ty in patients with this condition at baseline.

These results suggest that inhibition of TNF-� mayot be of benefit to patients with fibrosing lung condi-ions associated with RA. However, confounding by in-ication may explain these associations between TNF-lpha use and interstitial lung disease. Since there is somevidence that inflammation and inflammatory cytokinesre central to the pathology of pulmonary nonarticularA manifestations. However, further investigation (in-luding large-scale randomized studies) is required to bet-er delineate the role for inflammatory cytokines in thetiology of specific pulmonary nonarticular disease and,n particular, treatment with emerging biological thera-

ies.

D

Fantpqfcsaa

ecpfattdbwt

cmwmsqfm

setiftedsfbttvrscmmtwd

dej

R

1

1

1

1

1

1

1

1

1

L.W. Moreland and J.R. Curtis 141

ISCUSSION

uture research in this area is based on finding answers tonumber of key questions. A central question, which hasot been addressed extensively to date, concerns how pa-ients with systemic RA manifestations differ from RAatients without those manifestations. Answering thisuestion may provide insight into whether there are dif-ering levels of expression of particular pro-inflammatoryytokines between individuals with RA. The basis foruch heterogeneity, including genetics and environmentalnd iatrogenic (eg, treatment) factors, provides a second-ry avenue of research.

Another question relates to whether the incidence ofxtra-articular manifestations has decreased in the moreontemporary treatment era or if there are recognizableatterns in terms of geographic, gender-based, or otherorms of predilection to particular manifestations. Again,nswers to these questions may help uncover any associa-ions of nonarticular manifestations with patient-cen-ered or external factors. Furthermore, studies of bothisease mechanisms and epidemiological patterns shoulde able to inform researchers and clinicians how to predicthich patients will develop systemic manifestations over

he course of the disease.Last and importantly, there are several questions con-

erning treatment of nonarticular disease that requireore complete answers. These include the extent tohich currently available RA therapies prevent or pro-ote nonarticular disease. As this review has revealed,

everal smaller studies have made attempts to answer thisuestion, although the answer is far from complete evenor those extra-articular manifestations that have beenost rigorously investigated.The other question of practical concern to clinicians

hould be how nonarticular manifestations can be mostffectively treated, including the use of newer biologicherapies. To date, clinical trials of biologic agents thatnhibit pro-inflammatory cytokines have not principallyocused on extra-articular manifestations of RA. Studieso address the effect of these agents on extra-articular dis-ase need to be large and are potentially complex to con-uct. An added complication is that many extra-articularymptoms may be reported as adverse events or may occuror the first time during therapy and may not necessarilye related to treatment. Therefore, randomized con-rolled trials may have to be supplemented with observa-ional data. Challenges to interpreting results from obser-ational studies is that treatments are not assignedandomly, and patients with more severe RA are typicallyelected to receive more aggressive treatment regimens,reating spurious associations between extra-articularanifestations and the drugs used. Nevertheless, infor-ation from both types of data sources is needed to fur-

her inform us regarding whether cytokine inhibitionith agents currently available or in development can re-

uce the burden of extra-articular disease. If convincingly 1

emonstrated, this would represent a less anticipated ben-fit than the direct benefits of arthritis therapies to reduceoint inflammation.

EFERENCES

1. Turesson C, O’Fallon WM, Crowson CS, Gabriel SE, MattesonEL. Occurrence of extraarticular disease manifestations is associ-ated with excess mortality in a community based cohort of pa-tients with rheumatoid arthritis. J Rheumatol 2002;29:62-7.

2. Wilson A, Yu HT, Goodnough LT, Nissenson AR. Prevalenceand outcomes of anemia in rheumatoid arthritis: a systematicreview of the literature. Am J Med 2004;116(Suppl 7A):50S-7S.

3. Turesson C, O’Fallon WM, Crowson CS, Gabriel SE, MattesonEL. Extra-articular disease manifestations in rheumatoid arthritis:incidence trends and risk factors over 46 years. Ann Rheum Dis2003;62:722-7.

