Efficient removal of colistin A and B in critically ill patients undergoing CVVHDF and sorbent technologies

Colistin pharmacokinetics data are scarce regarding patients undergoing renal replacement therapy (RRT), or even absent as in patients treated with sorbent technologies potentially capable of removing colistin by extensive absorption on many polymeric materials.

Twelve septic shock patients with acute kidney injury (AKI) undergoing RRT [continuous venovenous hemodiafiltration (CVVHDF) n = 7, coupled-plasma filtration adsorption-HF (CPFA-HF) n = 4, hemoperfusion n = 1] treated with colistin methanesulfonate at a dose of 4.5 × 10(6) U bid were studied. Colistin A (Col-A) and colistin B (Col-B) concentrations on plasma and effluent at time 0, 0.2, 1, 3, 6, 12, 24 and 48 h were determined by the liquid chromatography-tandem mass spectrometry method.

With CVVHDF the sieving coefficient was lower for Col-A, peaked early (0.40 for Col-A at 10 min, and 0.59 for Col-B at 3 h) and declined after 48 h (0.22 and 0.30 for Col-A and Col-B, respectively). Colistin's filter clearance showed a similar pattern, with the highest clearance value of 18.7 ml/min for Col-B at 1 h. With CPFA-HF after the cartridge the Col-A and Col-B levels were negligible (<0.2 mg/l) or not detectable. The sum of the effluent and cartridge clearances reached values of 30 and 40 ml/min for Col-A and Col-B, respectively. With hemoperfusion the postcartridge concentrations for Col-A and Col-B were about 30 % lower than those determined precartridge.

During CPFA-HF and CVVHDF, the extent of colistin removal is high, and patients should receive an unreduced dosage. However, due to risk of accumulation in long-term administration colistin plasma levels determination is recommended.