Anti-neutrophil cytoplasmic antibody associated vasculitis: a retrospective epidemiological study in an Aotearoa New Zealand cohort

The anti-neutrophil cytoplasmic antibody (ANCA) associated vasculitis (AAV) are a heterogeneous group of diseases characterised by small vessel necrotising vasculitis. Any tissue may be affected, with the respiratory tract and kidneys most commonly involved.¹ AAV are classified into three clinical sub-groups: granulomatosis with polyangiitis (GPA), microscopic polyangiitis (MPA) and eosinophilic GPA (eGPA). Myeloperoxidase (MPO) and proteinase 3 (PR3) antibodies are specific types of ANCA, with PR3-ANCA typically associated with GPA and MPO-ANCA typically associated with MPA and renal disease. While known collectively as AAV, ANCA and the more specific antibodies PR3 and MPO can be negative, especially in eGPA. The clinical presentation of AAV ranges from non-specific constitutional symptoms to fulminant life- or organ-threatening disease. As a result, there can be a delay to diagnosis, and initial misdiagnoses with infection, malignancy or other inflammatory diseases can occur.

The aetiology of AAV is complex, but the involvement of the upper and lower airways suggest that inhaled antigens may have a role. Silica, which is found in concrete, bricks and stones, has been implicated in the aetiology of GPA^2,3 and MPA.³ The increase in severe renal and respiratory manifestations of MPO-positive AAV after the Great Hanshin earthquake of Kobe, Japan in 1995 was suggested to be due to the air pollution associated with the destruction and subsequent reconstruction of the city.⁴ A similar increase was not observed in the first 3 years after the large Ōtautahi Christchurch earthquake in 2011.⁵

Treatment of AAV is described in phases, with an initial remission induction phase followed by maintenance of remission and then long-term follow-up. Remission induction typically involves corticosteroids, which are gradually withdrawn as disease control is achieved. While corticosteroids are rapidly effective, their predictable adverse effect profile is associated with significant morbidity and mortality. Further treatment in the remission induction phase is based on disease severity and involves immunosuppressive therapy such as cyclophosphamide, rituximab, mycophenolate mofetil, methotrexate and azathioprine.⁶ Traditionally, cyclophosphamide has been the therapy of choice for remission induction in individuals with severe life- or organ-threatening disease. Rituximab, a monoclonal antibody directed against CD-20, is an alternative for remission induction.^7–9 In Aotearoa New Zealand, rituximab was first approved for remission induction in AAV in March 2014.

Although the incidence of AAV in the Waitaha Canterbury Region¹⁰ is relatively high, there are limited outcome data for Aotearoa New Zealanders with AAV,¹¹ and none since the introduction of rituximab. The primary aim of this study was to determine the incidence of AAV in the Waitaha Canterbury Region, and to describe the clinical characteristics, time to relapse, rate of remission and survival.

Methods

A retrospective cohort study of individuals diagnosed with AAV between 1 January 2016 and 31 December 2020 was undertaken. Ethical approval was gained from The University of Otago Human Health Ethics Committee (H21/151) and locality approval by Canterbury District Health Board (RO#21230). All ANCA tests reported by Canterbury Health Laboratories, the only laboratory providing ANCA testing in the region, were obtained. ANCA is routinely tested for in all patients where the diagnosis of AAV is under consideration, but it is important to note that there are other causes of a positive ANCA and AAV can rarely be ANCA-negative. For this reason, medical records of all individuals who had an ANCA test performed, irrespective of the result, were reviewed to identify incident cases as defined by the 2022 American College of Rheumatology (ACR)/European Alliance of Associations for Rheumatology (EULAR) classification criteria for GPA,¹² MPA¹³ and eGPA.^13,14 Date of diagnosis was based on the date of clinical confirmation of diagnosis in the medical record rather than date of ANCA testing. Incident cases in the 5-year period were followed until the last clinical encounter, 31 December 2021 or death. This allowed for at least a 1-year follow-up period for all cases.

Clinical characteristics including age at diagnosis, sex, ethnicity, comorbidities and date of death were collected. Clinical features on presentation were recorded. Investigations at diagnosis including full blood count, C-reactive protein, estimated glomerular filtration rate (eGFR), albumin, urine protein creatine ratio and red cells, site and biopsy results and imaging were collected, as well as therapies used for remission induction and maintenance treatment. Date of remission was defined as no detectable disease activity documented in the medical record accompanied by either no change to current treatment plan or de-escalation of therapy. Current incidence was compared with previously published incidence rates in the same region⁵ to determine if there had been any change over time.

