ObjectiveTo compare two forms of continuous renal replacement therapy, continuous venovenous hemofiltration (CVVH) vs. continuous venovenous hemodialysis (CVVHD), in terms of the removal of inflammatory mediators from the blood of patients with systemic inflammatory response syndrome and acute renal failure.DesignRandomized crossover, clinical study.SettingUniversity teaching hospital.PatientsThirteen patients with systemic inflammatory response syndrome and acute renal failure receiving continuous renal replacement therapy.InterventionPatients were randomized to receive either convective clearance using CVVH or diffusive clearance using CVVHD for the first 24 hrs, followed by the other modality for 24 hrs. All treatments utilized AN69 hemofilters. CVVH was performed with an ultrafiltration rate of 2 L/hr and CVVHD with a dialysis outflow rate of 2 L/hr.Measurements and Main ResultsPlasma and ultrafiltrate concentrations of tumor necrosis factor (TNF)-alpha, interleukin (IL)-6, IL-10, and sL-selectin were measured at 0, 1, 3, 6, 12, and 24 hrs by radioimmunoassay. Plasma endotoxin concentrations were also measured at 0, 12, and 24 hrs by chromogenic assay. CVVH was associated with a 13% decrease in plasma TNF-alpha concentrations compared with a 23% increase while on CVVHD (p < .05). Mean plasma concentrations of IL-6, IL-10, and sL-selectin were unchanged over time and between therapies. Only minimal amounts of mediators were recovered in the effluents with either therapy except for IL-6. The clearances for IL-6 were different between therapies, 1.9 +/- 0.8 (SD) mL/min for CVVHD and 3.3 +/- 1.5 mL/min for CVVH, (p < .01). Plasma endotoxin concentrations were not different between therapies.ConclusionCVVH resulted in a decrease in plasma TNF-alpha concentrations as compared with CVVHD, while the type of transport mechanism used did not influence plasma concentrations of IL-6, IL-10, soluble L-selectin, or endotoxin. Differences in clearance for IL-6 between CVVH and CVVHD did not translate into significant changes in circulating IL-6 concentrations. (Crit Care Med 1998; 26:1995-2000)