Microdialysis Monitoring of Biomarkers for Early Recognition of Intestinal Ischemia

studied using microdialysis sampling. A study in anaesthetised rats showed that of the stimulation of hepatic nerves result in glycogenesis via α -adrenergic mechanism and eicosanoids mediators (114). Other metabolites typically measured in the liver are glucose, lactate, pyruvate and glycerol. Dialysate concentrations of these analytes were measured in swine models during and after liver transplantation. The samples were collected and analysed at 20 minute intervals during both the donor operation and cold preservation, and 7 hours after reperfusion in the recipient (115). loop stoma decreased the probability of anastomosis leak (139). Another study distinguished a higher lactate/pyruvate ratio postoperatively in patients presenting anastomosis leakage before evidences of clinical symptom compared with those without (140).


Introduction
The intestinal tract is commonly involved in ischemia and reperfusion events, resulting from illness or surgical procedures that might lead to haemorrhage or shock. Ischemia is recognised as a poor blood supply and it is specifically known as intestinal or mesenteric ischemia when the bowel is involved. An ischemic insult can develop further into inflammation and injury. The mechanisms involved are still subject of study, but in general, an increase in intestinal cell membrane permeability provokes bacterial translocation and this in turn results in sepsis and multiple organ failure (MOF). The role of intestinal ischemia in septic shock will be discussed in this chapter as well as techniques commonly used to monitor gastrointestinal (GI) perfusion. However, most of these techniques depend on technical operators thus limiting the diagnosis of intestinal ischemia, which in most surgical departments still relies on clinical symptoms and examination.
Monitoring metabolic biomarkers has allowed not only the study of metabolic rates of tissues and organs, but the rapid detection of life-threatening events. During episodes of low blood flow, metabolic needs of the gastrointestinal tissue are not met. Monitoring this metabolic imbalance has the potential to provide an early diagnosis and prevent the injury from developing further into sepsis. Biosensors have been widely used to study tissue metabolism, proteins and nucleic acids with a wide range of applications. Biosensors as a diagnosis tool for gastrointestinal pathology will be revised in this chapter. The use of biosensors in clinical settings is, however, limited by issues such as biocompatibility, sterilisability and immunoreactivity and despite the extensive studies to overcome the immune-defensive reactions and toxicity effects due to the direct tissue implantation, other techniques have been used as an interface. Microdialysis is an extracting technique widely used to sample extracellular fluid of different human tissues. The use of microdialysis for the monitoring of digestive organs will be revised and intestinal microdialysis to monitor ischemia further described. Commonly off-line microdialysis pitfalls such as time lag and misplacement of samples are exposed as well as the use of recent technologies to overcome these issues. The synergetic combination of microdialysis and biosensors is the main reason for the position of this technology in the forefront of invasive monitoring. Biocompatibility issues of biosensors are extremely simplified when coupled with microdialysis and in turn biosensors confer microdialysis with a temporal resolution that it otherwise lacks. Although the combination of these techniques is still in research process, it presents the potential of close monitoring in clinical settings and home care scenarios for early diagnosis of conditions that can lead to sepsis if otherwise missed.

Intestinal ischemia and its role in septic shock
A poor blood supply to the intestine, also recognised as intestinal or mesenteric ischemia, generally leads to inflammation and injury (1) and easily develops into hypoxia (deprivation of oxygen supply) causing death of cells and tissue necrosis. This in turn can lead to sepsis and end in shock. Conversely, following haemorrhage or shock, the intestinal tract is the main organ to experience an ischemic/reperfusion injury. Intestinal cell membrane permeability can increase due to mesenteric ischemia (2), provoking bacterial translocation and injury of the gut barrier that in turn leads to sepsis and MOF (3)(4)(5). Hence, the measurement of intestinal permeability has been used as a valuable diagnosis of diseases affecting the bowel (6, 7).
