Pharmacokinetics of vancomycin in sputum of intubated patients: Optimized intravenous delivery vs. inhaled therapy

Abstract

Aims

Intubated patients with methicillin‐resistant Staphylococcus aureus pneumonia, fail optimized treatment with intravenous (IV) vancomycin (serum trough 15–20 μg/mL) in 38–79% of cases. Airway blood flow is diminished compared to alveoli and we hypothesized that vancomycin concentrations achieved in airway secretions are suboptimal and nonbactericidal. Targeted therapy by inhalation may overcome this deficit.

Methods

Airway pharmacokinetics of optimized IV and inhaled vancomycin in infected clinically stable prolonged mechanically ventilated patients were measured. First, IV vancomycin was given until optimized concentrations were achieved (15–20 μg/mL), and, at the same time point, sputum vancomycin concentrations were measured. Then, sputum concentrations were re‐assessed after 4 treatments of inhaled vancomycin (120 mg/2 mL) via a previously characterized nebulizing system that deposited 18 ± 2 mg in the lungs. Vancomycin post‐distribution phase serum peak and trough concentrations were also obtained. Serum albumin was measured to assess binding to vancomycin.

Results

Mean serum trough concentration was 18.4 ± 6.5 μg/mL. Sputum concentrations were affected by serum albumin. Only patients with severe hypoalbuminaemia had penetration of drug leading to therapeutic (15.7–17 μg/mL) sputum concentrations. Following inhaled vancomycin, sputum concentrations increased significantly to 199 ± 37.0 μg/mL (P = .002) exceeding minimum inhibitory concentration by 2 orders of magnitude.

Conclusion

Despite optimized serum concentrations, patients with albumin near normal had suboptimal concentrations of vancomycin in their sputum. Inhaled therapy may be clinically important for successful treatment of ventilator‐associated methicillin‐resistant Staphylococcus aureus infection. Further studies of inhaled therapy are needed to define their role as adjunctive therapy in ventilator‐associated pneumonia and as single therapy in tracheobronchitis.

What is already known about this subject

• Intubated patients with methicillin resistant Staphylococcus aureus pneumonia, fail optimized treatment with intravenous vancomycin (serum trough 15–20 μg/mL) in 38–79% of cases.

• Sputum concentrations which are important for airway infection are not well described and may be subtherapeutic with intravenous therapy.

What this study adds

• This investigation demonstrated that inhaled therapy raises concentrations to bactericidal levels and ensures effective therapy.

• Despite optimized serum concentrations, patients on intravenous therapy with albumin near normal had suboptimal concentrations of vancomycin in their sputum .

• Inhaled therapy may be clinically important for successful treatment of ventilator‐associated methicillin‐resistant Staphylococcus aureus infection.

1. INTRODUCTION

The current standard of care for treatment of methicillin resistant Staphylococcus aureus (MRSA) respiratory infection is intravenous (IV) vancomycin, or linezolid. ¹ However, the clinical cure rates are only 21.2–62% and mortality 16–38%. ² In theory, the in vivo efficacy of any systemic antibiotic depends on its in vitro activity, its local distribution to the site of infection, the amount bound to protein, the bacterial inoculum and the microenvironment of the infection. Protein‐bound vancomycin is inactive so protein concentrations in serum, the alveolar space and in sputum may have important effects on efficacy.

Pharmacokinetic studies of vancomycin for treating MRSA respiratory infection have focused on the alveolar compartment of the lung and measured drug concentrations in epithelial lining fluid (ELF) after IV treatment. ³ , ⁴ , ⁵ , ⁶ , ⁷ This compartment has very low albumin even in a setting of inflammation with antibiotics predominantly free and bioactive. However, in critically ill patients, ELF penetration is highly variable with as much as a 6‐fold drop of concentration from serum to ELF. ⁴ , ⁵ In infected, intubated patients Lamer and colleagues used dose‐adjusted IV vancomycin to raise serum concentrations to 24 μg/mL. ⁵ ELF concentration was measured at 4.5 μg/mL, presumably sufficient to treat a minimum inhibitory concentration (MIC) of 1 μg/mL. With that background, higher trough concentrations of vancomycin (concentrations of 15–20 μg/mL) have been recommended. ⁸ However, the clinical failure rate of IV vancomycin is unaffected by this latest recommendation. This observation suggests that reported ELF concentrations alone are not sufficient for clinical cure. ⁹ While the ELF compartment has been thought by many to represent the best drug exposure profile for the treatment of ventilator‐associated pneumonia (VAP), ³ , ⁵ , ⁶ others suggest that ELF is fraught with large variability and technical problems and should not be used at all. ¹⁰ , ¹¹ Even if correct for alveolar drug exposure, ELF drug concentrations will not reflect either the concentrations or bioactivity in the airway compartment of infection. Furthermore, to date, there are no data showing that successful clinical outcomes are directly related to ELF concentrations of vancomycin. ⁷ , ¹⁰

