Vasculitis nrrheum.2014.78

494 | AUGUST 2014 | VOLUME 10 www.nature.com/nrrheum

Department of Rheumatology, Klinikum Bad Bramstedt & University Hospital of Schleswig-Holstein, Oskar-Alexanderstrasse 26, 24576 Bad Bramstedt, Germany (E.C., F.M.).

Correspondence to: E.C. [email protected]

Current and emerging techniques for ANCA detection in vasculitisElena Csernok and Frank Moosig

Abstract | Detection of antineutrophil cytoplasmic antibodies (ANCAs) is a well-established diagnostic test used to evaluate suspected necrotizing vasculitis of small blood vessels. Conditions associated with these antibodies, collectively referred to as ANCA-associated vasculitides, include granulomatosis with polyangiitis (formerly known as Wegener granulomatosis), microscopic polyangiitis, and eosinophilic granulomatosis with polyangiitis (formerly known as Churg–Strauss syndrome). The diagnostic utility of ANCA testing depends on the type of assay performed and on the clinical setting. Most laboratories worldwide use standard indirect immunofluorescence tests (IFT) to screen for ANCA and then confirm positive IFT results with antigen-specific tests for proteinase 3 (PR3) and myeloperoxidase (MPO). Developments such as automated image analysis of immunofluorescence patterns, so-called third-generation PR3-ANCA and MPO-ANCA ELISA, and multiplex technology have improved the detection of ANCAs. However, challenges in routine clinical practice remain, including methodological aspects of IFT performance, the diverse antigen-specific assays available, the diagnostic value of testing in clinical settings and the prognostic value of serial ANCA monitoring in the prediction of disease relapse. This Review summarizes the available data on ANCA testing, discusses the usefulness of the various ANCA assays and advises on the clinical indications for the use of ANCA testing.

Csernok, E. & Moosig, F. Nat. Rev. Rheumatol. 10, 494–501 (2014); published online 3 June 2014; doi:10.1038/nrrheum.2014.78

IntroductionSince the 1980s, the presence of antineutrophil cytoplasmic antibodies (ANCAs) has proven to be a valuable serological marker that aids the diagnosis of smallvessel vasculitis. The principle association with ANCAs originally defined the group of ANCAassociated vasculitides, comprising granulomatosis with polyangiitis (GPA, formerly known as Wegener granulomatosis), microscopic poly angiitis (MPA) and eosinophilic granulomatosis with polyangiitis (EGPA, formerly known as Churg–Strauss syndrome), which have different frequencies of ANCApositivity.1–4 However, the spectrum of dis eases asso ciated with ANCAs has since broadened to include a range of other inflammatory and infectious dis eases, calling into question the diagnostic implications of ANCA positivity (reviewed elsewhere5–7). Furthermore, the application of ANCA testing as a clinical tool is still con sidered controversial,8 with issues around methodo logical aspects (such as assay performance and testing strategies), when to order ANCA testing (clinical ‘gating policy’), and the value and therapeutic implications of sequential ANCA testing. With respect to the last point, for instance, the reported prevalence of ANCAs in GPA varies widely, from 50% to 95% depending on disease stage, disease activity and therapy at the time of sampling, and the ANCA detection method used. Inconsistencies in ANCA detection frequently cause incorrect applications and interpretations in daily clinical practice.8

Despite these controversies, ANCA testing is widely used by clinicians when considering a diagnosis of ANCAassociated vasculitis (AAV). As indicated in the consensus guidelines for ANCA testing and reporting document, and according to the practice of most laboratories worldwide, screening for vasculitisassociated ANCAs is done by indirect immuno fluorescence technique (IFT).9,10 Three main patterns of fluorescence have been demonstrated with IFT on e thanolfixed neutrophils (Figure 1). The first, CANCA, is a diffuse, cytoplasmic granular fluorescence most prominent centrally, between the nuclear lobes, and is noted overall in 90% of patients with active, generalized GPA.11 The second, PANCA, is a perinuclear neutrophil staining pattern often with nuclear extension, noted in those with MPA and EGPA. The third, socalled atypical pattern, referred to as AANCA, is rare and combines both cytoplasmic and perinuclear or nuclear staining, with multiple antigen specificities; AANCA can occur in association with drug exposure, inflammatory bowel disease or rheumatoid arthritis, most often in the absence of vasculitis. The perinuclear or nuclear staining pattern of PANCA (or AANCA) is actually an artefact of ethanol fixation, as neutrophils show a cytoplasmic staining pattern after formalin fixation. The consensus guidelines require all sera to be examined by IFT on et hanolfixed neutrophils.9,10 However, it is technically difficult to differentiate PANCA (or AANCA) patterns from the staining shown by antinuclear antibodies (ANA) on this type of fixation. Several laboratories incorporate a protocol in which formalinfixed neutrophils are used,

Competing interestsThe authors declare no competing interests.

REVIEWS

© 2014 Macmillan Publishers Limited. All rights reserved

http://www.nature.com/doifinder/10.1038/nrrheum.2014.78

NATURE REVIEWS | RHEUMATOLOGY VOLUME 10 | AUGUST 2014 | 495

but the value of formalin fixation of neutrophils in differentiating antibodies is controversial. Comparing the IFT patterns of sera tested on ethanolfixed and formalinfixed neutrophil preparations, our group demonstrated that formalin fixation does adequately differentiate between ANA and PANCA (or AANCA) staining, and can been used in the routine diagnostic laboratory.12 The use of different fixation methods (ethanol and formalin) thus enables differentiation between PANCA (or AANCA) and the presence of ANA.

