Auto Antc y Enf Inflamatoria

8/4/2019 Auto Antc y Enf Inflamatoria

1/15

Clinical lectures

Evaluating inflammatory joint disease:how and when can autoantibodies help?

Olivier Meyer *

Rheumatology Department, Bichat Teaching Hospital, 46, rue Henri Huchard, 75018 Paris, France

Received and accepted 31 July 2003

Abstract

The diagnosis of inflammatory joint disease rests on a constellation of symptoms, signs, laboratory test results and, occasionally,histological findings. Classification criteria have been developed by national learned societies, international panels of experts or, more rarely,an expert working alone. These criteria are intended to provide a common language for therapeutic trials and international publications.Yet,they are often inappropriately used as diagnostic tools for the individual patient. Identification of an early seroimmunologic marker with highsensitivity and specificity for classifying patients with recent-onset joint disease is a daunting challenge.Test performance characteristics suchas sensitivity, specificity, positive and negative predictive values, and the positive or negative likelihood ratio help to assess the diagnosticusefulness of a laboratory test in a specific situation. The difference between the pretest and posttest likelihoods of obtaining a positive ornegative result measures the usefulness, or performance, of a laboratory test in a specific situation according to the prevalence of the disease.A higher positive likelihood ratio indicates a more useful test. In a patient with inflammatory joint disease, the diagnosis can be sought byassaying a limited number of autoantibodies according to a decision tree. Thus, IgM rheumatoid factors (latex test or ELISA) and antibodies

to filaggrin or other citrullinated proteins (antikeratin antibodies by indirect immunofluorescent assay or anticyclic citrullinated peptides byELISA) identify more than 70% of cases of early rheumatoid arthritis with greater than 98% specificity. If these markers are negative, testingfor antinuclear antibodies by indirect immunofluorescent assay on HEp-2 cells identifies 99% of cases of lupus and progressive systemicsclerosis. Confirmation of the diagnosis can be obtained by characterizing the autoantibodies: thus, presence of antidouble-stranded DNA(dsDNA, by the Farr radioimmunoassay, indirect immunofluorescent assay on Crithidia luciliae, or ELISA (IgG)) or of antinucleosomeantibodies (ELISA) indicates lupus, whereas anticentromere, antitopoisomerase I (Scl 70), and antinucleolar antibodies point to progressivesystemic sclerosis. A positive test for antibodies to soluble nuclear antigens of the U1 RNP type suggests mixed connective tissue disease orlupus but may indicate scleroderma. Anti-Sm antibodies are found in fewer than 10% of lupus patients but are highly specific.Anti-SSA (Ro)and anti-SSB (La) suggest lupus or primary Sjgrens syndrome. When tests are negative for ANA, several antibodies to cytoplasmicorganelles are valuable diagnostic tools, such as anti-J01 for polymyositis syndromes and antiribosome antibodies for lupus, although theirsensitivity is modest (2025%). Finally antineutrophil cytoplasmic antibodies (ANCAs) ensure the diagnosis of small-vessel vasculitides,which often involve the lungs and kidneys. Thus, in diffuse Wegeners granulomatosis, ANCAs exhibiting the classic cytoplasmic pattern andcorresponding by ELISA to anti-PR3 are found. In microscopic polyangiitis the ANCAs are peripheral and correspond by ELISA toantimyeloperoxidase antibodies. Tests for other antibodies are less often needed to evaluate inflammatory joint disease.

2003 ditions scientifiques et mdicales Elsevier SAS. All rights reserved.

Keywords: Inflammatory joint disease; Rheumatoid factors;Anticitrullinated protein antibodies;Antinuclear antibodies; Antineutrophil cytoplasmic antibodies

1. Introduction

The diagnosis of inflammatory joint disease rests on aconstellation of signs, symptoms, laboratory test results and,in some cases, histological findings. Classification criteriahave been developed by learned societies, international pan-

els of experts and, more rarely, experts working alone. Thesecriteria sets are intended to serve as a common language fortherapeutic trials and international publications, yet are oftenused as diagnostic tools in individual patients. They areill-suited for this purpose, particularly in patients with veryrecent manifestations, i.e. indeterminate early inflammatory

joint disease.

Identifying an early seroimmunologic marker that is bothsensitive and specific for classifying early inflammatory joint

* Corresponding author.E-mail address: [email protected] (O. Meyer).

Joint Bone Spine 70 (2003) 433447

www.elsevier.com/locate/bonsoi

2003 ditions scientifiques et mdicales Elsevier SAS. All rights reserved.doi:10.1016/j.jbspin.2003.07.004


2/15

disease is a major challenge for immunologists, epidemiolo-gists, and statisticians. Patients urge us to give them a definitediagnosis as fast as possible; they are probably right, sincewe now have potent biotherapies capable of stopping or

preventing bone and joint damage provided they are givenearly enough.However, not all autoantibodies are specific for a given

disease, and there is plenty of room for mistakes. We will usethree examples to show how an overly systematic approachthat ignores the limited specificity of autoantibodies can leadto diagnostic mistakes.

2. Three clinical examples

2.1. Example 1

A 75-year-old woman had been experiencing bilateralsymmetric polyarthralgia in her wrists and hands for severalmonths.Painful swelling of her hands and feet had developed3 weeks earlier, impairing her ability to carry out the activi-ties of daily living. Raynauds phenomenon since adoles-cence was the only other reported symptom. Her erythrocytesedimentation rate (ESR) was 45 mm/h and her blood cellcount was normal apart from a rather low leukocyte count, of3800 per mm3. Tests were positive for rheumatoid factors(RFs; latex, 320 IU/l; and WaalerRose, 140 IU/l).

A diagnosis of rheumatoid arthritis (RA) was given andmethotrexate started in a dosage of 10 mg/week, in combina-tion with prednisolone 7 mg/d. The development of skinlesions and vascular purpura indicating vasculitis prompteddiscontinuation of the methotrexate 3 months later. The ab-sence of joint erosions, evidence of sicca syndrome, andpositive tests for antinuclear antibodies (ANAs) in a titer of1:1000 with a speckled pattern, anti-SSA/Ro, and anti-SSB/La indicated that the correct diagnosis was primarySjgrens syndrome. Consequently, no further attempts atsecond-line drug therapy were made, and the prednisolonedosage was increased.

2.2. Example 2

This 60-year-old man had an unremarkable medical his-tory apart from recurrent ear, nose, and throat infectionsascribed to chronic sinusitis. He presented with a 3-monthhistory of bilateral polyarthritis involving the wrists, ankles,and knees. Nonsteroidal antiinflammatory drugs (NSAIDs)provided little relief. A trial of prednisolone treatment in adosage of 40 mg/d ensured resolution of the joint symptoms,which recurred a few weeks after discontinuation of the drug.

Physical findings included weight loss of 3 kg and low-grade fever (37.8 C) with no patent focus of infection apartfrom mild clear rhinorrhea tinged with blood. He reportedasthenia. Subcutaneous nodules, some covered with scabs,were felt over the elbows and ulnar ridges.

The ESR was 80 mm/h and the leukocyte count was11 000 per mm3. The latex and WaalerRose tests were both

positive in a titer of 160 IU/l. Tests were negative for ANAand antifilaggrin antibodies. The total complement level wasincreased to 160%.

The patient was given a diagnosis of RF-positive RA and

was started on sulfasalazine (2 g/d) and prednisolone(10 mg/d). The otorhinolaryngological manifestations wors-ened, and a chest radiograph showed nodular densities inboth lungs, including one with cavitation. Investigations fortuberculosis were negative. Wegeners granulomatosis wasconfirmed by a skin nodule biopsy at the elbow and by a testshowing cytoplasmic antineutrophil cytoplasmic antibodies(C-ANCAs) in a high titer (1:1000) shown by ELISA to bedirected against proteinase 3. Cyclophosphamide therapywas started and the prednisolone dosage was increased to1 mg/kg per d.

2.3. Example 3

A 59-year-old man presented with diffuse joint pain,swelling of the wrists, and moderately severe myalgia. Thesesymptoms had started 34 months earlier, showing littleresponse to NSAID therapy. He had skin lesions over thefingers predominating along the lateral edges and ascribed toatopic dermatitis or psoriasis. He was not a manual workerand had no manual activities in his leisure time. Musclestrength testing was normal, although he reported fatigabilityupon exertion with an impression that his performance wasdeclining. He ascribed these symptoms to weight gain of 6 kgafter he stopped playing tennis, which he had practiced at a

fairly high level.His ESR was 35 mm/h. Tests were negative for RF and

ANA. The creatine kinase level was three times the upperlimit of normal. The clinical symptoms and creatine kinaselevel were unchanged 6 weeks after discontinuation of thestatin he had been taking for 2 years to treat dyslipidemia. Adiagnosis of possible psoriatic arthritis was made, and pred-nisolone was started in a dosage of 10 mg/d.

A chest radiograph disclosed interstitial involvement ofboth lung bases, prompting lung function testing, whichshowed a restrictive pattern (with a 30% loss) and a diffusioncapacity of the lung for carbon monoxide of less than 40%.

Antisynthetase syndrome was considered and confirmed by amuscle biopsy and a positive ELISA for anti-J01. The pred-nisolone dosage was increased to 1 mg/kg per d.

