Ultrafiltration and Peritoneal Dialysis During Neonatal CPB

OBJECTIVES: To assess the impact of balanced ultrafiltration and peritoneal dialysis (PD) on plasma and urinary cytokines and renal dysfunction after
cardiopulmonary bypass (CPB) surgery in newborns and infants.

METHODS: Twenty-three newborns and infants weighing less than 7 kg and scheduled for operation on congenital malformation were enrolled in this descriptive open clinical study. All patients received conventional ultrafiltration in the CPB rewarming period. Eleven newborns underwent Tenckhoff-catheter implantation in the operation theatre as a routine institutional procedure and received PD after
admission to the ICU (the PD Ä+Å group). No PD was used in another 12 patients (the PD Ä-Å group). Interleukins (IL) 6 and 8 were measured four times pre- and post-operatively. Kidney function was assessed by creatinine clearances and urine protein and enzyme analyses.

RESULTS: All patients had an uneventful clinical course. Age (10+/-2 days, PD Ä+Å vs. 96+/-19 days, PD Ä-Å), CPB duration (215+/-23 vs. 143+/-20 min), and degree of hypothermia (26+/-1.3 vs. 31+/-0.1 degrees C) differed significantly between the groups. Age, CPB duration and ultrafiltration influenced post-operative IL-levels in an analysis of variance. While there were few differences immediately after the end of
ultrafiltration, post-operative levels of IL-6 and IL-8 were higher and more sustained in the newborns (PD Ä+Å) than in the older infants (PD Ä-Å). The
median amount of IL-6 and IL-8 removed by ultrafiltration came to 28 and 59% compared to the amount of IL-6 and IL-8 remaining in the blood at the end of CPB. IL-clearance by ultrafiltration was more than 1000-fold and by PD more than 100-fold as effective as IL-clearance by the kidney. While the kidneys showed an unselective mixed glomerular and tubular pattern of injury, during CPB higher
serum IL-concentrations correlated with lower urinary IL-clearances in both study groups.

CONCLUSIONS: Ultrafiltration and PD are highly effective in removing proinflammatory cytokines. Impaired kidney function was associated with
proinflammatory IL-serum concentrations. Thus, we raise the hypothesis that glomerular-filtered proinflammatory ILs damage the proximal tubular cells of the kidney in newborns and infants, thus contributing to post-operative renal
dysfunction. Conversely, we conclude that removing proinflammatory ILs by ultrafiltration and PD acts renoprotectively. A future prospective randomised
study could demonstrate whether this can indeed improve clinical outcome.