Objective: To determine whether plasmapheresis increases the risk for infection in immunosuppressed patients.
Design: Randomized, controlled trial.
Setting: Multicenter.
Patients: Eighty-six patients enrolled in a trial of plasmapheresis for severe diffuse proliferative lupus nephritis.
Interventions: Forty-six of the patients received high-dose steroid therapy plus cyclophosphamide therapy for 8 weeks. Thereafter, cyclophosphamide therapy was discontinued, and steroid therapy was tapered (standard treatment group). Forty patients received identical treatment and had 12 plasmapheresis procedures during the first 4 weeks of the treatment.
Measurements: Patients were examined for the development of infection.
Main results: No statistical difference in age, sex, race, serum creatinine level, proteinuria, or complement levels was found between the two groups. Over a follow-up period of 5376 patient-weeks, 74% of patients in the standard treatment group had 62 infections, yielding an aggregate infection rate of 1.15 infections per 100 weeks (median individual infection rate, 1.08; 25th and 75th percentiles, 0.0 and 2.44). This rate was comparable to that seen in the plasmapheresis-treated patients who were followed for 4187 patient-weeks: 68% had 51 infections, for an aggregate infection rate of 1.22 infections per 100 weeks (median individual infection rate, 0.94; 25th and 75th percentiles, 0.0 and 2.32). The infection rate was also comparable in the initial acute phase of the study, despite the fact that patients who received plasmapheresis then had significantly lower immunoglobulin (IgG) levels (P less than 0.001). Neither the site (superficial compared with systemic) nor the nature (conventional compared with unconventional) of infection differed statistically between the two groups. Of 14 patient deaths, 7 were from infection (4 in control group and 3 in the plasmapheresis group).
Conclusion: Plasmapheresis did not increase the risk for infection in immunosuppressed patients with severe lupus nephritis.