The intestinal barrier protects the gastrointestinal tract, which under normal conditions is colonised by the bacterial flora containing enough microorganism, bacteria, and endotoxins to kill the host (8). Epithelial cells covering the surface of the gastrointestinal tract forms this barrier that prevents the absorption of toxins, antigens, proteases and microorganism across the intestinal wall (9). However, bacteria penetrates the intestinal barrier with relative ease and it is a common occurrence in healthy patients. This bacterial translocation can lead to further damage, worsening the health situation in ill patients (8), hence, it is of utmost importance to control this translocation in patients with critical conditions. Some studies have focused their efforts on understanding the function and mechanism of the intestinal barrier (10,11), however, these are unclear and the correlation between bacterial translocation and intestinal permeability has raised controversies. A correlation between the loss of intestinal barrier function and bacterial translocation was found out in rodents, but not in humans, where bacterial translocation could not be related to the increase of permeability of the atrophied villous (6). Seemingly, no correlation was found after major gastrointestinal surgery between failure of the gut barrier function and septic complication (11). Hence, some studies have proposed a primary mechanism of translocation in the absence of damaged mucosal barrier, where migration of organism across the bowel occurred by pinocytosis in epithelial cells (12)(13)(14). Nevertheless, other investigations have identified alterations in intestinal permeability in critically ill patients, where the loss of tight junctions and cells at the villous tip was suggested to be the primary cause of changes in intestinal permeability (15,16).
During the first stage of the bacterial translocation process, bacteria adhere onto the enterocytes and cross the barrier via transcellular and paracellular mechanisms (12,17,18). Disrupted epithelial tight junctions ('leaky gut') are the cause of bacterial translocation in the paracellular route (19,20). While, intracellular trafficking follows the endocytic uptake during the transcellular route (21) (Figure 1). Nevertheless, regardless of the type of mechanism, bacterial translocation is triggered by common injuries such as ischemia/reperfusion, oxidative stress and bacterial action. Typically, the immune system will attack the bacteria once across the barrier. However, if this fails, sepsis or endotoxemia occurs, and further damage will develop into MOF (8,22). In fact, the bowel is the main organ involved in MOF and this typically occurs as a consequence of a systemic inflammatory response syndrome (SIRS). Insults such as haemorrhage, ischemia/reperfusion, infection or trauma generally lead to an inflammatory response and this further develops into shock, if untreated. Toxins induce the release of cytokines, leukotrienes and platelets-activating factors (PAF), which play a major role in the initiation of shock (23)(24)(25)(26). Hence, bacterial and toxin translocation have been suggested to be the main cause of all these intestinal dysfunctions (27,28). The increase of membrane permeability has been related to bacterial translocation mechanisms and ischemic process in the intestine. In fact, translocation, vasoconstriction and hypoxic villi are common mechanisms occurring in the bowel due to the proximity of this to luminal bacteria and toxins (29). Reduction in blood flow is also a cause of initiation of shock, as well as activation of endotoxins, which release vasoactive substances and stimulate the formation of oxygen free radicals in the tissue (25,27,30,31). The reduction of blood flow means a loss of plasma volume and hence a reduction of proteins into the interstitium space, which in turn leads to MOF (32)(33)(34)(35)(36). The sequence of shock varies with different pathologies (low perfusion, hypotension, hemoconcentration, vasocongestion), but the cause of death initiated due to the inflammatory response is very similar (37).
Certain diseases and common surgical procedures can lead to the weakness of blood vessels that with the continuous passage of blood can dilate forming an aneurysm. As the aneurysm expands, the risk of rupture increases and this can produce severe haemorrhage leading to other complications and ultimately to death. The most common abdominal aneurysm occurs in the aorta, hence the term abdominal aortic aneurysm (AAA). Ischemic colitis is the most common complication after abdominal aortic aneurysm surgery (38,39) presenting a high impact in mortality rates (40)(41)(42)(43). Another high risk postoperative complication is the leakage of a low rectal anastomosis connection, or the site where the two transacted bowel segments are joined again, closing the bowel lumen (44). Anastomosis leaks can easily develop into sepsis increasing the mortality toll. Although other variables contribute to the risk of anastomosis leakage, ischemia is the main factor related to anastomosis leak (45)(46)(47). During cardiac surgery with cardiopulmonary bypass (CPB), there is a high risk of patients developing intestinal mucosal ischemia. This may further lead to a disturbed mucosal integrity and increased intestinal permeability due to imbalance between splanchnic oxygen supply and demand which may contribute to systemic inflammation (48). Hence, an early diagnosis and the evaluation of an adequate splanchnic perfusion are crucial to improve outcome rates following surgical procedures such as aneurysm repair, anastomosis connections and CPB.

