Evidence‐based treatment of systemic vasculitis

Vasculitis/Series Editor: R. Watts The systemic vasculitides are potentially fatal if untreated and immunosuppressive therapy now saves lives and salvages organ function. Treatment has changed the outcome of vasculitis to that of a chronic disorder with accumulating morbidity and incapacity. Current treatment is toxic and contributes to morbidity and mortality. Balancing the dangers of disease against those of treatment requires detailed knowledge of both, knowledge that has been confused by differences in classification or is lacking due to a paucity of long‐term outcome studies. There have been relatively few randomized‐controlled trials in vasculitis, although their number is increasing. Much of the evidence supporting therapeutic decisions derives from small prospective studies or larger, usually single centre, retrospective experiences. More recently, consensus discussions have contributed to evidence on classification and existing treatment practice. This review will focus on the primary systemic vasculitides which are distinct from those related to infection, drugs, cancer or other systemic disorders, such as rheumatoid arthritis. While international consensus on disease definitions is not complete, most current investigators use the system derived from the Chapel Hill Consensus conference in 1992 [1]. Ongoing modifications to include antineutrophil cytoplasm autoantibodies (ANCA) in the classification system will group together those conditions with a pauci‐immune vasculitis predominantly affecting ‘microscopic’ blood vessels (Table 11) [2]. The term, ‘ANCA‐associated vasculitis’ has been applied to this subgroup although not all patients are ANCA positive at diagnosis; in particular, the frequency of ANCA positivity is lower in milder and organ‐limited presentations [3, 4]. Certain other problems with classification remain which complicate examination of the existing literature [5]. ‘Polyarteritis’ or ‘polyarteritis nodosa’ has been used to describe both polyarteritis nodosa and microscopic polyangiitis; following Chapel Hill the term polyarteritis nodosa is now restricted to disease where microscopic vessel involvement is absent [6]. The term ‘Wegener's granulomatosis’ is used by some writers when there is upper or lower respiratory tract involvement without the demonstration of destructive ear, nose and throat (ENT) lesions, lung cavities or granulomata in tissue biopsies, which would be regarded by others as microscopic polyangiitis. Lastly, renal vasculitis is now frequently used to describe patients with a crescentic necrotizing glomerulonephritis with few immune deposits, which has been previously termed ‘idiopathic rapidly progressive glomerulonephritis’ and can be regarded as a ‘forme‐fruste’ of primary systemic vasculitis [2]. Thus, classification remains based on clinical and pathological features at presentation and not on aetiology and there are inevitable overlaps between categories. Classification of primary systemic vasculitis according to ANCA positivity and predominant size of vessel involved Classification of primary systemic vasculitis according to ANCA positivity and predominant size of vessel involved Patients with more severe disease at presentation have a poorer outcome and therapeutic regimens have evolved to treat such presentations more aggressively, with, in consequence, more adverse effects. An international consensus group has formalized disease severity at presentation with the aim of designing regimens appropriate to each subgrouping (Table 2) [7, 8]. Such a system of subgrouping remains empirical and awaits the results of further studies to confirm its value. In a review of 342 patients with polyarteritis and Churg–Strauss angiitis, multiple potential prognostic factors were measured against outcome and vital organ involvement of kidneys, heart, gut, brain or lung were predictive and have been used to develop a simple ‘five factor score’ [9]. The same study group now classifies patients at diagnosis into ‘good’ and ‘poor’ prognostic groups and varies the intensity of treatment accordingly [10]. Other prognostic factors which have been identified are an increase in relapse rate with ANCA directed to proteinase 3 (PR3–ANCA), a diagnosis of Wegener's granulomatosis, the number of systems involved, race and serum creatinine level at diagnosis [11–14]. Increased age at diagnosis is also an important outcome predictor, with a high mortality in part due to infective complications of immunosuppression [15]. Further studies are required to identify prognostic markers, which will allow sophistication of subgrouping at diagnosis and better‐targeted therapy. Clinical subgrouping according to disease severity at presentation for ANCA‐associated vasculitis [7]. ANCA is negative in a minority of generalized and severe renal presentations. Refractory disease implies progressive disease despite at least 6 weeks of treatment with an appropriate regimen; ANCA may become negative with treatment Clinical subgrouping according to disease