Early Removal of Central Venous Catheter in Patients with Candidemia Does Not Improve Outcome: Analysis of 842 Patients from 2 Randomized Clinical Trials

Marcio Nucci () 2 3 Elias Anaissie 1 2 Robert F. Betts 0 2 Bertrand F. Dupont 2 5 Chunzhang Wu 2 6 Donald N. Buell 2 6 Laura Kovanda 2 6 Olivier Lortholary 2 4 5 0 University of Rochester , Rochester, New York 1 University of Arkansas for Medical Sciences , Little Rock 2 Received 12 February 2010; accepted 12 March 2010; electronically published 25 June 2010. Universita rio Clementino Fraga Filho, Universidade Federal do Rio de Janeiro , Rua Prof Rodolpho Paulo Rocco 255 Sala 4A 12-21941-013, Rio de Janeiro , Brazil 3 Hospital Universita rio Clementino Fraga Filho , Rio de Janeiro , Brazil 4 Institut Pasteur, Centre National de Re fe rence Mycologie et Antifongiques, Unite de Mycologie Mole culaire , Paris , France 5 Universite Paris Descartes, Hopital Necker-Enfants Malades, Service des Maladies Infectieuses et Tropicales, Centre d'Infectiologie Necker-Pasteur 6 Astellas Pharma , Deerfield, Chicago (See the editorial commentary by Brass and Edwards, on pages 304-306.) Background. Patients with candidemia frequently have a central venous catheter (CVC) in place, and its early removal is considered the standard of care. Methods. We performed a subgroup analysis of 2 phase III, multicenter, double-blind, randomized, controlled trials of candidemia to examine the effects of early CVC removal (within 24 or 48 h after treatment initiation) on the outcomes of 842 patients with candidemia. Inclusion criteria were candidemia, age 116 years, CVC at diagnosis, and receipt of 1 dose of the study drug. Six outcomes were evaluated: treatment success, rates of persistent and recurrent candidemia, time to mycological eradication, and survival at 28 and 42 days. Univariate and multivariate analyses were performed, controlling for potential confounders. Results. In univariate analysis, early CVC removal did not improve time to mycological eradication or rates of persistent or recurrent candidemia but was associated with better treatment success and survival. These benefits were lost in multivariate analysis, which failed to show any beneficial effect of early CVC removal on all 6 outcomes and identified Acute Physiology and Chronic Health Evaluation II score, older age, and persistent neutropenia as the most significant variables. Our findings were consistent across all outcomes and time points (removal within 24 or 48 h and survival at 28 and 42 days). The median time to eradication of candidemia was similar between the 2 study groups. Conclusions. In this cohort of 842 adults with candidemia followed up prospectively, early CVC removal was not associated with any clinical benefit. These findings suggest an evidence-based re-evaluation of current treatment recommendations. - Candidemia is a common nosocomial bloodstream infection and is associated with high mortality [1]. Patients with candidemia usually have a central venous catheter (CVC) in place, and prompt CVC removal is considered by some to be critical to a successful outcome. Recent guidelines for the management of candidiasis strongly recommend early CVC removal in all nonneutropenic patients with candidemia [2] or in patients with CVC-related candidemia [3]. These recommendations were based on selected studies suggesting that prompt CVC removal was associated with improved outcomes, including better treatment success, faster mycological eradication, decreased rates of recurrent and persistent candidemia, and improved survival [47]. However, these studies had serious limitations: small sample size, retrospective data collection, inclusion of patients without candidemia, lack of a definition for what constitutes early CVC removal, and suboptimal statistical analyses including no adjustment for important potential host confounders, such as a high severity of illness score and persistent neutropenia. Not cited in these guidelines are studies that made such adjustments and failed to confirm the reported association between early CVC removal and improved outcomes [811]. The findings in these latter studies prompted some to recommend against CVC removal and replacement in critically ill patients because of the risks of serious complications [12]. The optimal strategy to resolve this controversial issue is a randomized, controlled trial (RCT) in which patients with candidemia are randomized to have early CVC removal versus no removal, after patient stratification by key baseline variables; uniform antifungal therapy would be provided, and serial blood cultures would be performed at predefined time points to ascertain the time to mycological eradication. Because such a study is unlikely to be conducted in the near future, the next best approach relies on appropriate subgroup analysis of a large cohort of patients enrolled in recent RCTs of candidemia in which the effect of early CVC removal on various outcomes is evaluated [13]. The newer and more precise methods for grading evidence-based medicine give such subgroup analyses higher-quality grading than purely observational studies [14], provided the following requirements are fulfilled: (a) a representative patient population for which these evidencebased recommendations are applied, (b) direct comparison of 2 groups (eg, CVC removal or retention), (c) evaluation of clinically important outcomes, and (d) statistically significant results in a positive study or adequate sample size to rule out a b error in a negative study [15]. We present the results of a study in which we evaluated the effects of early CVC removal on clinically important outcomes among 842 patients enrolled in 2 recent large multicenter, multinational RCTs of treatment of candidemia [7, 16]. We specifically examined whether early CVC removaldefined as removal within 24 or 48 h after initiation of antifungal therapy was associated with the beneficial outcomes that form the basis for current recommendations for early CVC removal [2, 3] namely, better treatment success, faster mycological eradication, lower rates of recurrent and persistent candidemia, and improved survival. Characteristic Inclusion criteria Exclusion criteria Recruitment period No. of centers (geographic locations) Duration of therapy (minimum; maximum) CVC recommendations Frequency of repeat blood cultures Duration of follow-up after start of therapy Stratification Method of randomization Primary end point Clinical success Mycological success Recurrence 2-Arm studya 3-Arm studyb Remove CVC before the first dose of therapy 3 times per week until sterilization of blood cultures 2 and 6 weeks No recommendation given Daily until sterilization of blood cultures 12 weeks Center and neutropenic status Computer-generated, at each center Rate of overall treatment successc Resolution of symptoms of candidemia Eradication of baseline candidemia Reemergence of baseline fungal infection (same species) during follow-up APACHE II score ( 20 vs 120) and region (North America, Europe, Brazil, and India) Same as the 2-arm study Same as the 2-arm study (...truncated)