Gastrointestinal perfusion monitoring
The principal factors in the development of sepsis, systemic inflammatory syndrome and MOF are intestinal ischemia and gut barrier failure. These are also the main causes for the development of anastomosis leak and present a high rate of recurrence after AAA repair. During inadequate blood flow, the metabolic needs of the tissue are not met, leading to tissue injury. Consequently, ensuring both perfusion pressure and blood flow to maintain the necessary metabolic supply and demand are highly important to avoid diseases (49). It is in situations when blood flow has been partially compromised, where an early diagnosis could prevent the injury from developing further into organ dysfunction and death. In order to monitor these perfusion irregularities, several techniques have been used over the years. Human GI perfusion has been monitored by means of tonometry (50)(51)(52), laser Doppler flowmetry (53,54), reflectance spectrophotometry (55), near-infrared spectroscopy (56), orthogonal polarisation spectral imaging, (57,58), indocyanine green clearance (59), and measurements of plasma D-lactate (60). While the outcome of patients with septic shock can be predicted, these techniques cannot be compared to each other since they measure different constituents of GI perfusion (61). Alternatively, intestinal bleeding, which is undisputably an indication of organ damage, has been used to diagnose intestinal failure, however this fails to quantify the damage. Absorption markers of increasing permeability have also been extensively studied to assess gut functions, however, as they are highly invasive and require extensive nursing time, they are not appropriate for clinical monitoring in the intensive care unit (ICU). Gastrointestinal tonometry, due to its simplicity, is currently the most commonly used technique. It provides adequate information of GI perfusion and it is highly suitable for use in ICU. The technique consists of a silicone balloon inserted in the stomach that automatically infuses and samples gas every 10 minutes. The gas sample is analysed automatically using an infrared sensor within the measurement instrument (62,63). However, it has several disadvantages as it does not provide a continuous measurement, presents long response time and leads to potential systematic errors as the gas samples are drawn from the catheter and analysed using an external measurement device. Furthermore, this method depends on the technical operator, compromising its reproducibility, and the effect of perfusion countercurrents changes on its sensitivity is still obscure (64). In addition, improvement of gastrointestinal perfusion has failed to prevent mucosal failure and otherwise, increasing the perfusion might have an effect on the metabolic rate, modifying the balance between oxygen demand and supply (61).
Currently, techniques most commonly used in clinical practices for the diagnosis of leaks are routine blood tests, colonoscopy, ultrasound and abdominal x-rays such as, CT scans, mesenteric angiography and MRIs and as a last resort, an exploratory abdominal surgery. Regardless of all advances in clinical technology, the monitoring of gastrointestinal perfusion is still limited and in most surgical departments, the diagnosis of intestinal ischemia relies on clinical symptoms and examination (65). However, subjectivity, the lack of precision and the delayed manifestation of the symptoms account for the unreliability of this procedure.
Research based on microcirculation and secretion during shock has been broadly explored at cellular and molecular level, however, there is a lack of information at the organ level (66). The mechanism of ischemia before shock and MOF begin, is not only an important area of research but one with plenty of room for exploration. Monitoring organ function and metabolism could be the ultimate mode of measuring gut perfusion (64), since there are evidences of the existence of localised biochemical changes in specific segments of the patient's gut (67).
