Aerosolized delivery resulting in high polymyxin B concentration levels in epithelial lining fluid ensures efficacy in ventilator-associated pneumonia

Abstract

Background

Aerosolized polymyxin B delivery was a promising approach for the treatment of ventilator-associated pneumonia (VAP). However, there were little data on the concentrations of polymyxin B in epithelial lining fluid (ELF), which impedes the optimal use of aerosolized polymyxin B in clinics.

Methods

We present four cases of patients diagnosed with VAP caused by Gram-negative bacteria, who enrolled in a prospective, therapeutic drug monitoring (TDM) study of polymyxin B. The patients were treated with aerosolized and intravenous administration of polymyxin B. Polymyxin B concentrations in both ELF and plasma were determined using validated LC-MS/MS methods.

Results

All four patients achieved bacterial eradication, with three of them reaching clinical improvement or cure. Following aerosol administration (25 or 50 mg, q12h) and intravenous infusion (50−100 mg, q12h) of polymyxin B, it was observed that the concentrations of polymyxin B in ELF were significantly higher in ELF (20.6−97.6 mg/L) compared to those in plasma (1.19−5.16 mg/L) during the steady sate. The area under the concentration−time curve for 24 h (AUC_24h,ELF) ranged from 283.6 to 1872.9 mg•h/L.

Conclusions

This study presented polymyxin B concentrations in ELF following aerosolized delivery, supporting its clinical use from a PK/PD perspective. Following combined aerosol and intravenous administration, polymyxin B achieved notably higher concentrations in ELF than those observed in plasma.

Introduction

Ventilator-associated pneumonia (VAP) is commonly associated with invasive medical devices or surgical procedures, with an estimated attributable mortality rate of ∼10% or even higher among patients.¹ Gram-negative bacteria are responsible for ∼70% of VAP cases.² Polymyxins, which include polymyxin B and colistimethate (CMS, a pro-drug with colistin as the active component), are considered last-resort antibiotics for the treatment of Gram-negative bacterial infections. However, the intravenous administration of polymyxins has poor distribution into the epithelial lining fluid (ELF), inhalation of polymyxins was recommended.³ Aerosolized delivery of CMS has demonstrated significant advantages, including much higher ELF concentrations and increased rates of clinical cure (OR, 1.61) and microbiological eradication (OR, 1.37).⁴ The use of inhaled polymyxin B resulted in a favourable clinical outcome of cure and improvement in pneumonia patients.⁵ The ratio of the area under the ELF concentration curve to the minimum inhibitory concertation (AUC_ELF/MIC) is the most predictive pharmacokinetic/pharmacodynamic (PK/PD) index for describing the antimicrobial efficacy of aerosolized polymyxin B in the treatment of pulmonary infections in neutropenic mice.⁶ However, limited ELF concentrations for polymyxin B after aerosol treatment have been reported in clinics.

In the present study, we report four cases of VAP in which polymyxin B was administered through a combination of aerosolized and intravenous routes. Both bronchoalveolar lavage fluid (BALF) and plasma samples were collected to assess the concentrations of polymyxin B. Preliminary investigations into the PK and PK/PD of aerosolized polymyxin B were conducted.

Methods

Patient enrolment

Four patients who developed multi-drug-resistant Gram-negative bacteria infected VAP during stay in neurological ICU were enrolled in a prospective, TDM study of polymyxin B. The study was approved by the institutional review board of Huashan Hospital, Shanghai, China. Informed consent form was obtained from each patient’s next of kin before enrolling in the study. VAP was diagnosed according to the guidelines for the management of VAP patients.⁷ Each patient’s diagnosis was determined by a combination of positive sputum culture results of multi-drug resistant Gram-negative bacteria, the patient’s clinical manifestations, laboratory tests and CT imaging findings.

Drug treatments and sample collection

The four VAP patients received a combination of aerosol and intravenous administration of polymyxin B. For aerosol delivery of polymyxin B, 50 mg powder was dissolved in 5 mL of distilled water, and it was divided into two doses (25 mg each) or given as one dose. The nebulization was carried out using a NebuEz mesh nebulizer (model GUN300, GENTEC, Shanghai Corporation) for 30 min. BALF samples were collected from each patient during the steady state (after the third dose). The BALF samples were obtained at the steady state (following the third dose), and each patient had two or three samples collected: one before the next nebulization, as well as at 0.5 and 2 hours after nebulization. Specifically, for Patient 1, samples were collected at 0.5 and 2 hours after nebulization; for Patients 2 and 4, samples were collected before the next nebulization and at 2 hours after nebulization; and for Patient 3, samples were collected before the next nebulization, at 0.5 and 2 hours after nebulization. Mini bronchoalveolar lavage (BAL) was performed, and the lavage fluid was collected following previous protocols.⁸ Simply put, lavage was performed by rising through bronchoscopy with 20 mL of normal saline three times, and the second and third lavage fluids were combined and collected into sterile tubes with the first lavage fluid discarded. The lavage was evaluated by the total BALF volume recovery rate ≥30%. Then the lavage sample was centrifuged, and supernatant was collected, then adding bovine serum albumin (BSA) to a final concentration of 1% to avoid adsorption. Blood samples were obtained during the steady state (following the third intravenous dose) within 0.5 h both before and after the infusion. All samples were stored in a −70 °C fridge before LC-MS/MS analysis.

