rheumatoid arthritis: pathogenesis, clinical features, and ...arthritis) and the mhc molecules that...

RHEUMATOID ARTHRITIS: PATHOGENESIS, CLINICAL FEATURES, AND TREATMENT 1

Contents

1. Pathogenesis

2. Epidemiology

3. Clinical Features

4. Prognosis

5. Therapy

6. Specific Antirheumatic Drugs

7. References

Rheumatoid Arthritis:Pathogenesis, Clinical

Features, and TreatmentArthur Kavanaugh, MD

John J. Cush, MDRichard P. Polisson, MD, MHS

1. Pathogenesis

The pathobiology of RA involves a complex interac-tion of three different scientific domains: 1) a complexgenetic predisposition to the disease plus some envi-ronmental stimulus; 2) a self-perpetuating, self-ampli-fying, intra-synovial immune response; and at thefinal stage, 3) tissue injury mediated by pro-inflamma-tory cells, inflammatory effector molecules, anddegradative enzymes. In individuals with RA, thisprocess is arthrotropic and produces a characteristicpathologic lesion in the synovium as well as the hall-mark erosions of bone and destruction of cartilage atthe joint margin.

The histopathology of RA synovium has been welldescribed. The synovial lining, the interstitium, andthe microvasculature are all involved. Early on in theprocess, the synovial lining, which includes bothType A (macrophage-like) and Type B (mesenchy-mal or fibroblast-like) cells, becomes proliferative.The synovial lining increases in cell number andmass. Likewise, a diffuse and nodular inflammatorycell infiltrate is observed in the interstitium. Itincludes CD4+ and CD8+ lymphocytes, dendriticcells, and other antigen presenting cells. In somepatients, the histologic appearance is quite dramatic,showing focal aggregation of both T- and B-cells, aswell as the presence of germinal centers similar tothat which is seen in lymphoid tissues.

The synovium of the rheumatoid joint, although notmalignant, often times behaves as a local invasive“tumor.” The microvasculature initially revealsendothelial cell activation. As the process matures,plasma cells and multinucleated giant cells appear,and the vascular supply becomes exuberant. Finally,the growing synovium appears as granulation tissueas it advances to the hyaline cartilage at the margin ofthe joint. A local effect of degradative enzymes andactivated osteoclasts produces the classic erosion atthe bone and cartilage margin. These enzymes mayalso affect structures that are more distant, includingthe tendons, ligaments, and other musculoskeletalstructures. Erosions are produced by bone and matrixprotein resorbing osteoclasts, which may be inducedand activated by cytokines released into the inflam-matory milieu. Some of the scientific observationsthat explain these phenomena are noted below.

2 EDUCATIONAL REVIEW MANUAL IN RHEUMATOLOGY

Rheumatoid arthritis (RA) is a chronic, destructive,inflammatory arthropathy manifested by articularand extra-articular features. RA has profoundeffects on patient function and morbidity and exactsa substantial economic burden from affected per-sons. Although the pathology of the synovial inflam-mation and cartilage destruction that occurs inpatients with RA has been described for decades,many important developments in the understandingof genetic influences and immunopathophysiologicmechanisms have recently been defined. Basicresearch delineating the molecular mechanisms ofsynovial inflammation has driven the developmentof innovative therapies for patients with RA. Hope-fully, new genomic and proteomic information willallow further stratification identification of subsetsof RA patients who respond better, longer, and withfewer adverse reactions to targeted therapies.

Genetics and Environmental Stimulus

Genetic contributions to the pathogenesis of rheuma-toid arthritis (RA) are very complex. Genetic studiesin families of patients with RA suggest that monozy-gotic twins have a concordance rate for RA ofbetween 15%-30%. Thus, genetic factors contributeto susceptibility but are not the whole story. Obvi-ously, there must also be some triggering mechanism.It is certain that RA is a polygenic disease in whichsome genes may drive the initiation of the diseasewhile others may impact the response to therapy ordefine disease severity and prognosis.

Approximately 2 decades ago, certain alleles of theHLA-DR4 locus were found to be associated with RAin Caucasian patients. Most scientific evidence nowsuggests that certain genes within the major histo-compatibility complex (MHC), located on the shortend of chromosome 6 play the most significant role inthe initiation of the rheumatoid disease process. Sus-ceptibility to RA can be explained by the shared epi-tope hypothesis, ie, a 5-amino acid sequence found onseveral HLA-DR4 alleles defines a structural domainthat confers susceptibility to RA. Recent studies sug-gest that the clinical phenotype of RA (ie, destructivejoint disease and extraarticular manifestations) is alsoassociated with certain genes within the MHC. Evi-dence also suggests that there are RA susceptibilitygenes on other chromosomes that appear to correlatewith the phenotypic presentation of RA. Recently,allelic variations in PTPN22, a gene that encodes atyrosine kinase involved in inhibition of T-cell activa-tion, have been shown to be associated with RA.

There are several ways HLA-DR alleles could confersusceptibility to RA. In the simplest scenario, a pep-tide derived for example from a microbial proteincould bind to HLA-DR, and via standard antigen pre-sentation mechanisms trigger the proliferation ofantigen-specific T-lymphocytes. If the incitingmicrobe selectively infects synovial cells and theresulting immune response is unable to clear themicrobial infection, chronic synovial inflammationmight ensue. Alternatively, antigen-specific T-cellsactivated in response to a microbial peptide couldrecognize a self-peptide derived from a protein that isselectively expressed in synoviocytes, a processknown as molecular mimicry. However, althoughmyriad organisms have been proposed to predispose

to RA over the years, conclusive evidence for anysingle cause is lacking. Moreover, the dosage effectof HLA-DR in RA (ie, persons with 2 copies of theshared epitope containing MHC Class II allele aremore likely to have RA than those with a single copy)argues against simple antigen presentation being therelevant mechanism. More likely, presence of diseaseassociated HLA-DR alleles modifies the T-cell-receptor repertoire during thymic selection.

Other factors are also involved in the triggering ofRA in susceptible persons. One factor (not entirelyextragenetic) is female sex. Particularly at youngerages of onset, RA is more common in women than inmen, with a ratio of about 3:1. Whether this differ-ence reflects hormonal differences, genetic factorsinvolved in the presence or absence of the Y chromo-some, or genetic factors in the random inactivation ofone X chromosome is not known. Another factormay be the microbial flora (both pathogenic and com-mensurate) to which persons are differentiallyexposed, as mentioned above. Finally, it is possiblethat stochastic processes involved in the formation ofthe T-cell–receptor repertoire can, in a geneticallysusceptible person, produce a propensity for disease,the inheritance of which does not follow the rules ofMendelian genetics. At some point, the susceptiblehost’s genome, interacting perhaps with some exoge-nous factor, be it an infectious antigen or self-protein,induces the next phase in pathogenesis.

Intrasynovial Immune Response

The presence of activated T-cells, increased concen-trations of pro-inflammatory mediators, and endothe-lial activation within the synovium appear to impli-cate an aberrant self-perpetuating and self-amplifyingimmune response central to the initiation and suste-nance of rheumatoid synovitis. Animal models ofautoimmunity strongly support a role for the T-cell–receptor recognition of MHC–peptide com-plexes in the initiation and propagation of autoim-mune disease. In most animal models, the specificantigen responsible for the induction of autoimmunityis known (eg, type II collagen in collagen-inducedarthritis) and the MHC molecules that bind the pep-tide antigen, as well as the T-cell receptors that recog-nize the resulting complex, can be determined experi-mentally. Under these experimental conditions, inter-ventions that disrupt this quaternary complex (eg,


antibodies recognizing involved T-cell–receptor Vregions or specific MHC alleles) ameliorate autoim-mune disease. Both CD4+ and CD8+ T-cells can con-tribute to the initiation and propagation of autoim-mune disease in animal models. Often both cell typesare required to produce disease. Although it is likelythat human autoimmune disease is also initiated bythe recognition of self-MHC–peptide complexes byCD4+ or CD8+ T-cells, it is not yet possible to iden-tify triggering antigens. The identification of suchantigens would greatly facilitate the design of anti-gen-specific therapies for human autoimmune dis-ease. Although the evidence implicating CD4+ T-cells in the initiation of RA is strong, an essentialrole for other cell types is certain. Activatedmacrophages found in synovial tissues are a majorsource of cytokines such as tumor necrosis factor-alpha (TNF-�), which suggests an important role forthis inflammatory cytokine in this disease. Fibrob-lasts, mast cells and B-cells have also been shown tobe capable of significantly contributing to the gener-ation of rheumatoid inflammation.

Effector Molecules and Cartilage Damage

Once this biologic inflammatory process has matured,it appears to be regulated and amplified by inflamma-tory cells and their regulatory proteins, the mostnotable of which include the pro-inflammatorycytokines. T-cells and antigen presenting cells directly(by cell to cell contact) and indirectly (by other sig-nals) stimulate macrophages to secrete interleukin-1(IL-1) and tumor necrosis factor-alpha (TNF-�),among other inflammatory mediators. IL-1 and TNF-� or their mRNA have been identified in the serum,synovial fluid, and tissues of patients with RA. Both ofthese cytokines have broad biologic effects, includingthe induction by synoviocytes of matrix metallopro-teinases as well as the inhibition of tissue inhibitors ofmetalloproteinases (TIMPs). In vivo and in vitro stud-ies have documented that recombinant IL-1 and TNF-� injure normal hyaline cartilage. Chondrocytes, inparticular, enter a catabolic phase when exposed tothese cytokines, leading to an increase in collagenaseproduction and a decrease in matrix protein synthesis.TNF-� also induces adhesion molecules, eg, the intra-cellular adhesion molecule 1 (ICAM-1), which may inturn affect the homing and trafficking of inflammatorycells into the synovium. The “proof of concept” thatTNF is an important mediator in the pathobiology of

RA has been clearly shown by the profound clinicaleffects of the anti-TNF therapies (ie, infliximab, etan-ercept, and adalimumab), on the signs, symptoms andstructural damage that characterizes RA. Finally, theinvasive nature of the RA pannus has suggested tosome observers that the tissue behaves like a locallyinvasive tumor. Indeed, growth factors such as TGF-� and FGF produced locally in the synovium by resi-dent macrophages and fibroblasts may provide thedrive for tissue expansion.

Certain autoantibodies are associated with RA. Themost characteristic, certainly from a historical per-spective is rheumatoid factor (RF). The landmarkdiscovery of RF which is present in approximately75%-85% of people with RA, marked the begin-ning of the understanding that RA is a systemicautoimmine disease. Although it is relevant to thediagnosis of RA, RF may be of greater value to theclinician in prognosis. Epidemiologic observationssuggest that the presence of the RF is a marker thatpredicts more serious disease. Rheumatoid factorsare autoreactive antibodies that bind to the Fc por-tion of IgG molecules. RF is not specific forrheumatoid arthritis, as it is observed in a host ofother autoimmune disorders or infectious diseases.Also, by definition of how the test is performed, 5%of the normal population will have a positive testfor RF. RF of isotypes indicative of class switching(ie, IgG and IgA, rather than IgM) may be morespecific, as are higher titers. Although the potentialcontribution of RF to the pathogenesis of RA isunclear, it has been suggested that RF may inducetissue damage by acting as immune complexes.Over the years, a variety of other autoantibodieshave been reported to be associated with RA,including antikeratin antibodies, antiperinuclearfactor, and antifilaggrin antibody. It has beendemonstrated that the antigen recognized in com-mon by these antibodies is citrullinated protein.Citrulline is a nonstandard amino acid created bythe de-amination of arginine residues in proteins bythe enzyme peptidyl arginine deiminase (PADI).Interestingly, in some populations, allelic varia-tions in the genes encoding PADI have been shownto correlate with RA. Tests for anti-CCP (cyclic cit-rullinated peptide) antibodies are as sensitive as RFfor the diagnosis of RA and are more specific. Simi-larly, high titers of anti-CCP antibodies definepatients with a more aggressive disease phenotype.


2. Epidemiology

RA occurs worldwide in virtually all ethnic groups,with a prevalence estimated between 0.5% and 1%.Notable exceptions include a prevalence of 5% ormore in certain North American indigenous popula-tions (eg, the Yakima and Pima tribes) and a very lowprevalence in sub-Saharan Africa. It is significantlymore common in females than in males (about 3:1female: male ratio). Recently, an inception cohortstudy in Rochester, Minnesota established the annualincidence rate at 75.3 per 100,000 with a point preva-lence of ~1%. There are some interesting epidemio-logic trends from the Finnish Health Care system thatsuggest that over recent years the overall prevalenceof RA may be dropping and that it also may be occur-ring later in life.

Mortality rates of a cohort of 75 patients with RA fol-lowed over 15 years in a university hospital outpatientclinic have been reported to be 1.5- to 2-fold higherthan those of the United States population matched forage and gender. Similarly, an analysis of the MayoClinic and surrounding area population revealed thatalthough the life expectancy had increased in the gen-eral population from 1955 to the present, it had notchanged for those patients with RA. Death most oftenresults from infection, heart disease, respiratory fail-ure, renal failure, or gastrointestinal disease ratherthan from joint disease itself. Specifically, acceleratedatherosclerotic disease may afflict RA patients.Whether this is due to an inflammatory pathobiologyor to exposure to antirheumatic drugs (ie, glucocorti-coids) or both is unclear. A recent cohort study foundthat greater number of involved joints, decreasedfunctional status, presence of comorbid cardiovascu-lar disease, older age, and fewer years of formal edu-cation were all significant predictors of mortality inRA. Five-year survival of those patients with thepoorest functional status was 40%-60%, which iscomparable to that of patients with three-vessel coro-nary artery disease or stage IV Hodgkin’s disease.Notably, it seems that effective therapies may have apositive effect on this trend, and it has been shown thattreatment with methotrexate and TNF-inhibitors maypositively impact survival among RA patients.

3. Clinical Features

RA is characterized by symmetrical synovitisinvolving multiple diarthrodial joints as well asextra-articular features. Recent investigations haveclarified the severe effects that RA can have on a per-son’s function and mortality. Epidemiologic andgenetic studies have identified specific markers thatcorrelate with disease severity and prognosis. Withthe more widespread use of aggressive therapeuticregimens, including accelerated doses of methotrex-ate, combination DMARD therapy, new DMARDssuch as leflunomide, and the use of new biologicresponse modifiers (etanercept, infliximab, adali-mumab, and anakinra, abatacept and rituximab),some of the manifestations of RA appear to bechanging. For example, it has been suggested thatfeatures classically associated with very aggressivedisease, such as rheumatoid vasculitis, are less com-monly observed currently.

In 1987, the criteria for the classification of RA wererevised by a committee of the American College ofRheumatology (ACR). These diagnostic criteriawere not intended to establish a diagnosis clinically.Rather, they were designed to define homogeneouscohorts of patients for clinical research (Table 1).Likewise, in 1991 the Steinbrocker criteria for theclassification of functional capacity in RA wererevised by another committee of the ACR (Table 2).These criteria were designed to define the level ofclinical disability in RA patients, which is rapidlybecoming an important clinical endpoint and stratifi-cation variable in RA drug trials.