Splanchnic hypoperfusion is generally associated with a poor outcome in critically ill patients. Moreover, despite presenting a hyperdynamic systemic circulation, the mucosa hypoperfusion in gastric, small intestine and colonic tissue commonly develops into sepsis (68). Gastrointestinal mucosa has been described as one of the highest metabolically active tissues under conditions of sepsis (69,70) and although no clear association has been found, gut flow and metabolic changes appear to be heterogeneous (71)(72)(73)(74). During compromised mucosal tissue, the reaction mechanism has been proposed as an increase of oxygen extraction (75). However, regardless of adequate oxygen delivery, the cells are unable to use oxygen during sepsis (76). During the initiation of intestinal ischemia, whatever the cause of injury (bowel strangulation, intestinal perforation or impaired blood flow), a rise in intracellular biomarkers occurs due to the increase of intestinal permeability (77). Therefore, metabolic levels are presented as useful measurements for the monitoring of intestinal ischemia.
During ischemia, a change from aerobic to anaerobic tissue metabolism occurs, which results in a decrease of glucose and pyruvate levels and increased lactate and glycerol levels. During anaerobic metabolism, the support of glucose from sphlanchnic circuit is inhibited causing pyruvate to accumulate in the tissue. Due to the lack of oxygen, pyruvate cannot be incorporated into the citric acid cycle, metabolising pyruvate to lactate. Monitoring the gut metabolic markers provides information on the variation from aerobic to anaerobic metabolism, the balance between oxygen supply and demand and therefore the ischemic degree of that tissue. In order to obtain a more accurate assessment of tissue metabolism, determination of lactate/pyruvate or lactate/glucose ratio is preferred compared to lactate alone. In addition, during sustained ischemia, a breakdown of the cellular plasma membrane occurs resulting in the release of phospholipids into the extracellular fluid, which are degraded to free fatty acids and glycerol. Hence, glycerol is another common metabolic product monitored during ischemic studies (78).

Biosensors as a diagnosis tool of gastrointestinal pathology
Over the past few decades, a large number of biosensors (a biological sensing element directly interfaced to a signal transducer) have been developed for the detection of various metabolites, proteins and nucleic acids with a wide range of applications, such as for medical diagnosis of gastrointestinal disorders.
Currently, the most common minimally invasive method to monitor the perfusion of the gastrointestinal tract in clinical settings is gastric tonometry. However, the technique only provides discrete measurements. In order to overcome this limitation, an optical fibre sensor was developed (79). This is based on the utilisation of a CO2 sensing layer fixed at the end of an optical fibre catheter that is connected to an optoelectronic unit. The sensor allows the continous monitoring of the rapid changes in the gastric pCO2 and hence provides a better understanding of gastrointestinal physiology ( Figure 2).
Monitoring tissue adenosine triphosphate (ATP), in addition to other metabolic products, provides a further understanding of the cellular respiration rate. Furthermore, ATP has an important function within the immune system as a clotting signal molecule (80) and it is thought to regulate serotonin (5-HT) release and hence gastrointestinal motility (81)(82)(83). An ATP microelectrode biosensor with high long-term stability and selectivity was used to study the regulation of mucosal release mechanisms such as 5-HT based on the long-term in vitro variations of ATP release from isolated ileum and colon tissues (84).
Others have developed a microcapillary immunoassay, the Quantum-dot-Linked Immunosorbent Assay (QLISA), in order to differentiate between inflammatory bowel diseases such as ulcerative colitis and Crohn's disease and irritable bowel syndrome (IBS). The fecal levels of the gastrointestinal inflammatory disease's biomarkers, myeloperoxidase and lactoferrin, could be quantified using this biosensor (85). In another study, a filterpaper-based strip was used to develop a bacterial whole-cell biosensor that allows optical on-site detection of a chemical signalling molecule in Gram-negative bacteria, which plays an important role in the pathogenesis of several gastrointestinal disorders (86). Differentiation of old blood and new blood for the surveillance of rebleeding occurrence was possible with an endoscopically implantable wireless biosensor designed to detect blood labelled with fluoroscein. This has the potential of immediate, real-time detection of upper gastrointestinal bleeding. Furthermore, the wireless signal can be transmitted to external computers that relay the data to the patient's cell phone or to an emergency response network allowing an early-warning remote surveillance system. In this way, the detection of rebleeding of endoscopically defined sources of GI hemorrhage during periods of high risk is possible (87) (Figure 3).
