Acs0813 Multiple Organ Dysfunction Syndrome

Acs0813 Multiple Organ Dysfunction Syndrome

• 1.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 1 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME John C. Marshall M.D., F.A.C.S., F.R.C.S.C. Approach to Multiple Organ Dysfunction Syndrome The multiple organ dysfunction syndrome (MODS)—also In general terms, the dysfunction of a given organ system can known as progressive systems failure,1 multiple organ failure,2 be described in one of three ways: and multiple system organ failure3—is characterized by progres- 1. As a physiologic derangement (e.g., an altered ratio of arteri- sive but potentially reversible physiologic dysfunction of two or al oxygen tension [PaO2] to fractional inspiration of oxygen more organ systems that arises after resuscitation from an acute [FIO2] or an altered platelet count). life-threatening event. The term MODS was introduced by a 2. As the clinical intervention used to correct that derangement 1991 consensus conference of the American College of Chest (e.g., mechanical ventilation or blood component replace- Physicians (ACCP) and the Society of Critical Care Medicine ment therapy). (SCCM).4 The designation of MODS as a syndrome emphasizes 3. As a discrete clinical syndrome incorporating several descrip- that dynamic alterations in physiologic function in critically ill tive variables (e.g., acute respiratory distress syndrome patients may have common pathophysiologic underpinnings. [ARDS] or disseminated intravascular coagulation [DIC]). However, MODS is as much a paradigm as a syndrome—that is, it represents an approach to the care of the critically ill patient Whichever of these three perspectives is adopted, common that emphasizes intensive monitoring and support of organ sys- pathogenetic mechanisms underlie MODS, and common prin- tem function over specific therapies for isolated disease process- ciples direct its prevention and management. es and that focuses on preventing or minimizing iatrogenic injury RESPIRATORY DYSFUNCTION resulting from ICU interventions. MODS evolves in the wake of a profound disruption of sys- ARDS, initially described in the early 1960s,13,14 is the proto- temic homeostasis.5,6 It was originally described in patients with typical expression of respiratory dysfunction in MODS.15 In its overwhelming infection, multiple injuries, or tissue ischemia; however, it has many overlapping risk factors [see Table 1]. Table 1 Risk Factors for MODS Preexisting illness—in particular, chronic alcohol abuse7—pre- disposes to the development of organ dysfunction in patients Peritonitis and intra-abdominal infections exposed to these risk factors. Infection Pneumonia Necrotizing soft tissue infections Clinical Definitions of Inflammation Pancreatitis Organ Dysfunction Although MODS is readily Injury Multiple trauma recognized by experienced Burn injury clinicians, there is still no clear Ruptured aneurysm consensus on its description Ischemia Hypovolemic shock with respect to either the sys- Mesenteric ischemia tems whose function is deranged, the descriptors that Autoimmune disease best measure that derangement, or the degree of derangement Immune reactions Transplant rejection that constitutes organ dysfunction or failure. Graft versus host disease A systematic review of 30 published clinical studies evaluated the organ systems and variables used to describe MODS.8 Seven Delayed or missed injury organ systems—the respiratory system (all 30 reports), the renal Blood transfusion Iatrogenic factors system (29 reports), the hepatic system (27 reports), the cardio- Injurious mechanical ventilation vascular system (25 reports), the hematologic system (23 Total parenteral nutrition reports), the GI system (22 reports), and the CNS (18 Drug reactions reports)—were included in at least half of the studies. Scoring Intoxication Arsenic intoxication systems from both North America9,10 and Europe11,12 define Drug overdose MODS using six of these seven organ systems, eliminating the GI system because of the declining prevalence of stress-related Thrombotic thrombocytopenic purpura upper GI bleeding and the lack of satisfactory measures of GI Idiopathic factors Hypoadrenalism dysfunction. These scoring systems have many similarities [see Pheochromocytoma Table 2].
• 2.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 2 Recognize susceptible patient Patients at high risk for multiple organ dysfunction syndrome (MODS) are those who have experienced a disruption of systemic homeostasis resulting from one or more of the following: Approach to Multiple Organ • Infection • Immune system activation • Inflammation • Intoxication Dysfunction Syndrome • Injury • Iatrogenic factors • Ischemia Minimal organ dysfunction Prevent progression to MODS by optimizing support of hemodynamic, metabolic, and immunologic function, taking care to minimize iatrogenic injury during the provision of physiologic support. Hemodynamic support Maximize O2 delivery to tissues by the following measures: • Fluid replacement therapy • Vasoactive agents • Inotropic agents • Mechanical ventilation Metabolic support Reverse catabolic state with definitive intervention, including the following: • Debridement of devitalized tissue • Fixation of long bone • Burn wound excision and grafting fractures Provide early nutritional support by the enteral route. If gut function is inadequate, parenteral nutrition should be employed. Immunologic support Prevent nosocomial infection, treat documented infection, and minimize the consequences of injurious host defense responses by such measures as the following: • Timely and appropriate surgical intervention • Limiting breaches of mucosal defenses • Selective, targeted use of antibiotics Organ function is preserved or restored Organ function deteriorates Patient survives. Discharge patient from ICU.
• 3.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 3 Define prevalence, sites, and severity of organ dysfunction Organ dysfunction involves essentially any organ system but is classically identified in the respiratory, renal, cardiovascular, hepatic, hematologic, GI, and neurologic systems. It can be defined as a physiologic derangement, as the clinical support provided, or as a clinical syndrome comprising several abnormalities. Aggregate severity can be quantified by using readily available scores. Significant organ dysfunction Characterize physiologic derangement and institute supportive therapy. Single organ dysfunction Multiple organ dysfunction Search for correctable causes, including the following: Search for correctable causes, including the following: • Exacerbation of preexisting chronic disease • Occult infection • Adverse effects of medical therapy (e.g., • Complications from invasive devices and medications • Missed injuries transfusions, total parenteral nutrition) • Local complications (e.g., fluid overload, pneumothorax, biliary tract obstruction) • Local pathology (e.g., myocardial infarction, pulmonary embolism, pneumonia) Modify supportive care Minimize adverse consequences of supportive care through use of techniques such as pressure-limited ventilation and continuous hemodialysis. Manage infectious complications with local measures and sparing use of antimicrobial agents. Evaluate need for antibiotic therapy or surgical intervention. Antibiotic therapy Surgical intervention Adjuvant therapy • Sparing use of empirical antibiotic therapy Prepare patient for operation: Consider use of adjuvant therapies • Microbiologic diagnosis is usually possible • Preoperative optimization of physiologic function directed against coagulopathy • If organism cannot be isolated, discontinue • Safe, well-organized transport to OR (activated protein C) or acute adrenal antibiotics and repeat cultures • Surgical intervention in ICU when appropriate insufficiency (corticosteroids). and feasible Reevaluate clinical status MODS resolves, and patient’s condition improves MODS persists, and patient’s condition deteriorates or fails to improve despite support and in the absence Transfer patient from ICU. of correctable underlying disease Probability of death is high. Consider withdrawal of life support.
• 4.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 4 Table 2 Multiple Organ Dysfunction (MOD) Score9 Degree of Dysfunction Organ System Indicator of Dysfunction None Minimal Mild Moderate Severe (0) (1) (2) (3) (4) Respiratory PaO2/FIO2 ratio > 300 226–300 151–225 76–150 ≤ 75 Renal Serum creatinine level ≤ 100 µmol/L 101–200 µmol/L 201–350 µmol/L 351–500 µmol/L > 500 µmol/L Hepatic Serum bilirubin level ≤ 20 µmol/L 21–60 µmol/L 61–120 µmol/L 121–240 µmol/L > 240 µmol/L Cardiovascular Pressure-adjusted HR* < 10.0 10.1–15.0 15.1–20.0 20.1–30.0 > 30.0 3 Hematologic Platelet count > 120,000/mm3 81,000–120,000/mm3 51,000–80,000/mm 21,000–50,000/mm3 ≤ 20,000/mm3 Glasgow Coma Scale 15 13–14 10–12 7–9 ≤6 Neurologic score *Calculated as the product of HR and central venous pressure (CVP), divided by mean arterial pressure (MAP): (HR · CVP)/MAP. mildest form, respiratory dysfunction is characterized by tachyp- but supportive care, in the form of dialysis, did not become avail- nea, hypocapnia, and hypoxemia. As lung injury evolves, a com- able until the 1950s. bination of worsening hypoxemia and increased work of breath- Clinical or subclinical renal dysfunction is common in ing necessitates mechanical ventilatory support [see 8:6 Me- MODS. Early-onset renal dysfunction typically results from chanical Ventilator]. hypotension and decreased renal blood flow.The etiology of late- Increased capillary permeability and neutrophil influx are the onset renal failure is multifactorial and includes both pre- earliest pathologic events in ARDS. As the acute inflammatory renal factors (e.g., decreased cardiac output and hypovolemia) process resolves, further lung injury results both from the and the cumulative renal effects of nephrotoxic agents (e.g., process of repair, which involves fibrosis and the deposition of medications and radiocontrast material).24 Intrarenal vasocon- hyaline material, and from further lung trauma, resulting from striction results in a reduction in the glomerular filtration rate, positive pressure mechanical ventilation.16 hypoxic or oxidative injury to tubular epithelial cells, and des- Lung involvement in ARDS is inhomogeneous, with areas of quamation of injured cells into the tubules, causing leakage of fil- functional and aerated alveoli interspersed with areas of non- trate back into the renal interstitium and evoking neutrophil- functional alveoli.17 The distribution of injury reflects the seque- mediated inflammation that causes further local tissue injury.25 lae of care in the intensive care unit: consolidation occurs in the Intrarenal shunting of blood flow, coupled with occlusion of the posterior dependent regions of the lung, and cystic changes renal microvasculature by thrombi or aggregated blood cells, fur- develop from overdistention by the ventilator in the antidepen- ther contributes to ischemia and physiologic dysfunction. The dent regions.18 situation may be further aggravated by renal circulatory changes Impaired lung function is reflected in a reduced PaO2. To resulting from vasoactive agents administered to treat shock and ensure adequate oxygen delivery to the tissues, mechanical ven- by increased intra-abdominal pressure consequent to massive tilation must be instituted and FIO2 increased. The ratio of PaO2 fluid resuscitation. Histologic studies show acute tubular necro- to FIO2, therefore, provides a sensitive and objective measure- sis with disruption of the basement membrane, patchy necrosis ment of the degree to which oxygenation is impaired and so is a of the renal tubules, interstitial edema, and tubular casts; these reliable measure of physiologic respiratory dysfunction.19 microscopic changes correlate poorly with functional impair- Mechanical ventilation reflects the clinical intervention triggered ment.26 Activated neutrophils have also been implicated in the by impaired oxygenation, and the additional criteria for ARDS— pathogenesis of ARF,27,28 as has the induction of apoptosis in bilateral lung infiltrates and a normal pulmonary capillary wedge renal epithelial cells.29 pressure—serve to exclude such primary causes of acute hypox- Renal dysfunction in MODS is reflected physiologically in a emia as pulmonary embolism, atelectasis, and congestive heart decreased urine output, biochemically in a rising serum creati- failure. By consensus, ARDS is defined as a PaO2/ FIO2 ratio lower nine level, and therapeutically as the introduction of exogenous than 200 mm Hg in association with bilateral fluffy pulmonary renal replacement therapy or dialysis. infiltrates and a pulmonary capillary wedge pressure lower than HEPATIC DYSFUNCTION 18 mm Hg.20 ARDS is a robust model for the complex interactions that Hepatic dysfunction after trauma, like ARF, was first result in MODS. Lung injury in ARDS is the outcome of an described during World War II [see 8:9 Hepatic Failure].30 Two interaction between an insult, a susceptible host, and the clinical clinical syndromes have been described. The first, ischemic therapeutic response, and its severity reflects not only the degree hepatitis, or shock liver, characteristically follows an episode of of the initial insult but also various poorly defined genetic influ- profound hypotension with splanchnic hypoperfusion. Early ele- ences in the host21 and the inadvertent adverse consequences of vations of aminotransferase levels are striking and may be asso- the mode of respiratory support employed.22 ciated with an increased international normalized ratio and hypoglycemia; centrilobular necrosis is evident histologically. RENAL DYSFUNCTION Successful resuscitation of the shock state results in rapid nor- Acute renal failure (ARF) [see 8:7 Renal Failure] was first malization of the biochemical abnormalities.31 The second syn- described as a significant clinical problem during World War II,23 drome, ICU jaundice, is much more common than ischemic
• 5.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 5 HEMATOLOGIC DYSFUNCTION hepatitis and typically evolves many days after the inciting phys- iologic insult. Conjugated hyperbilirubinemia is a prominent The most common hematologic abnormality of critical illness feature, whereas elevation of aminotransferase levels and alter- is thrombocytopenia, which occurs in approximately 20% of all ations of hepatic synthetic function are less pronounced.32 ICU admissions.46,47 Causes include increased consumption, Histologic features include intrahepatic cholestasis, steatosis, intravascular sequestration, and impaired thrombopoiesis sec- and Kupffer cell hyperplasia. The pathogenesis is multifactorial ondary to suppression of bone marrow function. In addition, and includes ongoing hepatic ischemia, total parenteral nutrition heparin-induced thrombocytopenia resulting from antibodies to (TPN)–induced cholestasis, and drug toxicity. complexes of heparin and platelet factor 4 develops in as many An increased serum bilirubin level is the most commonly rec- as 10% of patients receiving heparin.48 The most fulminant ognized feature of the hepatic dysfunction of MODS. Although expression of hematologic dysfunction in MODS is DIC, which extracorporeal support devices have been used for patients with is characterized by derangements in platelet numbers and clot- end-stage liver disease, the hepatic dysfunction of critical illness ting times and the presence of fibrin degradation products in is not considered to be life-threatening in itself, and no specific plasma.49 The coagulopathy of critical illness is complex, involv- supportive therapy is indicated. ing multiple alterations in the biochemical mediators of coagu- lation and resulting in a shift to a procoagulant state.50 CARDIOVASCULAR DYSFUNCTION Mild anemia is common in critical illness, though the nature Both peripheral vascular and myocardial function are altered of abnormalities in red cell production and removal are less well in MODS. Characteristic changes in the peripheral vasculature characterized in this setting.51 Transient leukopenia may develop include a reduction in vascular resistance and an increase in in response to an overwhelming inflammatory stimulus, but neu- microvascular permeability, resulting in a hyperdynamic circula- trophilia is much more commonly encountered; total lympho- tory profile and peripheral edema. Both alterations jeopardize tis- cyte counts are reduced. Abnormalities in white cell populations sue oxygenation—reduced vascular resistance by facilitating reflect, at least in part, altered expression of apoptosis, which is shunting in the microvasculature and edema by increasing the inhibited in the neutrophil52 but accelerated in lymphoid cells.53 distance across which oxygen carried in the blood must diffuse to GASTROINTESTINAL DYSFUNCTION reach the cell. Shunting also occurs as a result of occlusion of the microvasculature by thrombi and aggregates of nondeformable Upper GI hemorrhage after burn injury was first described by red cells33; it is signaled by a reduction in arteriovenous oxygen Curling in 1842.54 Stress bleeding was once a relatively common extraction and an increase in mixed venous oxygen saturation complication, but improved techniques of resuscitation and (SmvO2). Biventricular dilatation with a reduction in the right and hemodynamic support, earlier diagnosis of infection, and the left ventricular ejection fractions has been described.34 Right ven- widespread use of stress ulcer prophylaxis have reduced the fre- tricular dysfunction is particularly prominent, perhaps as a con- quency of this event, to the point where it now is seen in fewer sequence of increased pulmonary vascular resistance secondary than 4% of ICU admissions.55 Other manifestations of GI dys- to