Clinical Presentation

Most patients with RA note a gradual onset of jointinflammation over weeks to months; however, 10%-20% may develop the classic joint stiffness, pain,and swelling of RA acutely. Some patients have awaxing and waning course sometimes referred to as“palindromic rheumatism”; it may be several yearsbefore the chronic, persistent features of RA becomeevident. Systemic symptoms, such as malaise,fatigue, and weakness often accompany the signsand symptoms of joint inflammation.

Based on cross-sectional studies of patients with RA,elderly-onset RA with onset at age 60 years or olderhas been considered to be a clinical entity distinctfrom younger-onset RA. However, the evidence from



Table 2

1991 American College of Rheumatology Revised Criteria for Classification of Functional Status in Rheumatoid Arthritis*

Class I Completely able to perform usual activities of daily living (self-care, vocational, and avocational)

Class II Able to perform usual self-care and vocational activities, but limited in avocational activities

Class III Able to perform usual self-care activities, but limited in vocational and avocational activities

Class IV Limited in ability to perform usual self-care, vocational, and avocational activities

* Usual self-care activities include dressing, feeding, bathing, grooming, and toileting. Avocational (recreational and/orleisure) and vocational (work, school, homemaking) activities are patient-desired and age- and sex-specific.

Table 1

1987 American College of Rheumatology Criteria forthe Classification of Rheumatoid Arthritis*

1. Morning stiffness in and around the joints lasting at least 1 hour before maximal improvement

2. At least 3 joint areas simultaneously have had soft tissue swelling or fluid (not bony overgrowth alone)observed by a physician. The 14 possible areas are: right or left PIP, MCP, wrist, elbow, knee, ankle, and MTP joints

3. At least 1 area swollen (as defined above) in a wrist, MCP, or PIP joint

4. Simultaneous involvement of the same joint areas (as defined in 2) on both sides of the body (bilateral involvement of PlPs, MCPs, or MTPs is acceptable without absolute symmetry)

5. Subcutaneous nodules over bony prominences, extensor surfaces, or in juxta-articular regions,observed by a physician

6. Demonstration of abnormal amounts of serum rheumatoid factor by any method for which the result has been positive in less than 5% of normal control subjects

7. Radiographic changes typical of rheumatoid arthritis on posteroanterior hand and wrist radiographs,which must include erosions or unequivocal bony decalcification localized in or most marked adjacentto the involved joints (osteoarthritis changes alone do not qualify)

* For the diagnosis of RA, 4 of the 7 criteria are required. MCP = metacarpophalangeal; MTP = metatarsophalangeal; PIP = proximal interphalangeal

observational studies has been conflicting. The clini-cal presentation of elderly-onset RA may be similar tothat of polymyalgia rheumatica, with an abrupt onsetof shoulder and hip joint inflammation. Other studieshave found the anatomic distribution of joint inflam-mation in patients with elderly-onset RA to be similarto that of patients with younger-onset RA. Rheuma-toid nodules and serum rheumatoid factor have beenobserved less often in patients with elderly-onset RAthan in those with younger-onset RA. Some authori-ties believe that elderly-onset RA usually follows arelatively benign course and generally responds wellto treatment. However, a recent study of consecutivepatients who presented to a university hospitalrheumatology clinic found few significant differencesbetween elderly-onset RA and younger-onset RA atthe time of disease onset.

Joint Manifestations

Morning gel describes a peculiar stiffness that is pro-nounced in the morning or after periods of inactivity.It plagues most patients with active inflammatoryarthritis, including those with RA. Its duration andquality oftentimes provide the clues to early diagno-sis of an inflammatory arthropathy. The clinical hall-mark of RA is a symmetrical polyarthritis involvingthe proximal interphalangeal joints (PIPs), metacar-pophalangeal joints (MCPs), wrists, elbows, shoul-ders, hips, knees, ankles, and metatarsophalangealjoints (MTPs). Although fully expressed RA mayinvolve multiple joints, it normally begins in one or afew joints and typically evolves in an additive fash-ion. The hands and wrists are affected in approxi-mately 90% of RA patients, and these joints mayshow unique changes over time. Early in RA, jointtenderness and subtle swelling are observed, butafter months to years the synovitis becomes prolifer-ative and destructive. The tissue within the jointbecomes boggy to the touch, then typical joint defor-mities appear, including ulnar drift at the MCPs,rotatory subluxation at the wrist (ulnar styloid), andthe “swan-neck” (flexion of DIP, hyperextension ofPIP) and “boutonniere” (hyperextension of DIP,flexion of PIP) deformities.

Although over 70% of patients with RA display radio-graphic evidence of erosions adjacent to small jointsof the hands and feet within the first 2 years after diag-nosis, about 25% of patients with early RA may not

develop erosive damage even after 8 years or more offollow-up. Most studies involving an inception cohortof patients suggest an overall, relatively linear rate ofjoint damage progression. The development of ero-sions correlates with the persistence of inflammation,as evidenced by clinical measures of morning stiff-ness, synovial swelling, and elevation of erythrocytesedimentation rate and other acute phase reactants.Magnetic resonance imaging (MRI) visualizes andquantifies tissue or “synovial load” and also detectsearly invasion of pannus into hyaline cartilage andsub-chondral bone. There is also increasing interest inthe use of ultrasonography, which can visualize bonyerosions, joint fluid, and synovial hypertrophy.

A syndrome of “pseudo-thrombophlebitis” can beseen in RA patients with active synovitis of the knee.Typically, a swollen, warm, tender calf causes prob-lems with weight bearing, and the physical examina-tion findings clearly mimic that of deep venousthrombophlebitis. The cause is typically a large,draining popliteal cyst, and for this reason, noninva-sive lower extremity studies of the deep venous sys-tem are typically negative.

The cervical spine is a common site of involvement inRA. In autopsy studies, up to 50% of patients with RAshow some cervical spine involvement. There is adirect relationship between cervical spine disease andthe presence of erosive peripheral joint disease. Prolif-erative synovitis can damage the ligaments and articu-lar cartilage of the cervical spine, causing severaltypes of cervical spine instability. These include: 1)atlantoaxial subluxation in 50%-70% of patients; 2)subaxial subluxation in 20%-25% of patients; and 3)basilar invagination into the foramen magnum aloneor in combination with atlantoaxial subluxation inapproximately 20% of patients. Symptoms of cervicalspine involvement in RA include neck pain, occipitalheadache, and paresthesias in the extremities. Basilarinvagination may cause symptoms of vertebrobasilarinsufficiency, such as tinnitus, vertigo, visual distur-bances, and dysphagia. Spinal cord compression maycause weakness or paralysis. Assessment of the poten-tial for neurologic injury is especially important inpatients with advanced RA requiring general anesthe-sia because of the risk of cervical cord compressionduring intubation. Plain lateral radiographs of the cer-vical spine with gentle flexion and extension views areused to quantify the degree of instability.



Pleuropulmonary Manifestations

Abnormalities of the pleura and lungs have beendetected by high-resolution computed tomography(HRCT) in up to 75% of patients with RA. However,pleura inflammation is often asymptomatic and maynot be detected until chest radiographs are done forother reasons. The prevalence of pleural effusion inRA is only about 5%. In some patients the pleuralfluid glucose may be dramatically low compared toserum levels. Although it may present rarely beforearthritis is manifested, pleural effusion usually occursin individuals with active established RA. About one-third of patients with RA with pleural effusions havecoexisting interstitial lung disease.

Interstitial pulmonary fibrosis occurs in 20%-40% ofpatients with RA, more commonly in patients withrheumatoid nodules and serum rheumatoid factor. Itcan be asymptomatic in its early stages because thedecreased physical activity of patients with RA maybe insufficient to induce dyspnea. The early radio-graphic changes of interstitial pulmonary fibrosis,consisting of ground-glass opacities predominantly inthe lower lungs, may not be evident on conventionalplain radiographs but appear as high-attenuationlesions at the periphery of the lungs on HRCT. Com-pared to HRCT, pulmonary function testing, espe-cially assessment of residual volume, is a more sensi-tive but relatively nonspecific indicator of interstitialpulmonary fibrosis in RA. Increased numbers of acti-vated T-lymphocytes, predominantly T-helper(CD4+) cells, are evident in the bronchoalveolar fluidof patients with RA with interstitial pulmonary fibro-sis, compared to patients with RA without lung dis-ease and healthy persons. Interstitial lung disease inpatients with RA may also be a complication of goldor methotrexate therapy, and there have also beenanecdotal reports associated with the TNF inhibitors.

Because transbronchial biopsies do not provideenough tissue for reliable diagnosis, the pattern oflung disease in patients with RA for whom treatmentfor pulmonary symptoms is being contemplated isbest determined by open-lung biopsy. Various histo-logic patterns have been described, including pul-monary rheumatoid nodules, usual interstitial pneu-monitis (UIP), bronchiolitis obliterans with patchyorganizing pneumonia (BOOP), lymphoid hyperpla-sia, and cellular interstitial infiltrates. Patients with

rheumatoid nodules, lymphoid hyperplasia, or non-specific cellular interstitial infiltrates have a betterprognosis than patients with UIP.

Pulmonary rheumatoid nodules may occasionally bevisualized on routine radiographs and confused withmalignancy. Occasionally, these nodules may cavi-tate. The unique entity of Caplan’s syndrome definesmultiple pulmonary nodules with cavitation in coalminers exposed to dust and other inhalants.

Cardiac Manifestations

Symptomatic cardiac involvement is rare in patientswith RA; however, echocardiography has shown theprevalence of asymptomatic cardiac involvement tobe higher than previously reported in autopsy studies.In a study of patients with RA and no cardiac symp-toms, there was a significantly increased prevalenceof posterior pericardial effusion (57.1%), mitral valveprolapse (34.3%), mitral valve thickening (22.9%),aortic root dilatation (34.3%), and aortic valve thick-ening (20.0%) compared to healthy persons.Myocarditis and endocarditis are not commonlyseen in patients with RA. Areas of focal myocarditisare seen at autopsy but are clinically insignificant.Arrhythmias secondary to involvement of the con-duction system by myocarditis, vasculitis, or nod-ules can occur but are exceedingly rare. Endocardialdisease in the form of nonspecific valvulitis has beenobserved in up to 70% of autopsied cases but is usu-ally asymptomatic and generally goes unrecognizedduring life. Rarely, valvular dysfunction may becaused by rheumatoid nodules. Aortitis, most com-monly involving the thoracic aorta, was identified in10 of 188 consecutive autopsied cases of patientswith RA and was associated with the presence ofserum rheumatoid factor, subcutaneous nodules, andrheumatoid vasculitis involving vessels other thanthe aorta.

Felty’s Syndrome and Pseudo-Felty’s Syndrome

Felty’s syndrome, which was reported in previousyears in as many as 1% of patients with RA, wasoriginally described as the association of RA withleukopenia and splenomegaly. It was frequentlyassociated with the presence of rheumatoid nodules,Sjögren’s disease, and other extra-articular manifes-

tations. Felty’s syndrome typically occurred in olderpatients with advanced erosive RA of at least 10 to20 years’ duration. Almost all Caucasian patientswith Felty’s syndrome were positive for HLA-DR4.Serum rheumatoid factor is almost always present,usually in high titer, and antinuclear antibodies(most commonly antihistone antibodies) are presentin 47%-100% of patients with Felty’s syndrome. Theprevalence of Felty’s syndrome and some othersevere extra-articular manifestations of RA appearto be decreasing over recent years, possibly in con-junction with the more common use of highly effec-tive treatment regimens.

Large granular lymphocytes (LGLs) normally com-prise 10%-15% of circulating peripheral bloodmononuclear cells. These LGLs contain azurophilicgranules scattered in the cytoplasm and can be dividedinto two subgroups: natural killer cells (CD3-,CD16+, CD56+) and CD3+ LGLs (CD3+, CD16+,CD56+), which are a subpopulation of T-cells. Bothsubgroups of LGL are capable of non-MHC-restrictedcytotoxicity and can secrete a variety of cytokines.Patients with clonal or polyclonal proliferations ofLGL, more commonly of CD3+ LGL, are consideredto have a form of chronic lymphoproliferative diseasecalled the large granular lymphocyte syndrome.Many of these patients’ cells display restricted T-cell–receptor gene rearrangements, but a specificcommon pattern of variable (V), diversity (D), andjoining (J) region T-cell receptor b-chain gene usage isnot observed when different patients are compared.

RA is the disease most frequently associated with theLGL syndrome (up to 39% of patients). Patients withthe LGL syndrome and RA exhibit idiopathic neu-tropenia and splenomegaly, a condition called pseudo-Felty’s syndrome. Because of an increased number ofLGLs, the total leukocyte count may be normal inpatients with pseudo-Felty’s syndrome. Similar topatients with Felty’s syndrome, patients with pseudo-Felty’s syndrome have serum rheumatoid factor andother autoantibodies and are susceptible to frequentinfections. However, extra-articular manifestationsand severe erosive arthritis are less common amongpatients with pseudo-Felty’s syndrome.

Rheumatoid Nodules

Rheumatoid nodules occur in approximately 15% to40% of patients with RA and are associated with thepresence of serum rheumatoid factor, erosive jointdestruction, necrotizing vasculitis, and other extra-articular manifestations of RA. They usually occur inareas that are repeatedly subjected to friction or pres-sure, such as the extensor surface of the forearm orthe Achilles tendon; however, nodules may alsodevelop in viscera such as the heart and lungs. Histo-logically, rheumatoid nodules contain a central areaof necrosis encircled by palisading histiocytes,which are surrounded by granulomatous tissue infil-trated with lymphocytes.

Patients receiving low-dose methotrexate therapy forRA may develop accelerated rheumatoid nodule for-mation despite good control of joint inflammation.These nodules are usually smaller than 5 mm and arelocated on the fingers, although they may be presentat other sites.

Malignancy

The overall incidence of cancer in patients with RA isnot increased; however, patients with RA have anincreased risk of developing lymphomas and lungcancer, and a decreased risk for developing cancers ofthe colon, rectum, and stomach. The low risk of col-orectal cancer among RA patients may be related totheir common use of nonsteroidal anti-inflammatorydrugs (NSAIDs).

Patients receiving methotrexate therapy for RA havedeveloped non-Hodgkin’s lymphoma, which mayregress with the discontinuation of methotrexatetherapy, as occurs with lymphomas in patientsreceiving immunosuppressive drugs after organtransplantation. These reversible lymphomas areusually diffuse, large-cell lymphomas of B-cell lin-eage. Epstein-Barr virus genome and latent mem-brane protein have been detected in lymphomas ofpatients with RA who were treated with methotrex-ate alone or with cyclosporin A, suggesting thatimmunosuppression may have contributed to thedevelopment of these lymphoid neoplasms.Recently, there has been some concern as to whetherpatients treated with TNF inhibitors may be atgreater risk of developing lymphomas, particularly


non-Hodgkin’s lymphoma (NHL). Analysis of anyincreased risk attributable to therapy is compoundedby the fact that patients with severe RA are at greaterrisk of developing NHL than the general population,and the risk correlates with the severity and activityof disease. Of note, this has been the subset of RApatients for whom TNF inhibitor therapy has beenmost widely utilized. At present, it seems that muchof the increased risk observed among patients treatedwith TNF inhibitors relates to the activity and sever-ity of RA, but this bears close observation.

