Cefepime MIC Breakpoint Resettlement in Gram-Negative Bacteria

Cefepime has recently drawn much attention, due mostly to a meta-analysis reported by Yahav et al. (3). They observed higher all-cause mortality for cefepime than for other beta-lactam antibiotics (risk ratio, 1.26; 95% confidence interval [CI], 1.08 to 1.49) and described neurotoxic adverse effects and inadequate in vivo antimicrobial efficacy as plausible reasons for increased mortality. Bhat et al. (1) observed increased mortality among cefepime-treated patients with bacteremia caused by gram-negative organisms when the cefepime MIC was ≥8 μg/ml (54.8%; 17 of 31 died) than when the MIC was <8 μg/ml (24.1%; 35 of 145 died). Based on pharmacodynamic and clinical grounds, they suggested that the current breakpoints (according to which organisms are considered susceptible if the MIC is ≤8 μg/ml by Clinical and Laboratory Standards Institute standards) for cefepime be lowered in countries where cefepime dosages of 1 to 2 g every 12 h is the licensed therapy for serious infections. However, we suppose that subgroup analysis excluding Pseudomonas aeruginosa and Acinetobacter spp. should reveal consistent results with statistical evidence to generalize their contention for other gram-negative organisms, because the MICs for P. aeruginosa and Acinetobacter spp. have been revealed to be much higher than those for other gram-negative organisms. Bhat et al. presented details on 204 bloodstream isolates from individual patients (the clinical outcomes for only 176 patients were analyzed because 21 patients were discharged within 28 days and 7 patients had two episodes of bacteremia), and P. aeruginosa and Acinetobacter spp. accounted for 79.4% (27 of 34) of the isolates for which the MICs were ≥8 μg/ml while they constituted only 15.9% (27 of 170) of those for which the MICs were <8 μg/ml. We could not identify the exact proportions of these two pathogens among isolates from nonsurvivors for which MICs were ≥8 μg/ml. However, Bhat et al. observed a high odds ratio (OR) for mortality among P. aeruginosa bacteremic patients through a subgroup analysis that compared outcomes associated with MICs of ≥8 μg/ml (mortality, 59.1%; 13 of 22 patients died) and those associated with MICs of ≤4 μg/ml (mortality, 20.8%; 5 of 24 patients died), and when a reconstituted population excluding P. aeruginosa was analyzed on the basis of MICs of ≥8 μg/ml versus MICs of ≤4 μg/ml, the analysis did not reveal statistical significance (4 of 9 patients [44.8%] with isolates for which MICs were ≥8 μg/ml died, and 30 of 121 [24.8%] with isolates for which MICs were <8 μg/ml died; OR = 2.4; 95% CI, 0.6 to 9.6). It seems quite reasonable to reconsider the current breakpoint MIC of cefepime (8 μg/ml) for P. aeruginosa. However, additional verification is required to resettle the cefepime MIC breakpoint for gram-negative pathogens other than P. aeruginosa and Acinetobacter spp. We believe that this verification may be accomplished through a subgroup analysis excluding P. aeruginosa and Acinetobacter spp. and providing statistical evidence. In 2006, the probability of target attainment (PTA) for the conventional dose of cefepime (2 g every 12 h) against common intensive care unit (ICU) pathogens in ICU patients was estimated (2). According to the results of the study, higher doses of cefepime (>4 g/day) are required to achieve the required PTA expectation value for P. aeruginosa, and even a higher dose (6 g/day) failed to achieve the bactericidal target for Acinetobacter baumannii, unlike that for other pathogens. Like various MIC distributions, we suppose that MIC breakpoint resettlement requires verifications for different pathogens.