Pharmacokinetics and Penetration of Sitafloxacin into Alveolar Epithelial Lining Fluid in Critically Ill Thai Patients with Pneumonia

Sitafloxacin showed potent activity against various respiratory pathogens. Blood and bronchoalveolar lavage (BAL) fluid samples were obtained from 12 subjects after a single oral dose of sitafloxacin 200 mg. The mean ± SD (median) maximum ratio of epithelial lining fluid (ELF) to unbound plasma concentration was 1.02 ± 0.58 (1.33).

ABSTRACT

Sitafloxacin showed potent activity against various respiratory pathogens. Blood and bronchoalveolar lavage (BAL) fluid samples were obtained from 12 subjects after a single oral dose of sitafloxacin 200 mg. The mean ± SD (median) maximum ratio of epithelial lining fluid (ELF) to unbound plasma concentration was 1.02 ± 0.58 (1.33). The penetration ratios based on the mean and median area under the curve from 0 to 8 h (AUC_0–8) were 0.85 and 0.79 μg · h/ml, respectively. Sitafloxacin penetrates well into ELF in critically ill Thai patients with pneumonia. (This study has been registered in the Thai Clinical Trials Registry [TCTR] under registration no. TCTR20170222001.)

TEXT

Sitafloxacin has been shown to have good in vitro activity against pathogens causing nosocomial pneumonia, especially Acinetobacter baumannii (1–3). Plasma pharmacokinetics (PK) of sitafloxacin exhibited good absorption and distribution into various tissues (4–7). However, sitafloxacin concentrations in epithelial lining fluid (ELF) have not been evaluated in humans. To determine the ELF concentrations and PK parameters, patients were enrolled to collect blood and bronchoalveolar lavage (BAL) fluid samples from June 2017 to May 2018. The protocol was approved by the Institutional Review Board (IRB) of the Faculty of Medicine, Ramathibodi Hospital, Mahidol University (MURA2016/646). The study was terminated by the IRB before the completion of subject enrollment because of a concern for increased mortality in patients with BAL fluid; these patients admitted to the intensive care unit (ICU) had a high risk of death. However, most subjects died from disease progression. Eligible criteria included patients aged ≥18 years admitted to the ICU, Ramathibodi Hospital. All patients had to be diagnosed with nosocomial pneumonia suspected to be caused by Gram-negative bacilli and on a ventilator. Subjects were enrolled after written informed consent from their closest relative. Patients were excluded if they were known to be hypersensitive to any quinolone antibiotics or had a baseline QT interval of >480 ms, HIV infection, or contraindication for bronchoscopy and BAL fluid sampling. Each subject received 200 mg of oral sitafloxacin (four 50-mg tablets), with 240 ml water given via tube under fasting conditions.

(This research was presented in part at the 28th European Congress of Clinical Microbiology and Infectious Diseases [ECCMID], Madrid, Spain, April 2018, and ID Week, San Francisco, CA, October 2018.)

Blood samples (3 ml) were collected into EDTA tubes before (time zero) and after administration for 0.5, 1, 2, 3, 8, and 12 h (main schedule). Some blood samples needed to be collected by BAL fluid sampling times. After that, the blood samples were immediately placed on ice and centrifuged at 3,000 × g for 10 min at 4°C. The plasma was removed and stored at −20°C. Standard fiberoptic bronchoscopy and BAL fluid sampling were performed by a pulmonologist once in each subject during 0.5 to 2, 3 to 4, 5 to 6, and 7 to 9 h (8). All samples were stored at −20°C until analysis for up to 6 months. The apparent volume of ELF in BAL fluid was determined by the urea dilution method (9). All samples were assayed at the Center of Analysis for Product Quality (CAPQ) Faculty of Pharmacy, Mahidol University. The bioanalytical procedure was developed and validated using liquid chromatography-tandem mass spectrometry analysis. The validated method was used to assay the amount of sitafloxacin by LCMS-8040 (serial no. 0105752; Shimadzu Corp.). Chromatographic separation was achieved on a reverse-phase C₁₈ column (ZORBAX SB C-18, 3.0 × 150 mm, 3.5 μm; Agilent Technologies, Inc.) and Poroshell 120 guard (SB-C18, 2.1 × 5 mm, 2.7 μm; Agilent Technologies, Inc.). The multiple-reaction monitoring transitions monitored were m/z 410.10 to 393.15 for sitafloxacin, and m/z 402.20 to 384.20 for moxifloxacin (internal standards). The mixture samples and moxifloxacin were applied to a solid-phase extraction cartridge (Oasis HLB, 60 mg/1 ml; Waters Corp., Milford, MA) washed with 1 ml of 5% methanol. Elution was done with 1 ml of 0.1% trifluoroacetic acid in methanol and then filtrated by a 0.22-μm filter membrane. Sitafloxacin quantification was assayed by weighted (1/X²) least-squares linear regression analysis of the peak area ratio of sitafloxacin to moxifloxacin. Plasma and BAL fluid standard curves were linear (r² > 0.99) over the concentration range 0.0025 to 0.5 μg/ml. The lower limit of quantification in plasma and BAL fluid was 0.0025 μg/ml. The intraday and interday precisions and accuracies in plasma were within 8.87% and ±12.26%. The intraday and interday precisions and accuracies in BAL fluid were within 8.15% and ±10.88%. Plasma and BAL fluid samples were stable up to 6 months when stored at −20°C, where the percentage of relative error was within 8.64 and 6.77, respectively. Urea concentrations were determined using the urea assay kit performed on Siemens Dimension EXL 200 (Siemens Healthcare Diagnostics, Inc.).