The major target antigens recognized by ANCAs in AAV are proteinase 3 (PR3) and myeloperoxidase (MPO). PR3ANCA usually cause a CANCA pattern and are mainly associated with GPA. MPOANCA is associated with a PANCA pattern and is predominantly seen in patients with MPA. Positive immuno fluorescence results should be followed by antigenspecific tests for PR3ANCA and MPOANCA.9 The tests most frequently used to identify ANCAs with different specificities are ‘direct’ and ‘capture’

ELISA. The combination of both IFT and antigenspecific assays (PR3ANCA and MPOANCA) has been confirmed in several studies to be the best strategy for ANCAdetection in vasculitis.11,13 In a specialized laboratory, ANCA detection uses both methods (IFT and ELISA), and contradictory results can be further analysed by other methods, by using other cell substrates, and by investigating other target antigens (elastase, cathepsin G, bactericidal permeability increasing protein and so on). It is important to emphasize that the extensive ANCAtesting approach used in these laboratories is rarely used in clinical practice, where IFT with or without direct ELISA is routine.

A variety of different immunoassays for the detection of ANCAs have been developed in the past 10 years, inclu ding thirdgeneration ELISA, dot blots and beadbased multiplex assays, along with automated systems for analy sing ANCA fluorescence patterns. This Review sum marizes the most clinically relevant developments in methodology and provides an overview on how ANCA testing in vasculitis could be improved and simplified by applying these techniques.

Automated analysis of ANCA patternsDespite the emergence of alternative screening methods, IFT remains the recommended method for ANCA screening in vasculitis. However, the reliability of this assay depends on a number of factors, including the substrate, conjugates and fixation methods used, the source of the cells, storage conditions, and procedures for incubation and washing.7 Furthermore, the assay is subjective, time consuming, requires expertise and is labour intensive. To address the limitations of conventional IFT, computerbased image analysis of indirect immuno fluorescence (IIF) patterns has been applied to the analysis of ANCA in neutrophilbased assays. Standardized, automated analysis of IFT for ANCA detection provides several advantages, including reduced consumption of samples and reagents, shortened analysis time and less extensive handling of samples. These systems, several of which are commercially available, generally use digital acquisition and computeraided analysis of IIF images using pattern recognition algorithms. These systems might distinguish the presence or absence of ANCAs, or might also be able to identify particular staining patterns.

One automated system used in ANCA detection was first developed for ANA detection on human epithelial type 2 (HEp2) cells. In contrast to HEp2 cells, neutrophils are characterized by nuclei of varying shapes, and algorithms for the identification of staining patterns in neutrophils are more complex than those for HEp2 cells. For instance, signal assessment and pattern classification cannot be performed inside or outside of the DAPIstained (that is, nuclear) area, as described for ANA pattern detection. Furthermore, CANCA and PANCA staining patterns are detected in the border regions of the neutrophil nucleus, rendering the detection of ANCAs a challenge for automation.14

Several studies have evaluated the usefulness of automated IIF systems for the objective interpretation of ANCA patterns. In a 2003 study, Boomsma et al.15 first

Key points

■ Screening for antineutrophil cytoplasmic antibodies (ANCAs) specific for proteinase 3 (PR3) and myeloperoxidase (MPO) remains a basic tool in the serological evaluation of systemic vasculitic disorders

■ Automated analysis of ANCA immunofluorescence demonstrates good agreement with conventional techniques when assessing patients with suspected vasculitis and, in the future, might improve efficiency and standardization in clinical laboratories

■ Antigen-specific ANCA assays that use bridging molecules exhibit superior performance compared with conventional assays and can be used for PR3-ANCA and MPO-ANCA detection

■ The new methods for ANCA detection and evaluation in ANCA-associated vasculitis should be urgently evaluated in multicentre studies, in anticipation of updating of standardization processes and a revision of existing strategies

a b

Figure 1 | ANCA IIF pattern differentiation by automated signal intensity analysis. IIF images, taken automatically by AKLIDES®, can discriminate a | P-ANCA-specific and b | C-ANCA-specific staining of neutrophils by use of mathematical algorithms for pattern differentiation. Chromatin is stained by DAPI (blue) and specific ANCA interactions are revealed by FITC (green)-labelled secondary anti-human IgG. DAPI and FITC fluorescence intensity signals of respective images were combined and illustrated in three dimensions (x, object length; y, object width; z, light intensity of fluorescence signal). Abbreviations: ANCA, antineutrophil cytoplasmic antibody; A-ANCA, atypical ANCA staining pattern; C-ANCA, cytoplasmic ANCA staining pattern; DAPI, 4',6-diamidino-2-phenylindole; FITC, fluorescein isothiocyanate; IIF, indirect immunofluorescence; P-ANCA, perinuclear ANCA staining pattern. Reproduced from Knütter et al. Arthritis Res. Ther. 14, R271 (2012),17 which is published under an open-access licence (http://creativecommons.org/licenses/by/2.0) by BioMed Central Ltd.

FOCUS ON VASCULITIS


http://creativecommons.org/licenses/by/2.0


compared a novel quantitative image analysis technique with conventional IFT and various ELISAs for monitoring ANCAs in a prospective cohort of 16 patients with PR3ANCApositive GPA. With respect to the detection of ANCA, the automated image analysis technique had a lower diagnostic sensitivity (75%) than conventional IFT (100%) or capture ELISA (100%). In determining the ability of ANCA levels to predict disease relapse, they found that increases in ANCA titres determined by automated image analysis were followed by relapse in the majority of patients with GPA (69%), a positive predictive value somewhat better than that of conventional IFT (56%) and comparable to that of direct ELISA (61–71%). Furthermore, the quantification of ANCA titres by image analysis was no different than by IFT, direct ELISA and capture ELISA.15 However, the image analysis method used in this study did not use different neutrophil fixation methods, but rather used the recommended method of ethanol fixation.