concomitant lung injury.35 Finally, a loss of normal heart rate function in MODS include ileus and intolerance of enteral feed- variability characterizes advanced cardiovascular dysfunction.36 ing,56,57 pancreatitis,58 and acalculous cholecystitis.59 The cardiovascular dysfunction of MODS is apparent clini- OTHER ORGAN SYSTEM DYSFUNCTION cally as increased peripheral edema with hypotension that is refractory to volume challenge and therapeutically in the use of MODS is associated with functional abnormalities of virtually vasoactive agents to support the circulation. Nitric oxide (NO) every organ system. Endocrine abnormalities include impaired has been implicated in both the peripheral vasodilatation37 and glucose regulation with hyperglycemia and insulin resistance60 the myocardial depression38 associated with critical illness. and hypercortisolemia with impaired responsiveness to adreno- corticotropic hormone (ACTH) stimulation.61,62 The sick euthy- NEUROLOGIC DYSFUNCTION roid syndrome, characterized by reductions in serum T3, with or Abnormalities of both central and peripheral nervous system without an increase in reverse T3 levels and a normal T4 level, is function are common in critical illness. CNS dysfunction occurs in another manifestation of the endocrine dysfunction of MODS.63 as many as 70% of critically ill patients, typically presenting as a Numerous derangements of immune function have been reduced level of consciousness without localizing signs. Its patho- described in MODS patients. Cell-mediated immunity is im- physiology is incompletely understood. Postulated mechanisms paired, as reflected by anergy to delayed hypersensitivity recall include the direct effects of proinflammatory mediators on cerebral skin testing64 and impaired in vitro lymphocyte proliferative re- function, the development of vasogenic cerebral edema, areas of sponses.65 The development of ICU-acquired infections caused cerebral infarction related to hypotension, and alterations in the by organisms of low intrinsic virulence can also be considered a blood-brain barrier resulting in changes in the composition of the manifestation of impaired immunity in MODS.66 interstitial fluid.39 Electroencephalography typically shows one of Abnormal wound healing also occurs in MODS. Common four patterns indicating increasingly abnormal activity: diffuse manifestations of impaired wound healing are the failure of an theta wave rhythms, intermittent rhythmic delta waves, triphasic open wound to develop satisfactory granulation tissue and the delta waves, and suppression or burst-suppression patterns.39 development of decubitus ulcers.67 Peripheral nervous system dysfunction, also known as critical illness polyneuropathy, is also common in MODS, though its clinical presentation tends to be more subtle than that of CNS Quantification of Organ Dysfunction dysfunction.40-42 Peripheral nervous system dysfunction may pre- Physiologic instability is the major indication for ICU admis- sent as failure of weaning from mechanical ventilation43 or as limb sion, and support of failing organ function is the ICU’s raison weakness with relative sparing of the cranial nerves. Endoneural d’être. The degree of physiologic derangement present at the edema and axonal hypoxia44 contribute to its pathogenesis, as do time of ICU admission is a potent determinant of ICU sur- the iatrogenic sequelae of neuromuscular blockade.45 vival,68,69 and irreversible organ dysfunction is the preeminent
• 6.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 6 100 scores).77 Alternatively, the aggregate severity of organ dysfunc- tion over time can be quantified by summing the worst values over time in each of the component systems. Such an approach 80 permits quantitation of attributable ICU morbidity as the differ- 28-Day Mortality (%) ence between the aggregate score and the score at baseline—thus identifying that component of ICU morbidity that can be pre- 60 vented by an effective ICU intervention. Finally, morbidity and mortality can be combined into a single value by using a mor- 40 tality-adjusted score that assigns a maximum number of points plus 1 to any patient who dies [see Table 3]. 20 Prevention of Organ Dysfunction in Critically Ill Patients 0 0 1–2 3–4 5–6 7–8 9–10 11–12 13–14 15–16 17–18 19–20 21+ Acute organ dysfunction is Baseline MOD Score the most common indication for admission to an ICU, and Figure 1 Increasing severity of organ dysfunction is directly any patient with significant correlated with increasing ICU mortality. physiologic instability is at risk for MODS. A number of risk mode of ICU death.6,70 Formal quantification of the severity of factors for the development of organ dysfunction have been iden- physiologic derangement or of the evolution of organ dysfunc- tified [see Table 1]: these reflect a common pathogenesis for the tion over time is not generally incorporated into individual syndrome through the activation of an innate immune response to patient care in the ICU. However, validated scoring systems have tissue injury. proved invaluable in describing patient populations, stratifying The first priority for optimal ICU care is to halt the progression patients for entry into clinical trials, and assessing ICU morbid- of existing organ dysfunction while preventing the development of ity in patient groups. new organ dysfunction. Prevention of organ dysfunction is perhaps There are a number of published systems for quantifying the best approached from the perspective of optimization of hemody- severity of organ dysfunction in the critically ill.9,10,12.66,71-74 These namic, metabolic, and immune homeostasis. There are a number systems are all structurally similar, evaluating dysfunction in each of interventions for which level 1 evidence of efficacy in reducing of six or seven organ systems on a numerical scale in which more mortality or preventing organ dysfunction exists [see Table 4]. points are assigned for greater degrees of physiologic severity; they vary primarily with respect to the variables used to describe OPTIMIZING HEMODYNAMIC HOMEOSTASIS dysfunction. A representative example of such a scoring system is The ability of the heart to pump blood is determined by (1) the the Multiple Organ Dysfunction (MOD) score [see Table 2].9 preload delivered to the right atrium, (2) the intrinsic contractility The numerical scores can be obtained and applied in a vari- of the myocardium, and (3) the afterload against which the heart ety of ways.75 Scores can be calculated on the day of ICU admis- must work—all of which may be deranged in critical illness. First, sion or at the start of the institution of a novel therapy during the reduction of intravascular volume as a consequence of hemor- ICU stay; such scores provide a measure of baseline illness sever- rhage, third-space loss, and increased microvascular permeability ity and correlate in a graded manner with the risk of ICU mor- reduces preload. Second, circulating mediators and NO depress tality [see Figure 1]. Scores can also be calculated daily, allowing myocardial contractility and thus impair the heart’s intrinsic the clinician to track net clinical improvement or deterioration pumping ability. Third, reduced vascular tone, mediated by NO, over time76 and to assess the progression or resolution of organ reduces afterload. The first two abnormalities reduce cardiac out- dysfunction (expressed as the area under the curve for the daily put; the third increases it but may, by altering resistance gradients Table 3 Approaches to Measuring Severity of MODS75 Objective Approach Uses To quantify baseline severity of Calculate organ dysfunction score on day of admission To establish baseline severity (e.g., for entry criteria for a organ dysfunction (admission MODS) clinical trial) or to ensure comparability of study groups To quantify severity of organ To determine intensity of resource utilization or evolution or Calculate score on particular ICU day (daily MODS) dysfunction at point in time resolution of organ dysfunction at discrete point in time To measure aggregate severity of Sum individual worst scores for each organ system over To determine severity of physiologic derangement over organ dysfunction over ICU stay defined time interval (aggregate MODS) defined time interval (e.g., ICU stay) To quantify new organ dysfunction Calculate difference between aggregate and admission To measure organ dysfunction attributable to events occur- arising after ICU admission scores (delta MODS) ring after ICU admission To provide combined measure of Adjust aggregate score so that all patients dying receive To create single measure that integrates impact of morbidity morbidity and mortality maximal number of points (mortality-adjusted MODS) in survivors and mortality for nonsurvivors
• 7.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 7 Table 4 ICU Interventions That Reduce intrinsic myocardial dysfunction can all alter the pressure at which optimal preload is obtained. Mortality or Attenuate Organ Dysfunction In practice, resuscitation should be titrated to optimize the balance between several parameters rather than targeting any one Objective Intervention parameter. It is sobering to recognize that current sophisticated approaches to resuscitation using the pulmonary arterial catheter Resuscitation Early goal-directed resuscitation243 have not been shown to yield net clinical benefit and may, in fact, Prophylaxis Selective digestive tract decontamination90 cause harm [see 8:4 Cardiopulmonary Monitoring].82 Hemodynamic resuscitation is most effective when it is early Restrictive transfusion strategy107 and rapid. A randomized trial of goal-directed therapy for sep- Low tidal volume ventilation22 sis, using a protocol comprising fluid administration, transfu- ICU support Daily wakening144 sion, and vasoactive support titrated to SmvO2 as measured from Tight glucose control60 the superior vena cava through a central venous catheter, found Enteral feeding86 that mortality was reduced from 46.5% to 30.5% when patients Activated protein C146 were resuscitated according to protocol in the emergency Mediator-targeted therapy Corticosteroids148 department within hours of their initial presentation. On the Antibody to TNF244 other hand, studies of goal-directed therapy initiated in the ICU have generally failed to demonstrate any evidence of benefit.83,84 Optimization of oxygen delivery presupposes the ability to in the microvasculature, alter nutrient flow to the tissues. oxygenate blood adequately in the lungs. Increased pulmonary The first priority in supporting cardiovascular homeostasis, capillary permeability, atelectasis, altered consciousness, and therefore, is to restore intravascular volume by administering flu- intrinsic lung disease can all reduce oxygen uptake in acutely ill ids.There is no convincing evidence that any particular resuscita- patients. Support can be provided through the administration of tion fluid is superior in all patients, though crystalloid is associat- oxygen to the spontaneously breathing patient, through the use ed with a lower mortality in trauma patients.78 Either normal of positive pressure ventilation by mask, or through endotracheal saline or lactated Ringer solution is an appropriate choice. The intubation and mechanical ventilation. Positive pressure ventila- volume of fluid needed to restore optimal preload may be signifi- tion can cause further lung injury, however, particularly when cant, reflecting not only acute losses but also the effective expan- the lung has been rendered vulnerable by early acute lung injury. sion of the vascular compartment because of vasodilatation and Limiting tidal volume during mechanical ventilation to 6 ml/kg the loss of fluids into the extravascular compartment because of has been shown to improve survival in patients with early increased capillary permeability. Blood loss should be corrected ARDS.22 by transfusing red cells, preferably in fresh, leukocyte-depleted OPTIMIZING METABOLIC blood. When hypotension is refractory to fluid administration, HOMEOSTASIS vasoactive agents, including vasopressors (e.g., dopamine and norepinephrine) and inotropes (e.g., dobutamine, epinephrine, The acute response to and amrinone) may help increase blood flow to the tissues.79 stress and injury is a complex, Given that the goal of hemodynamic stabilization is to sup- coordinated process charac- port organ function rather than to restore physiologic or bio- terized by increases in levels chemical normalcy, the best measures of the success of resusci- of catecholamines, glucocorti- tation are those that reflect either return of function (in particu- coids, antidiuretic hormone, lar, urine output) or adequate blood flow to the tissues (e.g., and hormones that regulate SmvO2 or lactate concentration). Each of these measures, howev- intermediary metabolism, including insulin, glucagon, and er, has shortcomings of which the clinician must be aware. Urine growth hormone [see 8:25 Metabolic Response to Critical Illness].6 output may be decreased because of intrinsic renal damage even The activation of this response results in a predictable series of in the face of adequate renal flow. SmvO2 may be artefactually metabolic alterations, including retention of salt and water, high because of shunting and abnormalities of oxygen uptake in increased production of glucose, enhanced lipolysis, increased the microvasculature. Lactate concentration is relatively insensi- protein catabolism, and an altered pattern of hepatic protein syn- tive to mild degrees of inadequate oxygen delivery and may be thesis known as the acute-phase response, characterized by elevated in patients with liver disease. increased synthesis of C-reactive protein, alpha1-anti-trypsin, and Gastric production of CO2 as measured with a gastric fibrinogen and reduced synthesis of albumin. tonometer has been proposed as a means of evaluating splanch- Metabolic prophylaxis of MODS is directed toward reversal nic blood flow, but the benefits of tonometry in improving out- of the stimuli responsible for the catabolic hormonal milieu and come are unproven.80 Microvascular flow can also be directly toward the provision of adequate biochemical substrate at a time visualized in the tongue or another exposed mucosal surface by of increased metabolic demand. Early definitive surgical therapy using orthogonal polarization spectral imaging.81 Neither of in the form of debridement of devitalized tissue, burn wound these approaches has been widely used to guide resuscitation. excision and grafting, and rigid fixation of long bone fractures Blood pressure is widely used as an index of the initial ade- can attenuate the postinjury hypermetabolic state and minimize quacy of resuscitation, but pressure measurements may not reli- the subsequent development of MODS, though the benefits of ably reflect flow in the microvasculature, particularly when sys- early definitive therapy must be weighed against the additional temic vascular resistance is low. Measurement of central venous stress of blood loss and hemodynamic instability. In the face of or pulmonary capillary wedge pressures provides an estimate of overwhelming injury, a policy of damage control to permit sta- the preload to the heart, though factors such as positive pressure bilization of the patient in the ICU is associated with an ventilation, the extent of capillary leakage in the lungs, and improved clinical outcome.85
