L. Harper and C. O. Savage

Transcription

1 Rheumatology 2005;44: Advance Access publication 21 December 2004 ANCA-associated renal vasculitis at the end of the twentieth century a disease of older patients L. Harper and C. O. Savage doi: /rheumatology/keh522 Objective. Antineutrophil cytoplasmic antibody (ANCA)-associated vasculitides are increasingly recognized in older patients. However, it is unknown whether disease presentation and response to treatment differs from younger patients. We aimed to examine the presentation, response to treatment and outcome of patients over 65 yr of age compared with a younger cohort. Methods. This retrospective, single centre, sequential cohort study reports presenting features and outcome of 233 consecutive new patients with ANCA-associated vasculitis between 1990 and Results. The median age of all patients was 65 yr (range yr). Older patients (>65 yr) presented with more severe renal involvement at presentation (P<01). Older patients were as likely to respond to treatment or undergo relapse as the younger patients. Older patients receiving immunosuppression had an increased risk of infection (P ¼ 027). was worse in the older group (P ¼ 16) and death occurred early. Mortality was associated with poor renal function (creatinine >400 lmol/l), infection and low serum albumin. Leucopenia was associated with severe renal impairment (P ¼ 048) and increased risk of infection (P ¼ 006). Multivariate analysis determined that serum creatinine >400 lmol/l and age were independent risk factors for poor prognosis. Conclusion. ANCA-associated vasculitis occurs frequently in older patients and physicians should maintain a high index of suspicion. Older patients have a poorer prognosis due to more severe renal involvement and increased sensitivity to adverse effects of treatment. This study highlights the importance of careful dosing of cyclophosphamide: in those aged over 65 yr a 25% dose reduction is safe and reduces the risk of leucopenia. This study further highlights the importance of renal function on prognosis and the need for less toxic treatment regimens. KEY WORDS: ANCA-associated vasculitis, Wegener s granulomatosis, Older age, Outcome, Treatment. Older patients are known to often present to medical practitioners with non-specific symptoms, often referred to as gone off their legs. It is important to recognize vasculitis, as it is a potentially treatable condition within this age group. Antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis affects more than 20 per million population per year with an age-specific increase in incidence: there is a peak age of yr (incidence 52.9/million population) [1]. Before the advent of immunosuppressive treatment, the prognosis of Wegener s granulomatosis was poor with a 2-yr mortality of 80%. The introduction of cyclophosphamide and steroids in the early 1970s has resulted in complete remission rates of greater than 90% and reported 5-yr survival rates of 60 80% depending on renal function [2, 3]. The morbidity and mortality nevertheless remains high, with many patients suffering severe adverse effects of treatment. Therapy causes severe adverse events in 25% of patients in the first year, irreversible damage in over 50% and is the major cause of death in the first 5 yr. However, clear prognostic factors are not available to allow tailoring of therapy to factors other than life-threatening organ involvement. Age and severity of renal disease are frequently cited as adverse prognostic indicators [4, 5]. It is thought that older patients are more likely to suffer severe adverse events due to treatment with increased risk of infection, although this is controversial [6, 7]. In any case, older patients present with more severe renal involvement, which is itself an adverse prognostic sign. Patients recruited into recent European Vasculitis Study Group (EUVAS) trials showed that those with more severe renal involvement (serum creatinine >500 mol/l) were older and had a higher mortality than those with less severe renal disease [8]. ANCA-associated vasculitis includes Wegener s granulomatosis, a granulomatous disease with systemic vasculitis, and microscopic polyangiitis, which is manifest by systemic vasculitis but no granuloma. These diseases are commonly grouped together due to their histological similarities, the absence of immune deposits, the contribution of ANCA to pathogenesis and their similar responses