MEMPHIS—Although treatment with a single antibiotic is standard for bacteremic pneumococcal pneumonia, it may be suboptimal, recent data suggest. In a retrospective study of 225 adults hospitalized for this severe form of pneumonia, the risk of death was significantly greater with monotherapy than with combination therapy, particularly among sicker patients.[1]

“These findings defy conventional wisdom,” Richard G. Wunderink, MD, one of the study authors, told PULMONARY REVIEWS. “We have always thought that if you knew the cause and sensitivity of pneumonia, one antibiotic ought to do.”

HOW PATIENTS WERE IDENTIFIED

Dr. Wunderink and his colleagues examined their health system’s records to identify all patients who had been treated for bacteremic pneumococcal pneumonia during a 4-1/2-year period. To be included in the study, patients had to have a diagnosis of community-acquired pneumonia (CAP) as well as a positive blood culture for Streptococcus pneumoniae obtained within 48 hours of presenting to the hospital. They also had to exhibit either one major or two minor clinical signs of acute pneumonia. The major signs included fever (37.8°C or higher), hypothermia (36.0°C or lower), cough, and sputum production. The minor signs were dyspnea, pleuritic pain, clinical evidence of lung consolidation, and a leukocyte count higher than 10,000 cells/µL or lower than 4,500 cells/µL.

To remove antibiotic resistance as a confounder, the authors excluded all patients not given at least one dose of an effective antibiotic within 24 hours of presentation and all patients whose pneumococcal isolate was resistant to the initial antibiotic chosen. They also used conservative definitions of resistance that would favor monotherapy. For example, a minimum inhibitory concentration of 2 µg/mL or higher was considered high-level resistance to penicillin, cefotaxime, ceftriaxone, and levofloxacin.

Patients hospitalized within the past 30 days were also excluded, as were immunocompromised patients.

Based on culture and sensitivity results, the patients’ empiric antibiotic therapy was classified as single effective therapy (SET), dual effective therapy (DET), or more than DET (MET). Since few patients received MET and those who did were significantly sicker than the rest, the authors’ analysis focused mainly on SET and DET.

GREATER MORTALITY WITH SET

Twenty-nine (12.9%) of the patients died, and all deaths occurred among those with a Pneumonia Severity Index (PSI) score above 90. The mortality rate was 18% in the 99 SET patients, 7% in the 102 DET patients, and 17% in the 24 MET patients. Overall, the odds of death in the SET group were three times greater than those in the DET group and more than twice those of the DET and MET groups combined.

The increased mortality risk in the SET patients was particularly surprising because these patients had a much lower predicted mortality than did the DET group. The authors therefore used logistic regression modeling to control for predicted mortality and found that the odds ratio for death was 6.4 when patients were given SET rather than DET. When the analysis was confined to the most severely ill patients (those with a PSI score above 90), the odds ratio for death associated with SET was still 5.5. Even after the analysis was controlled for possible confounders—including death within the first 48 hours, age, sex, and underlying chronic disease—the odds ratio for death was 4.9 in the patients given SET.

A variety of antibiotics had been administered to the patients in this study. About half the patients given SET had been treated with a quinolone; others had received a third-generation cephalosporin, macrolide, or other ß-lactam. The most frequently administered form of DET was the combination of a third-generation cephalosporin and either a macrolide or quinolone. The study found no survival advantage (or disadvantage) for any particular antibiotic or combination of antibiotics.

WHY DOES DET WORK?

DET may increase survival in patients who have bacteremic pneumococcal pneumonia through synergy—two antibiotics may kill more pneumococci than one—or by producing a better immune response than does monotherapy, the authors suggested. It is also possible, they said, that DET is more effective against the coinfection that appears to be common in CAP. Alternatively, it may simply be that monotherapy is inadequate.

Because their study was retrospective, the authors acknowledge that its results should be interpreted with caution. For example, they were unable to control for the amount of time until patients received the first dose of an effective antibiotic or for subsequent changes in antibiotic therapy.

The wide range of antibiotics and antibiotic combinations patients received was also a limitation because it made it impossible to draw conclusions about specific agents or regimens. Moreover, the authors were unable to determine the optimal duration of DET. Nonetheless, “our findings are at least suggestive that there is a reason to treat S pneumoniae with two antibiotics, especially if the patient is sicker,” said Dr. Wunderink, Director of Research at the Methodist Le Bonheur Healthcare Foundation in Memphis.

The results are strong enough to justify a prospective, randomized, double-blind trial of monotherapy versus combination treatment for bacteremic pneumococcal pneumonia, he added. Such a trial should focus on patients with a PSI score above 90, since SET appears to be adequate in those with a lower PSI score, he suggested.

—Timothy Begany

Reference
1. Waterer GW, Somes GW, Wunderink RG. Monotherapy may be suboptimal for severe bacteremic pneumococcal pneumonia. Arch Intern Med. 2001;161:1837-1842.

Return to table of contents