Statistical analysis

The annual incidence was calculated by dividing the number of new cases of AAV during the specified period by the size of the Canterbury District Health Board population in 2018 (N=539,631) from Stats NZ. The 95% confidence intervals (CI) for the incidence were calculated using the standard Poisson approximation. The incidence of MPO-positive and PR3-positive AAV was compared to historical data.⁵ Categorical data are presented in absolute numbers and percentages. Continuous variables are presented as mean and standard deviation (SD) or median and interquartile range (IQR). Kaplan–Meier plots for survival within the three disease phenotypes of AAV were estimated using IBM SPSS statistics (version 30.0).

Results

Incidence of AAV

In the 5-year period, 10,439 ANCA tests were requested from 9,319 individuals. Of these individuals, 46/9319 (0.45%) were confirmed incident cases of AAV, giving an average incidence of 1.70 cases per 100,000 per annum over the 5 years (95% CI 1.25–2.27). Of these 46 cases, four (8.7%) were ANCA-negative. Of the remaining 42 cases, 27 (64.3%) were MPO-positive—giving an incidence of 1.0 cases per 100,000 per annum (95% CI 0.66–1.46)—and 15 (35.7%) were PR3-positive, giving an incidence of 0.56 cases per 100,000 per annum (95% CI 0.31–0.92) (Table 1).

View Table 1–3, Figure 1–2.

All 46 cases were subsequently classified according to ACR/EULAR criteria as MPA, GPA or eGPA. Fifteen (32.6%) were GPA, giving an incidence of 0.56 cases per 100,000 per annum (95% CI 0.31–0.92), 24 (51.0%) were MPA, giving an incidence of 0.89 cases per 100,000 per annum (95% CI 0.57–1.32), and seven (14.9%) were eGPA, giving an incidence of 0.26 cases per 100,000 per annum (95% CI 0.10–0.53) (Table 1). There was one case with overlapping features of GPA and MPA. There was no significant increase in incidence over time.

Clinical characteristics

Demographics of the 46 individuals with AAV are outlined in Table 2. The mean (SD) age at diagnosis for any type of AAV was 66.8 (13.3) years. Hypertension was the most common pre-existing comorbidity, and eight individuals had a pre-existing malignancy. For all cases, the median (IQR) time from first patient-reported symptoms to physician diagnosis was 6 (2–17) weeks. Time from first patient-reported symptoms to physician diagnosis was 3 (1–8) weeks for those with MPA, 17 (8–35) weeks for GPA and 4 (2–8) weeks for eGPA. A biopsy was undertaken in 32/47 (68.1%) patients. The most common sites of biopsy were as follows: kidney 23/32 (71.9%), skin 3/32 (9.4%) and ear, nose and throat (ENT) 3/32 (9.4%). Biopsy revealed necrotising vasculitis in 21/32 (65.6%) and pauci-immune necrotising vasculitis in 14/32 (43.8%).

Treatment of AAV

All but one patient received oral glucocorticoid and 12/47 (25.5%) received intravenous glucocorticoid for remission induction (Table 3). Cyclophosphamide was the most common immunosuppressive therapy for remission induction (Table 3). Only three patients received rituximab for remission induction and one patient with MPA received plasmapheresis (PLEX). Azathioprine was the most common immunosuppressive therapy for maintenance of remission, followed by methotrexate and mycophenolate mofetil (Table 3). No patients received intravenous immunoglobulin. At 18 months, 26/37 (70.3%) patients with AAV remained on glucocorticoid therapy (Figure 1A). The mean dose of prednisone declined rapidly over the first 6 months of treatment, and by 12 months the mean (SD) dose was 7.0 (2.9) mg/day (Figure 1B).

Remission and survival

Remission was achieved by 38/46 (82.2%) patients at month 3 and 35/38 (92.1%) patients at month 18 (Figure 1C). During the study period there were a total of 11 deaths (23.4%), of which three occurred within the first 12 months of diagnosis (Figure 2).

Discussion

Herein we have shown that the incidence of AAV in Waitaha Canterbury was 1.70 cases per 100,000 per annum over the 5-year period of 2016–2021 (95% CI 1.25–2.27). Furthermore, the incidence has been stable over the last ~14 years, with a reported incidence of 1.87 (95% CI 1.23–2.72) per 100,000 per annum for the period 2007–2010 and 1.73 (95% CI 1.12–2.55) for the period 2011–2014.⁵ In the current study, 51.1% of cases were classified as MPA, 31.9% as GPA and 14.9% as eGPA, according to the 2022 ACR/EULAR criteria. One-year survival was approximately 94%.