Three studies published in 2012 evaluated the AKLIDES® (Medipan GmbH, Germany) automated IIF analysis system,16–18 which is able to quantify fluorescence intensity and interpret basic ANCA staining patterns on combined ethanolfixed and formalinfixed human neutrophils. Melegari et al.16 compared the diagnostic performance of AKLIDES® with that of both manual interpretation of IIF by laboratory experts and confirmatory tests (ANCAspecific ELISAs). The agreement between traditional visual reading and AKLIDES® of ANCA testing by IIF was 89.1%. However, this study involved only a small number of samples (n = 46), lacked any clinical information about the samples and reported only positive–negative discrimination between ANCA screening results.

Knütter et al.17 developed interpretation software for the AKLIDES® system using patternrecognition algorithms that enabled positive–negative discrimination and classification of CANCA and PANCA. In a study of sera from 342 patients with AAV and other rheumatic and infectious diseases, and from healthy individuals, they compared these algorithms and conventional fluorescence microscopy in the analysis of ANCA patterns on ethanolfixed and formalinfixed neutrophils (Figure 2). No significant difference was found between the performance of the visual and automated methods of ANCA detection in patients with AAV and other rheumatic and infectious diseases.17 The two methods showed strong agreement in the positive–negative classification of samples and in the interpretation of ANCA patterns on ethanolfixed and formalinfixed neutrophils, confirming the usefulness of automated patternrecognition algorithms for the assessment of ANCA IIF patterns. The main differences between these two methods for the interpretation of IIF patterns on ethanolfixed neutrophils were found for visual findings regarding the cytoplasmic and atypical patterns. The automated reading defined 22 (25.0%) of 88 visually determined CANCA patterns as negative results. Furthermore, 10 (18.2%) of 55 AANCA patterns determined visually were defined as nuclear patterns using the automated pattern recognition algorithms.17

To further clarify the diagnostic value of the AKLIDES® ANCA patternrecognition system on et hanolfixed and formalinfixed neutrophils, Damoiseaux et al.18 used sera from patients with AAV (n = 79) as well as distinct cohorts of relevant control sera (n = 117). Five series of control samples were used to examine the interference with other autoantibodies: ANApositive sera with different fluorescence patterns; antimitochondrial antibodypositive samples; AANCA positive sera; CANCA with BPI antigen specificities; and sera from healthy individuals. The data obtained in sera from patients with PR3ANCApositive AAV revealed that the automated system lacked sufficient sensitivity (74% compared with 92% for routine micro scopy) on ethanolfixed neutrophils, but the expected CANCA pattern was very well recognized on formalinfixed neutrophils (95% and 97% recognized by the automated system and routine microscopy, respectively). For MPOANCA positive sera, the expected PANCA pattern was best recognized on ethanolfixed neutrophils, whereas recognition of CANCA on fo rmalinfixed neutrophils was poor. Most interference with ANCA pattern recognition occurred in control sera with a homogeneous ANA pattern. In contrast to visual scoring, the AKLIDES® system could not differentiate between PANCA and the presence of ANA in most of the samples.18 Furthermore, the sensitivity of the AKLIDES® system for positive–negative discrimination was 94%, compared with 96% for visual scoring. This study was not appropriate to calculate sensitivity or specificity for AAV, but at least showed that the sensitivity of the automatic reading system might equal that of visual scoring.18 Moreover, the immunofluorescence pattern of ANCA predicts the ANCAantigen specificity, and for this purpose the reactivity on ethanolfixed neutrophils

ethN

formN

Cytoplasmic Perinuclear Nuclear Atypical Atypical Atypical

Cytoplasmic Cytoplasmic Negative Cytoplasmic Negative Negative

10 μm

Figure 2 | Characterization of neutrophils by automated pattern-recognition image analysis. IIF images, taken automatically by AKLIDES®, of serum samples demonstrate C-ANCA, P-ANCA and A-ANCA patterns on ethanol-fixed and formalin-fixed human neutrophils. Chromatin is stained by DAPI (blue) and specific ANCA interactions are revealed by FITC (green)-labelled secondary anti-human IgG. Abbreviations: ANCA, antineutrophil cytoplasmic antibody; A-ANCA, atypical ANCA staining pattern; C-ANCA, cytoplasmic ANCA staining pattern; DAPI, 4’,6-diamidino-2-phenylindole; ethN, ethanol-fixed neutrophils; FITC, fluorescein isothiocyanate; formN, formalin-fixed neutrophils; IIF, indirect immunofluorescence; P-ANCA, perinuclear ANCA staining pattern. Reproduced from Knütter et al. Arthritis Res. Ther. 14, R271 (2012),17 which is published under an open-access licence (http://creativecommons.org/licenses/by/2.0) by BioMed Central Ltd.

REVIEWS


http://creativecommons.org/licenses/by/2.0


is the determining factor. In this respect, the sensitivity of the automated system on ethanolfixed neutrophils needs to be optimized. Furthermore, data concerning the value of image analysis for monitoring ANCA levels are lacking.

Altogether, studies to date show that automated reading of IIF patterns is a promising technique for ANCA screening, but further improvement and evaluation in multicentre studies is required before it will be applicable in routine clinical praxis.