PK parameters were estimated by noncompartmental analysis using Phoenix WinNonlin (version 8.0) software (Certara). Area under the concentration-time curve (AUC) in ELF was determined from the composite of mean and median concentration values in each BAL fluid sampling time. The AUC_0–8 was calculated from the data from 7 to 9 h. Penetration was estimated for ratios of ELF to simultaneous plasma concentrations and the ratio of AUC_0–8 for ELF to AUC_0–8 for unbound plasma. Unbound plasma concentrations were calculated from the fraction unbound (f_u) related to serum albumin concentration in each subject as follows: f_u = 1/(1 + K_a · f_up · P_t), where K_a is the association constant, f_up is unoccupied binding sites, and P_t is total protein concentration (albumin) (10). The concentrations in ELF were assumed to be the unbound drug.

Most of the patients had schedule changes in blood sampling times as follows: 0.67, 3.2, 4, 5, 5.4, 5.8, 8.2, and 9 h. Characteristics of the patients are shown in Table 1. The plasma PK parameters are summarized in Table 2. The elimination half-life could not be predicted in two patients due to the insufficiency of observed concentrations. AUC_0–8 based on total plasma concentration was transformed to unbound AUC_0–8 determined by the average f_u (mean ± SD, 0.63 ± 0.04). The concentration-time profiles and individual concentrations are illustrated in Fig. 1 and 2, respectively.

TABLE 1.

Patient characteristics

Characteristic	Value
Male (n [%])	6 (50)
Age (mean, median [range] yr)	54, 57 (26–75)
Weight (mean, median [range] kg)	55, 52 (43–76)
APACHE II score (mean, median [range])	25, 21 (8–50)
Creatinine clearance (mean, median [range] ml/min)	76.25, 68 (9–235)
Albumin (mean, median [range] g/dl)	2.04, 2.0 (1.4–2.9)

TABLE 2.

PK parameters estimated from noncompartmental analysis in total plasma

Subjects	Tmaxa (h)	Cmax (μg/ml)	V/Fb (liter/kg)	CL/Fc (ml/min)	keld (h−1)	t1/2e (h)	AUC (μg · h/ml) at (h):
							0–8	0–12	0–24	0–∞
Mean	3.88	1.41	3.13	246.13	0.12	9.15	8.86	13.38	14.76	19.96
SD	2.46	0.69	2.51	136.00	0.09	7.11	7.36	12.85	8.32	18.67
Median	3.50	1.54	2.35	200.86	0.10	6.64	7.85	10.82	13.39	16.64
Geometric mean	3.21	1.22	2.42	210.41	0.10	7.21	6.87	10.01	13.10	15.84

^aT_max, time to reach maximum concentration.

^bV/F, apparent volume of distribution.

^cCL/F, apparent oral clearance.

^dk_el, elimination rate constant.

^et_1/2, elimination half-life.

FIG 1.

The median of sitafloxacin concentration-time profiles in plasma and ELF after a single dose of 200 mg during 0.5 to 2, 3 to 4, 5 to 6, and 7 to 9 h.

FIG 2.