to immunosuppressive treatment. However, patients with microscopic polyangiitis tend to be older than those with Wegener s granulomatosis [9]. It has been suggested that immunosuppressant regimes should be reduced in older patients to counteract the increased incidence of side-effects associated with therapy; however, it is not yet determined as to whether the disease in older patients is such that this is a reasonable option. The aim of this study was to investigate the presentation of ANCAassociated disease in older patients and disease outcome compared with a younger population. Patients and methods Inclusion Two hundred and forty consecutive patients presenting to a single centre from 1990 until 2000 with a diagnosis of ANCA-associated vasculitis with histological confirmation of pauci-immune necrotizing glomerulonephritis were included in the study. The study was conducted in accordance with the Declaration of Helsinki. Division of Immunology and Infection, The Medical School, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK. Submitted 27 August 2004; revised version accepted 23 November Correspondence to: L. Harper. L.Harper@bham.ac.uk 495 Rheumatology Vol. 44 No. 4 ß British Society for Rheumatology 2004; all rights reserved

2 496 L. Harper and C. O. Savage Patients were referred by their primary care physicians or local district general hospitals from a mixed urban and rural catchment area of approximately 1.5 million people. Patients were identified from a prospectively collected renal biopsy database. Patients with antiglomerular basement membrane disease or secondary causes of vasculitis were excluded. Patients were classified into diagnostic subgroups, namely Wegener s granulomatosis or microscopic polyangiitis. Wegener s granulomatosis was diagnosed if there was evidence of granulomatous disease on biopsy or presence of clinical signs strongly suggestive for granulomatous disease. This comprised involvement of the upper respiratory tract with nasal inflammation (purulent/bloody nasal discharge), sinusitis or otitis media or lower respiratory tract manifestation with pulmonary nodules or fixed infiltrates. Microscopic polyangiitis was diagnosed if evidence of small vessel vasculitis was present with absence of granuloma on biopsy specimens and absence of clinical signs strongly suggestive of granulomatous disease. ANCA antigen specificity was not used as a diagnostic criterion. Data were recorded retrospectively from patients notes. The time of diagnosis was defined as the date of renal biopsy. The time span from the onset of first symptoms related to small vessel vasculitis and inclusion was also recorded as self-reported by the patient. Presenting symptoms, co-morbidities and laboratory data, which included renal function and ANCA, were collected at inclusion. Episodes of disease relapse, as defined by the return of clinical signs or symptoms or laboratory evidence of disease activity sufficient to warrant a sustained increase in immunosuppressive therapy, were recorded. White cell counts were taken at each clinic visit to allow identification of episodes of leucopenia. ANCA were defined as either cytoplasmic (c-anca) or perinuclear (p-anca) as determined by indirect immunofluorescence. Disease activity was assessed using the disease extent index (DEI) [10]. Episodes of infection, leucopenia, diabetes, malignancy and other adverse events associated with treatment were collected from case records. Follow up was terminated in December Although this was a retrospective study, there was nevertheless an intention to treat all patients with a standard regime of 2 mg/kg oral cyclophosphamide until remission was achieved (defined by a Birmingham Vasculitis Activity Score of 0) and prednisolone 60 mg/day tapering to 7.5 mg/day by 6 months. The treatment regime was continued throughout the inclusion period. Twentyfive per cent dose reductions of cyclophosphamide were made for those aged over 65 yr or those who developed neutropenia (total leucocyte count less than 4000 cells/mm 3 ). No change in corticosteroid dose was made for age. Seventeen patients were treated with pulsed cyclophosphamide as per The European Study Group protocol for the CYCLOPS trial ( protocols/cyclops.doc) (10 under the age of 65 yr and seven over the age of 65 yr). In cases presenting as severe renal disease (creatinine >500 mol/l) or renal pulmonary syndrome, pulse methyl prednisolone (three