Low serum albumin levels are very common in critically ill patients, with reported incidences as high as 40-50%. This condition appears to be associated with alterations in the degree of protein binding of many highly protein-bound antibacterials, which lead to altered pharmacokinetics and pharmacodynamics, although this topic is infrequently considered in daily clinical practice. The effects of hypoalbuminaemia on pharmacokinetics are driven by the decrease in the extent of antibacterial bound to albumin, which increases the unbound fraction of the drug. Unlike the fraction bound to plasma proteins, the unbound fraction is the only fraction available for distribution and clearance from the plasma (central compartment). Hence, hypoalbuminaemia is likely to increase the apparent total volume of distribution (V(d)) and clearance (CL) of a drug, which would translate to lower antibacterial exposures that might compromise the attainment of pharmacodynamic targets, especially for time-dependent antibacterials. The effect of hypoalbuminaemia on unbound concentrations is also likely to have an important impact on pharmacodynamics, but there is very little information available on this area. The objectives of this review were to identify the original research papers that report variations in the highly protein-bound antibacterial pharmacokinetics (mainly V(d) and CL) in critically ill patients with hypoalbuminaemia and without renal failure, and subsequently to interpret the consequences for antibacterial dosing. All relevant articles that described the pharmacokinetics and/or pharmacodynamics of highly protein-bound antibacterials in critically ill patients with hypoalbuminaemia and conserved renal function were reviewed. We found that decreases in the protein binding of antibacterials in the presence of hypoalbuminaemia are frequently observed in critically ill patients. For example, the V(d) and CL of ceftriaxone (85-95% protein binding) in hypoalbuminaemic critically ill patients were increased 2-fold. A similar phenomenon was reported with ertapenem (85-95% protein binding), which led to failure to attain pharmacodynamic targets (40% time for which the concentration of unbound [free] antibacterial was maintained above the minimal inhibitory concentration [fT>MIC] of the bacteria throughout the dosing interval). The V(d) and CL of other highly protein-bound antibacterials such as teicoplanin, aztreonam, fusidic acid or daptomycin among others were significantly increased in critically ill patients with hypoalbuminaemia compared with healthy subjects. Increased antibacterial V(d) appeared to be the most significant pharmacokinetic effect of decreased albumin binding, together with increased CL. These pharmacokinetic changes may result in decreased achievement of pharmacodynamic targets especially for time-dependent antibacterials, resulting in sub-optimal treatment. The effects on concentration-dependent antibacterial pharmacodynamics are more controversial due to the lack of data on this topic. In conclusion, altered antibacterial-albumin binding in the presence of hypoalbuminaemia is likely to produce significant variations in the pharmacokinetics of many highly protein-bound antibacterials. Dose adjustments of these antibacterials in critically ill patients with hypoalbuminaemia should be regarded as another step for antibacterial dosing optimization. Moreover, some of the new antibacterials in development exhibit a high level of protein binding although hypoalbuminaemia is rarely considered in clinical trials in critically ill patients. Further research that defines dosing regimens that account for such altered pharmacokinetics is recommended.