Studies with biosensors are still commonly performed in animal models or rely on collecting patients' fluids and tissue samples. Implantable biosensors still have problems associated with biocompatibility. Damage of the tissue surrounding the biosensor is believed to affect the measurement levels. In addition, implanted biosensors suffer the reaction of the tissue towards the foreign body, the so-called biofouling. A cascade signal is sent by the immune system that triggers the release of platelets and immune proteins as well as the formation of new blood vessels, which encapsulate the biosensor and hence impair the sensing layer (88)(89)(90). Therefore, alternative techniques have been explored to monitor gastrointestinal biomarkers.

Microdialysis monitoring
In the early seventies, Ungertstedt and Pycock (91) developed microdialysis, an extraction technique to measure interstitial substance in the brain without causing extensive tissue damage. The original concept for microdialysis was to mimic a blood capillary (92). Microdialysis is based on the passive diffusion of the molecules through a hollow fibre membrane (93). A sterile physiological solution (perfusate) is infused by a syringe pump at very low flow rates (typically 0.3-2 μl/min). Molecules present in the extracellular fluid diffuse to the lumen of the microdialysis (dialysate) and this is sampled through the outlet ( Figure 4). The membrane cut-off confers selectivity to the technique, limiting the size of molecules diffusing through it. In clinical situation, this prevents the passage of virus into the dialysate and it eliminates problems associated with instability of metabolites due to sample preparation. Since the dialysate is free of proteins, this can be directly injected into analytical systems for analysis.
Since its beginning, microdialysis applications have broadened to the sampling of extracellular fluid of different human tissues including ear fluid (94), brain (95,96), liver (97), heart (98), lungs, muscle (99) and bowel (78). The reader is directed to some of the reviews on microdialysis found in the literature (92,(100)(101)(102)(103). Extensive studies have been carried out using microdialysis in the digestive system organs. This section describes some of the work published in stomach tissue, liver, pancreas and intestine.
Gastric microdialysis has been used to continuously monitor gastrin release to determine the response to food, acid blockage and acute vagal excitation (104). Microdialysis monitoring of gastric ischemia during temporary celiac artery occlusion in anesthetised rats showed a more remarkable response to ischemia in the relative changes of metabolic ratios of lactate/pyruvate and lactate/glucose compared to changes in the flow indicator of H2O efflux and glycerol (105). Histamine release from enterochromaffin-like (ECL) cells during ischemia has been commonly sampled from the stomach submucosa with a single microdialysis probe (106)(107)(108)(109). However, submucosal microdialysis sampling has been suggested to be unrepresentative of histamine levels released from ECL cells, since these are located in the mucosa layer, requiring a diffusion of histamine from the mucosa to the submucosa and finally to the probe (110). A more accurate sampling of stomach analytes was introduced using a multiple probe approach in which the probes were implanted in the stomach lumen, mucosa, submucosa and in the blood of a rat (111) ( Figure 5). This fourprobe microdialysis sampling was used to directly compare drug absorption between gastric ulcerated and healthy tissue in the same animal, where the dialysate was analysed using high performance liquid chromatography ultraviolet (HPLC-UV). The study showed a higher drug concentration in ulcer tissue, which was a function of ulcer size and thickness and probe location within the tissues when compared to healthy tissue (112). The reconstructed oesophagus of patients undergoing resection of carcinoma was monitored with microdialysis to investigate the risk of postoperative complications caused by ischemia. This study showed lactate/glucose ratio to be the most reliable parameter and sets microdialysis as a promising method for examination of free jejunal flaps (113). Liver metabolism has also been studied using microdialysis sampling. A study in anaesthetised rats showed that of the stimulation of hepatic nerves result in glycogenesis via α-adrenergic mechanism and eicosanoids mediators (114). Other metabolites typically measured in the liver are glucose, lactate, pyruvate and glycerol. Dialysate concentrations of these analytes were measured in swine models during and after liver transplantation. The samples were collected and analysed at 20 minute intervals during both the donor operation and cold preservation, and 7 hours after reperfusion in the recipient (115). This led to the monitoring of metabolic changes in liver grafts in a clinical pilot study with 10 patients (116). Metabolic products were also monitored during ischemia-reperfusion injury in the rat liver using microdialysis and commercially available bedside kits (117). This study proved that microdialysis is applicable for human liver surgery by continuous intraoperative monitoring of intrahepatic metabolism (118). Alternatively, liver slices and microsomes have been used to carry out in vitro microdialysis studies of hypoxia (119), steroid enzymes activity (120) and drug metabolic processes (121).