pulses of 1 g) or plasma exchange (intention to give seven 4 l exchanges over 2 weeks) were added. Maintenance therapy using azathioprine 1.5 mg/kg and prednisolone was given following induction therapy. The steroid dose was tapered for maintenance treatment (7.5 mg/day) and continued for 5 yr. All patients received Pneumocystis carinii (PCP) prophylaxis with Septrin while receiving induction therapy. The study was approved by the local ethics committee. Statistics Statistical analysis was performed with SPSS 12.0 for Windows (SPSS Inc., Chicago, IL, USA) Medians and ranges are reported for non-normally by distributed data and means S.D. are reported for normally distributed data. Differences between means were tested using Student s t-test where data were normally distributed and the Mann Whitney test was used for non-parametric data. The 2 test was used for categorical data and described with odds ratio (OR) and confidence intervals (CI). Kaplan Meier survival analysis was used to assess patient survival. Univariate survival analysis was performed using the log-rank test. Multivariate analysis of patient survival used Cox s regression model. Variables that did not affect survival significantly were removed by a stepwise procedure according to a likelihood ratio. All tests were two-tailed and P values of <5 were considered significant. Results Demographics Two hundred and forty patients with a new diagnosis of ANCAassociated vasculitis were included in the study. Seven patients were excluded from analysis due to dual positivity of ANCA and antiglomerular basement membrane antibodies and four patients excluded due to a diagnosis of Churg Strauss syndrome. Ten patients were lost to follow-up (median time 23 months, range 4 71 months). The median age was 65 yr (range yr). There was similar distribution of males and females (124 males and 105 females) (Table 1). As in other studies, those with Wegener s granulomatosis (WG) were younger than those with microscopic polyangiitis (MPA) (WG median age 63 yr, range yr; MPA median age 66 yr, range yr; P ¼ 11) [4, 11]. Ninety-eight per cent of patients were Caucasian. Remission was achieved in 198 (84%) patients by 3 months and in 23 (10%) further patients by 3 6 months. Presentation Older patients (those over 65 yr) presented with more severe renal failure (creatinine mol/l) compared with younger patients (creatinine mol/l), P<01. Forty-four patients over 65 yr and 24 patients under 65 yr presented with dialysis-dependent renal failure (OR 1.48, CI ; P ¼ 057). Renal function was weakly correlated with age (R ¼ 24, P ¼ 1). Younger patients had more extensive disease with a median DEI of 6 (range 2 14) compared with older patients, DEI ¼ 5 (range 2 12) (P<01), DEI was correlated negatively TABLE 1. Demographics and clinical characteristics Age >65 yr Age <65 yr Statistical significance No of patients Age (range) (yr) 70 (65 90) 50.5 ( ) Disease (WG:MPA) 36:78 45:70 NS Duration of symptoms 8 (1 520) 12 ( ) NS prior to diagnosis (range) (weeks) Sex (F:M) 58:56 47:68 NS Renal function P ¼ 003 (creatinine mmol/l) Disease extent 5 (2 12) 6 (2 14) P<01 index (range) Organ involvement NS Renal only Lung ENT Eye Skin Joints Peripheral neuropathy Diabetes prior to diagnosis 26% 18% NS Cardiovascular disease 18% 22% NS WG, Wegener s granulomatosis; MPA, microscopic polyangiitis; NS, not significant.

3 ANCA-associated vasculitis in older patients 497 with age (R ¼ 64, P<1). By definition all patients had renal involvement, there was no difference in ear, nose and throat, central nervous system or lung involvement between the two groups (Table 1). Eleven patients in total presented with lifethreatening pulmonary haemorrhage. There was no difference in the duration of reported symptoms prior to diagnosis between older and younger patients (P ¼ 0.9), the median time to diagnosis was 10 weeks (range weeks). All patients had microscopic haematuria. ANCA was present in 95% of patients by indirect immunofluorescence. Patients with p-anca were older than those with c-anca (median age p-anca 63 yr, range yr; median age c-anca 59 yr, range yr; P ¼ 42). There was no difference in the prevalence of diabetes mellitus or cardiovascular disease (such as myocardial infarct or cerebrovascular disease) between the two groups. Outcome was influenced by age: patients were divided into quartiles by age and survival compared. There were significant differences in