Preparing the bowel for surgery in the era of microbiome sciences

Microbiome Sciences-rethinking everything

The term “microbiome” encompasses all microbial life in a given ecological niche including bacteria, viruses, archea, protists and all the metabolites they produce while existing as a complex community ecosystem constantly exposed to the particulars of their living environmental context¹. Thus there is a skin, surface, hospital, water, sewage, gut, lung, urine “microbiome” that is shaped by their “exposome²,” i.e., all the environmental cues present in a given site to which they are exposed. The term “microbiome” includes all normal commensal functioning microbiota as well as the less abundant, but no less important, pathobiota. Like many ecosystems, the good members outnumber the bad members, yet the expression of “virulence” or harmfulness is primarily generated by the low abundance “pathobiota,” a term invented to indicate those “biota” that cause pathology. In many circumstances it is the normal microbiota that keep the pathobiota at bay providing “colonization resistance.” Yet a comprehensive readout of the entire system’s community structure, membership and function is now possible via genetic sequencing (metagenomics) or metabolite production (metabolomics) that has the potential to not only identify pathogenic features of the system, but also its potential benefits. For example, when sewage is tested for the COVID virus³, or the spike protein via PCR, a pathogenic feature of a community can be interrogated and its prevalence among the microbiota determined. Thus, looking for bad actors among a complex set of microbes, genes or biota is not only possible, but also practical. Such technology is forcing a rethinking of how we approach infection-related diseases. For example invoking the Koch’s postulates to identify the presence of an infection is no longer valid: the mere presence of a pathogen is no longer as important as the genes it expresses in vivo and if the gene products locally secreted are those that are actually causing the disease. This has permitted emergence of the “molecular Koch’s postulates⁴” to indicate that not only is the pathogen present, but it is expressing the virulence factors required to drive disease manifestation. In fact, the term “virulence,” harmfulness, clinical infection is now considered neither a property of the pathogen, nor that of the host, but rather a property of their interaction⁵. Much like two living bodies co-existing in a complex environment, it is the “chemistry” between them whose final output yields the harm. Both physico-chemical “cues” in the local environmental context as well as the host-related factors, can influence the interaction of microbes with their host and thus influence their pathologic or beneficial output. (see figure 1). Therefore the extent to which a microbe or pathogen becomes “in vivo expressed” by the local microenvironmental context and host factors (i.e. locally secreted “cues”), the interaction between microbe and host that this produces, and how the overall molecular bidirectional dialogue between host and pathogen plays out, will dictate whether “virulence,” or harmfulness, clinical infection or pathology develops⁶. In practical terms, the mechanisms that govern whether or not an infection develops from a pathogenic exposure remains poorly understood and defined as, in most cases, only the “infected” host is comprehensively examined. For example, a recent outbreak of Listeria occurred tied to processed meat⁷. Yet the number of individuals exposed who do NOT develop a clinical infection (i.e., asymptomatic carriers) is unknown, as they largely remained asymptomatic and therefore the extent and timing of their exposure remains unstudied. Similarly with the COVID virus, using PCR based probes to test for its presence in sewage can indicate the overall prevalence of the virus in expelled stool, but cannot estimate who is resistant, who is asymptomatic, who is a carrier and who can transmit the virus. Yet to determine what factors are involved when a patient develop symptoms, while another does not, requires a massive molecular analysis of exposed, unexposed, symptomatic and asymptomatic individuals tracked over time. Such an analysis rarely if ever exists and requires animal testing to invoke the molecular Koch’s postulates of disease. Neither epidemiologists, microbial scientists, infectious disease specialist or others can determine why one exposed person develops a clinically symptomatic infection, another becomes a carrier, another asymptomatic and another does not harbor or contract the infectious agent. Some type of molecular alignment of the “chemistry” between the pathogen, the native microbiome that provides colonization resistance and the context-dependent local microenvironmental “cues” tracked over time is necessary in order to understand why a clinical infection develop in one host and not another. Molecularly interrogating this “interactome” in real time, at each specific tissue site, i.e., the skin, mucous membranes, immune cells, etc., is not only a daunting task, but one that is likely impossible. For this reasons, clinicians prefer to broadly apply antibiotics, aseptics, vaccines and disinfectants, with the hope to eliminate both good and bad microbes and thus reduce infection-related complications. Yet in the era of personalized medicine and emerging antibiotic resistant organisms, the sustainability of this practice has come into question⁸. Like all approaches to eradicate the bad actors while preserving the beneficial ones so as to avoid the collateral damage to the innocents, the approach often ends up messy, empirical and untargeted.

