Background Antineutrophil cytoplasm antibodies (ANCA) are used as diagnostic markers for small-vessel vasculitis of the Wegener Granulomatosis-microscopic polyangiitis (WG-MPA) spectrum, but if testing is applied indiscriminately, its value is diminished. The authors measured the effect of a targeted ANCA testing policy introduced in our institution in an attempt to improve the diagnostic value of testing in patients with suspected vasculitis.
Methods The authors measured the rate of ANCA requests at a single regional centre in the year prior to and following the introduction of clinical guidelines to ensure appropriate test usage. The authors also audited clinical outcomes in patients in whom ANCA testing was declined.
Result Following implementation of the antineutrophil cytoplasm antibodies (ANCA) gating policy, the number of monthly ANCA tests carried out fell from 287±30 to 143±18 (p<0.0001) and was associated with an increased rate of positivity, from 18.5% (95% CI 17.0 to 20.1%) to 30.3% (27.5 to 33.1%; p<0.0001). The authors undertook a careful review of the case records from 263 patients in whom testing was declined according to the gating policy over an 8-month period. After 6 months' follow-up, no diagnoses of small-vessel vasculitis of the WG-MPA spectrum were reached.
Conclusions The rational use of ANCA testing to aid in the diagnosis of vasculitis should include a clinical gating policy to improve diagnostic performance. Adherence to a gating policy for ANCA testing coupled with close liaison between clinician and laboratory does not result in either a missed or delayed diagnosis of small-vessel vasculitis belonging to the WG-MPA spectrum.
- Small vessel vasculitis
- Wegener's Granulomatosis
- microscopic polyangiitis
- antineutrophil cytoplasm antibody
- clinical audit
Statistics from Altmetric.com
- Small vessel vasculitis
- Wegener's Granulomatosis
- microscopic polyangiitis
- antineutrophil cytoplasm antibody
- clinical audit
Antineutrophil cytoplasm antibodies (ANCA) are well-established markers of the small-vessel vasculitides belonging to the Wegener Granulomatosis (WG)-microscopic polyangiitis (MPA) spectrum.1 2 There is evidence that the introduction of widespread ANCA testing has raised awareness of systemic vasculitis.3 The value of the test as a diagnostic marker for necrotising small-vessel vasculitis is crucially dependent on the pretest probability of vasculitis. We have previously demonstrated the ineffectiveness of ANCA testing when applied indiscriminately,4 demonstrating that 88–100% of cases with a positive ANCA outside a rheumatology department, and excluding renal medicine, did not actually have any form of vasculitis. Diagnostic immunology laboratories continue to face a substantial burden of inappropriate requests leading to the occurrence of clinically insignificant ‘false-positive’ ANCA results. We have recommended local clinical guidelines (see Methods) for ANCA testing, but these have not been validated. In one study from Portsmouth,5 introduction of similar guidelines was tested in a sample of 287 patients, of whom 75 had the request for ANCA testing denied. In one case, a diagnosis of WG was made, but in all other cases, there was no diagnosis of systemic vasculitis after 2 years' follow-up as determined by a review of pathology laboratory records. In the current study, we assessed the impact of a gating policy for ANCA on patients in whom the request was denied, by auditing clinical outcomes in a group of unselected patients in whom requests for ANCA testing was declined. By so doing, it would be possible to determine the risk of overlooking a diagnosis of the WG-MPA spectrum of vasculitides.
Based on our previous audit,4 we introduced clinical guidelines for ANCA testing in a regional immunology laboratory with a catchment population of approximately 630 000. We circulated a letter stating that ANCA testing would only be performed in those clinical situations where there was a significant pretest probability of small-vessel vasculitis, indicating a list of conditions which satisfied this requirement (as shown in box 1); in addition, open access to ANCA testing was granted for the following specialities: rheumatology, nephrology, chest medicine, medical ophthalmology and ENT surgery.
