The Role of the Gut Microbiome on the Development of Surgical Site Infections

Abstract

Despite advances in antisepsis techniques, surgical site infection remains the most common and most costly reason for hospital readmission after surgery. Wound infections are conventionally thought to be directly caused by wound contamination. However, despite strict adherence to surgical site infection prevention techniques and bundles, these infections continue to occur at high rates. The contaminant theory of surgical site infection fails to predict and explain most postoperative infections and still remains unproven. In this article we provide evidence that the process of surgical site infection development is far more complex than what can be explained by simple bacterial contamination and hosts' ability to clear the contaminating pathogen. We show a link between the intestinal microbiome and distant surgical site infections, even in the absence of intestinal barrier breach. We discuss the Trojan-horse mechanisms by which surgical wounds may become seeded by pathogens from within one's own body and the contingencies that need to be met for an infection to develop.

Despite advances in antisepsis and surgical techniques, surgical site infection remains the most common and most costly reason for hospital readmission after surgery. ¹ Wound infections are conventionally thought to be directly caused by wound contamination from intraoperative fecal spillage or various environmental sources due to breaches in sterility. To combat wound infections, surgeons employ stringent decontamination protocols, rigorous infection control practices, minimally invasive surgery, and enhanced recovery pathways. ^{2 3 4 5} These interventions have led to significant decline in wound infections over the years. Despite strict adherence to the standard methods of antisepsis, surgical site infections continue to occur at high rates and following clean operations.

The premise that all surgical site infections arise due to an external contamination event remains unproven. We are slowly beginning to realize that this explanation is too simplistic and that far more is at play. Of course, in certain clinical situations such as massive intraoperative contamination from perforated viscous in a severely immunocompromised patient undergoing an emergency operation, it is not reductive to frame wound infection as the inability of the host to clear the contaminating microbial burden. ⁴ However, these extreme conditions are exceptions rather than the rule. And still, most such situations do not result in infection. ⁶ Most patients presenting with perforated diverticulitis with purulent or fecal peritonitis will go on to have emergency surgery and never develop a surgical site infection. Most immunosuppressed patients undergoing a solid organ transplant will go on to recover from the operation and never develop a surgical site infection. The contaminant theory of surgical site infection also fails to predict and explain the more common scenario of an infection that occur after an elective, clean operation in a healthy immunocompetent host.

In this article we provide evidence that the process of surgical site infection development is far more complex than what can be explained by bacterial contamination and hosts' ability to clear the contaminating pathogen. We further posit that the intestinal bacteria are a likely source of distant surgical site infections, even in the absence of intestinal barrier breach. We suggest that the occurrence, course, and outcome of these infections are directly influenced by the state of one's intestinal microbiome. Finally, we discuss the possible mechanisms by which surgical wounds may become seeded by pathogens from within one's own body and the contingencies that need to be met for an infection to develop.

Contamination Cannot be the Sole Explanation of Surgical Site Infections

Over the past few decades, we have seen a dramatic reduction in surgical site infections. ^{6 7} This is a direct result of improvements in infection control practices, decontamination protocols, minimally invasive surgery, and enhanced recovery pathways. ^{2 3 4 5 6 7} Yet, serious surgical site infections continue to occur despite these measures and their strict enforcement.

Why and how is it, that a subset of patients undergoing clean elective operations will go on to develop devastating wound infections, while majority of patients undergoing bowel resections with clear intraoperative source of bacterial contamination will never suffer infectious complications? These observations are usually explained in the context of the burden of contaminating bacteria weighed against the competence of the host's immune system. ⁸ However, numerous studies have shown that intraoperative bacterial contamination does not actually correlate with subsequent clinical infection. ^{9 10 11 12 13} In the largest culture-based study, 276 patients (out of 1,852 studied) had positive intraoperative wound cultures and 13.8% subsequently developed surgical site infections. ⁹ This means that more than 80% of patients with contaminated wounds did not go on to develop infections. Even more importantly, most pathogens recovered from the infected wounds did not match those found in the surgical site intraoperatively. Finally, a substantial number of patients with negative intraoperative wound cultures went on to develop wound infections. Although dated, the findings of this study are striking and match the results of more recent publications. ^{10 12 13} The evidence that most surgical site infections are a direct result of intraoperative contamination, whether it be by contiguous spread of organisms from the environment or the opened gastrointestinal tract, remains lacking. The contamination theory of wound infections does not appear to be the sole explanation or the main cause of infections following elective surgery. Surgical site infections continue to occur in the absence of intestinal content spillage, intestinal manipulation, and intraoperative contamination, and their cause remains elusive. ^{5 14}