Individual concentrations of sitafloxacin in plasma (right panel) and epithelial lining fluid (left panel).

This is the first study to determine ELF concentrations of sitafloxacin in humans. This study did not determine alveolar macrophage (AM) concentrations because we intended to determine the efficacy of sitafloxacin for treatment of extracellular pathogens, e.g., Streptococcus pneumoniae and A. baumannii. Moreover, financial support was too limited for analysis of AM concentrations. Our study found that the mean ± SD (median) maximum ratio was 1.02 ± 0.58 (1.33), and penetration ratios based on the mean and median AUC_0–8 were 0.85 and 0.79 μg · h/ml, respectively (Tables 3 and 4). The results indicate that sitafloxacin penetrates well into ELF. However, lower penetration was seen compared with that of other fluoroquinolones (8). This may be the result of the variability of plasma and ELF concentrations observed in critically ill patients compared with those in healthy subjects due to pathophysiological changes. Additionally, severe pneumonia or pathological lung disease may influence drug diffuse through alveolar epithelial cells (11–15). We found wide ranges in volume of distribution (V) and clearance (CL) rates, resulting in variability of concentrations (Table 2). Alterations in renal functions and V associated with critical illness may affect drug concentrations. The AUC/MIC target related to response of therapy for Gram-negative bacilli did not exceed the 125 target (16). The reported MIC₉₀ of sitafloxacin against A. baumannii clinical isolates in Thailand was 2 μg/ml (2). The estimated plasma AUC_0–24 (median, 13.39 μg · h/ml)/MIC₉₀ of sitafloxacin in ELF was 6.70 for A. baumannii isolates (MIC₉₀, 2 μg/ml). To exceed that target in ELF, the MIC needs to be <0.1 μg/ml. However, the AUC_0–24 of ELF could not be accurately calculated because we collected BAL fluid from 0 to 8 h. The reason for the lack of data for BAL fluid at 12 h was that the study was terminated before full recruitment. For maximum concentration (C_max), both plasma and ELF concentrations did not achieve PK/pharmacodynamic (PD) targets. Therefore, we recommend combination of sitafloxacin with other antibiotics, for example, colistin (17).

TABLE 3.

Concentrations of sitafloxacin in plasma and epithelial lining fluid

Time (h)	Concentration [mean ± SD (median) μg/ml]			ELF to plasma ratio [mean ± SD (median)]
	Total plasma	Unbound plasma	ELF	Total	Unbound
0.5–2	0.60 ± 0.46 (0.43)	0.39 ± 0.28 (0.28)	0.19 ± 0.20 (0.11)	0.30 ± 0.20 (0.37)	0.45 ± 0.28 (0.59)
3–4	1.37 ± 0.51 (1.56)	0.81 ± 0.34 (0.94)	0.48 ± 0.29 (0.55)	0.32 ± 0.11 (0.35)	0.54 ± 0.17 (0.59)
5–6	1.99 ± 2.34 (1.01)	1.32 ± 1.58 (0.64)	1.07 ± 0.93 (1.12)	0.63 ± 0.42 (0.42)	0.98 ± 0.66 (0.63)
7–9	0.85 ± 0.77 (0.67)	0.55 ± 0.49 (0.43)	0.61 ± 0.77 (0.17)	0.67 ± 0.38 (0.89)	1.02 ± 0.58 (1.33)

TABLE 4.

Pharmacokinetic parameters of sitafloxacin in plasma and epithelial lining fluid^a

Sample	Tmax (h)a	Cmax (μg/ml)a	Unbound AUC0–8 (μg · h/ml)b	ELF/unbound plasmac Cmax ratiob	ELF/unbound plasmac AUC0–8 ratiob
Plasma	3.5	1.54	5.58 (4.95)
ELF	5.4	1.12	4.77 (3.90)	0.73 (1.15)	0.85 (0.79)

^aData presented as medians.

^bData presented as means (medians).

^cUnbound fraction of sitafloxacin in plasma is 0.63.

The study had some limitations, such as the number of BAL fluid samplings from each subject, including lack of BAL fluid samples during 12 h. A small sample size and variability of the ELF concentrations in critically ill patients can affect the estimated AUC. Moreover, we did not collect the sample at steady state. Technique errors in the BAL fluid sampling procedure may have altered ELF volume estimation by the urea dilution method. Further PK/PD and clinical studies are needed to confirm clinical efficacy.