4. Feldmann M, Brennan FM, Maini RN. Rheumatoid arthritis.Cell 1996;85:307-10.

5. Anaya JM, Diethelm L, Ortiz LA, Gutierrez M, Citera G, WelshRA, et al. Pulmonary involvement in rheumatoid arthritis. SeminArthritis Rheum 1995;24:242-54.

6. Cohen MG, Li EK, Ng PY, Chan KL. Extra-articular manifesta-tions are uncommon in southern Chinese with rheumatoid arthri-tis. Br J Rheumatol 1993;32:209-11.

7. Cohen SB, Moreland LW, Cush JJ, Greenwald MW, Block S,Shergy WJ, et al. A multicentre, double blind, randomised, pla-cebo controlled trial of anakinra (Kineret), a recombinant inter-leukin 1 receptor antagonist, in patients with rheumatoid arthritistreated with background methotrexate. Ann Rheum Dis 2004;63:1062-8.

8. Goldring SR, Gravallese EM. Adverse effects of rheumatoid ar-thritis on bone remodeling. Arthritis Res Ther 2004;6:38.

9. Jongen-Lavrencic M, Peeters HR, Wognum A, Vreugdenhil G,Breedveld FC, Swaak AJ. Elevated levels of inflammatory cyto-kines in bone marrow of patients with rheumatoid arthritis andanemia of chronic disease. J Rheumatol 1997;24:1504-9.

0. Pollard LC, Choy EH, Gonzalez J, Khoshaba B, Scott DL. Fa-tigue in rheumatoid arthritis reflects pain, not disease activity.Rheumatology (Oxford) 2006;45:885-9.

1. Riemsma RP, Rasker JJ, Taal E, Griep EN, Wouters JM, Wieg-man O. Fatigue in rheumatoid arthritis: the role of self-efficacyand problematic social support. Br J Rheumatol 1998;37:1042-6.

2. Turesson C, Jacobsson L, Bergstrom U. Extra-articular rheuma-toid arthritis: prevalence and mortality. Rheumatology (Oxford)1999;38:668-74.

3. Wolfe F, Michaud K. Fatigue, rheumatoid arthritis, and anti-tumor necrosis factor therapy: an investigation in 24,831 patients.J Rheumatol 2004;31:2115-20.

4. Wolfe F, Michaud K. Anemia and renal function in patients withrheumatoid arthritis. J Rheumatol 2006;33:1516-22.

5. Gabriel SE, Crowson CS, Kremers HM, Doran MF, Turesson C,O’Fallon WM, et al. Survival in rheumatoid arthritis: a popula-tion-based analysis of trends over 40 years. Arthritis Rheum 2003;48:54-8.

6. Voulgari PV, Kolios G, Papadopoulos GK, Katsaraki A, SeferiadisK, Drosos AA. Role of cytokines in the pathogenesis of anemia ofchronic disease in rheumatoid arthritis. Clin Immunol 1999;92:153-60.

7. Yudkin JS, Kumari M, Humphries SE, Mohamed-Ali V. Inflam-mation, obesity, stress and coronary heart disease: is interleukin-6the link? Atherosclerosis 2000;148:209-14.

8. Davis MC, Zautra AJ, Younger J, Motivala SJ, Attrep J, IrwinMR. Chronic stress and regulation of cellular markers of inflam-mation in rheumatoid arthritis: implications for fatigue. BrainBehav Immun 2008;22:24-32.

9. Glossop JR, Dawes PT, Hassell AB, Mattey DL. Anemia in rheuma-

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

142 Systemic nonarticular manifestations of RA

toid arthritis: association with polymorphism in the tumor necrosisfactor receptor I and II genes. J Rheumatol 2005;32:1673-8.

0. Maury CP, Liljestrom M, Laiho K, Tiitinen S, Kaarela K, HurmeM. Tumor necrosis factor alpha, its soluble receptor I, and -308gene promoter polymorphism in patients with rheumatoid arthri-tis with or without amyloidosis: implications for the pathogenesisof nephropathy and anemia of chronic disease in reactive amyloid-osis. Arthritis Rheum 2003;48:3068-76.

