The Science of Anastomotic Healing

Abstract

Intestinal anastomotic tissue follows a similar pattern of healing that is seen in all tissues with characteristic inflammatory, proliferative, and remodeling phases. Several aspects of intestinal healing are distinct from other tissues, however, including its time course and interaction with the environment of the gastrointestinal tract. As the anastomosis progresses through each stage, initial inflammatory cells are replaced by collagen-producing fibroblasts that generate the anastomosis’ strength. A complex network of cell-to-cell signaling mediates this process through the release of cytokines and growth factors including platelet-derived growth factor (PDGF), transforming growth factor-β (TGF-β), and vascular endothelial growth factor (VEGF). Interventions based on these signaling pathways have been shown to improve anastomotic strength in animals, though methods for improving anastomotic healing in human patients remain unclear. Given the risks associated with anastomotic failure in patients, there is value in monitoring inflammatory markers and cytokines that can indicate the presence of a leak.

Stages of Anastomotic Healing

Anastomotic healing following intestinal surgery consists of a well-characterized series of events beginning with hemostasis and culminating in collagen remodeling and strength acquisition. Apposition of the serosal edges of divided bowel initiates a biological sequence that largely mimics healing in other parts of the body. While much of our knowledge regarding the involved cell types, cytokines, and growth factors in anastomotic healing comes from the study of cutaneous wound healing, anastomotic healing has multiple unique attributes. Despite these nuances in anastomotic healing, intestinal tissue follows the same broad stages of healing as the skin that will be addressed below: inflammation, proliferation, and remodeling (Figure 1).¹

Figure 1:

Stages of anastomotic healing with associated functions, dominant cell types, and prominent growth factors.

Inflammatory Phase

The inflammatory phase of wound healing, also referred to as the exudative or “lag” phase, begins with hemostasis from platelet aggregation and the creation of fibrin clots.² These clots act to both halt any bleeding and to create a temporary scaffold for the deposition of initial matrix proteins. In addition to forming these fibrin plugs, platelets congregating at the site of a mucosal injury release multiple cytokines and stored chemoattractant growth factors, including platelet-derived growth factor (PDGF), vascular endothelial growth factor (VEGF), and transforming growth factor β (TGF-β). This cytokine-rich environment at the site of injury induces increased permeability of the surrounding small vasculature, leading to the migration of inflammatory cells to the healing tissue. Following platelet aggregation, neutrophils and then macrophages are the earliest inflammatory cells to migrate to the site of injury. These cell types remain dominant at the healing anastomosis throughout the inflammatory phase, performing multiple functions including the phagocytosis of potential pathogens and the release of further growth factors to stimulate wound healing.³

Proliferative Phase

Following the inflammatory phase, healing intestinal tissue enters a proliferative stage in which fibroblasts are the dominant cell type present. In this phase of healing, fibroblasts’ main function is to deposit collagen in the wound base, replacing the initial temporary fibrin clot with granulation tissue.¹ Fibroblasts are initially attracted to the cut edge of the bowel due to the release of chemotactic factors by macrophages including PDGF, TGF-β, and fibroblast growth factor (FGF). In contrast to the collagen composition of mature bowel, which is approximately 20% type III collagen, this early granulation tissue formed by fibroblasts contains upwards of 30–40% type III collagen and requires remodeling in the next phase of healing.¹ Fibroblasts begin to appear in the late inflammatory stage and are the predominant cell type in the wound bed by day four, with the proliferation phase generally lasting 14 days. As fibroblasts accumulate, the previously dominant inflammatory cells including neutrophils and macrophages decrease in number and are replaced by lymphocytes.⁴ In addition to collagen deposition, this phase is characterized by angiogenesis and re-epithelialization of the bowel’s mucosal surface, each stimulated by specific growth factors discussed below. Smooth muscle cells and myofibroblasts appear late in this phase as the anastomosis acquires strength, contracting to decrease the size of the eventual scar.

