Résumé
There are very limited data on ticarcillin-clavulanate elimination by haemofiltration. We measured in vitro ticarcillin-clavulanate adsorption to polyacrylonitrile (PAN) filters and the sieving coefficient using a well-described bench model of haemofiltration. The dose of ticarcillin-clavulanate was 60/2 mg or 180/3 mg, and 0 or 12 g albumin was added to the 1 L of circulating blood-crystalloid mixture to produce four different experimental conditions. The experiment was repeated four times under each condition. Median (interquartile range [IQR] ) ticarcillin adsorption varied from 28 (27-30) mg to 85 (78-90) mg. Adsorption was increased when the dose of ticarcillin was higher (P<0.001), but was not affected by the addition of albumin. Median (IQR) adsorption of clavulanate ranged from 0.67 (0.55-0.75) mg to 1.8 (0.33-3.5) mg and was neither dose dependent (P = 0.505) nor significantly affected by the addition of albumin. Median (IQR) ticarcillin sieving coefficient ranged from 0.73 (0.67-0.75) to 0.99 (0.97-1.03). It was significantly higher with a higher dose of ticarcillin (P = 0.021) and without addition of albumin (P = 0.015). Median (IQR) clavulanate sieving coefficient ranged from 1.03 (1.00-2.24) to 2.0 (1.98-2.47). Clavulanate sieving coefficient was not significantly affected by dose or the addition of albumin.
These data indicate that significant adsorption of both ticarcillin and clavulanate occurs in vitro; however, this requires confirmation by clinical pharmacokinetic studies. The sieving coefficient data may help guide appropriate dosing of critically ill patients receiving haemofiltration until more extensive clinical pharmacokinetic data are available.