severity at presentation for ANCA‐associated vasculitis [7]. ANCA is negative in a minority of generalized and severe renal presentations. Refractory disease implies progressive disease despite at least 6 weeks of treatment with an appropriate regimen; ANCA may become negative with treatment The outcome of vasculitis has been assessed by death, development of end‐stage renal failure, disease relapse and acquisition of irreversible damage [8]. Therapeutic trials often use remission as an end‐point, but this term has not been standardized. Prior to the introduction of steroids, mortality of vasculitis with vital organ involvement was over 75%; an early Medical Research Council trial of cortisone found reduced mortality of polyarteritis at 1 yr, but after 3 yr there was no difference between treated and untreated patients due to steroid‐associated mortality [16]. Leib et al. [17] found a reduction in 5‐yr mortality to 50% with steroids in polyarteritis nodosa and a further reduction with the use of cytotoxics to 12%. Although cytotoxic drugs had been used in various forms since the mid‐1950s, the reports by Fauci et al. [18, 19] from the National Institutes of Health confirmed the position of cyclophosphamide in induction regimens for systemic vasculitis with a remission rate of 93% in 85 patients with Wegener's granulomatosis. Retrospective studies from single centres have reported survival rates with immunosuppressive therapy varying between 75% at 12 months and 87% at 8 yr with differences in disease presentations and possibly, therapeutic protocols accounting for much of the difference [20]. Recent studies of ANCA‐associated vasculitis point to a 2‐yr survival of approximately 79%, with 21% of initial diagnoses progressing to end‐stage renal failure; thus, at present, over 40% have a poor outcome by these two simple definitions (Table 3). Age and creatinine at presentation have been consistently associated with survival [11, 15, 23]. Similarly, the development of end‐stage renal failure is closely linked to creatinine; it occurs in around 20% of patients, and is also associated with features on renal biopsy [11, 23]. Relapse rates have varied between 11 and 60% and are influenced by disease subgroup and treatment; long‐term immunosuppressive therapy to prevent relapse contributes to late damage, and therapeutic regimens designed to induce disease remission also need to be judged by their subsequent disease relapse rates [13, 20, 23, 25, 26]. Inclusive scores to record accumulating ‘all cause’ damage have only recently been developed and are likely to play an increasing role in evaluating the medium‐ to long‐term efficacy of treatments in the future [8]. Outcome according to death and end‐stage renal failure for ANCA‐associated vasculitis. (Jayne) refers to unpublished data Outcome according to death and end‐stage renal failure for ANCA‐associated vasculitis. (Jayne) refers to unpublished data This subgrouping also includes ‘limited Wegener's granulomatosis’, patients with organ involvement confined to the respiratory tract [7]. Historical protocols have used the combination of cyclophosphamide and steroids, but several recent studies have substituted methotrexate for cyclophosphamide (Table 4) [27–30]. Remission rates of 60–70% have been achieved, when used as a remission‐maintaining agent after cyclophosphamide; relapse rates are low. Sneller et al. [27] and Stone et al. [30] observed higher relapse rates, possibly related to no previous cyclophosphamide exposure. Although most relapses were responsive to increases in the methotrexate and/or steroid doses, an inability to reduce the steroid dose and relapsing disease was predictive of more widespread vasculitis [29]. Adverse effects related to methotrexate included pneumonitis, although this can be difficult to diagnose in a patient with pulmonary Wegener's granulomatosis, hepatotoxicity and myelosuppression, but these events were reversible. The control of renal vasculitis with methotrexate is more controversial; patients with elevated serum creatinine have not typically been treated with methotrexate. In one series, those with presumed renal vasculitis had stabilization of excretory function; in another study, this subgroup were more likely to have refractory, progressive disease [27, 29]. The route of administration of methotrexate has varied between studies, oral or subcutaneous, which may influence tolerability and efficacy. Methotrexate is therefore an alternative component of initial therapy to cyclophosphamide for a patient with systemic vasculitis without threatened vital organ function. A proportion of patients will develop progressive disease and require conversion to cyclophosphamide and the effect of methotrexate on the control of relapse over longer time periods is unknown. None of the above studies have compared methotrexate to cyclophosphamide or studied microscopic polyangiitis, and this subject forms the focus of a current international trial (NORAM) [7]. A comparison of daily oral to pulse intravenous cyclophosphamide in polyarteritis