Renal Replacement Therapy for Septic Acute Kidney Injury

Intermittent or Continuous Modes

Whether RRT therapies are best applied as an intermittent or continuous mode to patients with septic AKI has been the matter of long-standing debate. iRRT techniques typically rely on larger extracorporeal blood flows and lead to more rapid correction of metabolic derangements. This mode uses dialysate and replacement fluid sourced from plumbed water with electrolyte solution added. However, iRRT may be poorly tolerated in hemodynamically unstable septic patients and in those who require slower restoration of metabolic equilibrium (e.g., cerebral edema). In the international BEST Kidney study, hypotension was reported more than twice as often during iRRT compared with CRRT, despite CRRT patients having a higher severity of illness.

In contrast, continuous RRT can be performed with lower blood flows and titrated to the patient's hemodynamic status. As patients with septic AKI often are receiving infusion of catecholamines to support mean arterial pressure, CRRT commonly is favored in these circumstances. Continuous RRT also provides more gradual correction of metabolic derangements. However, lower blood flows with CRRT may necessitate increased need for anticoagulation, which may provoke bleeding complications. Longer duration on an extracorporeal circuit also may limit mobility of the patient, may lead to greater clearance of medications and micronutrients, and costs more than intermittent RRT.

Multinational epidemiologic studies of ICU patients receiving RRT for AKI, of whom 41% to 48% had septic shock, reported that CRRT was used for 75% to 80%, iRRT for 17% to 25%, and peritoneal dialysis with slow continuous ultrafiltration in 1% to 3%. Although continuous modes appear to be most commonly applied to patients with severe septic AKI, there is strong preference for iRRT in some centers. For example, in the recent AKIKI trial investigating timing of RRT initiation in French ICUs, iRRT was the mode chosen for 50% of the patients, despite most having sepsis and 85% receiving catecholamines.

Evidence supporting the optimal “duration mode” of RRT is limited. Most studies comparing outcomes after intermittent and continuous modes have been observational, with varying forms of AKI, different severity of illness, excluded hemodynamically unstable patients, have had small cohort sizes, and have not considered cointerventions. Furthermore, other RRT variables, such as clearance mode, choice of dialysis/replacement fluids, anticoagulation, time to initiation, and type of filter membrane, have not been controlled.

Over the past decade controlled studies have examined the duration mode of RRT. A series of meta-analyses that incorporated the accumulating evidence have derived the same conclusion, that there is no apparent mortality benefit (overall 50% to 70%), no difference in hemodynamic instability requiring treatment, no difference in hospital length of stay, recovery of renal function, or need for chronic dialysis. These conclusions remained after adjusting for confounders and when analyzing just the subgroup of patients with septic AKI.

Since these meta-analyses, there has been a further randomized controlled trial that stratified patients to receive either iRRT (daily for 4 to 6 hours) or CRRT. This study enrolled 316 patients and revealed no difference in mortality, hospital length of stay, or renal recovery. Again, these outcomes were no different when analyzing the subgroup of patients with septic AKI.

Dialysis or Hemofiltration Modes

It remains unclear whether RRT is best delivered as dialysis, filtration, or a combination of both. Removal of low-molecular-weight solutes is similar with both clearance modalities, whereas larger molecules are cleared more effectively with convective therapies.

Cytokines, eicosanoids, endotoxins, and other inflammatory mediators are water soluble, largely unbound in the circulation, and normally eliminated by the kidney. Ultrafiltration has been shown to enhance clearance of inflammatory mediators and has been investigated as a potential therapeutic intervention for sepsis. However, when tested in randomized controlled studies, continuous hemofiltration at 25 to 30 mL/kg/hr applied at the first sign of organ failure in septic patients, even before severe AKI had developed, did not alter cytokine levels and either did not alter, or worsen, organ function. These findings temper enthusiasm for early hemofiltration in septic AKI pending further controlled trials.

The only controlled study to suggest a treatment effect of an RRT clearance mode randomized 206 AKI patients (34% with sepsis) to continuous hemofiltration (25 mL/kg/hr) with or without an added dose of dialysis (18 mL/kg/hr). The group with the added dialysis had greater clearance of urea and creatinine and higher survival rates (34% vs. 59%, p < .01). Although this was a single-center study, and the group with added dialysis received higher intensity RRT, it is reasonable to conclude that the combination of clearance modalities is preferable for septic AKI. A multicenter study that examines RRT clearance mode, while standardizing dose, would be required to determine whether this conclusion is valid.