DESIGN: This was a single-center double-blind randomized controlled trial comparing continuous venovenous hemofiltration-high cutoff to continuous venovenous hemofiltration-standard.
SETTING: Tertiary care hospital in Australia.
PATIENTS: Vasopressor-dependent patients in acute kidney injury who were admitted to the ICU.
INTERVENTIONS: Norepinephrine-free time were calculated in critically ill vasopressor-dependent patients in acute kidney injury, randomized to either continuous venovenous hemofiltration-high cutoff or continuous venovenous hemofiltration-standard.
MEASUREMENT AND MAIN RESULTS: A total of 76 patients were randomized with the following characteristics (continuous venovenous hemofiltration-high cutoff vs continuous venovenous hemofiltration-standard); median age of 65 versus 70 year, percentage of males 47% versus 68%, and median Acute Physiology and Chronic Health Evaluation scores of 25 versus 23.5. The median hours of norepinephrine-free time at day 7 were 32 (0-110.8) for continuous venovenous hemofiltration-high cutoff and 56 hours (0-109.3 hr) (p = 0.520) for continuous venovenous hemofiltration-standard. Inhospital mortality was 55.6% with continuous venovenous hemofiltration-high cutoff versus 34.2% with continuous venovenous hemofiltration-standard (adjusted odds ratio, 2.49; 95% CI, 0.81-7.66; p = 0.191). There was no significant difference in time to cessation of norepinephrine (p = 0.358), time to cessation of hemofiltration (p = 0.563), and filter life (p = 0.21). Serum albumin levels (p = 0.192) were similar and the median dose of IV albumin given was 90 grams (20-212 g) for continuous venovenous hemofiltration-high cutoff and 80 grams (15-132 g) for continuous venovenous hemofiltration-standard (p = 0.252).
CONCLUSIONS: In critically ill patients with acute kidney injury, continuous venovenous hemofiltration-high cutoff did not reduce the duration of vasopressor support or mortality or change albumin levels compared with continuous venovenous hemofiltration-standard.
METHODS: We measured plasma and post-filter levels of IL-6, TNF-alpha, IL-8, IL-1 beta, RANTES, IL-10, IFN-gamma and IFN-alpha in both study groups. We also measured cytokine levels in the ultrafiltrate and calculated sieving coefficients and clearances.
RESULTS: By 72 hours of treatment, IL-6 had decreased during both treatments (p = 0.009 and 0.005 respectively). In contrast, IL-10 had decreased with CVVH-Std (p = 0.03) but not CVVH-HCO (p = 0.135). None of the other cytokines showed changes over time. There were also no significant between group differences in plasma levels for each cytokine over the 72-hour treatment period. For all cytokines combined, however, the median sieving coefficient was higher for CVVH-HCO (0.31 vs. 0.16; p = 0.042) as was the mass removal rate by ultrafiltration (p = 0.027). While overall combined cytokine levels had fallen to 62.2% of baseline at 72 hours for CVVH-HCO (p<0.0001) and to 75.9% of baseline with CVVH-Std (p = 0.008) there were no between group differences.
CONCLUSIONS: CVVH-HCO achieved greater combined sieving coefficient and mass removal rate by ultrafiltration for a group of key cytokines than CVVH-Std. However, this effect did not differentially lower their plasma level over the first 72 hours. Our study does not support the use of CVVH-HCO to lower cytokines in critically ill patients with AKI.