4. Prognosis

The clinical expression of RA is usually established inthe first 2 years, although the disability measuresalmost always get worse with time. As mentioned pre-viously, mortality studies suggest that patients withsevere RA die at an accelerated rate due to infection,cardiopulmonary disease, and gastrointestinal bleed-ing. Results from a recent observational study suggestthat patients who are unresponsive to methotrexatemay have a significantly higher mortality risk com-pared to those patients who do respond to methotrex-ate. The presence of serum rheumatoid factor inpatients appears to best predict the subsequent devel-opment of RA for patients presenting with undifferen-tiated polyarthritis. Patients who have elevated titersof serum rheumatoid factor and anti-CCP antibodiesearly in the course of disease more often develop ero-sive joint disease than patients who are seronegative.High titers of rheumatoid factor are also associatedwith the appearance of extra-articular manifestations.Likewise, RA patients who have the HLA-DR4 hap-lotype usually have a poorer clinical outcome.

Most patients with RA have a slowly progressivecourse characterized by exacerbations and improve-ments. Long-term studies have shown that fewer than10% of patients with RA ever go into a prolongedremission. Greater number of tender or swollen jointsat disease onset has been correlated with increaseddisease activity and joint erosion. Lower socioeco-nomic status and fewer years of formal education,which is a component of or a surrogate marker forlower socioeconomic status, have been associatedwith higher morbidity and mortality in RA.

Defining markers of a bad prognosis early on in RAfacilitates therapeutic decision-making. Histori-cally, second-line therapy was reserved for severerefractory patients who had unremitting disease foryears. With recent evidence suggesting that erosionsappear early, there is a new emphasis on treating RApatients earlier with second-line agents. Bad prog-nostic features include: RF and anti-CCP seroposi-tivity; evaluated markers of inflammation (ESR andCRP); rheumatoid nodules; extra-articular features;larger numbers of active joints; early functionalimpairment; and early appearance of erosions.These features should trigger the instinct to treatmore aggressively. However, as always, the patientmust be educated about the risks and benefits ofsuch approaches and participate actively in the deci-


sion-making. Also, the above associations, whilecertainly true for large groups of patients, may notbe applicable for a single patient: The exceptionscan prove the rule.

5. Therapy

Background

RA is a chronic progressive disease, punctuated byperiods of exacerbation and remission. Over 5 to 10years, this “saw tooth” natural history leads to struc-tural joint damage and deteriorating musculoskeletalfunction. One subset of RA patients progress earlyand inexorably to functional disability and total jointreplacement despite aggressive treatment. On theother end of the clinical spectrum, some individualshave mild RA with long periods of inactivity, a pauci-articular disease expression, or a short course of activearthritis followed by total remission. Although thoseremitting patients have clinical findings that satisfypublished criteria for the diagnosis of RA during peri-ods of active disease, many experts believe that theymay in fact not have RA but some other reactive,inflammatory, self-limited arthropathy. Complicatingthis clinical heterogeneity is a lack of simple surrogatemarkers that indisputably predict long-term outcomein every single patient; therefore, in practice, cus-tomizing therapy by risk and benefit is difficult.

The history of anti-inflammatory drug developmentdates to the early part of the 20th century, when bedrest, painkillers, salicylates, and homeopathic reme-dies were standard practice. In the late 1940s and early1950s, corticosteroids were discovered. When admin-istered to patients with RA, the dramatic improve-ments they induced caused them to be considered“miracle drugs.” Soon thereafter, appreciation of theadverse effects related to their use tempered enthusi-asm. Other than acetylsalicylic acid (aspirin), the firstof many NSAIDs, phenylbutazone, was developedand also proved to be effective for patients with arthri-tis. Interestingly, it was during this same time thataminopterin, the parent compound of methotrexate,was initially identified and administered successfullyto patients with RA. Soon thereafter, gold salts wereproven to be effective in the first prototype of a well-designed trial of a disease-modifying antirheumaticdrug (DMARDs). Most recently, biologic drugsspecifically targeted to components of the immunesystem have been introduced into the clinic.


General Approach to Treatment

Although RA treatment approaches are highly indi-vidualized, according to practitioner experience andpatient preference, several trends deserve mention.Controlling pain is a critical objective, but RA withactive inflammation needs to be differentiated fromthe mechanical pain that can arise with joint defor-mity, because management strategies will differ.Specifically, pain from end-stage mechanical jointproblems may not necessarily require the use ofpotentially toxic immunosuppressive drugs. Thelevel of intensity of the therapeutic approach needs tobe developed with the cooperation of the patient.Treatment algorithms should always be used as aguideline to be modified by the actual clinical cir-cumstances—not as a rigid treatment protocol. Thecritical step in determining treatment guidelinesinvolves differentiating slowly progressive fromaggressive disease as outlined in Table 3.

Before 1989, when the early and widespread use ofmethotrexate became commonplace, the treatmentpyramid of RA was popularized. During this period,overly aggressive treatment with corticosteroids andother second-line drugs was eschewed because of theconcern for toxicity and because the disease wasregarded as nonfatal and only modestly disabling. Ini-tial drug therapy included months to years of high-dose salicylates or NSAIDs superimposed on a broadbase of patient education, rest, modest exercise, andthe use of assistive devices. Only with the widespreadappearance of erosions on radiographs would second-line drugs be prescribed. During the 1980s, mean dis-ease duration for many second-line drug trials was inthe 5- to 7-year range.

Traditional Treatment Pyramid

Prior to 1990, standard rheumatologic practiceinvolved a step-wise, conservative approach in theearly phases of the disease termed the RA treatmentpyramid. For the purpose of future discussion of RA


Table 3

Clinical Evaluation of Rheumatoid Arthritis

Classification of Disease Activity

Slowly Progressive Aggressive

Swollen joints Few Many

ESR or CRP Normal or modestly elevated Markedly elevated

Radiographic erosions Absent May be present

Rheumatoid nodules Absent May be present

Extra-articular manifestations Absent May be present

Rheumatoid factor or anti-CCP Normal or modestly elevated Markedly elevated

Functional status Preserved Impaired

CRP = C-reactive protein; ESR = erythrocyte sedimentation rate.Reprinted with permission from Lipsky, PE. Algorithms for the Diagnosis and Management of Musculoskeletal Complaints.Dallas, Tex: The University of Texas Southwestern Medical Center at Dallas; 1996.

treatment algorithms, this pyramid will be describedhere. The driving philosophy was that more toxic, butperhaps more effective, drugs would only be usedafter a prolonged period of persistent inflammation,often after patients had developed erosive disease. Atthe base of the pyramid lay the broad platform ofphysical medicine modalities, pain management,patient education, and “watchful waiting.” The nextlevel involved prescribing salicylates and non-steroidal anti-inflammatory drugs (NSAIDs) toreduce pain and suffering while optimizing exerciseand physical therapy approaches and awaiting (hope-fully) spontaneous disease remission. After a periodof unremitting disease (often lasting for years) fol-lowed by structural damage, the disease modifyingantirheumatic drugs (DMARDs) might be added asthe third tier of the pyramid. The rationale for thedelay here was that DMARDs (including intramuscu-lar gold, methotrexate, and immunosuppressivedrugs) were only worth the risk of adverse events inpatients with chronic, progressive disease who were atrisk of joint destruction. For the unfortunate patientwho did not respond to these modalities, the fourthlevel of treatment would include experimental thera-pies (usually under scientific protocol). Finally, gluco-corticoids (both systemic and intra-articular) could beused at all levels of the treatment pyramid for patientsat the early, middle, or later stages of disease.

Early Aggressive Therapy

Because patients treated by the pyramid approach fre-quently suffered through years of smoldering disease,a more aggressive approach has been embracedrecently. Although the rationale for this aggressivetreatment paradigm is being presently debated,emerging practice patterns suggest that rheumatolo-gists are adopting this approach. Several events haveforced a reevaluation of the traditional treatment pyra-mid. First, the toxicities (especially gastrointestinaland renal) of the traditional NSAIDs and salicylatesare now more apparent. Large databases of patientswith RA and other musculoskeletal diseases suggestan excessive rate of gastroduodenal ulcers, GI perfo-ration, or GI bleeding in individuals exposed chroni-cally to these drugs. So, the notion that these agents(used early and often in RA patients) have a “benign”toxicity profile is under increased scrutiny. Second,newer, more effective DMARD drugs are used earlierin patients with RA. Specifically, methotrexate shed

its negative aura as a cancer chemotherapy agent andrandomized, controlled trials showed impressive effi-cacy. As observational studies suggested, due to con-tinued long-term effectiveness and an acceptable toxi-city profile in the real world of rheumatology practice,methotrexate became the gold standard treatment forpatients with moderate RA. Third, long-term epi-demiologic studies now suggest that the conservative,step-wise pyramid treatment approach leaves manypatients with smoldering disease for long periods oftime, resulting in poor long-term functional outcome.In addition, several studies have documented initia-tion and significant progression of structural damageearly in the course of RA. Specifically, imaging tech-niques (especially MRI) have identified the presenceof early erosions and marrow effects of pannus in theabsence of radiographic abnormalities, suggestingthat irreversible damage may occur long before tradi-tional outcome parameters are able to detect it. Fourth,several simple prognostic clinical measures have beenlinked to poor outcome. These include: 1) persistentjoint inflammation or multiple affected joints; 2) ele-vated titers of rheumatoid factor or anti-CCP antibod-ies; 3) radiographic erosions; 4) extra-articular fea-tures; and 5) early functional impairment. Finally andmost importantly, the advances in molecular andgenetic medicine have elucidated a new understand-ing of RA pathogenesis that has now reached clinicalpractice. In the last decade, 6 new biologic therapiesthat have revolutionized the RA treatment schemehave been approved by regulatory authorities andintroduced into the clinic. These include: the p75-sol-uble TNF-� receptor-Ig Fc piece fusion protein (etan-ercept); a chimeric monoclonal antibody to TNF-�ligand (infliximab); a human anti-TNF-� antibody(adalimumab); an interleukin-1 receptor antagonist(anakinra), an inhibitor of T-cell costimulation(CTLA-4Ig; abatacept); and a chimeric monoclonalantibody to the B-cell molecule CD20.

These recent trends in rheumatology practice havedriven the concept of “inversion of the pyramid,” ie,DMARDs and the newer drugs are instituted early indisease and in novel combinations in RA patients hav-ing poor prognostic features. Concomitantly, the useof NSAIDs would be used for “breakthrough” pain,general palliation, or perhaps not at all (if pain is wellmanaged). NSAID drug holidays, therefore, are morefrequent, and hopefully, GI toxicity will be mini-mized. Likewise, COX-2 specific inhibitors, which


Diagnosis

Initiate NSAIDs

Outcome Assessment

Outcome Assessment

AcceptableResponse

· Initiate NSAIDs· Strongly Consider

- Oral Corticosteroids- Methotrexate

AggressiveDisease

Slowly ProgressiveDisease

AcceptableResponse

· Continued Disease Therapy· Progressive Disability· Progressive Joint Damage· Unacceptable Toxicity

Reevaluate:· Control Disease Activity· Functional Capacity· Joint Damage· Drug Toxicity

ContinueTherapy

Consider:· Altered Dosage of

Methotrexate· Intra-articular Corticosteroids· Another DMARD· Combination DMARDs· Biologic Agent· Immunosuppressive agent· Investigational Agent

ContinueTherapy

· ContinuedDisease Activity

· UnacceptableToxicity

Consider:· Another NSAID· Low Dose Oral

Corticosteroids· Intra-articular Corticosteroids· Hydroxychloroquine

· Continued Disease Activity· Progressive Disability· Progressive Joint Damage

OutcomeAssessment

ConsiderMethotrexate

Initial Evaluation

show an improved safety profile, are now widelyused. It has been theorized that this new approachwould both maximize benefit and minimize toxicity.Recent observations involving drug retention rate inrheumatology practice suggests that this practice pat-tern is finding its way into clinics. Based on publishedevidence, the medical management of RA is schema-tized in Figure 1. Critical to therapeutic success isconstant vigilance and systematic follow-up to evalu-ate effectiveness and toxicity. In this scheme,methotrexate is central. The newer drugs, anti-TNFagents, and other biologic agents may eventually beused in a different niche as post-marketing surveil-lance information becomes available.

Specific Therapeutic Approaches

Most rheumatologists agree that the judicious use ofNSAIDs or COX-2 specific drugs are still the usefuladjuncts, particularly during the early stages of activeRA for both slowly progressive and aggressive dis-ease. Full doses should be used and careful follow-uparranged so that second-line drugs can be prescribedwhen appropriate. Full doses of aspirin are accept-able but infrequently used because of poor patientcompliance and significant long-term tolerabilityproblems. Chronic NSAID use in RA has never beenshown to retard radiographic progression, and inaddition, NSAIDs may cause gastropathic sideeffects in high-risk patients. These drugs do, how-ever, partially reduce the pain and suffering that arethe hallmark of the disease.


Figure 1

Medical management of rheumatoid arthritis

If erosions appear early or if the disease showsaggressive features that are unabated despite NSAIDuse, then a decision to use a single second-line drugshould be made within 3 months or less (Table 4).Thus careful follow-up is crucial. Considerationshould be given to use hydroxychloroquine or sul-fasalazine if the disease is slowly progressive. Onthe other hand, if aggressive clinical features areobvious (erosions, sustained disease activity withmultiple joint involvement, or extra-articular fea-tures), then methotrexate alone or perhaps in combi-nation with hydroxychloroquine or sulfasalazinemight be a favored approach. Prednisone or equiva-lent is typically used during periods of very active

disease as “bridge therapy” and then tapered as thedisease comes under control with drug treatment.The early use of low-dose corticosteroids was shownto retard radiographic progression; however, nolong-term effect was observed on other key clinicalparameters such as functional status, and the long-term sequelae are still of concern. In most practicesin the United States, certain older DMARDs havebecome uncommon. Thus, because of adverseeffects related to their use, the introduction of neweragents and other factors, rheumatologists infre-quently use parenteral gold, oral gold, azathioprineand D-penicillamine.


Table 4

Rheumatoid Arthritis: Medical Management of Rheumatoid Arthritis with Disease-Modifying Antirheumatic Drugs

Slowly Progressive Disease Aggressive Disease

Consider in NSAID-nonresponsive patients Consider early in course of disease

Hydroxychloroquine Methotrexate, leflunomide

Sulfasalazine* Intramuscular gold (rarely used now)

Minocycline* Hydroxychloroquine

Sulfasalazine*

For patients refractory to the above, consider:

Etanercept, infliximab, or adalimumab

Combinations of DMARDs*

Cyclosporine*

Anakinra

For patients refractory to TNF inhibitors, consider:

Abatacept

Rituximab

* Not approved by the FDA for use in rheumatoid arthritis.DMARD = disease-modifying antirheumatic drug; NSAID = nonsteroidal anti-inflammatory drug.Reprinted with permission from Lipsky, PE. Algorithms for the Diagnosis and Management of Musculoskeletal Complaints.Dallas, Tex: The University of Texas Southwestern Medical Center at Dallas; 1996.