Polymyxin B concentration determination

Polymyxin B concentrations in plasma and BALF were determined by a previously validated LC-MS/MS method.⁹ The plasma sample was prepared by solid phase extraction with a 96-well Oasis WCX plate (Waters, MA, USA) according to the manufacturer’s protocol. The BALF sample was conducted by using a surrogate matrix (PBS containing 1% BSA).

Urea was served as an internal calibrate for calculations of polymyxin B concentrations in ELF. Urea levels in both plasma and BALF were determined using LC-MS/MS following a previously published method.¹⁰ To calculate polymyxin B concentrations in ELF, the concentrations in BALF samples were adjusted using the urea ratio of plasma to BALF, as described in a previous study.⁶

Pharmacokinetic parameters calculation

The calculation of the area under the plasma or ELF concentration curve (AUC) and elimination half-life (t_1/2) followed the first-order elimination method.¹¹

Results

Treatment outcome

The demographic information and baseline medical conditions, treatment regimens of polymyxin B, concomitant medications, microbiological, and clinical outcomes of the four patients are presented in Table 1. All patients were in a comatose state or under sedation during treatment. Bacterial cultures from sputum samples indicated the presence of Gram-negative bacteria, specifically K. pneumoniae, P. aeruginosa and E. cloacae. Antibiotic susceptibility tests revealed that the isolated bacteria were resistant to carbapenems and ceftazidime/avibactam but susceptible to polymyxin B in accordance with the EUCAST criteria in 2023.

Table 1.

Demographics, baseline bacteria, treatment regimens and clinical outcome of enrolled patients

Patient no.	1	2	3	4
Demographics and baseline disease
Age (y)	67	53	35	16
Gender	male	male	male	male
Body weight (kg)	60	77.5	85	51
Admission reasons	mechanical thrombectomy due to ‘acute occlusion of the right middle cerebral artery and hypertension;’ followed by a large area of right cerebral infarction and decompressive craniectomy	anterior cervical decompression, bone grafting, fusion, and internal fixation due to ‘Fracture of the 3rd, 4th and 5th cervical vertebrae and high traumatic paraplegia’	hydrocephalus, cerebral hernia, and hypoxic-ischaemic encephalopathy	hypovolemic shock and hypoxic-ischaemic encephalopathy
Baseline status	confused, bedridden for one month	quadriplegia, weak spontaneous breathing, poor spontaneous cough reflex for 3 months	in deep coma without spontaneous breathing, choking reflex, and bedridden for 5 months	in deep coma, no spontaneous breathing, no choking reflex
Infection and bacteria
Infection sites	pulmonary	pulmonary	pulmonary	pulmonary
Bacteria	CRKP	CRPA	CRKP; CRPA	CRE (E. cloacae)
PMB MIC (mg/L)	≤0.5	≤0.5	2; ≤0.5	≤0.5
PMB treatment
Inhalation dose (mg)	25	25	25 mg (Days 1–6), 50 mg (Days 7–12)	25
Intravenous dose	50 mg, q12h	100 mg, q12h	75 mg, q12h	75 mg, q12h
Treatment duration (d)	5	5	12	11
Co-medication	amikacin	meropenem (Days 1–4) piperacillin/tazobactam (Days 5–6)	tigecycline, levofloxacin	fosfomycin, SMZ
Note	none	none	CRKP and CRPA were negative after treatment for 6 days, then Serratia marcescens and Chryseobacterium meningosepticum showed positive	CRE was negative after 5 days of treatment, and Elizabethkingia meningoseptica showed positive
Clinical outcome and adverse events
Microbiology efficacy (EOT)	negative	negative	CRKP and CRPA negative (inhalation 25 mg, 6 days)	negative
Clinical outcome (EOT and TOC)	improved	check out	not improve	improved
Microbiology efficacy (TOC)	CRKP	negative	CRKP and CRPA negative	negative
28-day survival	yes	yes	no	yes
Adverse events	none	none	hyperpigmentation	hyperpigmentation

CRKP, carbapenem-resistant Klebsiella pnemoniae; CRPA, carbapenem-resistant Pseudomonas aeruginosa; CRE, carbapenem-resistant Enterobacteriaceae; EOT, end of treatment; TOC, test of cure visit (14-days post treatment).

