4.1 Ischaemia

Pre-capillary sphincters and intestinal arterioles help sustain convective and diffusive delivery of oxygen to the tissue (Figure 2).^181-185 The combination of these mechanisms provides substantial reserve when arterial pressure decreases, with recruitment of capillaries contributing more than arteriolar dilation.¹⁸⁶ Collateral blood flow also helps prevent intestinal ischaemia and tissue hypoxia. There is nonetheless a tissue oxygenation threshold below which mucosal permeability is inversely related to oxygen tension, typically about 40–45 mmHg.¹⁸⁶

IRI starts with ischaemia and evolves over the course of minutes to hours.²⁹ At the molecular level, structural and functional changes in the mucosa have been observed after as little as 5–15 min of ischaemia. Ultrastructural studies of dog ileum suggest that intracellular mucosal damage (damage to the endoplasmic reticulum) occurs after 10 min of arterial occlusion.¹⁸⁶ After 30 min of arterial occlusion, the intracellular spaces are widened, some epithelial cells have lifted from the basement membrane, and there is now overt cell failure characterised by mitochondrial vacuolisation, decreased oxygen uptake, loss of adenosine triphosphate and release of lysosomal enzymes.¹⁸⁶ The most notable reversible microscopic finding of acute gastrointestinal ischaemia is called Gruenhagen's space and can be histologically identified in the small bowel villi and colonic surface epithelium within 30–60 min after injury onset.^{3, 187} It is created by the accumulation of leaked intracellular fluid from injured cells and products of the topical response by fibroblasts and myocytes of the adjacent lamina propria.^{3, 177, 188} Gruenhagen's spaces resolve when the ischaemic cascade stops sufficiently early.¹⁷⁷ Arterial occlusions for two or more hours frequently produces massive epithelial lifting down the sides of the villi, completely denuded villi, disintegration of the lamina propria, increased mucosal permeability and net water loss into the bowel lumen.¹⁸⁶

Although the ischaemic process is initially local and often reversible, a far more serious and often irreversible, change to the epithelium is disruption of its barrier function which is largely determined by oxygen availability and/or utilisation. Intestinal oxygen uptake usually remains constant over a wide range of blood flows and pressures and is only compromised when perfusion is below a critically level (Figure 3). In hypoxic conditions, microcirculatory oxygenation is heterogeneous, with well-oxygenated microcirculatory units co-existing units that are hypoxic because of microcirculatory shunting.

Pd-porphyrin phosphorescence imaging and microsphere studies show, perhaps unsurprisingly, that hypoxic capillary units are found close to venules whereas oxygenated units are typically adjacent to arterioles.¹⁸⁹ The relatively hypoxic units, naturally, are first to suffer and last to recover after IRI.^{190, 191} Shunting of oxygen from the microcirculation explains shock conditions such as haemorrhage and sepsis during which regional hypoxia/dysoxia is evident despite sufficient systemic oxygen delivery (Table 2).^192-194

Low perfusion states, microcirculatory shock and loss of autoregulation alters the relation between mitochondrial oxygen consumption and cellular oxygen partial pressure.^{181-183, 195, 196} Specifically, graded reductions in blood flow produce concomitant reductions in cellular oxygenation without altering mitochondrial oxygen uptake.¹⁸⁶ The latter is reduced only when cellular oxygenation falls below a critical level; thereafter, low oxygen diffusion into cells decreases intracellular oxygenation to levels that no longer support normal oxidative metabolism.

Barrier dysfunction depends on the integrity of intercellular junctions between the intestinal epithelial cells, mainly tight junctions, adherens junctions and desmosomes¹⁹⁷—each of which is disrupted by sufficient ischaemia.^198-200 Blood flow reduction leading to a >50% decline in intestinal oxygen consumption is accompanied by mucosal damage, depression of mucosal Na⁺/K⁺-ATPase activity (which causes cell swelling and autolysis) and barrier dysfunction. The result is hypoxia/dysoxia-induced mucosal permeability, with injury related to the reduction in oxygen consumption (Figures 4 and 5).^{193, 201-203}