The biggest change in RA treatment in the last 5 yearsinvolves the use of multiple second-line agents concur-rently. Until recently this approach had been consid-ered overly aggressive and highly toxic. However, sev-eral well-designed, randomized, controlled trials ofcombination chemotherapy have shown improvedoutcomes. Almost all effective combination regimensuse methotrexate as the cornerstone. Combinations ofmethotrexate plus sulfasalazine and hydroxychloro-quine (or leflunomide, or cyclosporine) have allproven more efficacious than monotherapy. Mostimportantly, adverse effects with the combinationapproaches were comparable to those of monotherapy.The new TNF-� inhibitors have profound rapid effectson the clinical signs and symptoms of RA as well as onmood and constitutional symptoms. In addition, theyalso retard structural damage and improve quality oflife. While there has been a marked trend towardsincreasing use of these agents, eg, patients with earlierstages of disease, their ultimate practice niche is stillevolving. At present, etanercept, infliximab, and adali-mumab are most widely used for patients who fail torespond to methotrexate or who cannot take it becauseof comorbid diseases. Interestingly, recent data sug-gests that patients failing one TNF inhibitor, whetherdue to loss of efficacy or adverse effects, can stillachieve meaningful clinical responses when switchedto another TNF inhibitor. For patients failing TNFinhibitors or perhaps for those with contraindicationsto their use, the new biologic agents abatacept and rit-uximab may be tried.

A study in early RA from the Netherlands has pro-vided some insight as to the relative efficacy of varioustreatment approaches. In this study, known as the BeSttrial, 508 patients with early RA (defined as < 2 yearsof arthritis) were randomized to one of 4 treatmentarms: 1) sequential monotherapy (beginning withmethotrexate [MTX]); 2) step-up combination therapy(also beginning with MTX); 3) initial combinationtherapy with MTX, sulfasalazine (SSZ) and high doseprednisone (as was used in the earlier COBRA study);and 4) initial combination therapy with MTX plusTNF-inhibitor (infliximab). In all arms, the goal waslow levels of disease activity. At their 3 monthly fol-low-up visits, patients not achieving this goal, definedby a disease activity score (DAS) of 2.4 or less, wererequired to have their treatment altered according to analgorithm specific for each group. Eventually, groups1, 2, and 3 could end up on MTX plus TNF inhibitor. If

patients did achieve low disease activity on 2 succes-sive visits, treatment was tapered (to a minimum ofMTX 10 mg/week over the first 2 years, and off alltherapy after the second year). Data from the 1 yearand 2 years of follow-up showed that patients ingroups 3 and 4 achieved low disease activity and evenremission quicker than did those in groups 1 and 2,although as might have been expected given the studydesign with its mandatory changes in therapies, clini-cal efficacy was comparable across all groups by 2years. However, the progression of joint damage mea-sured radiographically was less in groups 3 and 4through the first 2 years of the study. Preliminaryassessment at year 3 has suggested that a substantialnumber of patients can have their therapy tapered oreven discontinued while maintaining low levels of dis-ease activity. This suggests that early aggressive ther-apy may be able to change the course of RA.

Surgery

Appropriately timed reconstructive orthopedicsurgery is an invaluable treatment strategy for selectedpatients with RA. Generally, surgery is most effectivefor patients with RA with the following conditions:impending tendon rupture in the hand and wrist (ten-don repair); functional limitation due to end-stagearthritis in large or weight-bearing joints (total jointreplacement of hip, knee, shoulder); joint instabilityor intractable pain with routine use (fusion of thumb,wrist, ankle, cervical spine).

Outcome Assessment

The optimal method for measuring outcome in thetreatment of patients with RA has always been an areaof discussion. Standard outcome parameters in RAtreatment trials include: tender and swollen jointcounts; global assessments of disease activity; assess-ment of pain by the patient; measurement of the acutephase response (eg, with the erythrocytsedimentationrate [ESR] or C-reactive protein [CRP]); assessmentof functional status (eg, using the Health AssessmentQuestionnaire [HAQ]) and assessment of joint dam-age; quantifying periarticular erosions; and joint spacenarrowing on radiographs.

Because individual measures are inexact, a trend hasbeen to use composite measures that incorporate sev-eral aspects of disease activity, and require important


improvements in a majority for a “response” to beachieved. The most commonly used are the responsecriteria developed by the ACR which require 20% ormore improvement in a variety of measures (com-monly called “ACR20 response”) and the diseaseactivity score (DAS) and its modifications (eg,DAS28). These composite indices are standard inclinical trials and are increasingly becoming morewidely used in practice.

This issue of quantifying outcome is important,because if measures are insensitive to change or donot describe clinically relevant conditions, then evalu-ations of therapies both in practice and in rigidlydesigned trials will be extremely difficult. It may bethat treatment in RA will follow the treatmentparadigm in oncology, such as induction of remis-sion/maintenance and relapse/reinduction strategy.One might then envision the newer, emerging biologictherapies to be used in conjunction with establishedsynthetic drugs. Emerging data suggest that thisindeed may be the case.

Because the TNF inhibitors have been shown to becapable of stopping the progression of radiographicjoint damage—a level of disease control previouslyconsidered unachievable—this has become increas-ingly utilized. Indeed, measures of radiographicdamage are standard in clinical trials now, andarresting such damage has become a goal in thetreatment of RA.

6. Specific Antirheumatic Drugs

NSAIDs and Salicylates

Mechanisms of action. Prostaglandins increase vas-cular permeability and sensitivity to bradykinins,thereby enhancing inflammation and increasing sensi-tivity to pain. NSAIDs inhibit the production ofprostaglandins of the E series, including the formationof prostacyclin and thromboxane, resulting in com-plex effects on vascular permeability and plateletaggregation. In the inflammatory cascade, polyunsat-urated fatty acids (including arachidonic acid), whichare constituents of all cell membranes and exist inester linkage with glycerols of phospholipids, areconverted to prostaglandins or leukotrienes through theaction of phospholipase A2 or phospholipase C. Freearachidonic acid acts as a substrate for theprostaglandin H (PGH) synthetase complex (PGHS),which includes both cyclooxygenase and peroxidase.These catalyze the conversion of arachidonic acid to theunstable cyclic-endoperoxide intermediates, PGG2 andPGH2, which then are converted to the more stablePGE2 and PGF2 compounds by specific synthetasesthat exist in specific tissues. NSAIDs specificallyinhibit cyclooxygenase activity and thereby reduce theconversion of arachidonic acid to PGG2.

One of the most important advances in the area ofinflammation involves recent elucidation ofcyclooxygenase (COX) biology. Accumulated evi-dence now suggests that there are at least 2 related iso-forms of the cyclooxygenase enzymes. They possesssignificant homology in those amino acid sequencesconsidered important for catalysis of arachidonic acid.The most important differences between the two iso-forms are in their regulation and expression of theenzymatic activity. To date, a constitutive form hasbeen described, prostaglandin synthetase H1 (PGHS-1 or COX-1), which is inhibited by all of the availableNSAIDs to varying degrees (depending on the appliedexperimental model system used to measure the drugeffects). COX-1 is expressed variably in most tissues;it has been described as a “housekeeping enzyme”that regulates normal cellular processes and is stimu-lated by hormones or growth factors. It is thought tobe important in maintaining the integrity of the gastricand duodenal mucosa, and its inhibition is consideredthe basic cause of many of the toxic effects of theNSAIDs. The other isoform is an inducible, highlyregulated enzyme, prostaglandin synthetase H2(PGHS-2 or COX-2), which is usually undetectable.


The expression of COX-2 is increased during states ofinflammation or experimentally in response to mito-genic stimuli. For example, in fibroblasts, variousgrowth factors, phorbol esters, and interleukin-1(IL-1) stimulate the expression of COX-2; in mono-cyte–macrophage systems, endotoxin stimulates it.The expression of COX-2 is inhibited by glucocorti-coids. Orally active NSAIDs and salicylates inhibitboth isoforms of cyclo-oxygenase, thereby reducingthe synthesis of prostanoids at sites of inflammation aswell as the “housekeeping” arachidonate mediators inrenal parenchyma and gastric mucosa.

Arachidonic acid can also serve as a substrate for 5- or12-lipoxygenase. These enzymes catalyze the conver-sion of arachidonic acid to biologically activeleukotriene and hydroperoxyeicosatetraenoic acids(HETEs). Although none of the currently availableNSAIDs inhibit 5-lipoxygenase directly, severalNSAIDs may indirectly reduce leukotriene levels byshunting arachidonic acid back into the triglyceridepool. Several new compounds may have inhibitoryeffects on both cyclooxygenase and lipoxygenase. Itremains to be seen whether these effects will be bio-logically important at the clinical level.

Data suggest that NSAIDs have other effects that maybe pertinent to their clinical effects; they inhibit acti-vation and chemotaxis of neutrophils as well asreduce the elaboration of toxic oxygen radicals fromstimulated neutrophils. There is also evidence thatseveral NSAIDs scavenge superoxide radicals.NSAIDs are lipophilic, becoming incorporated in thelipid bilayer of cell membranes and thereby interrupt-ing protein-protein interactions that are important forsignal transduction. Specific NSAIDs have beenshown to inhibit proteoglycan synthesis by chondro-cytes in vitro. There are in vitro data and a few casereports to suggest that the use of some NSAIDs accel-erates cartilage damage in OA, although few investi-gators believe the data to be compelling enough topreclude the use of NSAIDs as standard therapy forOA. Some NSAIDs have been shown to affect T-lym-phocyte function experimentally. Another newlydescribed biologic effect of NSAIDs not directlyrelated to prostaglandin synthesis involves the inter-ference with neutrophil-endothelial cell adherence,which is a critical step in the inflammation pathway.The effects of the NSAIDs on nitric oxide metabolism

have also been described. Nitric oxide synthetase,once induced by cytokines and other pro-inflamma-tory mediators, produces nitric oxide in largeamounts, leading to increased signs of inflammationsuch as vasocongestion, increased cytotoxicity, andincreased vascular permeability.

There is interesting evidence about the effects ofprostaglandins on cellular function. It has recentlybeen described that prostaglandins inhibit apoptosis(programmed cell death) and that NSAIDs, throughinhibition of prostaglandin synthesis, reestablish thereturn to a more normal cell cycle in some cell sys-tems. Thus, the suggestion is that in vitroprostaglandins are important modulators of apoptosisand that NSAIDs may alter this mechanism, thusaltering cell-cycle changes.

In summary, the literature is replete with reports ofpurported biologic efforts of NSAIDs; however, therole and importance of these prostaglandin- and non-prostaglandin-mediated processes in clinical inflam-mation are not entirely clear.

Pharmacology. All NSAIDs are well absorbed fromthe gastrointestinal tract, and because they are weaklyacidic, they become tightly bound to serum proteins.Therefore, alterations of circulating free drugs mayoccur when NSAIDs are co-administered with otherprotein-bound drugs or in patients with hypoalbu-minemia. In general, most NSAIDs are metabolizedin the liver and by-products excreted in the urine.Some NSAIDs are excreted in the bile and the urine.Others have prominent enterohepatic circulation.NSAIDs should be used with caution and vigilance inthe patient with liver or renal dysfunction.

Plasma half-life of the NSAIDs is highly variable; thispharmacokinetic heterogeneity reflects the wide vari-ability in dosing regimens. In general, the NSAIDs(and COX-2 specific anti-inflammatory drugs) withthe longest half-lives (eg, piroxicam, nabumetone,oxaprozin, and rofecoxib) are administered oncedaily; others (eg, ibuprofen) have short half-lives, andare administered 4 times per day in order to achieve abiologic and clinical effect. Some NSAIDs are lipidsoluble, readily penetrate the CNS, and occasionallycause side effects like HA, mentation difficulties, andmood alterations.



Table 5

Nonsteroidal Anti-Inflammatory Drugs and Dose Ranges by Chemical Classification

Daily Recommended Plasma Chemical Category Drug (trade name) Adult Dose (mg) Half-Life (hrs)

Carboxylic acids Acetylsalicylic acid (aspirin) 1300-5000 4-15

Choline magnesium 1500-3000 4-15Salicylate (Trilisate®)

Salsalate (Disalcid®) 1500-3000 4-15

Diflunisal (Dolobid®) 500-1500 7-15

Propionic acids Ibuprofen (Motrin®) 1600-3200 2

Naproxen (Naprosyn®) 500-1000 13

Fenoprofen (Nalfon®) 1200-3200 3

Ketoprofen (Orudis®) 150-300 2

Flurbiprofen (Ansaid®) 200-300 3-9

Oxaprozin (Daypro®) 600-1800 40-50

Acetic acids Indomethacin (Indocin®) 75-200 3-11

Tolmetin (Tolectin®) 800-1600 1

Sulindac (Clinoril®) 300-400 16

Diclofenac (Voltaren®) 100-200 1-2

Etodolac (Lodine®) 600-1200 2-4

Ketorolac (Toradol®)* 20-40 2

Fenamic acids Mefenamic acid (Ponstel®)† 500-1000 2

Meclofenamic acid (Meclomen®) 200-400 2-3

Enolic acids Roxicam (Feldene®) 10-20 30-86

Naphthylkanones Nabumetone (Relafen®) 500-1000 19-30

COX 2 inhibitors Celecoxib (Celebrex®) 200-400

Rofecoxib (Vioxx®) 12.5-25

Valdecoxib (Bextra®) 10-20

* Used for management of acute pain (not recommended for use for more than 21 days).† Used for management of pain or dysmenorrhea (not recommended for more than 7 days).

Two strategies presently exist to prevent GI ulcera-tion. The first involves the concomitant use of miso-prostol, a methylated, orally-active prostanoid. Ran-domized controlled trials and large post-marketing,population-based outcomes studies have proven thatmisoprostol reduces the risk of serious GI complica-tions including perforation, ulcers, and bleeding by upto 50%. However misoprostol at high doses itself pro-duces annoying GI side effects including abdominalpain and diarrhea. Recently, a combination ofdiclofenac (an NSAID) and misoprostol (in dailydoses lower than label suggestion) reduced the fre-quency of NSAID-induced ulcers by 60%-80%, whileshowing significant symptomatic efficacy. The sec-ond strategy builds upon the elucidation of the 2 sepa-rate isoforms of cyclo-oxygenase, COX-1 and COX-2. Specific inhibitors of COX-2, the isoform inducedby biologic mediators of inflammation, have beendeveloped into anti-inflammatory drugs that are selec-tive, and do not inhibit COX-1, when used in physio-logic doses.

Three COX-2 specific inhibitors, celecoxib, rofe-coxib, and valdecoxib, had been approved by variousregulatory agencies both in the United States andaround the world. Several other COX-2 inhibitors,including etoricoxib and lumaricoxib, progressedthrough late phases of the drug development pathwayand were even approved in a number of countries. Inclinical trials and in the clinic, all the COX-3inhibitors successfully achieved their main clinicaladvantage, namely the control of signs and symptomsof arthritis while reducing gastrointestinal complica-tions typical of nonselective NSAIDs, particularlyNSAID gastropathy. However, the use of COX-2inhibitors was dramatically altered when it was shownthat patients treated with rofecoxib experienced sig-nificant complications of atherosclerotic cardiovascu-lar disease more often than patients treated with olderNSAIDs. Subsequently, rofecoxib was removed fromthe market worldwide, valdecoxib was withdrawn byits manufacturer, and drugs in development had theirplans put on hold. At present, celecoxib remains theonly COX-2 inhibitor approved for use in the US andmany other countries. Analysis of this area is dynamicand evolving.