These three examples of the challenges raised by earlypolyarthritis show how tests for autoantibodies can help(examples 2 and 3) to identify the disease or, on the contrary,(examples 1 and 2), cause the diagnosis to err when theresults are interpreted too hastily. The possibility of a sero-logical latency phase should be borne in mind, particularlyfor systemic lupus erythematosus (SLE). Many pitfalls canarise with serological testing: for instance, the absence ofANA does not exclude the presence of other autoantibodiesuseful to the diagnosis, such as antiribosome, anti-SSA, orantiphospholipid antibodies in SLE or anti-J01 in polymyo-sitis. Knowledge of test characteristics such as sensitivity,

434 O. Meyer / Joint Bone Spine 70 (2003) 433447


3/15

specificity, positive and negative predictive values, and thelikelihood ratio (positive or negative) helps to assess thediagnostic usefulness of the test in a specific situation [1].The difference between the pretest and posttest likelihood ofa result (whether positive or negative) measures the useful-ness and performance of a laboratory test in a specific situa-tion, according to the prevalence of the disease (Table 1).

A higher positive likelihood ratio and a negative likeli-hood ratio near zero indicate an extremely useful test. Posi-tive likelihood ratios greater than 10 are excellent and greaterthan 5 very good, but clinicians must often make do withvalues between 2 and 5.

The likelihood that a test for a given autoantibody will bepositive in a given joint disease is lower early in the coursethan after a few years. In addition, the prevalence of positivetests for autoantibodies differs between patient cohorts re-cruited in hospitals and those recruited in the community farfrom specialized centers. Therefore, the results of metaanaly-ses that pool series of heterogeneous patients should beviewed with considerable skepticism.

3. Strategy for diagnosing early rheumatoid arthritis

[25]

In large prospective long-term studies of patients withearly inflammatory joint disease at baseline, 3565% of thepatients were classified as having RA after 12 months basedeither on 1987 American College of Rheumatology (ACR)criteria or on evaluation by a panel of experts [3,69]. Thisvariability in rates of RA can be ascribed in part to differ-ences in inclusion criteria, as some studies required at leastone and others at least two affected joints. Therefore, thepretest likelihood of RA ranges from 0.35 to 0.65. In thissituation, an immunological test with a likelihood ratio of2 would increase the posttest likelihood from 0.35 to 0.55 or0.65 to 0.80, which is useful. A likelihood ratio of 5 would beeven more helpful, increasing the posttest likelihood fromabout 0.35 to 0.78 or 0.65 to 0.92 [1].

Although the 1987 ACR criteria are often used for thediagnosis of RA, they perform poorly in the first weeks ormonths of the disease. In a study by Saraux et al. [3,10],among 270 patients with early inflammatory joint disease at

baseline, 98 (36%) were classified by a panel of experts ashaving RA 2 years later (although only 89 met 1987 ACRcriteria); at baseline, 33 (33.6%) of these 98 patients did nothave the four 1987 ACA criteria needed for RA classifica-tion. In addition, because no exclusion criteria were used inthis study, 31 (18%) of the 172 patients without RA after2 years met the 1987 ACA criteria for RA at baseline and 47(45%) at follow-up completion. These data indicate a press-ing need for defining other diagnostic criteria that are presentearly in the course of the disease and exhibit higher specific-ity for RA. Identical conclusions arise from the Norfolkregistry study [9] in a population-based cohort of 486 pa-

tients with early joint disease: 38% met the four 1987 ARAcriteria needed for RA classification yet 3 years later 76%still had synovitis and 40% had erosions.Among the patientsfollowed up for 3 years, 50% were classified by an expert ashaving RA. When used to predict progression to RA inpatients with early polyarthritis, the 1987 ACA criteria per-form poorly: sensitivity, 62%; specificity, 50%; positive pre-dictive value, 55%; negative predictive value, 57%; and posi-tive likelihood ratio, 1.2.

The other data from the literature, which have been re-viewed by Berthelot et al. for the CRI [4], show that meanvalues for sensitivity and specificity are 74% and 83%, re-

spectively. Thus, about one fourth of patients with early RAdo not meet 1987 ACR criteria for RA.

Autoantibodies, most notably those against citrullinatedfilaggrin, are good candidates as early criteria for RA. How-ever, their sensitivity is lower in the first months of thedisease (Table 2).

3.1. Rheumatoid factors [6,7,1116]

IgM RFs are found in 70% of patients with RA of at least1 year duration but may be present in only 5060% ofpatients with early disease. Positive RF tests often antedate

the first clinical manifestations (45% of patients in a Scandi-

Table 12 2 contingency table and main informative indices

Test DiseasePresent Absent

Positive True positives (A) False positives (B)Negative False negatives (C) True negatives (D)

Sensitivity = true positives/(true positives + false negatives) = A/(A+C).Specificity = true negatives/(false positives+ true negatives) = D/(B+D).Positive predictive value = true positives/(true positives + false negati-

ves) = A/(A+B) = (sensitivity prevalence)/[sensitivity prevalence + (1 specificity) (1 prevalence)].

Negative predictive value = true negatives/(false negatives + true nega-tive) = D/(C + D) = [specificity (1 prevalence)]/[(1 sensitivity) prevalence + specificity (1 prevalence)]

Positivelikelihood ratio= sensitivity/(1 specificity) = [A/(A+ C)]/[B/(B+ D)].

Negative likelihood ratio = (1 sensitivity)/specificity = [C/(A +C)]/[D/(B + D)].

Table 2Mean sensitivity (%) of serological markers for rheumatoid arthritis compa-rison between early- and long-standing rheumatoid arthritis

Early RA Long-standing RAAgglutinating RFs 40 70AKAs 26 55APF 50 70AFA, ELISA 50 54AFAs, WB 30 60Citrullinated peptides, ELISA 50 75Anti-RA 33 25 30Anti-Sa 20 45

RA, rheumatoid arthritis; RFs, rheumatoid factors; AKAs, antikeratinantibodies; AFAs, antifilaggrin antibodies; WB, western blot.

435O. Meyer / Joint Bone Spine 70 (2003) 433447


4/15

navian series), sometimes by several years. IgM anti-IgGRFs have limited specificity, as they are often found in otherconnective tissue diseases, liver diseases, lymphoprolifera-tive disorders, and infections (Table 3). However, amongpatients with musculoskeletal diseases, RFs have 97% speci-

ficity with a positive predictive value of 80% assuming a 16%prevalence of RA in the patient population [17]. The intro-duction of solid-phase immunoenzymetric assays (ELISA)has provided the ability to detect, not only IgMs, but alsoIgAs and IgGs with RF activity. RFs of any isotype may bepresent in about 80% of patients with advanced RA but only50% of those with early RA [16,18,19]. In some studies, theconcomitant presence of IgM RF and IgA RF was highlyspecific (specificity, 99%; positive predictive value, 9496%)and reasonably sensitive (50%) for RA [18,20,21]. In anotherstudy, however, detection of IgA RF did not add to theinformation provided by IgM RF detection [3].

3.2. Antiperinuclear factors (APF) [11,2227]

These autoantibodies are usually IgGs directed againstspherical keratohyaline granules 0.54 m in diameter foundin the cytoplasm of human buccal mucosa cells. An obstacleto widespread use of APF detection as a diagnostic test is theneed to obtain human buccal mucosa cells from gooddonors, on an extemporaneous basis. Sensitivities reported inthe literature range from 20% to 91%. The frequency of APFin patients with RF-negative RA has varied across seriesfrom 4% to 52%. In small series of patients with early RA

and negative tests for RFs, the prevalence of APF was1735% [12,28]. Specificity ranged from 73% to 99%, thisvariability being ascribable to the considerable heterogeneityof the patient populations. APF is also found in other rheu-matic diseases including primary Sjgrens syndrome (20%)and, less often, SLE (15%) and psoriatic arthritis with pe-ripheral polyarthritis (13%) [23]. APF is produced very earlyand, consequently, can serve as an early marker for RA [19].Similar to RFs, APF is detectable before the onset of clinicalmanifestations in 20% of patients with RA [24].

3.3. Antikeratin antibodies (AKAs) [11,12,14,2326,2931]

Antibodies to keratin, or rather to the stratum corneum,are IgG autoantibodies against a filamentous protein located

in the superficial layers of the keratinized epidermis. Untilnow, AKAs were detected by indirect immunofluorescentassay using rat distal esophagus sections as the substrate.AKAs have been found in 3657% of patients with RA and

IgM anti-IgG RF and in 640% of those without RFs by thelatex and WaalerRose tests. AKAs are highly specific (95100%) for RA in adults. We found positive tests for AKAs in40% of our patients within the first year after symptom onset.Lower proportions, ranging from 12% to 24%, have beenreported in patients with early RA and negative tests for RFs[7]. AKAs may be present during the preclinical phase in20% of RA patients [24].

3.4. Antibodies to filaggrin and to cyclic citrullinated

peptides (CCP) [6,7,3235]

Filaggrin, a filamentous protein associated with cytoker-atins, has been identified as the main target for APF andAKAs [36]. Specific assays for antifilaggrin antibodies(AFAs) using the purified protein and based on Westernblotting [26,28,34,37] or ELISA [33,35,38] have been devel-oped. AFAs have been detected in 5467% of patients usingELISA; with Western blotting, the proportions were 6065%in long-standing RA and 5030% in early RA [29,32]. How-ever, among patients with early RA and negative tests forRFs, only 827% had a positive ELISA and 13% a positiveWestern blot for AFAs. Epitopes recognized on the filaggrinmolecule contain large numbers of deiminated arginine resi-dues, which are converted to citrulline [39]. This posttrans-

lation change in the protein is not specific of filaggrin, so thatother citrullinated proteins may in fact be the target of antifil-aggrin antibodies, e.g. fibrin [36] and vimentin [40], both ofwhich are found in the synovium. An ELISA for antibodies toCCP has been developed and found positive in 75% ofpatients with long-standing RA and in 5060% of those withearly RA. Among patients with early RA and negative testsfor RF, 1743% (mean, 35%) had a positive ELISA foranti-CCP [6,27,32,41,42].

