Linezolid lock prophylaxis of central venous catheter infection

Prior to linezolid lock prophylaxis, the patient had 18 admissions for CVC-related infections, 28 positive blood cultures and eight CVC changes. He had spent 180 of the preceding 632 days in hospital. In 7 months of linezolid lock prophylaxis, he had one line change and one admission for 7 days due to CVC-related infection and a deep vein thrombosis in the common and external iliac veins. The tunnelled femoral line was removed and temporary venous access was achieved via a collateral vein. CVC-related infection was confirmed by culture of CNS from peripheral blood, a line culture and from the line tip. The patient had 44.3 infections per 1000 days (28 line infections in 632 days) prior to lock prophylaxis and 4.5 infections per 1000 days (one infection in the subsequent 221 days) during prophylaxis. The mean time to infection increased from 22.57 to 221 days when linezolid lock prophylaxis was used.

Prophylactic antibiotic lock solutions are only advised in special circumstances, such as recurrent catheter-related blood stream infections in a tunnelled line despite optimal aseptic technique (O'Grady et al., 2002). Although the infecting organisms and the antibiotics used were different, Onder et al. (2007) also showed that antibiotic lock prophylaxis reduced CVC-related infections and inpatient hospital stay in children with short bowel syndrome or intestinal transplantation, supporting the role of antibiotic lock prophylaxis in this patient group. Preventive lock therapy using vancomycin was shown to reduce line-infection rates in several prospective randomized controlled trials and a meta-analysis (Carratala et al., 1999; Garland et al., 2005; Safdar & Maki, 2006). Linezolid was selected for this study based on surprising in vitro results that demonstrated more rapid bacterial killing (approaching a 3-log reduction) than vancomycin or gentamicin against an established CNS biofilm in a model of catheter infection (Curtin et al., 2003). Although linezolid is bacteriostatic for planktonic-phase bacteria, this study suggests that it may be more effective in CNS biofilms. Linezolid was reported to be bactericidal in an in vitro model of endocarditis (Cha et al., 2003). Recent in vitro work suggests that bactericidal daptomycin is another excellent candidate for antibiotic lock prophylaxis (LaPlante & Mermel 2007; Raad et al., 2007). Concerns surrounding the long-term prophylactic use of antimicrobial therapy relate to selection of resistance and drug toxicity. Only 6 mg linezolid was administered per day and none of the adverse effects of systemic linezolid therapy were observed. Linezolid lock therapy has been used successfully in the treatment of catheter-related infection both in this patient and with systemic linezolid therapy in a similar patient (Castagnola et al., 2006). Resistance to linezolid was never observed in any of the cultures from our patient, despite 7 months of linezolid lock prophylaxis (MICs 0.25–1.5 mg ml^–1).

The expense of linezolid lock prophylaxis must be balanced against the cost of repeated hospital admissions, central line changes and therapeutic antibiotics. This patient has since left our care and is being assessed for an intestinal transplant. In patients with recurrent central line infections and long-term intravascular access problems where line removal is not feasible, linezolid lock prophylaxis should be considered but line surveillance, education and adherence to evidence-based guidelines for catheter insertion and manipulation (Centers for Disease Control and Prevention, 2005; Pronovost et al., 2006) remain the basis for the successful prevention of CVC-related infections.