Novel antigen-specific immunoassaysELISAAny new strategy for ANCA evaluation in vasculitis must aim to identify the target antigens PR3 and MPO, because the presence of PR3ANCA or MPOANCA correlates well with clinical, histological and genetic features.19–22 ‘Direct’ ELISA, developed in the late 1990s, uses PR3 or MPO antigens directly immobilized to the surface of the ELISA plate (Figure 3); however, this assay shows considerable variation in performance and often lacks sensitivity.23 This socalled firstgeneration ANCA ELISA has proven to be less sensitive than secondgeneration ELISAs,23 which uses secondary antibodies specific to PR3 or MPO that have been attached to the plate to ‘capture’ ANCA (Figure 3). The capture ELISA technique reduces the possible covering of epitopes by the plastic plate (as can occur with direct ELISA) and demonstrates overall superiority to direct ELISA in diagnostic performance.24–26 However, the capturing antibody might reduce the sensitivity of capture ELISA by masking epitopes of the antigen relevant to ANCA. Interestingly, PR3ANCA detected by capture ELISA correlates better with disease activity in patients with AAV than PR3ANCA detected by conventional ELISA.27 As a consequence, many commercially available capture ELISAs are in widespread clinical use.26,27

Thirdgeneration ELISA—‘anchor’ ELISA— immobilizes antigen to the plastic plate via a bridging molecule (Figure 3), preserving epitopes for binding; of note, almost all of the available anchor assays are PR3ANCA ELISA. The anchor assay has demonstrated superiority to direct ELISA and some capture ELISA assays.28 Thirdgeneration ANCA ELISA is now available (sometimes referred to by commercial companies as ‘high sensitivity’ ELISA) and its evaluation is ongoing28–33 (Table 1).

In a comparative study published in 2012, the diagnostic performance of 11 commercially available PR3ANCA and MPOANCA ELISA assays—notably including direct, capture and anchor ELISA systems—were ana lysed for their ability to distinguish sera from individuals with AAV.32 All the assays evaluated were highly sensitive and specific for GPA and MPA, and were able to differentiate active AAV from other diseases (systemic lupus erythematosus and rheumatoid arthritis). Maximum sensitivity for PR3ANCA in sera from patients with GPA was obtained with the combination of IFT and either capture or anchor ELISA, and for MPOANCA in sera from patients with MPA with the combination of IFT and either capture or direct ELISA (Table 1). In a second comparative study, published in 2013, Radice et al.33 investigated the diagnostic performance of nine commercially available PR3ANCA kits for the diagnosis of GPA (Table 1). Two of the assays were direct PR3ANCA ELISA and the remaining seven represented some of the tests developed in order to improve the performance of the antigenspecific tests (such as capture and anchor PR3ANCA ELISA, chemiluminiscence immuno assay, and fluorescence enzyme immuno assay). This study showed that the diagnostic value generally seems to be improved with anchor and cap ture ELISA, in comparison with direct ELISA. Nota bly, most of the anchor and capture assays improved the positive predictive value for diagnosis of GPA, and most enabled better discrimination between positive and nega tive samples owing to the reduction of borderline test results.33

Altogether, the antigenspecific anchor and capture ANCA assays exhibited superior performance and could be used for PR3ANCA and MPOANCA detection. As a consequence of the high sensitivity and specificity of these assays, a revision of the current algorithm (Figure 4) for ANCA detection is warranted. An alternative strategy has been proposed that uses such novel sensitive tests as the firstline screening assay, with confirmation of positive results by IFT—this approach is practically an inversion of the current sequence.34 However, although published data support the applica bility of these antigenspecific tests for initial screening,13,34 further studies in large cohorts of patients with inflammatory diseases are necessary before the ANCA consensus statement should be revised. An ongoing multi centre international study is focused on this goal. Over the next year, we expect to establish a more efficient test algorithm that provides superior sensitivity and specificity for the diagnosis of AAV.

Other immunoassaysOther immunoassays for antigenspecific ANCA detection include dotblot assays and beadbased multiplex testing, discussed briefly here, and other methods such as fluorescence enzyme immunoassays and chemi luminescence technology that are beyond the scope of this Review.

Dot blots are a variant of ELISA that use an antigencoated nitrocellulose strip. This assay is not quantitative, but it is especially suitable for examination of a single serum in case of emergency because it is a singlestrip preparation.35 A biochip technology (EUROPLUS®, EUROIMMUN AG, Germany) has also been developed for

Speci�cantibodiesin patient

serumPR3

Enzyme-conjugatedanti-human IgG

Bridgingmolecule

Monoclonalantibody

Direct ELISA Anchor ELISACapture ELISA

Figure 3 | Overview of ELISA procedures for ANCA detection. Using the example of antigen-specific tests for PR3, direct ELISA immobilizes the antigen to the solid phase, whereas in capture and anchor ELISA the antigen is bound to a bridging molecule (in the case of capture ELISA, an antibody) attached to the solid phase. Abbreviations: ANCA, antineutrophil cytoplasmic antibodies; PR3, proteinase 3.

FOCUS ON VASCULITIS



rapid ANCA testing.35 In this assay, IFT is com bined with a dotblot test for PR3ANCA and MPOANCA. However, the results obtained are only qualitative and should be confirm ed and quantified by other ANCA assays.

Beadbased multiplex assay is a variant of the solidphase assay in which capture antigens are bound to colourcoded beads in suspension that are then analysed using flow cytometry. The main advantage of this method is the ability to simultaneously detect multiple autoantibodies relevant to vasculitis (ANCAs and antiglomerular basement membrane [GBM] antibodies) in a small serum sample. The potential disadvantages are the same as with direct ELISA; that is, the antigenbinding potential of the PR3 and MPO and its potential loss after the coating process. A study by Trevisin and colleagues36 in sera from individuals with active and treated vasculitis and inflammatory bowel disease found a flow cyto metric immunoassay for PR3ANCA and MPOANCA had a specificity of 88%, compared with 96% and 94% for IFT and ELISA, respectively. The flow cytometric assay was almost as sensi tive as IFT, and more sensitive than, but just as specific as, most of the 12 commercial ELISAs tested in detecting ANCAs in sera from patients with vasculitis.36 However, the utility of this method for initial screening of

patients with suspected vasculitis needs to be evaluated in prospective studies.