As ischaemia progresses, intestinal epithelial cells detach from each other and from the basal membrane, propagating from the luminal to the basilar/crypt cells as expected from the normal oxygen gradient. Cell death mediated by necrosis and apoptosis exacerbate injury, while insufficient clearance of dying cells leads to increased inflammation and impaired tissue repair.⁴⁷ Loss of barrier function eventually results in shedding of the epithelial lining into the gut.^{3, 4, 158} A consequence is topical infiltration of luminally located toxins and bacteria which are part of the normal intestinal flora. Digestive enzymes produced by exocrine organs including the liver and pancreas also infiltrate and aggravate injury.²⁰⁴ The final manifestation of an ischaemic injury is dissemination of toxins and bacteria via the mesenteric lymphatics and consequent systemic inflammatory response, sometimes leading to multiple organ dysfunction.^{30, 138, 205-207}

4.2 Reperfusion

The progression of IRI does not conclude with correction of the initial causative pathophysiology; instead, restored perfusion and oxygenation provokes additional tissue injury.^{29, 35, 190, 208} When ischaemia is local, reperfusion is achieved by a spontaneous or therapeutic restoration of macrovascular and/or microvascular flow. In the case of systemic hypoperfusion, reperfusion occurs with the restoration of haemodynamic stability and/or resolution of conditions that compromise splanchnic perfusion.^{32, 40} The aetiologic factors along with processes implicated in IRI are depicted in Figure 6.

The biochemical basis for reperfusion injury is local production of reactive oxygen species.^{5, 97, 98, 102, 104-111, 125, 209-215} Reactive oxygen species are generated when ischaemic cells consume available adenosine triphosphate reserves and then produce metabolism byproducts including hypoxanthine and xanthine.²¹⁶ The enzyme xanthine dehydrogenase is a normal part of purine metabolism and uses nicotinamide adenine dinucleotide as a co-factor to produce reduced nicotinamide adenine dinucleotide and uric acid. But in hypoxic conditions, xanthine dehydrogenase undergoes irreversible proteolytic modification, mostly by trypsin, and converts to xanthine oxidase which uses the same substrates as xanthine dehydrogenase but consumes oxygen as a co-factor to produce hydrogen peroxide.^{107, 217} Upon restoration of oxygenation, xanthine oxidase metabolises the newly abundant oxygen which markedly increases the intracellular concentration of hydrogen peroxide. Accumulation of peroxide leads to the enzymatic and non-enzymatic generation of several other oxygen-based free radicals (Figure 7).^{96, 97, 211, 212}

Activation of nicotinamide adenine dinucleotide phosphate oxidases, a family of integral enzymes found in many tissues including intestine, results in an active protein complex that also produces extracellular hydrogen peroxide.^218-221 Furthermore, nitric oxide synthase functions as a homodimer and produces nitric oxide under normal conditions. But when the available tetrahydrobiopterin is exhausted, such as during oxidative stress induced by xanthine oxidase and nicotinamide adenine dinucleotide phosphate oxidases, the molecules uncouple and produce hydrogen peroxide instead of nitric oxide.^{202, 222} Hydrogen peroxide accumulation from various sources and subsequent reactive oxygen species production is deleterious both by disrupting intracellular structures and via strong chemoattraction of neutrophils.

An additional source of reactive oxygen species during reperfusion is mitochondrial dysfunction. The initial steps include disruption of normal electron transport chain function, most importantly complex I, exhaustion of antioxidant mitochondrial capacity, generation of superoxide anion and its rapid transformation to hydrogen peroxide, opening of mitochondrial permeability transition pore (mPTP) and intracellular release of hydrogen peroxide and cytochrome c.^{218, 223} In an environment rich in reactive oxygen species, mitochondrial DNA which is normally located in the mitochondrial matrix is released into the cytoplasm via the mPTP or via vesicle formation and cytoplasmic release.^224-228 Mitochondrial DNA functions as a damage-associated molecular pattern and its presence in the cytosol is implicated in the activation of several cellular pathways, analysed below.

