INTRAVITAL INTESTINAL VIDEOMICROSCOPY: TECHNIQUES AND EXPERIENCES PAUL J. MATHESON, Ph.D.,1 and R. NEAL GARRISON, M.D.1,2*
Intravital videomicroscopy (IVM) of the gastrointestinal (GI) tract is a sophisticated and powerful technique to directly observe the neurologically intact microvasculature of rats in naive and pathological conditions. We combine IVM with other techniques (i.e., vascular ring tension analysis and colorimetric microsphere determination of whole organ blood flow) to develop a strategy for the systematic analysis of the regulation of GI blood flow in healthy animals and in models of systemic sepsis and resuscitated hemorrhagic shock. We also study the molecular biology of the GI tract (enzyme- or radio-linked immunosorbent assays, fluorescent Greiss assay, and immunoblots) to correlate expression and levels of vascular mediators in tissue and arterial, venous, and portal blood with functional activity of the GI microvascular tree. When combined, these techniques develop a picture of gut pathophysiology at the level of the endothelium, vascular smooth muscle cells, and blood cells in the microcirculation. Our work led us to the general hypothesis that altered microcirculatory function in disease states lies primarily at the level of the interface between vascular and tissue physiology, i.e., the endothelial cell. This review focuses on methods and techniques for studying microvascular function, and concludes with focused reviews of pertinent findings. ª 2005 Wiley-Liss, Inc. Microsurgery 25:247ÿ257, 2005.
The
health of the gastrointestinal (GI) tract is prerequisite for the health of the individual. Numerous studies linked altered or impaired function in the microcirculation of the intestinal tract to pathophysiological conditions that can spiral into disease states like multiple system organ failure and death.1 Intravital videomicroscopy (IVM) of the GI tract is a sophisticated and powerful technique that allows for direct observation and study of the intestinal microvasculature in naive and pathological states. Regulation of intestinal blood flow is complex, as outlined in Table 1, which shows some of the proposed mediators of intestinal microvascular tone in the rat.2 Over the last 20+ years, we utilized several techniques such as vascular ring tension measurements, IVM, and colorimetric microsphere determination of whole organ blood flow to examine intestinal microvascular function in healthy Sprague-Dawley rats as well as in rat models of systemic sepsis and resuscitated hemorrhagic shock. We correlated these microvascular studies with molecular biology techniques such as enzyme- or radio-linked immunosorbent assays (ELISA and RIA, respectively), fluorescent Greiss assay, and various immunoblot techniques to measure the vascular mediators in tissue and in arterial, venous, and portal blood.3,4 When combined,
1 Department of Surgery, University of Louisville School of Medicine, Louisville, KY 2 Louisville Veterans Aairs Medical Center, Louisville, KY 40206 Grant sponsor: VA Merit Review Funding; Grant sponsor: Commonwealth of Kentucky Center for Excellence in Applied Microcirculatory Research; Grant sponsor: Ohio Valley Affiliate, American Heart Association. *Correspondence to: R. Neal Garrison, M.D., Department of Surgery, University of Louisville, Ambulatory Care Building, 2nd Floor, 550 S. Jackson St., Louisville, KY 40292. E-mail:
[email protected] Received 20 November 2004; Accepted 30 November 2004 Published online 2 June 2005 in Wiley InterScience (www.interscience. wiley.com). DOI: 10.1002/micr.20120
ª 2005 Wiley-Liss, Inc.
these techniques develop a picture of the pathophysiology of the gut at the level of the endothelium, vascular smooth muscle cells, and blood cells in the microcirculation. Our studies and the work of many other investigators led us to the generalized hypothesis that altered microcirculatory function in disease states lies primarily at the level of the interface between vascular and tissue physiology, specifically in deranged endothelial cell function, as shown in Figure 1.5 This review focuses on methods and techniques for studying microvascular function, and concludes with focused reviews of pertinent findings in septic shock and resuscitated hemorrhagic shock from our laboratory.
OVERVIEW OF IVM STRATEGY
In general, our laboratory strategy over the past 20+ years has revolved around the technique of IVM of the GI tract. However, IVM techniques are difficult to learn and time-consuming to perform, with each experiment requiring 4ÿ8 h or more of bench time, depending on the specific pathophysiological model and protocol, plus an equal or greater time commitment for data mining, analysis, graphical representation, and interpretation. Since each experiment often requires 4ÿ6 groups of 8ÿ10 animals each in order to achieve the appropriate statistical power, the time commitment for an investigator to perform an IVM experiment is usually many months. Frequently, the results of IVM experiments are unexpected and/or difficult to interpret. We utilize a system of analysis of IVM data that looks at gross measurements of function like central hemodynamic performance and vascular diameters and flow, as well as calculated measures such as dilator or constrictor
248
Matheson and Garrison
Table 1. Potential Vasoactive Mediators of Enteric Circulation Constrictors Neural mediators Increased sympathetic tone (adrenergic) Decreased parasympathetic tone (cholinergic)? Neuropeptide Y
Circulating humoral mediators Catecholamines (except in liver and muscle) Angiotensin II Vasopressin Serotonin Activated complement (C5a) Circulating paracrine and autocrine mediators Endothelin-1 Platelet-activating factor Constrictor Prostaglandins (F2a)
Dilators Decreased sympathetic tone Increased parasympathetic tone? Substance P Vasoactive intestinal polypeptide (VIP) Calcitonin-gene-related peptide (CGRPa) Catecholamines (only in liver and muscle) Histamine Bradykinin Activated complement (C3a, C5a) Adrenomedullin
Nitric oxide (NO; EDRF) Endothelial-derived hyperpolarizing factor (EDHF) Dilator Prostaglandins (I2 or prostacyclin)
‘‘Metabolic’’ vasodilators Decreased PO2 Increased PCO2 Decreased pH Increased metabolites (K+, lactate, adenosine)
capacity and dose of mediator that gives 50% of the maximal response, i.e., dose-response curves. Since IVM studies are time-consuming and difficult to perform well, we utilize techniques that increase the expected ‘‘payoff’’ for each proposed series of IVM experiments. This led to the strategy of screening potential studies by first performing routine ex vivo aortic ring or superior mesenteric artery (SMA) ring studies to assess mediator potency, effect, and derangement in disease states. These studies can be followed with in vivo microsphere blood flow experiments, using well-established colorimetric microsphere techniques to measure whole-organ blood flow responses in virtually any organ system. Microsphere blood flow studies allow us to visualize ‘‘snapshots’’ in time of the distribution of systemic blood flow to the GI and other organ systems. We also desire to maximize our interpretive ability with IVM studies, so we often combine IVM with molecular biology techniques in identically treated animals to understand more about the underlying physiology of microvascular mediators that control microvascular flow dynamics at a cellular or subcellular level. Our current strategy then is to screen proposed studies with aortic or SMA ring measurements, followed by micro-
sphere assessment of organ blood flow, and then to perform IVM and molecular biology studies of potential ‘‘high-yield’’ hypotheses based on the preliminary ring tension and microsphere results.