Intubated patients colonized with MRSA can develop ventilator‐associated tracheobronchitis (VAT) followed by VAP. Because of high failure rates in treating VAP, Stulik et al. examined the potential efficacy of early IV vancomycin in decreasing the progression of colonization to VAT or VAP. ¹² Of 39 patients with S. aureus colonization, 13 patients progressed to VAT and 15 to VAP demonstrating minimal effectiveness of IV vancomycin. The authors hypothesized that IV therapy failed because of insufficient ELF concentrations, and/or the presence of biofilm reducing antibiotic efficacy. We agree with these concerns. We believe that VAT and progression to VAP occur because IV antibiotics do not result in bactericidal levels in the inflammatory milieu in the airway provoked by the endotracheal tube. Consequently, clinical failure of IV therapy in treating VAT or VAP may be due to inadequate concentrations of active drug in airway secretions: a combination of poor penetration and/or low bioactivity due to protein binding or other molecular interference.

As a model for treatment, in this special situation, our group has favoured targeted therapy to the lungs and airways via aerosol in addition to conventional IV therapy. Delivering inhaled antibiotics directly to the lungs and airways by aerosol may safely raise local antibiotic concentrations beyond those maximally tolerated by IV therapy alone and thereby significantly increase the bioactive concentration. ¹³

The present study was designed to describe the total antibiotic concentrations achieved in the milieu of the intubated central airways first following optimized IV therapy and then after inhaled adjunctive therapy with vancomycin. ¹⁴ , ¹⁵ Defining airway pharmacokinetics of IV and inhaled vancomycin may provide new insights useful in the design of future clinical therapeutic trials.

2. MATERIALS AND METHODS

This Institutional Review Board‐approved study was carried out in long‐term hospitalized patients with respiratory failure maintained on mechanical ventilation via a tracheostomy. They were supported in a chronic ventilator unit for weaning at University Hospital, Stony Brook, NY, USA. They were chosen specifically because of their haemodynamic and renal function stability and because they did not have large fluid shifts, which would add variability to drug concentrations. Inclusion criteria included a clinical suspicion of MRSA infection, including purulent secretions containing Gram‐positive cocci on Gram stain. Patients were excluded if they produced no sputum. Additional exclusion criteria included allergy to vancomycin, serum creatinine ≥2 mg/dL or creatinine clearance ≤30 mL/min at the time of enrolment.

2.1. Experimental population

The experimental population is shown in Table 1 and enrolment flow chart is shown in Figure 1. A 10‐patient sample was felt to be sufficient for pharmacokinetic analysis. ¹⁶ All patients were treated with IV vancomycin. First, IV dosing was optimized by measuring serial total (free plus protein bound) vancomycin serum concentrations. ¹⁴ With optimized IV therapy established, vancomycin concentration in sputum was measured. Serum albumin, known to bind vancomycin, was also measured at the same time. Then, while IV vancomycin was continued, inhaled vancomycin aerosols were targeted to the lungs with a well‐characterized nebulizing system. It was administered every 8 h. After the fourth dose of inhaled vancomycin, concentration in the sputum and serum concentration of vancomycin were measured, (Figure 2)).

TABLE 1.

Demographics of ventilated patients.

Patient	Age (years)	Sex	Diagnosis	Airway type/duration	MRSA infection
1	53	F	Downs syndrome C3‐C4 injury	Tracheostomy/ >5 years	Pneumonia and bacteraemia
2	91	M	Myasthenia gravis	Endotracheal Tube/1 day	Pneumonia
3	55	M	COPD	Endotracheal Tube/2 weeks	Bilateral pneumonia
4	78	F	Acute CVA	Endotracheal Tube/1 day	Bilateral pneumonia
5	60	M	CVA	Tracheostomy/1 month	Pneumonia and bacteraemia
6	86	M	Myelodysplastic syndrome	Endotracheal Tube/1 day	Pneumonia
7	86	M	Bilateral basal ganglia infarctions	Endotracheal Tube/1 month	Pneumonia
8	63	M	Nasopharyngeal carcinoma	Tracheostomy/ ~1 month	Bilateral pneumonia
9	91	F	COPD	Endotracheal Tube/1 day	Bilateral pneumonia
10	60	M	COPD	Tracheostomy/1 day	Bilateral pneumonia

Abbreviations: COPD, chronic obstructive pulmonary disorder; CVA, cerebrovascular accident; F, female; M, male; MRSA, methicillin‐resistant Staphylococcus aureus.