Aggressive infiltration of affected tissues by neutrophils and their inflammatory activity is the most prominent immunologic component of the reperfusion phase of intestinal IRI. The first step in this process is chemotactic and cytokine-mediated attraction of the inflammatory cells. Attraction is mediated by various compounds. First, cytoplasmic mitochondrial DNA activates the endoplasmic Toll-like receptor-9,²²⁹ which usually functions as a pathogen recognising system in intestinal epithelial cells.^{230, 231} Activated receptors binds myeloid differentiation factor 88 and the complex further activates the NF-kB pathway which provokes release of cytokines.²²⁹ Second, increased cytosolic concentration of reactive oxygen species and the cytosolic presence of mitochondrial DNA activate the nucleotide-binding domain and leucine-rich repeat pyrin 3 domain inflammasome which has been identified in both intestinal epithelial cells and intestinal immune cells.^{232, 233} Inflammasome formation causes caspase 1 activation in multiple cellular pathways including ones that release IL-1β and IL-18.²³⁴ These cytokines are released along with other systemically proinflammatory mediators including transcription factors, hypoxia-inducible factor (HIF)-1, cyclooxygenase-2 and poly(ADP-ribose) polymerase to attract neutrophils to the affected intestine.^235-238 Digestive enzymes, toxins and bacteria that cross the compromised mucosal barrier also contribute.^{179, 239, 240}

Neutrophils infiltrate affected tissue through gaps they create between the endothelial cells and through the endothelial basal membrane, substantially increasing capillary permeability.¹⁶³ Neutrophil accumulation worsens tissue hypoxia because these metabolically activated cells demand much oxygen.²⁴¹ But the most deleterious effect of neutrophils is their inherent cytotoxicity which includes the release of chemical compounds, notably reactive oxygen species^{107, 112, 125} and enzymes such as elastase, myeloperoxidase, protease-3 and metalloproteinases (Figure 8).^{110, 114, 242} The putative purpose of releasing these compounds is to kill bacteria and other invading pathogens, but they also cause collateral damage to the extracellular matrix and surrounding cells.

Another contributor to IRI is microcirculation. Disruption of capillary and venule walls by neutrophilic infiltration causes perivascular oedema and endothelial injury which creates external pressure on micro-vessels. Swollen endothelial cells reduce luminal diameter and microvascular blood volume, a phenomenon that is well described in the cardiac circulation.^{108, 243, 244} Consequent reduction in local blood flow decreases the relative haematocrit in small vessels (Fåhræus effect; Table 3)^{245, 246} and increases endothelial shear stress.⁴³ Intravascular thrombus formation and exacerbated local hypoperfusion is aggravated by neutrophil clustering within capillaries, platelet aggregation, exposure to various tissue-based coagulation factors and subsequent activation of the coagulation cascade.^{99, 247}

Red blood cells also contribute by releasing ATP, nitric oxide and S-nitrosothiols. During deoxygenation, haemoglobin reacts with nitrite to form nitric oxide, but it also has a high affinity for nitric oxide. Scavenging of nitric oxide by haemoglobin can cause vasoconstriction, which is greatly enhanced by extracellular haemoglobin.²⁴⁸ And finally, low red cell volumes promote deformation and folding,^{249, 250} whereas high red cell volumes promote diffuse trajectories and shear.²⁵¹ Both derangements decrease delivery of oxygen to intestinal cells.

Many mechanisms thus conspire to seriously compromise local perfusion during ischaemia–reperfusion and create vicious cycles that can prevent recovery.

5.1 Regulation and function of HIFs

The linchpin of cellular responses to hypoxia is a group of transcription factors called HIFs. Initial studies of HIF evaluated the renal response to hypoxia and consequent production of erythropoietin which is mediated by HIF system activation.^252-255 HIFs have a heterodimer structure with an alpha and a beta subunit.²⁵⁶ The intracellular concentration of the active complex depends on the rate of hydroxylation of the alpha subunit and subsequent funnelling of the entire molecule through a ubiquitination and degradation process mediated by the von Hippel–Lindau tumour suppressor protein.^{257, 258}

In normoxic conditions, HIF-1α is degraded by hydroxylation at an oxygen-dependent rate that is mediated by the enzyme family of prolyl hydroxylase proteins which use oxygen as a substrate. Thereafter, von Hippel–Lindau E3 ubiquitin ligase complex completes the breakdown of HIF to proteasomes.^259-262 Prolyl hydroxylase protein activity, and thus the hydroxylation and degradation of HIFs, is suppressed in hypoxic environments which leads to intracellular accumulation of HIFs. When concentrations increase sufficiently, HIFs migrate into the nucleus²⁶³ where they bind specific histone acetyl-transferases (CBP and p300) and form a complex that enhances transcription of numerous genes.^264-267 HIF proteins are present in various types of cells, with HIF-1 and HIF-2 identified in intestinal epithelial cells, and in other intestinal cells including lymphocytes (Figure 9).^{268, 269}

HIF concentration also depends on enzymes called factors inhibiting HIF which are the only known hypoxia-sensitive asparagine hydroxylases. Post-translationally, factors inhibiting HIF modify an asparagine residue in HIF-1α, a change that prevents HIF-1α from binding p300/CRB, thereby depressing HIF activity.²⁷⁰ Like prolyl hydroxylase domain-containing proteins, factors inhibiting HIFs are less active under hypoxic conditions, thus enabling HIF-1α to effectively engage with its co-activators and enhance gene expression.

