Article Text


Novel insight into the role of microbiota in colorectal surgery
  1. Radu Bachmann1,2,
  2. Daniel Leonard1,2,
  3. Nathalie Delzenne3,
  4. Alex Kartheuser1,2,
  5. Patrice D Cani3,4
  1. 1Colorectal Surgery Unit, Cliniques universitaires Saint-Luc, Brussels, Belgium
  2. 2Cliniques des Pathologies Tumorales du Colon et de Rectum (CPTCR), Institut Roi Albert II, Cliniques universitaires Saint Luc, Brussels, Belgium
  3. 3Université catholique de Louvain, Louvain Drug Research Institute, Metabolism and Nutrition research group, Brussels, Belgium
  4. 4WELBIO (Walloon Excellence in Life sciences and BIOtechnology), Brussels, Belgium
  1. Correspondence to Professor Patrice D Cani, Université catholique de Louvain, LDRI, Metabolism and Nutrition research group, Av. E. Mounier, 73 Box B1.73.11, Brussels B-1200, Belgium; Patrice.cani{at} or Professor Alex Kartheuser, Colorectal Surgery Unit, Cliniques universitaires Saint-Luc, 10, avenue Hippocrate, B-1200 Brussels, Belgium alex.kartheuser{at}


Recent literature undeniably supports the idea that the microbiota has a strong influence on the healing process of an intestinal anastomosis. Understanding the mechanisms by which the bacterial community of the gut influences intestinal healing could open the door for new preventive and therapeutic approaches. Among the different mechanisms, data have shown that the production of specific reactive oxygen species (ROS) and the activation of specific formyl peptide receptors (FPRs) regulate intestinal wound healing. Evidence suggests that specific gut microbes such as Lactobacillus spp and Akkermansia muciniphila help to regulate healing processes through both ROS-dependent and FPR-dependent mechanisms. In this review, we will discuss the current knowledge and future perspectives concerning the impact of microbiota on wound healing. We will further review available evidence on whether mechanical bowel preparation and the use of specific antibiotics are beneficial or harmful procedures, an ongoing matter of debate. These practices have a profound effect on the gut microbiota composition at the level of both the mucosal and the luminal compartments. Therefore, a key question remains unanswered: should we continue to prepare the gut before surgical intervention? Current knowledge and data do not clearly support the use of one technique or another to avoid complications such as anastomotic leak. There is an urgent need for appropriate interventions with a deep microbiota analysis to investigate both the surgical technical benefits of a proper anastomosis compared with the potential effect of the gut microbes (beneficial vs harmful) on the processes of wound healing and anastomotic leakage reduction.


Statistics from


The human body exists in complete symbiosis with the gut microbiota, comprising hundreds of trillions of bacteria but also bacteriophagic particles, viruses, fungi and archaea. There is an increasing awareness that the microorganisms that reside within our gut are part of a complex, multidirectional communication network. The knowledge that the bacterial content of the intestine has a significant influence on the well-being of the human organism was postulated even in antiquity by the Egyptians and the Greeks. The first evidence of a treatment that aimed to change the composition of the microbiota was found in Chinese medicine during the fourth century by Ge Hong.1 However, it is worth noting that the recent development of culture-independent methods has revealed much greater diversity within the bacterial communities of the mucosa than were previously identified using traditional culture-based methods.2 Unequivocal experimental studies consisting mainly of comparing the physiology of germ-free versus conventionalised mice (meaning mice that have been treated with the microbiota of (un)healthy mice or humans) have shown that the gut microbiota may help to regulating host health.

Microbial dysbiosis (ie, altered intestinal microbial composition and activity) may contribute to the onset of several disorders, whereas increased microbial diversity and genetic richness are predominantly associated with a beneficial impact.3 ,4 Moreover, dysbiosis also consists of changes in the microbial metabolism, which consequently change the release of metabolites (eg, short-chain fatty acids (SCFA), bile acids, phenolic compounds) or bacterial components (eg, lipopolysaccharides, peptidoglycans) that act on host gut key functions (gut barrier, gut endocrine function, gut immunity).5

Currently, the ‘normal’ gut microbiota cannot be easily defined as it is influenced by pleiotropic factors such as nutrition, medication, GI surgery, smoking, age, gender and the area of living, plus stress, physical activity levels and other environmental factors. It is generally recognised that the intestinal microbiota of healthy adults is characterised by high subject specificity and stability over the years6 and only undergoes temporary minor changes upon fluctuations of environmental or dietary habits. However, there are important variations in the normal intestinal microbiota from one person to another. The indigenous microbiome plays an important role (positive or negative) in the regulation of systemic inflammation.7 This has been particularly studied in the context of nutritional disorders, such as obesity and related metabolic disorders (ie, including diabetes, NASH,8 cachexia or alcohol dependence9 ,10). In this review, we describe a novel context in which focusing on the gut microbiota could present new directions for the management of postsurgery outcomes. Indeed, the microbiota could modulate the restitution of the GI tract functions11 during the postoperative period and could also impact wound healing, even at sites outside the gut.

The philosophy behind the recent advances in clinical practice

In specific situations, such as intestinal surgery, the gut bacteria are still considered a putative source of pathobionts that can lead to further infections. Thus, mechanical bowel preparation (MBP) is still used. This procedure has an obvious, immediate impact on the composition of the gut microbiota, but it also alters the quality and the production of the protective mucous layer, which can permit an eventual bacterial translocation.