3.5. Antibodies to Sa [7,16,4143]

Anti-Sa antibodies are detected by line immunoassay us-ing human spleen or placenta extract as the substrate. Theantigen migrates at an apparent molecular weight of 50 kDa.IgG anti-Sa antibodies are present in 43% of patients withRA overall, 50% of thosewithRFs, and 27% of those withoutRFs. Anti-Sa seem to be produced early in the clinical courseof the disease (20% of patients with RA of less than 1 yearduration) [42,43] and have been found in 1528% of RF-negative patients [7,41]. The Sa protein is also a diminatedcitrullinated protein.

3.6. Antibodies to RA 33 [11,42,44,45]

Anti-RA 33 are antinuclear antibodies (ANA) that areundetectable by indirect immunofluorescent assay but can be

Table 3IgM rheumatoid factor. Rate of detection of IgM rheumatoid factors invarious situations that may suggest rheumatoid arthritis

Primary Sjgrens syndrome 60%SLE 20%Polymyositis 10%Hyperglobulinemic purpura 100%Mixed cryoglobulinemia 100%Cirrhosis and chronic hepatitis 50%Infective endocarditis 50%

RA, rheumatoid arthritis; SLE, systemic lupus erythematosus.



5/15

demonstrated by line immunoassay using a protein-richnuclear extract. Serum anti-RA 33 IgG are detectable in 35%of RA patients with or without IgM anti-IgG RF. Anti-RA33 are produced early in the clinical course of RA; thus,

positive tests have been found in 1528% of patients whohave RA without RFs [11,42]. Anti-RA 33 have only 85%specificity in patients with advanced inflammatory joint dis-ease: they are found in 5060% of patients with mixedconnective tissue disease (MCTD) and in 25% of adults andchildren with SLE. Specificity was only 75% in a Frenchcohort of patients with early inflammatory joint disease [3].

3.7. Antibodies to calpastatin [7,46,47]

Calpastatin is the natural specific inhibitor of proteolyticenzymes called calpains. The calpastatin molecule is an anti-genic protein, and its main epitopes recognized by autoanti-bodies are located on the C-terminal part of the molecule.Using fusion proteins, anticalpastatin autoantibodies can beassayed by Western blot or ELISA. Anti-RA-6 (calpastatin)antibodies have been detected by Western blotting in 57% ofpatients with long-standing RA and anti-RA-1 by ELISA in33% of patients with early RA overall and 28% of patientswith early RA but no RFs [7]. Specificity is low (71% bywestern blotting); in particular, the ELISA is positive in 37%of patients with early inflammatory joint disease other thanRA. Consequently, these antibodies have no diagnostic valuewhen assayed with the currently available methods.

Tables 4 and 5 recapitulate the diagnostic performance

data for the main autoantibodies in four of the largest pro-spective longitudinal studies of patients with early inflamma-tory joint disease. Requiring two positive tests for the diag-nosis had 9699% specificity for RA but only 3450%sensitivity (Table 5). Mean positive and negative predictivevalues were 85% and 75%, respectively.

4. Antinuclear antibodies and inflammatory joint

disease revealing connective tissue disease

ANA are a family of nonorgan-specific autoantibodies

that are extremely useful for refining or excluding a diagnosisof connective tissue disease. They should be evaluated usingthe cascade strategy [48], in which the first step is an indirectimmunofluorescent assay on a high-performance substratesuch as the human epithelioma cell line HEp2. In the secondstep, the antibodies are characterized by tests based on puri-fied or synthetic antigens.

4.1. ANA testing to screen for connective tissue disease

HEp2 cells are rapidly superseding rodent liver or kidneysections for global ANA detection. Because HEp2 cells are

more sensitive but also less specific than rodent substrates, ahigher cutoff must be used to define a positive test. Tan et al.[49] reported that 32% of normal individuals had a positiveHEp2-cell test with the1:40 cutoff used for rodent substrates;the percentage was 13% with a cut-off of 1:80, 5% with acutoff of 1:160, and only 3.3% with a cutoff of 1:320. Thus,1:160 is a reasonable cutoff, as it excludes 95% of normaladults. In children, 1:80 seems appropriate [50].

The numerous commercially available kits for ANA de-tection on Hep2 cells vary widely in their performance char-acteristics [51].

ANA detection remains the most sensitive immunologictest for diagnosing SLE (93%). However, ANAs are found in

many other conditions, so that specificity is low (57%). Thepositive predictive value is 11%, and the positive likelihoodratio is only 2.2. A negative test nearly rules out SLE; how-ever, the test may be negative early in the disease, andrepeated testing is in order in patients with suspected earlySLE and a negative ANA test. ANA-negative SLE has be-

Table 4Sensitivity of autoantibodies for the diagnosis of rheumatoid arthritis in patients with early inflammatory joint disease

Schellekens et al.(2000) [6]

GoldbachMansdky et al.(2000) [7]

Saraux et al.(2002) [3]

Jansen et al.(2002) [8]

Number of patients 288* 238 270 379Minimum number of affected joints 1 1 2

Median arthritis duration


6/15

come exceedingly rare since the widespread availability ofHep2 cell-based assays; nevertheless, a few SLE patientshave isolated anti-SSA/Ro antibodies, isolated antiribosomeantibodies, or apparently primary antiphospholipid syn-drome with an atypical presentation characterized by truepolyarthritis. Only when the results of ANA testing are avail-able should the clinician resort to tests that are more specific(>95%) but less sensitive, i.e. anti-dsDNA (sensitivity about60% and specificity 97.4%, positive likelihood ratio 16.4)and anti-Sm (sensitivity 10% in Caucasians and 2530% inCaribbean or African blacks) [52].

Table 6 reports data on the precision of ANA testing ininflammatory joint disease and connective tissue disease[53].

The performance characteristics of ANA testing for thediagnosis of other connective tissue diseases show that ANAsare valuable for detecting systemic scleroderma but are oflimited assistance in diagnosing Sjgrens syndrome, poly-myositis, or dermatomyositis.

4.2. Antibodies to double-stranded DNA

Several methods have been described for detecting anti-dsDNA antibodies. The Farr radioimmunoassay detects

high-affinity anti-dsDNA (sensitivity, 40100%; specificity,7099%). The immunofluorescent assay on Crithidia luci-liae is highly specific (93100%) and perhaps slightly moresensitive (detects both high-affinity and low-affinity anti-dsDNA, sensitivity 1397%). The ELISA is more sensitive(1982%) but less specific if the cutoff is too low (specificity9698%). Only high levels of IgGs or IgGs in combinationwith other isotypes indicate SLE [52,54].

4.3. Antinucleosome antibodies

The nucleosome is the fundamental subunit of chromatin.Each nucleosome is a plurimolecular complex composed of acore of four pairs of histone molecules (H2AH2B, H3, andH4), which are basic proteins, enclosed within a coil ofdouble-stranded DNA wrapped twice around the core andcontaining on average 146 base pairs. The DNA is fastened tothe core by another histone, H1, which serves to separate twoadjacent nucleosomes. Within the chromatin, the nucleo-somes are connected to one another like beads on a string,this last being a double-stranded DNA fragment composed of2060 base pairs.

Several published studies have evaluated the clinical rel-evance of antinucleosome antibodies detected using com-

Table 5Performance characteristics of autoantibodies for the diagnosis of early rheumatoid arthritis in four prospective longitudinal studies

Study Sensitivity(%)

Specificity(%)

PPV(%)

NPV(%)

PLR

1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4IgM RF 54 66 45 46.5 91 87 94 96.7 74 80 82 81 76 75 6 5 7.5 14IgA RF 30 52.7 94 76 71 5IgG RF 23 91 59 68 2.5AKA 26 47 84 94 56 82 76 1.6 7.8APF 52 79 68 74 2.5Anti-CCP 48 41 42.6 96 91 97.5 84 78 81 66 12 4.5 16.8Sa 22 98 88 61 11RA33 2 25 99 75 66 2 1IgM RF+ anti-CCP 39 33.3 98 97.5 91 78 19.5 13.3IgM RF or anti-CCP 63 73 55.4 88 83 96.7 72 83 5.2 4.3 15.8IgM RF + 1 other 50 34 96 99 90 79 70 80 12.5 34

Study 1, Schellekens; Study 2, Goldbach-Mansky; Study 3, Saraux; Study 4, Jansen. PPV, positive predictive value; NPV, negative predictive value; PLR,positive likelihood ratio; RF, rheumatoid factor; AKA, antikeratin antibodies; APF, antiperinuclear factor; CCP, cyclic citrullinated peptides.

Table 6Performance characteristics of ANA testing in the diagnosis of various inflammatory joint diseases

Condition Sensitivity

(%)

Specificity versus Likelihood ratioOther CTDs

(%)

Nonrheumatic diseases(%)

Healthyindividuals(%)

Overall

(%)

Positive Negative

SLE 93 43 75 78 57 2.2 0.11Scleroderma 85 44 75 71 54 1.86 0.27Sjgrens syndrome 48 44 91 71 52 0.99 1.01Polymyositis/DM 61 52 91 82 63 1.67 0.61RA in adults 41 38 85 82 56 0.93 1.06JIA 57 39 0.95 1.05JIA with uveitis 80 53 1.68 0.39Raynauds 64 48 8 15 41 1.08 0.88

CTD, connective tissue disease; SLE, systemic lupus erythematosus; DM, dermatomyositis; RA, rheumatoid arthritis; JIA, juvenile idiopathic arthritis.