FIGURE 1.

Flow diagram demonstrating the number of patients enrolled, the number disqualified, and the number that completed the study.

FIGURE 2.

Concentrations of vancomycin (μg/mL) in the serum and sputum before (intravenous vancomycin [IV Vanco] only) and after (post‐fourth aerosol) 4 inhaled doses of vancomycin (mean ± standard error of the mean). Left‐to‐right: vancomycin concentration in serum after the goal of 15–20 μg/mL was achieved (open squares), sputum concentration at the same time (open circles), serum vancomycin after the fourth inhaled dose (closed squares), and sputum concentration 30 min after the fourth inhaled dose (closed circles). Error bars in each group are partially obscured by data points.

2.2. Systemic vancomycin dosing

The dose of IV vancomycin was determined using a dosing protocol designed to achieve a steady‐state vancomycin trough concentration of 15–20 μg/mL based on estimated creatinine clearance (CL_Cr) and total body weight. To perform this calculation, renal function was estimated using the Cockcroft–Gault equation, which provides an estimated creatinine clearance for males:

$${\mathrm{CL}}_{\mathrm{Cr}}=\left[{\left(140–\mathrm{age}\right)}^{*}\text{weight}\right]/\left({72}^{*}\mathrm{Cr}\right)]$$ (1)

where, CL = creatine clearance, Cr = creatinine (mg/dL), age is in years, weight is in kg. Multiply male result by 0.85 for females.

Serum vancomycin peak and trough concentration were measured both prior to and during the period when subjects were receiving inhaled vancomycin. Vancomycin half‐life and maximum and minimum plasma concentrations were estimated using 1‐compartment linear pharmacokinetic equations as described by Rybak et al. ⁸ In this stable group of patients this methodology seemed appropriate and in concert with these guidelines. Measured vancomycin peak concentrations were not obtained in 3 subjects due to unavailability at designated sampling time.

2.3. Inhaled vancomycin protocol

Inhaled vancomycin was given via jet nebulizer, (AeroTech II, Biodex Medical Systems, Inc., Shirley, NY, USA). The device has been well characterized in our laboratory. In ventilated circuits, O'Riordan et al. reported a mass median aerodynamic diameter of 1.1 ± 1.8 μm (mean ± standard deviation), and, in intubated patients, an inhaled mass of 30.6 ± 6.3% and lung deposition of 15.3 ± 9.5%. ¹⁵ The nebulizer was breath‐actuated during the inspiratory phase by the ventilator (Evita 4, Drager Inc., Telford, PA, USA), placed in the inspiratory line 30 cm proximal to the patient Y‐piece with the humidifier bypassed. ¹³ Vancomycin was prepared by making a stock solution of 100 mg vancomycin (Baxter, Deerfield, IL, USA) per mL of H₂O. The nebulizer fill volume was made by using 1.2 mL of the stock solution bringing the volume to 2 mL with 0.8 mL of normal saline (120 mg/2 mL). This dose (120 mg) was nebulized every 8 h for 4 doses. The duration of nebulization was ~40 min. From previous reports we calculate that 18 ± 2 mg was deposited in the lungs and airways. ¹⁵

2.4. Sputum sampling protocol

Sputum specimens were obtained via suctioning through the tracheostomy tube just prior to initiation of aerosol therapy and 30 min after the fourth aerosol inhalation. Saline installation prior to suctioning was avoided. Palmer and colleagues, who reported the relationship between sputum sampled in this manner and nebulizer delivery, have described this approach in detail. ¹⁷ The sample was ultracentrifuged at 1 006 175 g for 1 hr at 40 ° C. The pellet was discarded, and the supernatant phase used for analysis. It was stored at −80°C until the time of analysis. Total vancomycin concentrations in the supernatant of the sputum and in the serum were determined by fluorescence polarization immunoassay (Emit 2000 Vancomycin Assay on a Beckman Coulter chemistry analyser, Beckman Coulter, Brea, CA, USA).

