Should Lock Therapy Always Be Avoided for Central Venous Catheter–Associated Fungal Bloodstream Infections?

Sir—One section of the recently published guidelines on the management of intravascular catheter—related infections [1] discussed the significance of the so-called “antibiotic lock” technique—that is, filling the catheter lumen with pharmacological concentrations of antibiotics and leaving them there for several hours—and raised several issues that we would like to address. Although small open trials have reported both a response and catheter salvage in >80% of cases of bacterial catheterrelated bloodstream infection (CRBSI), attempts to salvage catheters in cases of fungal CRBSI reportedly failed [2–5]. We recently administered antifungal lock therapy associated with systemic antifungal treatment to 2 patients who had Candida albicans CRBSI. Both patients were hospitalized and had a Groshong catheter that could not be removed, in one case due to coagulation disorders and in the other case due to opposition from the patient. The diagnosis of CRBSI was made, according to the method described by Blot et al. [6], on the basis of a finding of significantly earlier growth of C. albicans in a culture of a blood sample from the central venous catheter (CVC) than in a culture of a blood sample from the peripheral venipuncture site. In addition, both patients had fever without systemic inflammatory response syndrome [7]. Patient 1 was a 2-year-old girl with posttubercular obstructive hydrocephalus in whom a C. albicans CRBSI developed on day 40 after catheterization. She was treated with amphotericin B (AmB) desoxicolate both by use of the lock technique (concentration, 2.5 mg/mL for 12 h) and, systemically, by increasing the dosage from 0.3 to 1 mg/kg per day. Parenteral AmB was administered for 7 days. Fever resolved on day 3 of treatment. The patient completed a 14-day full course of lock treatment, and there was no relapse in the 6 months after catheterization. Patient 2 was a 65-year-old man with a large intestinal resection secondary to ischemia, in whom a C. albicans CRBSI developed on day 25 after catheterization. Due to moderate renal impairment (creatinine clearance rate, 42 mg/mL per day), systemic treatment with AmB was not considered, and the patient was treated with AmB desoxicolate, by use of the lock technique (concentration, 2.5 mg/mL), and with parenteral fluconazole (400 mg q.d.). The patient's temperature decreased significantly after 48 h of treatment, and fever completely disappeared on day 4 of treatment. Fluconazole treatment was stopped after 7 days, and AmB lock therapy was continued until day 14; there was no relapse in the 2 months after catheterization. AmB-related side effects were not observed in either patient. For both patients, 2 cultures of blood samples drawn from the CVC at the end of lock treatment had negative results. Even though caution should be taken when treating fungal CRBSI with use of the lock technique, our case reports suggest that the recommendation of catheter removal in case of fungal CRBSI should not be absolute. Consequently, we believe that, in cases of less severe CRBSI in patients without systemic inflammatory response syndrome and/or severe immunosuppression, use of the lock technique should be considered. In conclusion, as a first step, we would propose that, for selected patients with CRBSI, clinicians should attempt to salvage the catheter by using the lock technique for up to 72 h, to aid in the process of deciding whether to remove or maintain the catheter.