The sensitivity of prolyl hydroxylases and factors inhibiting HIF is expressed by the value of the Michaelis–Menten constant (K_M), or the oxygen tension at which the reaction rate is half-maximal.²⁷¹ Given an average intracellular oxygen tension of 7–20-mmHg (corresponding to 1.0–2.5% oxygen), factor inhibiting HIF (K_M = 50–80 mmHg/6.5–10.5% oxygen) may be more effective than prolyl hydroxylase (K_M = 120–210 mmHg/15.7–27.6% oxygen) as a physiological oxygen sensor.²⁷¹

The HIF system is an integral mediator of the cellular response to hypoxia. For example, mitochondrially generated reactive oxygen species appear to increase intracellular concentration of HIFs, although the specific chemical pathway remains unclear.²⁷² Carbon monoxide and ammonia similarly trigger the HIF cascade, eliciting the same physiological response.²⁷³ In contrast, nitric oxide, hydrogen sulphide and carbon dioxide inhibit the system.²⁷³

Immune mediators including tumour necrosis factor-a and interleukin-1b,^274-276 bacteria and bacterial components such as lipopolysaccharides all reduce HIF deactivation, thereby increasing intracellular HIF concentrations—even in normoxic conditions.^277-279 Many other proteins such as STAT3 and immunologic systems such as PI3K/AKT/mTOR, RAS/RAF/MEK/ERK and IKK/NF-κB also modify the complex HIF cascade.^280-283 Consequently, the effects of HIF activation are multifaceted, only partially understood, and depend on the duration and activating mechanism.²⁸⁴

5.2 The effect of HIFs on intestinal ischaemia–reperfusion injury

Activation of the HIF cascade leads to structural and functional changes in intestinal epithelial cells. Structural changes include maintenance and augmentation of tight junctions. For example, HIF promotes expression of the CLDN1gene and then production of claudin-1 protein which is a critical component of tight junctions. Functional changes include increased production and luminal secretion of mucus structural components (primarily mucins via augmentation of the MUC family genes) that stabilise mucus proteins such as intestinal trefoil family factor peptides and antimicrobial and immune mediating peptides including defensins, lysozyme and secretory phospholipase 2.^285-288 Combined, the net effect is to maintain tight junction integrity and mucosal barrier function, along with mucosal permeability under hypoxic conditions.²⁸⁹ However, HIF-1 can also augment apoptotic and inflammatory processes which can aggravate IRI-induced gut mucosal injury leading to barrier permeability, bacterial translocation and systemic inflammation.²⁸⁴ Which effect dominates depends on the circumstances and remains poorly understood.

The effect of hypoxia and increased HIF concentration on intestinal immune cells is also important. The main intestinal immune cells are T lymphocytes. ‘Conventional’ T cells (CD4- and CD8αβ⁺ cells) are located intra-epithelially and in the lamina propria and mainly defend against infections. In contrast, ‘unconventional’ T cells (CD8αα⁺, CD8αα⁻, CD4⁻ and CD8αβ⁻), which are also localised intra-epithelially, regulate immune responses and maintaining intestinal homeostasis.²⁹⁰ Activation of the HIF cascade has several effects on T cell function. First, it induces a shift from primarily using fatty acids when quiescent to intense glycolysis when activated. The shift is mediated by increased expression of the SLC2A1 gene, augmented function of the GLUT1 transporter which increases the intracellular availability of glucose, and activation of the genes encoding relevant glycolysis enzymes (including lactate dehydrogenase, pyruvate kinases and phosphofructokinase 1).²⁹¹ The second less understood effect pertains to cytokine production, transition to effector cells and immune cell recruitment from secondary lymphoid tissues.^{292, 293} Available data suggest that an increase in HIF promotes activation, cytotoxicity and altered transition to memory cells^{292, 294}—all of which can promote hypoxia-induced intestinal injury.^{295, 296}