ISOLATED VASCULAR RING STUDIES
Furchgott and Bhadrakam first measured aortic ring tension in 1950 to show that epinephrine caused constriction and nitrites produced dilation of rabbit aortae.6 The basic experimental setup for the measurement of arterial wall force generation in isolated vessel segments involves placing two stainless steel wires through the lumen of the vessel, one of which is anchored in a reservoir containing physiological salt solution, while the second is attached to a sensitive force transducer. In placing the wires through the lumen of the vessel, great care is needed to avoid damage to the endothelium; such damage can be unmasked by observing the response to topical acetylcholine placement. Healthy dilation suggests an intact endothelium, while no response or constriction reveals damaged or absent endothelial cells. The position of both the reservoir and the force transducer can be adjusted to align the plane of force across the vessel and to set the resting wall tension in the calibrated transducer. Traditionally, silk was used to tie the two wire segments to the bath and the transducer. Vankan et al.7 described the mathematical modeling of force in the vascular segment when mounted in the experimental setup. Figure 2 diagrams our current setup (using DigiMed Tissue Force Analyzers, MicroMed, Inc., Louisville, KY), which allows us to measure changes in force in the nanogram range, and to measure force generation in much smaller blood vessels than the rat aorta. The current limitation is the skill of the investigator at dissecting the blood vessel, placing the stainless steel wires in the lumen of the blood vessel, and mounting the vessel in the tissue bath. In our laboratory, we routinely measure force responses in rings of superior mesenteric arteries from rats. In the 50 years since those first experiments, numerous investigators have used and improved upon the vascular ring technique. We modified the technique of others8 extensively in the past 10 years. We currently use the aortae of 200-g Sprague-Dawley rats that are cut into 1.5-mm segments.9 This approach allows us to perform six different bath protocols for each animal, such that each rat serves as its own pharmacological control. The vascular response of the distal vs. proximal segments of rat aortae varies slightly, which necessitates the rotation of segments from each animal to avoid proximal vs. distal position error in the experimental results.
Intravital Intestinal Videomicroscopy
249
Figure 1. Overview diagram of central role of vascular endothelial cell in microvascular regulation of intestinal blood flow.
In the 1990s, Touiche Kawabe in our research laboratory (personal communication) found that using silk ties to anchor the aortic segment to the force transducer resulted in a stretch artifact that was dependent on the relative humidity of the silk, which is a significant source of experimental error, since the upper silk tie is partially submerged in physiological saline solution, partially at the air-surface interface, and partially in room air. Thus, the humidity artifact with silk ties varies with each individually mounted vascular ring as well as with slight changes in the contraction/relaxation state of the vessel with pharmacologic manipulation. Stainless steel wire is now the standard for force transducer anchors, since the tensile characteristics of stainless steel completely eliminate stretch artifacts, and they are easy to make and use. Since the transition to stainless steel, the importance of the magnitude of preload stretch of vessels has been unmasked as being of paramount importance. The preload stretch, as determined by the weight placed on the vessel at the outset of the protocol, can be considered to correlate with the hydrostatic pressure in the blood vessel in situ. We measured vascular responses to a variety of vasoactive mediators at incremental preloads in vessels hung on stainless steel anchors, which demonstrated that the response of the blood vessel, in both amplitude and direction (constriction vs. dilation), directly correlates with preload or vessel stretch. As yet unpublished and
ongoing mathematical modeling calculations suggest that the ‘‘physiological’’ preload in our experimental model needs to be in the 8ÿ10-g range to simulate blood pressures of 80ÿ120 mmHg. However, most studies in the literature to date reported preload values of only 0.5ÿ2 g, which, according to our mathematical models, correspond to mean arterial blood pressures below 60 mmHg. COLORIMETRIC MICROSPHERES
Another technique that we use as a method of experimental ‘‘screening’’ is the colorimetric microsphere technique, which provides a measure of wholeorgan blood flow. These studies utilize 15-lm-diameter polystyrene beads that are impregnated with one of several color dyes, as previously described by Hakkinen et al.10 In this technique, we cannulate the right carotid artery and place the tip of a polyethylene catheter into the left ventricle for microsphere bead solution injection. We inject about 1010 beads in 1 ml of saline at a rate of 0.6 ml/min. We also cannulate the left femoral artery to withdraw 0.7 ml of blood at a rate of 0.82 ml/min, which encompasses approximately two cardiac cycles during the microsphere injection phase. This arterial blood withdrawal at a known rate is used to standardize the unknown blood flow in organs of interest. Figure 3 outlines this phantom-organ technique.11
250
Matheson and Garrison
Figure 2. Schematic depiction of aortic ring setup utilized in investigators’ laboratory. This setup allows for two rats to be run concurrently with six segments from each aorta, which permits investigators to utilize completely paired protocols, with all animals serving as their own controls.