Fig 1 :-.

Physico-chemical “cues” in the local environmental context as well as the host-related factors, can influence the interaction of microbes with their host and thus influence their pathologic or beneficial output

The scientific rationale behind a bowel prep and the path forward

The history of preparing the bowel for surgery is a long one and unfortunately, today there are many opponents who do not believe it is necessary prior to intestinal surgery⁹. Initially in the early 1900’s the mortality from colon surgery approached 25% and was, most often due to an infection-related complications, i.e., sepsis, , anastomotic leak, wound infection and multiple other life-threatening infections. Given that during colon surgery, the intestinal track is opened and ostensibly the local tissues can become contaminated, the rationale behind purgative cleansing, oral non-absorbable antibiotics and systemic antibiotics seemed logical. The history behind this practice has been extensively reviewed by others^10,11. Yet today, in the era of enhanced recovery programs including early feeding of regular food, avoidance of opioids, asepsis techniques, clean and rotated wound instrumentation and the routine use of intravenous antibiotics, many have opined that the bowel prep, as traditionally applied, is unnecessary¹¹. Many have argued that for a minimally invasive (laparoscopic/robotic) right hemicolectomy in a low risk patient, in which it is anticipated that there will be minimal spillage and a wound protector is used, that a full bowel prep including purgative cleansing and oral antibiotics is unnecessary. This is especially relevant now that the rare occurrence of an anastomotic leak or a contained intrabdominal abscess that can be drained an interventional radiologist has become not only rare, but a non-operative intervention. It is even argued that for a simple left hemicolectomy, a full bowel prep is unnecessary. However for colorectal surgery in which a low colorectal anastomosis may be necessary, many argue that a bowel prep is necessary give that many such patients will require invasive biopsy, prior exposure to antibiotics, neoadjuvant chemoradiation, etc. Yet in light of progress in the field of surgery, one might argue that the pros and cons of imposing a bowel prep on all patients has reached equipoise. Given the rapid advance of technology (robotic surgery, wound protectors, advanced stapling, interventional radiology, serial endoscopy with augmented reality to detect ischemia, etc.), results continue to be mixed. Therefore, patient participation in their care, individual practicioners’ practice standards, hospital-based protocols, etc., need to be considered.

Evidence that a bowel prep is necessary/unnecessary prior to GI surgery.

While there much evidence exists to suggest that a bowel prep (purgative cleansing + oral non-absorbable antibiotics) can reduce the incidence of anastomotic leak, wound infections, and deep organ space infection, when examined in the aggregate, these results suffer from a classic “between comparison fallacy” as has been previously outlined¹². For example, when a randomized prospective trial (RCT) comparing bowel prep versus no bowel prep is reported, most patients receive an intravenous dose of a broad spectrum antibiotic. In addition, when anastomotic leaks are reported, they are underreported given that they are only counted when patients become symptomatic triggering imaging and a potential intervention¹³. In fact, much has been written about the underreporting of anastomotic leak given that unless a “procedure” is required (surgery, interventional radiology percutaneous drainage, nasogastric tube + antibiotics+ repeat imaging, etc.), most leaks go unnoticed and therefore unreported. Also, other SSIs such as wound infections, are often judged to be present using the CDC (Center for Disease Control) criteria¹⁴. This represents a major problem given that, using this criteria, an SSI (wound infection) is deemed to be present if a healthcare worker considers it so by by visual inspection alone, without culture and without any other verification. Thus wound infection, most of which are superficial and treated without culture or opening, are counted as such and deemed to be “decreased” by an intervention which is a very subjective assessment. In this scenario many SSIs (Surgical Site Infections- wound infections, anastomotic leaks, organ space infections) are miscounted and thus inaccurately assessed. With such confounders, when a one-size-fits-all antibiotic approach is tested in a “between-group” comparison study, assuming again that each patient’s “pathobiome” (the pathobiota that colonize a given patient) is the same, between-group comparisons become problematic. This latter concept is complicated by the recent observation that now, in the post-antibiotic era, multi-drug antibiotic resistant bacteria predominate in SSIs (if they are ever cultured); this is especially observed in patients undergoing colon surgery. This observation, coupled with a recent study in Science, Translational Medicine, demonstrated that 60% of patients undergoing instrumented spinal surgery, develop an SSI with a pathogen that is resistant to the antibiotics chosen for prophylaxis. Perhaps it is time to personalize our use of prophylactic antibiotics in elective surgery given the an increasingly complex and elderly patient population is emerging.

