Insufficient penetration of systemic vancomycin into the PermCath lumen.

Abstract

Catheter infection is a major cause of morbidity and catheter loss in chronic haemodialysis patients. There has been a large discrepancy in the catheter salvage rate, after an episode of documented bacteraemia, whether the patients receive systemic antibiotic alone or systemic antibiotics concomitant with 'antibiotic-lock technique' (20-30% vs 100%, respectively). To test the hypothesis that vancomycin may not adequately penetrate into the lumen of the catheter, despite therapeutic plasma levels, a series of in-vivo, ex-vivo, and in-vitro experiments were performed. We compared serum and intralumenal (0.3-0.5 ml aspirate from venous port of the catheter) vancomycin concentrations in 24 chronic haemodialysis patients, with documented bacteraemia, who had received prior systemic vancomycin therapy with 14 similar patients who had additionally received 'vancomycin-lock technique' (100 microg/ml of vancomycin in heparin solution) after each haemodialysis session. Despite serum vancomycin concentration of approximately 17 microg/ml in each group, the vancomycin concentration in the venous hub of the catheter was only 0.2+/-0.6 microg/ml in the former group, in sharp contrast to 125. 6+/-13 microg/ml in the latter group. In the ex-vivo experiment, four uninfected PermCaths which had been removed were immediately fixed and studied with scanning electron microscopy. No cellular or fibrin barrier could be found at the terminal pore of the catheter interfering with the diffusion of vancomycin from plasma into the catheter lumen. In the in-vitro experiments, three PermCaths filled with standard heparin solution were incubated for 48 h in 100 ml of plasma containing 20 microg/ml of vancomycin. Vancomycin concentration was measured in 0.3-0.5 ml solution aspirated from each port of the catheters. Vancomycin concentration was 0.2+/-0.1 microg/ml in the aspirated samples. Finally, two PermCaths filled with the standard heparin solution were incubated for 48 h in 100 ml of plasma containing 20 microg/ml of vancomycin, after which the catheters were sectioned at 4-cm intervals. Only the distal 4 cm of the catheters had vancomycin concentrations of 2 and 5 microg/ml, the remaining segments had levels </=0.5 microg/ml. CONCLUSION.: Our results indicate that diffusion of vancomycin from plasma into the haemodialysis catheter is negligible. Thus, haemodialysis patients with central venous catheter who have to be treated for bacteraemia with systemic antibiotic therapy must always receive 'antibiotic-lock technique' of the catheter after each haemodialysis session.