Clinical usefulness of ANCA testingAt present, ANCA testing is widely used by clinicians when AAV is suspected. Accurate identification of all patients with active AAV and the avoidance of mis diagnosis are best achieved by use of IFT combined with an antigenspecific assay.11 Using this strategy, PR3ANCA or MPOANCA are detectable in nearly all patients with active, generalized GPA or MPA, but only in approximately 60% of patients with limited (“initial phase”) dis ease.32,37 In cases of emergency, such as the pu lmonaryrenal syndrome (the differential diagnosis of which should include testing for antiGBM antibodies, which are speci fic for Goodpasture syndrome), several rapid screening assays for ANCA (dot blots, biochip technology) are available.35,38 If the results are clearly posi tive, for example, hightitre ANCA on IFT and proof of a defined speci ficity, immunosuppressive therapy should be started immediately. However, patients with ANCA should receive a careful workup because the presence of ANCA can indicate conditions other than vasculitis, such as infections or exposure to drugs, which might be worsened

Table 1 | Comparison of methods for testing for PR3-ANCA and MPO-ANCA in ANCA-associated vasculitis

Patient population (n) vs comparison group (n)

Method Sensitivity (%)

Specificity (%)

PPV (%)

NPV (%)

AUC/ROC Comments

GPA (86) vs non-vasculitic disease (450)28

IFTDirect PR3-ANCA ELISACapture ELISAAnchor ELISA

92607296

999999.398.5

Nd Nd 0.96 (0.94–0.98)0.80 (0.76–0.83)0.86 (0.82–0.89)0.96 (0.94–0.98)

Histological diagnosisRetrospective study

GPA (232) vs inflammatory diseases (661)29

IFTAnchor ELISA

77.980.4

90.997.4

7388

9393

Nd Histological diagnosisProspective study

GPA (59*) vs inflammatory and infectious diseases (585)30

Hn–hr PR3-ANCA ELISACapture ELISADirect (hn) PR3-ANCA

946664

99 (predefined)

Nd Nd Nd Histological diagnosis Retrospective study

GPA (34) vs SLE (65)31

Direct PR3-ANCAAnchor ELISA

97.1 98.4 Nd Nd 0.999 (0.947–1.00)

Clinical diagnosisRetrospective study

GPA (40) vs RA or SLE (20)32

IFTDirect PR3-ANCA (n = 5 kits)Capture ELISA (n = 2 kits)Anchor ELISA (n = 4 kits)

62.545–5560–62.560–62.5

95–100 Nd Nd Nd Histological diagnosisRetrospective study

MPA (40) vs RA or SLE (20)32

IFTDirect MPO-ANCA ELISA (n = 8 kits)Capture ELISA (n = 2 kits)Anchor ELISA (n = 1 kit)

82.562.5–85

8075

95–100 Nd Nd Nd Histological diagnosis Retrospective study

GPA (55) vs suspected vasculitis (175)33

IFTDirect PR3-ANCA ELISA (n = 2 kits)Capture ELISA (n = 2 kits)Anchor ELISA (n = 3 kits)Other assays (n = 2)

69.161.8–72.7

70.9–72.761.8–72.772.7–74.5

10095.4–96.4

95.9–99.598.5–99.095.9–97.9

Nd Nd Nd0.856–0.879

0.862–0.8780.833–0.8810.878–0.902

Clinical diagnosisRetrospective study

*47 of 59 patients in the GPA group had a cytoplasmic ANCA pattern on IFT. Abbreviations: ANCA, antineutrophil cytoplasmic antibody; AUC, area under the curve; GPA, granulomatosis with polyangiitis; hn, human native; hr, human recombinant; IFT, indirect immunofluorescence technique; MPA, microscopic polyangiitis; MPO, myeloperoxidase; Nd, not determined; NPV, negative predictive value; PPV, positive predictive value; PR3, proteinase 3; RA, rheumatoid arthritis; ROC, receiver operating characteristics; SLE, systemic lupus erythematosus.

REVIEWS



by immunosuppression. Thus, rapid ANCA tests might be helpful as an adjunct to urgent therapeutic decisions, but cannot supplant the critical appraisal of all findings including patient history, clinical assessment and imaging as well as other laboratory tests. Histological confirmation of vasculitis is still the goldstandard and should be sought in every patient.

The use of ANCA to monitor disease activity and guide treatment decisions in AAV is controversial (reviewed else where7). A metaanalysis published in 2012 con cluded that serial measurements of ANCA in patients in remission has limited value, as neither an increase in ANCA titre nor the persistence of ANCA during remission were highly predictive of relapse.39 Furthermore, the correlation of ANCA levels with disease activity is reportedly assaydependent, with the PR3ANCA capture ELISA showing some potential as a surrogate marker of disease activity in patients with GPA.27 At present, an increase in ANCA titre during f ollowup of patients with AAV, in the absence of signs of relapse, should prompt intensification of monitoring by measures such as urinalysis. The utility of novel antigenspecific immunoassays—anchor ELISA—for monitoring disease activity and guiding t reatment of patients with AAV requires longitudinal studies.