• 8.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 8 Nutritional support should be provided by the enteral route if Although nosocomial infections in critically ill patients usual- possible [see 8:22 Nutritional Support]. Enteral nutrition is feasi- ly arise from endogenous reservoirs, pathogens may also spread ble in most patients, particularly if feedings are initiated early. from patient to patient and from environment to patient. Certain The administration of even small quantities of enteral nutrition organisms—in particular, Acinetobacter, Xanthomonas, and Le- is considered advisable, even if it must be supplemented by some gionella—are transmitted through aqueous sources in the ICU, degree of parenteral nutrition. There is increasing evidence that and the isolation of these organisms from a critically ill patient is immunologically enhanced enteral formulas can yield better evidence of a potential problem in environmental infection con- clinical outcomes than standard enteral formulas.86 trol. Hand washing is an important but underutilized mode of Close regulation of glucose levels in accordance with an inten- infection prevention in the ICU [see 1:1 Prevention of Postopera- sive policy of monitoring and insulin administration has been tive Infection]. There is no clear evidence that protective isolation shown to improve clinical outcome.60 On the other hand, there of critically ill patients warrants the increased costs and is no evidence that administration of growth hormone offers any increased demands on nursing staff. significant benefit.87 The role of immunomodulation in the prophylaxis of MODS remains undefined. At present, there is no defined role for OPTIMIZING IMMUNOLOGIC chemoprophylaxis of MODS beyond the specific effects of HOMEOSTASIS drugs such as heparin (prevention of deep vein thrombosis Infection is an important [DVT]) and H2 receptor antagonists (prevention of upper GI risk factor for MODS, but the bleeding). converse is equally true: patients with MODS are at significantly increased risk for Evaluation of the Patient infection. This risk arises as a with Organ Dysfunction consequence both of impair- SINGLE ORGAN DYSFUNCTION ment of normal host defense mechanisms and of colonization with potentially infectious nosocomial pathogens [see 8:16 Single organ dysfunction Nosocomial Infection]. suggests local disease and Of the numerous derangements of normal immune function should trigger a search for with which critical illness is associated, impairment of mucosal potentially correctable causes defenses is probably the most important (and certainly the most in the organ system involved. Isolated organ dysfunction may preventable). Mucosal defenses are breached by surgical incisions reflect preexisting chronic disease or a local problem such as fluid and by invasive devices, including intravascular catheters, urinary overload, atelectasis, biliary tract obstruction, or elevated intracra- catheters, and endotracheal and nasogastric tubes. Limiting the nial pressure. Complications related to invasive devices or the number of such devices in use and paying rigorous attention to adverse effects of medications are common causes of single organ their insertion and maintenance are important for minimizing dysfunction; the diagnosis is often presumptive, established on the nosocomial infection rates.88 Gastric acid plays a primary role in basis of clinical improvement after discontinuance of the agent or maintaining the relative sterility of the stomach. Antacids ablate removal of the device. Finally, single organ dysfunction may indi- this defense and are a recognized risk factor for nosocomial pneu- cate acute disease in the involved organ, such as myocardial infarc- monia; they should not be used for stress ulcer prophylaxis. The tion, pulmonary embolism, or bone marrow suppression. declining incidence of clinically significant stress bleeding in the Acute respiratory dysfunction, for example, may be caused by contemporary ICU suggests that prophylaxis should be limited to pneumonia, atelectasis, pleural effusion, pneumothorax, or pul- patients who are at increased risk for stress ulceration.55 monary embolism. Central venous and pulmonary arterial Cytoprotective agents (e.g., sucralfate) appear to have no signifi- catheters may induce tachyarrhythmias as a result of mechanical cant advantages over H2 receptor antagonists in reducing the risk irritation of the conducting system. Isolated renal dysfunction of ventilator-associated pneumonia and are less efficacious in pre- may be a consequence of abdominal compartment syndrome or venting bleeding89; accordingly, H2 receptor antagonists appear to of the nephrotoxic effects of medications (e.g., acute tubular be the prophylactic agents of choice. necrosis caused by aminoglycosides and interstitial nephritis An alternative strategy for preventing pathologic gut coloniza- caused by penicillins and cephalosporins). Occasionally, renal tion and nosocomial infection involves prophylactic administra- dysfunction arises from a postrenal cause, such as blockage of a tion of a combination of systemic antibiotics (e.g., cefotaxime) Foley catheter. and topical nonabsorbed antibiotics (e.g., tobramycin, polymyx- Medications are important causes of liver dysfunction in the in, and amphotericin B). This approach, known as selective critically ill patient. Erythromycin, ketoconazole, and haloperidol, decontamination of the digestive tract (SDD), has proved effec- for example, can induce cholestatic liver injury. Thrombocyto- tive in reducing nosocomial infection rates and even ICU mor- penia is an important adverse effect of a number of medications, tality90; the effect is particularly evident in surgical patients who including heparin flushes to maintain the patency of arterial lines. receive both systemic and topical therapy.91 A decreased level of consciousness is usually the result of the Enteral feeding is beneficial in preventing nosocomial infec- poorly characterized metabolic encephalopathy of critical illness; tion. Systemic antibiotics suppress the indigenous flora of however, it is necessary to rule out local causes such as meningi- mucosal surfaces, promoting pathologic colonization with resis- tis, encephalitis, brain abscess, and subdural hematoma. Excessive tant organisms.92 Therefore, use of antimicrobial agents in criti- or prolonged use of narcotics or sedative-hypnotics may lead to cally ill patients must be selective and targeted, and the use of sustained alterations in level of consciousness, particularly when broad-spectrum empirical therapy should be minimized by reg- hepatic or renal function is impaired. Nondepolarizing muscle ular reviews of culture and sensitivity results and restrictions on relaxants (e.g., vecuronium) may cause prolonged neuromuscu- antibiotic prescription practices. lar blockade and peripheral neuropathy.45
• 9.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 9 MULTIPLE ORGAN DYSFUNCTION aid in establishing or excluding the diagnosis.96,97 MODS is rarely caused by urinary tract infections or device- Although it has been sug- related bacteremias, though these conditions are common in gested that there is a charac- patients with significant organ dysfunction. Clostridium difficile teristic temporal sequence in colitis or disseminated fungal infection may also present as dete- the development of MODS riorating organ function in critically ill patients.98 [see Table 5],3,9 the clinical course tends to be variable Iatrogenic factors MODS can be considered the quintes- and depends in part on the sential iatrogenic disorder, reflecting both the successes and the criteria used to define organ system dysfunction (see above).The failures of contemporary ICU practice. On one hand, the syn- specific pattern of dysfunction is much less important than the drome arose only because the supportive care available today course of the evolving syndrome. Resolution of dysfunction sug- permits the prolonged survival of critically ill patients who, in an gests an appropriate response to specific and supportive therapy. earlier era, would have died rapidly; on the other hand, poten- Worsening of dysfunction, on the other hand, should prompt a tially avoidable iatrogenic factors contribute prominently to the search for potentially correctable causes and a reevaluation of evolution of MODS. the methods of supportive care in use. Not infrequently, a treat- Technical or judgmental errors often set the stage for able cause of evolving MODS is found; often, however, no cause MODS.99-101 Whenever a patient manifests unexplained organ is evident. When no specific cause of the deterioration can be failure in the postoperative period, the surgeon must consider the identified, therapy should focus on optimizing supportive mea- possibility of an iatrogenic complication—for example, a missed sures to limit iatrogenic injury either until the patient recovers intestinal perforation in a trauma victim, a leak from a tenuous or, alternatively, until a considered decision is made that contin- anastomosis, or left colon ischemia after aneurysmectomy. uing active care is futile. Many of the therapeutic interventions that are the mainstay Search for Correctable Causes of ICU care have the potential to cause local and remote organ injury. In the experimental setting, mechanical ventilation with Occult infection Uncontrolled infection, particularly infec- high tidal volumes and low levels of positive end-expiratory tion arising within the abdomen, is an important risk factor for pressure (PEEP) can induce both pulmonary and remote MODS.2,3,93 The development of otherwise unexplained organ organ injury.102,103 A multicenter, randomized, controlled trial dysfunction should trigger a careful radiologic search for an confirmed that mechanical ventilation of patients with acute occult intra-abdominal focus.94 However, MODS also develops lung injury in accordance with a lung-protective strategy (i.e., in patients with pneumonia58 and other life-threatening infec- a tidal volume of 6 ml/kg) significantly improves survival22 and tions, and it sometimes evolves in patients in whom no infectious attenuates the local and systemic release of proinflammatory focus can be identified.66,95 mediators [see 8:6 Mechanical Ventilator].16 Oxygen in When MODS develops in the postoperative period, a careful high concentrations can produce pulmonary damage, probably search for infection must be undertaken, concentrating in partic- as a result of the generation of toxic oxygen intermediates [see ular on the operative site and on any invasive devices used. With 8:26 Molecular and Cellular Mediators of the Inflammatory appropriate attention to the clinical possibilities, aided by ultra- Response].104 sonography and CT scanning, the presence or absence of signif- Blood transfusion has been implicated in the development of icant intra-abdominal pathology can usually be established. Local organ dysfunction, an effect that occurs independent of the wound exploration may suggest the possibility of occult intra- effects of shock, blood loss, and fluid resuscitation.105,106 A mul- abdominal infection through the demonstration of impaired ticenter, randomized, controlled trial demonstrated a significant wound healing or fascial dehiscence or through the isolation of reduction in the severity of new organ dysfunction in a heteroge- typical intestinal microflora from a wound infection.The diagno- neous population of critically ill patients when transfusion was sis of pneumonia in intubated ICU patients is notoriously diffi- withheld unless the hemoglobin concentration was less than 70 cult; however, the use of quantitative techniques (e.g., protected g/L (7 g/dl).107 The age of the blood administered may be an specimen brush bronchoscopy and bronchoalveolar lavage) can underappreciated factor in defining optimal transfusion strate- gies.The effects of blood transfusion on splanchnic blood flow as Table 5 Temporal Evolution of MODS3,9 measured with a gastric tonometer are significantly dependent on the age of the transfused blood. Transfusion of blood that is more than 12 days old can have an adverse impact on oxygen Time from ICU Admission to Onset of System Significant Dysfunction (days) delivery.108 TPN can also contribute to the de novo development of organ Respiratory 1–2 dysfunction.TPN-associated alterations in hepatic function with intrahepatic cholestasis and fatty infiltration are relatively com- Hematologic 3 mon and are manifested by elevated aminotransferase and alka- line phosphatase levels.109 TPN may also give rise to glucose Central nervous 4 intolerance and can aggravate ventilatory impairment through Cardiovascular 4 increased CO2 production. In patients with borderline pul- monary function, this additional CO2 production may prevent Hepatic 5–6 weaning from ventilatory support.110 Parenteral nutrition is also associated with higher rates of postoperative and nosocomial Renal 4–11 infections after multiple trauma.111 Gastrointestinal 10–14 Medications—in particular, analgesics, sedatives, and antibi- otics—have also been associated with evolving organ dysfunction.
• 10.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 10 Support of Patients with ventilatory support, it may be necessary to accept an arterial oxy- Established Organ gen saturation (SaO2) as low as 80%. Dysfunction The fundamental challenge CARDIOVASCULAR SUPPORT • in providing intensive support- Tissue oxygen delivery (D O2) is a function of three variables— ive care to patients with estab- cardiac output, hemoglobin concentration, and SaO2 (oxygen dis- lished organ dysfunction is solved in the plasma makes • only a negligible contribution). how to support physiologic Tissue oxygen consumption (VO2) is a function of cardiac out- function while minimizing new put and oxygen extraction (defined as the difference between iatrogenic organ dysfunction. As more is understood about the arterial and venous oxygen content). In practice, multiple factors consequences of various treatment strategies, it is increasingly can impair oxygen delivery and consumption, and it is important apparent that optimizing function is not synonymous with maxi- that the clinician recognize these. mizing function and that the clinician must be acutely aware of the Oxygen uptake in the tissues is an entirely passive process, potential for causing more harm than good. resulting from the diffusion of oxygen toward the relatively hypoxic extravascular space along an oxygen saturation gradient VENTILATORY SUPPORT that is highest in the microvasculature and lowest at the cell.This Although ventilatory support is generally provided by endo- passive diffusion can be reduced if there is interstitial edema, tracheal intubation and mechanical ventilation, noninvasive pos- which makes the concentration gradient less steep, or if blood itive pressure ventilation may be appropriate for patients with flow through the microvasculature is rapid. Alternatively, a milder degrees of respiratory failure. A small randomized trial of reduction in the resistance of small arterioles may impede the patients with cardiogenic pulmonary edema found that when diversion of blood into the microvasculature. Moreover, nutrient compared with conventional administration of oxygen by mask, vessels in the microcirculation may be occluded by aggregates of noninvasive positive pressure ventilation shortened the time to neutrophils, platelets, and aged (and thus less deformable) red resolution of respiratory failure and reduced the need for subse- blood cells. The net result of these abnormalities is the shunting quent endotracheal intubation.112 Whether this approach is use- of oxygenated blood from the arterial side of the circulation to ful in patients with early ARDS is less clear, however; it may be the venous side. This phenomenon is readily detected through associated with a higher risk of complications.113 measurement of SmvO2, which is about 70% in normal persons Endotracheal intubation and positive pressure ventilation con- but typically is much higher in patients with sepsis and organ stitute the mainstay of support for critically ill patients with res- dysfunction. piratory failure. In unstable patients, it is best to use a controlled Paradoxically, each of the interventions commonly used to ventilatory mode (e.g., pressure control ventilation) rather than a increase tissue oxygen delivery can also decrease it. Fluid resus- spontaneous breathing mode (e.g., pressure support ventilation). citation can increase cardiac output by increasing preload, but in Oxygenation can be optimized through the use of PEEP, a patients with altered capillary permeability, it can create edema, maneuver that may also decrease the accumulation of interstitial thereby lengthening the distance across which oxygen must dif- fluid and minimize ventilator-associated lung injury.114 Ventilation fuse.Vasopressors can raise cardiac output by increasing periph- with large tidal volumes and high peak inspiratory pressures con- eral vascular resistance, but at the cost of reducing flow through tributes to lung injury, and it has been shown that the survival of nutrient vessels in the microcirculation. Inotropes, on the other ARDS patients can be improved by using low tidal volumes (~ 6 hand, directly increase cardiac output, albeit at the cost of ml/kg).22 Pressure-controlled ventilatory techniques limit peak increased myocardial oxygen consumption, but agents such as airway pressures to a maximum predetermined level, optimizing dobutamine may lead to further shunting by decreasing periph- gas exchange by inverting the inspiration-to-expiration ratio (I/E) eral vascular resistance. from its normal value of 1:2 to 1:1 or higher and by changing the Currently, intensive invasive monitoring of critically ill shape of the inspiratory flow curve (normally square) to one in patients with organ dysfunction is employed less frequently than which flow initially is rapid, then decelerates [see 8:6 it once was. The benefits of such monitoring remain somewhat Mechanical Ventilator].115 Although oxygenation can be main- uncertain. For example, a 1996 study suggested that the use of a tained with low tidal volumes, ventilation is jeopardized, with the pulmonary arterial catheter was associated with a 24% increase result that CO2 levels rise (so-called permissive hypercapnia).116 in mortality—not, presumably, because of complications of the Hypercapnia per se does not appear to be deleterious117; indeed, catheter itself but rather because the decisions made on the basis animal studies suggest that increased levels of CO2 may be inde- of the data provided led to greater harm than benefit.125 pendently beneficial to critically ill patients.118 For patients with Although this estimate of harm may be exaggerated, there is lit- refractory hypoxemia, high-frequency oscillation appears to be a tle countervailing evidence of benefit to justify routine use of promising ventilatory mode.119,120 pulmonary arterial catheters. A 2003 study of 1,994 high-risk Inhaled NO is selectively delivered to ventilated lung seg- patients undergoing major elective surgery found that Swan- ments and may effect early improvement of oxygenation in Ganz catheterization and preoperative optimization did not ARDS patients121; whether this early physiologic effect translates reduce mortality but was associated with a significant increase in into an improved clinical outcome is unknown. Extracorporeal the risk of pulmonary embolism.84 lung support by means of extracorporeal membrane oxygenation RENAL SUPPORT or extracorporeal CO2 removal can be lifesaving in patients with isolated severe respiratory failure that is refractory to other forms Renal replacement therapy in critically ill patients with of respiratory support.122,123 These techniques are resource MODS serves three functions: intensive, however, and have not been convincingly shown to 1. Regulation of fluid and electrolytes in patients in whom nor- yield better outcomes than conventional mechanical ventila- mal renal function is compromised and altered capillary per- tion.124 If the patient remains hypoxemic despite optimization of meability has led to total body fluid overload with edema.