survival, survival was worse in those aged over 65 yr (P ¼ 009). Ninety-one deaths occurred during follow-up, with the median time to death being 3 months (range months) (Fig. 1). Thirty-three deaths were due to sepsis (18 >65 yr and 15 <65 yr; P ¼ 0.11); treatment was withdrawn on nine patients, five patients died of malignancy and six died due to active disease. Using data provided by the Office of National Statistics (2002), death rates were analysed according to age at diagnosis; those over 65 had an increased risk of death compared with the normal population over 65 yr (OR 5.8, CI ). was not influenced by disease diagnosis, ANCA as assessed by IIF, or DEI. Two hundred and twenty patients in total achieved complete remission of their disease following induction therapy defined by the absence of clinically active disease, complete resolution of pulmonary infiltrates or stable chronic scarring and the absence of systemic inflammation. Nine patients failed to achieve complete remission of their disease and eventually died with active disease. There was no difference between younger and older patients achieving complete remission. Relapse occurred in 25.6% of patients at a median of 17 months (CI ). There was no difference in the risk of relapse between older and younger patients (P ¼ 9). As in other studies relapse was more common in Age influences survival age <53 years age >70 years age years age years patients with Wegener s granulomatosis (OR 2.1, CI ; P ¼ 001) [4, 12]. Renal function The median creatinine of the whole population was 415 mol/l (range mol/l). was worse in those presenting with a creatinine >400 mol/l (P<001) irrespective of age (Fig. 2a). Older patients with creatinine >400 mol/l did significantly worse (Fig. 2b). Older patients were more likely to progress to end-stage renal failure (P ¼ 39) (Fig. 3). In those patients who presented with severe renal failure (creatinine >500 mol/l), there was no difference in response to treatment between older and younger patients; older patients were as likely to become dialysis independent as younger patients following treatment at 3 months (P ¼ 8). In those who were dialysis dependent, risk of death was significantly greater in those aged over 65 yr; only 12 of 40 older patients who developed end-stage renal failure survived more than 3 months following commencement of renal (a) survival p<0001 (b) Renal function influences survival creatinine <400µmol/l >400µmol/l age and renal function Age <65 years Cr <400µmol/l Age <65 years Cr >400µmol/l Age >65 years Cr <400µmol/l Age >65 years Cr >400µmol/l p=02 FIG. 1. and age divided into quartiles. Patients aged over 65 yr had significantly poorer outcome than those under 65 yr. (Colour figure available at Rheumatology Online.) p< FIG. 2. and renal function (creatinine >400 mol/l at presentation) in the whole group (a). Serum creatinine >400 mol/l at presentation was associated with a significantly poorer outcome. This was also true for those over the age of 65 yr (b). (Colour figure available at Rheumatology Online.)

4 498 L. Harper and C. O. Savage Time to development of ESRF (a) Infection influences survival <65 years >65 years infection no infection survival p= duration FIG. 3. Age is an important predictor of end-stage renal failure (ESRF); those over the age of 65 were more likely to develop ESRF than patients under 65 yr of age (P ¼ 47). (Colour figure available at Rheumatology Online.) replacement therapy compared with 20 of 29 patients under 65 yr of age (OR 2.01, CI ; P ¼ 017). Infection was influenced by infection: the risk of death was increased in those who developed an infection requiring hospitalization (OR 1.58, CI ; P ¼ 041) (Figure 4). Older patients were more likely to develop infection than younger patients (OR 1.9, CI ; P ¼ 027). The risk of an infection in those who developed leucopenia was increased (OR 1.75, CI ; P ¼ 006); however, age was not a risk for leucopenia (P ¼ 71). Leucopenia was more likely to occur when patients were receiving cyclophosphamide (OR 1.24, CI ). The cyclophosphamide dosage was lower in the older group according to protocol (mean daily cyclophosphamide dose >65 yr mg/day, <65 yr mg/day; P ¼ 083). Further age-related reductions in cyclophosphamide dose did not seem justified as those over 75 yr showed no increased risk of leucopenia. Severe renal disease (creatinine >400 mol/l) also increased the risk of leucopenia (OR 1.43, CI ; P ¼ 