Antimicrobial Therapy of Ventilator-Associated Pneumonia

How To Select an Appropriate Drug Regimen

Ventilator-associated pneumonia (VAP) refers to nosocomial pneumonia developing in a patient receiving mechanical ventilation. VAP is recognized as being one of the leading causes of death from hospital-acquired infections in the ICU setting.^1,2 The estimated prevalence of VAP ranges from 10 to 65% with case fatality rates of 13 to 55%.^3,4 Additionally, several clinical studies have demonstrated an attributable mortality associated with VAP that is independent of patients’ underlying diagnoses and severity of illness at the time of ICU admission.^5,6 The occurrence of VAP also increases the costs associated with hospitalization. Several economic analyses estimate the excess medical costs attributed to an episode of VAP to be > $5,000 with prolongation of the hospital stay from 6 to > 30 days.^7,8,9 Recently, the importance of providing early effective antimicrobial therapy for patients with VAP has been highlighted by several clinical investigations.^10,11,12,13

Once VAP develops, treatment is usually supportive along with the administration of antibiotics. The selection of antimicrobial agents for the initial empiric treatment of VAP appears to be an important determinant of clinical outcomes, especially hospital mortality. Luna and colleagues¹⁰ examined 132 patients requiring mechanical ventilation with clinically suspected VAP. A total of 50 patients with positive BAL cultures received empiric antibiotic therapy prior to obtaining the BAL culture results. Patients who received adequate antibiotic therapy (n = 16), as defined by the BAL culture results, had a significantly lower mortality rate compared with patients receiving inadequate antibiotic therapy (n = 34) (37.5% vs 91.2%; p < 0.001). Alvarez-Lerma¹¹ evaluated the appropriateness of antimicrobial therapy in 430 patients with VAP receiving antibiotic treatment. The attributable mortality from VAP was significantly greater among patients receiving inadequate initial antimicrobial therapy compared with patients receiving adequate initial therapy (24.7% vs 16.2%; p < 0.039). Similarly, Rello et al¹² found that patients with VAP who received inadequate initial antibiotic therapy had significantly greater crude mortality rates (63.0% vs 41.5%; p = 0.06) and VAP attributable mortality rates (37.0% vs 15.6%; p < 0.05) compared with patients receiving adequate initial antibiotic therapy. Lastly, our own group confirmed these findings in a study employing mini-BAL to obtain lower respiratory tract cultures in patients with suspected VAP.¹³

Table 1 summarizes the pathogens associated with inadequate initial empiric antimicrobial treatment of culture-positive VAP in the four clinical studies noted above.^10,11,12,13 Most episodes of inadequate antimicrobial treatment were attributed to Gram-negative bacteria, with Pseudomonas aeruginosa, Acinetobacter species, Klebsiella pneumoniae, and Enterobacter species accounting for most cases. These same species of Gram-negative bacteria are often associated with antibiotic resistance and worse patient outcomes compared with more antibiotic susceptible strains of Gram-negative bacteria (eg, Haemophilus influenzae, Escherichia coli).^14,15 Methicillin-resistant Staphylococcus aureus (MRSA) was the next most common pathogen associated with the administration of inadequate antimicrobial treatment. Interestingly, only one of these four studies reported using specific methods for the isolation of anaerobic bacteria.¹² This probably accounts for the lack of identified anaerobic bacteria in the lower airway cultures from these studies.

The clinical importance of not specifically treating anaerobic bacteria in patients with suspected VAP is unknown. In this issue of CHEST Marik and Careau (see page 178) describe their experience with 185 episodes of suspected VAP or aspiration pneumonia (AP) in 143 patients. Despite using specific methods to isolate anaerobic bacteria, only one anaerobic microorganism was isolated from 75 episodes classified as either VAP or AP. These results differ somewhat from those of Dore’ and coworkers¹⁶ who examined 130 patients with microbiologically documented VAP using protected specimen brush (PSB) cultures and rigorous anaerobic culturing techniques. Among these patients, 100 (77%) had only aerobic bacterial strains isolated from PSB cultures. In 30 (23%) patients, PSB cultures resulted in anaerobic strains. Aerobic strains were associated with anaerobic strains in 26 patients, whereas 4 patients had only anaerobic strains isolated from PSB cultures. The 3-month mortality rates were reported to be similar for patients with and without anaerobic strains isolated (31% vs 36%). However, the influence of the adequacy of the initially prescribed antimicrobial treatment, for both aerobic and anaerobic bacterial strains isolated from the PSB cultures, on patient outcomes was not described in this investigation.

To date, and to our knowledge, no convincing clinical data are available supporting the hypothesis that routine treatment for anaerobic bacteria will improve the outcomes of patients with suspected VAP. Alternatively, several investigations have highlighted the problems associated with the overuse of anaerobic antibiotics, particularly clindamycin, in the hospital setting. These problems include antibiotic-associated diarrhea or colitis due to Clostridium difficile infection, direct end-organ drug toxicity, and unnecessary increases in medical care costs.^17,18 What appears to be currently needed are well-performed outcome studies aimed at determining the most effective, least toxic, and most cost-efficient approaches for the initial empiric treatment of suspected VAP. Included within such studies could be an evaluation of the routine administration of specific antimicrobial agents directed against anaerobic bacterial strains. However, as noted by the experience in patients with VAP described above,^10,11,12,13 developing new strategies aimed at providing improved initial antimicrobial coverage for potentially antibiotic-resistant Gram-negative bacteria and MRSA may yield greater clinical benefits.

At the present time, clinicians should be aware of the most common bacterial pathogens, and their antimicrobial resistance patterns, accounting for VAP at the hospitals where they practice. Prescribing an initial broad-spectrum antibiotic regimen to cover all likely pathogens will help to reduce the occurrence of inadequate treatment and may result in improved clinical outcomes. Initial combination antimicrobial therapy, particularly aimed against antibiotic-resistant Gram-negative bacteria (eg, P aeruginosa, Acinetobacter species) and MRSA, offers the greatest likelihood of providing adequate initial treatment. However, such broad-spectrum treatment should not be unnecessarily prolonged unless supported by appropriate culture data in order to avoid the emergence of antibiotic-resistant infections. The selection of the most effective combination of initial antibiotics for the treatment of suspected VAP is unknown at present. Nevertheless, emerging clinical data are becoming available to offer practical suggestions for consideration of what constitutes optimal initial antimicrobial therapy for VAP. One study found that imipenem plus amikacin plus vancomycin was the most effective drug regimen for patients at high risk of developing VAP due to antimicrobial-resistant bacteria.¹⁹ Other combinations of initial empiric antibiotics may also be effective and await identification in future clinical trials.