• 11.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 11 2. Removal of products of metabolism, medications, and other seen in patients with underlying cardiovascular disease.143 toxins that the failing kidneys are unable to clear. Thrombocytopenia is corrected by transfusion of platelet con- 3. Removal of circulating mediators of inflammation. centrates, but this generally is done only if the platelet count drops below 20,000/mm3. Coagulation factors can be replaced The first two are classic indications for dialysis, though the by giving fresh frozen plasma or cryoprecipitate. therapeutic objectives may differ; the third lies more in the realm Adequate analgesia and sedation are essential components of of promising experimental therapy. the care of MODS patients; however, the natural desire to allevi- Fluid overload is a common consequence of hemodynamic ate pain and discomfort can lead to oversedation. A policy of daily resuscitation during the early stages of acute illness. It results awakening can reduce morbidity and shorten the ICU stay.144 from increased capillary permeability, peripheral vasodilatation with expansion of the intravascular compartment, and impaired PHARMACOLOGIC THERAPY TARGETING HOST RESPONSE renal function. The use of continuous renal replacement thera- Experimental studies implicate an activated inflammatory pies to titrate fluid balance and reduce uremia is conceptually response in the pathogenesis of MODS. Despite extensive eval- appealing, but the benefits remain unproven. Both individual uation of a variety of novel strategies to target the host response randomized trials126,127 and a systematic review128 failed to show in sepsis, to date, only two approaches have demonstrated an that early and aggressive continuous renal replacement improved ability to improve survival. clinical outcome. On the other hand, a multicenter randomized Activated protein C is an endogenous anticoagulant molecule trial of more than 400 ICU patients with ARF showed that high- that inhibits factors V and VIII; in addition to its anticoagulant flow ultrafiltration (at a rate of 35 ml/kg/hr or higher) increased activities, it exerts significant anti-inflammatory activity [see 8:26 survival,129 and a prospective study demonstrated that daily (as Molecular and Cellular Mediators of the Inflammatory Response].145 opposed to alternate-day) intermittent hemodialysis improved Activated protein C has been produced as a recombinant protein survival and hastened the resolution of ARF.130 Another system- (drotrecogin alfa activated) and has been evaluated in a multicen- atic review suggested that imbalances between study groups ter randomized trial involving 1,690 patients with severe sepsis. In might have masked a potential benefit of therapy associated with this trial, treatment resulted in a 6.1% improvement in 28-day early continuous hemodialysis.131 survival146 and a more rapid resolution of cardiovascular, respira- Whether it significantly improves outcome or not, early con- tory, and hematologic dysfunction.147 The benefit appears to be tinuous renal replacement therapy does facilitate early manage- greatest in patients who have more severe illness (reflected in an ment of the patient with MODS by permitting the removal of elevated APACHE II [Acute Physiology and Chronic Health fluid, and it is generally better tolerated by hemodynamically Evaluation II] score or a greater number of dysfunctional organs), unstable patients than is intermittent hemodialysis. Evidence community-acquired infection, or coagulopathy. that dialytic therapy can accelerate the clearance of circulating Critical illness is associated with multiple abnormalities of mediators of sepsis is scant.132 endocrine function, including reduced responsiveness to endoge- nous glucocorticoids,62 a state that predicts an increased risk of SUPPORT OF OTHER ORGANS ICU mortality.61 In a 2002 study, administration of pharmaco- Enteral nutritional support has been shown to reduce the rate logic doses of corticosteroids (50 mg of hydrocortisone every 6 of infectious complications in patients with multiple trau- hours and 50 µg of fludrocortisone) to patients with refractory ma111,133 and in those with pancreatitis.134 A systematic review of septic shock and an impaired response to a short-course ACTH 15 randomized trials found that early enteral feeding reduced stimulation test reduced mortality by 10%.148 In contrast, earlier the infectious complication rate and shortened the ICU stay studies of high-dose corticosteroids in more heterogeneous without affecting mortality.135 The use of immunologically groups of patients with sepsis found no evidence of benefit.149 enhanced enteral formulas appears to be associated with a fur- Systematic reviews have suggested that neutralization of tumor ther reduction in infectious complications, ventilator days, and necrosis factor (TNF) or interleukin-1 (IL-1) can improve out- length of hospital stay.136 Paralytic ileus makes the provision of come in sepsis,150 but neither of these approaches is clinically enteral feeding more difficult. Erythromycin, which is a motilin available at present. Other anticoagulant or anti-inflammatory secretagogue, can facilitate bedside placement of enteral feeding strategies have been suggested but remain unproven. tubes137 and accelerate gastric emptying.138 Techniques for extracorporeal support of the failing liver have MODS AND ICU-ACQUIRED INFECTION been described, but their use is generally limited to a few centers Infection is a risk factor for organ dysfunction, but the con- with a particular interest in liver failure and organ transplanta- verse is equally true: organ dysfunction is a risk factor for noso- tion.139,140 Unlike primary liver failure, the hepatic dysfunction of comial infection, with the risk increasing as the severity of organ MODS does not lead to life-threatening organ system insuffi- dysfunction increases.66 The typical isolates are microbes of low ciency and rarely calls for specific support. Hypoalbuminemia is intrinsic pathogenicity, including coagulase-negative Staphy- common in MODS, occurring as a consequence of increased lococcus, Enterococcus, and Candida species and gram-negative vascular permeability, loss through the GI tract, and reduced organisms such as Pseudomonas and Enterobacter.151 These organ- hepatic synthesis from the activation of an acute-phase response. isms commonly colonize the upper GI tract of the critically ill Although hypoalbuminemia is associated with increased ICU patient,152 they emerge under antibiotic pressures, and they form morbidity and mortality, there is no convincing evidence that colonies on invasive devices—all of which may explain why they albumin supplementation improves clinical outcome.141 emerge as predominant infecting species in this setting. Studies A randomized trial of transfusion strategies in the ICU of SDD have demonstrated that preventing such infections demonstrated that organ function was improved by a restrictive reduces ICU morbidity and mortality,90,153 but there is scant evi- transfusion policy that withheld transfusion unless the hemoglo- dence that aggressive antimicrobial therapy to treat suspected bin level dropped below 70 g/L,107 a conclusion supported by a nosocomial infection improves outcome. In fact, two reports large European multicenter observational study.142 Benefit is also from 2000 suggested that a more restrictive approach to the pre-
• 12.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 12 scription of antimicrobial agents reduced mortality and morbid- Table 6 Prognosis in MODS ity.97,154 Worsening organ dysfunction should prompt a careful search for untreated foci of infection, but empirical therapy Number of Failing Systems Mortality (%) should be used cautiously; if such therapy is started, it should be discontinued promptly as culture data are obtained. Often, mere 0 3 removal of a colonized device (e.g., a central line or a urinary 1 15–30 2 50–60 catheter) amounts to definitive therapy for these infections. 3 60–100 The association of organ dysfunction with occult intra-abdom- 4 70–100 5 100 inal infection3,94 stimulated a period of enthusiasm for the prac- tice of so-called blind laparotomy—that is, laparotomy undertak- en to identify and treat an intra-abdominal infectious focus with- positive indicators according to those criteria had a mortality of out radiographic evidence that infection is present.155,156 The uni- close to 100%; today, a majority of such patients survive.93 formly disappointing results of such intervention,157 coupled with Organ dysfunction is potentially recoverable when the factors improvements in diagnostic imaging techniques, led to abandon- responsible for the persistence or progression of MODS can be ment of this approach except in certain unusual circumstances reversed. Identifying these factors, treating them appropriately, (e.g., clinically compelling evidence of a surgically correctable and providing optimal physiologic support can prove a daunting problem, suspicion of visceral ischemia, or the absence of the challenge, and it is often advisable to consider seeking indepen- appropriate imaging facilities). It goes without saying that the dent advice or transporting the patient to a center with the clin- classic physical findings of peritonitis—particularly when no ical expertise and facilities to manage the multidisciplinary prob- abdominal operation has been done—may be the sole indication lems faced by MODS patients. Given that MODS often evolves for surgical exploration. Moreover, in a complicated postopera- as a consequence of medical misadventure, early consultation or tive patient transferred from another institution because of wors- referral may be a sound approach from a medicolegal perspec- ening organ dysfunction, repeat laparotomy may legitimately be tive as well. On the other hand, it is a common contemporary considered a component of the admission physical examination. ICU scenario that MODS evolves and worsens despite optimal care, necessitating a decision whether to continue or discontinue active care. Outcome WITHDRAWAL OF LIFE SUPPORT PROGNOSTIC INDICATORS The most common mode of death for a patient with advanced The prognosis of MODS is MODS is limitation or withdrawal of life support in the face of directly related to the severity a persistent failure to respond to full, aggressive ICU care.160 The of the underlying organ dys- decision to withdraw or withhold life support is a complex one, function, which can be and there are considerable differences of professional opinion expressed in terms of either and practice regarding how best to make it.161 Factors that must the number of failing sys- be taken into consideration include the nature of the underlying tems3,73,158 [see Table 6] or the global severity of dysfunction as disease, the patient’s premorbid health status, the wishes of the determined by an organ dysfunction score [see Figure 1]. It must patient and the family regarding long-term ICU care, the be emphasized, however, that prognostic indicators reflect the patient’s ultimate prospects for an independent existence, and expected outcome of a group of patients and are of limited use the presence of active problems amenable to medical therapy. in making decisions about the care of an individual patient. Although end-of-life deliberations may be difficult for medical Moreover, the prognostic weight of these scales reflects stan- staff and family alike, careful and realistic consideration of the dards of care prevalent at a particular time and in a particular expectations of all involved can facilitate a decision to discontin- clinical setting. For example, at the time when Ranson’s criteria ue active therapy in a manner that is dignified and humane were developed,159 patients with pancreatitis and six or more rather than adversarial. Discussion MODS: Evolution of a Syndrome MODS.4 MODS is perhaps the classic instance of the new dis- The first ICU was established in Baltimore in the late ease paradigm, in that it develops only in patients who would 1950s.162 Its development marked much more than a simple have died without medical intervention and evolves because of advance in medical technology. The improvements in fluid the inadvertent consequences of that intervention. resuscitation achieved during World War II, followed by the Earlier reports had described the physiologic failure of dis- development of techniques of positive pressure mechanical ven- crete organ systems after trauma or acute illness. Stress-related tilation, hemodialysis, and central venous catheterization over upper GI bleeding was described in 184254 and trauma-related the subsequent decade, had set the stage for an entirely new dis- renal23 and hepatic30 dysfunction during World War II. ease paradigm—that of a disease that arose only in patients who Description of respiratory failure awaited the widespread use of would have died in the absence of resuscitation and exogenous mechanical ventilators, but a process termed high-output respi- support. The need for that support to sustain life became the ratory failure was described in patients with peritonitis in metaphor that described this new disorder, originally described 1963,13 anticipating the classic description of ARDS 4 years as sequential systems failure163 and now generally known as later.14 In each of these reports, however, the physiologic organ
• 13.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 13 abnormality was viewed as an isolated problem of a single organ, lipopolysaccharide (LPS), derived from the cell wall of gram-neg- albeit one that could have broader secondary consequences. The ative bacteria reproduces the physiologic features of sepsis in suggestion by Baue1 in 1975 that each organ abnormality was human volunteers,169 with larger doses producing life-threatening simply the local manifestation of a systemic process set the stage organ dysfunction.170 Moreover, endotoxin can be detected in the for attempts to identify common systemic pathologic processes circulation of patients at risk for MODS—not only patients with and hence to develop therapies that were not merely supportive sepsis,171,172 but also those who have experienced traumatic173 or but also targeted fundamental mechanisms in the pathogenesis thermal injury174 and those who are undergoing cardiopulmonary of MODS.The search for such therapies is, admittedly, still in its bypass or repair of an abdominal aortic aneurysm.175 infancy. Not only is the disease process complex, but our ability Animal studies, however, have shown that the sequelae of to describe and characterize it is limited as well. endotoxemia are an indirect consequence of the activation of MODS is invariably preceded by evidence of systemic activa- host innate immunity rather than a direct cytopathic effect of the tion of an adaptive host stress response to infection or tissue endotoxin molecule.The C3h HeJ strain of mice arose through a injury. This response, which includes changes in cardiorespirato- spontaneous mutation of the parental C3h HeN strain, involving ry function, increased microvascular permeability, evidence of an alteration in a single gene product. This defect, later recog- activation of innate immune mechanisms, and alterations in nized as a point mutation in the gene encoding Toll-like receptor intermediary metabolism, is termed sepsis when caused by infec- 4 (TLR4),176 conferred complete resistance to endotoxin lethal- tion and the systemic inflammatory response syndrome (SIRS) ity in C3h HeJ mice. Studies involving bone marrow irradiation when considered independent of cause.4,164 SIRS is mediated and crossover transplants of bone marrow cells between C3h through the release of a complicated network of host-derived HeN and C3h HeJ mice showed that endotoxin sensitivity was mediator molecules (see below).The name notwithstanding, des- transferred with bone marrow cells177 and confirmed that the ignation of SIRS as a syndrome may be somewhat presumptu- sequelae of endotoxin challenge arose indirectly, through the ous. There is no discrete or invariant pattern of clinical manifes- activity of marrow-derived cells from the host. Microarray stud- tations that identifies patients with activation of this complex ies have demonstrated that literally hundreds of genes are response, nor is there convincing evidence that the response is expressed in macrophages, endothelial cells, and neutrophils common to all patients who meet the clinical criteria for SIRS. after stimulation by LPS.178,179 The importance of this enor- It has been suggested that it is also possible to define a com- mously complex response is underlined by the observation that pensatory anti-inflammatory response syndrome (CARS) or a the lethality of murine endotoxemia can be prevented by neu- mixed acute response syndrome (MARS)165; however, these tralizing any one of several dozen of these gene expressions “syndromes” are more conceptual entities than they are diseases before endotoxin challenge.145 that can be diagnosed and treated.166 Endotoxin interacts with cells of the host innate immune sys- tem through TLR4.TLR4 is one of a family of 10 TLRs that have evolved to recognize danger signals in the extracellular environ- Theories of Pathogenesis ment and to activate cells to mount an appropriate response to a Organ dysfunction must ultimately be the consequence of the threat.180 TLRs recognize not only microbial products (e.g., malfunctioning or death of cells in that organ. Although cellular endotoxin) and components of the wall of gram-positive bacteria derangements are readily documented under experimental con- (e.g., lipoteichoic acid and peptidoglycan) (TLR2) but also bac- ditions, it remains largely unknown how these derangements terial DNA (TLR9) and even heat-shock proteins and structural translate into the physiologic changes that define the clinical syn- components of damaged cells (TLR2) [see Table 7]. Thus, the drome. Cellular dysfunction may reflect altered patterns of syn- response evoked appears not to be specific for the stimulus that thetic function, either because of activation of alternate patterns elicited it, just as the clinical syndrome of systemic inflammation of gene expression or because of relative cellular oxygen defi- and organ dysfunction is not unique to patients with infection ciency from defective cellular respiration (so-called cytopathic but can also be seen in association with other causes of tissue hypoxia).167 Mechanisms of fibrosis and repair may alter the nor- injury.164,181 mal cellular anatomic relationships and thereby impair function. The binding of endotoxin to TLR4 triggers a cascade of intra- Finally, cell death may result from mechanical or biochemical cellular signaling pathways leading to the expression of hundreds injury of sufficient severity to prevent oxidative metabolism and of genes whose products mediate innate immunity [see Figure 2]. produce anatomic disruption of the cell or necrosis, but it may The resulting alterations in normal patterns of cellular protein also occur through the activation of apoptosis (programmed cell synthesis are profound: not only does the initial stimulus trigger death). Each of these abnormalities has been described in a complex response, but the newly synthesized protein products patients with sepsis, and each has multiple overlapping causes. of this response also, in turn, are capable of acting on the cell to induce a further cascade of mediator molecules. Any attempt to INFECTION AND HOST SEPTIC RESPONSE classify the mediators involved is inevitably arbitrary and sim- The earliest descriptions of MODS emphasized its association plistic. It is useful, however, to consider the response as involving with occult infection,58,94 prompting the hypothesis that organ (1) early inflammatory mediators (e.g., TNF and IL-1), (2) late dysfunction arises through the direct effects of one or more inflammatory mediators (e.g., macrophage inhibitory factor and microbial toxins. However, the observations that MODS could high-mobility group box [HMGB]–1), (3) counterinflammatory also develop in patients with no identifiable focus of infection95 and tissue repair mediators (e.g., IL-10 and transforming growth and that treatment of infection did not necessarily reverse the factor [TGF]–β), (4) enzymes involved in the regulation of non- syndrome168 suggested that infection may be a cause of organ protein inflammatory mediators (e.g., inducible NO synthase, dysfunction in critical illness but is not necessarily the funda- phospholipase A2, and platelet-activating factor acetylhydro- mental mechanism. lase), (5) acute-phase reactants, and (6) cell surface adhesion or Microbial products, independent of bacterial viability, can signaling molecules (e.g., intercellular adhesion molecule evoke the clinical features of sepsis. Injection of endotoxin, or [ICAM]–1 and tissue factor).
