Pathophysiology Associated with Gastric Dilation-Volvulus Syndrome

Gastric dilatation - volvulus syndrome (GDVs) includes acute gastric dilation (GD), acute gastric dilation with gastric volvulus (GDV), and chronic gastric volvulus (cGV) [1]. Although the overall prevalence of this disease in the population at risk is low, animals affected by GD and GDV are commonly presented as emergencies, so this remains an important syndrome for any clinician involved in emergency practice. Treatment protocols based on best current evidence regarding the pathophysiology resulting from GD and GDV, has resulted in good survival rates [2-4]. It is likely that any future improvements in the treatment of this condition will be as a direct result of a better understanding of the disease mechanisms involved. The most profound functional changes that occur are associated with GD and GDV and vary according to the extent of gastric dilation, the degree of gastric rotation, and the duration of each. Although it is often helpful to consider the pathophysiology on the basis of an organ or body system, the changes that occur are complex and interdependent. For the body as a whole, the consequences of GDVs are reduced oxygen delivery to tissues. These changes can be so mild that they are barely evident or they can be so severe that they result in irreversible "sepsis syndrome", either through uncontrolled infection (sepsis) or noninfectious systemic inflammatory response syndrome (SIRS) [5]. The author will begin this review by describing the pathoanatomy of GDVs followed by a description of the effects this pathoanatomy has on body systems and organs based on data derived from experimental creation of GD or GDV. Data from clinical studies will be presented as confirmation, or otherwise, that each process is actually happening in dogs affected by GDVs. Again, however, it is important to remember that these processes are interdependent; no process or system can really be considered in isolation of other organs or systems. This is particularly germane to clinical treatment where a "global approach" to the patient aimed at optimizing oxygen delivery to tissues will yield the best clinical results in terms of overall outcome.

Pathoanatomy

Dilation of a normally positioned stomach, results in a 90- degree counterclockwise gastric rotation around the gastroesophageal junction [6]. The majority of dogs affected with GDVs undergo simultaneous dilation and clockwise rotation (between 180 and 360 degrees) of the stomach about the gastroesophageal junction [2-4]. It has been suggested that gastric malposition must be present before gastric dilation results in classic GDV [6]. What conditions or factors trigger gastric rotation and whether it preceded gastric volvulus or vice versa, in an individual, are rarely clear at the time of diagnosis. From a pathophysiologic perspective, GD and GDV can create similar systemic effects, but evidence exists that the deleterious effects to the blood supply of local organs (stomach and spleen) are greater when volvulus has occurred [6,7].

Splenic displacement has been documented in many patients with GDVs [2,6] and is considered inevitable because of its close anatomic association with the greater curvature of the stomach. The magnitude of displacement should, in theory, have a direct relationship with the degree of splenic vascular compromise.

Experimental Models of GDV

Attempts to precisely recreate the pathoanatomy of dogs with naturally occurring GDV have failed. In addition, all of the experimental models of this disease require general anesthesia, abolishing any conscious influence (such as release of endogenous catecholamines and cortisol) over events in these dogs. It seems that it is impossible to rotate the stomach of an otherwise normal dog and dilate it with air and have it remain in that state without additional "devices." Passi et al. [8] ligated the thoracic esophagus around an orogastric tube, via thoracotomy, and ligated the pylorus through a right subcostal laparotomy. They inflated the stomach to a mean pressure of 25 +/- 11 mm Hg to create a model of GD. Merkley et al. [9] used the experimental model of GD that was predominant at that time, using an intragastric balloon inflated to 80mm Hg. Orton and Muir [10] argued that such high intragastric pressures (80 mm Hg) should not be used because the intragastric pressure in dogs with naturally occurring disease varied from 9 to 62 mm Hg. They favored, therefore, using an intragastric balloon inflated to a pressure of 30 mm Hg [10,11]. In addition, these workers created a small celiotomy to allow concurrent rotation of the dilated stomach to create GDV. Subsequently, models of GDV have involved umbilical tape ligatures placed at the gastroesophageal junction and at the pylorus, in such a way that branches of the vagus nerve and the gastric vasculature are not compromised, to create an air-tight seal. The stomach is then inflated via a Foley catheter in the pyloric antrum, to a pressure of 30 mm Hg and the stomach is sutured in a rotated position to create GDV [7,12-14]. The length of time that these models of GD or GDV were maintained varied between 90 and 180 minutes. In addition to these models of GD and GDV, Lantz et al. [7] studied the effect of gastric volvulus alone by suturing the non distended stomach in position after rotating it 360 degrees. All models of GD or GDV that required celiotomy for their creation had the celiotomy closed before any measurements were made.

