The value of pulse cyclophosphamide in ANCA-associated vasculitis: meta-analysis and critical review

Background. The study aimed at Studying efficacy and adverse effects of pul se cyclophosphamide (pCyc) treatment and to compare it to continuous Cyclo- phosphamide (cCyc) for induction of remission in ANCA-associated vasculitid es from data in the published literature. Methods. A Medline search identified 14 studies, containing more than five patients. From the 11 nonrandomized studies. data on outcome following pCyc treatment were extracted. Results were given as fraction of the number of evaluable patients. A metaa nalysis was performed on the three prospective, randomized controlled trial s to compare outcomes concerning remission, relapses. infection. leucopenia , death and renal failure in patients treated with pCyc as opposed to cCvc. Results. The 11 non-randomized studies comprised 202 patients receiving pC yc. Cyc pulses of 375-1000 mg sqm pulse were applied at weekly to monthly i ntervals with different concomitant prednisolone regimens and variable adju nctive therapy. Complete remission was achieved in 112/191. partial remissi on in 23/191 evaluable patients. Relapses occurred in 68 135 patients, 40 1 15 patients were non-responders. Leucopenia, infections. haemorrhagic cysti tis, and deaths were rare. The meta-analysis, comprising 143 patients. show ed that pCyc compared with cCyc treatment was significantly less likely to fail to induce remission (OR 0.29: 95% CI 0.12 0.73) and had a significantl y lower risk of infection (OR 0.45: 95% CI 0.23-0.89) and leucopenia (OR 0. 36; 95% CI 0.17 0.78). Relapses occurred slightly, although not statistical ly significantly, more offer under pCyc treatment (OR 1.79: 95% CI 0.85-3.7 5). There were no differences in end-stage renal failure or deaths between the two regimens. Conclusions. The currently available, rather sparse data show that pCyc is less toxic than cCvc therapy and is an at least equally potent inductor of remission, but possibly at the expense of a higher relapse rate. The existi ng data do not give sufficient information on outcomes as time to remission and relapse, irreversible damage or quality of life without which a treatm ent regimen cannot satisfactorily be evaluated today. A large prospective r andomized controlled trial is needed to address these issues and their rela tive importance.