Editorial

MRSA ventilator-associated pneumonia: Myth or reality?

In intubated adult ICU patients, the diagnosis of MRSA ventilator-associated pneumonia, or VAP, is based on epidemiologic/microbiologic criteria. The characteristic clinical features of MRSA pneumonia are well-known from decades of clinical experience with MRSA community-acquired pneumonia in patients with influenza. This is the gold standard of diagnosing MRSA pneumonia.

Burke A. Cunha

Colonization does not mean infection

Patients with influenza who have bona fide MRSA pneumonia have several characteristic clinical features, such as otherwise unexplained high spiking fevers, hypotension, cyanosis, leukocytosis and rapidly cavitating infiltrates (< 72 hours) with elastin fibers in airway secretions (diagnostic of necrotizing pneumonia), hemorrhagic tracheobronchitis (due to influenza), and positive sputum culture for MRSA. In patients with influenza, if the methicillin-susceptible Staphylococcus aureus (MSSA) or MRSA strain is positive for Panton-Valentine leukocidin cytotoxin, outcomes are frequently fatal. This clinical presentation differs markedly from the typical intubated ICU patient with low-grade fever, leukocytosis and new infiltrates on chest X-ray with MRSA in respiratory secretion cultures.

Until proven otherwise, if the characteristic findings of MRSA pneumonia with influenza are not present, MRSA cultured from respiratory secretions represents colonization.

In the microbial milieu of the ICU, colonization of body fluids — urine, wounds and respiratory secretions, for example — is the rule rather than the exception. Although colonization may precede infection, clinical experience with nosocomial pneumonias suggests quite the opposite; that is, VAP is uncommon following even prolonged airway colonization with gram-negative bacilli (GNB), S. aureus or Candida spp. In intubated patients without influenza in the ICU, MRSA airway colonization rarely results in subsequent MRSA VAP. The clinical confusion results from assuming airway culture isolates are reflective of the lower lung pathogens causing VAP. However, the problem clinically is to differentiate the related respiratory secretion isolate from those that are unrelated. For this reason, clinical correlation of culture results with characteristic clinical findings is always the final determinant of the clinical significance of organisms cultured from any body site.

Airway colonization with gram-negative bacilli, such as Enterobacter spp., Stenotrophomonas maltophilia and Burkholderia cepacia, in intubated ICU adults is common, but these organisms rarely cause VAP in normal hosts.

Clinicians who do not specialize in infectious disease often treat patients based solely on culture results rather than correlating culture results with characteristic lung findings to determine if the organism cultured from airway secretions is the pathogen causing the lung infection. Often, empiric treatment for potential VAP (pVAP) is based on “guilt by association” in intubated ICU patients. It is a common daily occurrence to observe one or more potential pathogens (MSSA, MRSA, Pseudomonas aeruginosa, Enterobacter spp., Acinetobacter spp., Serratia spp.) persistently cultured from respiratory secretions from patients with fever, leukocytosis, and infiltrates due to other causes.

There are many mimics of VAP. Infiltrates may be due to noninfectious causes like congestive heart failure, chronic renal failure, systemic lupus erythematosus pneumonitis, pulmonary sarcoidosis, Wegener’s granulomatosis, bronchiolitis obliterans organizing pneumonia, and drug-induced pulmonary toxicity. If the underlying disorder causes fever, leukocytosis, etc., then the unwary clinician may relate it to the cause of pulmonary infiltrates. Without clinical correlation, the dilemma is further complicated when multiple potential organisms are cultured from respiratory secretions. Which, if any, are the cause of the lung infiltrates? P. aeruginosa VAP closely resembles that of MRSA VAP, with the same finding in a necrotic pneumonia with rapidly cavitating infiltrates. Fortunately, a negative MRSA respiratory secretion culture effectively rules out the diagnosis of MRSA VAP.

If MRSA VAP was as common as the literature suggests, it would have been frequently reported in autopsies. Clinically, from our influenza MRSA pneumonia experience, a fatal outcome is common. Most of the published studies on MRSA VAP are based on microbiologic/epidemiologic data, not clinical criteria. Many ICU intubated patients with fever, leukocytes and pulmonary infiltrates have MRSA airway colonization (30% to 40% reported as MRSA VAP). Adding MRSA-positive blood cultures does not affirm the diagnosis. The problem remains that positive blood cultures may be due to skin colonizers (gaining access to blood cultures) or reflect central venous catheter-associated bacteremia. In these same patients, a respiratory secretion culture that is positive for MRSA is common, but the bacteria may not be the cause of the infection.