Clinical use. Overall, the magnitude of therapeuticeffects of NSAIDs is similar when studied in popula-tions of arthritis patients in clinical trials. However,wide variability in response to particular NSAIDs hasbeen observed in individual patients. These drugs donot appear to retard the development or progression ofchondral erosions. However, NSAIDs do offer pallia-tion with respect to pain and suffering, as attested toby the “flare” phenomena frequently observed in clin-ical trials when these drugs are withdrawn frompatients with active RA. Despite the recent philosoph-ical change regarding the traditional treatment pyra-mid, NSAIDs are routinely used in RA patients toreduce the symptoms of pain. Most practitioners nowrecognize that they are purely soothing for muscu-loskeletal symptoms, but do not provide a “disease-modifying” biologic effect, and have significant atten-dant toxicity. Thus, there is increasing vigilanceregarding their use. As a class of drug, NSAIDs, espe-cially over-the-counter NSAID preparations, are usedroutinely in primary care practice. The NSAIDs andtypical dosages and regimens are outlined in Table 5.

Adverse effects. Adverse effects of traditional COX-1 inhibiting NSAIDs are a major issue, especially inpatients at high risk for toxicity. GI side effects in par-ticular are a major reason for discontinuing NSAIDs.Large databases suggest a hierarchy of toxicity, withtolmetin and indomethacin exhibiting the most toxic-ity and ibuprofen and nonacetylated salicylates theleast. Large population-based epidemiologic studiessuggest that patients exposed to NSAIDs exhibit a 3-10 fold higher risk of GI toxicity and death than thosewho are not exposed. GI side effects range in severityfrom mild dyspepsia and flatulence to death frommassive GI blood loss or visceral perforation. Com-plicating the surveillance issue for side effects is theobservation that up to 40% of patients with docu-mented gastroduodenal erosive disease may not com-plain of symptoms. Although observations gleanedfrom randomized controlled trials may not reflect thetrue natural history in large populations, endoscopicstudies in short-term trials have identified gastroduo-denal ulcers with a frequency of 15%-30%. Risk factors for GI adverse events in the RA populationinclude age, concomitant use of NSAIDs, priorNSAID related side effects, and level of disability.



Glucocorticoids

Since their discovery in the late 1940s by Hench andcolleagues at the Mayo Clinic, glucocorticoids havebeen used to treat patients with RA. Although earlyobservations suggested that these compounds exhib-ited profound clinical effects in patients with RA, thelong-term adverse event profile severely limited thewidespread and prolonged usage of these drugs.Because of the disastrous side effects observed in theearly 1950s, some academic centers and privaterheumatologists avoided the use of these drugs for anypatient with RA. What followed was a profoundantipathy on the part of rheumatologists to use gluco-corticoids for nonfatal inflammatory diseases. Gradu-ally, these agents were accepted as short- or intermedi-ate-term “bridge therapy” for active RA while await-ing disease amelioration by second-line agents. Inaddition, several studies have now shown that treat-ment with corticosteroids can slow the progression ofradiographic joint damage in RA.

Since those early days, the practice pendulum hasswung back, and glucocorticoids are now routinelybut cautiously prescribed for patients with active RA.

Mechanism of action. The mode of action of gluco-corticoids is complex. They bind specifically to acytoplasmic receptor, which belongs to a super familyof regulatory proteins (including the estrogen andvitamin D receptors). The glucocorticoid-receptorcomplex migrates to the nucleus and then affects genetranscription. Glucocorticoids have profound and pro-tean anti-inflammatory effects caused by the suppres-sion of immunomodulating proteins, including IL-1,IL-6, TNF-�, interferon-gamma, and GMCSF. Inaddition, glucocorticoids inhibit the production andexpression of pro-inflammatory prostaglandins andleukotrienes, inducible nitric oxide synthase, plas-minogen activator, and adhesion molecules (ICAM-1). Glucocorticoids also modulate cellular physiologyby reducing neutrophils at sites of inflammation;decreasing the number and function ofmonocytes/macrophages (perhaps by affectingchemokine physiology); suppressing antigen present-ing cells; and inhibiting the number of circulatinglymphocytes and their functions, including the prolif-erative response to mitogens, cytokine production,and immunoglobulin production. In addition, gluco-

corticoids also produce significant metabolic effectselsewhere. Generally speaking, they are catabolic andinduce breakdown of protein, diminished utilizationof glucose, and alterations in lipids, thereby facilitat-ing premature atherosclerosis.

Pharmacology. Glucocorticoids are well-absorbedfrom the gastrointestinal tract. Once absorbed, thedrug binds to corticosteroid binding globulin (CBG)and albumin. Glucocorticoids are metabolized largelyby the liver, where they are hydroxylated, conjugated,and then excreted in the urine. Prednisone is an inac-tive steroid, but once absorbed it is converted in theliver to prednisolone. The half-life and biologiceffects of various glucocorticoid preparations arehighly variable. Differences in plasma half-life andother factors contribute to significant differences inclinical potency. Most practitioners divide glucocorti-coids into 3 groups according to their duration ofaction. Short-acting glucocorticoids (8-12 hours)include cortisol and cortisone; intermediate-actingglucocorticoids (12-36 hours) include prednisone,prednisolone, and methylprednisolone; and finally,long-acting glucocorticoids (36-72 hours) includedexamethasone.

Clinical use. Low dosages of glucocorticoids co-administered with NSAIDs and DMARDs are usedroutinely as “bridge therapy” for some patients withvery active RA. In that setting, the plan is to firstquickly extinguish the signs and symptoms of the“flare” phenomena, but then to always taper gluco-corticoid treatment once the DMARD takes effect. Insome patients with chronic progressive disease, long-term low-dose glucocorticoid treatment is required tomaintain the disease at a certain level of control andto secure a reasonable quality of life. Glucocorticoidsare also used intermittently over 1-2 weeks in dosestapering rapidly from 20 mg per day for patientsexperiencing a “flare” of RA.

New information now exists regarding the magnitudeand durability of these effects as well as the potentialfor retarding cartilage damage and structural deterio-ration of the joint. In 1995, Kirwan showed that low-dose glucocorticoids resulted in a reduction of pro-gression of a joint damage in patients with RA whowere taking second-line antirheumatic drugs. Onehundred and twenty-eight RA patients with averagedisease duration of less than 2 years were randomized


in double-blind fashion to either placebo or low-doseprednisolone. All patients were receiving stable dosesof NSAIDs or second-line antirheumatic agents. Clin-ical improvement of the signs and symptoms of RA inthe prednisolone group was greater than in theplacebo group after the first year of therapy but disap-peared at 2 years. However, at 2 years, the erosionscores were significantly better in the prednisolonegroup compared to the placebo group. Kirwan’s obser-vations were the first to define a potential “diseasemodifying” capability of glucocorticoids.

Building upon that evidence, a recent 2-year double-blind randomized placebo-controlled trial was com-pleted in patients with early RA who were not concur-rently exposed to second-line antirheumatic drugs.Similar to the Kirwan study, the glucocorticoid groupshowed more clinical improvement than the placebogroup early on in the observation period, but the dif-ference between the groups was insignificant at 6 and12 months. However, radiologic outcome showed sig-nificant reduction in erosion and joint space narrow-ing scores in the prednisone group versus the placebogroup. These observations suggest that glucocorti-coids may have “structure-modifying” effects despitethe potential for long-term toxicity.

Given the recent additions to the pharmacologicarmamentarium to treat glucocorticoid-inducedosteoporosis and glucocorticoid-associated gastroin-testinal side effects, the concern about long-term sideeffects has been mitigated to some extent. However,given all of the other potential adverse effects of glu-cocorticoids, these agents should always be usedjudiciously. A thorough discussion of risk benefit andappropriate concern for patient expectations shouldbe part of the decision-making process when usingthese drugs.

Adverse effects. The adverse effects of glucocorti-coids are legion and are noted in Table 6. For the most part, these side effects are dose- and duration-dependent. Of particular importance in therheumatic disease population are immune suppres-sion, and osteoporosis. Other side effects may alsobe seen in RA patients, including aggravated hyper-tension, glucose intolerance, stimulation of cataractgrowth, difficulties with wound healing and, occa-sionally, CNS effects ranging from mild emotionallability to frank psychosis.

Table 6

Side Effects of Corticosteroid Therapy

Very common, should be considered in all patients

OsteoporosisIncreased appetiteCentripetal obesityImpaired wound healingIncreased risk of infectionSuppression of hypothalamic-pituitary axisArrest of normal growth in children

Frequently seen

MyopathyOsteonecrosisHypertensionThin fragile skin, striae, purpuraEdemaHyperlipidemiaPsychiatric symptoms, particularly euphoriaDiabetes mellitusPosterior subcapsular cataracts

Uncommon, but important to recognize early

GlaucomaSilent intestinal perforationPeptic ulcer diseaseGastric hemorrhageHypokalemic alkalosisHyperosmolar nonketotic coma

Rare

HirsutismPanniculitisSecondary amenorrheaImpotenceEpidural lipomatosis

Modified from Behrens TW, Goodwin JS. Glucocorticoids.In: McCarty D, Koopman W, eds. Arthritis and Allied Con-ditions. 11th ed. Philadelphia, Pa: Lea & Febiger;1989:604-21.


Table 7

Monitoring Drug Therapy in Rheumatoid Arthritis

Recommended monitoring strategies for drug treatment of rheumatoid arthritis*

Toxicities MonitoringRequiring Baseline System Review/

Drugs Monitoring† Evaluation Examination Laboratory

Salicylates, Gastrointestinal CBC, creatinine, Dark/black stool, CBC yearly, LFTs,nonsteroidal anti- ulceration AST, ALT dyspepsia, nausea creatinine testinginflammatory and bleeding or vomiting, abdominal may be required‡

drugs pain, edema,shortness of breath

Hydroxy- Macular damage None unless patient Visual changes, chloroquine is over age 40 or consider funduscopic

has previous eye and visual fields every disease 12 months

Sulfasalazine Myelosuppression CBC, and AST Symptoms of CBC every 2-4 or ALT in patients myelosuppression§, weeks for first 3 at risk, consider photosensitivity, rash months, then G6PD every 3 months

(continued on next page)

The effect of glucocorticoids on bone mineral densityis probably the most vexing complication for RApatients exposed to these compounds, especiallybecause many of these patients are post-menopausalwomen. Bone mineral density loss in these patientscan range from 4%-10% per year, and even low-doseglucocorticoid therapy may cause significant osteo-porosis in some patients. Most of the bone loss occursearly. Glucocorticoids cause osteoporosis via severalmechanisms, including stimulation of osteoclasticbone resorption, reduction of calcium absorption,reduction in sex hormone production, and finally, adirect negative effect on bone formation.

Strategies to minimize glucocorticoid-associatedosteoporosis have been developed recently andinclude calcium supplementation as a base withvitamin D. In addition, drugs that reduce osteoclas-tic bone resorption, (eg, the biphosphonates, calci-

tonin ((intranasal and parenteral)), PTH hormone,estrogen replacement for women, and testosteronereplacement for hypogonadal men), are the keytherapeutic agents that may reduce bone mineraldensity loss.

Acetaminophen

Acetaminophen has analgesic and antipyretic proper-ties. The exact mechanism of action is unknown, but itmay have both peripheral and central actions. Thismedication is well absorbed after oral administrationand may be effective within 30 to 60 minutes. Thehalf-life and duration of action is approximately 4hours. This drug is metabolized in the liver andexcreted in the urine. Overall, it is safe and producesless gastrointestinal dyspepsia and bleeding than theNSAIDs. Acetaminophen is used routinely in RApatients to manage pain due to mechanical joint prob-lems. This agent may cause hepatotoxicity in patients


Table 7 (continued)

Monitoring Drug Therapy in Rheumatoid Arthritis

Toxicities MonitoringRequiring Baseline System Review/

Drugs Monitoring† Evaluation Examination Laboratory

Methotrexate Myelosuppression, CBC, chest Symptoms of CBC, platelethepatic fibrosis, radiography myelosuppression§, count, AST cirrhosis, pulmonary within past year, shortness of breath, albumin, creatinineinfiltrates or fibrosis hepatitis B and C nausea/vomiting, every 4-8 weeks

serology in high- lymph node swellingrisk patients, AST or ALT, albumin, alkaline phosphatase, and creatinine

Leflunomide Hepatoxicity, myelosuppression, diarrhea (same as methotrexate for last 3 columns)

Gold, Myelosuppression, CBC, platelet count, Symptoms of CBC, platelet intramuscular proteinuria creatine, urine myelosuppression§, count, urine

dipstick for protein edema, rash, oral dipstick every 1-2 ulcers, diarrhea weeks for first 20

weeks, then at the time of each (orevery other) injection

Gold, oral Myelosuppression, CBC, platelet count, Symptoms of CBC, platelet proteinuria urine dipstick for myelosuppression§, count, urine

protein edema, rash, diarrhea dipstick for protein every 4-12 weeks

Corticosteroids Hypertension, BP, chemistry panel, BP at each visit, Urinalysis for (oral <10 mg of hyperglycemia bone densitometry polyuria, polydipsia, glucose yearlyprednisone or in high-risk patients edema, shortness of equivalent) breath, visual changes,

weight gain

*CBC = complete blood cell count (hematocrit, hemoglobin, white blood cell count) including differential cell and plateletcounts; ALT = alanine aminotransferase; AST = aspartate aminotransferase; LFTs = liver function tests; BP = blood pressure.† Potential serious toxicities that may be detected by monitoring before they have become clinically apparent or harmful tothe patient. This list mentions toxicities that occur frequently enough to justify monitoring. Patients with comorbidity, con-current medications, and other specific risk factors may need further studies to monitor for specific toxicity. ‡ Package insert for diclofenac (Voltaren®) recommends that AST and ALT be monitored within the first 8 weeks of treat-ment and periodically thereafter. Monitoring of serum creatinine should be performed weekly for at least 3 weeks inpatients receiving concomitant angiotensin-converting enzyme inhibitors or diuretics.§ Symptoms of myelosuppression include fever, symptoms of infection, easy bruisability, and bleeding.

American College of Rheumatology Ad Hoc Committee on Clinical Guidelines: Guidelines for monitoring drug therapy inrheumatoid arthritis. Arthritis Rheum. 1996;39:723-31. Reprinted with permission of the American College of Rheumatology.


with preexisting liver dysfunction. The issue of renaldysfunction secondary to acetaminophen hasreceived significant attention. A case-control study(telephone sampling method) examining 716 patientswith end stage renal disease from all causes and 361age-similar controls from the same geographic areasuggested that an average intake of acetaminophen ofmore than one pill per day and a cumulative intake of1000 or more pills in a lifetime were associated withan increased risk of end stage renal disease. Thisstudy should be interpreted with caution due topotential limitations, including recall bias, confound-ing factors associated with acetaminophen use in thecases that were not examined in the study, and thelack of control over accuracy of the informationreported by telephone interview. The true risk ofacetaminophen use in the development of end stagerenal disease remains unknown.

DMARDs

Antimalarial Drugs Mechanism of action and pharmacology. The anti-malarial drugs, chloroquine and hydroxychloroquine,are 4-aminoquinolones. Quinine, a related compound,has been used since the 1800s to treat rheumatic dis-eases. These drugs are well absorbed from the gas-trointestinal tract; hydroxychloroquine is excreted inthe feces and chloroquine in the urine. Although theexact mechanisms of action have not been completelyelucidated, the antimalarial drugs have been shown toinhibit the release of IL-1 by monocytes. They inhibitlysosomal enzyme release by stabilizing lysosomalenzymes, and they retard phagocyte chemotaxis andlymphocyte blastogenesis. They have also beenshown to inhibit phospholipase A2, a critical enzymein the prostaglandin cascade.