Furthermore, the use of bowel preparation agents (eg, antibiotics) and the method of administration also play a key role in the impact of bowel preparation on the colonic mucosa.

Striking experimental data12 obtained in rodents have shown that gut microbes play a crucial role in the vascularisation of the intestinal mucosa and in wound-healing processes. Moreover, gut bacteria produce numerous metabolites that contribute to the regulation of immunity and the renewal of epithelial cells and create an overall environment that limits pathogen development. Experimental and clinical trials show significant changes in the anastomotic tissue-associated microbiota as a result of changes in local environmental conditions. These changes can have either virulence potential or a beneficial effect depending on how the bacterial community composition is affected. As a consequence, the benefits of bowel preparation are currently a matter of debate.

Therefore, in light of recent advances in clinical practice, we will first discuss the major processes of intestinal repair and wound healing and the recently identified mechanisms involved in this interaction. We will next highlight the molecular mechanisms linking gut bacteria and host response in this context. We will present the current knowledge regarding the impact of standard colonic washout, which is commonly performed prior to colorectal surgery or colonoscopy, on the microbiota. We will then review studies analysing the correlation between the gut microbiota and the healing of a colorectal anastomosis. Finally, we will summarise and discuss the novel findings regarding the role of the gut microbiota on the outcome of colorectal surgery.

Physiology and the stages of anastomotic healing in relation to microbial interaction

Although clinicians often consider it an obstacle to proper healing, the microbiota present in the wound may also have positive effects. Certain usual components of the microbiota, especially butyrate-producing bacteria, assist in epithelial repair.13 Butyrate is a key SCFA that is considered a primary energy source for colonocytes and is responsible for cell proliferation. This bacterial metabolite is important for supporting epithelial cell viability and barrier integrity and can downregulate proinflammatory cytokines.14 Mucous layer-attached bacteria are usually considered more host-friendly than luminal bacteria.15 ,16 It is commonly accepted that the intestinal mucosa acts as a physical barrier, defending against bacterial invasion by forming a biofilm monolayer with intercellular contacts that prevents the translocation of pathogens (bacterial products). Other mechanisms, such as gut motility, which prevents prolonged contact between bacteria and potential mucosal defects, chloride secretion, cell–cell contacts, secretion of mucin by goblet cells, cytokines and the production of antimicrobial peptides and defensins, contribute to the protection against bacterial invasion.17 ,18 After a colonic surgical intervention, several of these mechanisms are temporarily disturbed, that is, the gut motility is ineffective because of the usual transitory postoperative ileus.

Li et al19 showed that the protection against mucosal infections provided by the mucous layer is complemented by the endogenous microbiota that fills the microbial niche in which enteric pathogens and opportunistic pathobionts are competing to generate infections. The same study further indicated that some bacterial mutants could not metabolise mucopolysaccharides and therefore act as weak colonisers compared with their parent strain. Therefore, the bacterium's survival or extinction in the host may be determined by its adaptation to life and the microbial community structure in this layer. Finally, the mucous layer is very dynamic and is subject to rapid renewal. Microbes have to adapt to maintain their place in such a system, while at the same time competing with one another for resources to survive within this challenging habitat.

The process of anastomotic repair

The process of tissue repair begins immediately after the anastomosis. Although it is a continuous process, it may be arbitrarily divided into the following chronological phases of tissue repair (figure 1): first, the haemostasis phase, which is marked by the production of a fibrin/fibronectin matrix that temporarily seals and connects the two bowel ends. Second, an inflammatory phase, which comprises the activation of the complementary cascade and the initiation of molecular events that lead to the infiltration of the wound site by neutrophils, which have the main task of preventing infection. Their first critical intervention is phagocytosis to destroy and remove bacteria, foreign particles and damaged tissue. In the later stages of this phase, macrophages appear in the wound and continue the process of phagocytosis. These cells are fundamental for the late stages of the inflammatory response, acting as key regulatory cells and providing an abundant reservoir of potent tissue growth factors. The last cells to enter the wound site in the late inflammatory phase are lymphocytes. Third, a proliferative phase, which is characterised by fibroblast migration, collagen synthesis, angiogenesis and granulation tissue. Fourth is the wound remodelling phase, during which the new epithelium is developed and the final scar tissue is formed (figure 1).

Figure 1

Microbiota influence the stimulation of reactive oxygen species (ROS) and thereby modulate the different anastomotic repair phases. Mechanisms and factors influencing the majors repair phases such as the haemostasis phase, the inflammatory phase, the proliferative phase and the wound remodelling phase. These phases are marked by the production of a fibrin/fibronectin matrix that temporarily seals and connects the two bowel ends, the activation of the complementary cascade and the initiation of molecular events that lead to the infiltration of the wound site by neutrophils, macrophages, lymphocytes fibroblast migration, collagen synthesis, angiogenesis and granulation tissue to developed the final scar tissue. Among the molecular mechanisms, formyl peptide receptors and neutrophilic NADPH oxidase (NOX2) are activated by specific formylated peptides as well as ROS production by certain taxa of the gut microbiota. This respiratory burst is a crucial reaction occurring in the phagocytes, inducing a rapid depletion of microenvironmental oxygen, resulting in a significant proliferation of anaerobic bacteria.