048). There was no difference in the types of infection that older patients developed: 20 patients in total developed an opportunistic infection, only one patient developed PCP. Other opportunistic infections included six patients with fungal pneumonia, three episodes of aspergillous pneumonia, six patients who developed cytomegalovirus, including one patient with uveitis, one patient with Listeria pneumonia and three patients with tuberculosis. Seventy per cent of infections were related to the lower respiratory tract. No dose reduction for age was made for corticosteroids. There was no difference in the median daily dose of steroids between older and younger patients (patients >65 yr 10 mg, range 60 0 mg; patients <65 yr 8 mg, range 60 0 mg; P ¼ 3). In this study, neither corticosteroid dose (high >20 mg, intermediate mg or low <10 mg) nor cumulative dose were risks for infection, irrespective of age. Multivariate analysis determined that serum creatinine >400 mol/l and age were retained as independent risk factors for poor prognosis. p=26 (b) p= Functions duration <65 infection occurred <65 no infection occurred >65 years infection occurred >65 years no infection occurred FIG. 4. Infection influences survival. (a) In the population as a whole an episode of infection increases the risk of death (P ¼ 26). However, age is a more important predictor of outcome (b). (Colour figure available at Rheumatology Online.) Adverse events due to therapy Sepsis and leucopenia were the most common events (Table 2). Malignancy occurred in 20 patients, eight had a skin malignancy and there were three bladder cancers. There was no association with duration or cumulative dose of cyclophosphamide. Twentyfive (10.7%) patients developed steroid-induced diabetes. There was no difference in the development of adverse events between older and younger patients except infection, which was increased in older patients. Discussion The present study analyses the pattern of presentation, outcome and associated side-effects of treatment in an older (age over 65 yr) population compared with a younger population (age under 65 yr). In this study, older patients were found to present with more severe renal disease, are more likely to have only renal involvement compared with younger patients and are more likely to have p-anca and a lower DEI. This did not appear to be due to delayed diagnosis as there was no difference in the duration

5 ANCA-associated vasculitis in older patients 499 TABLE 2. Adverse events Age >65 yr Age <65 yr Statistical significance Death OR 1.76 (CI ); P<001 Infection OR 1.67 (CI ); P ¼ 001 Leucopenia NS Myocardial infarction 3 1 NS Cerebrovascular disease 2 5 NS Diabetes 5 20 NS Thrombotic events (DVT NS or pulmonary embolus) Malignancy 11 8 NS Fractures 6 2 NS OR, odds ratio; CI, confidence interval; DVT, deep vein thrombosis; NS, not significant. of reported symptoms prior to diagnosis between those older than 65 yr and younger patients. Older patients are as likely to respond to treatment as younger patients and as likely to have relapsing disease, suggesting that there is no rationale to treatment with less effective immunosuppressant regimens although reduction of cyclophosphamide dose for age is advocated. We were able to reduce cyclophosphamide dose by 25% without an adverse effect on outcome in patients over 65 yr. Age over 75 yr was not an additional risk factor for leucopenia, suggesting that further reductions in cyclophosphamide dose based solely on age is not necessary. Renal function was retained as an independent risk factor for outcome, suggesting that, in those with active disease, treatment should be targeted at improving renal function with immunosuppressant treatment, irrespective of age. However, novel regimes are required to reduce associated toxicity, particularly in the older population. As in other recently published series, the age of patients with ANCA-associated vasculitis is increasing; in this study the median age was 65 yr (range yr) and 30% of patients were aged 70 yr or older, underscoring the importance of studying this segment of the patient population. The 1-yr mortality rate of 29% is similar to other studies with an older population and severe renal impairment [4, 5, 13, 14], highlighting the importance of age and renal function in survival. Mortality in the first 6 months was very steep, reflecting the aggressive nature of disease and treatment-associated