1. Papadaki HA, Kritikos HD, Valatas V, Boumpas DT, EliopoulosGD. Anemia of chronic disease in rheumatoid arthritis is associ-ated with increased apoptosis of bone marrow erythroid cells:improvement following anti-tumor necrosis factor-alpha anti-body therapy. Blood 2002;100:474-82.

2. Weinstein DA, Roy CN, Fleming MD, Loda MF, Wolfsdorf JI,Andrews NC. Inappropriate expression of hepcidin is associatedwith iron refractory anemia: implications for the anemia ofchronic disease. Blood 2002;100:3776-81.

3. Pietrangelo A, Trautwein C. Mechanisms of disease: the role ofhepcidin in iron homeostasis—implications for hemochromatosisand other disorders. Nat Clin Pract Gastroenterol Hepatol 2004;1:39-45.

4. Nemeth E, Rivera S, Gabayan V, Keller C, Taudorf S, PedersenBK, et al. IL-6 mediates hypoferremia of inflammation by induc-ing the synthesis of the iron regulatory hormone hepcidin. J ClinInvest 2004;113:1271-6.

5. Kaltwasser JP, Kessler U, Gottschalk R, Stucki G, Moller B. Effectof recombinant human erythropoietin and intravenous iron onanemia and disease activity in rheumatoid arthritis. J Rheumatol2001;28:2430-6.

6. Mercuriali F. Epoetin alfa for autologous blood donation in pa-tients with rheumatoid arthritis and concomitant anemia. SeminHematol 1996;33:18-20; discussion 21.

7. Han C, Rahman MU, Doyle MK, Bathon JM, Smolen J, Ka-vanaugh A, et al. Association of anemia and physical disabilityamong patients with rheumatoid arthritis. J Rheumatol 2007;34:2177-82.

8. Franck H, Rau R, Herborn G. Thrombocytopenia in patientswith rheumatoid arthritis on long-term treatment with low dosemethotrexate. Clin Rheumatol 1996;15:163-7.

9. Grove ML, Hassell AB, Hay EM, Shadforth MF. Adverse reac-tions to disease-modifying anti-rheumatic drugs in clinical prac-tice. QJM 2001;94:309-19.

0. Hamaguchi M, Kawahito Y, Ishino H, Yoshida M, Yoshikawa T.A case report of tumor necrosis factor-alpha antibody-inducedthrombocytopenia associated with emerging IgM anticardiolipinantibody in patients with scleroderma overlap/rheumatoid arthri-tis. Clin Rheumatol 2007;26:988-90.

1. Pathare SK, Heycock C, Hamilton J. TNFalpha blocker-inducedthrombocytopenia. Rheumatology (Oxford) 2006;45:1313-4.

2. Kirwan JR, Minnock P, Adebajo A, Bresnihan B, Choy E, de WitM, et al. Patient perspective: fatigue as a recommended patientcentered outcome measure in rheumatoid arthritis. J Rheumatol2007;34:1174-7.

3. Hewlett S, Hehir M, Kirwan JR. Measuring fatigue in rheumatoidarthritis: a systematic review of scales in use. Arthritis Rheum2007;57:429-39.

4. Mancuso CA, Rincon M, Sayles W, Paget SA. Psychosocial vari-ables and fatigue: a longitudinal study comparing individuals withrheumatoid arthritis and healthy controls. J Rheumatol 2006;33:1496-502.

5. Maini RN, Taylor PC, Szechinski J, Pavelka K, Broll J, Balint G,et al. Double-blind randomized controlled clinical trial of theinterleukin-6 receptor antagonist, tocilizumab, in european pa-tients with rheumatoid arthritis who had an incomplete responseto methotrexate. Arthritis Rheum 2006;54:2817-29.

6. Moreland LW, Genovese MC, Sato R, Singh A. Effect of etaner-cept on fatigue in patients with recent or established rheumatoid

arthritis. Arthritis Rheum 2006;55:287-93.