As microdialysis collects only the unbound fraction (the pharmacologically active fraction) of the drug, it eliminates the need for unnecessary extensive sample preparation and enzyme degradation processes. This is highly advantageous during pharmacokinetic studies, such as the intraperitoneal monitoring of a drug active concentration in freely moving rodents (122). Other studies have used microdialysis to investigate the pathophysiology of brain activity among irritable bowel syndrome (IBS) patients and braingut interactions. In vivo brain microdialysis monitoring during colorectal distention in rats showed that noradrenaline is released in the hippocampus during the distention thus suggesting a possible correlation with behavioural changes (123).

Intestinal microdialysis
Microdialysis possesses the advantage of its applicability to other organs in which needle puncturing is possible including the stomach and intestines. On the other hand, the puncturing performed by the probe implantation, as seen in Figure 5, can inflict an acute inflammatory response that may influence the measurement levels and interpretation (124). A common solution has been to establish equilibration periods after probe implantation. Nevertheless, there is a critical need for smaller microdialysis probes to reduce the damage in surrounding tissue and local blood supply. This trauma caused by the implantation of the microdialysis probe is one of the main controversial issues when deciding the probe placement in gastrointestinal tissue.
Intraluminal probe placement could be a good approach, because the injury that occurs in the bowel causes a bigger damage to the mucosa than to the seromuscular layer, and therefore, biomarkers in the lumen rise early during the injury period (125). However, intraluminal metabolites measurements have been found to be extremely low and the detection is not feasible due to the resting phase of the intestine and to the high volume of contents in the lumen (126). A common approach is to place the microdialysis probe in the human peritoneal cavity (127)(128)(129). The assumption is that due to the high amount of intestinal anaerobic products metabolised by the liver, the metabolic markers of an impaired circulation are greater in the peritoneum than in blood (77). However, the ischemic biomarkers are diluted by non-ischemic markers and by the systemic circulation supplied to the peritoneum (78). Sommer et al. compared results from microdialysis probes inserted intraperitoneally, intramurally and intraluminally in the bowel of a swine model (126) (Figure 6). Insertion of the probe intramurally provides faster detection of metabolic changes than intraluminal and peritoneal microdialysis due to the proximity of the probe to the damaged tissue and the lack of dilution artifacts (130). However, this is more invasive and presents a difficult challenge for the clinician thus requiring intensive training. A recent approach has been to evaluate the insertion of the microdialysis probe in pancreatic tissue guided by an endoscopic ultrasound device (131), which has the potential to be used in other tissues. In addition, the microdialysis probe has a fragile tip, which can be broken during the probe implantation. It is increasingly recognised the need for the production of less fragile probes and alternative geometries.
Microdialysis monitoring can be divided into off-line and on-line microdialysis depending on the coupling method between the microdialysis outlet and the detection technique.

Off-line Microdialysis
Several studies have used in vivo microdialysis monitoring in animal models to investigate splanchnic metabolic disorders by measuring glucose, lactate/pyruvate ratio and glycerol both intraluminally (132)(133)(134) and in the intraperitoneal space (128). From these studies, microdialysis was suggested as a valuable tool for surgeons to detect early signs of mesenteric ischemia and the postoperative complications associated with this. In another study, microdialysis was used to monitor several episodes of ischemia in the pig intestine. The results showed lower dialysate levels of lactate during the second ischemic event compared with the first. This suggested that extended periods of ischemia triggered a mechanism of protection against hyperpermeability for later ischemic events (135).