nodosa and Churg–Strauss angiitis in a good prognosis subgrouping according to the five factor score, similar to early systemic disease, found similar remission rates of 9/12 and 10/13, respectively, which are comparable to those obtained in Wegener's granulomatosis with methotrexate [10]. Side‐effects appeared more frequently, 41 events in 25 patients [10]. Therapeutic trials of methotrexate for Wegener's granulomatosis Therapeutic trials of methotrexate for Wegener's granulomatosis The empirical introduction of daily oral cyclophosphamide became popular during the 1970s but was only recently subjected to randomized trials. In a study of 71 patients with polyarteritis nodosa and Churg–Strauss angiitis who also received steroids and plasma exchange, cyclophosphamide led to improved disease control and fewer relapses [31]. The role of cyclophosphamide has been further confirmed by a meta‐analysis of three trials, including 182 cases, from the same study group [32]. Retrospective data from the North Carolina glomerulonephritis study group compared patients with microscopic polyangiitis treated with cyclophosphamide with those treated with steroids alone and found improved renal survival and a lower relapse rate in those receiving cyclophosphamide [11]. In a consensus statement by the European Vasculitis Study Group, the combination of cyclophosphamide for 1 yr and a tapering dose of prednisolone was regarded as the standard treatment for this subgroup [7]. Efforts have focused on minimizing cyclophosphamide exposure by using pulse rather than continuous administration (see below) or switching to an alternative drug once remission has been obtained. The efficacy of azathioprine for remission maintenance in vasculitis has been previously reported with relapse rates of 11–30%; smaller studies have used azathioprine in remission induction protocols [13, 17, 26, 33]. A large, international, randomized trial comparing azathioprine with continued cyclophosphamide for prevention of relapse, has recently been reported (CYCAZAREM) [34]. One hundred and fifty‐five patients were studied, and following induction of remission with oral cyclophosphamide and steroids for 3–6 months, they were randomized to continuing a lower dose of oral cyclophosphamide to 12 months or switching to oral azathioprine [34]. There was no difference in relapse rates, 17%, up to the end of the study, 18 months from treatment onset, and there was a trend to fewer serious adverse events in the azathioprine limb (Fig. 1). A surprising result of this study was the high remission rate with oral cyclophosphamide and prednisolone in this subgroup. Apart from the withdrawal of 10 patients prior to the start of the remission phase, largely due to death or treatment intolerance, all patients entered clinical remission. Thus, the ‘standard’ induction treatment with oral steroids and cyclophosphamide appears effective, but toxicity was high with over 160 adverse events reported and a serious or life‐threatening adverse event rate of 26%. Reversible leucopaenia was the most common adverse effect, azathioprine was reported in and has features similar to a relapse of vasculitis which has in diagnosis of severe and life‐threatening adverse effects in the trial [34]. remission‐maintaining drugs include for which a randomized trial found a higher relapse rate in the patients to as compared with continued but patients good control One study including patients with ANCA‐associated vasculitis also found good remission control in all with and prednisolone is an immunosuppressive which and is to azathioprine in the prevention of organ with A prospective study with a similar induction to the trial (see used in of azathioprine and one relapse was observed in 12 patients studied A for plasma in vasculitis was based on a presumed immune aetiology immune in most primary systemic vasculitides are typically or Further for plasma from its in disease where the renal is similar to that in renal vasculitis from the of immune The of plasma to steroids alone was assessed in patients with polyarteritis nodosa and Churg–Strauss difference in therapeutic efficacy was observed A further study from the same group using a similar patient subgroup compared the of plasma to steroids and cyclophosphamide in patients and found no therapeutic effect of plasma The of ANCA and for the of these autoantibodies has now a for plasma in ANCA‐associated In a randomized study of patients with Wegener's granulomatosis of varying plasma appeared to but this study is to be reported The of plasma to immunosuppression for rapidly progressive glomerulonephritis was not found to those with a creatinine level in a randomized study which according to renal A of this study was that this subgroup had a good outcome and plasma was more likely to those with renal failure prospective studies of pulse intravenous cyclophosphamide this of administration be to daily oral for the induction of remission not more in this those with more organ involvement and high ANCA had a poor therapeutic In continuing disease