7. Weinblatt ME, Keystone EC, Furst DE, Moreland LW, WeismanMH, Birbara CA, et al. Adalimumab, a fully human anti-tumornecrosis factor alpha monoclonal antibody, for the treatment ofrheumatoid arthritis in patients taking concomitant methotrex-ate: the ARMADA trial. Arthritis Rheum 2003;48:35-45.

8. Kanis JA, Johansson H, Oden A, Johnell O, de Laet C, Melton IL,et al. A meta-analysis of prior corticosteroid use and fracture risk.J Bone Miner Res 2004;19:893-9.

9. Fonseca JE, Cortez-Dias N, Francisco A, Sobral M, Canhao H,Resende C, et al. Inflammatory cell infiltrate and RANKL/OPGexpression in rheumatoid synovium: comparison with other in-flammatory arthropathies and correlation with outcome. ClinExp Rheumatol 2005;23:185-92.

0. Sato K, Suematsu A, Okamoto K, Yamaguchi A, Morishita Y,Kadono Y, et al. Th17 functions as an osteoclastogenic helper Tcell subset that links T cell activation and bone destruction. J ExpMed 2006;203:2673-82.

1. Kudo O, Sabokbar A, Pocock A, Itonaga I, Fujikawa Y, Athana-sou NA. Interleukin-6 and interleukin-11 support human oste-oclast formation by a RANKL-independent mechanism. Bone2003;32:1-7.

2. Abramson SB, Amin A. Blocking the effects of IL-1 in rheumatoidarthritis protects bone and cartilage. Rheumatology (Oxford)2002;41:972-80.

3. Eastgate JA, Symons JA, Wood NC, Grinlinton FM, di GiovineFS, Duff GW. Correlation of plasma interleukin 1 levels withdisease activity in rheumatoid arthritis. Lancet 1988;2:706-9.

4. Saidenberg-Kermanac’h N, Corrado A, Lemeiter D, deVernejoulMC, Boissier MC, Cohen-Solal ME. TNF-alpha antibodies andosteoprotegerin decrease systemic bone loss associated with in-flammation through distinct mechanisms in collagen-induced ar-thritis. Bone 2004;35:1200-7.

5. Suematsu A, Tajiri Y, Nakashima T, Taka J, Ochi S, Oda H, et al.Scientific basis for the efficacy of combined use of antirheumaticdrugs against bone destruction in rheumatoid arthritis. ModRheumatol 2007;17:17-23.

6. Kim MS, Day CJ, Selinger CI, Magno CL, Stephens SR, Morri-son NA. MCP-1-induced human osteoclast-like cells are tartrate-resistant acid phosphatase. NFATc1, and calcitonin receptor-pos-itive but require receptor activator of NFkappaB ligand for boneresorption. J Biol Chem 2006;281:1274-85.

7. Guler-Yuksel M, Bijsterbosch J, Goekoop-Ruiterman YP, deVries-Bouwstra JK, Hulsmans HM, de Beus WM, et al. Changesin bone mineral density in patients with recent onset, active rheu-matoid arthritis. Ann Rheum Dis 2008;67:823-28.

8. Chopin F, Garnero P, le Henanff A, Debiais F, Daragon A, RouxC, et al. Long term effects of infliximab on bone and cartilageturnover markers in patients with rheumatoid arthritis. AnnRheum Dis 2008;67:353-57.

9. Pixley JS, Patchin GM. Felty’s syndrome. BioDrugs 1997;7:356-65.

0. Bowman SJ, Hall MA, Panayi GS, Lanchbury JS. T cell receptoralpha-chain and beta-chain junctional region homology in clonalCD3�, CD8� T lymphocyte expansions in Felty’s syndrome.Arthritis Rheum 1997;40:615-23.

1. Meliconi R, Uguccioni M, Chieco-Bianchi F, Pitzalis C, BowmanS, Facchini A, et al. The role of interleukin-8 and other cytokinesin the pathogenesis of Felty’s syndrome. Clin Exp Rheumatol1995;13:285-91.

2. Sordet C, Gottenberg JE, Hellmich B, Kieffer P, Mariette X,Sibilia J. Lack of efficacy of rituximab in Felty’s syndrome. AnnRheum Dis 2005;64:332-3.

