Physiological changes, clinical features, and general management of infected patients
The host response to an infectious stimulus involves an intricate link between the inflammatory and coagulation systems, also mechanisms designed to limit damage to normal tissues. Key elements are: (1) the inflammatory cascade—antigens from infectious agents stimulate macrophages and monocytes (and other cells) via Toll-like receptors to release tumour necrosis factor α (TNFα), resulting in a cascade of pro-inflammatory cytokine release which is a vital component of the host’s attempt to control and eradicate infection, but unfortunately can also result in damage to both infected and uninfected host tissues; inflammatory mediators with prolonged actions or appearing later in the course of sepsis are likely to play an important role in determining prognosis; (2) the anti-inflammatory cascade—a compensatory response involving anti-inflammatory cytokines, soluble receptors, and receptor antagonists directed against pro-inflammatory cytokines that is intended to localize and control the systemic proinflammatory response to the infection; (3) the coagulation cascade—activated in an attempt to contain infection locally and prevent spread to other parts of the body; platelets are activated, procoagulant pathways are initiated, and anticoagulant mediators are down-regulated; (4) the anticoagulation cascade—the coagulation response to sepsis is regulated via antithrombin, tissue factor pathway inhibitor (TFPI), activated protein C (APC), and fibrinolysis.
Definitions—(1) Systemic inflammatory response syndrome (SIRS)—which can occur as a result of an infectious or noninfectious insult—requires the presence of at least two of the following: (a) hyper- or hypothermia, (b) tachycardia, (c) tachypnoea or hyperventilation, (d) leucocytosis, leucopenia or left shift. (2) Sepsis—a suspected or confirmed infection plus criteria for SIRS. (3) Severe sepsis—sepsis resulting in the acute dysfunction of at least one organ system. (4) Septic shock—infection resulting in hypotension despite adequate fluid resuscitation.
Management—key elements are (1) antibiotics—often initiated empirically before culture results are available; (2) control of the source of infection—searching for the site of infection so that it can be eradicated should begin as soon as haemodynamic and respiratory status are stabilized; antibiotics without source control often fail; (3) early goal-directed resuscitation—requiring (a) crystalloid infusions to maintain central venous pressure, (b) vasopressors if arterial pressure remains low, and (c) transfusion of packed red blood cells and/or infusion of dobutamine if central venous oxygen saturation remains low; (4) consideration of other treatments—many specialists advocate recombinant APC for patients with severe sepsis who have a low risk of bleeding and a high risk of death.
The term sepsis describes the physiological consequences of the activation of the systemic inflammatory cascade that occurs in infected patients. The cascade of events in response to infectious stimuli has been well characterized using both animal and human models. This response is the main focus of this chapter. Other aspects of sepsis, including epidemiology, clinical features, treatment, prognosis, and the controversial role of corticosteroids and tight glycaemic control will also be addressed.
Pathophysiology of infection
Initial investigations suggested that inflammatory cytokines mediated the physiological responses seen in patients with sepsis. However, information from studies of the coagulation system in sepsis and the subsequent examination of autopsy specimens demonstrated microthrombi in the arterioles and venules of various organs, suggesting that the coagulation system played at least some role in the pathophysiology. It is now clear that the inflammatory and coagulation systems are intricately linked and homeostasis of both is altered in infected patients.
Sepsis syndrome describes the physiological effects of the systemic inflammatory cascade produced by the human body in response to any of a variety of infectious stimuli. Antigens from infectious agents stimulate macrophages and monocytes to release tumour necrosis factor-α (TNFα) resulting in a cascade of cytokine release. Numerous cell wall antigens are able to stimulate this response, including lipopolysaccharide (LPS) or endotoxin from Gram-negative bacteria, lipoteichoic acid from Gram-positive bacteria, peptidoglycan and flagellin from both Gram-negative and Gram-positive bacteria, and mannan from fungi. These antigens initiate an inflammatory response via type I transmembrane receptors called Toll-like receptors found on the surface of a variety of cell types including macrophages, neutrophils, fibroblasts, and some epithelial and endothelial cells. Numerous Toll-like receptors have been identified, but subtypes 2 (TLR2) and 4 (TLR4) appear to play a major role in mediating the inflammatory response to infectious stimuli. TLR2 binds both peptidoglycan and lipoteichoic acid from Gram-positive bacteria. On the other hand, TLR4 serves as the signal transduction component for LPS from Gram-negative bacteria. In macrophages and neutrophils, binding of LPS to TLR4 results in the release of TNFα via NF-κB, a eukaryotic transcription factor. Circulating TNFα stimulates the release of other proinflammatory cytokines from macrophages and neutrophils. These proinflammatory cytokines, especially interleukins 1 and 6 (IL-1, IL-6), trigger numerous additional proinflammatory events within endothelial cells and leucocytes (Fig. 18.104.22.168). TNFα acts in conjunction with IL-1 to produce the fever, tachycardia, and tachypnoea seen with systemic inflammation. In addition, their synergistic effects are probably responsible for the hypotension and resultant organ dysfunction seen early in the course of severe sepsis. The purpose of this proinflammatory response, which represents a vital component of the host defence, is to control and eradicate the infection. Unfortunately, the response is often so exuberant and poorly controlled that it results in damage to both infected and uninfected host tissues.
