Pharmacokinetics of high-dose tigecycline in critically ill patients with severe infections

This was a prospective, observational study that was performed between 2015 and 2018 in the 20-bed ICU of a 1500-bed teaching hospital in Rome, Italy. The protocol was approved by the Catholic University’s Ethical Committee (approval number Prot.sf 8431/13). Written informed consent was obtained from the patients’ legally authorized representative. Critically ill adult patients were considered eligible for the study when the attending physician prescribed TGC as empirical treatment (within 12 h from microbiological sampling) of a possible MDR infection, or as targeted therapy based on definitive results, in the absence of any exclusion criteria: known TGC allergy, creatinine clearance less than 40 mL/min (calculated according to the Cocrockft–Gault formula) apart from those ones who were anuric and on continuous renal replacement therapy (CRRT), hyperbilirubinemia (bilirubin level higher than 3 mg/dL), severe hepatic failure (Child–Pugh C), little chance of survival as defined by the Simplified Acute Physiology 2 (SAPS 2) score > 80, concomitant treatment with other drugs that can potentially interfere with TGC (i.e., rifampin and cyclosporine). Patient without microbiologically confirmed infection were not excluded. TGC was administered intravenously at loading dose (LD) of 200 mg over 30-min, followed by 100 mg over 30-min bid. On day 4 after the commencement of the HD TGC, at steady state, pharmacokinetic analyses of the study group were performed. Clinical and demographic data were recorded upon enrolment. Safety and adverse events were determined through the observed biochemical abnormalities, documented according to the Department of Health and Human Services–Common Terminology Criteria for Adverse Events (DHHS-CTCAE v.3.0) classification [12].

Clinical cure was defined as the complete resolution of all signs and symptoms of the infection by the end of TGC therapy. In case of ventilator-associated pneumonia (VAP), improvement or lack of progression of all abnormalities on chest radiographs was also required [13]. Otherwise, the outcome was classified as treatment failure. Clinical outcomes were independently evaluated by two physicians (GDP, MSV) when judgments were discordant (two cases), the reviewers reassessed the data and reached a consensus decision. The quality of source control was considered adequate when it included drainage of infected fluid collections, debridement of infected solid tissue, removal of devices/foreign bodies, and definitive measures to correct anatomic derangements resulting in on-going microbial contamination and to restore optimal function within 48 h after diagnosis [14].

Although TGC concentration–time profiles are stable dose just on day 3, blood samples were collected after the seventh dose (on day 4 of treatment) at T0 (immediately before the initiation of the infusion) and 1, 1.5, 2, 4, 6, 8, 10, and 12 after the start of infusion. According to patients’ respiratory status, one mini-bronchoalveolar lavage (BAL) (40 mL sterile 0.9% saline solution was blindly instilled through a telescopic catheter and immediately aspirated in a trap) was performed on day 4, in case of suspected HAP.

Stock solution of TGC and the internal standard (IS), propranolol hydrochloride, were prepared by dissolving accurately weighed amounts of each compound in MeOH to obtain a final concentration of 0.1 mcg/mL. Calibration standards were prepared by diluting stock solutions of TGC in drug-free human plasma to yield TGC concentrations of 5000, 2500, 1250, 625, 312.5, 156.25, 78.125, 39.1, 19.5 and 9.76 ng/mL.

Tigecycline liquid/liquid extraction from plasma samples (see Additional file 1).

Tigecycline solid-phase extraction from BAL samples (see Additional file 1).

Chromatographic and Mass-Spectrometric Conditions (see Additional file 1)

A one-compartment model with first-order elimination determined pharmacokinetic parameters. The 0–12 h area under the time–concentration curve (AUC_0–12) was determined by the linear trapezoidal rule. TGC AUC_0–24 was calculated as AUC_0–12 X 2. TGC maximum and minimum concentrations (C_max, C_min) were directly obtained from observed peak and trough concentrations. Epithelial lining fluid (ELF) tigecycline (TGC_ELF) concentration was calculated from BAL concentration (TGC_BAL) using urea as dilution marker: TGC_ELF = TGC_BAL X urea dilution index (plasma urea concentration/BAL urea concentration) [15]. In all patients, distribution volume (Vd), drug clearance (CL), and elimination half-life (t_1/2) were calculated after a single 100-mg intravenous dose at steady state.

According to previous literature, based on early animal efficacy studies using a classification and regression tree approach, area under the concentration curve (AUC) _0–24/MIC ratio ≥ 4.5, 6.96 and 17.9 were used as PD targets for VAP, intra-abdominal infections (IAI) and skin–soft-tissue infections (SSTI), respectively [6]. Graphing of data was undertaken using Prism version 6.0 for Windows (graphPad Software, San Diego, CA).