morbidity, as seen in other studies [4, 5]. The increased mortality in older patients was evident early and reflected more severe renal disease and increased sensitivity to treatment toxicity. The commonest cause of death was infection, irrespective of age, being the cause of death in 36% of patients. Pneumonia was the most common infection. Only one of these patients had PCP this low incidence may be related to the use of Septrin as PCP prophylaxis during induction therapy. Previous studies have highlighted the increased risk of PCP in patients receiving cyclophosphamide and prednisolone [15, 16]. This study supports the view of Chung et al. [16] who have shown that PCP prophylaxis is cost-effective in patients with ANCA-associated vasculitis. Leucopenic episodes were more likely to occur in patients treated with cyclophosphamide. Controversy exists over whether the risk of myelotoxicity increases with age. Toxicity studies of cyclophosphamide have been performed in patients with neoplastic disease. Five retrospective studies failed to demonstrate any difference in incidence, severity or duration of myelodepression in older patients [17 21]. When cyclophosphamide dose was corrected for renal function, age was not a risk factor [17]. However, in further prospective studies age was a risk factor for myelodepression [22 24]. These studies suggested that the risk of neutropenia and death from neutropenic sepsis increases with age, particularly in those over 70 yr. In this study, we reduced the cyclophosphamide dose for age (25% reduction); this was an effective strategy as age was not linked as a risk for episodes of leucopenia, even in those aged over 75, suggesting that further reductions based on age are not necessary. Despite the frequent use of cyclophosphamide in patients with impaired renal function, dose alterations are controversial in renal failure. Some studies report no changes in pharmacokinetics or toxicity profiles of cyclophosphamide suggesting no dose reduction is required [25 27] whereas others have shown decreased clearance with enhanced toxicity [28, 29]. In a recently published well-conducted study there was clear evidence of reduced clearance of cyclophosphamide with increased systemic drug exposure, suggesting a need for dose reduction in those with reduced renal function [30]. No formal dose reduction was made for renal function in this study. Severe renal disease at entry significantly increased the risk of leucopenia. This study adds to the evidence that dosage of cyclophosphamide should be reduced for both age and renal function in patients with ANCA-associated vasculitis. Since the 1960s it has been known that the morbidity and mortality from infection in cancer patients who receive chemotherapy depends on the duration and severity of leucopenia or neutropenia [31]. It is now clear from this study that it is important to avoid even mild leucopenic episodes as they increase the risk of infection. Infection was common in both older and younger patients with 39.5% of patients having one or more episodes of infection requiring treatment with antibiotics, the risk being greater in older patients. Infection was associated with poor survival, indeed 47% of early deaths (<1 yr) in both age groups were due to infection. Age was a significant risk for infection in our study, as in other studies [31]. It has previously been suggested that the success of antibiotic treatment in leucopenic patients depends on leucocyte recovery [32]. Avoidance of cyclophosphamide-induced leucopenia is an important goal of future immunosuppressant regimes, but in those patients who do become leucopenic consideration should be given to the use of G-CSF, which has been safely used in leucopenic patients with ANCA-associated vasculitis [33]. Corticosteroids are fundamental to the treatment of vasculitis. Although these drugs increase the risk of infection, their impact on patients with ANCA-associated vasculitis is unclear. In a study of 75 patients with immunologically mediated renal disease treated with immunosuppression, no identifiable risk factors for infection were determined, although a subgroup analysis of 23 patients suggested impaired renal function and high steroid dose increased risk of severe infection [34]. In a study of patients with rheumatoid arthritis independent risk factors for infection included corticosteroid usage, but not dose, age, leucopenia or disease