The links between microbes and repair processes

Compelling evidence suggests that the gut microbiota also contribute to epithelial repair and wound healing. For instance, seminal papers have shown that axenic (germ-free) mice are characterised by a lower turnover of the intestinal epithelial cells.20 Moreover, the healing of specific defects in the gut barrier that occurs with inflammatory processes or mechanical injury is impaired in germ-free animals, which have epithelial regenerative responses that are greatly reduced compared with those of conventional animals.20 In addition, there is a strong correlation between different microbial communities in the wound and the ultimate repair of the wound, a point that will be developed further in this review.

The healing mechanism involves the migration of epithelial cells until the gap is closed and the barrier function is re-established. These biological processes are influenced by a variety of factors derived from the GI microenvironment, including host epithelial and lamina propria cells, the dietary and non-dietary proliferation components present in the GI lumen and the microbiota.11 It has been shown that the microbiota may have an important negative or positive impact on wound healing during the inflammatory phase7 and during cellular activation and the promotion of fibrosis.21 In fact, both bacterial competition and the cooperation of bacterial communities can either promote or repress wound healing. For example, it has been demonstrated that Lactobacillus reuteri strain RC-14 can inhibit Staphylococcus aureus infection in a rat surgical-implant model (box 1),22 whereas cooperation that leads to increased virulence involves siderophore-producing bacteria, such as Pseudomonas aeruginosa, that are able to ‘share’ the production of iron-scavenging siderophores with non-siderophore producers.23 Although those findings have been shown in specific models they suggest the existence of competition and or cooperation between microbes and repair, but this remain to be shown in humans.

Box 1

Microbiota and wound healing

Microbiota may hinder wound healing but may also have beneficial effects on wound healing

Akkermansia muciniphila and Lactobacillus spp play a major role in wound healing, stimulating pro-restitutive signalling and increasing migration and proliferation

The administration of oral antibiotics seems to have a beneficial impact on anastomosis healing

Reactive oxygen species and formyl peptide receptors are key molecular actors involved in the microbiota to wound healing process

The microbiota can hinder wound healing, but it might also have a beneficial effect on it. Although different than gut wound healing, surgical skin incision and suture was studied in germ-free and conventional mice. The conventional mice showed greater initial tensile strength of the wound, including higher hydroxyproline concentrations in the surrounding tissue, compared with the germ-free mice.24

In the next part of this review, we detail the molecular mechanisms by which the resident microbiome influences the restitution and wound repair response of the GI epithelium.

Molecular mechanisms linking microbiota signalling and wound healing

Patter recognition receptors, such as toll-like receptors (TLRs), are key partners in the innate immune system. The epithelial cells of the GI tract express multiple TLRs that recognise bacterial pathogen-associated molecular patterns.25 Therefore, TLRs have a crucial role in regulating interactions between specific microbes and the host response.26 Several reports suggest that they also play a major role in intestinal epithelial homeostasis and epithelial injury. However, the microbial ligands recognised by TLRs are not unique to pathogens and are produced by both pathogenic and commensal microorganisms. Rakoff-Nahoum et al26 demonstrated that commensal bacteria are recognised by TLRs under normal steady-state conditions and that acute GI epithelial damage and inflammation were aggravated in mice lacking TLR2, TLR4 or MyD88. Altogether, these observations suggest that the direct presence or interaction with gut microbes or specific receptors of innate immunity stimulate wound restitution and response to injury.

Epithelial restitution requires active cell migration, a process dependent on a constant turnover of focal cell-matrix adhesions. Data suggest that among the other mechanistic processes the production of reactive oxygen species (ROS) and reactive nitrogen species may be the key actors stimulating the beneficial effect of microbes on gut renewal and repair (figure 1).

H2O2 is one of the most relevant ROS in this context. It has been shown to be one of the key ROS produced during different processes, such as wound healing, vascularisation and epithelial restoration. In addition, the appropriate production of ROS helps to stimulate cell attraction, migration and adhesion and immune cell activation, all of which are involved in the regulation of wound healing (box 1) (figure 1).27 As soon as the intestinal tissue is injured, the production of ROS will confer a certain resistance to wound infection and will stimulate smooth muscle and epithelial cell proliferation to participate in wound healing and tissue regeneration (figure 1).28

Although the production and impact of mitochondrial ROS has been clearly shown, the mechanisms involved in the stimulation of ROS production are less understood. The role of specific microbial-derived molecules has been suggested. For example, formyl peptide receptors (FPRs) are activated by specific formylated peptides produced by certain taxa of the gut microbiota. These FPRs are expressed in intestinal epithelial cells, which are in direct contact with the gut luminal contents. Interestingly, specific microbial formyl peptides have been shown to stimulate FPRs and to induce ROS and the chemotaxis of immune cells (figure 1).29 ,30

A recent study demonstrated that enteric resident bacteria can potentiate epithelial restitution via the generation of ROS in epithelial cells, which in turn mediates the inactivation of focal adhesion kinase phosphatases.31 Microbiota–host epithelial cell interactions can also modulate β-catenin signalling, a key component in regulating epithelial cell proliferation.32 Several studies have demonstrated the possibility that certain normal members of the bacterial microbiome (eg, Lactobacillus species and butyrate-producing bacteria, which have a very potent activity) stimulate the generation of ROS in intestinal epithelial cells through FPR-mediated signalling that modulates the extracellular signal-regulated kinase pathway, which in turn regulates epithelial cell migration and epithelium restitution (figure 1).33–35

