Op-brhe150196 1339..1341

Granulomatosis with polyangiitis (GPA) is a potentially lifethreatening ANCA-associated vasculitis (AAV) that typically involves the respiratory tract and the kidney. The introduction of glucocorticoid and CYC therapy in the 1960s dramatically improved the prognosis of affected patients, with 78% surviving at 5 years compared with a 1-year mortality in 80% of untreated patients [1]. No conventional immunosuppressants have emerged as valid alternatives to CYC for the induction treatment of AAV, except for MTX in early systemic disease. The introduction of biologics such as rituximab has offered an exciting perspective for AAV treatment; however, trials comparing rituximab and CYC as initial therapy for AAV demonstrated that the former is not superior in inducing remission and that its short-term toxicity is also not negligible [2]. Thus, CYC still remains a cornerstone of AAV therapy and is unlikely to become of historical interest only. CYC has severe side effects, including bone marrow suppression, impaired fertility, haemorrhagic cystitis and increased susceptibility to infections and cancer. While some of these (e.g. leucopenia) usually occur early, others—particularly malignancies—are more frequent in the long term and their risk is related to the cumulative dose of CYC. This is particularly relevant for GPA patients, who are more prone to relapse than patients with other vasculitides and therefore require repeat treatment courses. Thus, several attempts have been made to minimize CYC exposure in AAV. In initial studies [1], CYC was used as both induction and maintenance treatment, which resulted in a considerable cumulative dose. The adoption of a staged approach based on an induction-of-remission phase (with high-dose glucocorticoids and CYC) followed by a maintenance period (where CYC is replaced by AZA or MTX) has reduced the cumulative amount of CYC. Subsequent studies have also demonstrated that CYC exposure can be reduced if a pulse i.v. regimen is used instead of daily oral administration. Finally, preventive measures such as the use of mesna are now recommended to limit the occurrence of haemorrhagic cystitis, which is associated with an increased risk of bladder cancer [3]. Changes in the way CYC is used in AAV have probably influenced its long-term cancer risk. To address this point, we searched PubMed without any date limits for full papers on CYC-related toxicity using the search terms ANCA-associated vasculitis, Wegener, treatment, cyclophosphamide, and cancer risk. Early studies covering observation periods that dated back to the 1970s to 1980s indicated that CYC-treated GPA patients had a 2.0to 2.4-fold increase in overall cancer incidence as compared with control populations [1, 4]. Cancer-typespecific analyses showed a significant excess risk for bladder cancer, non-melanoma skin cancer (NMSC) [1 , 4, 5], lymphoma and leukaemia [4]. More recent studies, reflecting treatment periods spanning the 1990s to 2000s, have provided more reassuring findings, showing standardized incidence ratios (SIRs) (i.e. observed number of cancers divided by the expected number) of 1.6 2.1 for all cancer types [6, 7]. These studies have also shown that lymphoma is no longer over-represented in the patient population [7]. Interestingly, in a recent work with 535 AAV patients enrolled in four trials conducted by the European Vasculitis Study Group, the SIR for non-NMSC dropped to 1.3 (95% CI 0.90, 1.80) and was not significantly increased; thus, the excess cancer risk observed in AAV was driven mainly by NMSC [6]. Cancer risk is also determined by other immunosuppressants, although their influence is difficult to assess because AAV is seldom treated with these agents alone. AZA used for >12 months was associated with a 3-fold increase in all-cancer incidence after 5 years of follow-up [5]. In a long-term follow-up study, cancer incidence in AAV patients treated with MTX for induction was comparable to that of patients who received pulse CYC [8]. Cancer risk attributable to MMF seems lower than with other agents, as demonstrated in renal transplantation; in a study comparing MMF with AZA in AAV, malignancies tended to be more frequent in the latter group [9]. In this issue of Rheumatology, Faurschou and colleagues [10] report the results of a registry-based study on the long-term cancer risk following conventional immunosuppressive therapy in GPA. The authors had previously shown a greater-than-expected occurrence of NMSC among CYC-exposed patients and an increased incidence of bladder cancer and myeloid leukaemia among patients treated with high cumulative CYC doses