With all the above problems, when an RCT is performed to determine the statistical difference (i.e. P<0.05) in the mean value of the % SSIs between patients randomly assigned to bowel prep versus no bowel prep, while the “between-group” P values are regularly reported, the “within-group” values are invariably never reported. For example, if the average percentage (%) SSI in the group assigned to “no prep” is 20% and the average % SSI in the “bowel prep” group is 10% yielding a between group P value of <0.05, the trial is considered to be significant in favor of a bowel prep. Yet the reasons that 80% of patients in the “no prep” group did NOT develop an SSI (superficial wound infection based on CDC criteria, anastomotic leak based on symptoms triggering imaging, organ space infection with culture, etc.), and the reasons 10% of the patients in the “bowel prep” group still developed an SSI (i.e. the within-group differences), is never addressed. Results of such a study might suggest that a bowel prep should thus be applied to all patients, independent of their individual risk. Sweeping suggestions such as these, based on interpreted studies communicated via consensus guidelines make the argument that if some is good for some, some is good for all¹⁵. For busy gastrointestinal surgeons, whose perception is that an SSI, while unfortunate, is rarely observed in their patients, applying a bowel prep to all patients seems unnecessary¹⁵. This is especially relevant given that patients find purgative cleansing to be undesirable and use of oral antibiotics distasteful. Many surgeons and patients will report that the oral antibiotics come out whole in the toilet as the purgative cleansing and antibiotic pill intake is mistimed.

Disrupting the microbiome with a bowel prep: The good goes out with the bad

Surgeons in Europe have introduced the concept of selective digestive decontamination as an approach using the reasoning that “if some is good, more must be better.” In some cases, as many as 5 antibiotics including anti-fungal agents are proposed to be ingested prior to gastrointestinal surgery^16,17. This empirical approach is rationalized based on the clear evidence that certain microbes can cause infection-related complications following surgery¹⁷ (anastomotic leaks, wound infections, organ space infections, pneumonia, etc.)¹⁶ and therefore, eliminating as many as possible should be the approach¹⁷. This practice assumes that antibiotics are both effective and safe and therefore should be applied¹⁸. For example, there is now uncontestable and unambiguous evidence that an anastomotic leak, a catastrophe for any patient and surgeon, is a microbe-mediated complication from anastomotic surgery¹⁹. Certainly avoiding this complication is warranted and if the empirical use of antibiotics, again perceived to be safe and efficacious¹⁸, can be proven to be useful, they should be used. Yet in the current era of the selective use of a bowel prep versus no bowel prep, widely applying this practice seems untenable¹⁵. A right hemicolectomy in a young otherwise healthy patient that is being performed laparoscopically or robotically seems to be a case in which a bowel prep is unwarranted²⁰. Not only is wide application of antibiotics to all patients problematic from its destructive effect on the microbiome, but its overuse can lead to the emergence of antibiotic resistant organisms¹⁸. The predominance of antibiotic-resistant organisms causing SSIs today should make surgeons take pause when considering the approach of one-size-fits-all in terms of applying a bowel prep²¹.

Data are now emerging to suggest that components of the normal gut microbiome are necessary to drive a recovery-directed immune response in the host²². Multiple studies in which the gut microbiome is eliminated with antibiotics demonstrate a worse outcome when a subsequent infectious stress is imposed²³. In contradistinction, when the microbiome is restored with a fecal transplant, recovery from an infectious or traumatic stress returns to its baseline²². Studies such as these continue to demonstrate that preserving the beneficial effects of the microbiome is preferable over the near complete “slash and burn” approach that multiple antibiotics impose.