The colorimetric microsphere technique provides a ‘‘snapshot’’ of blood flow in any organ that we choose to monitor, and we are able to perform three microsphere injections per animal in protocols that can last up to about 8 h. Typically, we monitor blood flow in the lungs, liver, both kidneys, stomach, small intestines, colon, spleen, and various muscles. We compare blood flow per gram of tissue in both kidneys as a measure of assay integrity and animal viability. If blood flow per gram of weight between kidneys is within 10% of the other kidney, then the animal and microsphere assays are considered viable. A greater than 10% variability in renal blood flow between left and right kidneys results in the animal not being included in the experimental results due to probable loss of beads and/or dye during the bead recovery and assay procedure. Less than 1% of animals have been excluded from experimental results due to renal blood flow mismatch.
INTRAVITAL VIDEOMICROSCOPY
The procedure that we perform that provides by far the most information about the regulation and status of blood flow to and within an organ is intravital videomicroscopy. Over the years, we have used this technique to examine microvascular blood flow in a number of tissue beds, including the cremaster muscle (skeletal muscle), the terminal ileum, the midjejunum, the liver, the kidney, and the lung of rats. In general, the tissue being studied can last up to 4ÿ5 h if it is placed in a pHand temperature-controlled bath with physiological salt solution (PSS),12 or about 8ÿ10 h if the tissue is suffused with PSS under a covering of saran polyvinylidene chloride (Saran Wrap).3 The composition of PSS, which is shown in Table 2, is standard Krebs solution for skeletal muscle, lung, and liver preparations, and ‘‘no glucose’’ Krebs solution for
Intravital Intestinal Videomicroscopy
251
Figure 3. Schematic depiction of instrumentation of Sprague-Dawley rat for measurement of whole-organ blood flow by colorimetric microsphere technique. This technique provides ‘‘snapshot’’ in time of blood flow distribution across many organ systems simultaneously. CA, carotid artery; FA, femoral artery; FV, femoral vein; HR, heart rate; LV, left ventricle; MAP, mean arterial pressure.
examination of kidney and intestinal preparations, since glucose is a very powerful vasodilator in these tissues. The PSS Krebs solutions are made fresh daily and are maintained at a temperature of 37.0 ± 0.5°C by a heating coil with a feedback controller at a pH of 7.40 ± 0.05, by bubbling the solutions with a mixture of N2 and CO2 gas. As already mentioned, these solutions can be used to bathe the tissue in a small tissue bath (volume, 50 or 60 ml), or can be kept in a reservoir that slowly suffuses by gravity across the tissue under Saran Wrap to maintain the physiological milieu. The two techniques have different strengths and weaknesses. The bath technique allows for easy monitoring of PSS and tissue conditions and also for easy introduction of topical pharmacological agents, but the tissue usually begins to degrade after 5ÿ6 h in the bath. On the other hand, the suffusion technique provides greater tissue stability and very clear, sharp images, but it is more difficult to introduce well-mixed pharmacological agents onto the tissue, and classically it was more difficult to monitor tissue conditions. Recent technological advances introduced commercially available, accurate, and small pH monitors that can be used in conjunction with the suffusion technique. Once the surgical preparation is complete and the animal is placed on the microscope stage, the image is projected through the microscope onto a digital camera that is connected to a personal computer and video storage device such as a Digital Video Disc (DVD) recorder or Digital Video Recorder (DVR) (Fig. 4). The image from the microscope can be time-stamped and stored on videotape or on recordable CD or DVD discs. The image can be annotated with information such as animal number, date, and protocol prior to storage and/or image-processing. We use a variety of Matrox digital imaging boards along with ImageOne
and other software for image analysis of blood vessel diameter, vasomotion, gray or color scale to assess vascular leak of fluorescently labeled albumin, or other parameters of interest. Simple calipers can also be manually compared to the video image on the monitor so that vessel diameter can be measured directly by the investigator as a form of quality control. This method is slow but very accurate, especially when multiple readings (15 per field) are averaged of each vessel image. Typically, we also pass the image through an optical Doppler velocimeter prior to the video camera on the microscope, to measure center-line red blood cell velocity in some of the larger blood vessels such as the inflow A1 arteriole of the cremaster muscle or terminal ileum. In general, each animal provides information on central hemodynamic status such as mean arterial blood pressure, heart rate, and cardiac output, as well as microvascular data such as inflow arteriolar blood flow, blood vessel diameter in a wide range of vessel types, and leak from postcapillary venules using fluorescentlabeled FITC-albumin. This provides a detailed picture of the macrovascular and microvascular status of the animal during a variety of physiological and pathophysiological conditions as well as during exposure to pharmacological agents. Often we examine the microvascular responses to incremental doses of a pharmacological agent (usually after a loading dose to ‘‘prime’’ the receptors present in the vessels) and then plot the diameters against dose to obtain dose-response curves, as shown in Figure 5.13,14 These dose-response curves provide data on the maximal response of the microvasculasture as well as information such as the dose that provides 50% of maximum response (EC50). These data can be compared between groups of animals to determine pathophysiological shifts in sensitivity and reac-
252
Matheson and Garrison
Table 2. Components of Physiological Salt Solutions for Tissue Perfusion* Solution
NaCl
KCl
KH2PO4
MgSO4
NaHCO3
CaCl2Æ2H2O
Glucose
Krebs ‘‘No glucose’’ Krebs Tris-saline Tris-glucose
118.0 112.9 94.11 88.98
4.7 4.7 5.87 5.87
1.2
1.2
25.5 25.5
2.55 2.55 2.55 2.55
11.10
11.10
TrisÆHCl
TrisÆbase
36.29 36.29
13.71 13.71
*All concentrations are in mM.
