Abstract
Aim
metagenomic next-generation sequencing (mNGS) has shown to be a useful method for pathogen detection in prosthetic joint infections (PJI). The technique promises to minimize the PJIs without the known causative agent. Our study aimed to compare diagnostic accuracies of cultures and mNGS.
Method
In this study, a meta-analysis following PRISMA recommendations was performed. PubMed and OVID Medline databases were used for article search. The studies using mNGS whole-genome sequencing method and the ones where PJI diagnosis was based on one of the currently recognized criteria were included. Studies were excluded if they comprised less than twenty cases, the ones with insufficient data for the analyses (true positive, true negative, false positive and false negative values for both mNGS and culture results) and publications with strong duplication bias. Univariate metanalysis using a random-effect model has been performed in R studio with a “meta” package. Pooled sensitivity and pooled specificity were calculated.
Results
Seven studies with a total of 822 cases were included in the meta-analysis, 476 cases defined as PJI and 346 controls. Two studies used IDSA (Infectious Diseases Society of America) diagnostic criteria and the Illumina HiSeq 2500 platform for sequencing and five studies used MSIS (MusculoSkeletal Infection society). Four of those used the BGISEQ-500 sequencing platform. For one study there was no data available. Studies were performed on prosthetic hip and knee joints. Through meta-analysis, it was observed that mNGS technique is more sensitive than cultures with 90% (CI 79%– 95%) and 74% (CI 68%-79%) respectively (p=0.006). The specificity between methods was similar, for mNGS reaching 94% (CI 89%-96%) and for cultures 97% (CI 90%-99%) (p=0.285). In the PJI group, 117 new possible pathogens that were not isolated by microbiological culture were detected by the mNGS, most frequently anaerobes and coagulase-negative staphylococci both in 20/117 (17.1%) cases. Fourteen new organisms were detected in the control group and were mostly regarded as contaminants.
Conclusions
Metagenomic sequencing has shown to be more sensitive than microbiological cultures in pathogen detection and thus has a great potential to improve the diagnosis and treatment of PJI. More studies on different prosthetic joints and comparing different diagnostic criteria for PJI would be needed to better understand the true diagnostic power of this method.
