Epidemiology of Anca Associated Vasculitis

ANCA associated vasculitis (AAV) comprises three syndromes with systemic vasculitis (Wegener's granu-lomatosis (WG), Churg Strauss syndrome (CSS) and microscopic polyangiitis (MPA)), which all involve small and medium sized vessels and are associated with antibodies against cytoplasmatic antibodies in neutrophils (ANCA). Polyarteritis nodosa (PAN) is included in this review as it also affects medium sized vessels, and has many clinical findings in common with the AAV. Since the recognition of ANCA, increasing data have become available on the epidemiology of these vasculitidis. WG constitutes half of the AAV and its prevalence has increased from 30/million in the late 1980's in the USA to 160/million in this century in northern Europe. The prevalence for the whole group of primary systemic vasculitides is now 300/million in Sweden. The annual incidence of WG increased from 6.0/million to 14/million during the 1990's in Tromsø, but it is unknown if this is a true increase or the result of an increased awareness of the diagnosis. For the whole group of AAV, the annual incidence in most more recent studies is relatively constant over time and by geographical location, ranging from 13 to 21/million. Nonetheless there are interesting differences in the prevalence of specific vasculitis between different geographical areas, as well as for sub specificities of ANCA. There seems to be a South-North gradient for WG and PR3-ANCA with high figures reported from northern Europe and southern New Zealand. In European studies WG is 90% PR3-ANCA positive. MPA which is predominantly MPO-ANCA associated are more frequent in the Mediterranean countries and also has an increasing gradient towards east-Asia, as almost all AAV in China and Japan are diagnosed as MPA, predominantly MPO-ANCA positive. There are also some ethnic and gender differences. WG is most prevalent among Caucasians in the USA and in people with European ancestors in Paris and in New Zealand, less frequent in Africans and Asians. Several studies have shown that the highest incidence of WG is in males 60-70 years old. Females are younger at onset, and in children WG is most frequent in girls. With better treatment (cyclophosphamide and corticosteroids), the survival of AAV has increased considerably. However disease control is not optimal, as most of the vasculitis present a remitting-relapsing course and organ damage is considerable. Hence, we are still looking to improve treatment regimens.


INTRODUCTION
The term vasculitis designates a group of now more than fifteen different diseases, in which inflammation of the wall of blood vessels occurs in veins, arteries and/or capillaries.Depending on the size and localisation of the affected vessels, patients present with more or less clinically distinct symptoms.While most types of vasculitis were previously considered to be rare, this is no longer the case (1).In addition, vasculitides may cause severe morbidity and increase mortality due to vascular obstruction with tissue ischemia and infarction of several organ systems (2,3).
Within the group of vasculitides, there is a distinct subgroup consisting of Microscopic Polyangiitis (MPA), Churg Strauss syndrome (CSS) and Wegener's granulomatosis (WG).These syndromes are all characterised by the presence of necrotizing lesions of small vessels and the regular presence of antibodies against cytoplasmatic antigens in neutrophils (ANCA) and therefore nowadays grouped together as ANCA associated vasculitis (AAV) (4)(5)(6).Polyarteritis nodosa (PAN) is not included in this group by most authors given its infrequent association with ANCA and its known (causal) relationship with viral hepatitis (7).However, as PAN affects similar vessels as AAV, PAN is often grouped along with the AAV.Increased scientific interest has resulted in enhanced management and has resulted in better outcomes in AAV over the last two decades (2,8).Thus the epidemiology of AAV is of common interest.

CLASSIFICATION CRITERIA
The AAV follow a relapsing-remitting time course and account for considerable morbidity and mortality.
They affect numerous organs, and are a major cause of kidney disease, caused by a necrotizing focal segmental glomerulonephritis, often leading to end stage renal failure, despite the best current treatment strategies.The lack of a unifying etiology on the one hand and the need for an accurate assessment to guide patient management on the other hand, is a constant challenge for clinicians involved in patient care.While it is essential to exclude infectious and other autoimmune diseases that can mimic primary vasculitis, the subsequent classification of the AAV is most often guided by classification criteria proposed by the American College of Rheumatology (ACR) (9)(10)(11)(12) (Table 1).These criteria were meant to help identify patients with specific vasculitic diagnoses within a large cohort of vasculitis patients and were based on expert diagnosis (9).Their main aim was to allow observational and interventional studies in groups of patients with similar disease in different regions.The resulting criteria had 82-88% sensitivity and 92-99% specificity for PAN (3 of 10 criteria), WG (2 of 4 criteria) and CSS (4 of 6 criteria) against the group of other vasculitis patients indicating at least 10% false negative and 5-10% false positive rates associated with the use of these schemes.Subsequent refinements to our thinking about vasculitis were made by an expert consensus meeting, the Chappel Hill Consensus Conference in 1994 that aimed to better define the various types of systemic vasculitis in order to increase our understanding of how various elements may contribute to the specific disease manifestations (13) (Table 2).This fundamentally different approach resulted in the description of MPA as a separate entity from PAN and also established the role for either histological proof or reliable surrogate markers of AAV in clinical practice.
The detection of ANCA has altered the clinical approach to patients with a possible vasculitic syndrome, but is neither incorporated in the ACR criteria nor in the CHCC definitions.This has led a group of doctors interested in epidemiology of vasculitis to develop a stepwise algorithm using both the ACR criteria, the CHCC definitions and ANCA result, as a method to permit epidemiological comparisons without confounding by classification (14).This algorithm, called EMEA algorithm, also defines the surrogate markers for the histological changes used in CHCC definitions (Table 3).