In summary, none of these models perfectly recreate the natural disease. In addition, anesthetic drug protocols have varied from model to model, which could have a significant effect on results. Because of the careful use of control animals, however, experiments using these models have provided a valuable insight into the events that follow GD and GDV.

Circulatory Pathophysiology

Passi et al. [8] using their acute GD model, demonstrated that caudal vena cava flow rates fell rapidly from 51 ml kg-1min-1 to between 0-9 ml kg-1min-1 within 31 minutes. Cranial caval flow rates fell from 28.9 ml kg-1min-1 to 24 ml kg-1min-1 during this time. Mean arterial pressure (MAP), during the same period, fell from 112 mm Hg to 90 mm Hg, and it took only a further 30 minutes for MAP to fall below 50 mm Hg. Pretreatment with lactated Ringer's solution (50 ml/kg), bilateral cervical vagotomy, and sympathectomy did not prevent the onset of these hemodynamic changes. Of 16 dogs, 12 required additional lactated Ringer's following gastric decompression to restore normal hemodynamics. These workers concluded that mechanical obstruction of the cava was responsible for an acute decrease in venous return, leading to systemic arterial hypotension, and that fluid administration was required to return these dogs to their predilation hemodynamic state. Orton and Muir [10,11] confirmed the earlier findings of Passi [8] and added further information regarding a fall in the maximum rate of left ventricular pressure change over time (dp/dtmax), increased heart rate, decreased cardiac output, and presumed neurohormonal reflex alterations in total peripheral resistance. Wingfield et al. [15] provided angiographic evidence of caudal venous attenuation and obstruction in their model of acute gastric dilation. Barnes et al. [16] provided indirect evidence of mechanical obstruction secondary to changes in intra-abdominal pressure in an experiment that investigated the hemodynamic response to elevations in intra-abdominal pressure (IAP). They documented hemodynamic changes similar to those that had been documented following experimental GD by increasing intraabdominal pressure to 40 mm Hg, alone. Together, the above experiments have demonstrated that acute gastric dilation, using intragastric pressures equivalent to those seen in dogs with natural disease, cause mechanical obstruction to the caudal vena cava with predictable consequences with respect to cardiac output. In addition, acute increases in IAP alone can be detrimental to blood flow through the low-pressure vascular systems (caudal cava and portal vein). Finally, because simply decompressing the stomach was not always enough to return hemodynamics to control levels, data from these experiments contained the clue that additional factors were mediating prolonged systemic hypotension in animals recovering from GD and GDV.

Impaired myocardial contractility has also been cited as a contributing factor in the poor circulatory status of dogs with GD and GDV. Such an influence could be secondary to an imbalance to the cardioinhibitory (e.g., myocardial depressant factor [MDF] or vagal mediated) and cardiostimulatory (circulating catecholamines) influences over myocardial function. In addition, myocardial ischemia could directly influence myocardial performance in these hypotensive patients. Orton and Muir [11] failed to detect cardioactive substances in their acute model of GDV, but they commented that, if MDF was being released by the pancreas, it could remain sequestered in the portal circulation until gastric decompression and mechanical release of the splanchnic circulation had occurred. Horne et al. [17] investigated the direct effect of experimental GDV on myocardial blood flow and myocardial oxygen extraction in addition to more global changes in hemodynamic indices. They documented reduced myocardial blood flow (50% below control level) increased heart rate (similar to what has been previously described) and corresponding increase in myocardial oxygen extraction (30% above controls). Of the 8 experimental dogs, 6 developed subendocardial necrosis. Unfortunately, the long-term effects of MDF and the myocardial ischemia were not determined by either investigation.

