Article Text

Download PDFPDF

Response to: ‘Circulating microbiome in blood of different circulatory compartments’ by Schierwagen et al
  1. Bastian Volker Helmut Hornung1,
  2. Romy Danielle Zwittink1,
  3. Quinten Raymond Ducarmon1,
  4. Ed J Kuijper1,2
  1. 1 Center for Microbiome Analyses and Therapeutics, Department of Experimental Bacteriology, Leiden University Medical Center, Leiden, the Netherlands
  2. 2 Netherlands Donor Feces Bank, Leiden University Medical Center, Leiden, the Netherlands
  1. Correspondence to Dr Bastian Volker Helmut Hornung, Center for Microbiome Analyses and Therapeutics, Department of Experimental Bacteriology, Leiden University Medical Center, Leiden 2312KT, The Netherlands; bastian.hornung{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

We read with interest the article by Schierwagen et al,1 who reported on their findings regarding the blood microbiome. Since our research group is working in a hospital setting, recently, we have also considered investigating the blood microbiome using collected blood samples that are potentially available. However, based on recent literature,2 3 we considered the presence of a blood microbiome very unlikely and did not further venture in this direction. We were therefore extremely excited to read this work, but the results quickly raised concerns in our research group. In another area of microbiota research, we recently investigated the usage of negative controls, and the reported results in this article look extremely similar to common contaminations derived from DNA extraction kits. From the 11 genera reported in figure 1D, 6 were previously reported as kit contaminants in the publications by Salter et al,4 Laurence et al,5 and Eisenhofer et al.6 The remaining five genera include Diaphorobacter, which is a close relative to Delftia,7 which is another kit contaminant,3 and Mucilaginibacter, a soil bacterium, but also three genera that cannot be excluded to be human pathogens. Further genera reported in the online supplementary, including the most abundant ones, can also be traced back to common contaminations (Pseudomonas, Janthinobacterium, Sphingomonas and Arthrobacter), as reported in the above-mentioned publications,4–6 and are comparable with contaminations found in other blood microbiome studies.3 Blauwkamp et al 8 recently showed that sepsis can be detected in cell-free DNA isolated from blood samples and reported only a very small number of identifications within these blood samples as not related to sepsis. All these results, together with the fact that Schierwagen et al 1 did not report using any negative controls, make the results reported in this article doubtworthy. We realise that the authors were able to culture multiple Staphylococcus and Acinetobacter strains from these blood samples. While both are clinically relevant organisms, the first one is also a known skin commensal, the second one potentially lives in water and both are known contaminants.6 Given that no negative controls for the culturing were reported, we also consider these results doubtful. The idea of having a blood microbiome is exciting, but we would like to ask the authors to confirm their results in a better controlled setting.



  • Contributors BVHH drafted the response. BVHH, RDZ, QRD and EJK reviewed and approved the response.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests EJK performed research for Cubist, Novartis and Qiagen, and participated in advisory forums of Astellas, Optimer, Actelion, Pfizer, Sanofi Pasteur and Seres Therapeutics. BVHH and EJK received an unrestricted grant from Vedanta Biosciences Inc. The companies and the Netherlands Donor Feces Bank had no role in writing the manuscript.

  • Provenance and peer review Not commissioned; internally peer reviewed.

  • Patient consent for publication Not required.