The patients received aerosolized polymyxin B at doses of 25 or 50 mg, q12h for a duration of 14 days. The microbiological efficacy, as assessed by sputum culture, indicated the absence of baseline bacteria after aerosolized and intravenous treatment of polymyxin B for 5 to 6 days in all four patients. Three of the patients exhibited alleviation of clinical symptoms, including reduced sputum secretion and improved laboratory examination. No cases of nephrotoxicity were reported during the polymyxin B treatment. There were no reports of neurotoxicity and hyperpigmentation was observed in two patients.

Concentrations of polymyxin B in ELF and plasma samples

The concentrations of polymyxin B are shown in Figure 1. For all four patients, plasma TDM of trough and peak concentrations demonstrated favourable AUC_24h, falling within the range of 56.7−85.7 mg•h/L when administering intravenous infusion of 50 to 100 mg polymyxin B.³ The t_1/2 was in a range of 9.1 to 10.8 h, which closely aligns with the elimination half-life previously reported in patients with bloodstream infections (12.5 ± 3.11 h).¹²

Figure 1.

ELF and plasma concentrations measured in patients after aerosol combined intravenous administration of polymyxin B. Arrows indicate the simultaneous administration of intravenous (blue arrow) and aerosol delivery (red arrow). For all patients, blood or BALF trough concentrations collected just before administration. Patient 1 (25 mg inhalation and 50 mg intravenous), Patient 2 (25 mg inhalation and 100 mg intravenous), Patient 3 (50 mg inhalation and 75 mg intravenous) and Patient 4 (25 mg inhalation and 75 mg intravenous).

In total, ten measurements of polymyxin B concentrations in ELF were obtained from four patients during the steady state. Notably, all the concentrations in ELF were in a range of 20.6−97.6 mg/L, significantly higher than those in plasma following both aerosol and intravenous administration of polymyxin B. The t_1/2 varied from 3.4 to 29.0 h, and AUC_24h ranged from 283.6 to 1872.9 mg•h/L across all four patients.

Discussion

Intravenous polymyxin B sulfate can quickly reach bactericidal concentrations in the bloodstream after administration, but it also poses the risk of inducing adverse reactions such as nephrotoxicity and neurotoxicity, and has the drawback of low tissue penetration in the lungs and brain. Inhaled polymyxin B can achieve high concentrations in the lungs, which is beneficial for treating HAP/VAP infections and may reduce the risk of nephrotoxicity. However, it may cause local airway irritation and cannot have a systemic bactericidal effect. Only a limited number of studies have evaluated the optimal aerosol doses of polymyxin B. It was reported that the aerosol doses of 50 mg (twice a day) polymyxin B resulted in 95% of high cure and improved rates in the treatment of VAP patients, and significantly higher eradication rates of K. pneumoniae compared to intravenous treatment alone (92.1% versus 70.1%, P < 0.003).^5,13 Liu et al. reported that low-dose intravenous and aerosolized polymyxin B treatment (1.82 mg/kg/day) for VAP patients in ICU was associated with favourable clinical outcomes. Besides, the Chinese expert consensus on the clinical use of polymyxin B has recommended aerosol doses of 25–50 or 50 mg administered twice daily.¹⁴

Aerosolized delivery plays a pivotal role in the treatment of pulmonary infections, with the advantage of ensuring adequate drug exposure at the site of infection. However, there are a few antibiotics formulations that have received approval for this route of administration, such as inhaled tobramycin, aztreonam and colistin for cystic fibrosis patients or HAP/VAP patients.¹⁵ Compared to placebo or systematic administration, inhaled antibiotics have shown multiple benefits including reducing bacterial load in sputum or the lower respiratory tract, decreasing the frequency of lung infection exacerbations, improving lung function, potentially reducing the duration of mechanical ventilation, enhancing quality of life and improving the survival rate in some cases.¹⁵ As a result, there is a scarcity of safety, efficacy and PK/PD evidence for the aerosol delivery of antibiotics. In particular, there are limited PK data available for aerosolized polymyxins. Several studies demonstrated 100- to 1000-fold higher concentrations of colistin in ELF compared to plasma after aerosol delivery of CMS.^16,17 In our present study, we observed ELF concentrations that were 10- to 20-higher than those in plasma following both aerosol and intravenous administration. These ELF concentrations surpassed the susceptible breakpoint for polymyxin B (2 mg/L). All patients exhibited alleviated symptoms of pneumonia, indicating the favourable outcomes associated with the combined aerosol and intravenous administration of polymyxin B in clinics.