7/15

mercial kits readily available to office-based practitioners.IgG antinucleosomes [55,56] were found in 5685% of SLEpatients overall and in 100% of those experiencing a flare oftheir disease. However, these antibodies were also found in

546% of patients with systemic scleroderma and in 2045%of those with MCTD, two conditions in which anti-dsDNAantibodies are negative (Table 7). A recent study showed thatthe presence of antinucleosome antibodies in patients withsystemic scleroderma was an artifact related to contamina-tion of the nucleosome preparation by topoisomerase I(Scl70) [57]. These data indicate high sensitivity of antinu-cleosome antibodies for SLE, often with persistently positivetests during remissions (62%) or when anti-dsDNA antibod-ies are or become undetectable (1030%). Calculated speci-ficity has ranged across studies from 94% to 97% [55,58]. Ina multicenter European study comparing the sensitivity andspecificity of four different commercial kits for antinucleo-some antibody detection, sensitivity ranged from 67% to84% and specificity from 92% to 98% [59].

4.4. Antinuclear antibodies and Raynauds phenomenon

In patients with isolated Raynauds phenomenon, a posi-tive test for ANA increases the risk of connective tissuedisease from 19% to 30%, and a negative test decreases therisk to 7% [60]. Thus, ANA testing is of minimal assistancein the evaluation of Raynauds phenomenon.

4.5. Antinuclear antibodies and systemic scleroderma

Several patterns of nuclear fluorescence strongly support adiagnosis of systemic scleroderma: thus, the anticentromerepattern is highly suggestive of limited systemic scleroderma(extending no higher than the elbows and knees), of which

CREST syndrome is one variant. Thus, anticentromere anti-bodies are found in 33% of scleroderma patients and have aspecificity of 99.9% and a positive likelihood ratio of 327 ascompared to healthy controls. However, performance is ex-

cellent also when scleroderma is compared to other connec-tive tissue diseases or when CREST is compared to otherforms of scleroderma (Table 8, as shown by Reveille et al.[61]). Anticentromere antibodies have been found in a verysmall number of patients with rheumatic diseases, includingSjgrens syndrome [62] and primary biliary cirrhosis.

A nucleolar pattern of fluorescence also suggests systemicscleroderma, with reported prevalences of 1540% in sclero-derma patients. Above all, this pattern is highly specific forsclerodermapolymyositis overlap syndromes or scleroder-matopolymyositis.

Some antibody specificities can be characterized by agarprecipitation, such as anti-Pm/Scl, but others require moresophisticated techniques that are not available on a routinebasis. Table 9 recapitulates the performance characteristicsof these antinucleolar antibodies for the diagnosis of sclero-derma.

Another ANA that is useful in the diagnosis of systemicscleroderma is antitopoisomerase I or Scl 70, which is highlyspecific of diffuse and severe variants of the disease. Sensi-tivity is 20% overall but reaches 40% in diffuse forms withnearly 100% specificity versus other connective tissue dis-eases (Table 10).

The data reported in Tables 810 indicate that anticen-tromere and Scl 70 antibodies should be looked for to supporta diagnosis of systemic scleroderma. In contrast, testing forantinucleolar antibodies is not recommended at present inthis situation [61].

Table 7Rate of detection of antinucleosome antibodies

Author [reference] Sample size Sensitivity SpecificitySLE Other CTDs Normal controls

Amoura et al. [55] SLE: 120, Scl: 37, MCTD: 20 71.7% 45.9% (Scl et MCTD) 3.5% 96%Min et al. [96] SLE: 129 86% 2% 97%Ben Hadj Hmida et al. [57] SLE: 32, Other: 55 Scl 81% 2%

Bruns et al. [58] SLE: 136, CTD: 30 56% 3% 97%SLE, systemic lupus erythematosus; CTD, connective tissue disease; MCTD, mixed connective tissue disease.

Table 8Performance characteristics of anticentromere antibodies in the diagnosis of systemic scleroderma

Scleroderma Sensitivity(%)

Specificity(%)

Likelihood ratioPositive Negative

Any scleroderma Normal controls 33 99.9 327 0.7Other CTD 31 97 12.5 0.7Raynauds 24 90 2.3 0.8

Crest Normal controls 65 99.9 650 0.4Other scleroderma 61 84 3.9 0.5Other CTD 61 98 29 0.4Raynauds 60 83 3.5 0.2

Limited scleroderma (ACR) Other scleroderma 44 93 6.1 0.6

CTD, connective tissue disease; ACR, American College of Rheumatology.



8/15

4.6. Antinuclear antibodies and undifferentiated connective

tissue disease

A fairly common clinical situation that leads to tests for

detecting and characterizing the main ANAs is undifferenti-ated connective tissue disease (UCTD). Among patients withUCTD, about 4070% have arthralgia and 2035% pol-yarthritis with Raynauds phenomenon [6365].

Longitudinal cohort studies have established that a limitedproportion of patients progress to a definite connective tissuedisease within 5 years. This proportion has ranged from 6%in a study by Danieli et al. [65] to more than 50%; it was 35%in the most recent series, reported by Bodolay et al. [63]. Insome studies, all the patients with definite diagnoses hadSLE, whereas in others the full range of connective tissuediseases was represented, including SLE, RA, Sjgrens syn-

drome, systemic scleroderma, MCTD, and polymyositis (in[64]). Autoantibodies that predict progression to SLE includeANA with a homogeneous pattern, anti-dsDNA, anti-Sm,and anticardiolipin antibodies. In the series by Danieli et al.[65], a nucleolar pattern of ANA predicted progression tosystemic scleroderma, whereas anti-SSA/Ro predicted pro-gression to Sjgrens syndrome. In another study, a differen-tiated connective tissue disease developed within 5 years in24% of patients with UCTD and anti-SSA/Ro antibodies; thediagnosis was Sjgrens syndrome in 50% of cases and SLEin 30% [66].

4.7. Inflammatory joint disease and antibodies to soluble

nuclear antigens

Some ANAs bind to soluble nuclear antigens and can helpin the diagnosis of various connective tissue diseases that canpresent as polyarthritis. The prevalence of these antibodiesvaries with the sensitivity of the assay. Agar precipitation isbeing increasingly replaced by ELISAs based on purified or

genetically engineered proteins. These ELISAs are moresensitive than Ouchterlonys double-diffusion method[6769], but a comparison of several ELISA kits showed thatperformance characteristics were not always satisfactory

[70].

4.7.1. Anti-U1 RNP (anti-70 kDa proteins and/or anti-A

and anti-C associated with U-ARN)

Anti-U1 RNP antibodies are found only in patients withMCTD (also called Sharp syndrome), since this antibody issufficient to define the disease (100% sensitivity). However,there is an initial serologically latent phase. Over the time,the disease can either burn out or progress to a major connec-tive tissue disease. In this situation, the test for anti-U1RNPmay be negative even with a highly sensitive method such asWestern blotting.

Specificity is poor, as anti-U1RNP are found in 30% ofpatients with SLE overall (with higher proportions in chil-dren and young adults) and in 1520% of patients withsystemic scleroderma, particularly those who are presentwith manifestations of MCTD. Thus, given the low incidenceof MCTD as compared to SLE or systemic scleroderma, thepositive likelihood ratio is very low for the diagnosis ofMCTD.

4.7.2. Anti-Sm (anti-BB and D proteins associated

with U1-RNA)

These antibodies have low sensitivity but nearly 100%specificity for SLE. They have been added to the classifica-tion criteria for SLE alongside anti-dsDNA and antiphospho-lipid antibodies. The prevalence of anti-Sm by agar diffusionin patients with SLE is 10% in Caucasians and 30% inCaribbean and African blacks and in Asians. Anti-Sm anti-bodies usually coexist with anti-U1RNP. Although extremelyuncommon, presence of anti-Sm without anti-dsDNA hasbeen reported. ANA, in contrast, are consistently present.

Table 9Performance characteristics of antinucleolus antibodies in the diagnosis of systemic scleroderma

Specificity Sensitivity(%)

Specificity(%)

Likelihood ratioPositive Negative

U3-RNP/fibrillarin and diffuse scleroderma 12 97 4.0 0.9RNA polymerase and diffuse scleroderma 38 94 6 0.7Pm/Scl and sclerodermatopolymyositis 50 98 31 0.5

Table 10Performance characteristics of anti-Scl70/topoisomerase I in the diagnosis of systemic scleroderma

Assay Comparison Sensitivity (%) Specificity (%) Likelihood ratioPositive Negative

ID Normal controls 20 100 >25 0.8Other CTD 40 99.6 52 1.5Raynauds 28 98 10 0.7

ELISA Normal controls 43 100 >55 0.6Other CTD 43 90 4.3 0.6

WB Normal controls 41 99.4 68 0.6Other CTD 40 99 40 0.6

ID, immuno-diffusion; CTD, connective tissue disease; WB, western blot.



9/15

4.7.3. Anti-SSA/Ro and anti-SSB/La

These two antibody types are discussed together becausethey can coexist in the same serum, although their diagnosticvalue differs slightly.

Anti-SSA/Ro recognizes (at least) two different proteins,Ro 60 and 52 kDa associated (or not) with nuclear andcytoplasmic hYARN.This intracellular location explains thatin a few patients, probably less than 1% [72], anti-Ro anti-bodies (usually 52 kDa) are found despite a negative HEp2test for ANAs [71]. The universal health insurance system inFrance accepts to reimburse tests for anti-SSA/Ro even inpatients without ANAs. However, anti-SSA/Ro antibodiesare of very little help for establishing a diagnosis. Three mainclinical situations are usually described:

SLE: 30% of patients have anti-SSA/Ro antibodiesoverall, and higher rates of occurrence are found in

patients who are older than 55 years of age (60%) and inthose with spontaneous (90%) or drug-induced [73]subacute cutaneous lupus, lupus syndrome with con-genital complement deficiency (C4, C2 or C1q) (>90%),or neonatal cutaneous lupus with or without congenitalatrioventricular block (>95%);

primary Sjgrens syndrome (50%) or Sjgrens syn-drome associated with SLE (>60%) or more rarely an-other connective tissue disease. Among patients withSjgrens syndrome, women are more likely than men tohave anti-SSA/Ro.