tivity to a given controlling agent over time or in response to a given condition such as feeding, sepsis, or resuscitated hemorrhagic shock. FOCUSED REVIEWS OF PATHOPHYSIOLOGICAL ALTERATIONS IN THE CONTROL OF INTESTINAL BLOOD FLOW
A number of pathophysiological conditions exist which alter intestinal macro- and microcirculatory regulatory mechanisms, such that alterations in intestinal blood flow could contribute to the genesis or maintenance of these disease processes.2 The mechanisms that produce these conditions can be difficult to elucidate because of the intricately woven involvement of many physiological systems and mediators, with multiple effects in different organ systems. The remainder of this review will summarize the pathological alterations associated with regulation of microvascular tone under conditions of septic shock and hemorrhagic shock that we have focused our research on over the past 20+ years. Shock is a pathophysiological state in which the circulatory system is unable to perfuse tissues and meet oxygen demand adequately. The ‘‘two-hit’’ hypothesis is one theory that evolved to describe the pathogenesis of multiple organ dysfunction syndrome (MODS) (Fig. 6), which suggests that uncorrected subacute shock events lead to end-organ hypoperfusion and deranged function, permanent organ damage, and finally organ failure. Reversal of shock can occur in one of two ways: physiological compensation or medical intervention.15 We examined both physiological responses in the microvasculature to shock states and pharmacological manipulation of the microvasculature during shock states. This review will focus on those investigations. REVIEW OF MICROCIRCULATION INVOLVEMENT DURING SEPTIC SHOCK
Despite decades of basic science and clinical research, septic shock with subsequent multiple system organ failure continues to be a significant cause of morbidity and mortality in the surgical intensive care unit.16 Hemodynamically, septic shock in humans is
characterized by an early hyperdynamic, high cardiac output phase and a delayed hypodynamic, low cardiac output phase.17 Despite appropriate treatment and care, sepsis can lead to MODS and death,17,18 which led to the implication of the gut as the motor or generator of MODS The gut experiences a combination of immune alterations19 and deranged blood flow during systemic inflammation. When mucosal blood flow decreases below a certain critical level required for cellular viability, free radical generation, cytokine production, and perhaps the loss of tissue integrity and mucosal barrier function can occur.20 In early, hyperdynamic states, mucosal vascular perfusion seems to be compromised in spite of decreased vascular resistance and hyperdynamic cardiac performance.21 This deranged intestinal function is thought to be due to myriad microcirculatory changes that occur during sepsis,1 which include altered adrenergic microvascular responses, deranged blood flow distribution, impaired coagulability, diminished red blood cell deformability, increased microvascular permeability, subsequent tissue edema, and impaired endothelial cell function. At some point in the progression of septic response, the hyperdynamic phase gives way to a period of decreased cardiac performance and vasoconstriction. Adrenomedullin, a potent vasodilatory peptide derived from endothelial cells, is thought to play a significant role in the development of both increased cardiac output and decreased total peripheral resistance during the early hyperdynamic phase of septic shock,22ÿ24 and the loss of adrenomedullin over time might contribute to the onset of the hypodynamic phase of shock. Cecal ligation and puncture (CLP) in the rat24 increased adrenomedullin in the serum at 2 h and was associated with elevated levels of adrenomedullin mRNA in the small intestine, left ventricle, and thoracic aorta.22,24 The time course observed in this expression pattern correlates well with the previously described cardiovascular phases of septic shock.25 Additionally, adrenomedullin blockade with specific anti-rat adrenomedullin antibodies prevented the development of the hyperdynamic phase of sepsis and normalized microvascular blood flow at 5 h after onset of CLP.23 Direct IVM studies to examine the role of adrenomedullin in the intestinal microcirculation
Intravital Intestinal Videomicroscopy
253
Figure 4. Schematic depiction of intravital videomicroscopy setups in use in investigators’ laboratory for study of regulation of microvascular blood flow distribution in rat intestinal microvasculature during various states of circulatory shock. This technique is most powerful of methods discussed, and provides detailed, continuous picture of arteriolar and venular distribution of blood flow within intestines and other organs during physiological and pathophysiological conditions. BP, blood pressure, HR, heart rate; RBC, red blood cell; T, temperature.
have not been reported but are currently underway in our laboratory. A number of models exist which allow for the study of infectious alterations in blood flow and distribution of flow, including CLP, intravenous or intraperitoneal bacteria or bacterial product infusion, and chronic abscess formation. Differences in antigen presentation, bacterial load, and toxin exposure inherent in these models confound the simple summarization and interpretation of data across models. Septic shock results in complex alterations in cells of the immune and vascular systems in a sequential and incremental response affecting all organ systems, including the GI tract, in ways that are unique to each model. This review will focus on findings that we made using our established models of Escherichia coli bacteremia and our chronic subcutaneous abscess model. Bacteremia causes intestinal microvascular vasoconstriction and hypoperfusion,26ÿ28 which could conceivably cause generation of free radicals and altered cytokine production. Lazaroid protects against intravenous E. coli-induced intestinal vasoconstriction and hypoperfusion20 and impaired renal blood flow,29 presumably by antioxidant protection via scavenging of free radicals and blockade of lipid radical chain reactions. This protection in the microvasculature was afforded in spite of a persistent increase in cardiac performance, which suggests that local
intestinal free radical production is important in intestinal microvascular sequellae during sepsis. Platelet-activating factor (PAF) is produced by many stimulated cells such as endothelial cells, neutrophils, basophils, monocytes, and macrophages via phospholipase A2-mediated cleavage of cell membrane phospholipids. Unlike the eicosanoids, which are also not stored in cells, PAF is not an arachidonic acid derivative, but rather an acetyl glycerol ether phosphocholine that is thought to be 10,000 times more potent than histamine with respect to its vasoactive properties. This vasoconstrictor is a potentially harmful mediator of blood flow that was shown to be upregulated during infection.21,30 We compared the microvascular events of early E. coli bacteremia in the absence and presence of topical PAF, both with and without the PAF receptor antagonist WEB2086.32 Both sepsis and topical PAF decreased arteriolar inflow and venular outflow that was due to spatially discrete areas of venular constriction in the group treated with topical PAF. Paradoxically, application of the PAF receptor antagonist prevented sepsis-induced vasoconstriction and reduction in blood flow in both the PAF-treated and E. coli sepsis groups. We also27 examined the involvement of arachidonic acid metabolites in the microvascular alterations of infection, and found that bacteremia caused increased cardiac output, with decreased intestinal blood flow secondary
254
Matheson and Garrison
Figure 5. Diagram of typical dose-response data, allowing assessment of both changes in mediator system sensitivity and reactivity in various pathological conditions.