Unnecessary use of vancomycin for empiric therapy

It makes little sense to empirically treat all patients with pVAP who have MRSA-positive respiratory secretion cultures with vancomycin (in addition to an anti-GNB antibiotic).

There are numerous implications for antimicrobial stewardship programs (ASPs) of unnecessary empiric vancomycin treatment of all patients with pVAP. First, in an economy with limited financial resources for health care, the needless cost of vancomycin is unjustifiable. Direct drug costs of vancomycin are only a portion of total costs. Indirect costs include IV administration ($10/IV dose), plus the cost of vancomycin levels.

The unnecessary use of vancomycin for empiric therapy of MRSA pVAP/VAP may have untoward microbiological consequences, as well. Because vancomycin does not penetrate well into respiratory secretions, it does not eliminate MRSA colonization. Low levels of vancomycin may select for multidrug-resistant GNB pulmonary pathogens. Second, vancomycin increases S. aureus cell wall thickness, which may result in permeability-mediated resistance to vancomycin as well as other antibiotics used to treat MRSA. Vancomycin permeability-mediated resistance in S. aureus manifests as “MIC creep,” increasing the minimum inhibitory concentration during or following therapy. Therapy with other MRSA antibiotics does not have these effects. Other potential collateral microbiologic effects include an increase in the prevalence of vancomycin-resistant enterococci in ICUs and hospitals.

Clinically, there are no differences in outcomes with or without empiric vancomycin added to an antibiotic with GNB activity, further suggesting that MRSA VAP is rare.

I have been long interested in pneumonias, and for decades I have diligently searched for cases of MRSA pneumonia. However, over the years, I have observed several cases of MRSA community-acquired pneumonia with influenza but only very rarely identified bona fide MRSA VAP. Using nonspecific diagnostic criteria, VAP has taken on mythic proportions. Although isolated cases certainly exist, MRSA VAP remains a rare event.

The reality is, if empiric monotherapy for GNB pVAP is selected wisely, there is no need for empiric MRSA coverage for pVAP/VAP. From a clinical perspective, carefully selected initial empiric monotherapy with a high degree of activity against GNB pulmonary pathogens, like P. aeruginosa, and low resistance potential remains the optimal clinical approach.

Disclosure: Cunha reports no relevant financial disclosures.

In intubated adult ICU patients, the diagnosis of MRSA ventilator-associated pneumonia, or VAP, is based on epidemiologic/microbiologic criteria. The characteristic clinical features of MRSA pneumonia are well-known from decades of clinical experience with MRSA community-acquired pneumonia in patients with influenza. This is the gold standard of diagnosing MRSA pneumonia.

Burke A. Cunha

Colonization does not mean infection

Patients with influenza who have bona fide MRSA pneumonia have several characteristic clinical features, such as otherwise unexplained high spiking fevers, hypotension, cyanosis, leukocytosis and rapidly cavitating infiltrates (< 72 hours) with elastin fibers in airway secretions (diagnostic of necrotizing pneumonia), hemorrhagic tracheobronchitis (due to influenza), and positive sputum culture for MRSA. In patients with influenza, if the methicillin-susceptible Staphylococcus aureus (MSSA) or MRSA strain is positive for Panton-Valentine leukocidin cytotoxin, outcomes are frequently fatal. This clinical presentation differs markedly from the typical intubated ICU patient with low-grade fever, leukocytosis and new infiltrates on chest X-ray with MRSA in respiratory secretion cultures.

Until proven otherwise, if the characteristic findings of MRSA pneumonia with influenza are not present, MRSA cultured from respiratory secretions represents colonization.

In the microbial milieu of the ICU, colonization of body fluids — urine, wounds and respiratory secretions, for example — is the rule rather than the exception. Although colonization may precede infection, clinical experience with nosocomial pneumonias suggests quite the opposite; that is, VAP is uncommon following even prolonged airway colonization with gram-negative bacilli (GNB), S. aureus or Candida spp. In intubated patients without influenza in the ICU, MRSA airway colonization rarely results in subsequent MRSA VAP. The clinical confusion results from assuming airway culture isolates are reflective of the lower lung pathogens causing VAP. However, the problem clinically is to differentiate the related respiratory secretion isolate from those that are unrelated. For this reason, clinical correlation of culture results with characteristic clinical findings is always the final determinant of the clinical significance of organisms cultured from any body site.

Airway colonization with gram-negative bacilli, such as Enterobacter spp., Stenotrophomonas maltophilia and Burkholderia cepacia, in intubated ICU adults is common, but these organisms rarely cause VAP in normal hosts.

