Therapeutic interventions aimed at preventing organ failure
The widespread use of blood and plasma for the first time during World War II resulted in the recognition of a new phenomenon — ‘irreversible shock’. It became apparent that numerous hypovolaemic casualties could have their vital signs restored with the administration of intravenous blood and plasma only to develop ‘secondary shock’, often days after the original insult, which rapidly became resistant to any form of treatment. Fifty years later the main thing that seems to have changed is the terminology used to identify the same condition. The terms multiple organ failure syndrome (MOFS), multiple systems organ failure (MSOF) and, most recently, multiple organ dysfunction syndrome (MODS) have all been used to describe ‘irreversible shock’ and organ failure. Rather like the elephant, it seems that we have difficulty describing it but we all know one when we see one.
In 1943 Blalock summarised the causes of irreversible shock seen in the Armed Forces as:
(1) haemorrhage uncomplicated by gross trauma; (2) burns; (3) trauma to large masses of muscle; and (4) the re-establishment of circulation in a damaged ischaemic area.
In the same dissertation he suggests that:
the initial phases (of shock) are certainly associated with and probably dependent upon a reduction in the volume of effective circulating blood … Investigations have indicated that at some time in the development of irreversible shock there appear effects that may be ascribed either to toxic substances elaborated in areas of tissue damage or to a derangement of metabolism produced by a circulation which, either locally or generally, has been compromised over a long time. The currently proposed theories of the pathogenesis of organ failure differ mainly in the fine detail, in particular many of the ‘toxic substances’ or mediators have been identified. Currently, we recognise and have tried to define more clearly the clinical conditions of the systemic inflammatory response syndrome (SIRS) and MODS.
It is postulated that, as a result of an insult, uncontrolled activation of inflammatory pathways may result in tissue destruction and subsequent organ failure. Inflammation is an essential component of the healing process. From a wide variety of stimuli (e.g. trauma, burns, infection) the final common pathway results in vasodilation, increased endothelial permeability, thrombosis, and leucocyte migration and activation. The systemic inflammatory response syndrome is the latest term proposed to describe a failure of localisation. MODS, due to tissue damage in organs distant to the site of the original injury, is the clinical manifestation of SIRS. Irrespective of the initiating stimulus (e.g. surgery, bacterial infection, pancreatitis), the morphology of necropsy specimens in both animal models and patients is remarkably constant. There is microvascular occlusion, endothelial destruction, interstitial oedema, leucostasis and thrombosis. Therefore, there would seem to be a dichotomy. Inflammation is essential for successful recovery from infection or injury, yet an excessive and uncontrolled inflammatory response can result in organ dysfunction or failure. One hypothesis is that there is a level of stimulation that, once exceeded, leads to uncontrolled activation of inflammatory pathways. There would appear to be a critical balance between activation and modulation (Fig. 4.1).
Fully established multiple organ failure is almost always fatal. A greater understanding of the pathogenesis of MODS has provided the basis for treatment regimens currently being used clinically or tested experimentally for the prevention of organ damage. (Fig 4.2 See above) summarises the various stages that are thought to lead to the development of MODS and therefore are targets for intervention. The rationale behind the apparent success, or failure, of a cross-section of these regimens is discussed below.
The primary insult
As always, the earliest treatment is most effective and prevention is better than cure. If we assume that the primary insult is unavoidable, has been correctly diagnosed and treated, then all subsequent forms of treatment are aimed at limiting the degree of tissue damage.