Another recent study by Alam et al36 elegantly shows that both mechanisms of action may coexist to induce intestinal repair upon injury (box 1). They found that Akkermansia muciniphila administration plays a major role in wound healing by creating a specific environment and local environmental changes. This state favours a transient consortium of microbes that can stimulate restitutive signalling pathways to increase both the migration and proliferation of the epithelial cells. The study demonstrated that among the molecular mechanisms, FPR1 and neutrophilic NADPH oxidase (NOX2) were required to induce the rapid depletion of local oxygen that resulted in the enrichment of anaerobic bacteria (figure 3). Using mice with a specific deletion of Fpr1 or Nox1 demonstrated that A. muciniphila modulate redox-dependent cellular signalling via these two specific actors (box 1). These data support previous findings that Akkermansia muciniphila reinforce the gut barrier function via mechanisms involving the production of mucous and antimicrobial factors in obesity and diabetes.16 ,37

Therefore, this example strongly highlights the need to preserve or restore the presence of specific bacteria at the mucosal level to induce proper complex homeostatic signalling and wound-healing processes (box 1).

The impact of preoperative mechanical bowel preparation with or without antibiotics on the microbiome

Mechanical bowel preparation: subject of debate

For more than half a century, the use of MBP before elective colorectal procedures has been a subject of debate. It has long been considered surgical ‘dogma’ and is still being used by a vast number of colorectal surgeons. In addition to the traditional effect of the so-called ‘bowel cleansing’, other technical factors point to the advantages of using MBP prior to surgical intervention. These include the fact that manipulating a colon full of stool is difficult and makes the palpation of small tumours impossible. In addition, the presence of excessive stool in the rectosigmoid prevents the performance of transluminal stapled anastomosis.

Since the emergence of the concept of enhanced recovery after surgery in the mid-1990s by Kehlet, the tendency to perform colonic surgery on the non-prepared colon has increased significantly without affecting the surgical outcome. Currently, many expert centres tend to operate on unprepared colons. Still, for rectal surgery requiring a low coloanal anastomosis, which is usually protected with a temporary ileostomy, the washout remains the rule.

Recently, the debate was reopened following the results of three large retrospective analyses from the USA published in 2015 that showed that preoperative MBP combined with oral antibiotics significantly improved the outcomes of colorectal resection.38–40 This combination reduces by nearly half the rate of complications after colorectal surgery, including surgical site infection and anastomotic leak and ileus (figure 2).39 These findings conflict somewhat with the results of the French prospective analysis GRECCAR 3 (a single-blinded, multicentre, randomised trial, which analysed the postoperative results in patients undergoing sphincter-saving rectal resection for cancer without preoperative MBP) completed several years ago.41 For this reason, a new large international prospective multicentre randomised four-arm trial is currently being designed and begin soon. This study would randomise patients into MBP versus no MBP groups with or without preoperative oral antibiotics.42

Figure 2

Microbiota shifting as a consequence of the injured mucosa accelerating the healing mechanism. Based on the findings of Alam et al,36 this graph shows the different changes that occur in the microbiota according to the intestinal injury. It shows the important role of the wound-mucosa-associated microbiota in initiating the early healing cascade by stimulating pro-restitutive signalling and increasing cell migration and proliferation.

Although a great number of randomised trials show the clear advantages of preparation with oral antibiotics, this practice has not been adopted in routine clinical practice; the actual standard is still the use of intravenous antibiotics just before the procedure begins.

Bowel-cleansing products

The efficacy of bowel-cleansing products and their safety to the patient have been studied extensively; however, the effects of lavage on the colonic microbiota have been poorly characterised to date.43 Polyethylene glycol (PEG) solutions are preferable to saline ones because they are easier to administer, more comfortable and acceptable to the patient and lead to fewer complications. However, the PEG solutions for MBP result in moderate-to-severe loss of superficial mucus, which could also affect the composition of colonic microbiota (box 2). The mechanisms of microbiota alterations are related to several factors. The ingested PEG electrolyte solution and the increased volume of fluids flush out the luminal bacteria, introduce oxygen into the normally anaerobic colonic ecosystem and reduce the nutrition supply for the intestinal bacteria (figure 3).44 All of these rapid changes may affect the microbial ecosystem.

Box 2

Microbiota and bowel preparation

Single-dose mechanical bowel preparation (MBP) (2 L) has a more severe effect on the microbiota than a double split dose (two separate 1 L dosages) does

Polyethylene glycol solutions for MBP result in an important loss of superficial mucosal mucus, which affects the microbial ecosystem

Almost restitutio ad integrum of the microbiota occurs a few weeks after MBP in healthy patients

A 24-hour clear liquid diet does not change the microbiota in the short term

Figure 3

Risk factors for anastomotic leakage after left-sided colorectal resection. Summary of the different factors influencing anastomotic healing and the gut microbiota composition. Mechanisms and taxa that have been associated with both positive (+, green arrow) and negative (−, red arrow) process of wound healing. Pathogenic bacteria overexploit microbiota nutrition sources and regulatory signals to promote their own growth and virulence. Bacterial proteases, that is, gelatinase (GeIE), disrupt the intestinal integrity and induce increasing inflammation. FPR, formyl peptide receptor; ROS, reactive oxygen species.

Impact of bowel preparation on microbiota

To date, few studies have investigated and described the impact that bowel preparation with or without oral antibiotics conducted prior to colorectal surgery may have on the mucosa-associated and luminal colonic microbiota (table 1). Even fewer studies have reviewed the direct impact of the gut microbiome on the healing of a colorectal anastomosis (summarised in table 1).