Remodeling Phase

The final and longest stage of anastomotic healing, remodeling, begins at approximately 14 days and continues for several weeks as the wound reorganizes into permanent scar tissue. Inflammatory cells that predominated in earlier stages of healing decrease further in number as the wound matures and, as in the proliferative phase, fibroblasts continue to be the main cell type present. The ratio of type I to type III collagen normalizes in the remodeling stage as matrix metalloproteinases (MMPs) work to remove type III collagen from granulation tissue and replace it with type I; following remodeling and maturation of the healing tissue, type I collagen represents 68% of total collagen content in the intestinal submucosa compared to 80% in healing dermis.^{1, 5} This more durable collagen mixture reorganizes into tight bundles while myofibroblasts contract to strengthen the healing wound. New vascular beds that began to form in the proliferative stage reorganize into mature networks as well, and after several weeks or months the anastomotic tissue stabilizes. Despite the time required for strength acquisition, healed wounds and anastomoses only partially regain the tensile strength of uninjured tissue.⁶

Differences in Anastomotic and Skin Healing

The main tenets of wound healing are conserved throughout the body as different tissues demonstrate the same stages of healing. While cutaneous wounds remain a focus for much of the literature on this topic, there have been an increasing number of studies focusing specifically on intestinal healing. Within the broader framework of wound healing established in dermatologic studies, there are several important characteristics distinct to anastomotic healing (Table 1).

Table 1.

Major differences between skin and anastomotic healing.

	Skin Wound	Intestinal Anastomosis
Time to Full Strength Acquisition	Gradual, up to 6 Months	Rapid, approximately 1 Month
Strength Layers	Dermis	Submucosa and Serosa
Collagen Synthesis
Collagen-Producing Cells	Fibroblasts	Fibroblasts and Smooth Muscle Cells
Main Collagen Subtypes (%)	Type I (80%), Type 111 (20%)	Type I (68%), Type III (20%), Type V (12%)
Collagenase Activity	Minimal	Increased in inflammatory phase (Colon > Small Bowel)
Wound Environment
pH	Stable	Wide range based on location in GI tract
Microbiome	Aerobic skin flora, infections generally treated topically. Rare instances of bacteremia.	Gut microbiome with aerobic and anaerobic bacteria. Potential for intra-peritoneal infection in setting of anastomotic leak
Mechanical Stress	Potential tension with movement depending on location. Minimal shear stress	Persistent shear stress from intraluminal contents, peristalsis
Perfusion	Remains constant	Splanchnic perfusion decreased in setting of hypovolemia

Perhaps the most important difference between the healing in the skin and the intestine is the time course of strength acquisition. While healing in the skin progresses gradually over several months, anastomotic tissue reaches near maximum strength within the first post-operative month.¹ Interestingly, strength acquisition is not uniform across the gastrointestinal tract as the rate of collagen synthesis and remodeling in an ileal anastomosis is quicker than in a colonic anastomosis in a rat model.⁷ Beyond gaining strength more rapidly, ileal and other small bowel anastomoses also attain near normal strength by four weeks post-operatively while colonic anastomoses acquire only 75% of normal tissue strength at four months.^{1, 8} These differences in healing between the small and large intestine may be at least partially due to increased collagenase activity in the colon, which causes a temporary decrease in anastomotic strength in the first three days post-operatively. This transient effect is more pronounced in the colon than the small intestine and may partially underlie differences in strength between the two tissue types. Clinically, this disparity in collagenase activity may also underpin the higher rate of anastomotic leak seen in colonic anastomoses compared to those in the small bowel, particularly in the descending colon.^9,10

The relatively rapid growth seen in anastomotic healing compared to cutaneous healing is essential due to the environment in the gastrointestinal tract. As the anastomosis gains strength and re-epithelializes, it is subject to continuous shear stress from intraluminal contents and peristalsis, both of which are not present in a cutaneous wound. Compounding this added stress, the healing bowel can experience relative ischemia as splanchnic blood flow is restricted in the setting of post-operative hypovolemia whereas the skin receives a relatively constant level of perfusion even in early hypovolemia.⁶

The gastrointestinal tract is also home to a broader and more diverse microbiome than the skin including anaerobic bacteria, which likely plays a role in the rate and efficacy of anastomotic healing.¹ This has been shown in animal models, with germ-free rats having notably lower bursting pressures at seven days after colectomy compared to conventional rats; ex-germ-free rats in this same study demonstrated significantly improved anastomotic strength after inoculation with a conventional microbiome but not after mono-contamination with Escherichia coli, confirming the beneficial effect of a diverse microbiome.¹¹ In humans, environmental factors including diet and antibiotic use can predispose patients to anastomotic leak mainly by altering the composition of the microbiome.¹² The promotion of a diverse and healthy microbiome may therefore be able to improve surgical outcomes. This has been shown in mice as prehabilitation with a low-fat diet decreases leak rates by preventing overgrowth of pathogenic bacteria.¹³