Avoiding tissue hypoxia — simple resuscitation with intravenous fluids
The commonest compounding insult is probably tissue hypoxia as a result of inadequate basic resuscitation. Treating the primary insult is almost a waste of time without restoration of an adequate circulating blood volume. The majority of patients will respond to the administration of intravenous fluids and supplementary oxygen via a face mask. There is little doubt that if the aim is to restore the intravascular volume then this can be done most efficiently using a colloid. As one of the central abnormalities in SIRS is thought to be
endothelial leak, one might hope that a colloid with a larger molecular weight than albumin might be more effectively retained within the vascular compartment and therefore more likely to maintain microvascular flow and organ perfusion. However, there are no human data that demonstrate that any particular solution (colloid or crystalloid) is better than another in terms of outcome, even though there is approximately a 100-fold price difference between the cheapest and most expensive alternatives. There are, however, good data to suggest that to avoid tissue hypoperfusion you need to give enough of whatever you choose and you need to give it promptly with the appropriate level of monitoring. For rapid restoration of haemodynamic function a colloid does the job more efficiently than a crystalloid. Albumin has no demonstrable advantages over cheaper alternatives such as the modified gelatins.
Treating tissue hypoxia — the global approach
The rationale here is that patients with SIRS are thought to have an occult tissue oxygen debt in spite of apparently normal global cardiovascular variables such as blood pressure and urine output, and that by increasing total body oxygen delivery, commonly by the administration of ionotropes, this debt will be repaid and hypoxic tissue damage will be avoided. In patients undergoing high-risk major surgery and critically ill patients on the intensive therapy unit (ITU) there appears to be a positive correlation between a high oxygen delivery and survival. In patients undergoing high-risk major surgery prophylactically increasing cardiac output and oxygen delivery to predetermined supranormal levels has been associated with a decrease in subsequent organ dysfunction and mortality. However, the same principles when applied to established critically ill patients on the ITU has met with very limited success. Once a critical degree of tissue hypoperfusion is established then the situation is apparently irreversible.
There is an ever increasing list of designer vasoactive drugs (e.g. dobutamine, dopexamine, enoximone, piroximone, milrinone, amrinone) that can be used to manipulate global haemodynamics in sepsis. They have attractive hypothetical advantages in terms of their effects on global oxygen flow variables that can be confirmed in animal models and humans. However, there is no evidence to suggest that they are any more effective than the cheaper alternatives such as adrenaline and noradrenaline in terms of outcome.
Treating tissue hypoxia — the regional approach
As a result of the enormous amounts of research that have focused on global oxygen delivery in SIRS and MODS, it has become evident that regional perfusion may be compromised in patients who have apparently adequate global oxygen delivery and consumption. Probably the most commonly monitored end organ in the ITU has been the kidney, as urine output is easy to measure and relates crudely to function. The production of 0.5 ml/kg/hour of urine is accepted as an
indicator of adequate regional perfusion. Unfortunately, although anuria carries a very poor prognosis the converse is not true. Many reno protective agents have failed to modify outcome in humans. In animal models the use of mannitol, frusemide and low-dose renal dopamine infusions has been shown to protect the kidney, but this has not been demonstrated in the clinical environment. This may be because the infusion of a renal dose of dopamine, for example, may well produce a diuresis but cannot be expected to protect the kidney from coexisting hypoxia and hypovolaemia unless it is used as just a part of a total patient management regimen. Many believe that such agents used in isolation do more harm than good as they maintain an adequate urine output in the face of tissue hypoperfusion and also increase myocardial work.
More recently much attention has focused on the splanchnic region. In particular, the measurement of gastrointestinal luminal PCO2, using a gastric or sigmoid tonometer, and calculation of gastrointestinal mucosal pH have become fashionable as indices of splanchnic perfusion. Splanchnic perfusion and, in particular, that to the gut mucosa is compromised early and preferentially in shocked states. There is also a hypothesis that gut mucosal hypoperfusion may result in leakage of gut luminal contents into the bloodstream and that this may be a proinflammatory factor tipping the balance in favour of SIRS and MODS. In support of this hypothesis gut mucosal hypoperfusion, as determined by the presence of a mural acidosis measured with a tonometer, has been shown to be the most sensitive predictor of MODS and death in patients undergoing major surgery and on the ITU. Certain therapeutic manoeuvres have been demonstrated to improve gut mucosal perfusion both in animal models and in humans. These include the administration of intravenous fluids, dobutamine, dopexamine and donor blood.
