News
Guest Editorial - Selecting Appropriate Agents For Treatment of Pneumonia
General guidelines like those from the American Thoracic Society can help in the selection of empiric treatment of pneumonia, both community- and hospital-acquired.
Cefepime has activity against many of the same pathogens as cefotaxime, with improved coverage against Pseudomonas aeruginosa and many strains of Enterobacter and Serratia, as well as better activity against methicillin-sensitive Staphylococcus aureus. However, its role in the treatment of pneumonia has not yet been defined.
Pneumonia remains a major cause of morbidity and mortality in the US. The annual incidence of community-acquired pneumonia (CAP) in the US is estimated to be approximately 2 to 4 million cases, with about 900,000 of these requiring hospitalization.[1] Mortality among hospitalized CAP patients is reported to be between 10% and 25%, but it may approach 50% among individuals treated in intensive care units (ICUs).[2] Despite aggressive and invasive diagnostic testing, a pathogen is not isolated in a majority of pneumonia cases; therefore, treatment is necessarily empiric.
Guidance in the empiric management of CAP in immunocompetent adults is provided in statements such as those released by the American Thoracic Society (ATS).[3] Patients are classified into loose groups based on clinical data (ie, age, presence of underlying disease, severity of infection) that provide clues to the potential microbiologic etiologies of their pneumonia. Given this group of most likely pathogens, rational antibiotic choices can be made.
The annual incidence of hospital-acquired pneumonia (HAP) is estimated at approximately 5 to 10 cases per 1000 hospital admissions[4]; the incidence increases as much as 6- to 20-fold among patients receiving mechanical ventilation.[5] The crude mortality for all patients with HAP ranges from 50% to 70%,[6] with deaths directly attributable to infection estimated at 33% to 50% of the total.[7] Delays in antibiotic treatment have been associated with higher mortality rates, especially in individuals infected with Pseudomonas aeruginosa.[8]
Following the earlier model for CAP, the ATS released a consensus statement on the management of HAP.[9] Patients are stratified into discrete groups, each with its own likely set of pathogens. The classifying factors include the severity of the current pneumonia, the presence or absence of risk factors for specific organisms, and the time of onset of the pneumonia. Again, this approach provides a basis for the selection of empiric antibiotics effective against a probable group of infecting organisms.
Third-generation cephalosporins such as ceftriaxone or cefotaxime have formed the backbone of empiric therapy for hospitalized CAP patients. Cefotaxime's spectrum of activity encompasses many of the likely pathogens; however, it has no activity against organisms such as Legionella or Chlamydia, which is an important consideration in patients with severe CAP being treated in the ICU.[3] Therefore, many practitioners have added a macrolide to their empiric regimen for hospitalized CAP patients.
Cefepime, the only fourth-generation cephalosporin commercially available in the US, has activity against many of the same pathogens as cefotaxime, with improved coverage against P aeruginosa andmany strains of Enterobacter and Serratia; it also has better activity against methicillin-sensitive Staphylococcus aureus (MSSA) than most third-generation cephalosporins. However, it too has poor activity against Legionella and Chlamydia.
In the current issue, a study by Willis and colleagues[10] compares the bacterial and clinical response of hospitalized pneumonia patients following treatment with either cefotaxime or cefepime. Unfortunately, the trial included a heterogenous mix of patients with CAP and HAP, with no indication of the severity of cases (ie, ICU vs non-ICU patients). The importance of such distinctions is directly related to the group of likely pathogens that the clinician should consider in choosing empiric coverage for the pneumonia. For instance, I have no problem choosing one or the other agent for the hospitalized CAP patient not requiring ICU treatment, but I would be reluctant to rely on either as monotherapy for the patient with severe HAP requiring mechanical ventilation. Patients in the latter group are at high risk for virulent organisms such as P aeruginosa, Acinetobacter species, and MSSA[9]; this list of likely pathogens makes the argument for multidrug combination therapy rather compelling.[11]
I acknowledge that the bacteria isolated from this heterogenous group are limited and that all seem to be sensitive to the chosen agents (see Table IV).[10] However, I continue to be skeptical of sputum culture results. Many people have access to antibiotics from past uncompleted therapy and frequently self-administer medications before being seen by a physician. It is estimated that 20% to 45% of CAP patients have had recent antibiotic treatment before presenting for medical attention,[12] and a single dose of appropriate antibiotic results in negative sputum cultures in patients with pneumonia caused by S pneumoniae.[13] Hospitalized patients are likely to have received several antibiotics before their current bout of HAP. Up to 30% of CAP patients are unable to produce a positive sputum sample.[14] In addition, sputum cultures may be negative or produce other bacteria even when a "gold standard" sample (blood culture, pleural fluid culture) has revealed a pathogen.[13]
Concomitant antibiotics were allowed in this study, introducing another factor that should be taken into account. Fourteen patients (5 in the cefepime group and 9 in the cefotaxime group) received 20 different antifungal and/or antimicrobial agents. Many of the antimicrobial agents have activity against pathogens likely to cause the patients' pneumonia, especially in the group with CAP. Thus, clindamycin, rifampin, trimethoprim/sulfamethoxazole, and vancomycin contribute to the clinical and microbiologic response of the patients. While none may be considered primary treatment for either CAP or HAP, their combination with either of the 2 study drugs may be sufficient for treatment of pneumonia, making it problematic to attribute cure or improvement solely to the study drug.
Four patients (1 in the cefepime group and 3 in the cefotaxime group) were judged to have failed to respond to therapy. The authors do point out that 2 of the 3 patients in the cefotaxime group had polymicrobial infections, although 4 other patients thought to have a polymicrobial infection were considered cured or improved. The other patient was thought to be infected with S aureus. I am curious to know if a pathogen was isolated from the single patient with treatment failure in the cefepime group; patients in this group with polymicrobial infections (2) were considered cured or improved.
The study design allowed switching from IV to IM study drug "to continue or complete a course of treatment." It is unclear to me what the total duration of therapy was in either study group: if patients were switched to oral drugs upon discharge from the hospital, or if patients continued receiving the IM study drug after discharge. In these days of limited reimbursement and capitation, there is great pressure brought to bear on clinicians for early discharge with a switch to oral medications. Home IV or IM therapy may be a limited option, possible only for patients unable to have PO intake.
This study does contribute to our knowledge about the use of cefepime and its safety compared with a frequently used third-generation cephalosporin, cefotaxime. However, this study alone is not sufficient to determine whether the agent represents adequate treatment for all hospitalized patients with pneumonia. The introduction of "extended-spectrum" fluoroquinolones will also have some impact on the choice of agents for the empiric therapy of such patients.