Abstract
Rapid identification of pathogens is critical for early and appropriate treatment of bloodstream infections. The various culture-independent assays that have been developed often have long turnaround times, low sensitivity and narrow pathogen coverage. Here, we propose a new multiplex PCR assay, MeltArray, which uses intact microbial cells as the source of genomic DNA (gDNA). The successive steps of the MeltArray assay, including selective lysis of human cells, microbial cell sedimentation, microbial cellular DNA extraction, target-specific pre-amplification and multiplex PCR detection, allowed the detection of 35 major bloodstream infectious pathogens in whole blood within 5.5 h. The limits of detection varied depending on the pathogen and ranged from 1 to 5 CFU/mL. Of 443 blood culture samples, including 373 positive blood culture samples and 70 negative blood culture samples, the MeltArray assay showed a sensitivity of 93.8% (350/373, 95% confidence interval [CI] = 90.7%-96.0%), specificity of 98.6% (69/70, 95% CI = 91.2%-99.9%), positive predictive value of 99.7% (95% CI = 98.1%-99.9%), and negative predictive value of 75.0% (95% CI = 64.7%-83.2%). The MeltArray detection results of 16 samples differed from MALDI-TOF and were confirmed by Sanger sequencing. Further testing of 110 whole blood samples from patients with suspected bloodstream infections using blood culture results revealed that the MeltArray assay had a clinical sensitivity of 100% (9/9, 95% CI = 62.8%-100.0%), clinical specificity of 74.5% (70/94, 95% CI = 64.2%-82.7%), positive predictive value of 27.3% (95% CI = 13.9%-45.8%), and negative predictive value of 100.0% (95% CI = 93.5%-100.0%). Compared with metagenomic next-generation sequencing, the MeltArray assay displayed a positive agreement of 85.7% (6/7, 95% CI = 42.0%-99.2%) and negative agreement of 100.0% (4/4, 95% CI = 39.6%-100.0%). We conclude that the MeltArray assay can be used as a rapid and reliable tool for direct identification of pathogens in bloodstream infections.