Several other types of immune cells are affected by the HIF cascade activation including dendritic cells, macrophages and neutrophils—all of which are present in the gastrointestinal track. Again, the best investigated change observed in these cells is the switch from aerobic to glycolysis-based metabolism via the same gene and protein upregulation described for lymphocytes. This shift constitutes an integral and critical step of immune cell activation, function and proliferation.²⁸³ Activation of HIFs in intestinal dendritic cells increases immune cell migration to affected tissues and maturation and antigen presentation. The mechanism appears to be upregulation of several cell membrane receptors, most prominently the Toll-like receptor-2 and -6 and the CCR7 surface receptor.^297-300 Changes to the dendritic cell immune mediator production under the influence of HIF activation remain under investigation and are not yet well understood.²⁸³

HIF system activation polarises macrocytes toward their proinflammatory sub-type which increases motility and bacteriophage activity.³⁰¹ Apart from the metabolic profile switch to glycolysis, this process includes alteration of membrane receptor constitution, notably with the increase of CD40 and CD206 receptors and the production and excretion of several different cytokines including IL-1β, TNF-α and IL-6.^302-305 Under IRI and hypoxic conditions, the HIF pathway prolongs neutrophil survival and increases phagocytic and cytotoxic function. A pronounced effect of HIF activation is to increase neutrophil survival by intensifying glycolysis which suppresses hypoxia-induced apoptosis.³⁰⁶ Oxidative killing of bacteria requires a respiratory burst and consequent increase in oxygen consumption³⁰⁷ which exacerbates tissue hypoxia (Figure 10). Local hypoxia stabilises HIF transcriptional machinery²⁴¹ which probably protects mucosa and helps resolve inflammation. Finally, HIF activation induces leukocyte β2 integrin expression, a critical structural and functional protein component for CD18 that mediates leukocyte adhesion and extravasation³⁰⁸ along with protein production.

Matrix metalloproteinases (MMPs) are zinc-dependent proteinases which regulate many cellular activities and can degrade nearly all extracellular matrix components. MMPs are produced in abundance once reperfusion is established and play a central role in disease progression.^{309, 310} For example, HIF directly increases various metalloproteinases including MMP-1,³¹¹ MMP-15³¹² and MMP-17.³¹³ Reactive oxygen species also activate MMPs both directly as observed in cell-free environments and indirectly as observed in experimental neutrophil-rich environments.^314-316

Disruption of mucosal barrier during IRI allows penetration of pancreatic trypsin into the intestinal wall.³¹⁷ Trypsin is a strong activator of the MMP-9 pro-form that is released from neutrophils,^{317, 318} while MMP-9 and myeloperoxidase play important roles in the physiological turnover of the extracellular matrix through degradation and remodelling.³¹⁹ Overexpression of MMP-9 during acute IRI stimulates inflammation and tissue injury and increases intestinal villous loss, particularly when MMP-9 is upregulated by thiobarbituric acid reactive substances.³¹⁸ MMP-9 activity also enhances lipid peroxidation in the gastrointestinal tract and in other end-organs including the kidney.^{318, 320}

7.1 Current clinical approaches

Treatment of intestinal hypoxia and dysoxia in IRI generally focuses on physiological mechanisms, regulatory functions and abnormalities at all biological levels. Oxygen supply to the intestine depends on appropriate functioning of convective and diffusive components of the transport system and intestinal mitochondrial function. Currently, intestinal oxygenation is difficult to directly assess at bedside and is best approached by surrogates such as lactate, microcirculatory perfusion and carbon dioxide pressure gradients (Table 4).^{343, 344}

Intestinal ischaemia and reperfusion are often caused or accompanied by hypotension which would normally prompt fluid administration. However, available studies^345-348 and pathophysiological rationale suggest that excessive fluid administration can produce excessive intra-abdominal pressure^{349, 350} which can impair intestinal oxygen availability and therefore aggravate IRI. Current evidence is insufficient to recommend specific fluid volume strategies for patients with or at risk for increased intra-abdominal pressure.