The symptoms of sepsis, specifically tachypnoea, tachycardia, fever, or hypotension, are what prompt most patients to seek medical attention. Unfortunately, these result from the early inflammatory cascade which is already well into its course by the time patients present for medical care. Administration of endotoxin to humans demonstrates that TNF is the primary protagonist of the inflammatory response as serum levels rise almost immediately. However, the presence of TNF in the serum is short-lived and the majority of patients with sepsis have undetectable levels at the time of presentation, even the most critically ill with organ failure or shock.
Many sepsis deaths occur later in the course, at least 48 to 72 h after the onset of symptoms. This has prompted many to speculate that inflammatory mediators with more prolonged actions or appearing later in the course of sepsis are likely to play an important role in determining prognosis. High-mobility group box protein 1 (HMGB1) is one such late mediator; it is a 30-kDa protein, named for its rapid migration on electrophoretic gels, which was purified along with histones from nuclei in the 1970s and is now classified as a nonhistone chromatin-associated protein. The critical role played by HMGB1 in gene transcription and DNA repair and replication has been known since shortly after its discovery. However, recent data suggest that HMGB1 also possesses inflammatory properties. Using nuclear pores, HMGB1 is able to move from the nucleus to the cytosol, making it available for extracellular secretion from macrophages when stimulated with LPS or TNF. Cell necrosis, but not apoptosis, results in the passive release of HMGB1 from all nucleated cells. Once outside the cell, HMGB1 functions as an important mediator of both inflammation and coagulation, stimulating the release of proinflammatory cytokines, including TNFα, IL-1, and IL-6 from endothelial cells and monocytes. HMGB1 also helps regulate coagulation by inducing the expression of adhesion molecules on endothelial cells, resulting in the secretion of plasminogen activator inhibitor type 1 (PAI-1) and tissue plasminogen activator.
Animal models of sepsis demonstrate that serum levels of HMGB1 increase after LPS administration. However, unlike TNF and IL-1 which peak early in the course of sepsis and become undetectable within a few hours, serum levels of HMGB1 remain undetectable until 8 h after LPS administration and continue to increase until they plateau 24 to 32 h later. Similar elevations in levels of HMGB1 can be detected in the serum of humans with severe sepsis or septic shock and in plasma and bronchoalveolar lavage fluid of patients with sepsis and acute lung injury. Serum levels of HMGB1 may also have prognostic significance as higher levels are associated with mortality.
Mice treated with sublethal doses of HMGB1 develop signs of endotoxaemia within 2 h and higher doses result in death within 18 to 36 h, even in mice resistant to the effects of LPS. Numerous animal models have also demonstrated that antagonizing the effects of HMGB1 improves survival in sepsis, even when the antagonism is considerably delayed. Anti-HMGB1 antibodies, ethyl pyruvate (a nontoxic food derivative that inhibits the release of HMGB1 from LPS-stimulated and TNF-stimulated macrophages), and competitive inhibition of the HMGB1 binding site on macrophages all result in improved survival in animal models in which sepsis was induced with endotoxin and caecal ligation and puncture. However, treatment with these antagonists did not just extend the time to death, but allowed many of the animals to survive until necropsy and ‘rescued’ animals that already exhibited signs of severe sepsis.
Although the initial inflammatory response in sepsis was originally believed to be largely uncontrolled, subsequent investigations have demonstrated that the body employs a compensatory anti-inflammatory response (CARS) in an attempt to maintain homeostasis. In addition to stimulating the release of other proinflammatory mediators, TNFα and IL-1 also stimulate leucocytes to release anti-inflammatory mediators, including IL-10, IL-13, and transforming growth factor β (TGFβ). These cytokines exert a direct anti-inflammatory effect on macrophages and endothelial cells and inhibit the synthesis of proinflammatory mediators. IL-10 and IL-13 also inhibit the ability of monocytes effectively to present antigens to other immune cells. Two new cytokines that might be therapeutic targets in the treatment of severe sepsis are IL-17 which promotes inflammation by triggering production of IL-1b, IL-6, and TNFα and macrophage migration inhibitory factor (MIF) which modulates TLR4. The inflammatory response is further controlled by the release of soluble receptors and receptor antagonists directed against proinflammatory cytokines. This compensatory anti-inflammatory response is intended to localize and control the systemic proinflammatory response to the infection. Unfortunately, the anti-inflammatory response often exceeds the proinflammatory response in the later phases of sepsis, resulting in a hyporesponsiveness of immune cells and an inability to mount an effective immune response to additional infectious insults. In a recent study, anti-CD3/anti-CD28- and lipopolysaccharide-stimulated splenocytes from patients dying of sepsis showed grossly impaired ability to secrete tumour necrosis factor, interferon, IL-6 and IL-10. Their spleens and lungs showed increased expression of inhibitory receptors and ligands and expansion of suppressor cell populations. Immunohistochemical staining showed extensive depletion of splenic CD4, CD8, and HLA-DR cells and expression of ligands for inhibitory receptors on lung epithelial cells. This lymphopenia and immunoparalysis late in the course of sepsis leads to delayed hypersensitivity, inability to clear infections, and an increased susceptibility to nosocomial infections, all of which contribute to the late morbidity and mortality in these patients.