Avoiding nosocomial infections.
Once a patient has some degree of organ dysfunction on an ITU they are thought to be at greater risk from nosocomial infection. In the prevention of secondary infection good hand washing and the avoidance of cross-infection carried by staff probably have the greatest impact. Assuming that such cross-infection is avoided then the patient is his or her own enemy and bacteria carried in the gastrointestinal tract provide the commonest source of secondary infection. Nosocomial pneumonia is thought to occur commonly as a result of spillage from the upper gastrointestinal tract into the lungs. It has been demonstrated that the administration of H2-receptor antagonists with the intention of reducing gastric acidity and avoiding stress ulceration also encouraged the growth of bacteria in the stomach and an increased incidence of nosocomial pneumonias. The use of sucralfate as stress ulcer prophylaxis has the advantage of also being bacteriostatic and has been associated with a decreased incidence of nosocomial pneumonia. Another approach to the problem of secondary infection from the gastrointestinal tract is selective decontamination of the digestive tract (SDD)
with the aim of reducing the incidence of secondary infection, from both overspill and translocation. SDD involves the use of a variety of topical and intravenous antimicrobial agents with the aim of removing the pathogenic gut flora but maintaining the commensal anaerobes. Numerous studies have demonstrated that SDD and/or modifying the gastric intraluminal pH reduces the incidence of nosocomial pneumonia. However, reducing the incidence of nosocomial pneumonias has had a far lesser effect on outcome than one might anticipate for a true cause and effect relationship.
Endotoxin is a recognised potent activator of various cellular and humoral pathways involved in the generalised inflammatory response. Endotoxins are mostly comprised of lipopolysaccharide, and most of their biological activity resides in the lipopolysaccharide section. The core region of lipopolysaccharide is nearly identical for most strains of Gram-negative bacterial endotoxins. Supranormal levels of naturally occuring endotoxin core antibodies have been associated with a reduction in organ failure in patients following high-risk surgery and on the ITU. However, in two large multicentre trials of patients with presumed Gram-negative infection, the results of giving donor antibodies against the core region of endotoxin have been inconclusive. Active immunisation would be an attractive alternative for patients scheduled for major surgery, and evidence from animal experiments has been encouraging. However, there is no currently available anti endotoxin vaccine. Other potential anti endotoxin strategies undergoing human testing include bactericidal/permeability-increasing protein (BPI), endotoxin-neutralising protein and dextran—polymixin B conjugate, all of which have the ability to protect animals from endotoxin-mediated toxicity.
Systemic inflammatory response syndrome
The organo-protective therapeutic regimens cited above seem to work if used prophylactically, as is the case in major surgery. However, the results of adopting similar regimens in established MODS are rather disappointing. This would suggest that once a systemic inflammatory response is unresponsive to cardiovascular manipulations and antimicrobials then if the progression to MODS is to be avoided organ protection must come from a different line of attack. An increased understanding of the host-derived mediators of the tissue destruction seen in MODS has opened up a whole new field of therapeutic agents directed against them. It is hoped that specific manipulation of key mediators will at least halt the tissue damage in its tracks and hopefully be curative.
Protein cytokines play an important part in the mobilisation, localisation and subsequent activity of leucocytes in the inflammatory reaction. Tumour necrosis factor alpha (TNFa) and the interleukins (IL) have emerged as prime targets for experimental manipulation. In animal models of septic shock, treatment with monoclonal antibodies to TNF-alpha and various interleukins (e.g. IL-1) have improved survival. There are ongoing trials of inhibitors of both TNF-alpha and IL in human sepsis. Unfortunately, the early reports have been inconclusive.