The composition of the intestinal wall represents a final difference between the two types of healing. The ratio of various collagen types differs in the bowel due to the presence of type V collagen, which is largely absent in the skin but makes up 12% of collagen in mature bowel.⁶ Pericellular type V collagen can be detected in colonic anastomoses in rats as early as post-operative day two and contributes to the basement membrane of regenerating capillary walls.¹⁴ The increased number of layers within the intestinal wall likely adds strength as well; in addition to its mucosal, submucosal, and muscular layers, the intestine includes an outer serosal layer that is not present in the skin. Expression of type I and type III collagen genes is seen in both the submucosal and serosal layers beginning on post-operative day one, suggesting that serosal healing plays an early role in strength acquisition.¹⁵ The overall arc of wound healing remains the same in the bowel as in the skin, but these intestine-specific nuances are important to consider in patients undergoing intra-abdominal surgery.

Growth Factors and Cell Signaling

The progression from each stage of healing to the next relies on a complex network of cell-to-cell signaling mediated in large part by growth factors and cytokines. This signaling begins early in the inflammatory stage of healing as platelets adhere to each other. In addition to providing hemostasis, platelets store multiple growth factors in pre-formed granules that are released in response to platelet aggregation. Essential growth factors in this step include PDGF, TGF-β, and VEGF, each of which plays a key role throughout healing and can be released by multiple inflammatory cell types.

Platelet-Derived Growth Factor (PDGF)

PDGF plays a particularly important role early in the inflammatory phase as it stimulates the release of TGF-β from macrophages while also increasing levels of VEGF and insulin-like growth factor I (IGF-I).¹⁶ Despite the central role that PDGF plays in signaling during the inflammatory phase, there is minimal evidence showing a causal link between its presence and improved anastomotic strength. Some indirect evidence actually shows that suppression of PDGF signaling improves anastomotic bursting pressure and hydroxyproline content in a rat model of intestinal healing. In a single study of rats treated with dexamethasone, those who were given post-operative intravenous trapidil, a competitive PDGF antagonist, showed improved anastomotic bursting pressure and hydroxyproline content; no difference was demonstrated in non-steroid treated rats, however.¹⁷ It is difficult to draw definitive conclusions about PDGF suppression based on this model, however, due to the multiple other functions of trapidil including blockage of thromboxane A₂ and prostacyclin synthesis.

Transforming Growth Factor-β (TGF-β)

Prior animal studies on the role of TGF-β have similarly conflicting evidence for the role of this growth factor in early anastomotic strength acquisition. Increased TGF-β signaling via adenoviral gene transfer in rats with healing colonic anastomoses increases bursting pressure on post-operative day six; interestingly, increased bursting pressure was only demonstrated when TGF was administered on postoperative day three whereas no benefit was observed with administration on post-operative day zero.¹⁸ Other findings have disputed this benefit, however, as suppression of TGF-β signaling with the intraperitoneal administration of TGF-β neutralizing antibodies in rats with healing small bowel anastomoses was shown to improve mucosal growth and revascularization by post-operative day five.¹⁹ These divergent responses to TGF-β modulation may be attributable to differential TGF-β expression in colonic and small intestine anastomoses. Compared to normal intestinal tissue, colonic anastomoses in rats have significantly lower levels of TGF-β expression on post-operative days one and seven, whereas expression in the small intestine remains at baseline.²⁰ The lack of clarity on the role of TGF-β in anastomotic healing is likely indicative of the complex nature of cell signaling during tissue healing, and further study is warranted.

Vascular Endothelial Growth Factor (VEGF)

With the introduction of new cell types in the proliferative phase, signaling pathways emerge that are essential in achieving the goals of neo-angiogenesis, mucosal regeneration, and formation of the extracellular matrix (ECM). Among these, revascularization through neo-angiogenesis plays a critical role in the progression of anastomotic healing by increasing oxygen supply and nutrient delivery to the wound bed. The key growth factors underlying neo-angiogenesis are VEGF and FGF. While VEGF also stimulates fibroblast migration in the proliferative phase, its primary role is to promote endothelial cell migration and proliferation, leading to an angiogenic effect in the healing anastomosis.¹⁶ VEGF administration in animal models has been shown to increase colonic anastomotic strength as early as post-operative day four.²¹ Gene therapy with recombinant VEGF in an opossum model of esophagogastrectomy similarly shown that increased expression of VEGF has a beneficial effect on anastomotic healing.²² This evidence for the central role of VEGF has been strengthened by studying animal models of bevacizumab administration, a VEGF inhibitor; in rabbits treated with bevacizumab, bursting pressure of intestinal anastomoses significantly decreases compared to controls.²³ While similar study in rats found no deleterious effects from bevacizumab administration, the majority of evidence from animal models suggests that the pro- angiogenic effects of VEGF are essential in anastomotic strengthening.²⁴

