LAUSR

We read with great interest the recent article on bacterial presence in intervertebral disc samples by Fritzell et al. The authors have performed a multicenter analysis exploring the correlation of bacterial presence/growth in discs and have investigated whether Modic changes were associated. We have specific concerns on the validity of the methodology used and the arguments put forth by the authors in the manuscript. We also present our viewpoints on this topic, according to the recent literature. In methodology, while bacterial culture has been performed on all samples (disc, vertebra, subcutis, and skin) DNA analysis for bacterial presence was performed only on disc/vertebra samples. It would have been more appropriate if all samples were investigated similarly to be fool proof. It is unclear how the surgeons ensured unbiased distribution of representative specimens for culture and DNA analysis. Samples for culture were transported from seven different centres to a university laboratory at Gothenburg and during this, transportation time and delay in plating is expected to decrease the diagnostic yield of routine culture methods. However, despite the fact that Cutibacterium acnes is a slowgrowing fastidious organism, the authors have managed to isolate it in 35% of disc/vertebra (14/40) samples from disc herniation group and 25% of samples (5/20) from scoliosis group by routine culture techniques. It is surprising to see with all the above possibilities of reduction in diagnostic yield by culture techniques, the isolation of Cutibacterium acnes was possible in 25–35% of disc/vertebra specimens compared to bacterial DNA detection in only 2% of disc/ vertebra specimens. This raises serious concerns on two aspects: firstly, whether adequate representative disc specimens were sent for DNA analysis; and secondly whether there was any bias in sample distribution. The huge discrepancy in diagnostic yield between 35% of routine culture versus 2% DNA detection in discs is questionable given the very well-known fact that 16S rRNA sequencing for bacterial detection is far superior and advanced, and it outperforms culture techniques even in polymicrobial specimens with higher sensitivity and specificity [1]. Cummings et al. [2] analysed a mixture-consisting genome of 20 known microorganisms and found out 16S rRNA to be superior in detecting even low-abundance pathogens. Xia et al. [3] analysed 55 patients with ventilator-associated pneumonia and again found 16S rRNA-based sequencing to be superior to routine culture. We feel the authors need to introspect their methodology of DNA analysis or adequacy of sampling for the same. Such a massive difference in sensitivity is not expected and also has never been reported since the advent of 16S rRNA sequencing. Secondly, the authors write that culture is the most sensitive method for detecting bacteria, and for the same reason, they had chosen it as the primary method. In this era of molecular microbiology, we feel this statement is controversial and outdated. It is globally accepted that the current cultured bacteria represent only a small portion of total bacterial diversity, and DNA sequencing techniques have helped us in identifying the unknown and uncultured bacteria [4]. The Human Microbiome project, which has received 170$ million funding from the National Institutes of Health, was initiated by the Government of the USA to explore and understand the role of microbes in human health and disease. They have used mainly advanced culture-independent methods such as 16S rRNA sequences, and over 650 peerreviewed publications have been associated with this project from 2009 to 2017. Nearly 10,000 bacterial species have been identified to be associated with the human ecosystem using DNA sequencing techniques, and many of them have never been isolated previously from human specimens. With undoubted recent literature on the higher sensitivity * S. Rajasekaran [email protected]