Diagnostic Host Gene Expression Analysis by Quantitative Reverse Transcription Loop-Mediated Isothermal Amplification to Discriminate between Bacterial and Viral Infections

• Published in: Clinical chemistry (Baltimore, Md.) Vol. 68; no. 4; pp. 550 - 560
• Main Authors: Remmel, Melissa C, Coyle, Sabrina M, Eshoo, Mark W, Sweeney, Timothy E, Rawling, David C
• Format: Journal Article
• Language: English
• Published: England Oxford University Press 31.03.2022
• Subjects: Algorithms; Antibiotics; Bacteria; Bacterial diseases; Bacterial infections; Biomarkers; Blood; Etiology; Gene amplification; Gene expression; Infections; Medical diagnosis; Patients; Reverse transcription; Ribonucleic acid; RNA; Test systems; Viral infections; host response; molecular diagnostics; qRT-LAMP; LAMP; point-of-care; gene expression
• ISSN: 0009-9147; 1530-8561
• DOI: 10.1093/clinchem/hvab275

Abstract Background Early and accurate diagnosis of acute infections can help minimize the overprescription of antibiotics and improve patient outcomes. Discrimination between bacterial and viral etiologies in acute infection based on changes in host gene expression has been described. Unfortunately, established technologies used for gene expression profiling are typically expensive and slow, confounding integration into clinical workflows. Here we report the development of an ultra-rapid test system for host gene expression profiling from blood based on quantitative reverse transcription followed by loop-mediated isothermal amplification (qRT-LAMP). Methods We developed 10 messenger ribonucleic acid-specific assays based on qRT-LAMP targeting 7 informative biomarkers to discriminate viral from bacterial infections and 3 housekeeping reference genes. We optimized qRT-LAMP formulations to achieve a turnaround time of 12 min without sacrificing specificity or precision. The accuracy of the test system was verified utilizing blood samples from 57 patients and comparing qRT-LAMP results to profiles obtained using an orthogonal reference technology. Results We observed a Pearson coefficient of 0.90 between bacterial/viral metascores generated by qRT-LAMP and the reference technology. Conclusions qRT-LAMP assays can provide sufficiently accurate gene expression profiling data to enable discrimination between bacterial and viral etiologies using an established set of biomarkers and a classification algorithm.