Review of: Contribution of the patient microbiome to surgical site infection and antibiotic prophylaxis failure in spine surgery

DOI: https://doi.org/10.58974/bjss/azbc055

Article review

Long, D. R. et al. Contribution of the patient microbiome to surgical site infection and antibiotic prophylaxis failure in spine surgery. Sci. Transl. Med. 16, eadk8222 (2024). https://www.science.org/doi/10.1126/scitranslmed.adk8222

Surgical site infection (SSI), historically was considered the greatest contributor to in-hospital mortality during the pre-antibiotic and pre-antiseptic surgery era. European SSI surveillance data today demonstrate a marked improvement in the rate of SSIs, with current rates ranging from 0.6%-1.2% for clean orthopaedic procedures to 1.5%-9.5% or more for clean-contaminated abdominal surgery.¹ Especially for clean procedures, preventative measures focus on antiseptic approaches directed at keeping exogenous infectious agents out of the surgical field. The elegant study by Long et al. forces surgeons, microbiologist, infection prevention experts and infectious diseases physicians to consider endogenous sources of infection for SSI pathogens and to consider prevention measures beyond those limited to the external environment alone.²

In this study, pre-operative nasal, rectal and skin swabs were taken from 210 patients that underwent instrumented posterior spinal fusion surgery. Clinical isolates from the 14 (6.8%) patients who developed an SSI were stored. Whole genomes of isolates and metagenomes of pre-operative swabs were compared using multiple bio-informatics approaches. Results indicated that 86% of SSI strains were present in the patient’s preoperative microbiome, raising the possibility that common exogenous sources of infection may be excluded as indicated, by comparing the metagenomes of clinical SSI isolates with those from 59 other patients that underwent spinal fusion surgery. Also, the antimicrobial resistance (AMR) genes present in the patient’s microbiome were comparable with the phenotypic resistance profiles of the SSI isolates. As mentioned by Drs Gilbert and Dr. Alverdy in their associated commentary, these findings might encourage surgeons to shift their choice of antibiotics from a “one-size fits all” approach towards one in which antibiotics are tailored to the individual patient using a pre-operative diagnostic approach involving metagenomics, microbiome prehabilitation and targeted prophylaxis.³

Establishing causation, however, is critical when interpreting these results. Members of the Enterobacterales order (Citrobacter, Enterobacter, Escherichia coli, Proteus and Serratia species) were most frequently identified as the cause of SSIs (55%), followed by Staphylococcus aureus (23%) and Enterococcus species (9%). All 14 patients who had an SSI had received cefazolin as peri-operative prophylaxis, which mainly covers virulent Staphylococci and Streptococci, but not Enterobacterales species. This study also demonstrated the existence of an anatomical cervico-lumbar gradient of microbes, ranging from gram-positive to gram-negative as dominant species. These findings could be considered as justification to extend the spectrum of prophylactic antibiotics depending on the location of the surgical wound. Interestingly, a significant portion of the pre-operative skin abundance consisted of anaerobes, however, no anaerobic coverage was applied against the SSI isolates. The authors did not use culture-independent methods to analyze wound cultures to rule out (co-)infection with uncultured anaerobes.

Surgeons and consulting (laboratory) specialists know how difficult it can be to 1) determine if a wound is truly infected, and 2) determine if a cultured microbe represents the disease-causing agent or a contaminant from the surrounding skin. When in doubt, physicians often decide on a defensive treatment strategy that includes prolonged antibiotic treatment directed at all cultured potential pathogens, especially when prosthetic material is implanted during surgery. During this study, clinical isolates were classified as SSI-causing pathogens if found in multiple independent wound cultures, preferably excised tissue or deep swab cultures. However, superficial swab cultures were included if the interdisciplinary treatment team considered them as potential pathogens that required antibiotic treatment. The authors did not describe how many times the SSI causing pathogen was based on a tissue biopsy, deep or superficial swab culture. Although reflective of generally accepted clinical practice, the lack of a standard algorithm in this study introduced the problem of inconsistent and misclassification of isolates as SSI causing pathogens versus wound contaminants. This is important since the authors argue that endogenous skin microbes cause the majority of SSIs.