• 14.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 14 Table 7 Toll-like Receptors and Their Ligands then able to extravasate by degrading the tight junctions between endothelial cells, probably through the action of the neutrophil enzyme elastase; this process can cause injury to the cell as the Toll-like Receptor Ligand(s) neutrophil transits.192 Once localized to the site of challenge, the neutrophil releases a variety of molecules (e.g., proteases and TLR1 Mycobacterium leprae lipopeptide, Borrelia outer reactive oxygen intermediates) that directly injure microorgan- surface protein isms and host tissue alike. Opsonization of bacteria by immu- Peptidoglycan, lipoteichoic acid, bacterial lipopro- noglobulin or complement enables the neutrophil to phagocytose TLR2 tein, Mycoplasma lipopeptide, zymosan, CMV, and kill invading pathogens. Activated neutrophils have been M. leprae lipopeptide, lipoarabinomannans, injured cells implicated in the pathogenesis of pulmonary,193,194 hepatic,195 GI,196 and renal28 injury in experimental models of inflamma- TLR3 Double-stranded RNA tion. Similarly, though ablation of fixed tissue macrophages increased the number of bacteria isolated from an animal model TLR4 Endotoxin, HSP60, β-glucan, neutrophil elastase of peritonitis, it nonetheless reduced the severity of septic symp- TLR5 Flagellin toms and improved survival.197 Other studies have implicated natural killer (NK) cells198,199 TLR6 Mycoplasma lipopeptide and CD8-positive T cells200 in the lethality of experimental sep- sis, though it is unclear whether lethality is a consequence of TLR7 Imidazoquinolones, guanine ribonucleosides direct cellular cytotoxicity or of the activation of other biologic TLR8 Unknown processes by secreted products of these cells.201 TLR9 Bacterial CpG DNA 1. LBP + LPS TLR10 Unknown Blockade of any of these molecules prevents lethality in mice that are subsequently challenged with endotoxin. None of these 2. mediators, however, is directly cytotoxic, which suggests that their role is to activate effector mechanisms of injury further Peptidoglycan 3. downstream. The identity of those downstream mechanisms of cellular dysfunction or death is still speculative, though a number Bacterial DNA of attractive possibilities have been proposed. GENETIC FACTORS IN HOST RESPONSE A Scandinavian population-based study of causes of premature mortality in adoptees revealed that genetic factors play a signifi- mCD14 TLR4 cant role in the outcome of infection. When one of an adoptee’s biologic parents died before the age of 50, the adoptee faced a six- fold increase in the risk of infectious mortality; this increased risk TLR2 was substantially greater than that associated with premature death from cardiovascular disease or cancer.182 It is now known Signal that polymorphisms in genes for TLRs and cytokines are common Transduction TLR9 in the general population183 and are associated with both altered Pathways Tyrosine and expression of the gene product and enhanced susceptibility to Threonine/Serine sepsis and organ dysfunction. A mutation in the gene for TLR4 Kinases Increased has been associated with a significantly increased risk of gram- NFκB negative infection in ICU patients.184,185 Polymorphisms in the genes for CD14,186 TNF,187,188 heat shock protein 70,189 IL-10,190 IL-6 and IL-1 receptor antagonist191 all have been associated with TNF-α greater degrees of organ dysfunction and a worse clinical outcome IL-1 Transcription in critical illness. TISSUE INJURY MEDIATED BY PHAGOCYTIC CELLS OF INNATE Figure 2 During lipopolysaccharide (LPS) signaling, (1) LPS IMMUNE SYSTEM binds to LPS-binding protein (LPB). (2) This complex is delivered Polymorphonuclear neutrophils and monocytes form the first to membrane-bound CD14 (mCD14). (3) The CD14-LPS complex line of innate host defenses against invading microorganisms and interacts with Toll-like receptor 4 (TLR4), which initiates the intracellular signal transduction pathway. The signal transduction injured tissue. Neutrophils are recruited to the site of tissue inva- pathway leads to NFκB translocation into the nucleus, with sion or injury by locally released chemokines (e.g., IL-8). They increased transcription of proinflammatory cytokines. TLR2 is adhere to the endothelium of the microvasculature through the known to signal membrane compounds of gram-positive bacteria interaction of the neutrophil adhesion molecule CD11b with such as peptidoglycan. TLR9 recognizes bacterial DNA. (IL— ICAM-1 on the endothelial cell [see 8:26 Molecular and Cellular interleukin; NFκB—nuclear factor κB; TNF-α—tumor necrosis Mediators of the Inflammatory Response]. Adherent neutrophils are factor–α)
• 15.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 15 ENDOTHELIAL INJURY inactivates thrombin and inhibits factors IXa, XIa, and XIIa. The network of endothelial cells lining the large and small ves- Levels of circulating antithrombin are reduced in sepsis. sels of the vascular tree is enormous, encompassing an area 600 Administration of recombinant antithrombin failed to improve times larger than that of the skin.202 Far from being a passive con- survival in a large multicenter study of patients with severe sep- duit for blood cells and plasma, the endothelium contributes sis, though interaction with heparin may have masked a thera- actively to the initiation, localization, and resolution of inflamma- peutic effect.211 tion.203 Circulating inflammatory mediators evoke an endothelial TFPI is synthesized by endothelial cells in response to cytokines cell response characterized by upregulation of adhesion molecules and shear stress. It too is a serine protease inhibitor whose target is such as ICAM-1 and vascular cell adhesion molecule (VCAM), the complex of tissue factor, factor VIIa, and factor Xa that initi- expression of TLR2 and tissue factor, and shedding of endothelial ates the coagulation cascade. Recombinant TFPI has been evalu- cell thrombomodulin.203,204 These changes result in neutrophil ated as a therapy for sepsis; however, a large multicenter trial failed localization and local activation of coagulation with secondary tis- to show any survival benefit from such therapy.212 sue injury. In addition, healthy endothelial cells appear to play a The interactions between coagulation and inflammation are role in limiting host injury during inflammation, as indicated by complex. Protein C binds to a receptor expressed on endothelial the observation that mice with a genetic deletion of syndecan-4 (a cells and neutrophils, and prevention of this interaction results in proteoglycan involved in endothelial cell adhesion and signaling) a disseminated inflammatory process.146 Binding of activated pro- exhibit increased mortality and exaggerated release of IL-1 after tein C to its receptor in brain endothelium inhibits endothelial cell endotoxin challenge.205 apoptosis and is neuroprotective during cerebral ischemia.213 Postmortem studies of patients with ARDS show that ICAM-1 APOPTOSIS is upregulated throughout the lung, whereas VCAM is expressed in larger vessels,206 and that circulating markers of endothelial acti- Apoptosis, or programmed cell death, is a process through vation are present in critically ill patients with organ failure.207 which cellular structural elements and DNA are degraded and Moreover, diffuse endothelial injury is suggested by the presence residual cellular constituents are converted into membrane- of circulating endothelial cells in the blood of patients in septic bound vesicles (apoptotic bodies) that are cleared by mac- shock.208 Therapeutic strategies that target the adhesive interac- rophages. Phagocytosis of apoptotic bodies prevents the uncon- tions of lymphocytes and endothelial cells have shown promise trolled release of cellular constituents and so prevents an inflam- against inflammatory conditions such as psoriasis and Crohn dis- matory response to the injured cell.214 Apoptosis is also an intrin- ease; studies of agents targeting the interaction of neutrophils with sically anti-inflammatory process, however, in that phagocytosis endothelial cells have so far yielded disappointing results.209 of an apoptotic cell by the macrophage evokes the expression of anti-inflammatory genes such as those for TGF-β and IL- INTRAVASCULAR COAGULATION 10.215,216 Apoptosis, like cell replication, is a physiologic process Although plasma contains all the factors necessary to induce that is essential for normal growth and development; either coagulation and formation of a fibrin clot, intravascular coagula- excessive or inadequate apoptosis can produce disease. tion is limited in healthy persons by the absence of a trigger and Increased rates of apoptosis of colonic epithelial cells and the presence of endogenous anticoagulant mechanisms.This bal- splenic lymphocytes are seen in autopsy specimens from patients ance is disrupted in persons experiencing systemic inflammation, dying as a consequence of sepsis or multiple trauma.217 resulting in microvascular coagulation and obstructing oxygen Conversely, whereas circulating neutrophils survive less than a delivery to the cell. Injury and inflammation can upregulate day in vivo in healthy persons, neutrophils isolated from the endothelial cell tissue factor expression and activate the coagula- blood of patients with sepsis show both phenotypic features of tion cascade by catalyzing the conversion of factor VII to factor activation and prolonged survival (a consequence of the inhibi- VIIa. Sequential activation of circulating coagulation factors tion of a constitutively expressed apoptotic program).52 Animal results in the conversion of prothrombin to thrombin and, in studies have shown that inhibition of lymphoid cell apoptosis turn, the conversion of fibrinogen to fibrin. The activation of improves survival after septic challenge.218 Conversely, induction coagulation is inhibited by three important endogenous antico- of neutrophil apoptosis improved survival in a rodent model of agulant pathways: the protein C pathway, the antithrombin path- intestinal ischemia-reperfusion injury.219 The potential role of way, and the tissue factor pathway inhibitor (TFPI) pathway. apoptosis in the pathogenesis of MODS is further suggested by Each of these is impaired in critical illness, leading to a net pro- animal studies of ventilator-induced lung injury, which demon- coagulant state. strated that injurious mechanical ventilation strategies can Protein C is synthesized in the liver and circulates as an inac- induce renal epithelial cell apoptosis and biochemical evidence of tive precursor. It is activated in the vascular tree through its inter- renal dysfunction.220 action with thrombomodulin on the endothelial cell; activated GASTROINTESTINAL DYSHOMEOSTASIS protein C blocks thrombin generation through its inhibitory interactions with factors V and VIII.210 Hepatic production of Because MODS not infrequently evolves despite apparently ade- protein C is impaired in sepsis. Moreover, shedding of endothe- quate control of infection or other inciting triggers, it has been sug- lial cell thrombomodulin results in a reduction in the activation gested that the GI tract may serve as the unseen motor of the patho- of available protein C, whereas acute-phase reactants such as logic state,221-224 both as a reservoir of microorganisms and their α1-antitrypsin accelerate its degradation. The importance of this products and as a mechanism for the evolution of inflammation. anticoagulant mechanism is underlined by a 2001 study of The GI tract of a healthy human being harbors upward of 600 recombinant activated protein C in patients with severe sepsis, in different microbial species, and bacteria outnumber human cells which treated patients experienced a significant improvement in 10 to 1.225 The composition of the GI flora is stable over a per- 28-day survival and those with the greatest degrees of organ dys- son’s lifetime, but intercurrent disease can produce profound function benefited most.147 changes in bacterial numbers and patterns of colonization. In crit- Antithrombin is a serine protease inhibitor that binds to and ically ill surgical patients, the normally sterile proximal GI tract
• 16.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 16 TWO-HIT HYPOTHESIS becomes heavily colonized with the same organisms that pre- dominate in nosocomial ICU-acquired infection—Staphylo- A complementary hypothesis of MODS pathogenesis sug- coccus epidermidis, Pseudomonas, Candida, and Enterococcus. gests that an acute insult, such as infection or trauma, primes the Colonization is significantly associated with the development of host so that a subsequent, relatively trivial, insult produces a nosocomial infection with the same organism.152 Extensive stud- markedly exaggerated host response.234 Such a model would ies in animal models have shown that acute insults (e.g., endo- account for the severity of MODS developing late after multiple toxemia, peritonitis, pneumonia, trauma, burns, biliary tract trauma181 as well as the substantial morbidity associated with obstruction, malnutrition, and lack of enteral feeding) can induce nosocomial infection in the ICU.151 Studies in animal models bacterial overgrowth in the gut and translocation of enteric bac- have revealed that previous hemorrhagic shock leads to an exag- teria to regional lymph nodes of the peritoneal cavity.226 Similar gerated response to subsequent endotoxin challenge.235 stimuli result in translocation in humans,227,228 and prevention of colonization with topical nonabsorbed antibiotics reduces the incidence of nosocomial infections, including bacteremias.153 MODS and Complexity Theory Detection of circulating endotoxin after burns174 or major vascu- None of the various theories of pathogenesis described so far lar surgery175 suggests that endotoxin is also absorbed from the is adequate to explain the evolution of the clinical syndrome of gut under conditions of altered gut barrier function. MODS. These models are not mutually exclusive; rather, they The role of the gut in MODS is not limited to that of a micro- reflect differing perspectives on a profoundly deranged state of bial reservoir. The GI tract and adjacent structures constitute systemic homeostasis—one that has evolved only because the the largest aggregation of immune cells in the body. Normal health care team has intervened to subvert an otherwise lethal interactions between the indigenous flora and the gut immune process. The enormously complicated interactions among mul- system serve to prevent the generation of immune responses to tiple predisposing patient factors, a series of physiologic insults, luminal antigens. Loss of these tonic inhibitory interactions may, and an endogenous response effected via multiple cell types and in turn, contribute to a state of systemically activated inflamma- many hundreds of biochemical mediators is best modeled by tion. Accordingly, hemorrhagic shock results in significantly ele- means of complexity theory.236,237 vated levels of TNF and IL-6 in portal venous blood,229 and por- According to the precepts of complexity theory, a complex tal endotoxemia replicates systemic features of MODS, such as system cannot be understood through isolated analysis of its hypermetabolism,230 impairment of cell-mediated immunity,231 component parts: it can be understood only through an appre- and activation of coagulation.232 Ablation of the hepatic Kupffer ciation of the various interactions of those parts. The result of cell population has improved survival in animal models of peri- these multiple interactions is neither a state of anarchy nor a lim- tonitis.197 Mesenteric lymph has been shown to contain factors itless series of unpredictable outcomes but, rather, a series of sta- that can evoke inflammatory changes in organs remote from the ble responses known as an emergent order. This stable order is gut.233 Thus, dysfunction of normal microbiologic and immuno- dependent on the interactions of all its parts but is much more logic