So what is it in the microbiome that drives a recovery directed immune response and how do we preserve it? Data are now emerging to suggest that key immune-activating metabolites released by the gut microbiota (i.e. indoles, tryptophan), become enriched in the plasma and target the aryl hydrocarbon receptor on immune cells such as macrophages²⁴, resulting in their activation to repair tissues and prevent infection. While it is becoming increasingly clear that preserving the metabolic output of keystone microbiota (i.e. the anaerobic population that produces key immune-stimulating microbiota), developing strategies that kill all the bad actors, while preserving the beneficial ones is no easy task. While many have proposed phage therapy^25–27, CRISPR-Cas 9²⁸ targeting of specific member of the gut microbiome, etc, today this is not yet clinically or practically possible. Yet widespread elimination with broad spectrum antibiotics seems to be the only method so far that is deliverable in the short term. However, this trend of applying broader and more powerful antibiotics to eliminate all microbiota, good and bad, is not sustainable as it is not a long-term evolutionarily stable strategy¹⁸. Resistant strains will emerge leading infectious disease specialists to plead for antibiotic stewardship. Yet herein lies the dilemma; if the only tool a surgeon has is antibiotics to prevent infection-related postoperative complications, they will be used more frequently and more broadly. Here we argue for the judicious use of antibiotic to slow this inevitable trend and await point-of-care diagnostics that will better identify those patients most at risk and thus those that might benefit from a full bowel prep.

What should be done today- a practical guide

Point-of-care diagnostics that can determine a patient’s “readiness” for surgery are needed²⁹. Not only is this possible today, emerging next generation technology using microbial sequencing and metabolomics is now possible. This has been demonstrated in mice²⁹ and certainly is possible in humans. Yet applying this approach in real time, in a costly and effective manner, has yet to be demonstrated, Therefore, the practicing surgeon is left to determine a patient’s risk based on his/her own assessment of the actual risks of an infection-related complication. While there are online tools that can assess a patients risk for a complication, much is left to be improved in this regard. There is no more devastating a complication than an anastomotic leak, where, while the patient is clearly the first victim, the surgeon is often the second victim. Despite all the confirmation bias that exists in this field, it has been clearly demonstrated that surgeons cannot predict which patients will leak, although they are sure that they can. The assertion that a right-sided anastomosis leaks less than a left-sided one, again has not been supported by evidence, yet the bias remains. In either event, avoid this devastating complication is always on the mind of the surgeon, and he/she will do anything to avoid it. A wound infection that can be treated with antibiotics, a wound vacuum device, and a minor opening of the wound or a organ space infection that can be treated by an interventional radiologist, while certainly not considered to be a minor complication, seems to be more acceptable in light of applying a bowel prep to all patients, independent of the patient’s perceived risk. While, in most practices, an anastomotic leak is a rare complication, when it occurs, it can be catastrophic. Therefore, surgeons must face the risks versus benefits of imposing a bowel prep on a given patient and must consider each patient as an individual case³⁰. This again becomes a judgement call requiring a candid conversation between doctor and patient. Despite consensus panels and guidelines suggesting that all patients should undergo a bowel prep, the ultimate decision is often based on weak evidence. Yet surgeons abjure on this particular issue given their perception of risk at the individual patient level. Unfortunately, guidelines regarding use of the bowel prep before colon surgery, are mixed^9,15. Depending on an individual’s practice, some say yes, while others say no⁹. While certainly a single dose of an intravenous antibiotic, seems safe, a patient’s dissatisfaction with purgative cleansing and oral antibiotics, perhaps unnecessary in many cases, needs to be balanced against the perceived risk. New data suggesting that preservation of the gut microbiome may drive a recovery-directed host response may make the current “slash and burn” approach to indiscriminately eliminating the microbiome with purgative cleansing and oral antibiotics to be balanced against emerging evidence that “bad actors” within the microbiome can cause SSIs. Therefore in low risk patients presumed to be candidates for enhanced recovery programs, dietary pre-habilitation and successful medical management of their co-morbidities, the bowel prep can be eliminated. However in high risk patients (i.e., colorectal surgery, neoadjuvant chemoradiotherapy, multiple medical co-morbidities), a bowel prep may avoid catastrophic infection-related complications³¹, which can be devastating for both patient and surgeon. Therefore, while recommendations may seem ambiguous, without with better “risk-assessment” tools to remove the uncertainty around whether to impose a bowel prep or not, concrete prescriptive guidelines cannot be generated. This situation creates a scenario of shared decision-making as to whether to offer a bowel prep between the surgeon and his/her patient.