Undifferentiated connective tissue disease (UCTD):16% of these patients have anti-SSA/Ro [64,65]. Bod-

olay et al. [63] reported that presence of anti-SSA/Roearly in the course of UCTD predicted progressionwithin 5 years to Sjgrens syndrome (66% vs. 6%; RR,18.5; CI, 7.522.2). Results were similar with anti-SSB.Among asymptomatic women who gave birth to babieswith congenital atrioventricular block or neonatal lupus,many had manifestations of UCTD or Sjgrens syn-drome 510 years later, whereas SLE was uncommon[74,75]. The introduction of commercially available kitsthat selectively detect anti-Ro 52 and 60 kDa has addedto (and complicated) our knowledge of the diagnosticvalue of these antibodies [76]. Most patients with SLE

or primary Sjgrens syndrome have both anti-Ro60 and 52 kDa. In a few patients, however, only one ofthese antibodies is detected: isolated presence of anti-Ro60 kDa may be mainly associated with SLE and isolatedpresence of anti-Ro 52 kDa with Sjgrens syndrome(40% of Sjgrens syndrome patients with anti-Ro and20% of lupus patients with anti-Ro) [77]. Isolatedanti-Ro 52 kDa are also found in patients with UCTD.Finally, anti-Ro 52 kDa are found in combination withanti-Jo1 in 50% of patients with polymyositis and anti-synthetase syndrome [78].

Anti-SSB/La recognize a 48-kDa protein associated withhYRNA. They are found in about 50% of patients withprimary Sjgrens syndrome and 15% of those with SLE.Among patients with Sjgrens syndrome and anti-SSA/Ro,

2060% also had anti-SSB/La, and among SLE patients withanti-SSA/Ro, 2040% also had anti-SSB/La in studies bySimmons OBrien et al. [79] and Sibilia et al. [80].

Polyarthralgia or polyarthritis are fairly often associated

with anti-SSA/Ro and/or SSB/La, in both Sjgrens syn-drome and SLE. In one study, 74% of Sjgrens syndromepatients had anti-SSA/Ro and 55% of SLE patients testedpositive for these antibodies; among these patients with anti-SSA/Ro, 34.5% of those with Sjgrens syndrome and 26.7%of those with SLE had polyarthritis [80]. In a study byOBrien of 100 patients with anti-SSA/Ro, 75% of the pa-tients had polyarthritis, which was related to SLE, Sjgrenssyndrome, or UCTD.

Anti-SSA/Ro and anti-SSB/La have been incorporatedinto the American and European classification criteria forSjgrens syndrome [81]. Anti-SSA/Ro has low specificity.

Anti-SSB/La, either in isolation or with anti-SSA/Ro, seemsto be more specific of Sjgrens syndrome. Thus, in a studyby Venables et al. [82], 83% of the patients with anti-La andanti-Ro antibodies met criteria for Sjgrens syndrome, ascompared to only 42% of the patients who had anti-Rowithout anti-La.

4.8. Inflammatory joint disease and antibodies

to HEp2-cell cytoplasm

The biologist may report absence of nuclear fluorescenceduringANA testing but presence of fluorescence confined to,or predominating in, the cytoplasm. In a patient with clinical

polyarthritis, this finding is associated mainly with threediagnoses.

Antimitochondrial antibodies related to primary biliarycirrhosis (99% sensitivity), with or without scleroderma.

Antiribosome antibodies directed against the P0, P1 andP2 proteins. These antibodies occur in patients with SLE(20% sensitivity). The vast majority of patients also haveANAs or, more rarely, anti-dsDNA antibodies; however,in fewer than 3% of patients, the antiribosome antibod-ies are the only markers for SLE.Antiribosome antibod-ies to the P0 protein have been described as specific forSLE (>95%) with either neurological involvement (de-

pression) or, above all, glomerular disease and pol-yarthritis. However, antiribosome antibodies can befound in other inflammatory joint diseases and connec-tive tissue diseases [83].

Anti-tARN synthetase antibodies should prompt asearch for polymyositis. These antibodies are usuallyanti-Jo1 antibodies (histidyl tARN synthetase), whichoccur in 2025% of patients with primary polymyositisand are associated with interstitial pulmonary fibrosisthat may reveal antisynthetase syndrome.

Antimitochondrial antibodies, antiribosome antibodies,and anti-J01 can each be characterized using a specificELISA.

Fig. 1 recapitulates the diagnostic value of ANAs in pa-tients with inflammatory joint disease.



10/15

5. Primary systemic vasculitis and polyarthritis

Among the main primary vasculitides, only HenochSchnlein purpura nearly always manifests as acute pol-yarthritis. No autoantibodies specific for HenochSchnlein purpura have been identified, and antibodiesto endothelial cells have no diagnostic value. Serum IgAlevels are high, suggesting predominant mediation by anIgA response.

Leukocytoclastic vasculitides manifest mainly as skinlesions, although polyarthritis may occur. No autoanti-bodies are associated with this disease, which is veryoften induced by a drug or an infection. In vasculitisassociated with primary cryoglobulinemia (i.e. not re-lated to hepatitis C virus infection), a consistent findingis the presence of monoclonal or polyclonal IgM RFsthat produce a positive latex or WaalerRose test. Thepresence of ANAs, anti-SSA/Ro, and/or anti-SSB/Lashould prompt investigations for another condition, ei-ther Sjgrens syndrome or, more rarely, SLE, revealedby hyperglobulinemic purpura.

Wegeners granulomatosis deserves careful discussionas a common pattern of primary vasculitis (annual inci-dence in Europe, 510/106 population) [84] that oftenpresents as polyarthritis resembling RA. Subcutaneousnodules mimicking rheumatoid nodules may be presentand 20% of patients seen early in the course of thedisease have positive tests for IgM RFs.

Among 17 patients with Wegeners granulomatosis evalu-ated in a teaching hospital, 13 (76%) had joint manifesta-tions: six had arthralgia and seven arthritis, and these symp-toms were inaugural in three and six patients, respectively.Nodules were found in two patients. Of 15 patients, five(33%) had positive tests for RFs [85]. ANCA testing is ofconsiderable interest in this situation: the result was positive

with a diffuse cytoplasmic pattern (C-ANCA) in five of six(83%) patients tested during a period with joint manifesta-tions.

More generally, C-ANCA detected by immunofluores-cence had sensitivity values ranging from 34% to 92% andspecificity values from 88% to 100% in large publishedseries [86], yielding a mean sensitivity of 66% (95% CI,5774%) and a mean specificity of 98% (95% CI,9699.5%). In patients with active Wegeners granulomato-

sis (and polyarthritis in this condition always indicates activedisease), mean sensitivity is 91% (95% CI, 8795%) andmean specificity 99% (95% CI, 9799.9%). During phases ofquiescence, the corresponding figures are 63% and 99.5%,respectively. In patients with limited forms (without renaland/or pulmonary involvement), C-ANCA are less commonthan in diffuse forms, being found in about 60% of casesduring phases of disease activity. The C-ANCA pattern re-flects the presence of antiproteinase 3 (PR3), which can beassayed using a specific quantitative ELISA. Peripheral AN-CAs (P-ANCA) are found by immunofluorescence in about20% of patients with Wegeners granulomatosis [87], and

their presence can be confirmed by a quantitative ELISAspecific for myeloperoxidase. There is general agreementthat immunofluorescence is more sensitive than ELISA forANCA detection. Therefore, when there is a clinical suspi-cion of Wegeners granulomatosis, an immunofluorescentassay should be obtained first and an ELISA used subse-quently for confirmation. Table 11 summarizes the perfor-mance characteristics of immunofluorescent assay andELISA for the diagnosis of Wegeners disease.

The numerical data reported in this chapter were obtainedin series of individuals with well-characterized diseases, ei-ther patients with a variety of inflammatory or infectiousconditions or healthy controls. This artificially elevates thesensitivity and specificity values. Few studies have evaluatedpatient populations with suspected systemic vasculitis. Stone

Fig. 1. Screening for and characterizing ANAs: prevalence in connective tissue disease.



11/15

et al. [88] studied 856 sera from patients with suspectedANCA-associated conditions. Only 8% met ChapelHill cri-teria for vasculitis; 49% of patients with C-ANCA and 62%of those with P-ANCA did not have vasculitis (Wegenersgranulomatosis or other primary vasculitides). The perfor-mance characteristics of ANCA for the diagnosis of vasculi-

tis are shown in Table 12. Performance was better still in astudy by Choi et al., in which the anti-PR3/C-ANCA andanti-MPO/ANCA-P combinations had the highest sensitivity(85.5%) and provided excellent specificity (98.6%) [89].Inflammatory joint diseases potentially associated withANCA production include RA (P-ANCA in 2040% of pa-tients overall and in higher proportions of patients withvasculitis or Feltys syndrome), SLE (29%), and Sjgrenssyndrome (1117%) [90].

This probably explains the disappointing predictive valuesin studies of ANCA testing in patients with a variety ofconditions: thus the positive and negative predictive values

for vasculitis were 59% and 84%, respectively, in a study byMcLaren et al. [91]. Raoetal. [92] reported on the usefulnessof C-ANCA for the diagnosis of Wegeners disease in212 patients with suspected vasculitis and a sufficiently longfollow-up to provide a definite diagnosis based on clinical orhistological criteria. Wegeners disease was confirmed basedon ACR criteria in 25 patients; only six of these patients hadbiopsy-proven disease. Of the 212 patients, 14 had positiveC-ANCA results. The sensitivity and specificity of C-ANCAfor Wegeners granulomatosis as defined by the ACR wereonly 28% (95% CI, 1046%) and 96% (95% CI, 9399%);the positive predictive value was 50% and the negative pre-

dictive value 91%, yielding a positive likelihood ratio of 7 forACR-defined disease. When only biopsy-proved cases were

considered, sensitivity was higher (83%; 95% CI, 53100%),but the positive predictive value remained low (36%).