Whereas PR3ANCA and MPOANCA are highly specific for AAV, they have very little diagnostic value for nonvasculitic conditions. ANCA testing performed in a population with a low pretest probability of vasculitis

is expected to yield a large number of ANCApositive results.13,40 Overestimating the diagnostic relevance of a positive ANCA test could misdirect clinicians and delay appropriate treatment. The diagnostic accuracy of ANCA testing should be improved by an increased pretest probability of vasculitis. The use of a ‘gating policy’ that adheres to clinical guidelines for limiting ANCA testing on the basis of clinical criteria9,10 makes ANCA screening more relevant clinically, as demonstrated in several studies.41–43 Investigating the effect of such a policy at a single regional centre, Arnold et al.43 followed up on patients for whom ANCA testing was deemed inappropriate and found that no diagnoses of AAV were delayed or missed.

Commercial ELISA screening can be used to detect ANCAs with specificities other than PR3 and MPO, such as elastase, cathepsin G, bactericidal permeability increasing protein, lactoferrin and others. However, these specific ANCAs have no diagnostic value for patients with AAV.

ConclusionsThe development of image analysis technology and novel antigenspecific immunoassays for detection of ANCAs is an important contribution to the diagnostic repertoire in the assessment of AAV. The evolution of ANCA testing needs to incorporate automated, multiplexed and increas ingly sensitive technologies that can correctly identify ANCAs in a routine clinical setting. Automated approaches to the performance and interpretation of ANCA IIF offer the opportunity to improve standardization and to address other shortcomings of manual analysis such as its labour intensiveness and its main di sadvantage: subjectivity.

This automation is also expected to have a substantial effect on diagnostic testing by requiring less sample and reagent, analysis time and sample handling. Image analysis technology has proven to be an interesting tool for ANCA screening, but we expect to see further standardization and improvement of this technique over the next few years before its routine application in clinical practice will be feasible.

Within specialized labs conducting clinical studies (as opposed to labs consulted in the course of clinical prac tice), the use of automated ANCA IIF analysis is expected to advance quickly and is already considered as an alternative to conventional manual IFT. Nonetheless, continuous efforts are necessary to ensure this method over comes all the disadvantages of the conventional manual approach. One important task for the near future is to assess the reliability of these automated assays in multicentre studies.

In the development of advanced PR3ANCA and MPOANCA assays, the binding of antigen onto the solid phase by means of bridging molecules improves the accessibility of the antigenrelevant epitope to the ANCA. Consequently, the sensitivity of capture as well as anchor ELISAs seems to be superior to that of conventional ELISA (which directly immobilizes antigen) and to IFT. However, despite several comparative studies, it is still unclear which immunoassays for the detection of PR3ANCA and MPOANCA provide the highest clinical

+

A-ANCA patternP-ANCA patternC-ANCA pattern

MPO-ANCA ELISAPR3-ANCA ELISAMPO-ANCA ELISA

P-ANCA (MPO-ANCA)

OtherANCAELISA

MPO-ANCAELISA

Other ANCAELISA

Cytoplasmicstainingon both

ethanol-�xedand formalin-�xed

cells

Mixed cytoplasmicand perinuclear/

nuclear staining onethanol-�xed cells

Cytoplasmicstaining on

formalin-�xed cells

Cytoplasmicstaining on


Nuclearstaining on


Perinuclear/nuclear stainingon ethanol-�xed cells

IFT on ethanol-�xed and formalin-�xed neutrophils

MPO-ANCA ELISAPR3-ANCA ELISA

Other ANCA ELISA

IFT on HEp2PR3-ANCA ELISA

ANA A-ANCAC-ANCA (PR3-ANCA)

– + – +– + –

Figure 4 | ANCA testing algorithm routinely used in the authors’ laboratory. The testing strategy for ANCA in patients with suspected vasculitis includes initial screening by IFT on ethanol-fixed and formalin-fixed neutrophils, and confirmation of positive results with ELISA. This algorithm conforms to international consensus guidelines for screening on ethanol-fixed neutrophils,9,10 but our laboratory uses additional tests on formalin-fixed neutrophils and HEp2 to better discriminate between P-ANCA (or A-ANCA) and ANA. Abbreviations: ANA, antinuclear antibodies; ANCA, antineutrophil cytoplasmic antibody; A-ANCA, atypical ANCA staining pattern; C-ANCA, cytoplasmic ANCA staining pattern; HEp2, human epithelial type 2 cells; IFT, indirect immunofluorescence test; MPO, myeloperoxidase; P-ANCA, perinuclear ANCA staining pattern; PR3, proteinase 3.

FOCUS ON VASCULITIS



accuracy. Furthermore, the clinical utility of measuring PR3ANCA and MPOANCA for monitoring disease activity remains a topic for further investigation.

The availability of highly sensitive PR3ANCA and MPOANCA assays has raised questions about the twostage diagnostic strategy currently recommended for ANCA detection.9,10 An alternative algorithm13,34 might use such highsensitivity solidphase assays as the initial step, with confirmation of positive results by IFT. Moreover, it has been postulated that the need to perform IIF–based assays might be eliminated altogether.7 How ever, at present, the evidence that IIF can be replaced by PR3ANCA and MPOANCA assays is insufficient, and future studies are needed to establish whether the al ternative ANCAdetection algorithm is feasible.

Altogether, these studies enable us to draw some conclusions. First, IFT performs consistently well but seems to be slightly less sensitive than anchor ELISA, which can also show a higher positive predictive value. Anchor ELISA, however, awaits further evaluation for its utility in monitoring treated and relapsing disease and with regard to its reproducibility in nonspecialist laboratories. The current practice of screening for ANCAs by

IFT and confirming IFTpositive sera in antigenspecific immuno assays is reasonable for now, but improvements in the methods for ANCA detection and the development of novel detection technologies could affect the validity of the current international recommendations, lead to updating of the standardization process and a revision of existing ANCA diagnostic strategies. Clearly, though, the development of any new testing strategy for ANCAs in vasculitis must identify the ANCA target antigens, as PR3ANCA and MPOANCA serotype correlate well with the disease expression.

