Abstract Background: Burkholderia cepacia complex (BCC) is a multidrug-resistant nosocomial pathogen associated with poor outcomes. Recent Clinical and Laboratory Standards Institute revisions have abolished clinical breakpoints (CBPs) for BCC, replacing them with epidemiological cutoff values (ECVs), creating uncertainty in interpreting antimicrobial susceptibility. This study evaluated the impact of these changes on susceptibility reporting and clinical outcomes of BCC bacteremia. Methods: A retrospective study was conducted at a tertiary-care center from July 2020 to June 2025. A total of 128 blood culture isolates underwent minimum inhibitory concentration (MIC) testing by broth microdilution. Susceptibility interpretation using pre-2025 CBPs was compared with the 2025 ECV-based classification. Clinical data for 91 patients were analyzed to determine treatment patterns, outcomes, and predictors of mortality. Results: Reclassification using ECVs led to additional isolates being categorized as wild-type for ceftazidime (9.4%), meropenem (10.1%), levofloxacin (9.2%), and minocycline (12%), with no change for trimethoprim–sulfamethoxazole (TMP-SMX). Clinical cure was achieved in 73.6% and microbiological cure in 69.1% of evaluated episodes. Day-14 and day-28 mortality were 14.3% and 23.1%, respectively. The lowest 28-day mortality was observed with TMP-SMX (0%), followed by minocycline (17.6%), levofloxacin (25%), and meropenem (26.5%). On multivariate analysis, septic shock was the only independent predictor of 28-day mortality (odds ratio 4.223; P = 0.008). Conclusions: BCC bacteremia is associated with high mortality. ECV-based reporting increases the proportion of isolates considered wild-type for most antimicrobials except TMP-SMX. Optimal therapy should incorporate MIC evaluation, pharmacokinetic-pharmacodynamic considerations, and infectious disease input. Prospective comparative studies are urgently needed to define the preferred agents and the role of combination therapy.
Prayag et al. (Thu,) studied this question.