Especially puzzling are infections that occur late in the process of surgical recovery, weeks to months following the operation. ¹⁵ These are most often associated with the presence of a foreign body such as a vascular graft or a joint prosthesis. Majority of these infections are due to bacteria also found on skin or found colonizing surfaces (i.e., Staphylococcus aureus ). However, the use of vancomycin powder prophylaxis in orthopaedic and neurosurgery has led to an increase in rates of polymicrobial Gram-negative prosthesis infections. ^{16 17} Furthermore, positive intraoperative wound and prosthesis cultures do not correlate with development of postoperative infections and contaminating bacteria rarely are the ones recovered from wound infections. ^{18 19 20} The exact source of these pathogens remains unclear and explanation for how they gain access to the surgical site and result in infection is lacking.

Underappreciated Reservoir of Pathogens Associated with Surgical Site Infections—The Gut

The intestinal microbiome plays an important role in the development of various physiologic systems, as well as establishing and maintaining homeostasis. Our physiology, in turn, imparts influence on the intestinal microbiome and its health. This occurs through an endless loop of ongoing signal exchange across the intestinal barrier. We are just beginning to understand how these interactions play a key role in recovery from surgical insult. There are reports from carefully designed animal studies, showing that supplementation of probiotics in animal feed can increase the speed of wound healing ²¹ and that disruption of normal intestinal flora can result in an exaggerated tissue injury caused by increase in overactive circulating neutrophil subsets. ²² Intestinal microbes can be linked to the occurrence, course, and outcome of postoperative infectious complications.

Beginning in the early 1900s, surgeons noticed a correlation between “intestinal antisepsis” with various oral preparations and improved surgical outcomes. ^{23 24} This eventually led to the development and widespread use of “bowel preparation,” a combination of mechanical purgatives and oral antibiotics. ²⁵ This practice combined with other forms of antisepsis led to a dramatic improvement in the incidence of postoperative infectious complications. ²⁶ As the complication rates continued to fall, the practice of bowel preparation before surgery was called into question and abandoned all together by many. A few randomized controlled trials comparing mechanical bowel preparation to no preparation at all (omitting oral antibiotics) failed to show benefit in reducing postoperative infections. ^{27 28 29} However, with renewed interest in the historical data, the practice of oral antibiotic bowel preparation combined with mechanical purgatives has been revived. Newer studies prove the importance of oral antibiotics, showing clearly a considerable reduction in infectious complications after intestinal surgery. ^{30 31 32} These studies clearly link intestinal microbes to postoperative wound infections. Yet, the precise route by which the intestinal pathogens gain access to the wound and the mechanism by which intestinal antisepsis prevents their translocation remain elusive. The reduction in infections after antibiotic bowel preparation is commonly explained in the context of a lower burden of contaminating bacteria following empiric killing of intestinal microbes and hence improved ability of the host to clear them. However, we have new evidence for an alternate route of translocation of intestinal microbes to the surgical site in the absence of a direct intestinal barrier breach—the Trojan-horse hypothesis of wound infections.

Trojan-Horse Hypothesis of Surgical Site Infections

In 1991, researchers at the University of Pittsburgh performed an ingenious set of experiments. They implanted sterile prosthetic biomaterials into the peritoneal cavity of otherwise healthy mice and assessed the biomaterials for bacterial presence. ³³ A subset of the biomaterials became colonized with intestinal microbes, despite no intestinal breach during the implantation procedure. These experiments brought forth an idea that intraperitoneal reactive stimuli can lead to bacterial translocation through the intact bowel wall. The researchers further proposed a mechanism involving immune cell phagocytosis of bacteria within the intestinal wall and chemotaxis to distant sites of inflammation. An opinion article published in Nature Reviews Microbiology in 2011 used the same concept to coin the Trojan-horse hypothesis of S. aureus metastatic infections. The authors opined that neutrophils may represent a privileged site for S. aureus in the bloodstream, may offer protection from circulating antibiotics, and may serve as shuttles for the bacteria to reach distant sites and cause infection (metastatic infection). ³⁴ It is well known that neutrophils and other phagocytes continuously sample, survey, and scavenge the environment at mucosal borders. Is it possible that the bacteria responsible for surgical site infections originate at the interface of the intestinal lumen and the mucosal surface of the gut and cause distant infection via the Trojan-horse mechanism? ( Fig. 1 )

Fig. 1.

