We have read the recent letter by Honore et al. [1] about our findings published in this journal regarding the influence of continuous renal replacement therapy (CRRT) on the pharmacokinetics of ceftolozane-tazobactam (C/ T) [2]. In our report, we decided to administer a 3 g/iv dose every 8 h taking into account two previous studies referenced in our paper [2] and another one which showed CRRT to be an independent predictor of clinical failure (OR 4.5, 95% CI 1.18–17.39, p = 0.02) when C/T is administered at 1.5 g every 8 h [3]. As Honore et al. explain in their paper, the C/T eliminitation was assumed by hemodiafiltration and the adsorption was not assessed [1]. However, there is a misunderstanding in this letter [1], because we used a polysulphone membrane (Fresenius, Germany) instead of an acrylonitrile 69 Multiflow (AN-69-M). In contrast to highly adsorptive membranes (HAM; e.g., AN69 surface-treated, AN69-ST), the antibiotic adsorption with polysulphone ones is negligible, which facilitates antibiotic adaptation during CRRT [4]. Our data should not be extrapolated to other clinical scenarios, as noted by Honore et al. [1]. In our report, ceftolozane and tazobactam plasma concentrations remained above the minimal inhibitory concentration (MIC), for MICs of up to 8 μg/mL, but we estimated that the administration of standard doses of 1 g/0.5 g, even with polysulphone membranes, could compromise the effectiveness of C/T for not reaching adequate tazobactam concentrations. Thus, the use of HAM would represent a real risk factor of clinical failure when a C/T dose of 1.5 g every 8 h is administered, especially in multidrug-resistant infections [3]. Therefore, we agree with Honore et al. [1] that therapeutic drug monitoring (TDM) is critical when using C/T for patients receiving CRRT, especially when MICs of bacteria like multidrugresistant (MDR) Pseudomonas aeruginosa are considered very high. However, the recommendation of continuous (over 24 h) vs extended (over 2 to 4 h) or intermittent (over 30 to 60min) infusion of beta-lactams is still under debate [5].

中文翻译：

---

在连续肾脏替代治疗期间优化头孢唑烷-他唑巴坦剂量：一些细微差别

我们已经阅读了 Honore 等人最近的来信。[1] 关于我们在本期刊上发表的关于连续性肾脏替代疗法 (CRRT) 对头孢唑烷-他唑巴坦 (C/T) 药代动力学影响的研究结果 [2]。在我们的报告中，考虑到我们论文中引用的两项先前的研究 [2] 和另一项表明 CRRT 是临床失败的独立预测因子（OR 4.5，95%），我们决定每 8 小时给予 3 g/iv 剂量。 CI 1.18–17.39，p = 0.02)，当 C/T 以 1.5 g 每 8 小时给药时 [3]。正如奥诺雷等人。在他们的论文中解释说，C/T 消除是通过血液透析滤过假设的，并且没有评估吸附 [1]。然而，这封信中有一个误解 [1]，因为我们使用了聚砜膜（Fresenius，德国）而不是丙烯腈 69 Multiflow (AN-69-M)。与高吸附性膜（HAM；例如，AN69 表面处理的 AN69-ST）相比，聚砜膜的抗生素吸附可以忽略不计，这有助于 CRRT 期间的抗生素适应 [4]。正如 Honore 等人所指出的，我们的数据不应外推到其他临床情况。[1]。在我们的报告中，头孢洛扎和他唑巴坦的血浆浓度保持在最低抑菌浓度 (MIC) 以上，MIC 高达 8 μg/mL，但我们估计，即使使用聚砜膜，标准剂量为 1 g/0.5 g，由于未达到足够的他唑巴坦浓度，可能会影响 C/T 的有效性。因此，当每 8 小时施用 1.5 g C/T 剂量时，使用 HAM 将代表临床失败的真正危险因素，尤其是在多重耐药性感染中 [3]。所以，我们同意 Honore 等人的观点。[1] 对接受 CRRT 的患者使用 C/T 时，治疗药物监测 (TDM) 至关重要，特别是当细菌如多重耐药 (MDR) 铜绿假单胞菌的 MIC 被认为非常高时。然而，关于持续（超过 24 小时）与延长（超过 2 至 4 小时）或间歇（超过 30 至 60 分钟）输注 β-内酰胺类药物的建议仍在争论中 [5]。