Objectives: To determine whether aerosolized antibiotics can be delivered efficiently to the lower respiratory tract in mechanically ventilated patients and to define possible clinical responses to these agents.
Design: Prospective serial study with cases as their own control.
Setting: A 10-bed respiratory care unit for patients with chronic respiratory failure in a tertiary university hospital.
Patients: Ventilator dependent patients who are otherwise medically stable. All subjects had a tracheostomy in place, were colonized with gram-negative organisms, and produced purulent secretions which could be sampled daily.
Interventions: Six patients received nine courses of nebulized therapy, which consisted of treatments every 8 hrs of gentamicin (80 mg) or amikacin (400 mg) for 14 to 21 days.
Measurements and main results: Doses to the lung were measured using radiolabeled aerosols and antibiotic concentrations in sputum. The response was assessed by a) changes in the volume of respiratory secretions; b) effect on bacterial cultures; and c) changes in the inflammatory cells and mediators of inflammation of the respiratory secretions (interleukin-1beta [IL-1beta], tumor necrosis factor-alpha [TNF-alpha], soluble intercellular adhesion molecule-1 [sICAM-1], and human leukocyte elastase). On average, patients inhaled 35.4 +/- 5.08% (SD) of the initial drug placed in the nebulizer (neb-charge). Of this neb-charge, 9.50 +/- 2.78% was found on the respirator tubing and tracheostomy tube and 21.9 +/- 7.15% was actually deposited in the lungs. The remainder of the neb-charge was sequestered in the nebulizer or exhaled. Trough sputum concentrations averaged 4.3 +/- 3.2 microg/mL/mg neb-charge (range 234 to 520 microg/mL) and increased to 16.6 +/- 8.1 microg/mL/mg neb-charge (range 1005 to 5839 microg/mL) immediately after therapy (p = .011). Serum concentrations were undetectable in most determinations except for a single patient who was in renal failure (8.7 microg/mL amikacin). Treatment caused a significant reduction in the volume of secretions (p = .002). Weekly cultures revealed eradication of Pseudomonas species, Serratia marcescens, and Enterobacter aerogenes in most of the trials. Before antibiotic treatment, concentrations of IL-1beta were higher than those reported in acute respiratory distress syndrome. Throughout the duration of the study, IL-1beta correlated significantly with the absolute number of macrophages, neutrophils, and lymphocytes, respectively (r2 = .55, p = .002; r2 = .50, p < .0004, r2 = .36, p = .005). TNF-alpha concentrations correlated with lymphocytes and neutrophils, respectively (r2 =.27, p = .013, r2 = .21, p = .033). sICAM-1 concentrations increased two-fold (p < .001) during treatment and then returned to baseline. The volume of secretions was related to neutrophil and IL-1beta concentrations, respectively (r2 = .25, p = .008, r2= .35, p = .006).
Conclusions: Nebulizer delivery of aerosolized aminoglycosides is efficient and predictable. In our clinical model, aerosolized antibiotics can make a significant impact on respiratory secretions. Their efficacy in treatment of critically ill patients remains to be determined.