Diagnóstico precoz y tratamiento preventivo de la infección relacionada con catéter venoso central tunelizado de pacientes en hemodiálisis

Resumen

Catheter related bloodstream infection (CRBI) is a major cause of morbidity and mortality in patients with end-stage renal disease treated with chronic hemodialysis. The incidence of CRBI in permanent tunnelled cuffed hemodialysis catheters oscillates from 1.5 to 5.5 episodes per 1000 catheter-days. Colonization is predominantly intraluminal in long-term catheters. In vivo colonization dynamics and biofilm formation have not been studied. Catheter colonization (CC) always precedes CRBI and infectious and mechanical complications. Our aim was to study colonization dynamics, to develop a continuous monitorization study on hemodialysis catheters and to evaluate a therapy in order to prevent CRBI. Colonization dynamics on polyurethane tunnelled catheters were evaluated in 24 female New Zealand White rabbits. Catheter lumen was filled with 0.1 mL of a log phase growth S. epidermidis (<3 log cfu/mL). Intraluminal colonization and systemic infection was daily evaluated, throughout 24 days, with intracatheter leukocyte stained by acridine orange (ILAO), intracatheter blood leukocyte culture (ILC) and quantitative blood cultures (QBC) simultaneously drawn through catheter and peripheral vein. Catheters were removed on days 3, 6, 9, 15 and 24, and semiquantitative and quantitative diagnostic techniques were performed. We demostrated that intraluminal infection leads to intra- and extra-luminal significative catheter colonization. Bacteremia was secondary to catheter manipulation. Infection disseminated to liver, lung, spleen and heart. We conducted an observational study in our hemodialysis unit, from July 2003 to January 2006, involving 35 patients with 45 catheters. The aim was to detect CC and establish a therapy based in catheter antibiotic lock in order to prevent development of CRBI. Monitorization was performed with ILAO and ILC every 15 days. A positive result involved a new QBC extraction through catheter and peripheral vein. There were 28 CC episodes in 15 patients, with 1.52 CC episodes per 1000 catheter-days. Main isolate involved was S. epidermidis. There were 16 CRBI episodes in 12 patients, with a CBRI incidence of 1.01 episodes per 1000 catheter-days. S. epidermidis was involved in 10 episodes and S. aureus in 5 episodes. Incidence of catheter related sepsis was 0.57 episodes per 1000 catheter-days. Time between CSC or CRBI episodes decreased within consecutive episodes. Arterial lumen was more likely to have positive cultures than venous lumen. Teicoplanin lock (10 mg/mL) therapy in colonized catheter success rate was 43.2%. Relapse or reinfection was evaluated with pulsed-field gel electrophoresis. A relapse rate of 32.4% and a failure rate of 16.2% were observed. Catheter removal rate was 0.38 catheter per 1000 catheter-days. In a historical control group, from January 2001 to June 2003, the rate of catheter related sepsis was 1.04 episodes per 1000 catheter-days and catheter removal rate was 0.43 catheter per 1000 catheter-days. A 42 days increase on catheter lifespan was observed during the monitorization and preventive therapy period. Teicoplanin lock therapy, on colonized catheters, was successful decreasing CRBI rate, catheter removal rate and increasing catheter lifespan. There were no statistically significant differences between the isolates¿ MICs before and after teicoplanin lock therapy. Antibiotic lock therapy did not affect bacterial antibiotic resistance patterns. Catheter-sparing diagnostic techniques were useful both in animal model and hemodialysis patients. ROC analysis showed that the optimum thresholds were 1500 cfu/mL for the ILC (88.2% sensitivity and 89.7% specificity), and 6 microorganisms/field for the ILAO (84.3% sensitivity and 92.8% specificity). Risk factors for CSC and CRBI were assessed. S. aureus nasal carriage was a significant risk factor for S. aureus CRBI in our hemodialysis patients.