DIAGNOSTIC CONSIDERATIONS
The gold standard for a diagnosis of AAV is histologic evidence of small vessel vasculitis in the context of a clinical constellation with respiratory and renal symptoms together with a positive ANCA result (15).Biopsy findings include leukocyte invasion of the vessel wall, ensuing vessel wall damage and extravasations of red blood cells, fibrinoid necrosis, thrombosis, and sometimes a granulomatous reaction.Importantly, negative biopsy findings do no necessarily exclude AAV given the potential for skip lesions in vessels (16).
The clinical constellation is often that of a patient with a (sub)acute presentation of a systemic inflammatory reaction with constitutional symptoms as fever, fatigue, tiredness and weight-loss together with more specific organ findings such as pulmonary-renal syndrome with nasal discharge, cough, dyspnoea and a uraemic/nephritic syndrome.Clinical findings in the different AAV are depicted in Table 4. Laboratory findings often reflect the degree of systemic inflammation with increased levels of acute phase reactants, the presence of anaemia of chronic disease and more  (13).

WG
Granulomatous inflammation involving the respiratory tract, and necrotizing vasculitis affecting small to mediumsized vessels (e.g.capillaries, venules, arterioles and arteries).Necrotizing glomerulonephritis is common.

MPA
Necrotizing vasculitis, with few or no immune deposits, affecting small vessels (i.e.capillaries, venules or arterioles).Necrotizing arteritis affecting small and medium-sized arteries may be present.Necrotizing glomerulonephritis is very common.Pulmonary capillaritis often occurs.

CSS
Eosinophil-rich and and granulomatous inflammation involving the respiratory tract, and necrotizing vasculitis affecting small to medium-sized vessels, and associated to asthma and eosinophilia.

PAN
Necrotizing inflammation of medium-sized or small arteries without glomerulonephritis or vasculitis in arterioles, capillaries or venules.specific signs of target organ dysfunction such as low levels of oxygenation, increased levels of serumcreatinin and the combined presence of hematuria and proteinuria.Ocular, skin and peripheral nerve affection are also often present (17,18).The detection of a positive result for ANCA by either immunofluorescence or ELISA technique in a patient with systemic disease has a considerable clinical impact as it is often used as a surrogate marker in cases of suspected vasculitis (19).Nonetheless, ANCA testing by itself has only limited sensitivity and specificity for the diagnosis of AAV in a clinical setting (20).

Incidence and prevalence
Classic PAN is the oldest of these systemic vasculitides, and was described by Kussmaul 22) may be the explanation for the observed increase in incidence of WG, also seen in our study from north Norway (23).However, a true increase in incidence has not yet been ruled out.Incidences of WG in the adult population in various areas and time periods are shown in Table 5.More convincing data exist on a true increase in prevalence for the AAV which can be explained both by earlier recognition of the diseases and with earlier and better treatment resulting in increased survival (23,24).Prevalence data on AAV are listed in Table 6.