6. Polyarthritis and antiphospholipid antibodies

Tests for anticardiolipin and antiphospholipid antibodiesare of limited usefulness in the diagnosis of polyarthritis.Although 2030% of patients with SLE have detectableantiphospholipid antibodies, they also haveANAs, which arefar more sensitive for the diagnosis.

Anticardiolipin and antiphospholipids may be useful inpatients with primary antiphospholipid syndrome, but pol-yarthritis is virtually never present in this condition. Piette etal. [93] even suggested that polyarthritis may rule out thisdiagnosis. Only arthralgia and myalgia are widely acceptedamong internists as manifestations of primary antiphospho-

lipid syndrome. Among populations seen by rheumatolo-gists, the situation is more complex, and we have describedborderline syndromes that have more in common with pri-mary antiphospholipid syndrome than with SLE [94]. How-ever, antiphospholipid antibodies have low specificity, and inrheumatology populations they are not infrequently found inwell-characterized conditions such as pseudomyalgia rheu-matica and giant cell arteritis [95], at the initial inflammatoryphase (3050% of cases); their presence should prompt in-vestigations for extensive arteritis, in particular with involve-ment of the abdominal aorta.

Antiphospholipid antibodies have been reported also inother connective tissue diseases including primary Sjgrens

syndrome (237%), RA (1823%), and systemic sclero-derma, as well as in patients with articular and other extrahe-patic manifestations of hepatitis C virus infection (25%),where they showed no correlation with thrombotic events[90].

Thus, testing for antiphospholipid antibodies as a first-lineinvestigation is of no help in the diagnosis of early inflamma-tory joint disease.

In conclusion, Fig. 2 suggests a decision tree for autoanti-body testing aimed at obtaining diagnostic orientation in apatient with early inflammatory joint disease. Although onlythe main diagnoses are indicated in this figure, this strategy

should be helpful to the practitioner, whose clinical judgmentremains the best diagnostic tool.

Table 11Antineutrophil cytoplasmic antibodies in the diagnosis of Wegeners granu-lomatosis (metaanalysis)

Sensitivity SpecificityC-ANCA 64 95ELISA PR3 66 87C-ANCA + ELISA PR3 73 99P-ANCA 21 ELISA MPO 24

ANCA, antineutrophil cytoplasmic antibody; MPO, myeloperoxidase.

Table 12Performance characteristics of antineutrophil cytoplasmic antibodies (C or P) in the diagnosis of primary systemic vasculitis (from Stone et al. [87])

IF ANCA Sensitivity(%)

Specificity(%)

PPV(%)

NPV(%)

PLR

IF, ANCA 67 93 45 97 9.4IF, C-ANCA 42 96 51 95 11.8IF, P-ANCA 25 96 38 94 6.9ELISA (PR3 or MPO) 55 99 83 96 54.2ELISA PR3 35 99 80 95 45.5ELISA MPO 20 99.7 88 93 79.8IF and ELISA 52 99 88 96 82.1

ANCA,antineutrophil cytoplasmic antibody; IF, immunofluorescent assay; MPO,myeloperoxidase; PPV, positive predictive value; NPV, negative predictivevalue; PLR, positive likelihood ratio.



12/15

References

[1] American College of Rheumatology ad hoc committee on immuno-logic testing guidelines. Guidelines for immunologic laboratory test-ing in the rheumatic diseases: an introduction. Arthritis Rheum[Arthritis Care Res] 2002;47:32933.

[2] Vittecoq O,de Bandt M,MeyerO, HachullaE, leLotX, les membresdu CRI. Les facteurs rhumatodes sont-ils utiles au diagnostic noso-logique dun rhumatisme inflammatoire voluant depuis moins de12 mois en labsence de signes cliniques dorientation? Rev Rhum2002;69:1358.

[3] Saraux A, Berthelot JM, Chales G, Le Henaff C, Mary JY,Thorel JB, et al. Value of laboratory tests in early prediction ofrheumatoid arthritis. Arthritis Rheum 2002;47:15565.

[4] Berthelot JM, Wendling D, Combe B, Le Lot X, Saraux A, pourle CRI. Performances des critres 1987 de polyarthrite rhumatode delAmerican College of Rheumatology dans le contexte des arthritesdbutantes: tude de la littrature. Rev Rhum [ed Fr] 2002;69:12834.

[5] Meyer O. Les auto-anticorps devant un rhumatisme inflammatoiredbutant. In: Blotman F, Combes B, Leroux JL, Sany J, editors.Journes montpelliraine de Rhumatologie. Montpellier: Sauramps

medical ed; 2001. p. 5160.[6] Schellekens GA, Visser H, de Jong BAW, van den Hoogen FHJ,

Hazes JMW, Breedveld FC, et al. The diagnostic properties of rheu-matoid arthritis antibodies recognizing a cyclic citrullinated peptide.Arthritis Rheum 2000;43:15563.

[7] Goldbach-Mansdky R, Lee J, McCoy A, Hoxworth J, Yarboro C,Smolen JS, et al. Rheumatoid arthritis associated autoantibodies inpatients with synovitis of recent onset. Arthritis Res 2000;2:23643.

[8] Jansen AMIE, van der Horst-Bruinsma IE, van Schaardenburg D, vande Stadt RJ, de Koning MHMT, Dijkmans BAC. Rheumatoid factorand antibodies to cyclic citrullinated peptide differentiate rheumatoidarthritis from undifferentiated polyarthritis inpatients with earlyarthritis. J Rheumatol 2002;29:20746.

[9] Harrison BJ, Symmons DPM, Barrett EM, Silman AJ. The perfor-

mance of the 1987 ARA classification criteria for rheumatoid arthritisin a population based cohort of patients with early inflammatorypolyarthritis. J Rheumatol 1998;25:232430.

[10] Saraux A, Berthelot JM, Chals G, Le Henaff C, Thorel JB,Hoang S, et al. Ability of the American College of Rheumatology1987 criteria to predict rheumatoid arthritis in patients with earlyarthritis and classification of these patients 2 years later. ArthritisRheum 2001;44:248591.

[11] Cordonnier C, Meyer O, Palazzo E, De Bandt M, Elias A, Nicaise P.Diagnostic value of anti-RA33 antibody, antikeratin antibody, anti-

perinuclear factor and antinuclear antibody in early rheumatoidarthritis: comparison with rheumatoid factor. Br J Rheumatol 1996;35:6204.

[12] Von Essen R, Kurki P, Isomaki H, Okubo S, Kautiainen H, Aho K.Prospect for an additional laboratory criterion for rheumatoid arthri-tis. Scand J Rheumatol 1993;22:26772.

[13] Schmerling RH, Delbanco TL. How useful is the rheumatoid factor?An analysis of sensitivity, specificity and predictive value.Arch InternMed 1992;152:241720.

[14] SarauxA,VallsI, VoisinV, KoreichiA, Baron D,Youinou P, et al.Howuseful are tests for rheumatoid factors, antiperinuclear factors, antik-eratin antibody, and the HLA DR4 antigen for the diagnosis of rheu-matoid arthritis? Rev Rhum 1995;62:1620.

[15] Shmerling RH, Delbanco TL. The rheumatoid factor: an analysis ofclinical utility.Am J Med 1991;91:52834.

[16] Vittecoq O, Jouen-Beades F, Krzanowska K, Bichon-Tauvel I,Mnard JF, Daragon A, et al. Les facteurs rhumatodes, les anticorpsantifilaggrine et la faible production in vitro dinterleukine 2 etdinterfron-gamma sont des marqueurs immunologiques utiles audiagnostic prcoce de la polyarthrite rhumatode dans une populationde recrutement libral. Etude prliminaire. Rev Rhum 2001;68:23949.

[17] Wolfe F, Cathey MA, Roberts FK. The latex test revisited. Rheuma-toidfactor testing in 8287rheumatic disease patients.Arthritis Rheum1991;34:95160.

[18] Jonsson T, Steisson K, Jonsson H, Geirsson AJ, Thorsteinsson J,Valdimarsson H. Combined elevation of IgM and IgA rheumatoidfactor has high diagnostic specificity for rheumatoid arthritis. Rheu-matol Int 1998;18:11922.

[19] Berthelot JM, MaugarsY, CastagneA, Audrain M, Prost A. Antiperi-nuclear factors are present in polyarthritis before ACR criteria forrheumatoid arthritis are fulfilled. Ann Rheum Dis 1997;56:1235.

Fig. 2. Autoantibodies of assistance in the diagnosis of inflammatory joint disease. Decision tree.



13/15

[20] Swedler W, Wallman J, Froelich CJ, Teodorescu M. Routine measure-ment of IgM, IgG and IgA rheumatoid factors: high sensitivity, speci-ficity and predictive value for rheumatoid arthritis. J Rheumatol 1997;24:103744.

[21] Visser H, Gelinck LBS, KampfraathAH, Breedveld FC, Hazes JMW.

Diagnosticand prognostic characteristicsof the enzyme linked immu-nosorbent rheumatoid factor assays in rheumatoid arthritis. AnnRheum Dis 1996;55:15761.

[22] Berthelot JM, Garnier P, Glmarec J, Flipo RM. Valeur des anticorpsanti-prinuclaires au 1:100 pour le diagnostic de polyarthrite rhuma-tode. Etude de 600 patients et mta-analyse de la littrature. RevRhum 1998;65:915.

