Comparison of clinical metagenomics with 16S rDNA Sanger sequencing for the bacteriological diagnosis of culture-negative samples

Background: Currently, diagnosis of bacterial infections is based on culture, possibly followed by the amplification and sequencing (Sanger method) of the 16S rDNA- encoding gene when cultures are negative. Clinical metagenomics (CMg), i.e. the sequencing of a sample's entire nucleic acids, may allow for the identification of bacteria not detected by conventional methods. Here, we tested the performance of CMg compared to 16S rDNA sequencing (Sanger) in 50 patients with suspected bacterial infection but negative cultures. Methods: This is a prospective cohort study. Fifty patients (73 samples) with negative culture and a 16S rDNA sequencing demand (Sanger) were recruited from two sites. On the same samples, CMg (Illumina NextSeq) was also performed and compared to 16S rDNA Sanger sequencing. Bacteria were identified using MetaPhlAn4. Results: Among the 73 samples, 20 (27 %, 17 patients) had a clinically relevant 16S rDNA Sanger sequencing result (used for patient management) while 11 (15 %, 9 patients) were considered contaminants. At the patient level, the sensitivity of CMg was 70 % (12/17) compared to 16S rDNA. In samples negative for 16S rDNA Sanger sequencing (n = 53), CMg identified clinically-relevant bacteria in 10 samples (19 %, 10 patients) with 14 additional bacteria. Conclusions: CMg was not 100 % sensitive when compared to 16S, supporting that it may not be a suitable replacement. However, CMg did find additional bacteria in samples negative for 16S rDNA Sanger. CMg could therefore be positioned as a complementary to 16S rDNA Sanger sequencing.