While it is not surprising that wounds can become contaminated with strains from the patient’s own flora, making this distinction, although difficult, is necessary to claim causality. The authors could have reduced this bias by excluding isolates from superficial swabs and by solely relying on multiple tissue and /or fluid cultures. As a matter of practical importance, treatment decisions concerning antibiotic treatment are often not based on culture results alone, but are also influenced by the patient medical history, the physician and patient preference and many other factors. Also, there is no description of a protocol that recommends the minimum number of tissue biopsies and biopsy handling (clean surgical instrument for each biopsy), which is customary in the treatment of fracture-related and prosthetic joint infections.^4-6 Taken together, generalizations and reproducibility of the results, especially considering the small number of events remains problematic. The proximity of the incision of especially lumbar spinal fusion to the rectal microbiome, and the presence of foreign material (plates, screws, etc.) limit extrapolation of the current findings to SSIs related to other surgical procedures.

Nonetheless, studies such as presented by Long et al. have the potential to cause a paradigm shift in SSI prevention measures. Pre-operative screening for Staphylococcus aureus carriage and subsequent decolonization has shown to decrease SSIs and is recommended by the WHO for cardiothoracic and orthopaedic procedures.⁷ One might consider extending this strategy to, for example, Enterobacterales in thoracic and lumbar spinal fusion procedures. Since the main reservoir for Enterobacterales is the gastrointestinal tract (in contrast to S. aureus which mainly colonized the nasal cavity), selective decontamination of the digestive tract (SDD) might be justified in selected patients. However, the expected high number needed to treat (if effective) and micro-ecological effects must be considered as important downsides. Also, coverage of anaerobes is (intentionally) not included in SDD regimes with the goal of preserving the commensal microbiome to promote colonisation resistance. Finally, determining pre-operative antimicrobial resistance among strains in one’s microbiome might be useful as an approach to customise peri-operative antibiotic prophylaxis to the individual patient.

Strategies to nourish the beneficial commensal members of the microbiome while at the same time inhibiting or attenuating the virulence of potential pathogenic microbes seems to be a more sustainable approach to SSI prevention. For example, faecal microbiota transplantation is successfully used to restore gastro-intestinal microbial balance in patients with therapy-resistant Clostridium difficile infection. However, conceptually it is not a rational strategy to reduce colonisation with Enterobacterales as they are, in small abundance, part of the healthy commensal flora. Modulating the skin microbiota before surgery is an attractive alternative. Topical applications of coagulase-negative Staphylococci to provide protection against S. aureus infection is currently under investigation but remains experimental.⁸ Pre-habilitation of the surgical patient including dietary changes, smoke cessation and medication changes offer additional opportunities, given that a patient’s lifestyle is associated with the changes in gastrointestinal⁹ as well as skin microbiota composition.¹⁰ Next-generation antimicrobials, agents that target bacterial virulence without growth inhibition or bacterial killing, have gained attention over the last several decades but have not yet been fully established to be clinically useful.¹¹ Application of these strategies for SSI prevention have not been investigated so far.

Future pre-operative interventions tailored to the individual patient’s micro-environment seem to be more promising and most eco-neutral (i.e. sustainable). While expensive and time-consuming, the benefit of metagenomic screening, as used in this study, versus culture-based techniques, remains to be proven. Fast, cheap and small size sequencing platforms, such as Oxford Nanopore sequencing, combined with advanced, easy-to-use bioinformatic pipelines are needed. These next generation approaches will need to include surgeons, molecular and medical microbiology and infectious diseases specialists to advise on treatment decisions, including specific prophylaxis, lifestyle adjustments or postponing surgery based on individual dynamic risk profiles.

Today, limiting contamination by infectious agents from the surgical field has led to a drastic reduction of SSIs. Yet at the same time, it is clear that surgeon-scientists need to explore the role of the microbiome on SSI prevention given its emerging role as a potential source of infection. Approaches such as personalized SDD and peri-operative prophylaxis based on anatomical location of the surgical incision and microbial tracking using next generation sequencing is now possible and should remain under investigation. One lesson learned from this study is the importance of procuring samples from SSI wounds and to identify the causative organisms so that one can differentiate between those that are contaminants versus those that are actually driving the pathology. Future studies must be optimised and standardised so that the representative materials procured are properly analysed to avoid their misclassification. Despite its limitations, this well executed study shows the potential of unleashing advanced NGS strategies to drive SSI rates toward zero.