to arteriolar and venular constriction. Cyclo-oxygenase blockade with mefenemate did not prevent the arteriolar constriction and decreased inflow, but mefenemate produced a profound venular dilation. These studies suggest that microvascular control mechanisms in response to a septic challenge vary considerably, not only by organ system, but also by distal-to-proximal position in the arteriolar and venular trees. While it is true that hyperdynamic cardiac output with decreased total peripheral resistance is a clinical hallmark of sepsis, relatively few studies have addressed the underlying cellular mechanisms that bring about these cardiovascular changes. While adrenomedullin was implicated in the hyperdynamic phase and perhaps in the switching to the hypodynamic phase, another mediator that has received significant attention is endothelial-derived relaxing factor, or nitric oxide (NO). NO is chronically elevated during infection, and NO blockade might improve survival rate.33 We found21 that NO blockade during acute E. coli bacteremia actually worsens the microvascular constriction and hypoperfusion of the gut. The effect of continued overexpression of NO on endothelial and vascular smooth muscle cells is not clear.34 Vascular ring studies demonstrate decreased reactivity of large-conduit blood vessels after prolonged infectious exposure.35ÿ37 We addressed this issue by studying28 in vivo intestinal microvascular responses shortly after IV E. coli infusion, and observed impaired magnitude and frequency of vasomotion (a normally occurring rhythmic process of dilation and contraction) in both inflow and premucosal arterioles. Topical application of acetylcholine partially restored the vasomotor activity of the premucosal arterioles and fully restored the vasomotion of inflow arterioles. This study revealed no change in sensitivity or reactivity of endothelium-dependent relaxation to acetylcholine or
Figure 6. One theoretical interpretation of two-hit hypothesis of deranged intestinal function contributing to progression to multiple organ dysfunction during shock states, which envisions vicious cycle promulgating deranged blood flow to intestine.
endothelium-independent relaxation to nitroprusside in the bacteremic group. However, constriction with norepinephrine was impaired in the bacteremic group in both inflow and premucosal arterioles, suggesting that acute bacteremia causes both endothelial alterations (vasomotion) as well as vascular smooth muscle cell changes (a-adrenergic constriction). The alterations briefly discussed here deal with mechanisms that center on the involvement of the endothelial cell in the microvascular derangement in blood flow seen during septic shock, in particular at the level of the endothelial cell membrane. Another mediator whose release is controlled at the level of the endothelial cell in response to local conditions is endothelin, a potent endothelial-derived vasoconstrictor peptide. We found increased endothelin release38 to be important in the balance of microvascular tone toward a more tonically constricted state during bacteremia.39 Other endothelial cell alterations are extensively described, such as the upregulation and expression of adhesion molecules in the presence of chemotactic bacteria or bacterial products.40 Thus, another endothelium-centered mechanism that might contribute to impaired intestinal blood flow is altered neutrophil trafficking with subsequent capillary and/or postcapillary venule plugging. These changes might further cultivate hypoxic conditions that could stimulate free radical formation and proinflammatory cytokine production, thus feeding back in a vicious cycle leading to deranged flow and organ dysfunction. While no gold standard treatment exists for the treatment of sepsis and MODS, potential therapies that stabilize the endothelial cell membrane with the intent of normalizing intestinal blood flow seem worthy of further pursuit.