[23] Hoet RM, Van Venrooij WJ. The antiperinuclear factor (APF) andantikeratin antibodies (AKA) in rheumatoid arthritis. In: Smolen JS,Kalden JR, Maini RN, editors. Rheumatoid arthritis. Berlin:Heidelberg: Springer; 1992. p. 299318.

[24] Aho K, Von Essen R, Kurki P, Palosuo T, Heliovaara M. Antikeratinantibody and antiperinuclear factor as markers for subclinical rheu-matoid disease process. J Rheumatol 1993;20:127881.

[25] Alnigenis NNY, Warikoo S, Barland P. Antiperinuclear factor andantikeratin antibody: diagnostic use in a clinical laboratory setting. JRheumatol 2000;27:20567.

[26] Vincent C, de Keyser F, Masson-Bessire C, Sebbag M, Veys EM,Serre G. Anti-perinuclear factor compared with the so called antik-eratin antibodies and antibodies to human epidermis filaggrin, in thediagnosis of arthritis. Ann Rheum Dis 1999;58:428.

[27] Van Jaarsveld CHM, ter Borg EJ, Jacobs JWG, Schellekens GA,Gmeling-Meyling FH, van Booma-Frankfort C, et al. The prognosticvalue of the antiperinuclear factor, anti-citrullinated peptide antibod-ies and rheumatoid factor in early rheumatoid arthritis. Clin ExpRheumatol 1999;17:68997.

[28] Forslind K, Vincent C, Serre G, Svensson B. Antifilaggrin autoanti-bodies in early rheumatoid arthritis. Scand J Rheumatol 2000;29:3202.

[29] Aho K, Palosuo T, Lukka M, Kurki P, Isomaki H, Kautiainen H, et al.Antifilaggrin antibodies in recent-onset arthritis. Scand J Rheumatol1999;28:1136.

[30] Le Goff P, Youinou P, Jouquan J. Etude de la prcocit dapparitiondes anticorps antikratine. Rev Rhum 1986;53:699700.

[31] Meyer O, Fabregas D, Cyna L, Ryckewaert A. Les anticorps anti-kratine. Un marqueur des polyarthrites rhumatodes volutives. RevRhum 1986;53:6015.

[32] Kroot EJJA, de Jong BAW, van Leeuwen MA, Swinkels H, van denHoogen FH, vant Holf M, et al. The prognostic value of anti-cycliccitrullinated peptide antibody in patients with recent-onset rheuma-toid arthritis. Arthritis Rheum 2000;43:18315.

[33] Paimela L, Palosuo T, Aho K, Lukka M, Kurki P, Leirisalo-Repo M, et al.Association of autoantibodiesto filaggrinwith an active

disease in early rheumatoid arthritis. Ann Rheum Dis 2001;60:325.

[34] Union A, Meheus L, Humbel RL, Conrad K, Steiner G,Moereels H, et al. Identification of citrullinated rheumatoidarthritisspecific epitopes in natural filaggrin relevant for antifilag-grin autoantibody detection by line immunoassay. Arthritis Rheum2002;46:118595.

[35] Vincent C, Nogueira L, Sebbag M, Chapuy-Regaud S, Arnaud M,Letourneur O, et al. Detection of antibodies to deiminated recombi-nant rat filaggrin by enzyme-linked immunosorbent assay. A highlyeffective test for the diagnosis of rheumatoid arthritis. ArthritisRheum 2002;46:20518.

[36] Serre G.Autoanticorps antifilaggrine/antifibrine dimine. Outre leurintrt diagnostique et pronostique, les autoanticorps

antifilaggrine/antifibrine dimine (AFA) sont probablement impli-qus dans la physiopathognie rhumatode. Rev Rhum2001;68:2013.

[37] Vincent C, Simon M, Sebbag M, Girbal-Neuhauser E, Durieux JJ,Cantagrel A, et al. Immunoblotting detection of autoantibodies tohuman epidermis filaggrin: a new diagnostic test for rheumatoidarthritis. J Rheumatol 1998;25:83846.

[38] Palosuo T, Lukka M, Alenius H, Kalkkinen N, Aho K, Kurki P, et al.

Purification of filaggrin from human epidermis and measurement ofantifilaggrin autoantibodies in sera from patients with rheumatoidarthritis by an enzyme-linked immunosorbent assay. Int Arch AllergyImmunol 1998;115:294302.

[39] Schellekens GA, de Jong BAW, Van den Hoogen FHJ, Van dePutte LBA. Citrulline is an essential constituent of antigenic determi-nants recognized by rheumatoid arthritis-specific autoantibodies. JClin Invest 1998;101:27381.

[40] Menard HA, Lapointe E, Rochd MD, Zhou ZJ. Insights into rheuma-toid arthritis derived from the Sa immune system.Arthritis Res 2000;2:42932.

[41] Hayem G, Chazerain P, Combe B, Elias A, Haim T, Nicaise P, et al.Anti-Sa antibody is an accurate diagnostic and prognostic marker inadult rheumatoid arthritis. J Rheumatol 1999;26:713.

[42] Hueber W, Hassfeld W, Smolen JS, Steiner G. Sensitivity and speci-

ficity of anti-Sa autoantibodies for rheumatoid arthritis. Rheumatol1999;38:1559.

[43] Desprs N, Boire G, Lopez-Longo FJ, Mnard HA. The Sa system: anovel antigenantibody system specific for rheumatoid arthritis. JRheumatol 1994;21:102733.

[44] Hassfeld W, Steiner G, Hartmuth K, Kolarz G, Scherak O,Graninger W, et al. Demonstration of a new antinuclear antibody(anti-RA 33) that is highly specific for rheumatoid arthritis. ArthritisRheum 1989;32:151520.

[45] Hassfeld W, Steiner G, Graninger W, Witzmann G, Schweitzer H,Smolen JS. Autoantibody to the nuclear antigen RA33: a marker forearly rheumatoid arthritis. Br J Rheumatol 1993;32:199203.

[46] Desprs N, Talbot G, Plouffe B, Boire G, Mnard A. Detection andexpression of a cDNA clone that encodes a polypeptide containingtwo inhibitory domains of human calpastatin and its recognition byrheumatoid arthritis sera. J Clin Invest 1995;95:18916.

[47] Mimori T, Suganuma K, Tanami Y, Nojima T, Matsumura M,Fujii T, et al. Autoantibodies to calpastatin (an endogenous inhibitorfor calcium-dependent neutral protease, calpain) in systemic rheu-matic diseases. Proc Natl Acad Sci USA 1995;92:726771.

[48] Homburger HA. Cascade testing for autoantibodies in connectivetissue diseases. Mayo Clin Proc 1995;70:1834.

[49] Tan EM, Feltkamp TEW, Smolen JS, Butcher B, Dawkins R, Frit-zler MJ, et al. Range of antinuclear antibodies in healthy individu-als. Arthritis Rheum 1997;40:160111.

[50] Craig WY, Ledue TB, Johnson AM, Ritchie RF. The distribution ofantinuclear antibodytiters in normal children and adults. J Rheuma-tol 1999;26:9149.

[51] Emlen W, ONeill L. Clinical significance of antinuclear antibodies.

Comparison of detection with immunofluorescence and enzyme-linked immunosorbent assays. Arthritis Rheum 1997;40:16128.[52] Kavanaugh AF, Solomon DH, American College of Rheumatology ad

hoc committee on immunologic testing guidelines. Guidelines forimmunologic laboratory testing in the rheumatic diseases: anti-DNAantibody tests.Arthritis Rheum [Arthritis Care Res] 2002;47:54655.

[53] Solomon DH, KavanaughAJ, Schur PH, American College of Rheu-matology ad hoc committee on immunologic testing guidelines.Evidence-based guidelines for the use of immunologic tests: anti-nuclear antibody testing. Arthritis Rheum [Arthritis Care Res] 2002;47:43444.

[54] Smeenk RJT. Detection of antibodies to dsDNA: current insights intoits relevance. Clin Exp Rheumatol 2002;20:294300.

[55] Amoura Z, Koutouzov S, Chabre H, Cacoub P, Amoura I, Mus-set L, et al. Presence of antinucleosome autoantibodies in a restricted

set of connective tissue diseases. Antinucleosome antibodies of theIgG3 subclass are markers of renal pathogenicity in systemic lupuserythematosus. Arthritis Rheum 2000;43:7684.



14/15

[56] Chabre H, Amoura Z, Piette JC, Godeau P, Bach JF, Koutouzov S.Presence of nucleosome-restricted antibodies in patients with sys-temic lupus erythematosus. Arthritis Rheum 1995;38:148591.

[57] Ben Hadj HmidaY, Schmit P, Gilson G, Humbel RL. Failure to detectantinucleosome antibodies in scleroderma: comment on the article by

Amoura et al. Arthritis Rheum 2002;46:2802.[58] Bruns A, Bl S, Hausdorf G, Burmester GR, Hiepe F. Nucleosomes

are major T and B cell autoantigens in systemic lupus erythematosus.Arthritis Rheum 2000;43:230715.

[59] Goetz J, HumbelRL, Monier JC,Cohen J,Andr C, BacqueryA, et al.Diagnostic value of nucleosome specific antibodies in recent activenon treated SLE. Arthritis Rheum 2000;43:S379 (abstract 1878).

[60] Luggen M, Belhorn L, Evans T, Fitzgerald O, Spencer-Green G. Theevolution of Raynauds phenomenon: a long-term prospective study.JRheumatol 1995;22:222632.