The Trojan-horse mechanism of metastatic surgical site infections. The neutrophils continuously sample, survey, and scavenge the environment at mucosal borders of the gut. Once the bacterium finds itself in the neutrophil phagosome, it may express a dormant avirulent state, evading host's defenses. Microbe-carrying neutrophil returns to the circulation and then homes to the damaged tissues where it delivers its bacterial payload. The environment of the damaged tissue then cues the bacterium to express a virulent phenotype in a context-dependent manner that may result in an infection.

S. aureus is one of the most successful opportunistic pathogens. This is in part due to its antimicrobial resistance, ability to hide in protective niches, and ability to evade, manipulate, and subvert the host's immune defenses. ^{34 35 36 37} Not by accident, it is also one of the most common culprits of surgical site infections. ³⁸ Furthermore, the gut serves as an underappreciated and understudied reservoir of S. aureus , with carriers accounting for approximately 20% of population. ^{39 40} Among its evasive strategies, S. aureus developed an ability to survive dormant and unharmed within the phagosome of a nutrophil. ^{37 41 42 43} Interkingdom cooperation is thought to be key in allowing for this molecular truce between the microbe and the leukocyte, where the bacterium expresses a dormant avirulent state and is tolerated by host's defenses. ⁴¹ Additionally, neutrophils containing intact S. aureus within their phagosomes have been shown to be capable of releasing their bacterial payload and cause clinically relevant infections. ⁴⁴

The Trojan-horse hypothesis posits that microbe-carrying neutrophils return to the circulation and then home to the damaged tissues where they deliver their bacterial payload ( Fig. 1 ). The environment of the damaged tissue then cues the bacterium to express a virulent phenotype in a context-dependent manner that may result in an infection. ^{3 34 45 46} This phenomenon is well known to surgeons, where more trauma to the tissue correlates with a higher chance of infection. ^{45 47} This hypothesis could provide an explanation for the mechanism of infections that have eluded surgeons to date. Examples of these include infected pancreatic necrosis and prosthesis infections (mesh, breast implants, vascular grafts, joint replacement prosthesis).

Plausibility of the Trojan-Horse Mechanism of Surgical Site Infection

Several contingencies must be satisfied for the Trojan-horse mechanism of wound infection to be considered plausible. First, enteric-derived bacteria must be able to enter and subvert an immune cell. Second, immune cell harboring the pathogen must home to the area of injury and release the bacterium. Third, the wound environment must cue the bacterium to express a virulent phenotype capable of causing an infection.

An original study published by our group was the first to model the Trojan-horse mechanism of surgical site infections. ⁴⁵ Mice were intestinally colonized with a bioluminescent tracer labeled methicillin-resistant Staphylococcus aureus (MRSA) and then subject to a partial hepatectomy with a traumatic insult to the rectus muscle. The intestinal barrier was never breached. A large subset of the animals developed clinically apparent abdominal wall infections caused by bioluminescent strain of MRSA. All the wounds were swabbed prior to closure to rule out intraoperative contamination. Subsets of animals were sacrificed early and their blood cultured to rule out MRSA bacteremia. Reiterative experiments were performed purposefully trying to contaminate abdominal wounds postoperatively with MRSA. Remarkably, there was no evidence of bacteremia or intra- and postoperative contamination.

Another group of researchers from Shanghai, China performed a set of similar experiments modeling prosthetic joint infections. ⁴⁶ Their results mirrored the above findings, but these researchers went a step further. They proved that neutrophils are capable of hosting intestinal-derived bioluminescent MRSA, which can later be reanimated to grow in culture. Furthermore, they showed that intravenous injection of neutrophils containing gut-derived MRSA, can in fact result in prosthetic graft infection. At the same time, MRSA injected directly intravenously is not capable of causing such infections.

Conclusion

It is therefore plausible to conclude that the interplay of the ongoing molecular dialogue between the pathogen, the immune cell, and the traumatized tissue is the main predictor that governs the occurrence, course, and outcome of clinical infection and that the Trojan-horse hypothesis of metastatic surgical site infections may be their mechanistic explanation. Compelling evidence exists that preserving a healthy intestinal microbiome perioperatively is key to surgical recovery and prevention of postoperative infectious complications. Rather than utilizing a broad-based kill strategy of intestinal antisepsis, it is crucial that we develop ways to preserve the microbial diversity and limit the opportunities for colonization by opportunistic pathogens.