Intravital Intestinal Videomicroscopy
ROLE OF THE MICROCIRCULATION IN HEMORRHAGIC SHOCK
Resuscitated hemorrhagic shock with subsequent multiple organ failure remains a costly and significant cause of morbidity and mortality.41 Current resuscitation paradigms consist of rapid IV fluid correction of the hemorrhagic volume deficit.42 Despite restoration of adequate blood pressure and cardiac output, some patients develop major alterations in organ perfusion and tissue metabolism.43 Numerous interventions have been used to protect organ perfusion and cellular integrity from ischemic injury.44ÿ47 Primarily, these measures work via three mechanisms: 1) improvement of organ blood flow and tissue perfusion and oxygenation; 2) enhancement of metabolic processes to stabilize cellular integrity; or 3) administration of specific inhibitors or antagonists of inflammatory modulators to alter the inflammatory process and preserve tissue function and integrity. In the intestinal circulation, resuscitated hemorrhagic shock produces vasoconstriction and hypoperfusion in a similar fashion to that previously described for septic shock.46 The endothelial cell appears to play a pivotal role in the development of this condition. In the past, our primary focus was on altered production of endothelium-derived mediators of intestinal blood flow after resuscitation from hemorrhagic shock. The complete etiology of MODS after resuscitated hemorrhage is not fully described, but one theory suggests that resuscitated hemorrhage can lead to MODS as a result of decreased intestinal blood flow from impaired endothelial cell function. Diminished intestinal microvascular blood flow might then lead to tissue hypoxia or anoxia, release of reactive oxygen-derived free radicals via oxidative burst, and cytokine release, which can all contribute to deranged inflammatory processes in the gut. The interpretation of laboratory studies of vascular function following resuscitated hemorrhagic shock has been complicated by the use of a wide range of experimental hemorrhagic shock models and protocols. For example, until recent years, many hemorrhagic shock models included systemic heparinization to prevent thrombus formation in the shed blood, which greatly alters platelet and endothelial cell function and improves microvascular blood flow. In those studies, animals essentially received a therapeutic agent, heparin, which profoundly alters the experimental results. For this reason, many hemorrhagic studies that were performed prior to the observations about heparin provide different experimental interpretations than previously thought. Another factor that might alter interpretation of data from hemorrhagic shock studies is that several protocols of hemorrhage and resuscitation are com-
255
monly utilized. Hemorrhage can be performed either by quick withdrawal from an artery or by a relatively slow withdrawal from the venous side of the circulation. Additionally, hemorrhage can be characterized as a fixed-pressure hemorrhage wherein the pressure is held at a predetermined level by withdrawal and infusion of blood for a fixed time period, or as a fixed volume hemorrhage over time. These different methods significantly alter the development, severity, and outcome of a hemorrhage. The method of resuscitation varies widely from no resuscitation, to return of shed blood or shed blood plus one or two equal volumes of fluid, to intravenous infusion of multiple volumes of fluid without any returned shed blood. Moreover, resuscitation regimens are not standardized between different research groups who favor different formulations of colloid or crystalloid solutions. This review will focus on recent studies utilizing models of nonheparinized, fixed-pressure hemorrhagic shock with shed blood and/or crystalloid fluid resuscitation. Hemorrhage studies of this kind demonstrated that microvascular blood flow in the liver, small intestine, kidneys, spleen, and skeletal muscle is significantly impaired despite the restoration of central venous pressure and central hemodynamic parameters of cardiac function and arterial pressure.48 Numerous studies3,49,50 demonstrated that intestinal microvascular blood flow is significantly impaired following hemorrhage and resuscitation, despite the confirmation of adequate fluid resuscitation by measurement of cardiac performance and blood pressure. Intravital videomicroscopy studies showed that diameters of inflow and premucosal arterioles return to normal values immediately following resuscitation, but that arteriolar constriction progressively worsens over time, such that by 2 h after resuscitation, inflow and premucosal arterioles are 25ÿ30% constricted, with an accompanying 45ÿ50% decrease in blood flow. Fruchterman et al.3,46 demonstrated impaired endothelial-dependent vasodilation to acetylcholine in these vessels, as well as histological injury and neutrophil influx and impaired nitric oxide synthase function, but no apparent oxidant injury. Furthermore, these changes were prevented or diminished by the blockade of the complement cascade with soluble complement receptor treatment prior to the resuscitation regimen. The mechanism of this microvascular protection is not known, but it might involve protection of endothelial cell function and/or decreased neutrophil activation and influx. Several studies45,51 showed that pentoxifylline administration during and after resuscitation improves microvascular tissue perfusion. Pentoxifylline has a wide range of therapeutic benefits after hemorrhage and resuscitation, which include improved cardiac perfor-
256
Matheson and Garrison
Figure 7. Peritoneal resuscitation maintains dilation and perfusion in intestinal microvasculature during resuscitated hemorrhagic shock, demonstrating possible role for peritoneal resuscitation in shock regimens. Reproduced with permission from Zakaria et al.54 A1, inflow arteriole; CR, conventional resuscitation; dA3, distal premucosal arteriole; HS, hemorrhagic shock; pA3, proximal premucosal arteriole; PR, peritoneal resuscitation.
mance and intestinal and renal microvascular blood flow. Pentoxifylline is thought to improve these conditions by one of its myriad effects, including increased red blood cell deformability, endothelial cell membrane fluidity, and altered neutrophil activity and cytokine production. It is not known which of these functions produce the improvement in microvascular blood flow, or if all of the effects are necessary for microvascular perfusion to be restored to prehemorrhage levels. In addition to pentoxifylline and soluble complement receptor, heparan49 and chemically modified heparin,52 which are heparin analogues with little or no anticoagulant activity, also appear to restore microvascular blood flow after centrally resuscitated hemorrhagic shock. All of these agents appear to similarly protect the endothelial cell membrane after hemorrhagic shock, suggesting that the endothelial cell does indeed play a central role in the pathogenesis of complications that can develop after resuscitated hemorrhagic shock. A recent focus of interest in our laboratory is on the role of the intestinal interstitial milieu in the regulation of small-intestinal blood flow in hemorrhagic shock, and on the manipulation of that environment as a therapeutic modality. In particular, our recent studies42,53,54 focused on the use of peritoneal resuscitation solutions of differing compositions as adjuncts to blood or fluid (crystalloid) resuscitation. In particular, we showed that adding peritoneal resuscitation to standard resuscitation protocols of shed blood plus two equal volumes of saline reverses the persistent vasoconstriction that develops in the intestinal microvasculature in spite of restored central hemodynamic performance after hemorrhagic shock. Figure 753 is a summary of our recent studies that
demonstrated good microvascular perfusion of both inflow arterioles (A1) and premucosal arterioles (proximal and distal A3) after resuscitated hemorrhagic shock in animals that received peritoneal resuscitation in addition to blood and saline. We postulated42 that direct peritoneal resuscitation (DPR), in conjunction with conventional IV fluid resuscitation (CR), provides therapeutic benefits that include: 1) reversal of intestinal vasoconstriction and hypoperfusion that occur after CR alone; 2) reversal of microvascular no-flow phenomena after CR; 3) suppression of proinflammatory cytokines and neutrophil activation in the gut; 4) prevention of CR-mediated interstitial fluid shifts and electrolyte imbalances; and 5) correction of lactic acidosis by improved tissue perfusion and the buffering power of DPR. In conclusion, peritoneal resuscitation, alone or in conjunction with blood and fluid resuscitation, helps maintain a dilated, perfused gut microcirculation, suggesting that the peritoneal resuscitation might be a viable adjunct therapy in patients with hemorrhagic shock in order to prevent gut hypoperfusion, which can contribute to the development of MODS and death.