Table 1

Studies conducted to verify the effect of bowel preparation on the microbiota composition

Mai et al45 were the first to examine the effect of bowel preparation and colonoscopy on the intestinal microbiota composition. Bucher et al46 analysed the differences in the stool microbiota in patients who did or did not undergo preoperative MBP after left-sided colorectal surgery for benign or malignant disease. Harrel et al47 split the subjects into three groups to compare the impact on the microbiota of MBP or a liquid diet for 24 hours with no bowel preparation (box 2). O'Brien et al48 analysed the microbiota in stool samples collected 1 month and 1 week before and 1 week, 1 month and 3 months after MBP using denaturing gradient gel electrophoresis (DGGE) and high-throughput sequencing (HTS) methods. Jalanka et al49 was the first (and at present, the only) study to examine the effect of two cleansing regimens on faecal microbiota composition. This was also the first study to analyse the different effects associated with two different methods for administering the PEG solution (Moviprep): as separate 1 L doses or as a single 2 L dose (box 2). Drago et al50 compared the microbiota composition immediately and 1 month after MBP; contrary to most other studies, they found that MBP has long-lasting effects on the gut microbiota composition and homeostasis, particularly in terms of a decrease in the abundance of Lactobacillaceae, a population of protective bacteria (figure 3). Shobar et al51 conducted one of the most thorough studies, which simultaneously analysed the composition of the microbiota at the levels of the mucosal and luminal compartments, differentiating between healthy subjects and patients with IBD and using a high-resolution technique.

As summarised in table 1, all of the studies clearly showed a massive impact on the diversity of the gut microbiota composition in the colonoscopy samples taken after MBP. Moreover, MBP does not decontaminate the GI tract as the amount of bacteria remains almost the same as when MBP is not performed. Finally, the most deleterious effect of MBP is that it upsets the microbiota enough to change the gut microbiota composition, decreasing the diversity and allowing opportunistic pathogens to thrive.52 In contrast, in healthy subjects, an almost complete restitution of the microbiota composition was noted just a few weeks after MBP (box 2).

In general, it is accepted that the average microbiota composition in an ordinary normal subject is composed of Firmicutes, Bacteroidetes, Actinobacteria, Verrucomicrobia and Proteobacteria.53

All of the studies describing the modification of the gut microbiota immediately after the MBP showed important short-term distortions in the composition and an important reduction in diversity without being able to determine one unique general pattern. Most often, they indicated an increased level of Proteobacteria and Bacteroidetes, a significant reduction in Lactobacillaceae, an increase in Enterobacteriaceae abundance at the family level and a drastic change in the ratio of Gram-positive to Gram-negative species. Similar changes were observed when studying the microbiota in diarrhoea diseases in general (figure 3).54

Purging leads to the introduction of oxygen into the normally anaerobic colonic ecosystem, resulting in an increased amount of Proteobacteria. In addition, the increase in pH might reflect the loss of specific bacterial metabolites, such as SCFA, and thereby could allow Proteobacteria to flourish.49

Impact of antibiotics use on intestinal anastomotic healing

Without any doubt, the discovery and introduction of antibiotics was a huge step forward in the success of colorectal surgery. The first experimental study to demonstrate the influence of the microbiota on intestinal anastomotic healing was performed over 70 years ago, when Cohn and Rives55 managed to prevent anastomotic leak in a dog model by instilling intraluminal antibiotics (ie, achromycin) at the site of a bowel anastomosis that was rendered ischaemic by ligation of the mesenteric vessels. Interesting findings have also been found with regard to the impact that the use of antibiotics with MBP may have on the microbiota. Not all commonly used oral antibiotics appear to have positive effects.56 Antibiotics that induce bacterial translocation have been found to lead to increased inflammation in response to coincident epithelial injury.57 Additionally, it is well established that excessive uncontrolled inflammation results in a variety of pathological conditions, including, at the extreme, septic shock.

Conversely to MBP, the administration of antibiotics with broader coverage than necessary provokes an immediate and long-lasting effect on microbial diversity (figure 4), often leading to definitive incomplete recovery and increased antibiotic resistance.58 A significant reduction in Bacteroidetes and a concurrent increase in Firmicutes was found after the administration of broad spectrum antibiotics.59 Probiotic microorganisms, prebiotic substrates that enrich certain bacterial populations or symbiotic combinations of both of these will most likely comprise an alternative approach to non-selective antibiotic therapy in the near future.

Figure 4

The Impact of mechanical bowel preparation (MBP) and antibiotics on microbiota diversity and total bacterial load. The curves that are based on the very few, quite heterogenic studies agree on a significant decrease of the total bacterial load as well as of their diversity with a fast recovery of the absolute bacteria number and a regeneration of the diversity almost ad integrum within 2 weeks after the MBP. Antibiotic preparation leads to an almost complete drop of the total bacterial load with a similar fast recovery of the total number of bacteria but with a very slow, incomplete regeneration of the diversity, even months after and most probably with permanent consequences.

The same considerations must also to be taken into account with regard to colon-cleansing procedures performed independently from a medical intervention. Colon detox is increasingly believed to promote a healthy lifestyle, and a whole industry has developed on this premise. However, very limited scientific research is available regarding the impact of such procedures. It is very likely that they lead to imbalances in the microbiota with negative health effects.