Box 1 Selection criteria for current gating policy, adapted from McLaren et al4
Chronic necrotising large airways disease
Cavitating pulmonary nodules
Suspected Churg–Strauss syndrome
Rapidly progressive glomerulonephritis
Cutaneous vasculitis accompanied by systemic symptoms
We measured the number of ANCA requests ordered in 2005–2006 and 2006–2007, the year before and the year after introduction of the gating policy, respectively. When a request for ANCA testing was declined, the requesting clinician was invited to contact the immunology laboratory to discuss any change in clinical features which warranted ANCA testing.
In order to measure the impact of the gating policy, we reviewed clinical outcomes in a group of unselected patients in whom requests for ANCA testing had been declined, from a series of requests made during an 8-month period in 2007–2008, so that we could ascertain whether or not a diagnosis of the WG-MPA spectrum of vasculitides had been overlooked. All records were reviewed at least 2 months after the date of the ANCA request.
Serum samples for ANCA were assayed by indirect immunofluorescence (IIF) using a screening dilution of 1 in 20 on ethanol-fixed human neutrophil slide preparations (Innova) with characterisation of positive samples using antigen-specific enzyme-linked immunosorbent assays (ELISA) to detect antibodies to proteinase-3 (PR3) and/or myeloperoxidase (MPO) (Phadia Varelisa). The laboratory's performance in ANCA testing by IIF and anti-PR3 and anti-MPO ELISA assays was validated by satisfactory performance in the UK National External Quality Assurance Scheme operated by the Department of Immunology, Sheffield Teaching Hospitals. The hospital records of all patients in whom requests for ANCA were declined due to failure to meet the selection criteria over an 8-month period (2007–2008) were scrutinised to ascertain demographic data, clinical features and final diagnosis.
The data analysis was performed with commercially available software packages (Medcalc version 18.104.22.168, Mariakerke, Belgium). All continuous variables were tested for normal distribution using the D'Agostino–Pearson test and compared with independent t tests. Normally distributed variables are presented as mean±SD. Non-normally distributed data are presented as median and interquartile range (25th and 75th percentiles).
Prior to introduction of the ANCA gating policy, 287±30 ANCA tests were performed per month (2005–2006). In the year after introduction of the policy, this figure had fallen to 143±18 tests/month (p<0.0001). The rate of positivity increased from 18.5% (95% CI 17.0 to 20.1%) to 30.3% (95% CI 27.5 to 33.1%) following introduction of the ANCA-gating policy (p<0.0001).
Of a total of 1248 requests for ANCA testing received during an 8-month period following introduction of the ANCA-gating policy (2007–2008), 263 requests were rejected according to the policy because these requests did not meet the selection criteria. The clinical presentations in these individuals are shown in figure 1. Clinical records were evaluated at a median of 154 days (interquartile range 118–190) from the date of request of the test, thus enabling sufficient time for a diagnosis of vasculitis to be confirmed or refuted. The final diagnoses were: vasculitis associated with cerebral lupus (following rheumatology review) and Henoch–Schonlein purpura in five (2%), rheumatoid arthritis, systemic lupus erythematosus, mixed/undefined connective tissue disease in 16 (6%), inflammatory bowel disease, liver dysfunction and pancreatitis in 28 (11%), non-vasculitic neurological disease in 38 (14%), non-vasculitic renal disease in 16 (6%) and miscellaneous conditions in 130 (49%). In 30 patients (11%), symptoms resolved with no cause identified or diagnosis made.
In 26 of 263 cases in whom requests for ANCA testing were initially rejected, the test was performed, following discussion with the requesting clinician. Twenty-one were negative for ANCA, but five tested positive and were described as typical ANCA (table 1). However, in none of these five cases did ANCA positivity make any difference to the final clinical diagnosis. None of the 26 patients developed small-vessel vasculitis of the WG-MPA spectrum.