Keratinocyte Growth Factor (KGF) and Insulin-Like Growth Factor-I (IGF-I)

Mucosal regeneration occurs simultaneous to neo-angiogenesis during the early proliferative phase via keratinocyte growth factor (KGF). The role of KGF in wound healing has primarily been studied in cutaneous wounds, though there is some evidence for its role in intestinal healing. Intraperitoneal delivery of KGF in rats with left-sided colon anastomoses significantly thickens the mucosal layer of the anastomosis, reduces inflammation, and increases bursting pressure on post-operative days two, four, and seven.²⁵ The same group also showed that IGF-I had a similar effect as KGF, though co-administration did not lead to further acceleration of anastomotic healing.²⁶ Further studies have showed that exogenous IGF-I administration improves anastomotic strength and healing in animals treated with a variety of immunosuppressive medications including 5-fluorouracil (5-FU), corticosteroids, and mycophenolate.^27–29 Taken together, these studies suggest that this positive effect from IGF-I may act to counteract the negative affect of immunosuppression via increased mucosal production.

Other Growth Factors

A host of other growth factors and cell signaling pathways also play a role during the proliferative phase. Fibroblast growth factor, which increases through the proliferative phase in conjunction with multiplying fibroblasts, has been shown to improve healing in multiple types of anastomoses; prior animal models demonstrating this effect include a rat model of esophageal anastomosis with direct FGF application by gelatin film, as well as a dog model of pancreaticojejunostomy with local FGF administration using hydrogel microspheres.^{30, 31} Epidermal growth factor (EGF) and human growth hormone (hGH) have each been shown to improve anastomotic healing and bursting pressure as well, specifically in rats treated with steroids or other immunologic agents.³² Heparin-binding EGF-like growth factor (HB-EGF) has likewise been shown by one group to improve anastomotic healing in mice with direct application; importantly, HB-EGF knockout mice in the same study were shown to have decreased anastomotic strength, suggesting a critical role for HB-EGF.³³ While the exact mechanisms underlying the impact of many of these growth factors remain undefined, these recent studies have helped to develop a basic understanding of the signaling pathways that guide anastomotic healing.

Tissue Remodeling

Collagen Synthesis and Degradation

Fibroblasts are the most common cell type present in the wound bed during the proliferative phase, carrying out several functions including recreation of the ECM and collagen deposition in granulation tissue. During the proliferative phase, high levels of IL-1β and PDGF stimulate the activation of myofibroblasts, a form of activated fibroblast that participates in wound contracture.³⁴ The number of fibroblasts peaks during the proliferative phase and then gradually decreases as collagen concentration and wound breaking strength increase.¹ This reduction in fibroblast number is in part due to the actions of TGF-β1 and FGF, which induce apoptosis of granulation tissue fibroblasts.³⁵ With this decline in fibroblast function, wound healing transitions from collagen production to remodeling.

Collagen content in colonic anastomoses has been shown to peak at six days post-operatively in a rat model, with collagen deposition slowing after day six. Notably, this model showed that collagen content in the healing anastomosis was tenfold higher than in healing skin wounds.³⁶ Small bowel anastomoses may reach peak collagen content significantly sooner with one animal model showing a tenfold maximum increase in collagen activity in ileal anastomoses by day four; this response to wounding in the small intestine was both faster and stronger than in colonic anastomoses.⁷ Increased collagen content alone is not necessarily indicative of anastomotic strength however; strength is derived from the reorganization of collagen, and inhibition of cross-linking has been shown to decrease anastomotic strength without altering overall collagen content.³⁷