Few have translated this technology to human subjects. Some pilot studies have used microdialysis for intraoperative monitoring. Intestinal luminal microdialysis and tonometry was used to monitor the rectal mucosa of patients undergoing elective cardiac surgery with cardiopulmonary bypass. This combination allowed the monitoring of both circulation and metabolism to indicate the adequacy of splanchnic perfusion in the colon (136). Others have used blood, urine and interstitial fluid microdialysis samples to investigate the phamacokinetic effect of antibiotic concentrations at the sites of infection during AAA open repair surgery (137). But, in general, studies in patients are carried out postoperatively. The safety of intraperitoneal microdialysis was evaluated to monitor metabolic and inflammatory changes in infants after surgery for necrotising enterocolitis (138). Peritoneal microdialysis was used postoperatively to assess the anastomosis leak rate in patients undergoing a low rectal resection due to cancer. In a medical trial for defunction stoma procedure, where patients were randomised, results showed that showed that defunctioning loop stoma decreased the probability of anastomosis leak (139). Another study distinguished a higher lactate/pyruvate ratio postoperatively in patients presenting anastomosis leakage before evidences of clinical symptom compared with those without (140).
From these clinical studies, scientists have agreed on the advantages of microdialysis as an early indication of intraoperative and postoperative complications compared with customary devices for the monitoring of splanchnic circulation. However, most of the microdialysis studies presented here used either bulky laboratory techniques such as liquid chromatography and electrophoresis or the commercially available bedside kit analyser from CMA/Microdialysis (141). Since the detection is off-line, samples are required to be collected in vials and manually stored in ice by a trained technician or the ward nurse. This time lag and the problem associated with sample misplacement are serious pitfalls of offline microdialysis.

On-line Microdialysis
In order to have an early diagnosis, a tight monitoring is required. Metabolic events are rapid so a continuous monitoring is necessary. In recent years, compact instrumentation and microfluidic devices are being coupled online with microdialysis probes to overcome the temporal resolution of microdialysis. Lab-on-chip technology, such as microchip electrophoresis and liquid chromatography (142,143) and microfluidic devices that increase temporal resolution by creating segmented nanolitres dialysate samples (144) are increasingly being investigated. However, the synergetic combination of microdialysis and biosensors has put this technology to the forefront of invasive monitoring. Sensitivity and biocompatibility of the biosensors increased when coupled to microdialysis and in turn biosensors enhance the temporal resolution of microdialysis to the millisecond scale (145,146). This technology plays an important role in both point-of-care testing and home care (147).
An on-line rapid sampling microdialysis biosensor system has been used to monitor ischemia at the bowel level. The system couples selectively glucose and lactate electrochemical biosensors on-line with the microdialysis probe using a flow injection analysis (FIA) system. This allows in vivo monitoring of metabolic substrates changes in the colon at a high time resolution, every 30 seconds, without the need for extensive manipulation, running by itself during 24 hours up to 5 days (Figure 7).
Rapid sampling microdialysis was used to monitor rapid changes of glucose and lactate levels in patients undergoing an elective colectomy, most typically due to cancer. The microdialysis probe was inserted in the seromuscular layer of the colon and sutured to ensure fixation and allow the surgeon to proceed with the resection. A stabilisation period of 10-15 minutes was allowed before the transection of the main feeding artery, where glucose and lactate levels decrease and increase, respectively. These metabolic changes were not immediate as expected, but after an average interval of 12 minutes, attributed to the colon collateral flow, however, these agreed with the mechanism of blood flow reduction  (158). A preliminary work in animal models was carried out to understand the mechanism of ischemia in compromised bowel, such as that after an anastomosis procedure. Here, the microdialysis was also tunnelled in the seromuscular layer, but in parallel with the site of the anastomosis construction, within a distance of a few millimeters. In this case, glucose and lactate changes occurred almost immediately after the feeding artery transection, which reveals that the tissue cannot rely on the same collateral flow once it has been compromised (by the resection previous to the anastomosis) (157). This has been further illustrated when comparing both data, where a therapeutical window is clearly observed in healthy patients but not in compromised tissue (159) (Figure 8).