severity [35]. In our study steroid use, irrespective of age, was not a risk for infection. The present study occurred in patients with impaired renal function who were also on additional immunosuppressive agents, which may explain the lack of association with steroid use. It has previously been suggested that older patients are more likely to present with CNS involvement, with a 5-fold higher incidence of CNS involvement in those aged over 60 [6]. Our study differs from that of Krafcik et al. [6] since we concentrated on an older group of patients aged >65 yr, as the median age of all patients at presentation was 65 yr and all patients had renal involvement. In our study older patients were more likely to present with isolated renal involvement but there was no difference in the profile of other organ involvement, although younger patients were more likely to have more generalized disease with a higher DEI. Renal function declines progressively with age, increasing the risk to older patients; 30% of healthy older people (aged over 70 yr) have a glomerular filtration rate of <70 ml/min

6 500 L. Harper and C. O. Savage [36], which may explain the increased severity of renal disease in older patients. In conclusion, this study highlights the importance of renal function and age as predictors of outcome in ANCA-associated vasculitis. ANCA-associated vasculitis is becoming increasingly common in older people and early diagnosis is essential. Ninetyfive per cent of patients were ANCA positive in this study, and this was not influenced by age. The combination of microscopic haematuria, progressive renal impairment and a positive ANCA has a positive predictive value of 95% for necrotizing glomerulonephritis. Clinicians treating older patients should maintain a high index of suspicion regarding the diagnosis of ANCAassociated vasculitis as these patients have reduced renal reserve and are more likely to present with severe renal disease. This study provides no rationale for using less effective immunosuppression in older patients, but careful cyclophosphamide dosing is required, considering age and renal function, to reduce the risk of treatmentassociated mortality. A 25% reduction in cyclophosphamide dose in those aged over 65 yr is safe and does not have an adverse impact on outcome; a further reduction should be considered for those with severe renal involvement. Rheumatology Key messages ANCA-associated vasculitis occurs in older patients. is worse due to increased renal involvement and sensitivity to adverse effects of treatment. Acknowledgements The authors thank Dr P. Nightingale for generous support and guidance in the statistical analysis of the work presented, Dr A. Howie for renal biopsy data and Dr D. Adu for clinical support. The authors have declared no conflicts of interest. References 1. Watts RA, Lane SE, Bentham G, Scott DG. Epidemiology of systemic vasculitis: a ten-year study in the United Kingdom. Arthritis Rheum 2000;43: Fauci AS, Haynes BF, Katz P, Wolff SM. Wegener s granulomatosis: prospective clinical and therapeutic experience with 85 patients for 21 years. Ann Intern Med 1983;98: Hoffman GS, Kerr GS, Leavitt RY et al. Wegener s granulomatosis: an analysis of 158 patients. Ann Intern Med 1992;116: Booth AD, Almond MK, Burns A et al. Outcome of ANCAassociated renal vasculitis: a 5-year retrospective study. Am J Kidney Dis 2003;41: Mahr A, Girard T, Agher R, Guillevin L. Analysis of factors predictive of survival based on 49 patients with systemic Wegener s granulomatosis and prospective follow-up. Rheumatology 2001;40: Krafcik S, Covin R, Lynch J, Sitrin R. Wegener s granulomatosis in the elderly. Chest 1996;109: Higgins RM, Goldsmith DJ, Connolly J et al. Vasculitis and rapidly progressive glomerulonephritis in the elderly. Postgrad Med J 1996;72: Jayne D. Current attitudes to the therapy of vasculitis. Kidney Blood Press Res 2003;26: Hauer HA, Bajema IM, van Houwelingen HC et al. Renal histology in ANCA-associated vasculitis: differences between diagnostic and serologic subgroups. Kidney Int 2002;61: de Groot K, Gross WL, Herlyn K, Reinhold-Keller E. Development and validation of a disease extent index for Wegener s granulomatosis. Clin Nephrol 2001;55: Weidner S, Geuss S, Hafezi-Rachti S, Wonka A, Rupprecht HD. ANCA-associated vasculitis with renal involvement: an outcome analysis. Nephrol Dial Transplant 2004;19: Jayne D, Rasmussen N, Andrassy K et al. A