homeostasis in the GI tract and the liver results in the gut than their simple sum. It is characterized both by a resilience to serving as a secondary source of the stimuli that induce and per- potentially disruptive external influences and by a degree of petuate MODS. intrinsic variability. Loss of intrinsic variability is a marker of a A clinically relevant role for the gut in the pathogenesis of failing or diseased system. The healthy human heart exhibits MODS is supported by the results of randomized trials of normal variability in rate and rhythm, which is lost in patients SDD.153 with significant congestive heart failure.238 Loss of intrinsic heart rate and blood pressure variability is also evident in patients with septic shock.239,240 Table 8 Characteristics of Linear The idea that concepts inherent in complex nonlinear systems and Complex Systems241 can be applied to the understanding of MODS is compelling.241 Clinicians are most familiar with linear models of disease, and Assumptions of Linear Assumptions of Complex the assumptions on which these models are based generally lead Systems Systems to effective clinical responses. For example, diabetes results from the deficiency of a single protein and is treated effectively by Mediators have unique, consis- Mediators have redundant, variable, tent biologic effects and context-dependent effects replacing that protein; cholangitis is the consequence of obstruc- tion of the bile duct and is treated by relieving that obstruction. Antagonism of effect is mediated Antagonism of effect is not attribut- For linear diseases such as these, a single abnormality produces by specific inhibitor able to single mediator or process the clinical phenotype, and definitive therapy involves correcting that abnormality. Disorders of biologic pathways such as the Biologic processes occur Biologic processes occur sequentially concurrently coagulation cascade represent a somewhat more elaborate vari- ation on the same theme. For example, pathologic coagulation Biologic response is reliably pre- Biologic response is not consistently resulting in DVT can be treated by means of a variety of differ- dicted by measurement of predicted by measure of single responsible mediator mediator ent strategies aimed at disrupting the clotting cascade. Similarly, hypertension can be managed by means of various pharmaco- Modulation of biologic process Modulation of process is not pre- logic strategies that modify different aspects of the regulation of occurs in dose-dependent fash- dictably dose-dependent; small vasomotor tone. ion; small intervention has small perturbation may have large effect effect Complex disorders such as MODS, however, are not so pre- dictable, and the therapeutic modulation required is much Normal behavior of system results Normal integrated behavior of sys- more difficult [see Table 8]. The consequences of manipulating tem results in physiologic variabili- in physiologic stability ty or oscillations over time a particular mediator may be highly context dependent, vary- ing with the nature of the insult, the predisposition of the host,
• 17.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 17 and the timing of administration.145 Moreover, intervention making a normal or supranormal physiologic state the thera- carries the potential cost of inadvertent iatrogenic harm. A peutic target in a critically ill patient whose homeostasis is pro- sobering consequence of high-quality clinical trials performed foundly disrupted may, on occasion, be detrimental rather than in the ICU over the past decade has been the realization that beneficial.22,107,242 References 1. Baue AE: Multiple, progressive, or sequential sys- Crit Care Med 149:818, 1994 tions. Crit Care Med 24:1408, 1996 tems failure: a syndrome of the 1970s. Arch Surg 21. Marshall RP, Webb S, Bellingan GJ, et al: 41. Hund E: Neurological complications of sepsis: 110:779, 1975 Angiotensin converting enzyme insertion/deletion critical illness polyneuropathy and myopathy. J 2. Eiseman B, Beart R, Norton L: Multiple organ polymorphism is associated with susceptibility Neurol 248:929, 2001 failure. Surg Gynecol Obstet 144:323, 1977 and outcome in acute respiratory distress syn- 42. de Letter MA, Schmitz PI, Viseer LH, et al: Risk 3. Fry DE, Pearlstein L, Fulton RL, et al: Multiple drome. Am J Respir Crit Care Med 166:646, factors for the development of polyneuropathy system organ failure: the role of uncontrolled 2002 and myopathy in critically ill patients. Crit Care infection. Arch Surg 115:136, 1980 22. Brower RG, Matthay MA, Morris A, et al:Ventila- Med 29:2281, 2001 4. Bone RC, Balk RA, Cerra FB, et al: ACCP/SCCM tion with lower tidal volumes as compared with 43. Coronel B, Mercatello A, Couturier JC, et al: Consensus Conference. Definitions for sepsis and traditional tidal volumes for acute lung injury and Polyneuropathy: potential cause of difficult wean- organ failure and guidelines for the use of innova- the acute respiratory distress syndrome. N Engl J ing. Crit Care Med 18:486, 1990 tive therapies in sepsis. Chest 101:1644, 1992 Med 342:1301, 2000 44. Bolton CF: Neuromuscular complications of sep- 5. Deitch EA: Multiple organ failure: pathophysiolo- 23. Bywaters EGL, Beall O: Crush injuries with sis. Intensive Care Med 19(suppl 2):S58, 1993 gy and potential future therapy. Ann Surg impairment of renal function. Br Med J 1:427, 1941 45. Segredo V, Caldwell JE, Matthay MA, et al: 216:117, 1992 Persistent paralysis in critically ill patients after 6. Beal AL, Cerra FB: Multiple organ failure syn- 24. Morris JA Jr, Mucha P Jr, Ross SE, et al: Acute long-term administration of vecuronium. N Engl J drome in the 1990’s: systemic inflammatory posttraumatic renal failure: a multicenter perspec- Med 327:524, 1992 response and organ dysfunction. JAMA 271:226, tive. J Trauma 31:1584, 1991 46. Baughmann RP, Lower EE, Flessa HC, et al: 1994 25. Lameire N, Vanholder R: Pathophysiologic fea- Thrombocytopenia in the intensive care unit. 7. Moss M, Parsons PE, Steinberg KP, et al: Chronic tures and prevention of human and experimental Chest 104:1243, 1993 alcohol abuse is associated with an increased inci- acute tubular necrosis. J Am Soc Nephrol 12:S20, dence of acute respiratory distress syndrome and 2001 47. Gando S, Nanzaki S, Kemmotsu O: Disseminated severity of multiple organ dysfunction in patients intravascular coagulation and sustained systemic 26. Tilney NL, Lazarus JM. Acute renal failure in sur- inflammatory response syndrome predict organ with septic shock. Crit Care Med 31:869, 2003 gical patients: causes, clinical patterns and care. dysfunctions after trauma: application of clinical 8. Marshall JC: Multiple organ dysfunction syn- Surg Clin North Am 63:357, 1983 decision analysis. Ann Surg 229:121, 1999 drome (MODS). Clinical Trials for the Treatment 27. Lauriat S, Linas SL: The role of neutrophils in of Sepsis. Sibbald WJ, Vincent J-L, Eds. Springer- 48. Aster RH: Heparin-induced thrombocytopenia acute renal failure. Semin Nephrol 18:498, 1998 and thrombosis. N Engl J Med 332:1374, 1995 Verlag, Berlin, 1995, p 122 28. Heinzelmann M, Mercer-Jones MA, Passmore 49. van der Poll T, de Jonge E, Levi M, et al: Patho- 9. Marshall JC, Cook DJ, Christou NV, et al: JC: Neutrophils and renal failure. Am J Kidney Multiple organ dysfunction score: a reliable genesis of DIC in sepsis. Sepsis 3:103, 1999 Dis 34:384, 1999 descriptor of a complex clinical outcome. Crit 50. Marshall JC: Inflammation, coagulopathy, and the Care Med 23:1638, 1995 29. Faraco PR, Ledgerwood EC, Smith KGC: pathogenesis of the multiple organ dysfunction Apoptosis and renal disease. Sepsis 2:31, 1998 syndrome. Crit Care Med 29(suppl):S106, 2001 10. Bernard G: The Brussels score. Sepsis 1:43, 1997 30. Bywaters EGL: Anatomical changes in the liver 51. Todd JC III, Mollitt DL: Effect of sepsis on eryth- 11. Vincent JL: The sepsis-related organ failure after trauma. Clin Sci 6:19, 1946 assessment (SOFA) score. Intens Care Med 1996 rocyte intracellular calcium homeostasis. Crit 31. Hawker F. Liver dysfunction in critical illness. Care Med 23:459, 1995 12. Le Gall JR, Klar J, Lemeshow S, et al:The logistic Anaesth Intens Care 19:165, 1991 organ dysfunction system—A new way to assess 52. Jimenez MF, Watson RWG, Parodo J, et al: organ dysfunction in the intensive care unit. 32. Schwartz DB, Bone RC, Balk RA, et al: Hepatic Dysregulated expression of neutrophil apoptosis JAMA 276:802, 1996 dysfunction in the adult respiratory distress syn- in the systemic inflammatory response syndrome drome. Chest 95:871, 1989 (SIRS). Arch Surg 132:1263, 1997 13. Burke JF, Pontoppidan H,Welch CE: High output respiratory failure: an important cause of death 33. Langenfeld JE, Machiedo GW, Lyons M, et al: 53. Hotchkiss RS, Tinsley KW, Swanson PE, et al: ascribed to peritonitis or ileus. Ann Surg 158:581, Correlation between red blood cell deformability Sepsis-induced apoptosis causes progressive pro- 1963 and changes in hemodynamic function. Surgery found depletion of B and CD4+ T lymphocytes in 116:859, 1994 humans. J Immunol 166:6952, 2001 14. Ashbaugh DG, Bigelow DB, Petty TL, et al: Acute respiratory distress in adults. Lancet 2:319, 1967 34. Parrillo JE, Parker MM, Natanson C, et al: Septic 54. Curling TB: On acute ulceration of the duodenum shock in humans. Advances in the understanding in cases of burns. Med-Chir Tr London 25:260, 15. Kollef MH, Schuster DP: Medical progress: the of pathogenesis, cardiovascular dysfunction, and 1842 acute respiratory distress syndrome. N Engl J Med therapy. Ann Intern Med 113:227, 1990 55. Cook DJ, Fuller H, Guyatt GH, et al: Risk factors 332:27, 1995 35. Vincent JL, Gris P, Coffernils M, et al: Myocardial for gastrointestinal bleeding in critically ill 16. Ranieri VM, Suter PM, Tortorella C, et al: Effect depression characterizes the fatal course of septic patients. N Engl J Med 330:377, 1994 of mechanical ventilation on inflammatory media- shock. Surgery 111:660, 1992 56. Chang RWS, Jacobs S, Lee B: Gastrointestinal tors in patients with acute respiratory distress syn- drome: a randomized controlled trial. JAMA 36. Yien HW, Hseu SS, Lee LC, et al: Spectral analy- dysfunction among intensive care unit patients. 282:54, 1999 sis of systemic arterial pressure and heart rate sig- Crit Care Med 15:909, 1987 nals as a prognostic tool for the prediction of 57. van der Spoel JI, Oudemans-van Straaten HM, 17. Gattinoni L, Pesenti A, Bombino M: Rela- patient outcome in the intensive care unit. Crit Stoutenbeek CP, et al: Neostigmine resolves critical tionships between lung computed tomographic Care Med 25:258, 1997 illness-related colonic ileus in intensive care density, gas exchange, and PEEP in acute respira- tory failure. Anesthesiology 69:824, 1988 37. Lorente JA, Landin L, Renes E, et al: Role of patients with multiple organ failure—a prospective, nitric oxide in the hemodynamic changes of sep- double-blind, placebo-controlled trial. Intensive 18. Pinhu L, Whitehead T, Evans T, et al: Ventilator- sis. Crit Care Med 21:759, 1993 Care Med 27:822, 2001 associated lung injury. Lancet 361:332, 2003 19. Murray JF, Matthay MA, Luce JM, et al: An 38. Finkel MS, Oddis CV, Jacob TD, et al: Negative 58. Bell RC, Coalson JJ, Smith JD, et al: Multiple expanded definition of the adult respiratory dis- inotropic effects of cytokines on the heart mediat- organ system failure and infection in adult respi- tress syndrome. Am Rev Respir Dis 138:720, ed by nitric oxide. Science 257:387, 1992 ratory distress syndrome. Ann Intern Med 1988 39. Bolton CF, Young GB, Zochodne DW: The neu- 99:293, 1983 20. Bernard GR, Artigas A, Brigham KL, et al: The rological complications of sepsis. Ann Neurol 59. Glenn F, Becker CG: Acute acalculous cholecysti- American-European consensus conference on 33:94, 1993 tis: an increasing entity. Ann Surg 195:131, 1982 ARDS. Definitions, mechanisms, relevant out- 40. Bolton CF: Sepsis and the systemic inflammatory 60. Van den Berghe G, Wouters P, Weekers F, et al: comes, and clinical trial coordination. Am J Respir response syndrome: Neuromuscular manifesta- Intensive insulin therapy in the surgical intensive
• 18.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 18 care unit. N Engl J Med 345:1359, 2001 effectiveness of right heart catheterization in the factor? Am J Respir Crit Care Med 157(6 pt 61. Annane D, Sebille V, Troche G, et al: A 3-level initial care of critically ill patients. JAMA 276:889, 1):1721, 1998 prognostic classification in septic shock based on 1996 104. Deneke SM, Fanburg BL: Normobaric oxygen cortisol levels and cortisol response to corti- 83. Gattinoni L, Brazzi L, Pelosi P, et al: A trial of toxicity of the lung. N Engl J Med 303:76, 1980 cotropin. JAMA 2834:1038, 2000 goal-oriented hemodynamic therapy in critically 105. Maetani S, Nishikawa T, Tobe T, et al: Role of 62. Cooper MS, Stewart PM: Corticosteroid insuffi- ill patients. N Engl J Med 333:1025, 1995 blood transfusion in organ system failure follow- ciency in acutely ill patients. N Engl J Med 84. Sandham JD, Hull RD, Brant RF, et al: A ran- ing major abdominal surgery. Ann Surg 203:275, 348:727, 2003 domized, controlled trial of the use of pulmonary- 1986 63. Vasa FR, Molitch ME: Endocrine problems in the artery catheters in high-risk surgical patients. N 106. Sauaia A, Moore FA, Moore EE, et al: Early risk chronically critically ill patient. Clin Chest Med Engl J Med 348:5, 2003 factors for postinjury multiple organ failure.World 22:193, 2003 85. Pape HC, Giannoudis P, Krettek C:The timing of J Surg 20:392, 1996 64. Christou NV, Meakins JL, Gordon J, et al: The fracture treatment in polytrauma patients: rele- 107. Hebert PC,Wells G, Blajchman MA, et al: A mul- delayed hypersensitivity response and host resis- vance of damage control orthopedic surgery. Am J ticentre randomized controlled clinical trial of tance in surgical patients—20 years later. Ann Surg 183:622, 2002 transfusion requirements in critical care. N Engl J Surg 222:534, 1995 86. Heyland DK, Novak F, Drover JW, et al: Should Med 340:409, 1999 65. Grbic JT, Mannick JA, Gough DB, et al: The role immunonutrition become routine in critically ill 108. Marik PE, Sibbald WJ: Effect of stored blood of prostaglandin E2 in immune suppression fol- patients? A systematic review of the evidence. transfusion on oxygen delivery in patients with lowing injury. Ann Surg 214:253, 1991 JAMA 286:944, 2001 sepsis. JAMA 269:3024, 1993 66. Marshall JC, Christou NV, Horn R, et al: The 87. Takala J, Ruokonen E, Webster NR, et al: 109. Grant JP, Cox CE, Kleinman LM, et al: Serum microbiology of multiple organ failure. The proxi- Increased mortality associated with growth hor- hepatic enzyme and bilirubin elevations during mal GI tract as an occult reservoir of pathogens. mone treatment in critically ill adults. N Engl J parenteral nutrition. Surg Gynecol Obstet Arch Surg 123:309, 1988 Med 341:785, 1999 145:2398, 1977 67. Eachempati SR, Hydo LJ, Barie PS: Factors influ- 88. Maki DG: Risk factors for nosocomial infection in 110. Askanazi J, Rosenbaum SH, Hyman AI, et al: encing the development of decubitus ulcers in intensive care: devices vs nature and goals for the Respiratory changes induced by the large glucose critically ill surgical patients. Crit Care Med next decade. Arch Intern Med 149:30, 1989 loads of total parenteral nutrition. JAMA 29:1678, 2001 89. Cook DJ, Guyatt GH, Marshall JC, et al: A ran- 243:1444, 1980 68. Knaus WA, Wagner DP, Lynn J: Short-term mor- domized trial of sucralfate versus ranitidine for 111. Moore FA, Feliciano DV, Andrassy RJ, et al: Early tality predictions for critically ill hospitalized stress ulcer prophylaxis in critically ill patients. N enteral feeding, compared with parenteral, adults: science and ethics. Science 254:389, 1991 Engl J Med 338:791, 1998 reduces postoperative septic complications: the 69. Lemeshow S, Le Gall JR: Modeling the severity of 90. D’Amico R, Pifferi S, Leonetti C, et al: Effec- results of a meta-analysis. Ann Surg 216:172, illness of ICU patients: a systems update. JAMA tiveness of antibiotic