[61] Reveille JD, Solomon DH, American College of Rheumatology adhoc committee on immunologic testing guidelines. Evidence-basedguidelines for the use of immunologic tests: anticentromere, Scl-70and nucleolar antibodies. Arthritis Rheum [Arthritis Care Res] 2003;

49:399412.[62] Tubach F, Hayem G, Elias A, Nicaise P, Ham T, Kahn MF, et al.Les anticorps anticentromre en pratique rhumatologique nesont pas toujours associs une sclrodermie. Rev Rhum 1997;64:42530.

[63] Bodolay E, Csiki Z, Szekanecs Z, Ben T, Kiss E, Zeher M, et al.Five-year follow-up of 665 Hungarian patients with undifferentiatedconnective tissue disease (UCTD). Clin Exp Rheumatol 2003;21:31320.

[64] Mosca M, Neri R, Bombardieri S. Undifferentiated connective tissuediseases (UCTD): a review of theliterature anda proposal forprelimi-nary classification criteria. Clin Exp Rheumatol 1999;17:61520.

[65] Danieli MG, Fraticelli P, Franceschini F, Cattaneo R, Farsi A, Passal-eva A, et al. Five-year follow-up of 165 Italian patients with undiffer-entiated connective tissue diseases. Clin Exp Rheumatol 1999;17:58591.

[66] Cavazzana I, Franceschini F, Belfiore N, Quinzanini M, Caporali R,Calzavara-Pinton P, et al. Undifferentiated connective tissue diseasewith antibodies to Ro/SSA: clinical features and follow-up of148 patients. Clin Exp Rheumatol 2001;19:4039.

[67] Meilof JF, Bantjes I, De Jong J, van Dam AP, Smeenk RJT. Thedetection of anti-Ro/SSA and anti-La/SSB antibodies. A comparisonof counterimmunoelectrophoresis with immunoblot, ELISA, andRNA-precipitation assays. J Immunol Methods 1990;133:21526.

[68] Morozzi G, Bellisai F, Simpatico A, Pucci G, Bacarelli MR, Cam-panella V, et al. Comparison of different methods for the detection ofanti-Ro/SSA antibodies in connective tissue disease. Clin Exp Rheu-matol 2000;18:72931.

[69] Manoussakis MN, Kistis KG, Liu X, Aidinis V, Guialis A, Moutso-

poulos HM. Detection of anti-Ro (SSA) antibodies in autoimmunediseases: comparison of five methods. Br J Rheumatol 1993;32:44955.

[70] Tan EM, Smolen JS, mcDougal JS, Butcher BT, Conn D, Dawk-ins R, et al. A critical evaluation of enzyme immunoassays for detec-tion of antinuclear autoantibodies of defined specificities. I. Precision,sensitivity and specificity. Arthritis Rheum 1999;42:45564.

[71] Pourmand N, Blomberg S, Rnnblom L, Karlsson-Parra A, Petters-son I, Wahren-Herlenius M. Ro 52 kDa autoantibodies are detected ina subset of ANA-negative sera. Scand J Rheumatol 2000;29:11623.

[72] Peene I, Meheus L,Veys EM, de Keyser F. Diagnostic associations ina large and consecutively identified population positive for anti-SSAand/or anti-SSB: the range of associated diseases differs according tothe detailed serotype. Ann Rheum Dis 2002;61:10904.

[73] Srivastava M, Rencic A, Diglio G, Santana H, Bonitz P, Wat-son R, et al. Drug-induced, Ro/SSA-positive cutaneous lupus erythe-matosus. Arch Dermatol 2003;139:459.

[74] Brucato A, Franceschini F, Gasparini M, de Juli E, Ferraro G, Quin-zanini M, et al. Isolated congenital complete heart block: long-termoutcome of mothers, maternal antibody specificity and immunoge-netic background. J Rheumatol 1995;22:53340.

[75] Julkunen H, Kurki P, Kaaja R, Heikkil R, Immonen I,

Chan EKL, et al. Isolated congenital heart block. Long-term outcomeof mothers and characterization of the immune response to SS-A/Roand to SS-B/La. Arthritis Rheum 1993;36:158898.

[76] Peene I, Meheus L, de Keyser S, Humbel R, Veys EM, de Keyser F.Anti-Ro52 reactivity is an independent and additional serum markerin connective tissue disease. Ann Rheum Dis 2002;61:92933.

[77] Ben-Chetrit E, Fox RI, Tan EM. Dissociation of immune responses tothe SSA (Ro) 52 and 60 kDa polypeptides in systemic lupus erythe-matosus and Sjgrens syndrome. Arthritis Rheum 1990;33:34955.

[78] Rutjes SA, Vree Egberts WTM, Jongen P, van den Hoogen F,Pruijn GJM, van Venrooij WJ. Anti-Ro 52 antibodies frequentlyco-occur with anti-J01 antibodies in serafrom patients with idiopathicinflammatory myopathy. Clin Exp Immunol 1997;109:3240.

[79] Simmons-OBrien E, Chen S, Watson R, Antoni C, Petri M, Hoch-berg M, et al. One hundred anti-Ro (SSA) antibody positive patients:

a 10-year follow-up. Medicine 1995;74:10930.[80] Sibilia J, Goetz J, Berthel C, Koegler A, Maloisel F, Javier RM.

Signification clinique et biologique des anticorps anti-Ro/SSA. Etudeet suivi dun groupe de 127 patients avec anticorps anti-Ro/SSA. In:Abuaf N, Pierron D, Rouquette AM, editors. Anticorps anti-SSA/RoMthode de dpistage. Valeur diagnostique, BMD; 1996. p. 91112.

[81] Vitali C, Bombardieri S, Jonsson R, Moutsopoulos HM, Alex-ander EL, Carsons SE, et al. Classification criteria for Sjgrenssyndrome: a revised version of the European criteria proposed by theAmericanEuropean Consensus Group. Ann Rheum Dis 2002;61:5548.

[82] Venables PJW, Shattles W, Pease CT, Ellis JE, Charles PJ, Maini RN.Anti-La (SSB): a diagnostic criterion for Sjgrens syndrome? ClinExp Rheumatol 1989;7:1814.

[83] Johanet C, Andr C, Sibilia J, Baquey A, Oksman F, SanMarco M, et al. Signification clinique des anticorps antiribosomes.Rev Md Interne 2000;21:5106.

[84] Watts RA, Scott DGI. Epidemiology of the vasculitides. Curr OpinRheumatol 2003;15:116.

[85] Alcalay M, Azais I, Pallier B, Touchard G, Patte F, Brugier JC, et al.Les manifestations articulaires de la maladie de Wegener. A propos de13 observations. Rev Rhum 1990;57:84553.

[86] Rao JK, Weinberg M, Oddone EZ, Allen NB, Landsman P, Feuss-ner JR. The role of antineutrophil cytoplasmic antibody (c-ANCA)testing in the diagnosis of Wegener granulomatosis. A literaturereview and meta-analysis. Ann Intern Med 1995;123:92532.

[87] Hagen EC, Daha MR, Hermans J, Andrassy K, Csernok E,Gaskin G, et al. Diagnostic value of standardized assays for anti-neutrophil cytoplasmic antibodies in idiopathic systemic vasculitis.

Kidney Int 1998;53:74353.[88] Stone JH, Talor M, Stebbing J, Uhlfelder ML, Rose NR, Car-son KA, et al. Test characteristics of immunofluorescence and ELISAtests in 856 consecutive patients with possible ANCA-associatedconditions.Arthritis Rheum [Arthritis Care Res] 2000;13:42434.

[89] Choi HK, Liu S, Merkel PA, Colditz GA, Niles JL. Diagnostic perfor-mance of antineutrophil cytoplasmic antibody tests for idiopathicvasculitides: metaanalysis with a focus on antimyeloperoxidase anti-bodies. J Rheumatol 2001;28:158490.

[90] Hachulla E, De Bandt M, Dubucquoi S, Vittecoq O, Le Lot X,Meyer O, et al. Intrt du dosage des anticorps antinuclaires, desanticorps antiphospholipides et des anticorps anticytoplasme des neu-trophiles dans le diagnostic nosologique des rhumatismes inflamma-toires chroniques dbutants sans signe clinique dorientation. RevRhum [Ed Fr] 2002;69:13946.

[91] McLaren JS, Stimson RH, McRorie ER, Coia JE, Luqmani RA. Thediagnostic value of anti-neutrophil cytoplasmic antibody testing in aroutine clinical setting. Q J Med 2001;94:61521.



15/15

[92] Rao JK,Allen NB, Freussner JR, Weinberger M. A prospective studyof antineutrophil cytoplasmic antibody (c-ANCA) and clinical criteriain diagnosing Wegeners granulomatosis. Lancet 1995;346:92631.

[93] Piette JC, Wechsler B, Frances C, Papo T, Godeau P. Exclusion criteriafor primary antiphospholipid syndrome. J Rheumatol 1993;20:18024.

[94] Weber M, Hayem G, De Bandt M, Seifert B, Palazzo E, Roux E, et al.Classification of an intermediate group of patients with antiphospho-lipid syndrome and lupus-like disease: primary or secondaryantiphospholipid syndrome? J Rheumatol 1999;26:21316.

[95] Meyer O, Nicaise P, Moreau S, de Bandt M, Palazzo E,Hayem G, et al. Antibodies to cardiolipin and beta 2 glycoprotein I inpatients with polymyalgia rheumatica and giant cell arteritis. RevRhum [Engl Ed] 1996;63:2417.

[96] Min DJ, Kim SJ, Park SH, Seo YI, Kang HJ, Kim WU, et al.Anti-nucleosome antibody: significance in lupus patients lackinganti-double-stranded DNA antibody. Clin Exp Rheumatol 2002;20:138.


Auto Antc y Enf Inflamatoria

Documents

Transcript of Auto Antc y Enf Inflamatoria