16S RRNA V3-V4 AMPLICON NEXT-GENERATION SEQUENCING IN THE DIAGNOSIS OF PERIPROSTHETIC JOINT INFECTIONS

Aim To date, no ultimate diagnostic gold standard for prosthetic joint infections (PJI) has been established. In recent years, next generation sequencing (NGS) has emerged as a promising new tool, especially in culture-negative samples. In this prospective study, we performed metagenomic analysis using 16S rRNA V3-V4 amplicon NGS in samples from patients with suspected PJI. Methods A total of 257 (187 culture-negative (CN) and 70 culture-positive (CP)) prospectively collected tissues and sonication fluid from 32 patients (56 revisions) were included. 16S rRNA V3-V4 amplicons were sequenced using Illumina's MiSeq (California, USA) followed by bioinformatic analysis using nf-core/ampliseq pipeline. Results We successfully sequenced 255 samples and detected a total of 105 microorganisms. These were mainly environmental microorganisms present in a small number of reads (≤100), indicating possible contamination. Pseudomonas spp. (non-aeruginosa species) was detected most frequently in 73% (187/255) of samples. The test showed limitations in species classification and identified microorganisms mainly at genus level. Significant differences in the number of reads were observed when comparing CN (≤100) and CP (≥1000) samples. In two CP, no bacteria were identified with sequencing, which is probably due to low bacterial load (1 CFU. Haemophilus spp. was detected with a significant number of reads (≥10000) in five samples from a single patient, in whom infection was considered likely according to EBJIS criteria, changing it to confirmed infection. Staphylococcus spp. was identified with ≥10000 reads in two CNs from an individual who was receiving antibiotic treatment at the time, had clinical signs of infection, and had a confirmed infection with S. lugdunensis one month earlier. Cutibacterium spp. with 36% (93/257) and Staphylococcus spp. with 34% (87/257) were detected with a minimal number of reads (≤100) in several CN, indicating possible contamination with normal skin microbiota. In one patient, Facklamia spp., an opportunistic pathogen, was detected in two samples by sequencing, but not by culture. Conclusions We consider 16S rRNA V3-V4 amplicon sequencing to be a promising tool; however, further studies are needed to clarify uncertainties regarding the interpretation of the results in combination with other criteria. Using this method, we were able to successfully confirm infection in two patients whose microbiological results were initially negative, leading to a change from likely to confirmed infection in one case. The thresholds and interpretation of the results are currently unclear, therefore the method is being used experimentally rather than diagnostically at the time of writing.