3. Ghavami A, Genevay S, Fulpius T, Gabay C. Etanercept in treat-ment of Felty’s syndrome. Ann Rheum Dis 2005;64:1090-1.

4. del Rincon ID, Williams K, Stern MP, Freeman GL, Escalante A.High incidence of cardiovascular events in a rheumatoid arthritiscohort not explained by traditional cardiac risk factors. Arthritis

Rheum 2001;44:2737-45.

5

5

5

5

5

6

6

6

6

6

6

6

6

6

6

7

7

7

7

7

7

7

7

7

7

8

8

8

L.W. Moreland and J.R. Curtis 143

5. Turesson C, Jarenros A, Jacobsson L. Increased incidence of car-diovascular disease in patients with rheumatoid arthritis: resultsfrom a community based study. Ann Rheum Dis 2004;63:952-5.

6. Turesson C, McClelland RL, Christianson TJ, Matteson EL. Se-vere extra-articular disease manifestations are associated with anincreased risk of first ever cardiovascular events in patients withrheumatoid arthritis. Ann Rheum Dis 2007;66:70-5.

7. De Pablo P, Dietrich T, Karlson EW. Antioxidants and othernovel cardiovascular risk factors in subjects with rheumatoid ar-thritis in a large population sample. Arthritis Rheum 2007;57:953-62.

8. Roman MJ, Moeller E, Davis A, Paget SA, Crow MK, LockshinMD, et al. Preclinical carotid atherosclerosis in patients with rheu-matoid arthritis. Ann Intern Med 2006;144:249-56.

9. Dessein PH, Joffe BI, Singh S. Biomarkers of endothelial dysfunc-tion, cardiovascular risk factors and atherosclerosis in rheumatoidarthritis. Arthritis Res Ther 2005;7:R634-43.

0. Lindmark E, Diderholm E, Wallentin L, Siegbahn A. Relation-ship between interleukin 6 and mortality in patients with unstablecoronary artery disease: effects of an early invasive or noninvasivestrategy. JAMA 2001;286:2107-13.

1. Ridker PM, Rifai N, Stampfer MJ, Hennekens CH. Plasma concen-tration of interleukin-6 and the risk of future myocardial infarctionamong apparently healthy men. Circulation 2000;101:1767-72.

2. Gerli R, Sherer Y, Bocci EB, Vaudo G, Moscatelli S, Shoenfeld Y.Precocious atherosclerosis in rheumatoid arthritis: role of tradi-tional and disease-related cardiovascular risk factors. Ann NYAcad Sci 2007;1108:372-81.

3. Choi HK, Seeger JD. Lipid profiles among US elderly with un-treated rheumatoid arthritis—the third National Health and Nu-trition Examination Survey. J Rheumatol 2005;32:2311-6.

4. Boers M, Nurmohamed MT, Doelman CJ, Lard LR, VerhoevenAC, Voskuyl AE, et al. Influence of glucocorticoids and diseaseactivity on total and high density lipoprotein cholesterol in pa-tients with rheumatoid arthritis. Ann Rheum Dis 2003;62:842-5.

5. Allanore Y, Kahan A, Sellam J, Ekindjian OG, Borderie D. Effectsof repeated infliximab therapy on serum lipid profile in patientswith refractory rheumatoid arthritis. Clin Chim Acta 2006;365:143-8.

6. Popa C, Netea MG, Radstake T, Van der Meer JW, StalenhoefAF, van Riel PL, et al. Influence of anti-tumour necrosis factortherapy on cardiovascular risk factors in patients with active rheu-matoid arthritis. Ann Rheum Dis 2005;64:303-5.

7. Seriolo B, Paolino S, Sulli A, Fasciolo D, Cutolo M. Effects ofanti-TNF-alpha treatment on lipid profile in patients with activerheumatoid arthritis. Ann NY Acad Sci 2006;1069:414-9.