We report on the effect of instigating a policy to restrict ANCA testing to those patients in whom there is a significant pretest probability of having a diagnosis of WG-MPA spectrum of vasculitis. There was a dramatic reduction in requests for ANCA being made, from an average of almost 290 per month to around 140 per month, effectively halving the throughput and cost of ANCA testing. This change resulted from voluntary reduction in requests by clinicians following instigation of the new policy, and from scrutiny of clinical request cards by the laboratory staff. The rate of positive ANCA results also increased from 19% to 30% following introduction of the policy, indicating improved efficiency of screening. We have assumed that the number of actual cases of WG and MPA in this cohort of ANCA positive individuals has not changed significantly between 2005 and 2007, but this has not been formally evaluated.
Despite the instigation of a gating policy, we report that there has been minimal impact on patients in whom the test was denied. We suggest that the absence of clinical features suggestive of small-vessel vasculitis can be reliably used to decline ANCA testing without a significant risk of missing a diagnosis of the WG-MPA spectrum of vasculitides. Sinclair et al used a similar symptom-based gating policy to guide ANCA testing.5 Crucially, however, we have validated our testing policy by documenting the absence of either WG or MPA in all patients in whom ANCA testing was declined. Of the 26 patients in whom a request for ANCA testing was initially declined, there was no objective histological evidence of small-vessel vasculitis in the sole patient who was shown to have an elevated anti-MPO (table 1) on a background of confusion, probable Guillain–Barré syndrome and subsequent renal impairment. Given the presence of ANCA activity, including anti-MPO and anti-PR3 reactivity in many preparations of IVIg,6 7 we cannot discount the possibility that this patient's anti-MPO may, in part, at least be due to passive transmission via IVIg. We acknowledge that our study could have been strengthened by analysis of the remaining 237 sera in whom ANCA testing was declined. Unfortunately, this was not possible due to lack of saved serum samples. However, given the absence of clinical features suggestive of vasculitis at presentation and the lack of subsequent progession at notes review at a median of 154 days, we are confident that our testing policy has not overlooked a possible diagnosis of ANCA-associated small vessel vasculitis. Of the five patients who did have clinical diagnoses of vasculitis, two were diagnosed as having cerebral lupus, and three had Henoch–Schonlein purpura.
The criteria in our ANCA testing policy are based on clinical criteria suggested in the international consensus statement for ANCA testing, which have been validated in the clinical setting.8 9 These clinical characteristics comprise glomerulonephritis, pulmonary haemorrhage, pulmonary–renal syndrome, cutaneous vasculitis, multiple lung nodules, chronic destructive disease of the upper airways, longstanding sinusitis or otitis, subglottic stenosis, mononeuritis multiplex and retro-orbital masses. Despite restricting the availability of widespread ANCA testing, we wanted to avoid unnecessary delay in making a diagnosis of WG or MPA. By ensuring close liaison between the laboratory and referring clinician and offering requesters the option of further discussion with the immunology laboratory, we are confident that this testing policy can be reliably implemented in the setting of a large regional immunology laboratory.
Although ANCA assays continue to be performed daily in our laboratory, implementation of this ANCA gating policy has led to significant savings in the consumable costs associated with a halving of the test workload. These savings have to be weighed up against the investment in medical and scientific staff time required to maintain and implement the gating policy.
The implementation of evidence-based selection criteria for ANCA testing has major benefits for the requesting clinician, the immunology laboratory and, most importantly, the patient. The policy minimises the occurrence of false-positive ANCA results with its attendant risk of inappropriate immunosuppressive treatment for patients misdiagnosed as having vasculitis. For the laboratory, it maximises efficiency by preventing unnecessary testing.
The rational use of ANCA testing to aid in the diagnosis of vasculitis should include a clinical gating policy to improve diagnostic performance.
A clinically validated gating policy ensures that ANCA testing is targeted at those patients with a significant pretest probability of ANCA-associated small-vessel vasculitis.
Adherence to a gating policy for ANCA testing coupled with close liaison between clinician and laboratory does not result in either a missed or delayed diagnosis of small-vessel vasculitis belonging to the WG-MPA spectrum.
We thank the staff of the Immunology laboratory, for performance of ANCA tests and help with implementation of the ANCA gating policy, and J Dempsey in the Laboratory IT Department, for extracting workload data on ANCA testing. The department receives support from the NIHR Biomedical Research Centre Programme.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.