prophylaxis in critically ill 1992 272:1049, 1994 adult patients: systematic review of randomized 112. Masip J, Betbese AJ, Paez J, et al: Non-invasive controlled trials. Br Med J 316:1275, 1998 pressure support ventilation versus conventional 70. Vincent JL, De Mendonca A, Cantraine F, et al: Use of the SOFA score to assess the incidence of 91. Nathens AB, Marshall JC: Selective decontamina- oxygen therapy in acute cardiogenic pulmonary organ dysfunction/failure in intensive care units: tion of the digestive tract in surgical patients. Arch oedema: a randomised trial. Lancet 356:2126, results of a multicenter, prospective study. Crit Surg 134:170, 1999 2000 Care Med 26:1793, 1998 92. Van Der Waaij D:The ecology of the human intes- 113. Delclaux C, L’Her E, Alberti C, et al: Treatment 71. Goris RJA, te Boekhorst TPA, Nuytinck JKS, et tine and its consequences for overgrowth by of acute hypoxemic nonhypercapnic respiratory al: Multiple organ failure. Generalized autode- pathogens such as Clostridium difficile. Annu Rev insufficiency with continuous positive airway pres- structive inflammation? Arch Surg 120:1109, Microbiol 43:69, 1989 sure delivered by a face mask: a randomized con- 1985 trolled trial. JAMA 284:2352, 2000 93. Le Mee J, Paye F, Sauvanet A, et al: Incidence and 72. Moore FA, Moore EE, Poggetti R, et al: Gut bac- reversibility of organ failure in the course of ster- 114. Parker JC, Hernandez LA, Peevy KJ: Mechanisms terial translocation via the portal vein: a clinical ile or infected necrotizing pancreatitis. Arch Surg of ventilator induced lung injury. Crit Care Med perspective with major torso trauma. J Trauma 136:1386, 2001 21:131, 1993 31:629, 1991 94. Polk HC, Shields CL: Remote organ failure: a 115. Munoz J, Guerrero JE, Escalante JL, et al: 73. Hebert PC, Drummond AJ, Singer J, et al: A sim- valid sign of occult intraabdominal infection. Sur- Pressure-controlled ventilation versus controlled ple multiple system organ failure scoring system gery 81:310, 1977 mechanical ventilation with decelerating inspira- predicts mortality of patients who have sepsis syn- tory flow. Crit Care Med 21:1143, 1993 95. Meakins JL,Wicklund B, Forse RA, et al:The sur- drome. Chest 104:230, 1993 gical intensive care unit: current concepts in infec- 116. Hickling KG, Henderson SJ, Jackson R: Low 74. Vincent J-L, Moreno R, Takala J, et al: The SOFA tion. Surg Clin North Am 60:117, 1980 mortality associated with low volume pressure (sepsis-related organ failure assessment) score to limited ventilation with permissive hypercapnia in 96. Heyland DK, Cook DJ, Marshall JC, et al: The describe organ dysfunnction/failure. Intensive severe adult respiratory distress syndrome. clinical utility of invasive diagnostic techniques in Care Med 22:707, 1996 Intensive Care Med 16:372, 1990 the setting of ventilator-associated pneumonia. 75. Marshall JC: Charting the course of critical ill- Chest 115:1076, 1999 117. Hickling KG, Walsh J, Henderson S, et al: Low ness: prognostication and outcome description in mortality rate in adult respiratory distress syn- 97. Fagon J-Y, Chastre J, Wolff M, et al: Invasive and the intensive care unit. Crit Care Med 27:676, drome using low-volume, pressure-limited ventila- noninvasive strategies for management of suspect- 1999 tion with permissive hypercapnia: a prospective ed ventilator-associated pneumonia: a randomized study. Crit Care Med 22:1568, 1994 76. Cook RJ, Cook DJ, Tilley J, et al: Multiple organ trial. Ann Intern Med 132:621, 2000 dysfunction: baseline and serial component 118. Laffey JG, Tanaka M, Engelberts D, et al: 98. Nieto-Rodriguez JA, Kusne S, Mañez R, et al: scores. Crit Care Med 29:2046, 2001 Therapeutic hypercapnia reduces pulmonary and Factors associated with the development of can- systemic injury following in vivo lung reperfusion. 77. Ferreira FL, Bota DP, Bross A, et al: Serial evalu- didemia and candidemia-related death among Am J Respir Crit Care Med 162:2287, 2000 ation of the SOFA score to predict outcome in liver transplant recipients. Ann Surg 223:70, 1996 critically ill patients. JAMA 286:1754, 2001 119. Derdak S, Mehta S, Stewart TE, et al: High-fre- 99. Henao FJ, Daes JE, Dennis RJ: Risk factors for quency oscillatory ventilation for acute respiratory 78. Choi PT,Yip G, Quinonez LG, et al: Crystalloids multiorgan failure: a case control study. J Trauma distress syndrome in adults: a randomized, con- vs. colloids in fluid resuscitation: a systematic 31:74, 1991 trolled trial. Am J Respir Crit Care Med 166:801, review. Crit Care Med 27:200, 1999 100. Muckart DJ,Thomson SR: Undetected injuries: a 2002 79. Dellinger EP: Cardiovascular management of sep- preventable cause of increased morbidity and 120. Ferguson ND, Stewart TE: The use of high-fre- tic shock. Crit Care Med 31:946, 2003 mortality. Am J Surg 162:457, 1991 quency oscillatory ventilation in adults with acute 80. Gomersall CD, Joynt GM, Freebairn RC, et al: 101. Davis JW, Hoyt DB, McArdle MS, et al: The sig- lung injury. Respir Care Clin North Am 7:647, Resuscitation of critically ill patients based on the nificance of critical care errors in causing pre- 2001 results of gastric tonometry: a prospective, ran- ventable death in trauma patients in a trauma sys- 121. Dellinger RP, Zimmerman JL, Taylor RW, et al: domized, controlled trial. Crit Care Med 28:607, tem. J Trauma 31:813, 1991 Effects of inhaled nitric oxide in patients with 2000 102. Muscedere JG, Mullen JBM, Gan K, et al: Tidal acute respiratory distress syndrome: results of a 81. De Backer D, Creteur J, Preiser JC, et al: ventilation at low airway pressures can augment randomized phase II trial. Crit Care Med 26:15, Microvascular blood flow is altered in patients lung injury. Am J Respir Crit Care Med 149:1327, 1998 with sepsis. Am J Respir Crit Care Med 166:98, 1994 122. Lewandowski K: Extracorporeal membrane oxy- 2002 103. Slutsky AS, Tremblay LN: Multiple system organ genation for severe acute respiratory failure. Crit 82. Connors AF Jr, Speroff T, Dawson NV, et al: The failure. Is mechanical ventilation a contributing Care 4:156, 2000
• 19.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 19 123. Mols G, Loop T, Geiger K, et al: Extracorporeal 143. Hebert PC, Yetisir E, Martin C, et al: Is a low consensus conference definitions of the systemic membrane oxygenation: a ten-year experience. transfusion threshhold safe in patients with car- inflammatory response syndrome (SIRS) and Am J Surg 180:144, 2000 diovascular diseases? Crit Care Med 29:227, allied disorders in relation to critically injured 124. Morris AH, Wallace CJ, Menlovet RL, et al: 2001 patients. Crit Care Med 25:1789, 1997 Randomized clinical trial of pressure-controlled 144. Kress JP, Pohlman AS, O’Connor MF, et al: Daily 165. Bone RC: Sir Isaac Newton, sepsis, SIRS, and inverse ratio ventilation and extracorporeal CO2 interruption of sedative infusions in critically ill CARS. Crit Care Med 24:1125, 1996 removal for adult respiratory distress syndrome. patients undergoing mechanical ventilation. N 166. Marshall JC: Rethinking sepsis: from concepts to Am J Respir Crit Care Med 149:295, 1994 Engl J Med 342:1471, 2000 syndromes to diseases. Sepsis 3:5, 1999 125. Connors AF Jr, Speroff T, Dawson NV, et al: The 145. Taylor FB Jr, Stearns-Kurosawa DJ, Kurosawa S, 167. Fink MP: Cytopathic hypoxia. Crit Care 6:491, effectiveness of right heart catheterization in the et al: The endothelial cell protein C receptor aids 2002 initial care of critically ill patients. JAMA 276:889, in host defense against Escherichia coli sepsis. 1996 Blood 95:1680, 2000 168. Norton LW: Does drainage of intraabdominal pus reverse multiple organ failure? Am J Surg 126. Mehta RL, McDonald B, Gabbai FB, et al: A ran- 146. Bernard GR, Vincent J-L, Laterre PF, et al: 149:347, 1985 domized clinical trial of continuous versus inter- Efficacy and safety of recombinant human acti- mittent dialysis for acute renal failure. Kidney Int vated protein C for severe sepsis. N Engl J Med 169. Suffredini AF, Fromm RE, Parker MM, et al:The 60:1154, 2001 344:699, 2001 cardiovascular response of normal humans to the administration of endotoxin. N Engl J Med 127. Bouman CS, Oudemans-van Straaten HM, 147. Vincent J-L, Angus DC, Artigas A, et al: Effects of 321:280, 1989 Tijssen JG, et al: Effects of early high-volume con- drotrecogin alfa (activated) on organ dysfunction tinuous venovenous hemofiltration on survival in the PROWESS trial. Crit Care Med 31:834, 170. Taveira Da Silva AM, Kaulach HC, Chuidian FS, and recovery of renal function in intensive care 2003 et al: Brief report: shock and multiple organ dys- patients with acute renal failure: a prospective, function after self administration of salmonella 148. Annane D, Sebille V, Charpentier C, et al: Effect endotoxin. N Engl J Med 328:1457, 1993 randomized trial. Crit Care Med 30:2205, 2002 of treatment with low doses of hydrocortisone and 128. Tonelli M, Manns B, Feller-Kopman D: Acute fludrocortisone on mortality in patients with sep- 171. Danner RL, Elin RJ, Hosseini JM, et al: Endo- renal failure in the intensive care unit: a systemat- tic shock. JAMA 288:862, 2002 toxemia in human septic shock. Chest 99:169, ic review of the impact of dialytic modality on 1991 149. Cronin L, Cook DJ, Carlet J, et al: Corticosteroid mortality and renal recovery. Am J Kidney Dis 172. Bates DW, Parsonnet J, Ketchum PA, et al: treatment for sepsis: a critical appraisal and meta- 40:875, 2002 Limulus amebocyte lysate assay for detection of analysis of the literature. Crit Care Med 23:1430, 129. Ronco C, Bellomo R, Homel P, et al: Effects of 1995 endotoxin in patients with sepsis sydrome. Clin different doses in continuous veno-venous Infect Dis 27:582, 1998 150. Marshall JC: Such stuff as dreams are made on: haemofiltration on outcomes of acute renal fail- 173. Hoch RC, Rodriguez R, Manning T, et al: Effects mediator-targeted therapy in sepsis. Nat Rev ure: a prospective randomised trial. Lancet of accidental trauma on cytokine and endotoxin Drug Discov 2:391, 2003 356:26, 2000 production. Crit Care Med 21:839, 1993 151. Vincent JL, Bihari DJ, Suter PM, et al:The preva- 130. Schiffl H, Lang SM, Fischer R: Daily hemodialy- 174. Winchurch RA, Thupari JN, Munster AM: lence of nosocomial infection in intensive care sis and the outcome of acute renal failure. N Engl Endotoxemia in burn patients: Levels of circulat- units in Europe: results of the European J Med 346:362, 2002 ing endotoxins are related to burn size. Surgery Prevalence of Infection in Intensive Care (EPIC) 131. Kellum JA, Angus DC, Johnson JP, et al: Con- Study. JAMA 274:639, 1995 102:808, 1987 tinuous versus intermittent renal replacement ther- 175. Roumen RMH, Frieling JTM, van Tits HWHJ, et 152. Marshall JC, Christou NV, Meakins JL: The gas- apy: a meta-analysis. Intensive Care Med 28:29, al: Endotoxemia after major vascular operations. J trointestinal tract: the “undrained abscess” of 2002 Vasc Surg 18:853, 1993 multiple organ failure. Ann Surg 218:111, 1993 132. Cole L, Bellomo R, Hart G, et al: A phase II ran- 176. Poltorak A, He X, Smirnova I, et al: Defective 153. Nathens AB, Marshall JC: Selective decontamina- domized, controlled trial of continuous hemofil- LPS signaling in C3H/HeJ and C57BL/10ScCr tion of the digestive tract (SDD) in surgical pa- tration in sepsis. Crit Care Med 30:100, 2002 mice: mutations in the Tlr4 gene. Science 282: tients. Arch Surg 134:170, 1999 133. Kudsk KA, Croce MA, Fabian TC, et al: Enteral 2085, 1998 154. Singh N, Rogers P, Atwood CW, et al: Short- versus parenteral feeding: effects on septic mor- 177. Michalek SM, Moore RN, McGhee JR, et al: The course empiric antibiotic therapy for patients with bidity after blunt and penetrating abdominal trau- primary role of lymphoreticular cells in the medi- pulmonary infiltrates in the intensive care unit: a ma. Ann Surg 215:503, 1992 ation of host responses to bacterial endotoxin. J proposed solution for indiscriminate antibiotic 134. McGregor CS, Marshall JC: Enteral feeding in prescription. Am J Respir Crit Care Med 162(2 pt Infect Dis 141:55, 1980 acute pancreatitis: just do it. Curr Opin Crit Care 1):505, 2000 178. Zhao B, Bowden RAS, Stavchansky SA, et al: 7:89, 2001 Human endothelial cell response to gram-nega- 155. Ferraris VA: Exploratory laparotomy for potential 135. Marik PE, Zaloga GP: Early enteral nutrition in abdominal sepsis in patients with multiple-organ tive lipopolysaccharide assessed with cDNA acutely ill patients: a systematic review. Crit Care failure. Arch Surg 118:1130, 1983 microarrays. Am J Physiol Cell Physiol 281: Med 29:2264, 2003 C1587, 2001 156. Hinsdale JG, Jaffe BM: Re-operation for intra- 136. Beale RJ, Bryg DJ, Bihari DJ: Immunonutrition in abdominal sepsis. Indications and results in mod- 179. Fessler MB, Malcolm KC, Duncan MW, et al: A the critically ill: a systematic review of clinical out- ern critical care setting. Ann Surg 199:31, 1984 genomic and proteomic analysis of activation of come. Crit Care Med 27:2799, 1999 the human neutrophil by lipopolysaccharide and 157. Bunt TJ: Non-directed relaparotomy for intraab- its mediation by p38 mitogen-activated protein 137. Griffith DP, McNally AT, Battey CH, et al: dominal sepsis: a futile procedure. Am Surg Intravenous erythromycin facilitates bedside kinase. J Biol Chem 277:31291, 2002 52:294, 1986 placement of postpyloric feeding tubes in critical- 180. Aderem A, Ulevitch RJ: Toll-like receptors in the ly ill adults: a double-blind, randomized, placebo- 158. Knaus WA, Draper EA, Wagner DP, et al: induction of the innate immune response. Nature controlled study. Crit Care Med 31:39, 2003 Prognosis in acute organ system failure. Ann Surg 406:782, 2000 202:685, 1985 138. Berne JD, Norwood SH, McAuley CE, et al: 181. Faist E, Baue AE, Dittmer H, et al: Multiple Erythromycin reduces delayed gastric emptying in 159. Ranson JHC, Rifkind KM,Turner JW: Prognostic organ failure in polytrauma patients. J Trauma critically ill trauma patients: a randomized, con- signs and nonoperative peritoneal lavage in acute 23:775, 1983 trolled trial. J Trauma 53:422, 2002 pancreatitis. Surg Gynecol Obstet 143:209, 1976 182. Sorenson TI, Nielsen GG, Andersen PK, et al: 139. Rozga J, Podesta L, LePage EA, et al: A bioartifi- 160. Prendergast TJ, Claessens MT, Luce JM: A Genetic and environmental influences on prema- cial liver to treat severe acute liver failure. Ann national survey of end-of-life care for critically ill ture death in adult adoptees. N Engl J Med Surg 219:538, 1994 patients. Am J Respir Crit Care Med 158:1163, 318:727, 1988 1998 140. Mitzner SR, Stange J, Klammt S, et al: 183. Lazarus R, Vercelli D, Palmer LJ, et al: Single Extracorporeal detoxification using the molecular 161. Cook DJ, Guyatt GH, Jaeschke R, et al: Deter- nucleotide polymorphisms in innate immunity adsorbent recirculating system for critically ill minants in Canadian health care workers of the genes: abundant variation and potential role in patients with liver failure. J Am Soc Nephrol decision to withdraw life support from the criti- complex human disease. Immunol Rev 190:9, 12(17 suppl):S75, 2001 cally ill. JAMA 273:703, 1995 2002 141. Alderson P, Bunn F, Lefebvre C, et al: Human 162. Safar P, DeKornfeld T, Pearson J, et al: Intensive 184. Agnese DM, Calvano JE, Hahm SJ, et al: Human albumin solution for resuscitation and volume care unit. Anesthesia 16:275, 1961 toll-like receptor 4 mutations but not CD14 poly- expansion in critically ill patients. Cochrane 163. Tilney NL, Bailey GL, Morgan AP: Sequential morphisms are associated with an increased risk of Database Syst Rev (1):CD001208, 2002 system failure after rupture of abdominal aortic gram-negative infections. J Infect Dis 186:1522, 142. Vincent JL, Baron JF, Reinhart K, et al: Anemia aneurysms: an unsolved problem in postoperative 2002 and blood transfusion in critically ill patients. care. Ann Surg 178:117, 1973 185. Lorenz E, Mira J-P, Frees KL, et al: Relevance of JAMA 288:1499, 2002 164. Muckart DJJ, Bhagwanjee S: The ACCP/SCCM mutations in the TLR4 receptor in patients with
• 20.
