Challenges and opportunities from the revised international consensus on ANCA testing

In common with many assays of autoantibodies, patient serum samples tested for antineutrophil cytoplasmic antibodies (ANCA) are highly heterogeneous. Each ANCA assay is an attempt to quantify an array of polyclonal antibodies with multiple avidities and often variable epitope and antigen specificities. Additionally, different testing methods and platforms commonly differ in epitope and antigen expression, and the range of binding avidities detected. The consequences of this are seen in comparative studies.^1,2 For ANCA testing, this variability has been addressed by standardization of test methodology using indirect immunofluorescence (IIF) on ethanol fixed neutrophils, encapsulated in published international consensus statement from 1999.^3,4 The recently published Revised 2017 International Consensus recommends that ‘high quality antigen-specific assays for Proteinase 3 (PR3-ANCA) and myeloperoxidase (MPO-ANCA) antibodies should be used as the primary screening method for ANCA’, replacing IIF as the primary autoantibody test in ANCA-associated vasculitis (AAV). ⁵ Given the good performance of IIF ANCA and the poor standardization of immunometric assays (IMA) for PR3/MPO-ANCA, this is a controversial recommendation. It should be emphasized that the 2017 consensus statement does not claim to present evidence-based guidelines or meta-analysis, and the authors acknowledge that the recommendations require evaluation in prospective studies. When considering ANCA testing options, these limitations must be considered.

Three autoantibody tests are commonly used in relation to AAV: IIF and IMA for PR3-ANCA and MPO-ANCA antibodies. IIF has been the test of choice since the 1999 consensus statement, with PR3/MPO-ANCA reflex testing. ³ Use of IIF according to these recommendations has transformed ANCA testing, and most laboratories using commercial kits are now able to reliably categorize patients as C/P-ANCA positive or negative by IIF. This has been achieved by careful standardization of ANCA IIF testing on a methodology proven in large multicentre clinical validation studies.^6–8

In contrast to IIF, IMA assays for PR3/MPO-ANCA, though much improved since 1999, are poorly standardized. A range of reporting units are used, but even assays reporting in identical units have widely differing cut-offs for positivity, notable differences in reported quantitative values, poor linearity and differences in classification of serum samples as positive or negative.^1,2 Recently established reference materials for MPO-ANCA and PR3-ANCA antibodies (ERM-DA476/IFCC and ERM-DA483/IFCC, respectively) should improve this but, given the variability in patient serum samples and assay design, significant assay differences will remain. There is also no universally agreed definition for a positive test result.

The trial data behind the 2017 statement compared only two IIF methods with eight IMA methods and excluded eosinophilic granulomatous polyangiitis (previously Churg-Strauss syndrome), so are only applicable to granulomatous polyangiitis (GPA, previously Wegener’s granulomatous) and microscopic polyangiitis (MPA). Although recommending use of a high-quality assay, no definition of what this means is provided. The authors also recommend prospective evaluation of their recommendations. ⁵ Any move to adopting use of PR3/MPO-ANCA IMA as primary tests for AAV requires consideration of the risks these issues pose to clinical efficacy in routine practice.

The 2017 statement makes six recommendations, including primary testing using ‘high-quality’ PR3/MPO-ANCA assays (recommendation 2). ⁵ The data supporting the 1999 consensus showed approximately equal numbers of patients with AAV in the IIF +ve/IMA –ve and IIF-ve/IMA + ve groups, supporting the idea that IIF and IMA have similar clinical sensitivities when used as the primary test.^3,8 This was confirmed by data supporting the 2017 statement. ⁹ Despite these good and comparable sensitivities for AAV across most IIF and IMA methods, it is clear from other studies that there is often poor agreement in classifying patients as positive or negative.^1,2 However, a review of the last 12 months of results from UK NEQAS for Autoimmune Serology ANCA shows that, despite the use of a wide range of manufacturers and kits and a wide range of reported quantitative values, there is good agreement in classifying samples as either positive or negative for PR3/MPO-ANCA.

The new recommendations offer comparable and possibly improved sensitivity but come with these significant standardization issues. If considering adoption of PR3/MPO-ANCA as the primary ANCA screening method, these limitations must be balanced against any advantages from using automated IMA. The workflow advantages of automated platforms are well known (reduced turnaround time, easier out-of-hours use, immediately available quantitative results). Ensuring tests used are fit for purpose is an essential function of clinical laboratories and, given the poor standardization of these assays, appropriate local clinical validation of the specific test to be used is needed. Otherwise we risk compromising the efficacy of a well-characterized testing pathway. There is also a danger that automation, by making the tests easy to introduce and perform, results in more inappropriate testing, poorer predictive values and use in centres without the expertise to provide appropriate clinical advice and interpretation.

There are clear challenges but what of the opportunities created by the new consensus statement? An area often neglected in laboratory medicine is the use of stratified quantitative results to give better, result specific, predictive values for test results. This works best for tests designed for a specific disease group, such as ANCA for AAV. Recommendation 6 of the 2017 consensus states ‘taking into account antibody levels improves clinical interpretation’. This is equally valid regarding PR3/MPO-ANCA results performed as reflex test under the 1999 recommendations. However, using PR3/MPO-ANCA as the primary test makes quantitative results available more quickly and consequently stratified predictive values of more immediate use. In my laboratory at Hull & East Yorkshire Hospitals, we have used primary testing with PR3/MPO-ANCA in this way since 2014 and this is the most obvious benefit – the provision of patient-specific predictive values when they are most useful, during the diagnostic work-up. Given the poor standardization of assays, these will not be comparable across assays and will need to be recalculated for each kit, and possibly locally derived because of the effect of patient selection.

There is clear evidence that PR3/MPO-ANCA can perform at least as well as IIF as the primary test in AAV testing for GPA and MPA, but the 2017 statement does not address issues of PR3/MPO-ANCA test standardization. In particular, how to identify a ‘high quality’ test, what constitutes a positive result and how to monitor test performance in order to guarantee continued clinical efficacy. Manufacturers and users must address these issues. These limitations do not negate the recommendations, but the associated risks must be considered in local context before implementation. Clinical validation of each test and testing protocol, monitoring of batch to batch variability and continued vigilance by clinical audit are all required. Careful control and monitoring of these risks are essential if we are to avoid compromising the established and important clinical role of ANCA testing in AVV.