Clinical use. Although antimalarial drugs fell out offavor for the treatment of rheumatic diseases in the1950s because of the retinal toxicity that was observedwith long-term and high-dose therapy, these drugshave made a comeback in recent years. Several con-trolled trials have proven hydroxychloroquine to beeffective for patients with mild SLE whose manifesta-tions are primarily dermatitis and arthritis. Thesedrugs are also used in early RA (usually beforemethotrexate, parenteral gold, and anticytokine ther-apy), or in more aggressive RA as combinationchemotherapy with methotrexate. A recent controlled

trial of hydroxychloroquine showed it to be signifi-cantly more effective than placebo in patients withearly RA. Clinical effects are usually observed at 8 to12 weeks; starting dosage is 400 mg/d in divideddoses, with a dose reduction to 200 mg/d at 8 to 12weeks if a good response is observed.

Adverse effects. Adverse events are common andprobably dose related, thus providing some ratio-nale for the lower doses presently in use. Gastroin-testinal intolerance with dyspepsia, diarrhea, andnausea are common; however, gastrointestinalbleeding is not a usual sequelae of these symptoms.Peripheral neuropathic symptoms, myopathy, andskin rashes (including exfoliative dermatitis) havealso been reported. Of most concern is the potentialfor ocular toxicity, which includes central effects onvisual accommodation, corneal deposits, andretinopathy. The latter is associated with permanentvisual loss, but at the doses prescribed today, thisside effect is very rare in clinical practice. Symp-toms include central scotomata, peripheral fielddeficits, or frank blindness. Regular ophthalmologicexamination (according to American Academy ofOphthalmology Guidelines), which should includevisualization of the macula for pigmented stipplingas well as peripheral field testing to a red test object,may define a state of premaculopathy that predictsthe eventual irreversible changes (Table 7).

SulfasalazineSulfasalazine is 5-aminosalicylic acid linked by adiazo bond to the antibacterial sulfonamide, sulfapyri-dine. In Western Europe, it has been used since the1940s for patients with RA. Because of its favorableeffects and relative lack of toxicity, it is now morewidely used in the United States, especially in earlymild RA.

Mechanism of action. The exact mechanism of actionof sulfasalazine is poorly understood and is compli-cated by the fact that the metabolites, 5-aminosalicylicacid and sulfapyridine, are also biologically active.Most compelling is the fact that the parent molecule,sulfasalazine, can inhibit 5-lipoxygenase, thus reduc-ing production of leukotrienes. It has also been shownto inhibit lymphocyte mitogenesis and nuclear factorkappa B, a transcription factor important in the geneticregulation of inflammatory pathways.

Pharmacology. Because of its insolubility, sul-fasalazine is poorly absorbed in the stomach and smallbowel. In the colon, the diazo bond is reduced by bac-teria, thus liberating 5-aminosalicylic acid, most ofwhich remains in the colon and is excreted in thefeces. Sulfapyridine is absorbed, metabolized in theliver, and excreted in the urine.

Clinical use. Sulfasalazine is widely used inpatients with early mild RA, and most long-termstudies suggest a favorable efficacy profile. Mostrecently, sulfasalazine has also been shown to bemarginally effective in patients with the seronega-tive spondyloarthropathies, especially in those withpsoriatic arthritis. It apparently has little effect onaxial spine disease. It is commonly prescribed as partof a combination chemotherapy protocol withhydroxychloroquine and methotrexate. The startingdosage is usually 500 mg twice a day, with gradualescalation to a total of 3 g/d in two divided doses.

Adverse effects. Adverse effects are common butusually minor in magnitude or severity. Most long-term RA studies of sulfasalazine have recorded veryfew serious adverse events. Previous sulfa allergyprecludes its use, and glucose-6-phosphate dehy-drogenase (G6PD) screens may be considered inappropriate patients. Gastrointestinal intolerance,including nausea and abdominal discomfort, areusually seen in the first 2 to 3 months of therapy andmay be minimized by gradually increasing the dose.Cutaneous adverse events include urticaria, pruri-tus, maculopapular rash, and photosensitive erup-tion. As with any sulfa drug, serum sickness or toxicepidermal necrolysis rarely can be seen. Leukope-nia and neutropenia, which are usually clinicallyinsignificant, are observed in 1% to 5% of exposedpatients; therefore, periodic hemogram screening issuggested (Table 7).

Methotrexate

Methotrexate is a methylated analogue of tetrahydro-folate and inhibits dihydrofolate reductase. Its historyas an antirheumatic drug dates to 1950, when a caseseries of patients with inflammatory arthritis improvedsignificantly after taking aminopterin, a drug related tomethotrexate. Although it remained in use for psoria-sis, methotrexate was largely abandoned as a treatmentfor RA until the 1970s, when it began to see limited use

by individual practitioners. Subsequently, it was sys-tematically evaluated and proven efficacious in severallarge, randomized, controlled trials in the 1980s.Today, its clinical utility in the treatment of patientswith RA is undisputed, and it stands alone as the goldstandard second-line antirheumatic drug against whichothers are compared. Even the new TNF inhibitors aretypically added to methotrexate “partial responders” orafter methotrexate has failed or caused toxicity.

Mechanism of action. At the cellular level,methotrexate is transported across the cell membrane,binds to dihydrofolate reductase, reduces intracellularfolate pools, and inhibits purine synthesis. Intracellu-lar methotrexate is metabolized to polyglutamates,which serve to amplify inhibition of dihydrofolatereductase. The antimetabolic effects of methotrexatecan be reversed with exogenously applied reducedfolates, such as leucovorin (folinic acid).

In patients with RA, methotrexate may not function asit does in cancer patients receiving antimetabolitechemotherapy. Indeed, its prompt clinical effects sug-gest inhibition of inflammation as a possible mecha-nism of action. However, methotrexate does notinhibit cyclooxygenase. It does inhibit IL-1 andleukotriene-4 (LTB4), and reduces levels of prote-olytic enzymes. Methotrexate, by virtue of the bio-logic effects of intracellular polyglutamates, mayinduce extracellular expression of adenosine, which inturn may have an inhibitory effect upon cells that par-ticipate in inflammation.

Pharmacology. Unlike most other second-lineantirheumatic drugs, methotrexate’s pharmacokineticprofile and mechanisms of action have been well stud-ied. In low oral doses, the drug is well absorbed, withpeak plasma levels occurring in 1 to 2 hours; at higherdoses, however, the fraction of drug absorbed actuallydecreases because of saturation of carrier-mediatedgastrointestinal absorption. In addition, the presenceof food may decrease bioavailability. When the drugis given parenterally, its pharmacokinetic profile issomewhat different. Methotrexate is protein boundand may be displaced by other protein-bound drugs,including NSAIDs. Methotrexate is excreted in theurine by a process of filtration, reabsorption, and tubu-lar secretion, and thus blood levels might be affectedby drugs that interfere with these processes.



Clinical use. Widely utilized in many connective tis-sue and autoimmune diseases, methotrexate has beenshown to be an excellent steroid-sparing agent. It iseffective in the spondyloarthropathies, psoriaticarthritis, inflammatory myopathies, systemic vasculi-tis, and nonrenal SLE. However, it has been mostextensively studied in patients with RA. Multipleshort-term and long-term studies have shown supe-rior clinical efficacy. Patients with RA takemethotrexate for years, attesting to its favorable effi-cacy-to-toxicity profile. In most studies, methotrex-ate produces a predictable response at 6 to 8 weeks,which peaks at 6 months and then requires graduallyincreasing dosages of the drug in order to maintainthe improvement. Although methotrexate may notproduce remission in many patients, it may retardradiographic progression. Dosing usually begins at7.5 to 10 mg orally per week, with gradual increasesup to maximum dosage of 20 to 25 mg orally perweek depending on clinical response. If the clinicaleffect is suboptimal or if gastrointestinal adverseeffects occur, methotrexate at similar doses can begiven parenterally. Most recently it has been tested incombination with other antirheumatic drugs, such ashydroxychloroquine and sulfasalazine in combina-tion, and with cyclosporine alone.

Adverse effects. Adverse events may be classified asminor and major. Minor adverse effects, which arecommon, include stomatitis, gastrointestinal upset,headache or minor central nervous system distur-bance, and elevations of levels of liver transaminases.All of these minor adverse effects may be averted byconcomitant use of folic acid. Alternatively, changingdose, dose regimen, or route of administration mayalso reduce these annoying side effects.

The more serious or major side effects include livertoxicity, which is especially well described in the pastamong patients with psoriasis. In patients with RA,mild transient elevation of liver function tests is com-monly observed, but serious liver disease, includingcirrhosis, is extremely rare. Risk factors for seriousliver disease may include other contributors to liverdysfunction, such as alcohol, diabetes mellitus, non-alcoholic fatty liver disease, and other medications.Briefly, complete blood count, renal function, liverfunction, and serum albumin are monitored every 1 to

2 months during therapy with methotrexate (Table 7).A significant fall in serum albumin or persistent eleva-tion of liver function tests may warrant a liver biopsyif therapy is to be continued. Dosages must be low-ered for patients with renal insufficiency.

Bone marrow suppression manifested as pancytope-nia is an infrequent but clinically significant adverseevent in patients with RA. Macrocytosis may be aharbinger of this toxicity, and concomitant renal insuf-ficiency resulting in increased levels of and prolongedexposure to methotrexate may predispose to marrowtoxicity. Folinic acid (leucovorin rescue) effectivelyreverses this.

An acute hypersensitivity to pneumonitis has beenobserved in approximately 1%-2% of patients and attimes may be severe or even fatal. Clinical character-istics include dry cough, fever, dyspnea, and severehypoxemia. Radiographs show diffuse interstitialprominence or frank infiltrates and occasionaladenopathy. Because methotrexate may predisposepatients to opportunistic organisms, infection must beruled out and the drug stopped. Patients usuallyrespond to symptomatic support and drug withdrawal.

Finally, methotrexate is a known teratogen, and thecharacteristic fetal abnormalities in babies of mothersexposed to this drug have been termed “aminopterinsyndrome.” The children exhibit craniofacial dys-plasias, mental retardation, and cardiovascular anoma-lies. Methotrexate is also a potent abortifacient; incombination with misoprostol, it was shown to be aneffective abortion-inducing therapy. Thus, methotrex-ate should never be given to pregnant or lactatingmothers. Although there exists little consensus regard-ing an algorithm to discontinue methotrexate if a preg-nancy is desired, conventional wisdom suggests thatthe drug should be discontinued in women for at least3 months (to include at least two ovulatory cycles)before conception. Methotrexate in doses used for RAprobably does not have significant effect on fertility orsterility, although this is poorly studied.

Cyclosporine

Mechanism of action. Cyclosporine, a metabolicproduct of fungal organisms, has immunologicactivity affecting calcium-associated signalingevents in the regulation of cytokine gene expres-sion. Cyclosporine inhibits T-cell interaction withmacrophages, reduces IL-2 synthesis, prevents pro-duction of IL-1 receptor, and suppresses synthesisand release of IL-2 by CD4+ cells. Recent studiesalso indicate that cyclosporine increasesinsulin–like growth factor and bone GLA (gamma-carboxyglutamic acid) protein and may stimulatethe androgen axis. Growth of bone marrow-derivedB-lymphocytes, myeloid, and erythroid cell linesare not inhibited by cyclosporine.

Pharmacology. Cyclosporine is lipophilic and maybe given orally or intravenously. Absorption of oraldoses of cyclosporine is variable. It is bound toplasma proteins, lipoproteins, and erythrocytes. Vol-ume of distribution includes tissues that accumulatehigher concentrations than serum, such as body fat,liver, spleen, pancreas, kidneys, lungs, and lymphnodes. The drug is metabolized in the liver to multi-ple metabolites (some of which have immunosup-pressive activity) by cytochrome P450 mixed func-tion oxidase. Drug interactions may occur, and grape-fruit juice may increase cyclosporin A serum levels.Ketoconazole and diltiazem raise cyclosporine lev-els, whereas phenytoin, phenobarbital, and rifampinmay lower drug levels. The drug must be carefullyadministered in patients with renal dysfunction, andcombination with other potentially nephrotoxic med-ications such as NSAIDs are discouraged or donewith considerable caution and monitoring.

Clinical use. Use of cyclosporin A in rheumatologicconditions has thus far included RA, Sjögren’s dis-ease, Behçet’s-related inflammatory eye disease,SLE, scleroderma, polymyositis, and systemic vas-culitis. Dosages of <5 mg/kg daily reduce potential forside effects. A number of combination therapy studieshave been reported, among them a study comparingthe use of cyclosporin A to placebo in RA patientswho were inadequately controlled with low-doseweekly oral methotrexate. This study showed a signif-icant drug effect compared to placebo, although con-cerns about this continue. However, because of the

potential for adverse reactions (see below) even at thelower doses, cyclosporin A use is decreasing as theanti-TNF inhibitors penetrate routine practice.

Adverse effects. Toxic effects of cyclosporin Ainclude: nausea; anorexia; flushing; headaches;tremors; sleep disturbance; hypertrichosis; gingivalhyperplasia; hypertension; and hyperuricemia. Renalinsufficiency may result from reversible acute toxic-ity likely to be due to vasoconstriction or fromchronic nephropathy with findings of interstitialfibrosis, tubular atrophy, and vasculopathy. Baselinerenal insufficiency, age above 30 years, and higherdosing place the patient at increased risk for theseside effects. Potassium-sparing diuretics should beavoided, because cyclosporine may cause hyper-kalemia. Frequent monitoring of blood pressure,blood chemistries, and leukocyte counts, as well asurinalysis should be done. Because the potential formalignancy exists in patients treated withcyclosporin A, appropriate long-term monitoring ofthese patients is important.

Leflunomide

Leflunomide was initially developed to prevent trans-plant rejection; it was developed clinically for RApatients as an alternative to methotrexate and sul-fasalazine. It was approved by regulatory authoritiesfor use in patients with active RA.

Mechanism of action. The active metabolite ofleflunomide inhibits dihydroorotate dehydrogenase,thereby interfering with pyrimidine synthesis. Thedown-stream is inhibition of activated lymphocytes(T-cells and B-cells) and other cells that participatein the inflammatory process. In animal models, itinhibits the expression of adjuvant and collagen-induced arthritis and allograft rejection. It inhibitsT-cell proliferation and B-cell antibody formation.Leflunomide also inhibits tyrosine kinases,although the in vivo clinical effect of this action isnot completely elucidated.

Pharmacology. Leflunomide is orally administered,shows good bioavailability, and has a long plasmahalf-life of roughly 2 weeks. It is highly protein bound,metabolized in the liver, exhibits significant enterohep-atic circulation, and is excreted by the kidney and gut.



Clinical use. In 3 12-month phase III controlled clini-cal trials involving 1839 patients, leflunomide wasfound to be significantly superior to placebo, andequivalent to methotrexate and sulfasalazine in reduc-ing the signs and symptoms of active RA. The onset ofaction of leflunomide appeared to be earlier than withthe other active comparator treatments. Importantly,in these same trials, leflunomide retarded the progres-sion of radiographic erosions, but this structural modi-fication effect was similar in magnitude to thatobserved in both methotrexate and sulfasalazine treat-ment groups. It has also been shown to improve qual-ity of life.