The revised Starling equation indicates that trans-endothelial fluid movement is related to the plasma-sub-glycocalyx colloid osmotic pressure. Consequently, colloids such as albumin may delay trans-vascular fluid escape.³⁵¹ On the other hand, colloids may increase hydrostatic capillary pressure which can augment fluid filtration. Crystalloid solutions decrease colloid osmotic pressure and increase hydrostatic capillary pressure, theoretically leading to higher fluid filtration than colloids.³⁵¹ As with fluid volume, there is currently no evidence-based guidance for selecting crystalloid or colloid for IRI.

Abdominal perfusion pressure is estimated as the mean arterial pressure minus the intra-abdominal pressure which is usually 0–5 mmHg. Elevated intra-abdominal pressure reduces blood flow to the abdominal viscera including intestines. Clinical^352-354 and experimental^355-357 evidence suggest that a target abdominal perfusion pressure of at least 60 mmHg improves survival after intestinal hypoperfusion.^358-360

Based on the rationale provided by the Starling curves and Guyton theory of cardiac function,³⁶¹ elevated central venous pressure also increases splanchnic venous pressure and decreases intestinal perfusion and capillary exchange capacity. Central venous pressure is normally low, and even below zero when sitting quietly.³⁶² When intestinal perfusion is insufficient, a reasonable strategy is to first reduce central venous pressure which will also reduce splanchnic congestion and improve lymphatic flow.³⁶³ Thereafter, as necessary, arterial pressure can be increased.

It is equally important to consider that the veins of patients with severe hypovolemia are usually maximally constricted. Further activation of alpha-adrenergic receptors in veins with exogenous pure α-1 adrenergic agonists thus has little effect on venoconstriction, but does constrict arteries potentially leading to intestinal hypoperfusion and hypoxia.^364-369 Compared with α-1 adrenergic agonists, norepinephrine might exert an additional benefit by stimulating β-2 adrenoceptors which may facilitate emptying of the venous system and possibly enhance visceral blood flow. We note though that catecholamines may affect coagulation through pathways other than those involving adrenergic receptors³⁷⁰ and inhibit cellular respiration in a dose-dependent manner,^{371, 372} thus aggravating microvascular flow and cellular metabolism.⁴³ Reducing endogenous and exogenous adrenergic stimulation (decatecholaminisation) and using a multimodal vasopressor strategy may prevent or reduce the immune- and metabolism-modulating properties of exogenous catecholamines³⁷³ and ameliorate metabolic stress.³⁷⁴

Common treatment strategies based on the correction of macrohaemodynamic variables often do not account for downstream microcirculatory failure. Indeed, relying purely on macrohaemodynamic targets when trying to restore intestinal tissue oxygen delivery may be futile since haemodynamic coherence may be lost. Consequently, classical haemodynamic optimisation may fail to improve microvascular perfusion and even prove deleterious. Therefore, a microcirculation-guided resuscitation strategy might improve intestinal perfusion and haemodynamic coherence during IRI—recognising that we currently lack reliable ways of evaluating intestinal microcirculation in patients (Table 4). Vasodilators or more sophisticated strategies such as blood purification and antioxidants may also be effective.^375-377 But as with other aspects of fluid management for IRI, there is currently little evidence to suggest that any given approach is preferable. Importantly, macro- and micro-circulatory responsiveness to treatments should be simultaneously assessed to determine whether improved macrohaemodynamics improve intestinal capillary flow.

7.2 Experimental therapeutics and potential interventions

Various treatment approaches for post-IRI hypoxia and dysoxia have been assessed in research settings with mixed results. Oxygen microbubbles and other chemicals that off-load oxygen can be delivered intra-luminally.^{378, 379} However, there is little evidence of benefit³⁷⁸ while hyperoxia might induce vasoconstriction in the microcirculation, thereby worsening perfusion heterogeneity, and decrease endothelial cell viability and proliferation.^380-383

The mPTP is a calcium-dependent, ion non-selective membrane channel that mediates inner mitochondrial membrane permeability, allowing diffusion of molecules up to 1.5 kDa in size.³⁸⁴ Short-term (reversible) opening of the mPTP contributes to mitochondrial bioenergetics, protection from oxidative damage, enabling the efflux of calcium from the mitochondrial matrix, and cell signaling.³⁸⁵ Sustained (irreversible) mPTP opening causes mitochondrial swelling, which ruptures the outer mitochondrial membrane and induces processes leading to cell death. Intestinal IRI induces calcium overload in mitochondria and the production of reactive oxygen species, which causes the mPTP to open.^224-226 As a result, molecules with small molecular weights and H⁺ anions enter the mitochondrial matrix, dissipating the mitochondrial membrane potential, uncoupling the electron transport chain and inhibiting adenosine triphosphate synthesis,³⁸⁶ while water seeps into the organelles which causes them to swell and rupture.³⁸⁷