Clinical Evidence

Good objective clinical data regarding the circulatory status of dogs with GDVs before fluid resuscitation is not contained in any of the large reports of this condition [2-4]. One small prospective study by Wagner et al. [19] evaluated cardiopulmonary variables in 6 dogs that underwent gastropexy without partial gastrectomy for GDV. Unfortunately, 4 of these 6 dogs had intravenous fluid therapy before any measurements were made and all had intraoperative fluid therapy "according to the perceived needs of the patient, made by the attending clinician and anaesthetist." Wagner et al. [19] concluded that not all dogs with naturally occurring GDVs have severe circulatory derangements. The study population did not include any dogs that required partial gastric resection -- creating study population bias -- and with so many uncontrolled variables, such as degree of gastric distention and fluid administration, and the small sample size, it is hard to interpret these data as being sound scientific evidence. Other evidence of the degree of systemic hypoperfusion was provided indirectly by De Papp et al. [4] who retrospectively evaluated venous plasma lactate concentration as a predictor of gastric necrosis and survival in 102 dogs. In this study, the dogs with gastric necrosis were compared with dogs that did not have gastric necrosis, and mean venous plasma lactate concentrations were 6.6 mmol/L and 3.3 mmol/L, respectively. Originally, these workers hypothesized that the gastrointestinal tract was the source of the lactate; but ultimately they concluded that any ischemic tissue in the body could contribute to the venous plasma lactate concentration, especially if the sample was retrieved prior to gastric decompression and release of the splanchnic (portal venous) reservoir of gut-derived lactate. These authors suggested that high lactate was more likely to reflect systemic hypoperfusion than gut production alone and, as such, should be considered a marker of disease severity rather than a direct indicator of gastric-wall ischemia. In addition, dogs with naturally occurring GDVs frequently have rupture of the short gastric vessels and hemoperitoneum, both of which can affect regional organ blood flow (see later) and systemic hemodynamic parameters.

The prevalence of cardiac arrhythmias in dogs recovering from GDVs has been well documented [2,3,19]. Evidence of myocyte injury in the form of serum cardiac troponin I (cTI) and cardiac troponin T (cTT) has been demonstrated in dogs clinically affected with GDVs [20]. In the same study, 10 of 16 dogs that died had arrhythmias; all 16 had detectable cTI and CTT, but, it was not reported how many of the dogs that died did so as a direct result of a fatal arrhythmia [20]. In 4 of 5 dogs necropsied, myocardial necrosis was evident. In the absence of the identification of a direct myocardial depressant factor, another cardiotoxic element, or electrolyte derangements in the plasma of dogs affected with GDVs, reduced coronary blood flow and increased myocardial oxygen demand could be considered as a reflection of the systemic hypotension and hypoperfusion. Because coronary blood flow occurs predominantly during diastole, low diastolic pressures, in particular, would reduce coronary blood flow. Because diastolic pressure depends on systemic vascular resistance, perturbations of this in addition to total intravascular volume could have a detrimental effect on myocardial oxygen delivery.

In emergency practice, the clinician must rely on subjective indicators of perfusion such as mucous membrane color, capillary refill time, and pulse rate and quality. It would be an error to ignore these basic physiologic parameters when assessing the circulatory status of any patient.

Respiratory Pathophysiology

The effects of increasing gastric volume or increasing intra-abdominal pressure on total thoracic volume and diaphragmatic excursions have been studied infrequently in an experimental setting [15]. These effects include a decrease in total thoracic volume, decreased diaphragmatic excursions, and partial lung lobe collapse (causing ventilation-perfusion mismatching), all resulting in hypoventilation and hypoxia. Compensatory mechanisms such a tachypnea and alterations in the lateral thoracic dimensions preserve lung function initially. Ultimately, the pulmonary blood flow falls secondary to systemic hypovolemia, magnifying the effect of ventilation-perfusion mismatching.

Clinical Evidence

No firm clinical evidence exists that these mechanisms occur in dogs clinically affected by GDVs, but anecdotal evidence of respiratory compromise abounds. It is understandable to anyone involved in emergency practice why objective documentation of this respiratory impairment has not been acquired. Improved pulmonary function, however, must remain one of the potential benefits of gastric decompression.

Dogs clinically affected by GDVs often make frequent attempts to vomit, and regurgitate any oral intake of fluid or food. This might place them at risk for aspiration pneumonia. Aspiration pneumonia is a finding that has been reported as a postoperative complication in several studies of this disease [2,3]. Aspiration of pharyngeal contents can have an acute and a longer term effect on pulmonary gas exchange and should be considered as one of the potential causes of poor lung performance in these animals.