8. Tam LS, Tomlinson B, Chu TT, Li TK, Li EK. Impact of TNFinhibition on insulin resistance and lipids levels in patients withrheumatoid arthritis. Clin Rheumatol 2007;26:1495-8.

9. van Halm VP, Nurmohamed MT, Twisk JW, Dijkmans BA,Voskuyl AE. Disease-modifying antirheumatic drugs are associ-ated with a reduced risk for cardiovascular disease in patients withrheumatoid arthritis: a case control study. Arthritis Res Ther

2006;8:R151.

0. Hurlimann D, Forster A, Noll G, Enseleit F, Chenevard R, Dis-tler O, et al. Anti-tumor necrosis factor-alpha treatment improvesendothelial function in patients with rheumatoid arthritis. Circu-lation 2002;106:2184-7.

1. Jacobsson LT, Turesson C, Gulfe A, Kapetanovic MC, PeterssonIF, Saxne T, et al. Treatment with tumor necrosis factor blockersis associated with a lower incidence of first cardiovascular events inpatients with rheumatoid arthritis. J Rheumatol 2005;32:1213-8.

2. Dixon WG, Watson KD, Lunt M, Hyrich KL, Silman AJ, Sym-mons DP. Reduction in the incidence of myocardial infarction inpatients with rheumatoid arthritis who respond to anti-tumornecrosis factor alpha therapy: results from the British Society forRheumatology Biologics Register. Arthritis Rheum 2007;56:2905-12.

3. Liang KP, Liang KV, Matteson EL, McClelland RL, ChristiansonTJ, Turesson C. Incidence of noncardiac vascular disease in rheu-matoid arthritis and relationship to extraarticular disease manifes-tations. Arthritis Rheum 2006;54:642-8.

4. Gonzalez A, Maradit Kremers H, Crowson CS, Ballman KV,Roger VL, Jacobsen SJ, et al. Do cardiovascular risk factors conferthe same risk for cardiovascular outcomes in rheumatoid arthritispatients as in non-RA patients? Ann Rheum Dis 2008;67:64-9.

5. Prada J, Noelle B, Baatz H, Hartmann C, Pleyer U. Tumournecrosis factor alpha and interleukin 6 gene expression in kerato-cytes from patients with rheumatoid corneal ulcerations. Br JOphthalmol 2003;87:548-50.

6. Atchia II, Kidd CE, Bell RW. Rheumatoid arthritis-associatednecrotizing scleritis and peripheral ulcerative keratitis treated suc-cessfully with infliximab. J Clin Rheumatol 2006;12:291-3.

7. Fujita M, Igarashi T, Kurai T, Sakane M, Yoshino S, TakahashiH. Correlation between dry eye and rheumatoid arthritis activity.Am J Ophthalmol 2005;140:808-13.

8. Bilgici A, Ulusoy H, Kuru O, Celenk C, Unsal M, Danaci M.Pulmonary involvement in rheumatoid arthritis. Rheumatol Int2005;25:429-35.

9. Turesson C, Matteson EL, Colby TV, Vuk-Pavlovic Z, VassalloR, Weyand CM, et al. Increased CD4� T cell infiltrates in rheu-matoid arthritis-associated interstitial pneumonitis comparedwith idiopathic interstitial pneumonitis. Arthritis Rheum 2005;52:73-9.

0. Atkins SR, Turesson C, Myers JL, Tazelaar HD, Ryu JH, Matte-son EL, et al. Morphologic and quantitative assessment ofCD20� B cell infiltrates in rheumatoid arthritis-associated non-specific interstitial pneumonia and usual interstitial pneumonia.Arthritis Rheum 2006;54:635-41.

1. Wolfe F, Caplan L, Michaud K. Rheumatoid arthritis treatmentand the risk of severe interstitial lung disease. Scand J Rheumatol2007;36:172-8.

2. Dixon WG, Watson KD, Lunt M, Hyrich KL. Rheumatoid ar-thritis, interstitial lung disease, mortality and anti-TNF therapy:results from the BSR Biologics Register (BSRBR) In: ACR/ARHPAnnual Scientific Meeting 2007. Boston (MA): November 6-11,

2007. (Abstract).