Thirty years of cyclophosphamide: assessing the evidence.

The ideal therapy for lupus nephritis should reduce mortality and end-stage renal disease in the long term, induce early response and remission, prevent flares, have minimal side-effects and not compromise fertility. It should also be active in all ethnic groups, widely available and cost effective. Despite 30 years' clinical experience, the ability of cyclophosphamide to meet these needs is not supported by robust evidence. The first National Institutes for Health (NIH) trial in 1986 led to a shift from oral to intravenous cyclophosphamide. The three NIH trials together then led to the dogma that high-dose intravenous cyclophosphamide is the only cytotoxic agent superior to steroids alone in lupus nephritis and to its general acceptance as the 'standard of care'. Since then, high-dose intravenous cyclophosphamide has been shown to have no impact on survival, to be less effective in black patients and to have many side-effects, particularly an unacceptable risk of premature menopause. The Euro-Lupus Nephritis Trial found that low-dose intravenous cyclophosphamide could be used as an alternative to a high-dose regimen and was associated with half as many severe infections. Other advantages include no hospitalisation and virtually no risk of premature gonadal failure. Other studies have looked at regimens in which cyclophosphamide is entirely replaced--for example, with mycophenolate mofetil--and have found fewer side-effects and better induction of remission. Intravenous cyclophosphamide is the only therapy with long-term data for reduction of end-stage renal disease. As data on other therapies accumulate, however, intravenous cyclophosphamide might no longer be considered the standard treatment for lupus nephritis.