2.5. Serum albumin measurements

Serum albumin was measured in our clinical laboratory at the time of enrolment in the study.

2.6. Vancomycin measurement

The vancomycin assay was colorimetric and performed on a COBAS 600 Analyser (Roche Diagnostics North America, Indianapolis, IN, USA). To determine the range of reliability of total vancomycin concentration in the supernatant phase of sputum, a dilutional curve was performed. Samples of 0, 5, 20, 40 and 60 μg/mL were made by adding the appropriate amount of vancomycin to known amount of supernatant. The assay was reliable between the concentrations of 8 and 100 μg/mL. Patient samples were diluted if they were >100 μg/mL.

2.7. Vancomycin penetration

The penetration ratio of IV vancomycin was calculated by dividing the sputum concentration of total vancomycin by the serum concentration of total vancomycin prior to addition of inhaled therapy.

2.8. Statistics

Data are expressed as mean ± standard error of the mean. Wilcoxon matched pairs test was used for the analysis of the pharmacokinetic part of this investigation for concentrations in different distributive compartments of the same patient. Significance was defined as a P‐value ≤.05.

Linear correlation was used to determine the relationship between concentrations in the serum and the sputum of vancomycin and the relationship of penetration to serum albumin.

3. RESULTS

3.1. Pharmacokinetic data

Individual data points for vancomycin concentrations in serum and sputum are shown in Figure 2. Optimized IV therapy resulted in a serum concentration of 18.4 ± 2.0 μg/mL. Sputum levels were lower with mean sputum vancomycin of 10.8 ± 1.3 μg/mL (P = .016). Following 4 treatments of inhaled vancomycin, sputum vancomycin concentration increased significantly to 199.7 ± 37 μg/mL (P = .002, n = 8 instead of 10 because 2 patients had peak sputum that was too viscous to analyse). Vancomycin concentration in the serum was not significantly changed after 4 doses of inhaled vancomycin: (18.4 ± 2.0 μg/mL vs. 17.5 ± 1.1 μg/mL, P = .75.).

All indices of pharmacokinetics of vancomycin for individual patients are listed in Table 2.

TABLE 2.

Pharmacokinetics of individual patients (mean ± confidence interval [CI]).

ID	CLCr (mL/min)	Serum Vanco pre‐AA (μg/mL)	Trough sputum Vanco (μg/mL)	Sputum Vanco after inhaled vanco (μg/mL)	Albumin	Penetration of vancomycin
1	148	11.4	9.7	NA	3.2	0.85
2	171	20.1	11.2	119	3.8	0.56
3	21	19.1	17	267	1.8	0.89
4	93	15.6	15.7	263	2.2	1.01
5	100	34.9	7.6	NA	2.8	0.22
6	43	16.7	16	188	1.6	0.96
7	31	15.8	7.1	78	2.5	0.45
8	193	18.5	9.2	49	2.4	0.5
9	81	12.5	7.1	298	3.6	0.57
10	94	19.2	7.5	336	2.8	0.38
Mean	98	18.4	10.8	200	2.7	0.55
±CI	113	12	7.5	208	0.44	0.16

Abbreviations: AA, aerosolized antibiotics; CL_Cr, creatinine clearance; vanco, vancomycin.

3.2. Sputum vs. serum concentrations

The concentration of vancomycin in the sputum had no correlation to serum vancomycin concentration (R = .15, P = .892). The range of serum albumin was large (1.6–3.7 g/dL). As shown in Figure 3 only the 3 patients with the lowest serum albumin had therapeutic concentrations in the sputum and they had the highest penetration ratio (patients 1, 4 and 6; Table 2). Overall, there was a significant correlation between serum albumin and sputum concentration (R = .64, P = .04; Figure 4).

FIGURE 3.

Serum albumin vs. sputum concentration of vancomycin (R = .64 and P = .04).

4. DISCUSSION

Treating physicians adjust drug dosing attempting to reach an optimum serum level of vancomycin. Our data suggest that, for airways, this approach will often result in suboptimal levels in the sputum. The relationship between serum and sputum concentration of vancomycin depends on serum albumin, suggesting that binding of drug in the serum reduces penetration into the airway. The range of serum albumin measured in our patients reflected the diverse nutritional status of these chronically ill patients, and our results indicate that optimal serum antibiotic levels alone do not predict sputum levels.

Our data demonstrate that sputum levels can be increased approximately 20‐fold following inhaled vancomycin, an increase in level clearly unattainable by IV therapy alone.