despite intravenous cyclophosphamide has to conversion to a daily oral randomized trials have the as to cyclophosphamide is and as as daily oral administration for the induction of remission (Table None was to efficacy in vasculitis, although one study a higher relapse rate after intravenous pulse use the studies that adverse effects were more in continuous oral cyclophosphamide although this was only the primary in the study by et al. In this study, azathioprine was substituted for cyclophosphamide once remission had been and the difference in adverse effects events was not The studies by et al. and et al. were both early due to more adverse effects in the continuous The high number of adverse events has been associated with the steroid dose used in these trials and with the cyclophosphamide tapering in the et al. study, which may have the of Further studies are required can be between the two of cyclophosphamide administration the demonstration that of cyclophosphamide are effective, some of the over cyclophosphamide exposure will be reduced [34]. of intravenous cyclophosphamide in primary systemic vasculitis. polyarteritis Churg–Strauss microscopic polyangiitis; Wegener's granulomatosis. of intravenous cyclophosphamide in primary systemic vasculitis. polyarteritis Churg–Strauss microscopic polyangiitis; Wegener's granulomatosis. patients with vasculitis in renal failure with an for to renal is likely to damage and this is by the that the level of serum is related to long‐term renal survival Such have received more therapy with steroids or plasma to to renal function. have been reported to increase the of renal but only plasma has been subjected to In the randomized study by et al. those with a creatinine over or on appeared to from plasma with renal as compared with not treated with plasma Other studies have included a of but when patients with renal failure are there appears to be an improved outcome with plasma (Table Other for the effect of plasma include the of and and with has been used in the for this The European Vasculitis Study has regarded both prednisolone and plasma as used to oral steroids and cyclophosphamide for this and is comparing the two in a current study [7]. ANCA‐associated vasculitis is the most of pulmonary usually in the of the pulmonary renal and this presentation a high mortality. prospective therapeutic studies are although both plasma and prednisolone have been used in retrospective reports survival in randomized trials including patients with renal failure due to renal vasculitis. Treatment with or without plasma (Jayne) refers to unpublished data survival in randomized trials including patients with renal failure due to renal vasculitis. Treatment with or without plasma (Jayne) refers to unpublished data alternative treatments have been studied for progressive disease to standard therapy or where standard therapy is not using or with or without has led to remission in small some of these have been While the role of this awaits studies, these results have the of the to the of vasculitis and they are the drugs since cyclophosphamide to induce remission of intravenous has several of of potential in it which has effects at both and and with factors such as and randomized trials have confirmed a for in the treatment of disease, a vasculitis of related to exposure. The dose for disease is over and this has been used in to cytotoxics and steroids for ANCA‐associated vasculitis prospective studies in ANCA‐associated vasculitis have found treatment in of patients after of the of is by continuing immunosuppressive when as therapy to patients, entered remission for up to 1 yr a trial of relapsing or ANCA‐associated vasculitis found improved disease control in patients This effect was only observed up to 3 months, that possibly at would be for a effect Treatment have also been after in other forms of vasculitis, including polyarteritis nodosa and Churg–Strauss angiitis therapeutic may be to the clinical to in This systemic vasculitis is by of and of immune increases the of immune and of Adverse effects of therapy are and include and renal failure, as as vasculitis Vasculitis may as a of infection, such as and can relapse of primary vasculitis of the upper respiratory tract by in Wegener's granulomatosis increases the of disease relapse, and this has to the of an aetiology for this vasculitis, by therapy with in Wegener's granulomatosis reduced the of respiratory tract relapse in a study when to immunosuppression The study was yr and the difference in relapse frequency was in the 6 a high rate of drug was observed was in disease in Wegener's granulomatosis when used in of immunosuppression at various of disease in a prospective study of patients factors have an alternative using of the which is clinical An between ANCA and relapse but the of this and the role in ANCA play in treatment remain patients are at higher of relapse than those with and with largely renal vasculitis patients had lower relapse rates than those where Wegener's granulomatosis with respiratory tract involvement An