REFERENCES 1. Hinshaw LB. Sepsis/septic shock: participation of the microcirculation: an abbreviated review. Crit Care Med 1996;24:1072ÿ1078. 2. Matheson PJ, Wilson MA, Garrison RN. Regulation of intestinal blood flow. J Surg Res 2000;93:182ÿ196. 3. Fruchterman TM, Spain DA, Wilson MA, Harris PD, Garrison RN. Selective microvascular endothelial cell dysfunction in the small intestine following resuscitated hemorrhagic shock. Shock 1998;10:417ÿ422. 4. Spain DA, Fruchterman TM, Matheson PJ, Harris PD, Garrison RN. Complement activation mediates intestinal injury after resuscitation from hemorrhagic shock. J Trauma Injury Infect Crit Care 1999;46:224ÿ233. 5. Wanacek M, Weitzberg E, Rudehill A, Oldner A. The endothelin system in septic and endotoxin shock. Eur J Pharmacol 2000;407:1ÿ15. 6. Furchgott RF, Bhadrakam S. Reactions of strips of rabbit aorta to epinephrine, isopropylarterenol, sodium nitrite and other drugs. J Pharmacol Exp Ther 1953;108:129ÿ143. 7. Vankan WJ, Huyghe JM, Slaaf DW, van Donkelaar CC, Drost MR, Janssen JD, Huson A. Finite-element simulation of blood perfusion in muscle tissue during compression and sustained contraction. Am J Physiol 1997;273:1587ÿ1594. 8. Cohen R, Shepherd J, Vanhoutte P. Inhibitory role of endothelium in the response of isolated coronary arteries to platelets. Science 1983;221:273ÿ274. 9. Kawabe T, Harris PD, Zakaria EL, Garrison RN. Sepsis alters vessel contraction by adrenoceptor-induced nitric oxide and prostanoid. J Surg Res 2003;110:352ÿ359. 10. Hakkinen JP, Miller MW, Smith AH, Knight DR. Measurement of organ blood flow with coloured microspheres in the rat. Cardiovasc Res 1995;29:74ÿ79. 11. Matheson PJ, Lusco V, Wilson MA, Garrison RN. Omega-3 fatty acids in immune-enhancing enteral diets selectively increase blood flow to the ileum by a bile acid-dependent mechanism. Surgery 2002;132:673ÿ681.
Intravital Intestinal Videomicroscopy 12. Matheson PJ, Wilson MA, Harris PD, Anderson GL, Garrison RN. Glucose-induced intestinal hyperemia is mediated by nitric oxide. J Surg Res 1997;72:146ÿ154. 13. Zhao H, Spain DA, Matheson PJ, Harris PD, Garrison RN. Progressive decrease in constrictor reactivity of the non-absorbing intestine during sepsis. Shock 2001;16:40ÿ43. 14. Zhao H, Spain DA, Matheson PJ, Vaughn C, Harris PD, Garrison RN. Sustained infection induces two distinct microvascular mechanisms in the splanchnic circulation. Surgery 2000;128:513ÿ519. 15. Revell M, Greaves I, Porter K. Endpoints for fluid resuscitation in hemorrhagic shock. J Trauma Injury Infect Crit Care [Suppl] 2003;54:63ÿ67. 16. Dremsizov TT, Kellum JA, Angus DC. Incidence and definition of sepsis and associated organ dysfunction. Int J Artif Organs 2004;27:352ÿ359. 17. Mainous MR, Ertel W, Chaudry IH, Deitch EA.. The gut: a cytokine-generating organ in systemic inflammation. Shock 1995;4:193ÿ199. 18. Nieuwenhuijzen GA, Deitch EA, Goris RJ. Infection, the gut and development of the multiple organ dysfunction syndrome. Eur J Surg 1996;162:259ÿ273. 19. Beishuizen A, Vermes I, Haanen C. Endogenous mediators in sepsis and septic shock. Adv Clin Chem 1998;33:55ÿ131. 20. Krysztopik RJ, Bentley FR, Spain DA, Wilson MA, Garrison RN. Lazaroid improves intestinal blood flow in the rat during hyperdynamic bacteremia. Br J Surg 1997;84:1717ÿ 1721. 21. Spain DA, Wilson MA, Bar-Natan MF, Garrison RN. Role of nitric oxide in the small intestinal microcirculation during bacteremia. Shock 1994;2:41ÿ46. 22. Zhou M, Chaudry IH, Wang P. Adrenomedullin is upregulated in the heart and aorta during the early and late stages of sepsis. Biochim Biophys Acta 1999;1453:273ÿ283. 23. Wang P, Ba ZF, Cioffi WG, Bland KI, Chaudry IH. The pivotal role of adrenomedullin in producing hyperdynamic circulation during the early stage of sepsis. Arch Surg 1998;133: 1298ÿ1304. 24. Wang P, Zhou M, Ba ZF, Cioffi WG, Chaudry IH. Up-regulation of a novel potent vasodilatory peptide adrenomedullin during polymicrobial sepsis. Shock 1998;10:118ÿ122. 25. Wang P. Adrenomedullin in sepsis and septic shock. Shock 1998;10:383ÿ384. 26. Wilson MA, Steeb GD, Garrison RN. Endothelins mediate intestinal hypoperfusion during bacteremia. J Surg Res 1993;55: 168ÿ175. 27. Gosche JR, Spain DA, Garrison RN, Lubbe AS, Cryer HG. Differential microvascular response to cyclo-oxygenase blockade in the rat small intestine during acute bacteremia. Shock 1994;2: 408ÿ412. 28. Spain DA, Wilson MA, Krysztopik RJ, Matheson PJ, Garrison RN. Differential intestinal microvascular dysfunction occurs during bacteremia. J Surg Res 1997;67:67ÿ71. 29. Krysztopik RJ, Bentley FR, Spain DA, Wilson MA, Garrison RN. Free radical scavenging by lazaroids improves renal blood flow during sepsis. Surgery 1996;120:657ÿ662. 30. Karasawa K. Kato H. Setaka M. Nojima S. Accumulation of platelet-activating factor acetylhydrolase in the peritoneal cavity of guinea pig after endotoxin shock. J Biochem 1994;116: 368ÿ373. 31. Koltai M. Hosford D. Braquet PG. Platelet-activating factor in septic shock. New Horizons 1993;1:87ÿ95. 32. Bar-Natan MF. Wilson MA. Spain DA. Garrison RN. Plateletactivating factor and sepsis-induced small intestinal microvascular hypoperfusion. J Surg Res 1995;58:38ÿ45. 33. Wolkow PP. Involvement and dual effects of nitric oxide in septic shock. Inflammation Res 1998;47:152ÿ166.