Review criteria

A search for original articles published between January 2000 and October 2013 was performed in MEDLINE without language restrictions. Search terms used, in various combinations, included: “vasculitis”, “antineutrophil cytoplasmic antibodies”, “proteinase 3-ANCA”, “myeloperoxidase-ANCA”, “methods”, “ELISA”, “indirect immunofluorescence”, “automation” and “diagnostic utility”. All articles identified were full-text papers. The reference list was last updated February 2014.

1. Davies, D. J. et al. Segmental necrotizing glomerulonephritis with antineutrophil antibody: possible arbovirus aetiology? Br. Med. J. (Clin. Res. Ed.) 285, 606 (1982).

2. Van der Woude, F. J., Rasmussen, N. & Lobatto, S. Autoantibodies against neutrophils and monocytes: tool for diagnosis and marker of disease activity in Wegener’s granulomatosis. Lancet 1, 425–429 (1985).

3. Jennette, J. C. et al. Nomenclature of systemic vasculitides. Proposal of an international consensus conference. Arthritis Rheum. 37, 187–192 (1994).

4. Jennette, J. C. et al. 2012 revised international Chapel Hill consensus conference nomenclature of vasculitides. Arthritis Rheum. 65, 1–11 (2013).

5. Schmitt, W. H. & van der Woude, F. J. Clinical applications of antineutrophil cytoplasmic antibody testing. Curr. Opin. Rheumatol. 16, 9–17 (2004).

6. Csernok, E., Lamprecht, P. & Gross, W. L. Diagnostic significance of ANCA in vasculitis. Nat. Clin. Pract. Rheumatol. 2, 174–175 (2006).

7. Cohen Tervaert, J. W. & Damoiseaux, J. Antineutrophil cytoplasmic autoantibodies: how are they detected and what is their use for diagnosis, classification and follow-up? Clin. Rev. Allergy Immunol. 43, 211–219 (2012).

8. Specks, U. Controversies in ANCA testing. Cleve. Clin. J. Med. 79 (Suppl. 3), S7–S11 (2012).

9. Savige, J. et al. International Consensus Statement on testing and reporting of antineutrophil cytoplasmic antibodies (ANCA). Am. J. Clin. Pathol. 111, 507–513 (1999).

10. Savige, J. et al. Addendum to the International Consensus Statement on testing and reporting of anti-neutrophil cytoplasmatic antibodies. Quality control guidelines, comments and recommendations for testing in other autoimmune diseases. Am. J. Clin. Pathol. 120, 312–314 (2003).

11. Hagen, E. C. et al. Diagnostic value of standardized assays for anti-neutrophil cytoplasmic antibodies in idiopathic systemic

vasculitis. EC/BCR Project for ANCA Assay Standardization. Kidney Int. 53, 743–753 (1998).

12. Csernok, E. & Holle, J. U. Twenty-eight years with antineutrophil cytoplasmic antibodies (ANCA): how to test for ANCA—evidence-based immunology? Autoimmunity Highlights 1, 39–43 (2010).

13. Russell, K. A., Wiegert, E., Schroeder, D. R., Homburger, H. A. & Specks, U. Detection of anti-neutrophil cytoplasmic antibodies under actual clinical testing conditions. Clin. Immunol. 103, 196–203 (2002).

14. Willitzki, A. et al. New platform technology for comprehensive serological diagnostics of autoimmune diseases. Clin. Dev. Immunol. http://dx.doi.org/10.1155/2012/284740 (2012).

15. Boomsma, M. M. et al. Image analysis: a novel approach for the quantification of ANCA levels in patients with Wegener‘s granulomatosis. J. Immunol. Methods 274, 27–35 (2003).

16. Melegari, A. et al. A comparative study on the reliability of an automated system for the evaluation of cell-based indirect immunofluorescence. Autoimmun. Rev. 11, 713–716 (2012).

17. Knütter, I. et al. Automated interpretation of ANCA patterns—a new approach in the serology of ANCA-associated vasculitis. Arthritis Res Ther. 14, R271 (2012).

18. Damoiseaux, J., Mallet, K., Vaessen. M., Austen, J. & Cohen Tervaert, J. W. Automatic reading of ANCA-slides: evaluation of the AKLIDES system. Clin. Dev. Immunol. http://dx.doi.org/10.1155/2012/762874 (2012).

19. Franssen, C. et al. Disease spectrum of patients with antineutrophil cytoplasmic autoantibodies of defined specificity: distinct differences between patients with anti-proteinase 3 and anti-myeloperoxidase autoantibodies. J. Intern. Med. 244, 209–216 (1998).

20. Schönermarck, U., Lamprecht, P., Csernok, E. & Gross, W. L. Prevalence and spectrum of rheumatic diseases associated with proteinase 3-antineutrophil cytoplasmic antibodies (ANCA)

and myeloperoxidase-ANCA. Rheumatology (Oxford) 40, 178–184 (2001).

21. Hauer, H. A. et al. Renal histology in ANCA-associated vasculitis: differences between diagnostic and serologic subgroups. Kidney Int. 61, 80–89 (2002).

22. Lyons, P. A. et al. Genetically distinct subsets within ANCA-associated vasculitis. N. Engl. J. Med. 367, 214–223 (2012).