In practice, the use of leflunomide increased after itsintroduction, particularly among patients who failedmethotrexate or who were not considered candidatesfor the anticytokine treatments. A loading dose of 100mg po qd for 3 days was recommended in the past, butits use has decreased as a means to minimize sideeffects. Some people begin with the maintenance doseof 20 mg po qd. Dosage reduction to 10 mg po qd isoccasionally a useful strategy when adverse effectsare encountered. Individual practitioners are nowexperimenting with leflunomide in combination withother second-line agents.

Adverse effects. Side effects seen with leflunomideinclude reversible alopecia, skin rash, diarrhea, andliver transaminase elevations. The pivotal trials ofleflunomide for RA showed that approximately 15%of patients had mild elevations of liver transami-nases (1.2-2.0 x upper limit of normal). Severe ele-vations (ie, >3 times the upper limit of normal) wereonly seen in approximately 1%-4 % of patients inthese trials. Many of the hepatic adverse eventsoccurred early in the treatment phase (usually withinthe first 6 months). Many of the patients who hadsevere hepatic reactions to leflunomide were takingother medications that may have been hepatotoxic orhad significant comorbid diseases that might havecontributed to the outcome.

The general recommendations for monitoring liverfunction in patients taking leflunomide are similar tothe guidelines developed for monitoring RA patientstaking methotrexate therapy. It is now recommendedthat ALT and AST be measured prior to drug initia-tion, monthly for 6 months, and finally, every 1 to 3months thereafter if the drug is well tolerated. More

frequent monitoring should be done if the patientshows elevations of liver transaminases. If elevationsare observed, the dose should be reduced or stoppedand liver function tests should be repeated until theyare normal. Persistently abnormal or any severe eleva-tion of liver transaminases (ie, >3 x upper level of nor-mal) should prompt withdrawal of the drug. In theearly trials, infections were infrequent and wereequally prevalent across methotrexate and sul-fasalazine treatment groups. Leflunomide is terato-genic and contraindicated for pregnant women; thus,it should be used with great caution in women ofchildbearing potential.

Tumor Necrosis Factor Inhibitors: Etanercept, Infliximab and Adalimumab

In many animal models of arthritis, and in in vitro and exvivo experiments assessing human rheumatoid tissue,TNF-� has been proven to be of profound importance indriving inflammation in the RA synovium. Transgenicmice transected with the human TNF-� gene developchronic arthritis, and the treatment of these animals witheither the TNF-� receptor fusion proteins or the mono-clonal antibody to TNF-� have abrogated the disease invivo. TNF-� induces other pro-inflammatory cytokines,(including IL-1 and IL-6), stimulates production ofmatrix metalloproteinases (eg, collagenase), andincreases expression of adhesion molecules.

EtanerceptEtanercept is a dimeric fusion protein of the extracel-lular p75 soluble TNF-� receptor linked to IgG1. Itspecifically binds to TNF-� and prevents its interac-tion with the TNF-� receptor.

Pharmacology. It is administered subcutaneously,either 25 mg twice per week or 50 mg once a week.After a single 25 mg injection, the median half-life is115 hours.

Clinical use. Etanercept has shown profound effi-cacy in a number of double-blind randomized con-trolled trials. In the first Phase III trial of 234 patientswith active RA, an impressive reduction of clinicalsigns and symptoms was observed in the etanerceptgroup compared to placebo. In another Phase III studyof 6 months duration, RA patients who were inade-quately controlled with methotrexate were random-ized to either etanercept or placebo. The addition of

etanercept markedly improved the clinical status ofRA patients compared to those who received placebo.

Adverse effects. As with all TNF inhibitors, therethere is concern regarding the potential for immunesuppression and increased risk of infection. In clinicaltrials, while overall infections were observed morecommonly, there was no increase in the frequency ofsevere infections. As with all TNF inhibitors, thedevelopment of autoantibodies, in particular ANAand anti-dsDNA, was observed in higher frequency inetanercept-treated patients compared with placebo.However, development of clinical lupus is rare.

Infliximab Infliximab is a chimeric IgG1 monoclonal antibody toTNF-�, thereby inhibiting its biologic effects in thesynovium.

Pharmacology. It is administered intravenously, andafter a single intravenous administration of 3 mg/kg,the half-life is 8 to 9.5 days. Infliximab is approved foruse, in conjunction with concomitant methotrexate, atdoses of 3 to 10 mg/kg given every 4 to 8 weeks.

Clinical use. Infliximab was initially approved byFDA for treating patients with active Crohn’s disease.The efficacy of infliximab in patients with active RAdespite concomitant therapy with methotrexate wasevaluated in several randomized placebo-controlledtrials. Infliximab at 3 mg/kg was profoundly superiorto placebo in reducing the signs and symptoms of RA.The usual dose regimen is 3 mg/kg by intravenousinfusion at 0, 2, and 6 weeks, followed by 3 mg/kgintravenously every 8 weeks. Infliximab should begiven to RA patients taking methotrexate, and dosageadjustment of up to 10 mg/kg can be made if there isinsufficient clinical response.

Adverse effects. As with all TNF inhibitors, there is aconcern regarding the potential for immune suppres-sion and increased risk of infection. In clinical trials,while overall infections were observed more com-monly, there was no increase in the frequency ofsevere infections. As with all TNF inhibitors, thedevelopment of autoantibodies, in particular ANAand anti-dsDNA, was observed in higher frequency inetanercept-treated patients compared with placebo.Development of clinical lupus is rare.

AdalimumabAdalimumab is a human IgG1 monoclonal antibodyto TNF-� that specifically binds to circulating andcell-bound TNF, thereby inhibiting its biologiceffects.

Pharmacology. Adalimumab is given subcutaneouslyand has a half-life of approximately 13 days. It is givenat dose of 40 mg every other week, with the potential toincrease weekly doses.

Clinical use. The efficacy of adalimumab in patientswith active RA has been shown in several randomizedplacebo controlled trials, both in conjunction withmethotrexate and as monotherapy.

Adverse effects. As with all TNF inhibitors, there is aconcern regarding the potential for immune suppres-sion and increased risk of infection. In clinical trials,while overall infections were observed more com-monly, there was no increase in the frequency of severeinfections. As with all TNF inhibitors, the develop-ment of autoantibodies, in particular ANA and anti-dsDNA, was observed in higher frequency in etaner-cept-treated patients compared with placebo. Develop-ment of clinic lupus is rare.

Interleukin-1 Receptor Antagonist Anakinra

Interleukin-1 (IL-1) is a pro-inflammatory cytokinethat contributes to synovial inflammation, in part bythe induction of other inflammatory cytokines and pro-tein-degrading enzymes such as collagenase. Anakinra(IL-1ra) is a homologue of the naturally occurring IL-1receptor antagonist (IL-Rra) that competes with IL-1for binding to type 1 IL-1 receptors.

Pharmacology. Anakinra, which is given typically asa daily subcutaneous injection, has a half-life ofapproximately 4-6 hours.

Clinical uses. Anakinra has been shown to be effec-tive in several double blind placebo controlled trials. Ingeneral, the extent of responses observed withanakinra is less than that seen with TNF inhibitors.

Adverse events. The overall safety profile of thisagent seems to be acceptable with the most frequentadverse events being injection site reactions. Infec-



tions are always a concern because of the active pro-inflammatory nature of IL-1; however, significantincrease in frequencies of opportunistic infectionswere not seen in the clinical studies. In a study inwhich anakinra was used in combination with theTNF inhibitor etanercept, a significantly greater num-ber of serious infections developed; therefore, combi-nation therapy is not recommended.

T-cell Costimulation Inhibitor Abatacept

Abatacept, previously known as CTLA-4-Ig, is afusion protein consisting of the extracellular portionof the T-cell molecule CTLA-4 linked with the Fcpiece of human IgG1 that has been modified so as tonot activate complement. CTLA-4 is a naturalinhibitor of the interaction between CD28, a keystimulatory molecule on T-cells, and its ligandsCD80 and CD86 on antigen presenting cells.

Pharmacology. Abatacept is given as an intravenousinfusion over 30 minutes once a month, after an initi-ation phase of treatment at weeks 0, 2 and 4. The doseis weight based, approximating 10 mg/kg.

Clinical use. The efficacy of abatacept has beenshown in RA patients who had active disease despiteconcomitant MTX or who have failed previous treat-ment with TNF inhibitors.

Adverse effects. As with all immunomodulatoryagents, patients treated with abatacept should beobserved for signs and symptoms of infection,although in clinical trials these occurred only slightlymore commonly than in the comparison arms. In astudy in which abatacept was used in combinationwith the TNF inhibitor etanercept, a significantlygreater number of serious infections developed;therefore, combination therapy is not recommended.

��-cell Inhibitor Rituximab

Rituximab is a chimeric monoclonal antibodydirected against CD20, a molecule whose expressionis limited to �-cells. Rituximab was approved as atreatment for non-Hodgkin’s lymphoma in 1997.Because of the role of �-cells in various autoimmunediseases, it has been under study in several rheumaticdiseases.

Pharmacology. In RA, rituximab is typically givenas 2 intravenous infusions of 1,000 mg, given 2weeks apart over 3-4 hours. Preinfusion corticos-teroids (eg, 100 mg methylprednisolone) are com-monly used and seem to increase the tolerability ofthe first dose particularly. Following treatment withrituximab, there is a profound and long-lasting deple-tion of circulating �-cells. Because not all patientsrespond, this cannot be used as a marker of efficacy.

Clinical use. The efficacy of rituximab has beenshown in several studies in RA. Its initial approval inRA was for patients who had failed previous treat-ment with TNF inhibitors.

Adverse effects. As with all immunomodulatoryagents, patients treated with rituximab should beobserved for signs and symptoms of infection,although in clinical trials these occurred only slightlymore commonly than in the comparison arms. Whileinfusions of rituximab appear to be better tolerated inRA as compared to lymphoma, infusion related sideeffects can occur and can be severe.

7. References

Pathophysiology

1. Stastny P. Association of the B-cell alloantigenDRw4 with rheumatoid arthritis. N Engl J Med.1978;298:869-871.

2. Weyand CM, Hicock KC, Conn DL, Goronzy JJ.The influence of HLA-DRB1 genes on diseaseseverity in rheumatoid arthritis. Ann Intern Med.1992;117:801-886.

3. Arend W, Dayer J. Inhibition of the productionand effects of Interleukin-1 and tumor necrosisfactor in rheumatoid arthritis. Arthritis Rheum.1995;38:2,151-160.

4. Mojcik C, Shevach E. Adhesion molecules. Arthritis Rheum. 1997;40:6,991-1004.

5. Gregersen PK, Silver J, Winchester RJ. The sharedepitope hypothesis: an approach to understandingthe molecular genetics of susceptibility to rheuma-toid arthritis. Arthritis Rheum. 1987;30:1205-1213.

6. Nishioka K, Hasunuma T, Kato T, et al. Apoptosisin rheumatoid arthritis, A novel pathway in the regulation of synovial tissue. Arthritis Rheum.1998;41:1,1-9.

7. Koch A. Angiogenesis. Arthritis Rheum. 1998;41:6,951-962.

8. Paleolog E, Young S, Stark A, et al. Modulation of angiogenic vascular endothelial growth factorby tumor necrosis factor and Interleukin-1 in rheumatoid arthritis. Arthritis Rheum. 1998;41:7,1258-1265.

9. Weyand CM, McCarthy TG, Goronzy JJ. Correla-tion between disease phenotype and genetic het-erogeneity in rheumatoid arthritis. J Clin Invest.1995;95:2120-2126.

10. Wagner UG, Koetz K, Weyand CM, Goronzy JJ.Perturbation of the T-cell repertoire in rheumatoidarthritis. Proc Natl Acad Sci USA. 1998;95:14447-14452.

11. Carson DA, Chen PP, Kipps TJ. New roles forrheumatoid factor. J Clin Invest. 1991;87:379-383.

12. Kong YY, Feige U, Sarosi I, et al. Activated T-cells regulate bone loss and joint destruction inadjuvant arthritis through osteoprotegerin ligand.Nature. 1999;402:304-309.

13. Choy E, Panayi G. Cytokine pathways and jointinflammation in rheumatoid arthritis. N Engl JMed. 2001;344:907-916.

Epidemiology

1. Wolfe F, Mitchell D, Sibley J, et al. The mortalityof rheumatoid arthritis. Arthritis Rheum.1994;37:4, 481-494.

2. Yelin E, Wanke L. An assessment of the annualand long-term direct costs of rheumatoid arthritis.The impact of poor function and functionaldecline. Arthritis Rheum. 1999;42:6,1209-1218.

3. Gabrial S, Crowson C, O’Fallon W. The epidemi-ology of rheumatoid arthritis in Rochester, Min-nesota. Arthritis Rheum. 1999;42:3,415-420.

4. Kaipiainen-Seppä T, Aho K. Incidence of chronicinflammatory joint diseases in Finland in 1995. J Rheumatol. 2000;27:94-100.

5. Wållberg-Jonsson S, Johansson H, Ohman ML, etal. Extent of inflammation predicts cardiovasculardisease and overall mortality in seropositiverheumatoid arthritis: a retrospective cohort studyfrom disease onset. J Rheumatol. 1999;26:2562-2571.

Clinical Features

1. Tuesson C, Jacobsson L, Bergstrom U. Extra-articular rheumatoid arthritis: prevalence and mor-tality. Rheumatology. 1999;38:668-674.

2. McQueen FM. Magnetic resonance imaging inearly inflammatory arthritis: what is its role?Rheumatology. 2000;39:700-706.



3. Gridley G, McLaughlin JK, Ekbom A, et al. Inci-dence of cancer among patients with rheumatoidarthritis. J Natl Cancer Inst. 1993;85:307-11.

4. Boden SD. Rheumatoid arthritis of the cervicalspine. Surgical decision making based on predic-tors of paralysis and recovery. Spine. 1994;19:2275-80.

5. Alarcón G. Predictive factors in rheumatoid arthri-tis. Am J Med. 1997;103 (Suppl. 6A):319.

6. Furst DE. Predictors of worsening clinical vari-ables and outcomes in rheumatoid arthritis.Rheum Dis Clin N Am. 1994;20:309-19.

7. Felson DT, Anderson JJ, Boens M, et al. AmericanCollege of Rheumatology. Preliminary definitionof improvement in rheumatoid arthritis. ArthritisRheum. 1995;36:727-735.

8. Wolfe F, Zwillich S. The long-term outcomes ofrheumatoid arthritis. A 23-year prospective, longi-tudinal study of total joint replacement and its pre-dictors in 1,600 patients with rheumatoid arthritis.Arthritis Rheum. 1998;41:6,1072-1082.

9. Riedemann JP, Munoz S, Kavanaugh A. The use ofsecond generation anti-CCP antibody (anti-CCP2)testing in rheumatoid arthritis—a systematicreview. Clin Exp Rheumatol. 2005;23 (Suppl.39):S69-S76.

Analgesics

1. Bradley JD, Brandt KD, Katz BP, Kalasinski LA,Ryan SI. Comparison of an anti-inflammatorydose of ibuprofen, an analgesic dose of ibuprofen,and acetaminophen in the treatment of patientswith osteoarthritis of the knee. N Engl J Med.1991;325:87-91.