early study in Wegener's granulomatosis during remission randomized patients ANCA to treatment or no in therapy relapse was in the group and was not in the group treated on the of the ANCA the exposure to immunosuppression was lower in the early group Other studies have consistently reported a high frequency of ANCA positivity at the time of relapse, and an relapse in those with ANCA positivity during remission or in those ANCA positive during remission [13, 26]. While further studies based on ANCA and are a common to the of relapse is to reduce the between in to diagnose relapse as early as the rate and are of in but prognostic for vasculitis relapse Other drugs for which have been include in Wegener's and for Churg–Strauss angiitis developed for other clinical such as and the factor and are also in vasculitis. was not to be an treatment in vasculitis, although have of plasma including with or to ANCA without of plasma and of comparable efficacy to plasma The of ANCA will the development of which has the of factors without or using cytotoxic by has led to remission in a few of vasculitis has also vasculitis in a The further development of this in vasculitis on the of the and the of patients organ damage has The toxicity of treatment contributes to the chronic morbidity and mortality of vasculitis. The National Institutes of Health with Wegener's granulomatosis reported a of treatment toxicity to damage in over 50% of their patients [20]. The trial an adverse effect frequency of patient with severe or life‐threatening adverse effects the 18 months [34]. adverse effects are the most common of death or severe morbidity and their frequency is associated with age and steroid [15]. rates of up to 20% have been found which have for with in centres where this is common toxicity of cyclophosphamide is to and In the to be studied to developed and cancer [20]. patients were over a time when daily oral cyclophosphamide was standard therapy. The frequency of was related to the or dose of cyclophosphamide with a 50% rate after months or None of the patients without developed is the in cancer with longer which was in this study to be at 10 yr and at yr is in pulse patients, reported in only one from the studies (Table A study found an increase in cancer rates in patients receiving oral cyclophosphamide for more than 1 yr, and an increase in related to azathioprine and steroid exposure failure is associated with the cyclophosphamide dose and is therefore likely to be more in daily oral This toxicity has been assessed in when comparable oral regimens led to in and reduced rates The was in that were higher with oral from the use of cyclophosphamide in with have that the of is related to age and of treatment with an of in those 25 yr and 60% in those over yr from and for this in studies A high of steroid related irreversible damage has been reported in retrospective [20]. disease is common due to the high exposure and the age of the patient effect with was found in a relatively small randomized trial Consensus have and a recent randomized trial found a effect of on Patients with chronic disease have an of disease which is likely to be further by steroids, due to effects on blood and and possibly other This remains to be in vasculitis. nodosa associated with of has to steroids, the drugs and and plasma in over with of in 50% has also been associated with polyarteritis nodosa and microscopic polyangiitis, but is more related to vasculitis randomized trials have of vasculitis with in combination with steroids for this although disease relapse often drug More recent studies have reduced relapse rates with higher and treatment periods A small randomized study found no with A in to prednisolone for steroid effect was with methotrexate in a trial including patients with or The of vasculitis and to classification have the of evidence to therapy. clinical trials have had small to allow to be and treatment protocols have developed empirical more recently, consensus The to which primary systemic vasculitis need to treatment is Vasculitis in the of Wegener's granulomatosis, microscopic polyangiitis, Churg–Strauss angiitis and polyarteritis nodosa appears to in a similar to therapy and it is that disease severity the rather than subgroup. threatened vital organ damage is present, early treatment will long‐term as has been for renal vasculitis, where renal at remission long‐term renal In this early organ damage is is more important than therapeutic The of treatment efficacy with toxicity in the existing literature and has the use of protocols of intensity for severe vasculitis. The protocols also to efficacy over the remission rate was high treatment toxicity contributed to in and serious adverse effects in over one [34]. as the need for toxic there is a more for the to be regarded as a subgroup and appropriate protocols designed to a more with The of and immune for the vasculitis with these drugs in vasculitis is usually they are for a typically or rheumatoid arthritis. and the of these will also require the of in vasculitis they can be