Clinicians who do not specialize in infectious disease often treat patients based solely on culture results rather than correlating culture results with characteristic lung findings to determine if the organism cultured from airway secretions is the pathogen causing the lung infection. Often, empiric treatment for potential VAP (pVAP) is based on “guilt by association” in intubated ICU patients. It is a common daily occurrence to observe one or more potential pathogens (MSSA, MRSA, Pseudomonas aeruginosa, Enterobacter spp., Acinetobacter spp., Serratia spp.) persistently cultured from respiratory secretions from patients with fever, leukocytosis, and infiltrates due to other causes.

PAGE BREAK

There are many mimics of VAP. Infiltrates may be due to noninfectious causes like congestive heart failure, chronic renal failure, systemic lupus erythematosus pneumonitis, pulmonary sarcoidosis, Wegener’s granulomatosis, bronchiolitis obliterans organizing pneumonia, and drug-induced pulmonary toxicity. If the underlying disorder causes fever, leukocytosis, etc., then the unwary clinician may relate it to the cause of pulmonary infiltrates. Without clinical correlation, the dilemma is further complicated when multiple potential organisms are cultured from respiratory secretions. Which, if any, are the cause of the lung infiltrates? P. aeruginosa VAP closely resembles that of MRSA VAP, with the same finding in a necrotic pneumonia with rapidly cavitating infiltrates. Fortunately, a negative MRSA respiratory secretion culture effectively rules out the diagnosis of MRSA VAP.

If MRSA VAP was as common as the literature suggests, it would have been frequently reported in autopsies. Clinically, from our influenza MRSA pneumonia experience, a fatal outcome is common. Most of the published studies on MRSA VAP are based on microbiologic/epidemiologic data, not clinical criteria. Many ICU intubated patients with fever, leukocytes and pulmonary infiltrates have MRSA airway colonization (30% to 40% reported as MRSA VAP). Adding MRSA-positive blood cultures does not affirm the diagnosis. The problem remains that positive blood cultures may be due to skin colonizers (gaining access to blood cultures) or reflect central venous catheter-associated bacteremia. In these same patients, a respiratory secretion culture that is positive for MRSA is common, but the bacteria may not be the cause of the infection.

Unnecessary use of vancomycin for empiric therapy

It makes little sense to empirically treat all patients with pVAP who have MRSA-positive respiratory secretion cultures with vancomycin (in addition to an anti-GNB antibiotic).

There are numerous implications for antimicrobial stewardship programs (ASPs) of unnecessary empiric vancomycin treatment of all patients with pVAP. First, in an economy with limited financial resources for health care, the needless cost of vancomycin is unjustifiable. Direct drug costs of vancomycin are only a portion of total costs. Indirect costs include IV administration ($10/IV dose), plus the cost of vancomycin levels.

The unnecessary use of vancomycin for empiric therapy of MRSA pVAP/VAP may have untoward microbiological consequences, as well. Because vancomycin does not penetrate well into respiratory secretions, it does not eliminate MRSA colonization. Low levels of vancomycin may select for multidrug-resistant GNB pulmonary pathogens. Second, vancomycin increases S. aureus cell wall thickness, which may result in permeability-mediated resistance to vancomycin as well as other antibiotics used to treat MRSA. Vancomycin permeability-mediated resistance in S. aureus manifests as “MIC creep,” increasing the minimum inhibitory concentration during or following therapy. Therapy with other MRSA antibiotics does not have these effects. Other potential collateral microbiologic effects include an increase in the prevalence of vancomycin-resistant enterococci in ICUs and hospitals.

Clinically, there are no differences in outcomes with or without empiric vancomycin added to an antibiotic with GNB activity, further suggesting that MRSA VAP is rare.

I have been long interested in pneumonias, and for decades I have diligently searched for cases of MRSA pneumonia. However, over the years, I have observed several cases of MRSA community-acquired pneumonia with influenza but only very rarely identified bona fide MRSA VAP. Using nonspecific diagnostic criteria, VAP has taken on mythic proportions. Although isolated cases certainly exist, MRSA VAP remains a rare event.

The reality is, if empiric monotherapy for GNB pVAP is selected wisely, there is no need for empiric MRSA coverage for pVAP/VAP. From a clinical perspective, carefully selected initial empiric monotherapy with a high degree of activity against GNB pulmonary pathogens, like P. aeruginosa, and low resistance potential remains the optimal clinical approach.

Disclosure: Cunha reports no relevant financial disclosures.