Experimental evidence suggests that tight serum glucose control may preserve mitochondrial function and attenuate organ dysfunction independently of organ perfusion.³⁸⁸ Metformin may help attenuate mPTP opening, stimulate mitochondrial biogenesis and reduce mitochondrial reactive oxygen species production. However, the drug can also cause metabolic acidosis with hyperlactatemia, vasoplegia and hypoglycemia.^{389, 390} Imeglimin inhibits mitochondrial permeability transition, possibly with less toxicity than metformin.^{389, 391} Animal studies also suggest that cyclosporine A inhibits the mPTP in different conditions, that is, sepsis and traumatic brain injury. Studies targeting restoration of cytochrome-c-oxidase (Complex IV) activity and maintenance of mitochondrial inner membrane integrity suggest that antioxidant treatments may preserve intestinal function by maintaining mitochondrial function.^392-394

An intriguing therapeutic approach is ischaemic pre- and post-conditioning which might enhance the HIF cascade and improve resistance to intestinal IRI. Specifically, remote ischaemic preconditioning, achieved through repeated and brief ligation of a limb prior to the induction of intestinal ischaemia, reportedly reduces microscopic damage to the intestine. Reduced damage correlated with reduced concentrations of myeloperoxidase and reduced HIF-1a expression, supporting the hypothesis that ischaemic preconditioning ameliorates intestinal IRI.^{395, 396} Ischaemic preconditioning induces upregulation of miR-21 through HIF-1α, inhibits apoptosis and leads to downregulation of the apoptotic mediators PDCD4 and Fas-L, thereby mitigating intestinal I/R injury.³⁹⁷

Ischaemic postconditioning is safe and clinically feasible and may improve the morphology and respiratory function of intestinal mucosal cell mitochondria and increase mitochondrial transmembrane potential.³⁸⁷ For example, three cycles of 5-min ischaemia/5-min reperfusion induced by a blood pressure cuff placed on the upper limp may attenuate intestinal injury in patients undergoing major non-cardiac surgery without any potential risk.^{398, 399} Experimental evidence suggests that ischaemic postconditioning alleviates intestinal mucosal injury and oxidative stress by regulating mPTP formation to ameliorate intestinal I/R injury.⁴⁰⁰ Additionally, ischaemic postconditioning reduces intestinal injury by inhibiting apoptosis of intestinal mucosal cells.^{401, 402}

Manipulation of intravascular volume as well as the coagulation profile in the setting of haemorrhage with fresh frozen plasma, modification of microvascular shear stress and use of novel vasoactive agents may be protective against IRI and IRI-associated hypoxia/dysoxia.^{43, 403-405} Downstream regulation of the hypoxia cascade by chelation of iron-catalysed reactive oxygen species has been attempted with the novel iron-chelator DIBI, with promising results.⁴⁰⁶ Also, melatonin, although having no effect on macrohaemodynamics variables, may exert anti-inflammatory and anti-oxidative action and attenuate the shock-induced decrease of microcirculatory oxygenation.⁴⁰⁷

Other experimental studies suggest that the IRI-induced vascular permeability is ameliorated by hypothermia during the ischaemia and reperfusion.⁴⁰⁸ Increased permeability following combined ischaemia and hypothermia was observed only when reperfusion was accompanied by rewarming.⁴⁰⁸ These data suggest that hypothermia may be a protective physiologic response to injury and that hypothermic reperfusion may limit capillary endothelial damage and oedema formation, possibly by decreasing the synthesis of chemical mediators and oxygen radicals. Furthermore, hypothermia has been shown to reduce the degradation of adenosine triphosphate to hypoxanthine and the rate of conversion of xanthine dehydrogenase to xanthine oxidase,⁴⁰⁹ thereby protecting against mucosal permeability and intestinal IRI. Finally, dexmedetomidine and other anaesthetics may modulate gene expression, channel activation, transmitter release, inflammatory processes and cell death, thus exerting protective effects in intestinal IRI.^410-412

Experimental studies suggest various intriguing approaches. However, we currently lack convincing clinical evidence to support any particular management strategy.