2. Perneger TV, Whelton PK, Klag MJ. Risk of kid-ney failure associated with the use ofacetaminophen, aspirin, and nonsteroidal anti-inflammatory drugs. N Engl J Med. 1994;331:1675-9.

Nonsteroidal Anti-inflammatory Drugs

1. Brooks PM, Day RO. Nonsteroidal anti-inflamma-tory drugs: differences and similarities. N Engl JMed. 1991;324:1716-25.

2. Hoppman RA, Reden JG, Ober SK. Central ner-vous system side effects of nonsteroidal anti-inflammatory drugs. Ann Intern Med. 1991;115:787-96.

3. Graham DY, White RH, Moreland LW, SchubertTT, Katz R, Jaczewski R, et al. Duodenal and gas-tric ulcer prevention with misoprostol in arthritispatients taking NSAIDs. Ann Intern Med.1993;119:257-62.

4. Schlondorff D. Renal complications of nons-teroidal anti-inflammatory drugs. (Clinical Con-ference) Kidney Int. 1993;44:643-53.

5. Silverstein FE, Faich G, Goldstein JL, et al. Gas-trointestinal toxicity with celecoxib vs. non-steroidal anti-inflammatory drugs for osteoarthritisand rheumatoid arthritis: the CLASS study: a ran-domized controlled trial. Celecoxib long-termarthritis safety study. JAMA. 2000;284:1247-1255.

6. Furst DE. Are there differences among nons-teroidal anti-inflammatory drugs? Comparingacetylated salicylates, nonacetylated salicylates,and nonacetylated nonsteroidal anti-inflammatorydrugs. Arthritis Rheum. 1994;37:1-9.

7. Johnson AG, Nguyen TV, Day RO. Do nons-teroidal anti-inflammatory drugs affect blood pressure? A meta-analysis. Ann Intern Med.1994;121:289-300.

8. Kim JG, Graham DY. Helicobacter pylori infec-tion and development of gastric or duodenal ulcerin arthritis patients receiving chronic NSAID ther-apy. The Misoprostol Study Group. Am J Gastroenterol. 1994;89:203-207.

9. Mitchell JA, Akarasereenont P, Thiemermann C,Flower RJ, Vane JR. Selectivity of nonsteroidalanti-inflammatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc Natl Acad Sci USA. 1994;90:11693-11697.

10. Bombardier C, Peloso PM, Goldsmith CH. Salsalate-Diclofenac Study Group. Salsalate, anonacetylated salicylate, is as efficacious asdiclofenac in patients with rheumatoid arthritis. J Rheumatol. 1995;22:617-624.

11. Diaz-Gonzalez F, Gonzalez-Alvero I, CompaneroMR. Prevention of in vitro neutrophil-endothelialattachment through shedding of L-selectin by non-steroidal anti-inflammatory drugs. J Clin Invest.1995;1756-65.

12. Fries JF. ARAMIS and toxicity measurement. J Rheumatol. 1995;22:995-7.

13. Bombardier C, Laine L, Reicin A, et al. Compari-son of upper gastrointestinal toxicity of rofecoxiband naproxen in patients with rheumatoid arthritis.N Engl J Med. 2000;343:1520-1528.

14. Lichtenstein DR, Syngal S, Wolfe MM. The dou-ble-edged sword: nonsteroidal anti-inflammatorydrugs and the gastrointestinal tract. ArthritisRheum. 1995;38:5-18.

15. Silverstein FE, Graham DY, Senior JR, et al.Misoprostol reduces serious gastrointestinal com-plications in patients with rheumatoid arthritisreceiving nonsteroidal anti-inflammatory drugs.Ann Intern Med. 1995;123:241-249.

16. Ostensen M, Ostensen H. Safety of the non-steroidal anti-inflammatory drugs in pregnantpatients with rheumatic disease. J Rheumatol.1996;23:1045-1049.

17. Simon L, Weaver A, Graham D, et al. Anti-inflam-matory and upper gastrointestinal effects of cele-coxib in rheumatoid arthritis. JAMA.1999;282:20:1921-1927.

18. Crofford LJ, Oates JC, McCune WJ, et al. Throm-bosis in patients with connective tissue diseasestreated with specific cyclooxygenase 2 inhibitors:a report of four cases. Arthritis Rheum. 2000;43:1891-1896.

Glucocorticoids

1. Kirwan JR. The effect of glucocorticoids on jointdestruction in rheumatoid arthritis. The Arthritisand Rheumatoid Council Low-Dose Glucocorti-coid Study Group. N Engl J Med. 1995;333:142-146.

2. Baxter JD. Minimizing the side effects of gluco-corticoid therapy. Adv Intern Med. 1990;35:173-193.

3. Boumpas DT, Chrousos GP, Wilder RL, CuppsTR, Balow JE. Glucocorticoid therapy forimmune-mediated diseases: basic and clinical cor-relates. Ann Intern Med. 1993;119:1198-208.

4. Saag KG, Koehnke R, Caldwell JR Brashington R.Burmeister LF, Zimmerman B, Kohler JA, FurstDE. Low-dose, long-term corticosteroid therapyin rheumatoid arthritis: an analysis of seriousadverse events. Am J Med. 1994;96:115-123.

5. van Everdingen AA, Jacobs JWG, van ReesemaDRS, Bijlsma JWJ. Low-dose prednisone therapyfor patients with early active rheumatoid arthritis:clinical efficacy, disease-modifying properties, andside effects. A randomized, double-blind, placebo-controlled clinical trial. Ann Inter Med. 2002;136:1-12.

Antirheumatic Drugs

1. Tugwell P, Pincus T, Yocum D, Stein M, Gluck O,Kraag G, et al. Combination therapy withcyclosporine and methotrexate in severe rheuma-toid arthritis. The Methotrexate-CyclosporineCombination Study Group. N Engl J Med.1995;333:137-141.

2. van der Heide A, Jacobs JW, Bijlsma JW,Heurkens AH, van Booma-Frankfort C. Theeffectiveness of early treatment with second-lineantirheumatic drugs. A randomized, controlledtrial. Ann Intern Med. 1996;124:699-707.

3. O’Dell JR, Haire CE, Erikson N, Drymalski W,Palmer W, Eckhoff PJ. Treatment of rheumatoidarthritis with methotrexate alone, sulfasalazine andhydroxychloroquine, or a combination of all threemedications. N Engl J Med. 1996;334:1287-1291.



4. Fries JF, Williams CA, Ramey D, Bloch DA. The relative toxicity of disease-modifyingantirheumatic drugs. Arthritis Rheum.1993;36:297-306.

5. Cash JM, Klippel JH. Second-line drug therapy for rheumatoid arthritis. N Engl J Med.1994;330:1368-1375.

6. Esdaile JM, the HERA Study Group. A random-ized trial of hydroxychloroquine in early rheuma-toid arthritis. Am J Med. 1995;98:156-168.

7. Anderson RJ. Hydroxychloroquine therapy inrheumatic diseases. Bull Rheum Dis. 1995;44:6-7.

8. Clark P, Casas E, Tugwell P, et al. Hydroxychloro-quine compared with placebo in rheumatoid arthri-tis. A randomized controlled trial. Ann Intern Med.1993;119:1067-1071.

9. Weisman MH, Weinblatt ME. Treatment of theRheumatic Diseases. Philadelphia, Pa: WB Saunders; 1995.

10. Weinblatt M, Kremer J, Coblyn J, et al. Pharma-cokinetics, safety, and efficacy of combinationtreatment with methotrexate and leflunomide inpatients with active rheumatoid arthritis. ArthritisRheum. 1999;42:7,1322-1328.

11. Strand V, Cohen S, Schiff M, et al. Treatment ofactive rheumatoid arthritis with leflunomide com-pared with placebo and methotrexate. Arch InternMed. 1999;159:11,2542-2550.

12. Maxime D, Combe B, Cantagrel A, et al. Combi-nation therapy in early rheumatoid arthritis: a ran-domized, controlled, double blind 52-week clini-cal trial of sulfasalazine and methotrexate com-pared with the single components. Ann RheumDis. 1999;58:220-225.

13. Sokka T, Hannonen P. Utility of disease modifyingantirheumatic drugs in “sawtooth” strategy. Aprospective study of early rheumatoid arthritispatients up to 15 years. Ann Rheum Dis.1999;58:618-622.

14. Breedveld FC, Dayer JM. Leflunomide: mode ofaction in the treatment of rheumatoid arthritis. AnnRheum Dis. 2000;59:841-849.

15. Tugwell P, Wells G, Strand V, et al. Clinicalimprovement as reflected in measures of functionand health-related quality of life following treat-ment with leflunomide compared with methotrex-ate in patients with rheumatoid arthritis: sensitiv-ity and relative efficiency to detect a treatmenteffect in a twelve-month, placebo-controlled trial.Leflunomide Rheumatoid Arthritis InvestigatorsGroup. Arthritis Rheum. 2000;43:506-514.

16. Sharp JT, Strand V, Leung H, et al. Treatment withleflunomide slows radiographic progression ofrheumatoid arthritis: results from three random-ized controlled trials of leflunomide in patientswith active rheumatoid arthritis. LeflunomideRheumatoid Arthritis Investigators Group. Arthri-tis Rheum. 2000; 43:495-505.

17. Tsakonas E, Fitzgerald AA, Fitzcharles MA, et al.Consequences of delayed therapy with second-line agents in rheumatoid arthritis: a 3 year followup on the hydroxychloroquine in early rheumatoidarthritis (HERA) study. J Rheumatol. 2000;27:623-629.

Methotrexate

1. Kamel OW, van de Rijn M, Weis LM, et al. Briefreport: reversible lymphomas associated withEpstein-Barr virus occurring during methotrexatetherapy for rheumatoid arthritis and dermato-myositis. N Engl J Med. 1993;328:1317-1321.

2. Kremer JM. The mechanism of action ofmethotrexate in rheumatoid arthritis: the searchcontinues. J Rheumatol. 1994;21:1-5.

3. Walker AM, Funch D, Dreyer NA, Toman KG,Kremer JM, Alarcon GS. Determinants of seriousliver disease among patients receiving low-dosemethotrexate for rheumatoid arthritis. ArthritisRheum. 1993;36:329-335.

4. Kremer JM Alarcon GS, Lightfoot RW, et al.Methotrexate of rheumatoid arthritis: suggestedguidelines for monitoring liver toxicity. ArthritisRheum. 1994;37:316-328.

Anticytokine Therapy

1. Maini R, Breedveld F, Kalden J, et al. Therapeuticefficacy of multiple intravenous infusions of anti-tumor necrosis factor monoclonal antibody com-bined with low-dose weekly methotrexate inrheumatoid arthritis. Arthritis Rheum. 1998;41:9,1552-1563.

2. Moreland L, Schiff M, Baumgartner S, et al. Etan-ercept therapy in rheumatoid arthritis: a random-ized, controlled trial. Ann Intern Med.1999;130:478-486.

3. Weinblatt M, Kremer J, Bankhurst A, et al. A trialof etanercept, a recombinant tumor necrosis factorreceptor: Fc fusion protein, in patients withrheumatoid arthritis receiving methotrexate. N Engl J Med. 1999;340:4:1,453-259.

4. Moreland L, Baumgartner S, Schiff M, et al. Treat-ment of rheumatoid arthritis with a recombinanthuman tumor necrosis factor receptor (p75)-Fcfusion protein. N Engl J Med. 1997;337:3:7,141-147.

5. Maini R, St Clair E, Breedveld F, et al. Infliximab(chimeric anti-tumour necrosis factor monoclonalantibody) versus placebo in rheumatoid arthritispatients receiving concomitant methotrexate: arandomized phase three trial. Lancet. 1999;354,12:1932-1939.

6. Heck L, Koopman W. Biologic agents for treatingrheumatoid arthritis. Concepts and progress.Arthritis Rheum. 1997;40:3:7,397-409.

7. Jones R, Moreland L. Tumor necrosis factorinhibitors for rheumatoid arthritis. Arthritis Foundation. 1999;48:3,1-4.

8. Kremer J. Combination therapy with biologicagents in rheumatoid arthritis: Perils and promise.Arthritis Rheum. 1998;41:9,1548-1551.

9. Lipsky PE, van der Heijde DM, St Clair EW, et al.Infliximab and methotrexate in the treatment ofrheumatoid arthritis. N Engl J Med.2000;343:1594-1602.

10. Bathon JM, Martin RW, Fleischmann RM, et al. Acomparison of etanercept and methotrexate inpatients with early rheumatoid arthritis. N Engl JMed. 2000;343:1586-1593.

11. Jiang Y, Genant HK, Watt I, et al. A multicenter,double-blind, dose-ranging, randomized, placebo-controlled study of recombinant human inter-leukin-1 receptor antagonist in patients withrheumatoid arthritis radiologic-progression andcorrelation of Genant and Larsen scores. ArthritisRheum. 2000;43:1001-1009.

12. Bresnihan B, Alvaro-Gracia JM, Cobby M, et al.Treatment of rheumatoid arthritis with recombi-nant human interleukin-1 receptor antagonist.Arthritis Rheum. 1998;41:2196-2204.

Miscellaneous Therapies

1. Barnett M, Kremer J, St Clair E, et al. Treatment ofrheumatoid arthritis with oral type II collagen.Results of a multicenter, double-blind, placebo-controlled trial. Arthritis Rheum. 1998;41:2,290-297.

2. Snowden J, Kearney P, Kearney A, et al. Long-term outcome of autoimmune disease followingallogenic bone marrow transplantation. ArthritisRheum. 1998;41:3,453-459.

3. McKown K, Carbone L, Kaplan S, et al. Lack ofefficacy of oral bovine type II collagen added toexisting therapy in rheumatoid arthritis. ArthritisRheum. 99:42;6,1204-1208.



4. O’Dell J, Paulsen G, Haire C, et al. Treatment ofearly seropositive rheumatoid arthritis withminocycline. Arthritis Rheum. 1999;42:8,1691-1695.

5. Felson D, LaValley M, Baldassare A, et al. Theprosorba column for treatment of refractoryrheumatoid arthritis. A randomized, double-blind,sham-controlled trial. Arthritis Rheum.1999;42:10,2153-2159.

6. Conn D, Arnold W, Hollister J. Alternative treat-ments and rheumatic diseases. Arthritis Foundation. 1999;48:7:12,1-2.

7. Goekoop-Ruiterman YPM, de Vries-Bouwstra JK,Allaart CF, et al. Clinical and radiographic out-comes of four different treatment strategies inpatients with early rheumatoid arthritis (the BeStStudy). Arthritis Rheum. 2005;52:3381-3390.

8. Genovese MC, Becker J-C, Schiff M, et al. Abata-cept for rheumatoid arthritis refractory to tumornecrosis factor inhibition. N Engl J Med.2005;353:1114-1123.

9. Emery P, Fleischmann R, Filipowicz-SosnowskaA, et al. The efficacy and safety of rituximab inpatients with active rheumatoid arthritis despitemethotrexate treatment: results of a phase IIb dou-ble-blind, placebo-controlled, dose ranging trial(DANCER). Arthritis Rheum. 2006;54:1390-1400.

rheumatoid arthritis: pathogenesis, clinical features, and ...arthritis) and the mhc molecules that...

Documents