In all intestinal tissues, organization and remodeling of the immature collagen matrix deposited by fibroblasts is largely mediated by MMPs that degrade the ECM. This family of proteins has over 20 distinct subtypes in humans divided into subgroups based on which ECM components each enzyme breaks down (i.e. collagenases, gelatinases).³⁸ Effective remodeling depends on a balance of activity between MMPs and tissue inhibitors of metalloproteinases (TIMPs), allowing for the recycling of collagen and creation of a mature scar. Various MMPs and TIMPs are highly expressed at different points in wound healing, with MMP-3, MMP-8, MMP-9, TIMP-2, and total MMP activity shown to peak early and decrease over time while MMP-2 and TIMP-1 gradually increase.³⁸ Expression varies by tissue type as well, with MMP expression patterns varying considerably between ileum and colon samples; zymography from rat tissue extracts shows bands likely corresponding to pro-MMP-2 and active MMP-2 that are only found in the ileum, while different pro-MMP-1 subtypes are expressed in each tissue. ³⁹

Remodeling of the anastomosis is dependent on MMPs, but over-expression of these enzymes early in anastomotic healing can lead to unbalanced collagen degradation resulting in anastomotic dehiscence. Pre-operative ischemia, which is a risk factor for anastomotic dehiscence, induces more widespread MMP and TIMP-1 expression in peri-anastomotic colonic tissue in a rabbit model during the first several days post-operatively. This overexpression drew an indirect association between MMP activity and anastomotic failure, which has been further bolstered by evidence that microbiome-stimulated activation of MMP-9 contributes to anastomotic leak as well.⁴⁰ Consistent with these findings, inhibition of MMP activity significantly improves bursting strength in rat anastomoses.⁴¹ Different factors within the anastomotic microenvironment may then influence rates of dehiscence by altering MMP activity. In normal healing tissue, these effects are likely attenuated by inhibition of MMP activity by TIMPs and other growth factors present in early healing such as TGF-β, which acts to suppress MMP-9 function.⁴²

Interventions to Improve Anastomotic Strength and Healing

Failure of anastomotic healing and leakage of intraluminal contents can have devastating consequences for patients undergoing intra-abdominal surgery. Given the high risks associated with anastomotic leak, there have been many studies on interventions to improve anastomotic healing in animal models. Due to the central role that MMPs play in collagen remodeling, they have been a frequent target for inhibition in these studies. Specific, systemic inhibition of MMP-8, MMP-9, and MMP-12 has been shown to improve bursting strength in colonic anastomoses in rats, with similar benefits seen with non-selective inhibition in other rat models.^{41, 43–45} Local inhibition using sutures coated with doxycycline, an MMP inhibitor, have demonstrated improved anastomotic strength as well.⁴⁶

Direct application of several growth factors discussed in the previous section have been studied as well. VEGF gene therapy or administration in particular seems to improve anastomotic healing, with animal models of esophagogastrostomy and colonic anastomosis both demonstrating this beneficial effect likely due to accelerated angiogenesis.^{21, 22, 47} In animal models using immunosuppressive medications such as corticosteroids or 5-FU there is a benefit from systemic IGF-I administration as well.^27–29 This improvement in immunosuppressed mice correlates with evidence that IGF-1 coated sutures improve anastomotic strength and hydroxyproline content in a rat model of colitis.⁴⁸ The positive effect of growth factor administration in immunosuppressed animal models is of particular interest considering the wide variety of patients on immune-modulating therapy who require intestinal surgery.

Other interventions from previous studies have had mixed results. Non-steroidal anti-inflammatory drugs (NSAIDs) have been proposed as a novel intervention to limit inflammation and subsequently promote healing, and one rat model did show that NSAIDs can limit collagen degradation in colonic anastomoses.⁴⁹ Other animal models have shown no effect on anastomotic strength from NSAIDs, however, and their clinical impact seems to be minimal.⁵⁰ Other systemic interventions are less well-studied. Preparation with glutamine and synbiotics was shown to increase bursting pressure in a rat model, while platelet rich plasma and a superoxide dismutase mimic improved bursting pressure in rats receiving 5-FU and chemoradiotherapy, respectively.^51–53

Previous systematic reviews have compared various less well-studied agents, though almost exclusively in animal models. In a 2016 meta-analysis of animal studies by Nerstrom et al. hyperbaric oxygen was the only identified beneficial intervention, improving bursting pressure in rat models of anastomotic ischemia; granulocyte macrophage-colony stimulating factor showed no benefit in chemotherapeutic models, and there was inadequate data to evaluate any other agents in the metaanalysis.⁵⁴ A separate meta-analysis from Oines et al. in 2014 identified several other agents that increased bursting pressure in animal models including iloprost, tacrolimus, erythropoietin, hGH, and IGF-I. This group again pointed to the benefit of MMP inhibition as well. Only one agent in this meta-analysis, aprotinin, was used in human patients, and it showed no impact on the rate of anastomotic leak in a cohort of 113 patients.⁵⁵