Current recommended dosing regimens are based on ELF data but there are no data for optimization of vancomycin dosing based on airway concentration and MICs. If the MIC of the organism is 1 μg/mL in serum, why does IV therapy fail? And what concentrations might be bactericidal in the sputum? Our data provide a possible mechanism for treatment failure, for VAP in spite of the best efforts of treating physicians, e.g. lower antibiotic levels in the milieu of the tracheostomy tube, similar to Gram‐negative infections with reduced penetration and increased inactivation of vancomycin.

The effects of serum albumin on penetration of vancomycin into sputum were significant. Protein binding effects on antibiotic distribution are not often considered by intensivists but are clinically relevant. Thirty to 70% of vancomycin in serum is bound to albumin and only unbound vancomycin is active and penetrates tissues. ⁸ , ¹⁸ Our results indicate that serum albumin significantly affects penetration into the airway compartment. The relationship of serum protein to tissue penetration of vancomycin is well known; however, this relationship has not been explored in sputum. ¹⁹ , ²⁰ As a determinant of sputum concentrations of vancomycin, serum albumin was more important than serum vancomycin. This observation suggests that patients with normal serum albumin treated only with IV therapy may be at even greater risk than hypoalbuminaemic patients for nonbactericidal vancomycin concentrations because they will have the lowest antibiotic concentrations in airway sputum. This finding may have important clinical significance and requires further clinical study.

Is our focus on the central airway compartment clinically relevant? These pharmacokinetic data correlate with our clinical observations with adjunctive inhaled therapy. In 2 placebo‐controlled studies of adjunctive inhaled antibiotics in intubated patients with VAT and VAT/VAP, our group has reported that weaning from the ventilator is facilitated, bacteria are eradicated, and signs of infection reversed more effectively following inhaled adjunctive therapy. ¹⁷ , ²¹ Those observations indicate more effective reduction in airway and systemic inflammation and infection following inhaled antibiotics targeted to the central airways.

4.1. Limitations

Bioactivity assays of vancomycin in the sputum were not done. Future studies including bioactivity assays could further address the importance of vancomycin binding in the sputum. and bioactivity.

5. CONCLUSIONS

Following optimized IV vancomycin therapy, sputum levels of vancomycin were significantly reduced when compared to serum levels. Vancomycin concentration in the airway compartment was directly impacted by the serum albumin concentration.

The pharmacokinetic and drug exposure profile in the airway may be an important component defining clinical efficacy. Delivery via inhalation overcomes reduced penetration from serum to the airway. This could have important ramifications for treatment of tracheobronchitis with inhaled therapy alone and as adjunctive therapy in VAP. The inhaled route raises concentrations in sputum orders of magnitude beyond that of IV therapy reducing effects of serum albumin and other factors affecting airway drug levels.

AUTHOR CONTRIBUTIONS

L.B.P.: study design, data collection, data analysis, literature search, manuscript preparation; G.C.S.: study design, data collection, data analysis, manuscript preparation; M.M.: study design, data collection, data analysis, manuscript preparation.

The authors confirm that the PI for this paper is Lucy Palmer, and that she had direct clinical responsibility for patients. The State University of New York at Stony Brook owns patents on the treatment of respiratory infection in the intubated patient licensed to InspiRx, Inc., Somerset, NJ. Dr Smaldone consults to InspiRx and is a member of the Advisory Board. Dr Palmer declares no conflicts of interest.

CONFICT OF INTEREST STATEMENT

The State University of New York at Stony Brook owns patents on the treatment of respiratory infection in the intubated patient licensed to InspiRx, Inc., Somerset, NJ. Dr Smaldone consults to InspiRx and is a member of the Advisory Board. Dr Palmer declares no conflicts of interest.

INSTITUTIONAL REVIEW BOARD STATEMENT

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of Stony Brook University (80 274–4, February 25, 2009).

INFORMED CONSENT STATEMENT

Informed consent was obtained from all subjects (or their Health Care Proxy) involved in the study.

DISCLOSURE

Stony Brook University holds patents on the treatment of respiratory infection in the intubated patient.

Palmer LB, Monteforte M, Smaldone GC. Pharmacokinetics of vancomycin in sputum of intubated patients: Optimized intravenous delivery vs. inhaled therapy. Br J Clin Pharmacol. 2025;91(1):127‐133. doi: 10.1111/bcp.16225

DATA AVAILABILITY STATEMENT

Data can be made available on request.