257
34. Kirkeboen KA, Strand OA. The role of nitric oxide in sepsis—an overview. Acta Anaesthesiol Scand 1999;43:275ÿ288. 35. Young JS, Headrick JP, Berne RM. Endothelial-dependent and independent responses in the thoracic aorta during endotoxic shock. Circ Shock 1991;35:25ÿ30. 36. McKenna TM, Titius WA. Role of monokines in altering receptor and non-receptor mediated vascular contraction in sepsis. Prog Clin Biol Res 1989;286:279ÿ303. 37. McKenna TM, Martin FM, Chernow B, Briglia FA. Vascular endothelium contributes to decreased aortic contractility in experimental sepsis. Circ Shock 1986;19:267ÿ273. 38. Chou MC, Wilson MA, Spain DA, Hadjiminas D, Anderson GL, Cheadle WG, Garrison RN. Endothelin-1 expression in the small intestine during chronic peritonitis. Shock 1995;4:411ÿ414. 39. Wilson MA, Steeb GD, Garrison RN. Endothelins mediate intestinal hypoperfusion during bacteremia. J Surg Res 1993;55: 168ÿ175. 40. Zimmerman GA, McIntyre TM, Prescott SM. Adhesion and signaling in vascular cell-cell interactions. J Clin Invest [Suppl] 1997;100:3ÿ5. 41. Hassoun HT, Kone BC, Mercer DW, Moody FG, Weisbrodt NW, Moore FA. Post-injury multiple organ failure: the role of the gut. Shock 2001;15:1ÿ10. 42. Garrison RN, Conn AA, Harris PD, Zakaria ER. Direct peritoneal resuscitation as adjunct to conventional resuscitation from hemorrhagic shock: a better outcome. Surgery 2004;136:900ÿ908. 43. Ivatury RR, Simon RJ, Islam S, Fueg A, Rohman M, Stahl WM. A prospective randomized study of end points of resuscitation after major trauma: global oxygen transport indices versus organspecific gastric mucosal pH. J Am Coll Surg 1996;183:145ÿ154. 44. Childs EW, Smalley DM, Moncure M, Miller JL, Cheung LY. Effect of LFA-1beta antibody on leukocyte adherence in response to hemorrhagic shock in rats. Shock 2000;14:49ÿ52. 45. Flynn WJ, Cryer HG, Garrison RN. Pentoxifylline restores intestinal microvascular blood flow during resuscitated hemorrhagic shock. Surgery 1991;110:350ÿ356. 46. Fruchterman TM, Spain DA, Wilson MA, Harris PD, Garrison RN. Complement inhibition prevents gut ischemia and endothelial cell dysfunction after hemorrhage/resuscitation. Surgery 1998;124: 782ÿ791. 47. Wang P, Ba ZF, Chaudry IH. Endothelial cell dysfunction occurs after hemorrhage in nonheparinized but not in preheparinized models. J Surg Res 1993;54:499ÿ506. 48. Smail N, Catania RA, Wang P, Cioffi WG, Blank KI, Chaudry IH. Gut and liver: the organs responsible for increased nitric oxide production after trauma-hemorrhage and resuscitation. Arch Surg 1998;133:399ÿ405. 49. Watkins JM, Spain DA, Krysztopik RJ, Downard PJ, Wilson MA, Garrison RN. Heparan preserves intestinal perfusion after hemorrhage and resuscitation. J Surg Res 1996;66:154ÿ158. 50. Flynn WJ Jr, Gosche JR, Garrison RN. Intestinal blood flow is restored with glutamine or glucose suffusion after hemorrhage. J Surg Res 1992;52:499ÿ504. 51. Wang P, Ba ZF, Zhou M. Pentoxifylline restores cardiac output and tissue perfusion after trauma-hemorrhage and decreases susceptibility to sepsis. Surgery 1993;114:352ÿ359. 52. Wang P, Ba ZF, Chaudry IH. Chemically modified heparin improves hepatocellular function, cardiac output, and microcirculation after trauma-hemorrhage and resuscitation. Surgery 1994;116: 169ÿ176. 53. Zakaria ER, Garrison RN, Spain DA, Matheson PJ, Harris PD, Richardson JD. Intraperitoneal resuscitation improves intestinal blood flow following hemorrhagic shock. Ann Surg 2003;237:704ÿ713. 54. Zakaria ER, Spain DA, Harris PD, Garrison RN. Generalized dilation of the visceral microvasculature by peritoneal dialysis solutions. Periton Dialysis Int 2002;22:593ÿ601.