Gastrointestinal, Hepatic, and Splenic Pathophysiology

The effect of 360-degree rotation of the blood supply of the non distended dog stomach was investigated by Lantz et al. [7]. These workers concluded that gastric venous obstruction contributed to mucosal and submucosal edema throughout the stomach, and that stretching of the short gastric arteries along the greater curvature of the stomach exacerbated the ischemic injury at that site. Davidson et al. [13] demonstrated reduced blood flow to all tissues except liver and heart ventricles after experimental GDV lasting 270 minutes. In this experiment, pancreatic and jejunal blood flow were among those measured. The implication of the results of the two aforementioned experiments, is that gastric dilation causes reduced gastrointestinal blood flow either by direct compression of the portal vein or via reduced cardiac output. The compromise to gastric blood flow is worsened by gastric rotation and compression (or avulsion) of the short gastric and splenic arteries. It is important also to realize that the perfusion pressure in the gastric wall arterioles and capillaries must exceed the compressive force applied to these vessels by the tension in the gastric wall in order for blood to flow. The potential for gastric necrosis is, therefore, influenced by several factors: degree of gastric rotation, duration of rotation, intragastric pressure (gastric wall tension), systemic arterial pressure, and compression or injury to the short gastric branches of the splenic artery.

Increased levels of circulating endotoxin along with histologic evidence of intestinal villus injury were demonstrated by Davidson et al. [13] in their experimental GDV group. The finding of high level of circulating endotoxin could indicate increased production of bacterial lipopolysaccharide (LPS) in the gut or increased permeability to LPS. Peycke et al. [21] investigated the effect of experimental GDV on adenosine triphosphate (ATP) and conductance of the canine gastric and jejunal mucosa. They concluded that the jejunal mucosa showed more profound changes and that the conductance values and ATP levels in jejunal mucosa were suggestive of cell-membrane dysfunction. These experiments could be taken as evidence for intestinal mucosal barrier compromise secondary to GDVs, resulting in systemic release of both bacterial products (e.g., LPS) and, potentially, whole bacteria.

Histologic evidence of hepatocellular damage following experimental GDV was provided by Davidson et al. [13]. Several mechanisms could be responsible for this injury: ischemia-reperfusion injury, damage caused by endotoxin, and hypoxic damage. Alterations in hepatic function have not been assessed in experimental models of GDV but, in theory, acute diminution of hepatic function could have serious consequences. The effect of experimental GDV on the spleen has not been reported in detail.

Clinical Evidence

Gastric necrosis following naturally occurring GDVs has a predictable pattern, with the gastric fundus most commonly affected [22]. In clinical studies of this disease, rupture of the short gastric vessels and hemoperitoneum is often recorded, demonstrating that the natural disease causes changes that experimental models have not included, such as blood loss and irretrievable vascular compromise. In addition to necrosis of the fundus, necrosis at the cardia has been reported [2], suggesting that twisting of the cardia can also directly affect local tissue integrity, probably by direct vascular occlusion. The reported prevalence of gastric necrosis in patients with naturally occurring GDV is between 10% and 30% in most studies [2-4,22]. This fact may give an indirect indication of the prevalence of "severe" disease among affected dogs, since the survival rate for dogs that do not require partial gastric resection as part of their treatment approached 100% in one study [4]. The intermediate-term effects of the gastric-wall insult are uncertain. Perhaps the abnormalities in gastric myoelectrical activity reported by Stampley et al. [14] and Hall et al. [23] in dogs recovering from GDV and gastropexy represent these effects. Postoperative gastrointestinal ileus is an occasional complication in clinical patients recovering from GDVs, but the potential causes of this include anesthetic drug use, analgesic drugs, anxiety, and disease-related gastrointestinal compromise, so it is not possible to identify a single cause and effect relationship.

Splenic vascular compromise (vessel avulsion, intravascular thrombosis, infarction) occurs in approximately 16% of dogs that survive and 40% of dogs that die following acute GDVs [2]. Again, perhaps the presence of splenic complications implies more "severe" disease.