Conclusion

In summary, new knowledge from sequencing and metabolomics of a fecal sample is emerging to suggest that among a given microbiome’s community structure, membership and function, are highly beneficial microbes that are necessary to drive a recovery directed response following major injury and exposure to infectious agents. As a result, leveraging next generation technology to guide our approach to selectively apply an effective method to prepare the bowel for surgery is now possible. Yet from a practical standpoint, given that we cannot predict who will develop a postoperative infection and who will not, similar to parenteral prophylactic antibiotics prior to surgery, it is often recommended that all patients undergoing gastrointestinal surgery, especially those undergoing colon surgery, have their bowel prepared prior to surgery with purgative cleansing and oral non-absorbable antibiotics. Yet give the immense dissatisfaction of patients to undergo such preparation, the bowel prep, as currently configured, is under siege. In many cases, surgeons feel it to be unnecessary, while patients are reluctant to undergo it. Surprisingly today, the field still cannot define what constitutes a “normal” microbiome³². Several investigators are attempting to define “enterotypes³³” that typify one patient versus another, however it turns out that microbiota, especially among those patients who have undergone neoadjuvant chemoradiotherapy, have been exposed to a western diet and antibiotics, who harbor cancer, correspond to microbiota whose community structure, membership and function are anything but “normal.” The challenge is to leverage next generation technology to guide us to an effective and palatable bowel prep 2.0, so that patients can be protected from the pathogens that lie beneath while preserving those microbiota that are beneficial for recovery. Like all decisions made with incomplete information, there being no other, whether to mandate a bowel prep for all patients undergoing colon surgery, to this day, remains a judgement call that should involve a robust discussion between the surgeon and patient³⁴.

Key Points.

• Microbiome sciences and next generation technology of stool analysis with genetic sequencing and metabolomics can interrogate the gut microbiota and inform the most optimal way to prepare the bowel for colorectal surgery

• Understanding which microbiota are “friends” and which are “foes” is essential to be able to properly prepare the bowel for surgery.

• Elucidating the scientific basis for preparing the bowel for surgery is a critical step in being able to rationally design a bowel prep for an individual patient.

• A one-size-fits-all approach to preparing the bowel for surgery is inappropriate- experts in the field and consensus panels drafting guidelines remain divided on patient selection and who should receive what.

Synopsis.

Precisely how to prepare the gastrointestinal track prior to colorectal surgery remains controversial. To date, there is no consensus on whether purgative cleansing is needed, oral antibiotics are needed, both are needed or whether intravenous antibiotics alone are sufficient. The concept that a “one-size-fits-all” approach can be universally applied to all patients undergoing colorectal surgery is waning in enthusiasm especially given the variation in risk perception among patients and surgeons. Furthermore, both patient and surgeon are dissatisfied by the discomfort caused by purgative cleansing and oral antibiotics. Next generation technology in microbiome sciences has revealed that pathogenic strains can complicated anastomotic healing and can cause a wound or deep organ space infection (surgical site infection-SSI) via mechanisms that occur beyond direct contamination of the tissue site. In this review, we discuss the scientific rationale in generating a microbiome readout that will inform how to selectively eliminate problematic pathogens while preserving those beneficial bacterial strains that may enhance recovery from surgery. We also provide the reader with practical decision points with which to contemplate the need for a bowel prep with your patient.

Clinics Care Points.

• Clinicians need to risk-stratify patients and anticipate the extent and duration of the proposed surgery before deciding whether to prescribe purgative cleansing, oral antibiotics, both or intravenous antibiotics alone.

• In young, otherwise healthy patients undergoing a limited right-sided colon resection performed in a minimally invasive way, a full bowel prep (purgative cleansing, oral antibiotics) is probably unnecessary.

• Imposing a one-size-fits-all approach to preparing the bowel for surgery is inappropriate. In the era of personalized medicine and individual care plans, each patient, depending on their real and/or perceived risk of an infection and the anticipated extent and length of surgery, should be prescribed a specified plan.

• Evidence is mounting to suggest that oral antibiotics alone, without purgative cleansing, may be as efficacious at preventing a surgical site infection and anastomotic leak compare to purgative cleansing + oral antibiotics. In some cases, purgative cleansing can be harmful.

• Preserving the beneficial function and recovery of the gut microbiome should be a goal of preparing the bowel for surgery. When appropriate, all efforts should be made to minimize complete and durable elimination of all gut microbiota. Understanding that both the host tissues of the patient, as well as their microbiota, need to recover from the stress of surgery, is critical.