The system was recognised as a potential candidate for monitoring early diagnosis of ischemia after AAA repair surgery. A set of data was obtained during the monitoring of aneurysm repair elective surgery patients for up to 2 days in ICU. The probe was implanted in the bowel of seven subjects in the mesenteric border of the sigmoid colon just at the junction of the mesentery with the colon. Although changes in plasma values were observed, dialysate levels for both glucose and lactate were steady. This confirms the fact that microdialysis levels are a good indicator of local changes, but it does not reflect systemic conditions. The lactate/glucose ratio was observed to be constant for all patients for up to 2 days after probe implantation, which strongly indicates the lack of acute ischemic events. Some episodes of transient metabolic ischemia were observed due to a compromised blood supply, which allows for pattern recognition during further studies (160).

Conclusion and future trends
Mesenteric ischemia has been known to increase intestinal cell membrane permeability, provoking bacterial translocation and ultimately leading to sepsis and multiple organ failure. An adequate splanchnic perfusion and evaluation of intestinal permeability are crucial to improve outcome rates following surgical procedures. However, current techniques for the diagnosis of leaks are unreliable due to the subjectivity, lack of precision and the delayed manifestation of the symptoms.
Microdialysis has been used to monitor biomarkers in a range of tissues and organs. Clinical studies using intestinal microdialysis have agreed on the advantages of the technique as an early indication of intraoperative and postoperative complications compared with common monitoring of splanchnic circulation. However, off-line detection carries a limitation for the early diagnosis of ischemic insults, since metabolic events are rapid. On-line rapid microdialysis technology takes advantage of the synergetic combination of microdialysis and biosensors, which plays an important role in both point-of-care testing and home care.
Microdialysis has recently been evaluated as a tool to assess the metabolic changes during liver resection, stomach ischemia and for the diagnosis of novel biomarkers of the pancreas that are undetectable in plasma. Preliminary work has been done with microdialysis to monitor key metabolites during and after organ transplantation. Although still in early stage, this presents the potential of monitoring the organ's condition during the transportation. This will allow not only the determination of the suitability of the organ for transplantation, but will also decrease the possibility of occurrence of an ischemic insult by perfusing the organ when it reaches a certain predetermined threshold.
Among some limitations of microdialysis is the trauma caused by the implantation of the probe, which causes an inflammatory response and compromises the results of the investigation. A common solution has been to establish equilibration periods after probe implantation and to develop miniature probes. Since smaller sized probes present high back-pressure, a droplet microdialysis probe has been designed to overcome this issue (161). Alternatively, flat microdialysis geometries can be envisioned, for those tissues where tunnelling of the probe is not necessary. Temporal resolution also limits microdialysis use, hence microfluidic and lab-on-chip devices have been investigated to directly connect the microdialysis with biosensors (162)(163)(164). As the microfabrication technology advances, more chips and miniaturised devices will be seen such as micropumps to be implanted in line with the microdialysis inlet and wireless detection systems embedded in the probe design, reducing in this way bulky instrumentation, otherwise required. One of the paramount targets of biomedical research is to be able to create close-loop systems that combine sensing devices with drug delivery systems. Microdialysis probes are currently used for drug delivery and fluid sampling for diagnosis. Hence, it can be envisioned a device where microdialysis is coupled with biosensors for diagnosis and depending on the levels recorded, the mechanism switches to a reverse mode delivering the drug required.
Most microdialysis publications focus on the monitoring of neurochemicals, however, general reviews and book chapters are helpful to understand the fundamentals of the technique and learn how to handle the probes and perform calibration experiments (165)(166)(167). Originally, microdialysis probes were fabricated from dialysis fibres connected to an inlet and outlet tubing. With the expansion of microdialysis, few major trades are handling the commercialisation of microdialysis products, producing CE approved sterile probes for human use (141) and for in vitro and animal studies (168,169).

Author details
Radha Swathe Priya M. and Emma P. Córcoles Faculty of Health Science and Biomedical Engineering, Universiti Teknologi Malaysia, Skudai, Malaysia