The incidence of AAV observed in this study is generally consistent with other studies. A systematic review of studies that evaluated incidence and prevalence of AAV in people aged >16, which included a meta-analysis, reported an overall pooled incidence of AAV of 1.72 per 100,000 person years (95% CI 1.33–2.16) (n=7 studies).¹⁵ In the current study, annual incidence rates were reported using an antibody-based classification in order to facilitate the examination of temporal trends by allowing comparison with a previous study from this region.¹⁵ The predominance of individuals classified as having MPA in our study is of interest. While geographical variation exists in AAV incidence,^1,16 we would usually expect to see more GPA in countries where the population is predominantly of European ancestry¹⁷ and, in contrast, more MPA in Asian regions.¹⁸ Of note in our previous study, which identified prevalent cases from 1999 to 2003, ~70% were GPA and 30% MPA.¹⁰ Further studies over time would be useful to confirm these findings.

It has previously been suggested that environmental pollution from an earthquake, specifically silica dust from destruction and then reconstruction, might be associated with AAV. This was based on a study of the incidence of MPA, which reported a doubling of MPA after the 1995 Kobe earthquake (mean incidence 1.74/100,000/year [95% CI 0.77–3.96] pre-earthquake and 3.31/100,000/year [95% CI 1.77–6.17] post-earthquake).¹⁹ We previously reported no difference in the incidence of AAV overall, nor in PR3- or MPO-positive AAV in the 3 years before and 3 years after the 2011 Ōtautahi Christchurch earthquake.⁵ The current study, which includes incident AAV cases 5–10 years post-earthquake in the same region, showed a non-significant increase in PR3-positive and decrease in MPO-positive AAV compared to the other two time periods. However, as noted, both are different from the 1999–2003 period where GPA dominated.¹⁰ However, the relatively small numbers of people with AAV likely limits our ability to detect a significant change in incidence.

The high use of oral cyclophosphamide for remission induction in our cohort reflects evidence from the CYCLOPS trial, which reported similar rates of remission rates in those who received oral versus intravenous cyclophosphamide but with fewer relapses, albeit with the potential for more adverse effects.²⁰ Rituximab emerged as an alternative agent for remission induction with the RAVE⁸ and RITUXVAS⁷ trials. In Aotearoa New Zealand, rituximab became available for treatment of AAV in 2014 but is limited to those refractory to 3 months of cyclophosphamide, those where cyclophosphamide is contraindicated, females of child-bearing age and those with a previous history of haemorrhagic cystitis, urological malignancy or haematological malignancy.²¹ Thus, its use is limited, as reflected in the low numbers of individuals receiving rituximab for remission induction or maintenance therapy. It is important to note that in places where access to rituximab is not restricted, clinicians and patients choose rituximab for induction therapy given the efficacy is comparable to cyclophosphamide but the ease of administration and the long-term side effect profile favour the use of rituximab. Furthermore, as the cost of biologic therapies reduces with access to biosimilar agents, there may be a role for widening access in Aotearoa New Zealand for induction therapy in AAV.

Glucocorticoids have a key role in treatment AAV, albeit associated with serious adverse effects with long-term use. In our cohort, the majority of patients remained on glucocorticoids with a mean dose of 14mg daily at month 6, which had reduced to 7mg daily by 12 months. The use of a reduced-dose glucocorticoid regimen in patients with severe AAV was only recently shown to be non-inferior to a standard-dose regimen for the composite outcome of end-stage renal disease or death and was associated with a decreased risk of infection.²² This is reflected in the 2024 EULAR guidelines that recommend tapering the glucocorticoid dose to a target of 5mg prednisone daily within 4–5 months.²³

The survival rates in our study are similar to those in the literature. The overall survival at 2 years in a population-based study from the United States of America was 91% (95% CI 84–99).²⁴ A recent study of the long-term outcomes of 848 people with AAV enrolled in seven clinical trials reported 1- and 5-year survival rates of 88.2% (95% CI 85.8–90.2) and 78.2% (95% CI 75.1–81).²⁵ Similar to our cohort, most individuals in these studies received cyclophosphamide.

There are several limitations to our study. Firstly, the date of diagnosis was based on clinical documentation rather than date of ANCA testing, and the median time from patient-reported symptoms to physician diagnosis was 6 weeks. Secondly, AAV are a rare condition, and the small numbers of incident cases make it difficult to show changes in incidence over time. The main strength of this study is the likely complete capture of people with AAV, given that Canterbury Health Laboratories is the only laboratory testing for ANCA.

In summary, we have shown no clear change in incidence of AAV over time despite the Ōtautahi Christchurch earthquake. Despite the limited availability of rituximab, survival at 5 years is similar to international studies.