At present, very little is known about how to re-establish the microbiota after sudden iatrogenic modification. The question of whether the humans possess a distinct source or niches for re-establishing normal microbiota has not yet been answered. The appendix has been suggested to represent a ‘safe house’ for gut microbes.60 Gut-associated lymphatic tissue is present in the appendix, suggesting that it has important immune functions for regulating microbiota composition. In contrast, inflammation in the appendix has been associated with local invasion by Fusobacterium nucleatum.61 However, it is not very plausible to conceive that all patients without an appendix will develop serious problems with recovery after a bowel preparation. It is clear that the real drivers of partial recovery of the initial ecosystem are the restoring forces within a dynamic system rather than inertia.

Influence on and changes in the gut microbiota after intestinal anastomotic injury and the impact of microbes on intestinal anastomotic healing

The most frequent postoperative surgical complications after colorectal resections (aside from bleeding) are the surgical site infection, anastomotic leakage (AL), and intra-abdominal abscess, and the microbiota has a significant role in triggering each of these complications (figure 2). Although AL is usually considered a result of poor surgical technique, there is increasing evidence that the gut bacteria play a key role in its development, even if the mechanisms are not entirely understood.

Other risk factors have been already well reviewed, and different preventive strategies have been proposed (figure 3).

Kirchhoff et al62 classified them according to factors that can be influenced (eg, obesity, nutritional status, preoperative bowel cleaning, experience of the surgeon, preoperative anaemia, intraoperative blood loss, drainage and type of surgical approach, including type of anastomosis) and risk factors that cannot be influenced (eg, age, gender, prior radio-chemotherapy, comorbidities) (figure 3). Strikingly, many of the risk factors that may be influenced are deeply impacted by the microbiota, while at the same time they also influence the composition of the microbiota. Novel techniques to improve some of these factors in a reasonably short period have been successfully tested in rodent studies and ongoing clinical studies. This is the case of Akkermansia muciniphila, for example (box 1). Everard et al16 ,63 have shown that this mucin-degrading bacteria, which lives in the mucous layer, restore gut barrier function and reduce bodyweight gain, fat mass development and low-grade inflammation by acting on both the mucous layer (ie, increasing it) and the production of specific antimicrobial proteins.

The incidence of anastomotic leak after colorectal resection varies from 3% to 20%, depending mostly on the anastomosis localisation; the highest leakage rate of up to 24%64 was observed in low colorectal or coloanal anastomosis (figure 2). Anastomotic leak is associated with an increased risk of morbidity, short-term mortality and permanent ostomy.65 Furthermore, in patients with cancer, anastomotic leak has been associated with increased local recurrence and poor survival after colorectal cancer surgery.66 In recent decades, the number of colorectal anastomotic leaks has remained constant. This important observation is a fact, despite significant technical surgical advances (for instance, the use of novel surgical devices, different staplers, laparoscopic-guided or robotics-guided procedures or even perioperative arterial vascular control by indocyanine green-enhanced (ICG) fluorescence-guided colorectal resection).

The benefits of using fluorescence angiography to control vascularisation of the anastomosis after intestinal resection were clearly demonstrated in the PILLAR II study.67 By using the ICG fluorescence, the frequency of AL dropped to <2%. More recent published data confirm this finding. According to the work of Cohn and Rives55 and Reinhardt et al,12 we can conclude that the microbiota-induced vascular remodelling is of huge importance for the adequate healing of an anastomosis.

A potential way to reduce the impact of perioperative microbiota disturbance would be the development of an enhanced recovery protocol (ERP) that was developed to reduce physiological stress and postoperative organ dysfunction after surgery. This protocol, also known as ‘Fast-Track’, gained significant adherence of many surgical teams in the last 20 years. It is based on preoperative, intraoperative and postoperative measures, for example, avoidance of MBP, use of minimal invasive surgery, avoidance of perioperative fasting, standardised anaesthetic and analgesic regimens and early mobilisation.

ERP have been shown to significantly improve the morbidity and mortality. Although so far data are lacking, it can be assumed that this protocol, by promoting a safe postoperative recovery with early return of bowel function, might also reduce negative influence on the loss of microbiota diversity and specificity.

At present, only a few studies have examined the impact of an intestinal anastomotic injury on the microbiota composition. Therefore, we believe that more scientific efforts should be made to determine whether a specific dysbiosis of the microbiota composition resulting from surgery trauma or an eventual bowel preparation might favour the development of AL.

In clinical routine, the effect of an intestinal anastomotic injury on the microbiota is likely not caused by surgical stress alone but is the result of the perioperative management as a whole. Thus, there are many potential sources of factors that may influence the microbiota. We may cite, for instance, MBP, the use of antibiotics, modulation (or lack thereof) of the preoperative and postoperative diet, the use of postoperative analgesia and opioids, and the administration of proton pump inhibitors or any gastric acid reducers (figure 3). Finally, transanal drainage tube placement or the use of vasoactive medicines are also specific confounding factors (figure 3). In addition, the initial composition of the gut microbiota might already be dysbiotic and therefore not ‘normal’ because of the specific disease to be treated. However, there is no doubt, for example, that the microbiota in patients with IBD suffers from significant changes in its bacterial composition.51 Even though data are scarce for the moment, we can presume that other diseases that will finally point towards a colorectal surgery such as repeated diverticulitis or colorectal cancer are also characterised by changes in the microbial diversity also leading to a chronic inflammatory status, as recently shown by Barbara et al.68 All these changes threaten the microbiota diversity and allow pathogenic bacteria that contribute to postsurgical complications to thrive. Hence, pushing the research to try to find clinical viable solution to ameliorate the preoperative microbiota should be a priority.