Geographical variations
WG has the highest annual incidence rate, of about 10 per million in northern Europe, while the incidence is only half of that in southern Europe (Table 5).This has led to a theory of a latitude effect on AAV characteristics, which was recently confirmed in the Southern hemisphere in a study that found WG to be much more prevalent in southern than in northern New Zealand (25).MPA, in contrast, has a higher incidence in southern Europe, and is also found more common in Asia.A recent study from Japan on AAV with renal involvement found a high annual incidence of MPA of 14.8/million, whereas there were no patients with WG (26).Most (91%) patients were MPO-ANCA positive, while none were PR3-ANCA positive.Similar findings was seen in Bejing, China, where most patients with ANCA related disease had MPO-ANCA (213 anti-MPO versus 32 anti-PR3) (27).This is in contrast to Swedish findings of 35% MPO-ANCA positive and 59% PR3-ANCA positive in renal involved WG and MPA patients (87% of WG patients tested positive for PR3-ANCA) (28).
A referral hospital in Mexico City has observed a tripled incidence of WG over the last seven years (29).The Mexican WG patients have similar ANCA pattern as seen in northern Europe and in New Zealand with 80-90% of WG patients being PR3-ANCA positive and 10-15% MPO-ANCA positive.For MPA the reverse is found, 50-90% are MPO-ANCA positive while 5-40% are PR3-ANCA positive (28)(29)(30).Also in India there seems to be a preponderance of WG (54%) compared to MPA (18%), however the frequency of PR3-and MPO-ANCA are more equal (48% and 43% respectively) (31).These findings all indicate latitude and east-west related difference both in the frequency of various vasculitides, but also in the ANCA pattern for different geographic areas.

ANCA and ethnicity
The reason for the different specificities of ANCA is unclear, but genetic variations may be important.It has been shown that WG, associated with PR3-ANCA, is more prevalent among Caucasian than black Americans (32).In New Zealand WG is twice as common among people of European ancestry compared to New Zealand Maori or Asian, which again has twice the prevalence of people identifying themselves as Pacific Islanders (25).In a multiethnic urban area in France, the prevalence of primary systemic vasculitis (AAV and PAN) was twice as high in individuals of European descent (107.4/million) as in non-European (52.5/million) (33).
* Time given for point prevalence (pp) or period prevalence (n-year prevalence).
** The EMEA algoritm is used to for classification, including ANCA and surrogate markers for granulomatous involvement in respiratory tract and for glomerulonephritis (14).

Genetic contributions
Despite uncertainty about its aetiopathogenesis, AAV are considered to be autoimmune diseases where a genetic inclination to autoimmune reactions can be triggered by environmental factors.Although case reports have illustrated that familial clustering does occur, there have so far been no systemic reporting on familial aggregation in AAV (34).Genome-wide microarray scanning studies have not been reported, although single centre studies have reported on candidate genes in AAV.Alpha-1 anti-trypsin (AAT) is a physiological inhibitor of proteinase-3 (PR3) which is one of the antigens targeted by c-ANCA.Carriage of a defective allele in the coding region for AAT at the protease inhibitor locus (Pi) was the first reported genetic risk factor for the development of anti-PR-3 positive AAV (35) and may lead to increased PR-3 expression and increased risk for relapse in WG patients, while PR-3 gene polymorphism does not seem to contribute to PR-3 expression (36).Other association studies have found no or only weak associations between AAV and polymorphic cytokine, HLA, and Fcgamma receptor genes (37).

Age and gender
Mean age at onset of WG is 50 years in most studies and was stable over 15 years in our study from northern Norway (23,33,38,39).However, in Germany, age at onset has increased with 15 years from 40-49 years (1966-1993) to 60 years (1998-2002) (40).MPA patients is older at diagnosis compared to WG patients.
In European studies MPA is diagnosed at 60 years (33,38), while in New Zealand and in Japan the MPA patients are 70 years at diagnosis (26).There are less epidemiological data on CSS and PAN, but mean age at onset for CSS is about 50 years, while onset of PAN is at age 45-48 years (33,38).
In both WG and MPA the age and gender-specific incidence rate show increase with age, most obvious in males.There is an overall peak in the 65-74 year age group in Europe, in the 70-74 year age group of MPA and RLV in Japan, and 70-79 year group of WG in New Zealand (25,26,30,41).In the study from northern Norway we found a clear age-specific incidence top only in males (65-74 years).The less prominent peak incidence in females occurred at younger age (45-54 years), and moreover, in the whole cohort of WG patients with onset of disease from 10-84 years of age, all patients with childhood onset were girls (23).Also other studies in children report more WG in girls than in boys (42).
In the adult population the male to female ratio of 1.3-1.6 in WG (23,33,38).In studies from Paris and Sweden there is an opposite male-female ratio of 0.5-0.9 in MPA, while the data on PAN and CSS are more conflicting, with a male-female ratio of 1.5-0.3 and 1.7-1.0,respectively (33,38).

Annual and seasonal variations
The etiology of the AAV and PAN is unknown except for the association of PAN with Hepatitis B virus infection (43).The onset of WG, very often involving the upper airways has led to the hypothesis that inhaled agents or infections could be triggering factors.Some studies have showed a seasonal variation, supporting an infectious etiology (44,45).However there are now conflicting results concerning seasonal variation in onset of AAV (23,32,46,47).Also an annual variation in onset has been shown in WG (23), as in giant cell arteritis (48), but this phenomenon has to be confirmed in more studies.

