[What are the advances in the treatment of acute kidney failure?]

Mortality due to acute renal failure has remained high over the last three decades despite a better understanding of the pathophysiologic mechanics involved and advances in the management of critically ill patients. This paradoxical situation raises many questions concerning the criteria used to assess treatment and disease severity as well as the effectiveness of certain recently proned "advances". In the 1980s, much progress was made in preventing and/or limiting the extent of acute renal failure. The use of very early and vigorous fluid administration associated with alkaline diuresis prevents traumatic rhabdomyolysis; saline hydration before and after radiocontrast administration protects against the acute decrease in renal function in high-risk patients; maintenance of an adequate intravascular volume and of blood pressure helps prevent acute renal failure. After the onset of acute failure, low-dose dopamine can increase urine output, whereas dobutamine improves creatinine clearance but there is no evidence that increasing urine output lowers morbidity or mortality. New techniques have been proposed for acute renal replacement therapy, but except for use of bicarbonate dialysis and biocompatible membranes, none have been shown to be superior. It has been claimed, on the basis of uncontrolled or retrospective studies, that continuous hemofiltration or hemodiafiltration could have a beneficial effect on survival and/or the course of infectious complications. However, convincing evidence of this beneficial effect is lacking since these techniques are incompatible with clinically pertinent removal of proinflammatory cytokines. The type of renal support may have no effect on outcome. Can further progress be expected in the future? Antagonists of NO receptors and growth factors have a protective effect on renal function in animal models, results which may be of clinical relevance. Their clinical potential should be evaluated in prospective randomized trials involving patients where severity of illness is assessed at inclusion using a multiparametric model combining a severity score and relevant prognostic factors.