There are three commonly applied nebulizers in clinics: the jet nebulizer, ultrasonic nebulizer and vibration mesh nebulizer. The jet nebulizer has a simple structure and is durable, being widely used clinically. However, it is noisy, requires a compressed gas or power supply and has a low drug deposition rate (∼15%). The ultrasonic nebulizer is quiet, with small and uniformly sized aerosol particles, but it is prone to clogging, has la ow drug deposition rate (30%–40%) and may cause drug denaturation. The vibration mesh nebulizer is quiet, lightweight, allows for adjustable drug inhalation dose and has a high drug deposition rate (40%–60%) with low residual drug.¹⁸ In the present study, the applied vibration mesh nebulizer has a high delivery efficiency and high drug deposition rate, which leads to high concentrations in ELF.

The optimal antibiotic dosage should be optimized with the diverse organisms and their minimum inhibitory concentration (MIC) levels, according to the pharmacokinetics/pharmacodynamics (PK/PD) principles. As PK/PD at infection site as a treatment target have gained widespread acceptance, the PK/PD targets (AUC_ELF/MIC) of polymyxin B associated with bacteriostasis against P. aeruginosa strains were found to range from 1326 to 1506 in an animal model.⁶ In the present study, three patients achieved significantly higher or comparable PK/PD targets (ranging from 1451.2 to 3475.8). Nevertheless, only the PK/PD target of P. aeruginosa in animal lung infection was reported, poses a significant obstacle to fine-tune the dosage precisely according to other pathogen and their MICs. In the meantime, concerns regarding lung toxicity arising from the high concentrations of polymyxin B in ELF were raised. In the animal model, when the maximum ELF concentrations of polymyxin B achieved 184.0 ± 35.1 mg/L, histopathological results revealed reduced lung inflammation and preserved lung epithelial integrity.⁶  In vitro studies demonstrated that a much higher concentration of 1 mmol/L polymyxin B induced toxic metabolic response perturbations in epithelial cells.¹⁹ Our study further demonstrated the recovery of chest CT imaging in patients, indicating the safety of combined aerosol and intravenous polymyxin B treatment.

BAL is a widely accepted procedure for collecting respiratory secretions and serves as a diagnostic tool for inflammatory airway and lung disease. All four patients could potentially benefit from the removal of mucus during BAL, since the initial retrieved BAL instillation typically contained abundant mucus. The second and the third instillations showed similar polymyxin B concentrations (29.4 versus 24.4 mg/L) and were combined to reduce sampling error. However, one limitation of our study is the lack of the BAL cytology diagnosis.

Furthermore, polymyxin B, which shares similar structure to colistin, could also potentially bind to mucin present in sputum, thereby reducing its antimicrobial activity.²⁰ Consequently, there is an urgent need for clinical studies involving a substantial number of patients to investigate the PK of polymyxin B within the ELF, which are essential for optimizing aerosol delivery dosing regimens.

Conclusions

This is the first report of ELF concentrations following aerosolized administration of polymyxin B in critically ill patients and help leverage the gap in the PK/PD aspects of aerosolized delivery. Notably, the AUC values are significantly elevated in the ELF when compared to plasma (283.6−1872.9 mg•h/L versus 56.7−86.7 mg•h/L). For further validation, additional studies involving a larger cohort of patients and a more comprehensive assessment of the pharmacokinetics of aerosolized polymyxin B are warranted.

Funding

This research was supported by Shanghai Municipal Commission of Commerce (Shanghai Talent Awards, LJ2016-01); the Shanghai Municipal Science and Technology Commission (19411964900) and the National Natural Science Foundation of China (82173896).

Transparency declarations

The authors declare no conflicts of interest.

Author contributions

X.L. and L.Y. wrote the main manuscript text; X.L. designed the experiment and measured concentrations; L.Y. recruited patients and J.Z. and J.H. conceived the study. All authors revised and reviewed the manuscript.

Data availability

The data underlying will be available upon reasonable request.

Ethical approval

The study was approved by the institutional review board of Huashan Hospital, Shanghai, China (Review No. 2020-669). Informed consent form was obtained from each patient’s next of kin before enrolling in the study. The clinical trial was registered at: https://www.chictr.org.cn (registration number ChiCTR2200062227, registered 30 July 2022).