Isolates were identified by matrix-assisted laser desorption ionization-time-of-flight (MALDI-TOF) mass spectrometry (MALDI Biotyper, Bruker Daltonics GmbH, Leipzig, Germany). The in vitro susceptibility of the isolates was assessed with the Vitek 2 system (bioMérieux, Marcy l’Etoile, France) or with panels manufactured by MERLIN Diagnostica GmbH (Bornheim, Germany). Results were interpreted in accordance with the European Committee on Antimicrobial Susceptibility Testing (EUCAST) clinical breakpoints. The presence of carbapenemase genes of blaKPC, blaNDM, blaVIM, blaOXA-48, blaOXA-23, and blaOXA-58 types was determined by polymerase chain reaction and DNA sequencing analysis using previously described protocols (Endemiani, Poirel, Woodford) [16].

All statistical analyses were performed using MedCalc software, version 12.2.1 (MedCalc^®, MariaKerke, Belgium). Kolmogorov–Smirnov test was used to value the variables distribution. The data with a non-Normal distribution were assessed with Mann–Whitney test and the median and selected centiles’ (25th–75th) value were given (interquartile range, IQR). The data with a Normal distribution were assessed with Student’s test. Categorical variables are presented as proportions and were analysed with the use of the Chi square test or Fisher’s exact test, as appropriate. A p value < 0.05 was considered significant. Due to the PK/PD design of the study, a sample size was not calculated, foreseeing the recruitment of at least 30 patients during the predefined study period (July 2015–July 2017).

The clinical details of the 32 non-obese patients included in the study are listed in Table 1. Albumin levels were quite low with an overall positive fluid balance at enrolment. Median SAPS II score was 53.5 and the most relevant comorbidities were cardiovascular diseases, chronic obstructive pulmonary disease, chronic renal failure and neoplasm (Table 1). Median SOFA score was 7 and many patients were in septic shock or presented with acute respiratory failure (ARF) and acute kidney injury (AKI) requiring mechanical ventilation (MV) and CRRT, respectively. More than half of the patients had VAP, followed by intra-abdominal infections and skin and soft-tissue infections: in the microbiological case-mix Gram-negative bacteria were mostly represented (90.6%). Median duration of TGC therapy was 12 days and it was started empirically in half of the cases. The use of vasopressors and MV during TGC therapy was high and 30-day mortality rate was 28.1%. Eleven patients had a treatment failure, whilst the other 21 successfully eradicated the infection. There were no between-group differences in terms of demographic aspects and main comorbidities. Further the two groups were similar in terms of presenting features and outcomes with the exception of VAP rate which was higher in the treatment success group (76.2% vs. 27.3%, p = 0.02) and a trend to a higher percentage of skin and soft-tissue infections and source control amongst patients who failed TGC treatment (p = 0.06 and p = 0.07, respectively).

A one-compartment model with first-order disposition processes adequately described the concentration–time curve, although significant interindividual variability was observed. Vd, Cl and t_1/2 were 438.6 L, 42.1 L/h and 7.2 h, respectively. Median and IQR values of C_max and C_min were 0.34 [0.15–1.03] mcg/mL and 0.09 [0.05–0.26] mcg/mL (Table 1). Figure 1 shows the mean ± SD time–concentration profile at different time points of plasma tigecycline concentrations, compared with most frequently observed MIC values (0.12–0.25–0.5 mcg/mL). AUC_0-24 and IQR were calculated for each patient and the percentage of target attainment was also computed for nosocomial pneumonia (NP) (AUC_0-24/MIC breakpoint of 4.5), complicated intra-abdominal infections (cIAI) (AUC_0-24/MIC breakpoint of 6.96) and complicated skin and soft-tissue infections (cSSTI) (AUC_0-24/MIC breakpoint of 17.9). Considering a target AUC_0-24/MIC of 4.5, 75% of the patients would be successfully treated in presence of 0.5 mcg/mL MIC value and all the patients obtained the PK target with MIC ≤ 0.12 mcg/mL. Considering a target AUC_0-24/MIC of 6.96, at least 50% of the patients would be adequately treated against bugs with MIC ≤ 0.5 mcg/mL, whilst only 15.6% obtained the PK target with MIC of 2 mcg/mL. Finally, with a target AUC_0-24/MIC of 17.9 only 25% of the patients obtained the PK target at MIC values of 0.5 mcg/mL and less than 10% were adequately treated against germs with MIC value ≥ 1 mcg/mL (Fig. 2).

Tigecycline pulmonary concentrations were measured in 12 (1 h) and 7 (12 h) patients, respectively. Main reasons to exclude samples were presence of blood and excessive dilution, whilst five patients did not undergo BAL due to severe respiratory failure. Median and IQR ELF C_max was 0.42 [0.15–1.2] and ELF C_min was 0.32 [0.17–0.43]; median and IQR ELF/plasma ratio (%) was 152.9 [73.5–386.8] (Table 2). Mean ± SE ELF concentrations were similar at 1 h and 12 h (0.78 ± 0.2 mcg/mL vs. 0.36 ± 0.1 mcg/mL; p = 0.19) (Fig. 3). Conversely, no significant differences were found comparing mean ± SE ELF/plasma ratio at 1 h and 12 h (281 ± 107.6 vs. 298.3 ± 60.7; p = 0.9) (Fig. 4).