Although elevations in serum alkaline phosphatase and alanine aminotransferase have been documented in dogs recovering from GDVs, there is no clinical evidence of either acute or chronic hepatic dysfunction in these dogs.

Ischemia-Reperfusion Injury, Sepsis, Sirs, and Ards (See The Systemic Inflammatory Response Syndrome and Shock)

Experimental evidence that ischemia-reperfusion injury occurs in dogs with acute GDVs was provided by Badylak et al. [12] and Lantz et al. [25]. Ischemia-reperfusion mediated production of reactive oxygen species directly through lipid peroxidation and indirectly through neutrophil activation will fuel the genesis of local and systemic inflammation. Endotoxin is a potent mediator of inflammation and has been demonstrated to be released in experimental models of GDVs [13]. Hypoperfusion and direct ischemic damage to many organs could also result in activation of neutrophils, monocytes, macrophages, and platelets, all of which contribute cytokines to the inflammatory cascade. In addition, inflammation associated with bacterial infection will feed local and systemic inflammation. Together these factors can trigger systemic inflammatory response syndrome (SIRS).

Clinical Evidence

High levels of circulating endotoxin have not been demonstrated in clinical patients with GDVs, nor have indicators of lipid peroxidation been measured from tissues of clinical patients. Large studies of GDVs do not frequently cite massive bacterial infections in dogs that die following treatment [2-4]. The most commonly cited causes of death are euthanasia because of massive gastric necrosis, and continued cardiovascular instability characterized by hypotension, despite appropriate resuscitation attempts, followed by multiple organ failure, notably pulmonary [2,3]. Such findings could be interpreted as circumstantial evidence, at least, that SIRS is a consequence for animals severely affected by GDVs.

Clinical Indicators of Disease Severity

Several preoperative clinical parameters that have been studied might best be considered indicators of the severity of disease in dogs suffering acute GDV rather than being a direct cause of morbidity or mortality. Such parameters include plasma lactate [4], abnormal haemostatic profiles [25], and the presence of cardiac arrhythmias [3,19]. Venous plasma lactate is considered an indicator of the degree of circulatory system compromise. Of course, other clinical data (pulse rate, capillary refill time, mucous membrane colour, packed cell volume, plasma protein levels) can also aid in the clinical evaluation of circulatory status. High lactate was associated with increased chance of having gastric necrosis in one study [4]. This study did not imply cause and effect, simply statistical association between a measure of perfusion status (plasma lactate) and the presence of gastric necrosis, an event influenced by many factors. It would seem, however, that high plasma lactate pre-resuscitation and, perhaps more importantly, elevated plasma lactate that is refractory to treatment, should be considered an indicator of severe disease.

The presence of abnormal laboratory hemostatic profiles in dogs clinically affected with GDV was associated with increased chance of gastric necrosis, in another study [25]. The authors of this study implied that microthrombi that either caused or resulted from the abnormal coagulation status in these dogs accelerated the demise of the gastric wall by local vessel occlusion [25]. Although it is difficult to determine whether stasis precedes thrombus formation in gastric-wall arterioles and capillaries or vice versa in dogs with GDV, abnormalities of hemostasis and evidence of disseminated intravascular coagulation (DIC) should be considered as clear evidence of severe disease.

The presence of cardiac arrhythmias was a negative prognostic indicator in one study [3] but was not thought to influence overall outcome in another [2]. Given that the majority of experimental and clinical evidence so far points to myocardial ischemia as a key factor in the development of cardiac arrhythmias, it would seem that dogs that have the most severely compromised circulation are going to be most likely to develop arrhythmias. Again, in this scenario, the presence of cardiac arrhythmias could be considered an indicator of disease severity.

The three parameters mentioned in this section are representative of many that could be considered in a similar way. Indeed, intraoperative findings such as gastric necrosis and splenic complications could also be considered markers of disease severity, as could postoperative findings such as intractable hypotension, intractable elevation in plasma lactate, fall in the partial pressure of oxygen in arterial blood, and intractable arrhythmias. The sensitivity of each with respect to the severity of disease is unknown and, given the potential complexity of the pathophysiologic events in these patients, the author would caution against using any single number or event as a reason for withdrawing treatment from any patient. Rather, the patient should be considered as a whole in respect of clinical decision making.