Decreasing cytokine synthesis and secretion
Corticosteroids reduce TNF-alpha mRNA translation in response to a stimulus and thus reduce secretion. Numerous studies have demonstrated the protective effects of corticosteroids in animal models of septic and haemorrhagic shock. The use of low-dose dexamethasone has been shown to improve outcome in paediatric patients with meningitis. From a purely hypothetical viewpoint, steroids should be the answer to the treatment of SIRS. However, two large multicentre, randomised trials of high dose dexamethasone used in the treatment of septic shock failed to demonstrate any improvement in survival.
In 1987 it was reported that the endothelium-derived relaxant factor was identical to the free radical nitric oxide (NO). NO is synthesised from r-arginine by a constitutive enzyme present in the endothelium, which has a physiological role in the control of blood pressure, and by an inducible nitric oxide synthase, which is expressed in vessel walls and phagocytic cells in response to endotoxin or cytokines. NO has a myriad of actions but is predominantly a vasodilator and can modify the neutrophil—platelet interactions that may result in the microvascular occlusion seen in MODS. These effects create a therapeutic dilemma. Should you give NO in an attempt to restore microvascular flow or block its effects to restore the blood pressure in septic shock? In patients with severe acute respiratory distress syndrome inhaled NO has been demonstrated to reduce pulmonary artery pressure and improve pulmonary oxygenation without affecting systemic vascular resistance. However, it has also been demonstrated that blocking the production of NO from its precursor Larginine is possible by the administration of arginine analogues such as NG-monomethyl-r.-arginine (L-NMMA). In animal studies the administration of NO antagonists has been shown to restore the vascular response to catecholamines and improve survival.
Arachidonic acid metabolites
There is a multitude of animal and human evidence to suggest that metabolites of arachidonic acid play key roles in the pathogenesis of MODS, both protective (e.g. prostaglandin E2) and deleterious ones (e.g. leukotrienes and thromboxane). Cyclo oxygenase inhibitors such as ibuprofen or indomethacin, which are nonsteroidal anti-inflammatory drugs, have been shown to reduce tissue damage and improve survival in animal models of sepsis.
Degranulating neutrophils as part of a systemic inflammatory response are said to cause microvascular injury and promote organ dysfunction by the release of destructive enzymes and the generation of oxygen free radicals. Free radical scavengers such as superoxide dismutase, allopurinol and even vitamin C are universally successful in reducing the tissue damage seen in septic and haemorrhagic shock models. Provisional reports from at least two human studies claim success from free radical scavenging.
Contact, coagulation and complement activation
A common clinical feature of SIRS is a coagulopathy. Histologically the microvascular abnormality seen in MODS is not unlike clot. This has led to the assumption that there is a disturbance of the balance between procoagulant and anticoagulant pathways in SIRS that can manifest itself most vividly in the disseminated intravascular coagulation seen in meningococcal meningitis. There are preliminary results suggesting that the administration of clinical concentrates of inhibitors of the contact system such as antithrombin III and C1 -esterase inhibitor may modify the outcome in established SIRS.
Endogenous anti-inflammatory agents
It is now recognised that most proinflammatory acute-phase reactants are balanced by the production of endogenous anti-inflammatory acute-phase reactants. For example, antagonists to soluble IL-1 and TNF-alpha are produced by hepatocytes and released into the circulation, thereby reducing the inflammatory response. The anti-inflammatory cytokine IL-10 is a potent macrophage-deactivating factor, and injection of recombinant IL-10 has been shown to protect mice from endotoxic shock. IL-10 is thought to regulate the effects of other cytokines (e.g. TNF-alpha), rather than block them completely, and therefore has the potential for maintaining optimal balance in the inflammatory system.
Multiple organ dysfunction syndrome
There are no animal or human data to suggest that fully established MODS is treatable. This does not mean that all patients who have MODS die. However, the small percentage who survive have probably done so because supportive care has given them a chance to get better.
Prevention of MODS by the prompt diagnosis and treatment of the primary insult coupled with cardiovascular resuscitation and supportive care has an extremely high success rate in patients who have some hope of long-term survival. However, once the same group of patients has established organ failure the outlook is extremely gloomy.
No comments yet.