© 2006 American Thoracic Society doi: 10.1164/rccm.200604-535ED
A CAP on Antibiotic DurationNorthwestern University Feinberg School of Medicine Chicago, Illinois Why do we use antibiotics for community-acquired pneumonia (CAP)? Although this seems like a question that any second-year medical student could readily answer, the antibiotic management of CAP suggests that many clinicians have forgotten the reasons. In this issue of the Journal (pp. 84–93), Christ-Crain and colleagues (1) present results of a clinical trial that focuses on antibiotic management of CAP. They confirmed their primary hypothesis that antibiotic management based on serum procalcitonin (PCT) levels would decrease overall antibiotic use for patients with CAP presenting to the emergency department, as compared with usual practice. In addition, their results provide insights into issues relating to antibiotic therapy for CAP. The primary principle of antibiotic treatment of CAP is that antibiotics are designed to kill or at least decrease proliferation of bacteria. The purported advantage of measuring serum PCT is that only bacterial infections cause its elevation. This same group has previously demonstrated that PCT levels can select which patients presenting with a febrile respiratory illness can safely avoid antibiotics or have them withdrawn (2). In the present study, they found that approximately 15% of patients with CAP had very low PCT levels, consistent with the absence of bacterial infection. Avoiding or stopping antibiotics in this group was not associated with adverse outcome. In other CAP studies, about 18% of infections have been positively identified as being caused by viruses (3). Since there are no bacteria to kill, this group should derive no benefit from antibiotic therapy.
Is this reasoning correct? Maybe. In the previous clinical trial (2), patients admitted with status asthmaticus had low PCT levels. However, a very recent study has demonstrated that 61% of these same patients have seroconversion for Mycoplasma pneumoniae or Chlamydophilia pneumoniae (4). These patients clearly respond to antibiotic treatment in placebo-controlled trials. Unfortunately, evidence for atypical pathogens other than Legionella was not systematically sought in the study of Christ-Crain and colleagues (1) nor was antibiotic therapy broken down by PCT level. The possibility remains that PCT is more of a marker of "typical" bacterial CAP and may not be elevated with some atypical pathogens, especially C. pneumoniae and Mycoplasma (5). Antibiotic therapy directed at these atypical infections has not been conclusively demonstrated to alter clinical outcome (6), even in patients with documented seroconversion. A low PCT level may be the appropriate screening criterion for entry into a prospective randomized, placebo-controlled trial of antibiotic treatment of atypical CAP, with subsequent confirmation of pathogens by seroconversion. In addition, the findings of this study can be seen as suggesting that a The threatened avian influenza pandemic may provide another opportunity to demonstrate whether antibiotics affect the course of viral CAP. Mortality in influenza is generally recognized to be more dependent on bacterial superinfection than the original influenza infection. Do 24 hours of prophylactic antibiotics decrease this risk? Since many of the patients in both of the PCT studies of Christ-Crain and colleagues (1, 2) may have received one dose of antibiotics by the time that PCT levels were available, their data don't exclude the possibility that this one dose was beneficial. Epidemiologic data indicating that CAP mortality has increased in England after emphasizing avoidance of antibiotics in patients with viral-type illnesses (7) suggest that the question of the potential prophylactic utility of antibiotics in viral pneumonias is a critical one to settle before avian influenza or some other influenza epidemic hits. The availability of rapid diagnostic tests for influenza makes a randomized placebo-controlled trial of a short-course prophylactic antibiotic regimen clinically feasible. The pattern seen in the severe acute respiratory syndrome (SARS) epidemic suggests that the present default mode is routine antibiotics even when a positive viral diagnosis is made (8). The other forgotten principle is that antibiotics only rarely "cure" pneumonia. Their role is to tip the balance in the battle between bacteria and host defense toward the host. If host response is inadequate, as in chemotherapy-induced neutropenia, resolution of the pneumonia often does not occur until return of adequate neutrophil counts, no matter how potent or long the duration of antibiotics. One of the incongruities of CAP has been the persistent mortality of bacteremic pneumococcal pneumonia even when exquisitely penicillin sensitive. In patients dying in the first few days of admission, no viable bacteria are found at autopsy. If the bacteria have been eliminated in these patients who die, why should antibiotics be continued for 2 weeks in patients who are less ill? Unfortunately, PCT levels do not return to low levels for 2 weeks in bacteremic patients and normalization of levels is also delayed in severe CAP (1). The pattern of PCT response in such patients appears to be more of a marker of ongoing inflammation than persistent bacterial infection. PCT-based therapy therefore did not shorten the duration of antibiotics. Even more disturbing is the apparent finding that PCT levels were not consistently elevated in patients with empyema. Whether other localized metastatic sites of infection, such as meningitis, pericarditis, or septic arthritis, will result in elevations of PCT is also called into question. Unfortunately, patients with severe or complicated CAP are precisely those in whom an accurate marker of response to therapy would be valuable (9), similar to what has previously been suggested for ventilator-associated pneumonia (10). PCT does not appear to reliably discriminate these patients. The major finding of the present study of PCT-based antibiotic therapy was that the duration of antibiotic therapy was decreased to an average of 7 days versus 14 days in the control group (1). No other secondary variables, including hospital length of stay and mortality, were different when therapy was based on PCT levels. However, 14-day treatment of CAP is no longer the standard. Clinical trials and newer guidelines for management of CAP suggest that shorter courses, 7 days or less, should be used (11, 12). Therefore, the use of PCT-guided therapy is unlikely to make a major impact if current guidelines are followed. The cost-effectiveness of measuring PCT levels was marginal in patients who received the standard course of antibiotics and would disappear entirely if the duration of treatment were closer to 7 days.
The major implication of this study (1) is to support putting a cap on the duration of antibiotic therapy for CAP, similar to that of ventilator-associated pneumonia (13). Needless prolongation of the duration of antimicrobial treatment is more likely to select for antibiotic resistance or, even worse, place false hope in a failing regimen if ongoing infection really is the cause. The greatest clinical benefit of PCT-directed therapy still appears to be selection of patients who may not need antibiotics at all or, at least, not FOOTNOTES Conflict of Interest Statement: R.G.W. participated as an investigator in the Orthro-McNeil study mentioned. He has also served on an advisory board for Orthro-McNeil (total of $3,500 over 2 years) and Sanofi-Aventis (total of $3,000 in the last year). REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
-lactam is unlikely to benefit patients with low PCT levels. 