23. Holle, J. U., Hellmich, B., Backes, M., Gross, W. L. & Csernok, E. Variations in performance characteristics of commercial enzyme immunoassay kits of the detection of antineutrophil cytoplasmatic antibodies: what is the optimal cut-off? Ann. Rheum. Dis. 64. 1773–1779 (2005).

24. Westman, K. et al. Clinical evaluation of a capture-ELISA for PR3-ANCA. Kidney Int. 53, 1230–1236 (1998).

25. Sun, J. et al. Capture-ELISA based on recombinant PR3 is sensitive for PR3-ANCA testing and allows detection of PR3 and PR3-ANCA/PR3 immunocomplexes. J. Immunol. Meth. 211, 111–123 (1998).

26. Csernok, E. et al. Evaluation of capture ELISA for detection of antineutrophil cytoplasmic antibodies against proteinase-3 in Wegener‘s granulomatosis: first results from a multicenter study. Rheumatology (Oxford) 43, 174–180 (2004).

27. Gisslen, K., Wieslander, J., Westberg, K. & Herlitz, H. Relationship between anti-neutrophil cytoplasmatic antibody determined with conventional binding and the capture assay, and long-term clinical course in vasculitis. J. Int. Med. 251, 129–135 (2002).

28. Hellmich, B., Csernok, E., Fredenhagen, G. & Gross, W. L. A novel high sensitivity ELISA for detection of antineutrophil cytoplasm antibodies against proteinase-3. Clin. Exp. Rheumatol. 25 (1 Suppl. 44), S1–S5 (2007).

29. Holle, J. U., Csernok, E., Fredenhagen, G. & Gross, W. L. Clinical evaluation of hsPR3-ANCA ELISA for detection of antineutrophil cytoplasmic antibodies directed against proteinase 3. Ann. Rheum. Dis. 69, 468–469 (2010).

REVIEWS


http://dx.doi.org/10.1155/2012/284740

http://dx.doi.org/10.1155/2012/762874

http://dx.doi.org/10.1155/2012/762874


30. Damoiseaux, J. et al. A novel enzyme-linked immunosorbent assay using a mixture of human native and recombinant proteinase-3 significantly improves the diagnostic potential for antineutrophil cytoplasmic antibody-associated vasculitis. Ann. Rheum. Dis. 68, 228–233 (2009).

31. Roggenbuck, D. et al. High sensitivity detection of autoantibodies against proteinase 3 by a novel third-generation enzyme-linked immunoabsorbent assay. Ann. NY Acad Sci. 1173, 41–46 (2009).

32. Holle, J. U., Hermann K., Gross, W. L. & Csernok, E. Comparative analysis of different commercial ELISA systems for detection of anti-neutrophil cytoplasm antibodies in ANCA-associated vasculitides. Clin. Exp. Rheumatol. 30 (Suppl. 70), S66–S69 (2012).

33. Radice, A., Bianchi, L., Maggiore, U., Vaglio, A. & Sinico, R. A. Comparison of PR3-ANCA specific assay performance for the diagnosis of granulomatosis with polyangiitis (Wegener’s). Clin. Chem. Lab. Med. 51, 2141–2149 (2013).

34. Vermeersch, P. et al. Determination of anti-neutrophil cytoplasmic antibodies in

small vessel vasculitis: comparative analysis of different strategies. Clin. Chim. Acta. 397, 77–81 (2008).

35. Damoiseaux, J. et al. EUROPLUS ANCA BIOCHIP mosaic: PR3 and MPO antigen microdots improve the laboratory diagnostics of ANCA-associated vasculitis. J. Immunol. Meth. 348, 67–73 (2009).

36. Trevisin, M., Pollock, W. & Savige, J. Evaluation of a multiplex flow cytometric immunoassay to detect PR3- and MPO-ANCA in active and treated vasculitis, and in inflammatory bowel disease. J. Immunol. Meth. 336, 104–112 (2008).

37. Finkielman, J. D. et al. ANCA are detectable in nearly all patients with active severe Wegener’s granulomatosis. Am. J. Med. 120, 643.e9–643.e14 (2007).

38. Westman, K. W. et al. Rapid screening assay for anti-GBM antibody and ANCAs; an important tool for the differential diagnosis of pulmonary renal syndromes. Nephrol. Dial. Transplant. 12, 1863–1868 (1997).

39. Tomasson, G. et al. Value of ANCA measurements during remission to predict a relapse of ANCA-associated vasculitis:

a meta-analysis. Rheumatology (Oxford) 51, 100–109 (2012).

40. McLaren, J. S., Stimson, R. H., McRorie, E. R., Coia, J. E. & Luqmani, R. A. The diagnostic value of anti-neutrophil cytoplasmic antibody testing in a routine clinical setting. QJM 94, 615–621 (2001).

41. Mandl, L. A. Using antineutrophil cytoplasmic antibody testing to diagnose vasculitis: can test-ordering guidelines improve diagnostic accuracy? Arch. Intern. Med. 162, 1509–1514 (2002).

42. Sinclair, D., Saas M. & Stevens, J. M. The effect of a symptom related ‘‘gating policy’’ on ANCA requests in routine clinical practice. J. Clin. Pathol. 57, 131–134 (2004).

43. Arnold, D. F. et al. Does a gating policy for ANCA overlook patients with ANCA associated vasculitis? An audit of 263 patients. J. Clin. Pathol. 63, 678–680 (2010).

Author contributionsE.C. researched data for and wrote the article, F.M. made a substantial contribution to discussion of content, and both authors reviewed/edited the manuscript before submission.

FOCUS ON VASCULITIS


Documents

Vasculitis nrrheum.2014.78