A number of studies have described bacterial wound communities, especially in chronic open wounds, and they reported the following observations:69–73

  • The number and proportion of bacterial species can range greatly among individual wounds.

  • Bacterial diversity determined with culture-based methods is significantly lower than that determined with new culture-independent methods, such as PCR amplification of the 16S rRNA gene.

  • The microbiota can differ among different wounds, while the bacterial communities at different sites within an individual wound are significantly more similar to one another than to those of different wounds.69

  • The reliability of both culture-based and non-culture-based analysis depends heavily on the sampling method used. For example, certain sampling techniques will not detect anaerobic bacteria, which are common in chronic wounds.74 Therefore, when studying the human microbiota, important controls need to be assured to guarantee that the chosen sampling techniques are as unbiased and comprehensive as possible.

Shogan et al75 analysed the changes in luminal versus tissue-associated microbiota at anastomotic sites created in the colons of a rodent model using 16S rRNA gene amplicon sequencing of samples collected on the day of surgery and the sixth day following surgery. It is worth noting that they did not use antibiotics either preoperatively or postoperatively, and they did not perform bowel preparation. The results indicated that anastomotic injury induced significant changes in the anastomotic tissue-associated microbiota with minimal differences in the luminal microbiota (stool).

A 500-fold and 200-fold increase in the relative abundance of Enterococcus and Escherichia/Shigella, respectively, were identified as the most significant differences. Functional profiling predicted the prevalence of bacterial virulence-associated pathways in postanastomotic tissues, including the production of haemolysin,76 ,77 cytolethal toxins,78 cytotoxic necrotising factor,79 bacterial attachment to mammalian cells via AIDA-I adhesin-like protein, cytotoxicity to eukaryotic cells as a result of cytotoxic necrotising factor,79 increased motility (outer membrane usher protein, fimbrial-like protein, type 1 fimbrial protein, fimbrial chaperone protein)80 and invasion of epithelial cells that can be attributed to Enterobacteriaceae, particularly the Escherichia genera.81 Proteolytic cleavage by coccolysin, also known as gelatinase, can be attributed to Enterococcus.76 ,82 Importantly, the gelatinase GelE of Enterococcus faecalis has been shown to degrade collagen, a critical component of anastomotic tissue healing (figure 3).83 Although little is known about E. faecalis and its possible influence on leakage, this bacterium is highly prevalent in leaking anastomotic tissues.84 ,85 Enterococcus has a high adherence affinity to extracellular cellular matrix proteins such as fibronectin, laminin and various different types of collagens,86–88 including collagen IV.89

In a recent clinical study, Komen et al85 proposed a feasible fast-screening tool for the early detection of colorectal AL based on the presence of Escherichia coli and E. faecalis in drained fluid using RT-PCR. They showed that the increased presence of E. faecalis in drained fluid may be a sensitive screening tool for the early detection of colorectal AL, while negative test results virtually rule out the presence of an anastomotic leak.

As previously mentioned, age is an important risk factor for colorectal surgery. Interestingly, it has been shown that the microbiota composition at the extremes of life, in infants and the elderly, is different from that of adults.90 The core microbiota of elderly subjects is distinct from that previously established for younger adults, with a greater proportion of Bacteroides spp and distinct abundance patterns of Clostridium groups,91 which may contribute to the known increase in the AL risk among the elderly.

A critical loss of cytoprotective microbiota in anastomotic tissues may further facilitate colonisation and invasion by disease-associated microbiota, again with the potential to complicate healing (figure 3). Levels of the commonly cytoprotective bacteria belonging to the family of Lactobacillaceae have also been shown to be reduced in anastomotic tissues.92–94 The lack of such taxa may play a role in the development of leaks when levels are critically decreased and additional physiological stress occurs. As described earlier, these taxa contribute to wound healing and repair via mechanisms involving ROS production and specific formylated peptides acting on FPRs (figure 1).36 ,95 Nevertheless, in the study by Okada et al,94 conventional rats had a significantly higher anastomotic bursting pressure in the ileum compared with both rats whose colons had been mono-contaminated with Lactobacillus acidophilus and with germ-free rats, suggesting the importance of a whole bacterial community and the interactions within the community.

In a 2002 rodent study, Stappenbek et al demonstrated the key role of the microbiota in (re)constructing the mesenchymal microvascular network through the intermediary of Paneth cells, a key cellular component of the innate immune system. They showed that Bacteroides thetaiotaomicron stimulates angiogenesis by doubling the capillary network in the small intestine via Paneth cells.96

In 1997, Schardey et al97 suggested that P. aeruginosa might play a causative role in anastomotic leak and performed the first randomised, prospective, double-blind and placebo-controlled trial with antibiotics, confirming that microbes played a role in human anastomotic leak.98 In this trial, patients received either placebo or decontamination with oral antibiotics just before and 1 week after the surgical intervention. All of the patients received perioperative intravenous antibiotic prophylaxis. The authors concluded that this decontamination during anastomotic wound healing is safe and effective for preventing esophagojejunal AL after total gastrectomy.