Monitoring of Anastomotic Healing

Anastomotic dehiscence leads to a broad spectrum of post-operative complications, ranging from subclinical leaks to life-threatening sepsis. Given the potential for devastating complications, there have been many studies aimed at predicting or diagnosing anastomotic leak early in its course. Classical indicators of anastomotic leak, such as altered vital signs, are often non-specific with exceedingly small positive predictive value for a leak.⁵⁶ More promising data has emerged from studying markers in either serum or perianastomotic peritoneal fluid. These diverse indicators are generally inflammatory markers, cytokines, or metabolites that indirectly indicate inflammation or ischemia at the healing anastomosis.

Inflammatory markers and Cytokines

The most well-studied biological markers of anastomotic leak after intestinal surgery have been inflammatory markers such as C-reactive protein (CRP), calprotectin, and procalcitonin. CRP and calprotectin are elevated in the setting of intestinal inflammation while procalcitonin is a marker for bacterial infection, both of which are prominent features of anastomotic leak.^{57, 58} Their utility in the detection of anastomotic leak is best as a screening test with high negative predictive value (NPV) for normal values. Serum CRP levels, for example, have been shown to have a NPV of 89–97% in colorectal surgery patients when using a cutoff value between 135 and 180 mg/L on post-operative day four.^{59, 60} Procalcitonin and calprotectin serum levels have been shown to be similarly useful, though the best NPV is observed when using these markers in combination with CRP.^{61, 62} Combining serum CRP levels with intraperitoneal MMP-9 levels was also shown to increase its predictive value.⁶³ Given these repeated findings, normal values for serum inflammatory markers may have a role in guiding early discharge after colorectal surgery. It is important to note, however, that these markers tend to be less accurate in predicting leak following laparoscopic surgery when compared to open surgery.⁶⁴ As minimally invasive techniques continue to proliferate, the usefulness of measuring inflammatory markers to rule out leak may become less clear.

Cytokines involved in immune cell activation have also been identified as markers for anastomotic leak, both in serum and intraperitoneal fluid. In particular, TNF-α, IL-1β, IL-6, IL-8, and IL-10 have been frequently studied in this capacity. Individually, elevation of each of these cytokines is nonspecific for anastomotic leak, but combining them into models may help screen patients for dehiscence.⁶⁵ This is particularly true in the early post-operative period; IL-6, IL-10, and TNF-α have all been observed to be higher on post-operative day 1 in the intraperitoneal fluid of patients with anastomotic leak.^{63, 66, 67} Interestingly, patients with early anastomotic leak have elevated levels of these intraperitoneal cytokines as well as IL-1β and IL-8 during the first five post-operative days, yet those with leaks later in their course do not have similar early cytokine responses.⁶⁸ These findings suggest that the utility of intraperitoneal cytokines for detecting leak is highest immediately post-operatively and then diminishes.

Acid/Base Status & Metabolites

In addition to measuring serum markers, clinicians can use output from drains placed at the time of surgery to surveil the anastomotic environment. Peritoneal fluid output from either traditional surgical drains or smaller microdialysis catheters have been used with success in a variety of studies to characterize changes seen in anastomotic leak. Fluid pH and the concentration of glucose metabolites have shown the best efficacy in predicting leak, likely due to the effect of ischemia on carbohydrate metabolism.

In the absence of adequate oxygen supply, anaerobic metabolism depends on glycolysis to break down glucose with fermentation of the resultant pyruvate, producing lactic acid. Decreased oxygen delivery due to ischemia at the site of a leak should then logically lead to increased fermentation and decreased pH as the concentration of lactic acid increases. In a recent prospective study of 753 patients undergoing rectal surgery, the pH of drain effluent was shown to be a reliable marker of anastomotic leak; using a pH cutoff of 6.798 led to a remarkable high sensitivity and specificity of 98.7% and 94.7%, respectively, for anastomotic leak.⁶⁹ This group did not evaluate concentrations of glucose metabolites, but this evidence for pH as a marker corresponds well with previous findings regarding peritoneal concentrations of glucose, lactate, and pyruvate in anastomotic leak. Numerous studies have shown that peritoneal lactate increases in the setting of leak, though with somewhat low predictive values.^{70, 71} Lactate to pyruvate ratio, which increases in the setting of ischemia, is similarly elevated in the setting of anastomotic leak.^{66, 72}