Is there a role for albumin-based fluid replacement in the critical patient?

Albumin accounts for 80% of colloid osmotic pressure and is the carrier of several endogenous molecules and drugs. Albumin shows anti-oxidant properties, acts as nitric oxide modulator and it is the most relevant "buffer" of the extra-cellular space. Low plasma albumin concentrations are associated to worse outcome. After 3 metanalysis (1998, 2001 and 2003) suggesting that albumin was harmful, neutral and advantageous, respectively, the SAFE study found that the infusion of 4%-albumin solution in a general ICU population was as safe as normal saline. The subgroup analysis of the study showed a beneficial trend in septic patients treated with albumin and unfavourable trend in trauma patients with brain injury. Recently, the ALBIOS trial randomized 1818 patients with severe sepsis or septic shock to receive either 20% albumin and crystalloids or crystalloids alone for volume replacement. In addition in the treated group albumin was given for 28 days if a threshold of 30 g/L was not achieved. The trial did not find any difference in 28-day and 90-day mortality rates between the 2 groups. A post-hoc analysis of the severe septic shock patients showed a significantly lower 90-day mortality in the Albumin group. During the volume replacement phase an equal amount of fluid was given suggesting that the oncotic effect was extremely modest, if any. Therefore the possible benefit of albumin likely relies on its non-oncotic functions. The available data suggest that albumin must be avoided in trauma patients with brain injury, while its use in patients with septic shock could be indicated, particularly if we consider that the possible alternative carries side effects on coagulation, kidney function, tissue edema, absent in albumin. The use in case of liver cirrhosis and spontaneous bacterial peritonitis, type 1 hepatorenal syndrome should be considered and recently terminated RCT in these patients seem extremely encouraging.