In another study, a novel model of AL was developed in which rats that had been preradiated (as is usually necessary in rectal surgery) underwent distal colon resection and then were intestinally inoculated with P. aeruginosa, a common coloniser of the radiated intestine.99 A significantly higher incidence of AL was found when the rats were colonised by P. aeruginosa compared with radiation alone. Moreover, the study showed that the inoculated strain mutated in cases of AL, showing altered pyocyanin production and enhanced collagenase activity, high swarming motility and destructive activity towards cultured intestinal epithelial cells (ie, apoptosis, barrier function, cytolysis). In contrast with the radiated rats, the non-radiated rats maintained a normal microbiota and an intact mucous layer, which may shield against the virulent effects of the mutated phenotype. This work demonstrates that the in vivo transformation of microbial pathogens into a tissue-destroying phenotype may have important implications in the pathogenesis of anastomotic leak.

Finally, the use of the novel virulence-suppressing compound PEG/Pi (high-molecular-weight PEG) prevented P. aeruginosa from transforming to the tissue-destructive phenotype and prevented anastomotic leak in rats.100 Wu et al100 demonstrated that PEG/Pi has the potential to function as a surrogate mucin within the intestinal tract of a stressed host by inhibiting key interactive events between colonising microbes and their epithelial cell targets.

Van Praagh et al conducted a unique clinical study that investigated the microbiota composition in the area of anastomotic fistulas. The study was conducted with patients included in the C-seal trial. The C-seal is a biodegradable sheath that protects the stapled colorectal anastomosis from leakage.101 The researchers compared the mucosa-associated microbiota collected at the anastomosis level during the surgical intervention from eight patients who developed AL requiring reintervention with that of matched individuals from the same trial who had a favourable evolution.102 They found a correlation between the bacterial family Lachnospiraceae, low microbial diversity and AL, possibly in association with the body mass index. The relative abundance of the Lachnospiraceae family (phylum Firmicutes) may be explained by the higher abundance of mucin-degrading Ruminococci (phylum Firmicutes) within that family in AL cases, which is also observed in IBD (figure 3).

The studies discussed above suggest that commensal microbes are influenced by intestinal resection and anastomosis, along with other aspects of perioperative management. This results in the bloom of specific species that are normally present in the normal commensal microbiota (ie, Enterococcus and Pseudomonas) but that display a more aggressive, tissue-destroying phenotype. This could contribute to the pathogenesis of anastomotic breakdown and clinical leakage.103

Conclusions, perspective and recommendations

To sum up, to date and in the light of all the available evidence so far, there is no clear answer regarding whether gut preparation prior to surgical intervention should be performed.

Recent research shows that the gut microbiota acts as an ecosystem with a very important role in the well-being of the entire organism. It has also been shown that the microbiota has a great influence on the healing process of an intestinal anastomosis.

Compelling research proves that standard bowel preparation with or without the administration of antibiotics disturbs the diversity and the fragile balance of the normal and putatively healthy microbiota. This disturbance is associated with an immediate impact on the gut barrier function and changes in intestinal innate immunity. The imbalances caused by the preparation methods may have an impact on the healing process of the anastomosis, thereby calling their supposed benefit into question. At the same time, there are no proven alternatives to the use of antibiotics, even though its well-proven positive effects on healing go hand-in-hand with the disturbance of the microbiota.

Research should strive for targeted approaches that would ensure the preservation of the microbiota in a way that can selectively inhibit the pathobionts and simultaneously preserve the good bacteria. Such a targeted treatment prior to surgery would create the optimal ecosystem and minimise the risk of complications.

We believe that the findings available to date are still excessively descriptive, and we recommend a better focus on the type of preparation and on descriptions of the microbiota upon the onset of diseases. Therefore, a deeper analysis is clearly needed to fully understand the specific interactions between the bacteria inhabiting the gut and their role in the complex wound-healing processes. However, it is now clear that more attention is being paid and will be likely paid in the near future to both the beneficial and pathological roles of the gut microbes in the healing of anastomotic wounds. Developing tools and adequate protocols evaluating the clinical relevance of gut microbiota modulation to positively influence the wound healing are needed in the future. However, there is an urgent need for future research that should focus not only on the composition but also on the specific management of the microbiota. Finally, targeted use of new generation of probiotics, prebiotics and synbiotics represents a interesting research area with a huge potential.


View Abstract


  • Twitter Follow Patrice Cani @MicrObesity

  • Contributors RD, DL and PDC conceived and drew the figures. RB, DL, NMD, AK and PDC have all contributed to the design and the writing of the review.

  • Funding This work was supported by the FRFS-WELBIO under grant WELBIO-CR-2012S-02R. This work was supported in part by the Funds Baillet Latour (Grant for Medical Research 2015), a FIRST Spin-Off grant (convention 1410053). PDC is a research associate at FRS-FNRS (Fonds de la Recherche Scientifique), Belgium, and the authors have no competing interest to declare. The recipient of grants from FNRS (convention J.0084.15, convention 3.4579.11), PDR (Projet de Recherche, convention: T.0138.14) and ARC (Action de Recherche Concertée—Communauté française de Belgique convention: 12/17-047). PDC is a recipient of the POC ERC grant 2016 (European Research Council, Microbes4U_713547) and the ERC Starting Grant 2013 (starting grant 336452-ENIGMO). ND is recipient of grants from EU (European Union's Seventh Framework Program (MYNEWGUT project agreement 613979), from Wallonia-Belgium (Food4Gut Excellence Project; NUTRIGUTIOR project, convention 6918), from Flanders Region – Belgium (Agency for Innovation by Science and Technology, Branding project).

  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.

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.