Blood biomarker algorithms for the diagnosis of mycoplasma pneumoniae respiratory infections

The correct diagnosis of acute infections as to bacteria, mycoplasma or virus is a clinical challenge and has a great impact on the therapeutic decisions. Current diagnostic tests of mycoplasma pneumoniae infections of the respiratory tract such as PCR and serology are either somewhat unreliable or slow and do not entirely meet the clinical needs of accurate and fast diagnosis. The aim of this report was to examine a panel of candidate biomarkers and their capacity to distinguish mycoplasma pneumoniae respiratory infections from respiratory infections caused by either bacterial or virus.Method: Patients with confirmed etiology of their acute respiratory infections (n = 156) were included of which 28 patients were diagnosed with mycoplasma pneumoniae. Blood was taken before any antibiotics treatment and analysed for Azurocidin (HBP), Calprotectin, CRP, Human Neutrophil Lipocalin (HNL), Interferon gamma-induced Protein 10 kDa (IP-10), Procalcitonin (PCT), Thymidine Kinase 1 (TK1), THE-Related Apoptosis-Inducing Ligand (TRAIL).Results: Individually the concentrations of IP-10, TK1 and P-HNL distinguished mycoplasma pneumoniae from bacterial infections with AUCs of 0.79-0.85. However, in combination, TK1 with either IP-10 or P-HNL showed an AUC of 0.97-0.95. In the distinction between mycoplasma pneumoniae and viral respiratory infections CRP, Calprotectin and TRAIL showed individual AUCs of 0.94-0.84. Together with either P-HNL dimer or PCT, CRP showed AUCs of 0.97.Conclusion: Our results indicate that it may be possible to design useful diagnostic algorithms of biomarkers that could help distinguish mycoplasma pneumoniae from respiratory infections caused by bacteria or virus. The development of rapid point-of-care assays based on such algorithms could be clinically useful tools in the therapeutic decision-making.