In the June 15th 2014 issue of Clinical Infectious Diseases, Huang and colleagues identified and validated a novel biomarker called Lipocalin-2 (Lpc-2) that may help distinguish children with severe bacterial pneumonia or bacteremia from those with non-bacterial illnesses (Huang H. Clin Infect Dis. 2014; 58: 1707-1715.). Plasma samples from Gambian infants and children aged 2 to 59 months were obtained and underwent mass spectrometry. When samples from healthy patients, patients with non-severe pneumonia, and patients with severe pneumonia were compared, eight proteins were identified that incrementally increased from healthy controls to patients with severe pneumonia. Lpc-2, a protein known to be expressed in neutrophils, which functions to sequester iron and thus limit bacterial growth, was among these proteins. They determined that elevated Lpc-2 levels had a sensitivity and specificity of approximately 70% in predicting severe disease and a sensitivity of 70% and specificity of 94% in predicting likely bacterial pneumonia. When used with clinical criteria, this marker increased diagnostic specificity for severe pneumonia to 82%.
Though this study has several limitations, I applaud the authors in their efforts to identify objective markers of bacterial infection. Among bacterial infections, pneumonia presents unique challenges in diagnosis in that there is no absolute “gold standard” clinical, laboratory or radiographic finding, often leaving the clinician uncertain about need for antibiotic therapy.
Many overlapping clinical syndromes, including bronchitis, bronchiolitis and upper respiratory infection may cause many of the same signs as bacterial pneumonia. Laboratory criteria, including white blood cell count and C-reactive protein, are non-specific. Finally, chest X-ray findings of atelectasis, chronic lung disease, pulmonary edema and viral pneumonia can be indistinguishable from bacterial pneumonia.
In the developing world, as pointed out by the authors, use of these biomarkers could (with more study) result in development of point-of-care testing to identify those children who require referral and/or antibiotic therapy, and thus better allocate scarce resources. Here in the United States, we also face a resource allocation issue in treating respiratory infections, though ours is more likely due to antibiotic over prescription. Among children admitted to US children’s hospitals, antibiotic therapy for pneumonia accounts for more antibiotic days than any other condition. In US intensive care units, suspected lower respiratory tract infection is the most common reason for antibiotic use. If specific markers can be developed that allow identification of those children with bacterial who will benefit from antibiotic therapy, we can likely save a great deal of unnecessary antibiotic exposure. This and other biomarkers warrant further investigation.