Environmental factors
The geographical variations in AAV and the possible seasonal and annual variations in onset may reflect environmental influence.A number of factors have been reported to be associated with the development of vasculitis, including various infections, silica, hydrocarbons, inhaled fumes and particulates, drug, allergy, vaccination and farming.

Infections
Infection has been postulated as the cause of AAV vasculitis, for WG, since the first description of the disease, yet no proof exists.However, classic PAN is closely related to Hepatitis B virus (HBV) and cryoglobulinaemia to Hepatitis C virus (49,50).A number of other viruses have been related to vasculitis.Cytomegalovirus has been reported to mimic WG, and parvovirus has been associated with WG and PAN (31).Also bacteria have been proposed as triggering factors for vasculitis, especially is staphylococcus aureus associated to disease relapse in WG (51).

Silica
AAV have been associated with exposure to particulate silica, like quartz, granite, sandstone or grain dust.The high levels of air pollution and silica dust after the earthquake in Kobe, Japan, in 1995, and the subsequent increase in MPO-ANCA-associated vasculitis, suggested a causal relationship (52).Six case-control studies have found significant associations between AAV and silica exposure, but with some conflicting results for the specific diagnoses of WG and MPA (53)(54)(55)(56)(57).The most recent study from the USA found significant association only with higher lifetime exposure to silica, and indicated that crop harvesting might imply special risk (58).

Farming
Farming in the year prior to onset of AAV was associated with an increased risk of developing WG and MPA (OR 2.7 and 6.3 respectively), but not CSS in a study from the UK (55).The study could not distinguish between crops and livestock, but the association appeared to be stronger for livestock.An earlier study from the USA failed to find association between WG and farming (46).

Inhaled fumes and hydrocarbons
A case-control study from the USA reported association of inhaled fumes and particulate materials, with WG compared to healthy and rheumatic disease controls, but not to respiratory disease controls.For pesticides there was significant difference also to respiratory disease controls (46).Conflicting results exist for exposure to metals and welding fumes (54,55), as for exposure to occupational hydrocarbons and solvents (55,59).

Drugs, allergy, vaccinations
A large number of drugs have been associated with vasculitis, most often hypersensitivity vasculitis, but propylthiouracil and hydralazine have also been reported in association with AAV.In a review of 250 MPO-positive systemic vasculitis patients, 30 patients with the highest MPO-ANCA titer were examined for exposure to hydralazine, propylthiouracil, allopurinol, penicillamine or sulphasalazine.Among these high titered AAV patients 60% had used one or two of these drugs within 9 months prior to onset of disease (60).However, a possible association was not confirmed in another study of 586 patients treated with propylthiouracil, as only one case of vasculitis was found (61).
Leukotriene receptor antagonists have been linked to onset of CSS in several patients.However, recent studies raises some doubt about the relation, as it seems that patients experience a flare in their asthma requiring intensifying treatment, including leukotriene receptor antagonists prior to onset of the vasculitis (62,63).
Two case-control studies have shown significantly more allergy during the last year prior to onset of primary systemic vasculitis than in healthy controls.This also holds true for the specific diagnosis of WG, but with conflicting results for MPA and CSS (55,64).Autoimmune reactions, including AAV have been reported to follow vaccinations (65), but more data is needed to state a clear relation.

Mortality
The prognosis for AAV has changed dramatically during the last 50 years, most notable for WG.In the 1950s the mean survival of untreated WG was 5 months with less than 10% surviving 2 years (66).After the introduction of cyclophosphamide and corticosteroids as standard treatment, patients now experience a one year survival of 93% and a 5-year and 10-year survival of 79% and 75% respectively (24).Still the mortality risk ratio is 4 compared to the general population (67), and 5-10% of WG patients die in active disease during the first 3 months after diagnosis (24).
The increased mortality risk is associated with higher age; a rise of each decade in age increasing the risk of death in WG patients with a hazard ratio of 2.18 (24).This is illustrated with the increased mortality in patients aged > 50 years from northern Norway (Figure 1) and is confirmed in other studies (68,69).Also renal involvement at diagnosis (68), especially severe renal impairment in a degree demanding dialysis, is predictor of reduced survival (24).Patients who had developed permanent organ damage at diagnosis, were also at increased risk of death in our study from northern Norway.The 5-year survival rate in MPA is between 45% and 75% which is worse than in WG (2,70,71).Patient survival in CSS is 60-97% at 5 years (72,73) depending on the presence at diagnosis of one or more of the five factors related to worse outcome: proteinuria > 1 g/day, creatinine > 1.58 mg/dl, gastrointestinal involvement, cardiomyopathy or neurological involvement (74).This five factor score has been shown to predict reduced survival also in MPA and PAN, though PAN (without HBV) has the best survival of these diagnoses, with a 10-year survival rate close to 80% (75).