  © 2003 WebMDInc. All rights reserved. ACS Surgery: Principles and Practice 8 CRITICAL CARE 13 MULTIPLE ORGAN DYSFUNCTION SYNDROME — 20 gram-negative septic shock. Arch Intern Med vation. J Immunol 166:2018, 2001 testinal tract. Annu Rev Med 31:107, 1977 162:1028, 2002 205. Ishiguro K, Kadomatsu K, Kojima TY, et al: 226. Wells CL, Maddaus MA, Simmons RL: Proposed 186. Gibot S, Cariou A, Drouet L, et al: Association Syndecan-4 deficiency leads to high mortality of mechanisms for the translocation of intestinal bac- between a genomic polymorphism within the lipopolysaccharide-injected mice. J Biol Chem teria. Rev Infect Dis 10:958, 1988 CD14 locus and septic shock susceptibility and 276:47483, 2001 227. O’Boyle CJ, MacFie J, Mitchell CJ, et al: mortality rate. Crit Care Med 30:969, 2002 206. Muller AM, Cronen C, Muller KM, et al: Microbiology of bacterial translocation in hu- 187. Mira J-P, Cariou A, Grall F, et al: Association of Heterogeneous expression of cell adhesion mole- mans. Gut 42:29, 1998 TNF2, a TNF-a promoter polymorphism, with cules by endothelial cells in ARDS. J Pathol 198: 228. MacFie J, O’Boyle C, Mitchell CJ, et al: Gut ori- septic shock susceptibility and mortality. JAMA 170, 2002 gin of sepsis: a prospective study investigating 282:561, 1999 207. Ueno H, Hirasawa H, Oda S, et al: Coag- associations between bacterial translocation, gas- 188. Appoloni O, Dupont E, Vandercruys M, et al: ulation/fibrinolysis abnormality and vascular endo- tric microflora, and septic morbidity. Gut 45:223, Association of tumor necrosis factor-2 allele with thelial damage in the pathogenesis of thrombocy- 1999 plasma tumor necrosis factor-α levels and mortal- topenic multiple organ failure. Crit Care Med 30:2242, 2002 229. Deitch EA, Xu D, Franko L, et al: Evidence favor- ity from septic shock. Am J Med 110:486, 2001 ing the role of the gut as a cytokine generating 189. Schroder O, Schulte KM, Ostermann P, et al: 208. Mutunga M, Fulton B, Bullock R, et al: Cir- organ in rats subjected to hemorrhagic shock. Heat shock protein 70 genotypes HSPA1B and culating endothelial cells in patients with septic Shock 1:141, 1994 HSPA1L influence cytokine concentrations and shock. Am J Respir Crit Care Med 163:195, 2001 interfere with outcome after major injury. Crit 230. Arita H, Ogle CK, Alexander JW, et al: Induction 209. Harlan JM, Winn RK: Leukocyte-endothelial of hypermetabolism in guinea pigs by endotoxin Care Med 31:73, 2003, interactions: clinical trials of anti-adhesion thera- infused through the portal vein. Arch Surg 190. Reid CL, Perrey C, Pravica V, et al: Genetic varia- py. Crit Care Med 30(suppl):S214, 2002 123:1420, 1988 tion in proinflammatory and anti-inflammatory 210. Esmon C: The protein C pathway. Crit Care Med cytokine production in multiple organ dysfunction 231. Marshall JC, Ribeiro MB, Chu PTY, et al: Portal 28(suppl):S44, 2000 syndrome. Crit Care Med 30:2216, 2003 endotoxemia stimulates the release of an immuno- 211. Warren BL, Eid A, Singer P, et al: High-dose suppressive factor from alveolar and splenic 191. Ma P, Chen D, Pan J, et al: Genomic polymor- antithrombin III in severe sepsis: a randomized, macrophages. J Surg Res 55:14, 1993 phism within interleukin-1 family cytokines influ- controlled trial. JAMA 286:1869, 2001 ences the outcome of septic patients. Crit Care 232. Sullivan BJ, Swallow CJ, Girotti MJ, et al: 212. Abraham E, Reinhart K, Opal S, et al: Efficacy Bacterial translocation induces procoagulant Med 30:1046, 2002 and safety of tifacogin (recombinant tissue factor activity in tissue macrophages: a potential mecha- 192. Ginzberg HH, Cherapanov V, Dong Q, et al: pathway inhibitor) in severe sepsis. JAMA nism for end-organ dysfunction. Arch Surg Neutrophil-mediated epithelial injury during 290:283, 2003 126:586, 1991 transmigration: role of elastase. Am J Physiol 213. Cheng T, Liu D, Griffin JH, et al: Activated pro- 233. Gonzalez RJ, Moore EE, Ciesla DJ, et al: Post- Gastrointest Liver Physiol 281:G705, 2001 tein C blocks p53-mediated apoptosis in ischemic hemorrhagic shock mesenteric lymph activates 193. Yum HK, Arcaroli J, Kupfner J, et al: Involvement human brain endothelium and is neuroprotective. human pulmonary microvascular endothelium for of phosphoinositide 3-kinases in neutrophil acti- Nature Med 9:338, 2003 in vitro neutrophil-mediated injury: the role of vation and the development of acute lung injury. J 214. Thompson CB: Apoptosis in the pathogenesis and intercellular adhesion molecule-1. J Trauma Immunol 167:6601, 2001 treatment of disease. Science 267:1456, 1995 54:219, 2003 194. Lee WL, Downey GP: Leukocyte elastase. Physio- 215. Fadok VA, Bratton DL, Konowal A, et al: 234. Moore FA, Moore EE: Evolving concepts in the logical functions and role in acute lung injury. Am Macrophages that have ingested apoptotic cells in pathogenesis of postinjury multiple organ failure. J Respir Crit Care Med 164:896, 2001 vitro inhibit proinflammatory cytokine production Surg Clin North Am 75:257, 1995 195. Jaeschke H, Smith CW. Mechanisms of neutro- through autocrine/paracrine mechanisms involving 235. Fan J, Marshall JC, Jimenez M, et al: Hemor- phil-induced parenchymal cell injury. J Leukoc Biol TGFb, PGE2, and PAF. J Clin Invest 101:890, rhagic shock primes for increased expression of 61:647, 1997 1998 cytokine-induced neutrophil chemoattractant in 196. Kubes P, Hunter J, Granger DN: Ischemia/reper- 216. Byrne A, Reen DJ: Lipopolysaccharide induces the lung: role in pulmonary inflammation follow- fusion induced feline intestinal dysfunction: rapid production of IL-10 by monocytes in the ing lipopolysaccharide. J Immunol 161:440, 1998 importance of granulocyte recruitment. Gastro- presence of apoptotic neutrophils. J Immunol enterology 103:807, 1992 168:1968, 2002 236. Godin PJ, Buchman TG: Uncoupling of biological oscillators: a complementary hypothesis concern- 197. Nieuwenhuijzen GAP, Haskel Y, Lu Q, et al: 217. Hotchkiss RS, Schmieg RE Jr, Swanson PE, et al: ing the pathogenesis of multiple organ dysfunc- Macrophage elimination increases bacterial Rapid onset of intestinal epithelial and lympho- tion syndrome. Crit Care Med 24:1107, 1996 translocation and gut origin septicemia but atten- cyte apoptotic cell death in patients with trauma uates symptoms and mortality rate in a model of and shock. Crit Care Med 28:3207, 2000 237. Seely AJE, Christou NV: Multiple organ dysfunc- systemic inflammation. Ann Surg 218:791, 1993 tion syndrome: exploring the paradigm of complex 218. Hotchkiss RS, Tinsley KW, Swanson PE, et al: non-linear systems. Crit Care Med 28:2193, 2000 198. Carson WE, Yu H, Dierksheide J, et al: A fatal Prevention of lymphocyte cell death in sepsis cytokine-induced systemic inflammatory response improves survival in mice. Proc Natl Acad Sci 238. Ivanov PC, Nunes Amaral LA, Goldberger AL, et reveals a critical role for NK cells. J Immunol 162: USA 96:14541, 1999 al: Multifractality in human heartbeat dynamics. 4943, 1999 Nature 399:461, 1999 219. Sookhai S, Wang JH, McCourt M, et al: A novel 199. Badgwell B, Parihar R, Magro C, et al: Natural mechanism for attenuating neutrophil-mediated 239. Korach M, Sharshar T, Jarrin I, et al: Cardiac vari- killer cells contribute to the lethality of a murine lung injury in vivo. Surg Forum 50:205, 1999 ability in critically ill adults: influence of sepsis. model of Escherichia coli infection. Surgery 132: Crit Care Med 29:1380, 2001 220. Imai Y, Parodo J, Kajikawa O, et al: Injurious 205, 2002 mechanical ventilation and end-organ epithelial 240. Annane D, Trabold F, Sharshar T, et al: Inap- 200. Sherwood ER, Lin CY, Tao W, et al: b2 Micro- cell apoptosis and organ dysfunction in an experi- propriate sympathetic activation at onset of septic globulin knockout mice are resistant to lethal mental model of acute respiratory distress syn- shock: a spectral analysis approach. Am J Respir intra-abdominal sepsis. Am J Respir Crit Care drome. JAMA 289:2104, 2003 Crit Care Med 160:458, 1999 Med (in press) 221. Carrico CJ, Meakins JL, Marshall JC, et al: 241. Marshall JC: Complexity, chaos, and incompre- 201. Sempowski GD, Lee DM, Scearce RM, et al: Multiple organ failure syndrome. The gastroin- hensibility: parsing the biology of critical illness. Resistance of CD7-deficient mice to lipopolysac- testinal tract: the ‘motor’ of MOF. Arch Surg Crit Care Med 28:2646, 2000 charide-induced shock syndromes. J Exp Med 121:196, 1986 242. Hayes MA, Timmins AC, Yau EHS, et al: Ele- 189:1011, 1999 222. Marshall JC, Nathens AB: The gut in critical ill- vation of systemic oxygen delivery in the treatment 202. Henneke P, Golenbock DT: Innate immune ness: evidence from human studies. Shock 6:S10, of critically ill patients. N Engl J Med 330:1717, recognition of lipopolysaccharide by endothelial 1996 1994 cells. Crit Care Med 30(5 suppl):S207, 2002 223. Deitch EA. The role of intestinal barrier failure 243. Rivers E, Nguyen B, Havstad S, et al: Early goal- 203. Aird WC: The role of the endothelium in severe and bacterial translocation in the development of directed therapy in the treatment of severe sepsis sepsis and multiple organ dysfunction syndrome. systemic infection and multiple organ failure. and septic shock. N Engl J Med 345:1368, 2001 Blood 101:3765, 2003 Arch Surg 125:403, 1990 244. Panacek EA, Marshall JC, Fischkoff S, et al: 204. Faure E, Thomas L, Xu H, et al: Bacterial lipo- 224. Fink MP: Gastrointestinal mucosal injury in Neutralization of TNF by a monoclonal antibody polysaccharide and IFN-gamma induce Toll-like experimental models of shock, trauma, and sepsis. improves survival and reduces organ dysfunction receptor 2 and Toll-like receptor 4 expression in Crit Care Med 19:627, 1991 in human sepsis: results of the MONARCS trial. human endothelial cells: role of NF-kappa B acti- 225. Savage DC: Microbial ecology of the gastroin- Chest 118:88S, 2000