randomized trial of maintenance therapy for vasculitis associated with antineutrophil cytoplasmic autoantibodies. N Engl J Med 2003;349: Westman KW, Bygren PG, Olsson H, Ranstam J, Wieslander J. Relapse rate, renal survival, and cancer morbidity in patients with Wegener s granulomatosis or microscopic polyangiitis with renal involvement. J Am Soc Nephrol 1998;9: Hogan S, Nachman P, Wilkman A, Jennette J, Falk R. Prognostic markers in patients with antineutrophil cytoplasmic autoantibodyassociated microscopic polyangiitis and glomerulonephritis. J Am Soc Nephrol 1996;7: Sen RP, Walsh TE, Fisher W, Brock N. Pulmonary complications of combination therapy with cyclophosphamide and prednisone. Chest 1991;99: Chung JB, Armstrong K, Schwartz JS, Albert D. Cost-effectiveness of prophylaxis against Pneumocystis carinii pneumonia in patients with Wegner s granulomatosis undergoing immunosuppressive therapy. Arthritis Rheum 2000;43: Gelman RS, Taylor SG. Cyclophosphamide, methotrexate, and 5-fluorouracil chemotherapy in women more than 65 years old with advanced breast cancer: the elimination of age trends in toxicity by using doses based on creatinine clearance. J Clin Oncol 1984;2: Christman K, Muss HB, Case LD, Stanley V. Chemotherapy of metastatic breast cancer in the elderly. The Piedmont Oncology Association experience [see comment]. J Am Med Assoc 1992;268: Ibrahim NK, Frye DK, Buzdar AU, Walters RS, Hortobagyi GN. Doxorubicin-based chemotherapy in elderly patients with metastatic breast cancer. Tolerance and outcome. Arch Intern Med 1996;156: Begg CB, Carbone PP. Clinical trials and drug toxicity in the elderly. The experience of the Eastern Cooperative Oncology Group. Cancer 1983;52: Giovanazzi-Bannon S, Rademaker A, Lai G, Benson AB 3rd. Treatment tolerance of elderly cancer patients entered onto phase II clinical trials: an Illinois Cancer Center study. J Clin Oncol 1994;12: Tirelli U, Errante D, Van Glabbeke M et al. CHOP is the standard regimen in patients > or ¼ 70 years of age with intermediate-grade and high-grade non-hodgkin s lymphoma: results of a randomized study of the European Organization for Research and Treatment of Cancer Lymphoma Cooperative Study Group. J Clin Oncol 1998;16: Rowe JM, Andersen JW, Mazza JJ et al. A randomized placebocontrolled phase III study of granulocyte-macrophage colonystimulating factor in adult patients (>55 to 70 years of age) with acute myelogenous leukemia: a study of the Eastern Cooperative Oncology Group (E1490). Blood 1995;86: Schiffer CA. Hematopoietic growth factors as adjuncts to the treatment of acute myeloid leukemia. Blood 1996;88: Grochow LB, Colvin M. Clinical pharmacokinetics of cyclophosphamide. Clin Pharmacokinet 1979;4: Moore MJ. Clinical pharmacokinetics of cyclophosphamide. Clin Pharmacokinet 1991;20: Bramwell V, Calvert RT, Edwards G, Scarffe H, Crowther D. The disposition of cyclophosphamide in a group of myeloma patients. Cancer Chemother Pharmacol 1979;3:253 9.

7 ANCA-associated vasculitis in older patients Bagley CM Jr, Bostick FW, DeVita VT Jr. Clinical pharmacology of cyclophosphamide. Cancer Res 1973;33: Juma FD, Rogers HJ, Trounce JR. Effect of renal insufficiency on the pharmacokinetics of cyclophosphamide and some of its metabolites. Eur J Clin Pharmacol 1981;19: Haubitz M, Bohnenstengel F, Brunkhorst R, Schwab M, Hofmann U, Busse D. Cyclophosphamide pharmacokinetics and dose requirements in patients with renal insufficiency. Kidney Int 2002;61: Bradley JD, Brandt KD, Katz BP. Infectious complications of cyclophosphamide treatment for vasculitis. Arthritis Rheum 1989; 32: Link H, Maschmeyer G, Meyer P et al. Interventional antimicrobial therapy in febrile neutropenic patients. Study Group of the Paul Ehrlich Society for Chemotherapy. Ann Hematol 1994;69: Hellmich B, Schnabel A, Gross W. G-CSF treatment for cyclophosphamide-induced severe neutropenia in Wegener s granulomatosis. Arthritis Rheum 1999;42: Cohen J, Pinching AJ, Rees AJ, Peters DK. Infection and immunosuppression: a study of the infective complications of 75 patients with immunologically mediated disease. Q J Med 1982;51: Doran MF, Crowson CS, Pond GR, O Fallon WM, Gabriel SE. Predictors of infection in rheumatoid arthritis. Arthritis Rheum 2002;46: Fehrman-Ekholm I, Skeppholm L. Renal function in the elderly (>70 years old) measured by means of iohexol clearance, serum creatinine, serum urea and estimated clearance. Scand J Urol Nephrol 2004;38:73 7.