Relapse
Up to 90% of patients with AAV go into remission during the first 3-6 months of treatment.Most AAV however, shows a chronic relapsing course, which demands close follow-up and long-term therapy with immunosuppressive therapy.
The relapse rate is highest in WG, with a cumulative relapse rate of more than 60% depending on time of follow-up and the intensity of maintenance treatment (18,73).The relapse rate is increased in PR3-ANCA positive patients (76), and a rise in cANCA or PR3-ANCA titer predicts relapse (77).The relapse rate in MPA is lower than in WG (78).In a French study following patients for more than 7 years, the relapse rate in MPA was 35% and in CSS 20%.PAN revealed the lowest relapse rate, with 20% in non-HBV associated PAN and 8% in the HBV associated PAN (75).
In WG patients from northern Norway we found that less intensive treatment during the first 6 months was related to increased risk of relapse.A cumulative cyclophosphamide dose of < 10 g, and high dose prednisolone (≥ 20mg/day) for less than 2.75 months, increased the hazard ratio for relapse with 2.8 and 2.4 respectively (18).
Chronic nasal carriage of staphylococcus aureus is found in many patients with WG, and is related to increased risk of relapse (79).Adjuvant remission maintenance treatment with trimethoprim-sulfamethooxazole reduced the relapse rate in one study (76).

Organ damage
Organ damage occurs early in AAV, best studied in WG (24,80).Damage can be studied using the vasculitis damage index (VDI) (80).Using the VDI, we showed that the increase in damage was 6 times higher during the first 6 months after diagnosis compared to the next 18 months.This early damage was mainly due to disease activity (93%), while treatment related damage increased later in the disease course (18).The early damage was related to disease activity at onset of therapy, and could in our study be reduced by longer time on cyclophosphamide during the first 6 months.Treatment protocols using pulse cyclophosphamide resulted in less late damage compared to treatment with daily oral cyclophosphamide (24).
Nasal and sinus dysfunction is the most frequent damage in WG (61% in our study), hearing loss is seen in about a third of WG patients, visual loss in 8-9% and saddle nose in 16% (24,81).Peripheral neuropathy is frequent (43%), but complaint on pulmonary symptoms was only reported by 24% of patients after 5 years, although 80% of patients had previous lung involvement (82).End stage renal disease (ESRD) occurs in about 20% of patients with AAV (24,83).Reduced renal survival was related to severe renal involvement in terms of dialysis dependence at start of treatment in our study (24).The relationship between ESRD and initial serum creatinine and dialysis dependency was also seen in another Norwegian study (67).

Malignancy
The increased life expectancy in AAV is due to more rapid recognition and aggressive treatment of the disease.In most patients this will include the use of cytotoxic drugs, in particular the use of oral or intravenous cyclophosphamide, that exerts its action by forming crosslinks between and within DNA strands and ultimately lead to cell death.While most active metabolites are oxidized, a portion is converted into acrolein that is a known bladder toxicant.Not surprisingly, increased longevity for WG is thus associated with an increased risk of late onset cancer, which mainly can be attributed to the use of cyclophosphamide (84,85).Cancer risk is not increased in patients that never receive cyclophosphamide, while high cumulative doses of cyclophosphamide put AAV patients at risk for hematological, bladder and skin malignancies after 7-18 years of disease (84).The known carcinogenic effects of acrolein on bladder epithelium are mainly a problem in patients treated with oral cyclophosphamide (8,24,86).

CONCLUSION
With a population prevalence of vasculitides of 300/ million, they represent big challenges not only to the treating physician, but also to the total health care system.Remission rates and survival in AAV have increased during the last 50 years, but disease relapses, organ damage and late malignancy still represents big challenges, and the optimal treatment is yet to be found.

Figure 1 .
Figure 1.Kaplan Meier survival plots for 79 WG patients with a mean follow up of 76 months, from northern Norway (unpublished).
Surrogate markers for renal vasculitis (glomerulonephritis)*Haematuria associated with red cell casts or >10% dysmorphic erythrocytes 2+ haematuria and 2+ proteinuria on urinalysis * Only one surrogate marker is necessary to support the diagnosis.In all cases other causes must be excluded.

Table 4 .
Clinical findings in ANCA associated vasculitis.

Table 5 .
Incidence of ANCA associated vasculitis and PAN in the adult population.