Earlier investigations implicated the inflammatory cascade in sepsis, but more recent work has demonstrated the involvement of the clotting and fibrinolytic systems which are intricately related to the inflammatory response (Fig. 22.214.171.124). As well as disrupting proinflammatory and anti-inflammatory homeostasis, sepsis also disturbs coagulation homeostasis. In response to infectious stimuli, the body rouses the coagulation cascade in an attempt to contain infection locally and prevent spread to other parts of the body. To accomplish this, platelets are activated, procoagulant pathways are initiated, and anticoagulant mediators are down-regulated. Unfortunately, this often results in a ‘sepsis-associated coagulopathy’, which may range from mild thrombocytopenia or increase in prothrombin time to overt disseminated intravascular coagulopathy. If the procoagulant response becomes too exuberant, microvascular thromboses can form resulting in local tissue hypoxia and subsequent organ dysfunction.
The activation of the coagulation response is intricately linked to the inflammatory response. TNFα and IL-1 both stimulate the release of tissue factor from monocytes and neutrophils and cause tissue factor normally present on endothelial cells, but not exposed to the circulation, to be ‘unveiled’ (Fig. 126.96.36.199). Tissue factor acts as the bridge between the inflammatory and coagulation pathways by activating the extrinsic clotting system and stimulating the formation of thrombin and fibrin clots. Specifically, the highly thrombogenic tissue factor (TF) combines with circulating factor VII to form a TF:VIIa complex. This complex activates factor X, which subsequently produces thrombin from prothrombin. Thrombin stimulates the formation of fibrin clots in the microcirculation with the aim of confining the infection to the local site. The disadvantage created by diffuse intravascular microthrombi is the creation of areas of regional hypoperfusion, resulting in tissue ischaemia, coagulation necrosis, and organ dysfunction.
Thrombin, TF:VIIa complex, and activated factor X also provide positive feedback to the inflammatory cascade. All three function as potent inflammatory mediators, stimulating neutrophil migration and release of additional proinflammatory cytokines. In turn, this positive feedback loop eventually fuels additional tissue factor release and more thrombin and fibrin clot formation.
In similar fashion to its compensatory anti-inflammatory response, the body also regulates the coagulation response to sepsis. The anticoagulant response is mediated via four mechanisms: (1) antithrombin, (2) tissue factor pathway inhibitor (TFPI), (3) activated protein C (APC), and (4) fibrinolysis. Unfortunately, inflammation in sepsis suppresses many of these counter-regulatory measures, promoting a procoagulant environment. In sepsis, infection-induced inflammatory responses are procoagulant through tissue factor activation of factor VII and thence factor X and thrombin. This is enhanced by complement activation. Anticoagulant mechanisms such as protein C, antithrombin, and tissue factor pathway inhibitor are simultaneously impaired. Tissue factor, a transmembrane glycoprotein expressed by adventitial fibroblasts and vascular smooth muscle cells, comes into contact with blood through vascular damage, stimulation of mononuclear and endothelial cells by bacterial endotoxin or proinflammatory cytokines, or via microparticles shed from leucocytes, endothelial cells, vascular smooth muscle cells, and platelets. Neutrophils localize fibrin deposition to small blood vessels so limiting the spread of pathogens. Protease-activated receptors (PARs) 1–4 are activated by proteases such as thrombin, activated protein C, plasmin, trypsin, cathepsin-G, mast cell tryptase, leucocyte proteinase-3, and bacteria-derived enzymes. PARs increase vascular permeability via sphingosine 1 phosphate (S1P) receptors.
Antithrombin is synthesized in the liver and secreted into the circulation. It occurs free in the plasma and attached to platelets and endothelial cells, where it functions as an inhibitor of the coagulation system. Antithrombin directly suppresses thrombin-induced fibrin formation, inhibits factors IXa, XIa, and XIIa of the intrinsic coagulation pathway, and decreases the activation of factor Xa which is common to both the intrinsic and extrinsic coagulation pathways. The serum half-life of antithrombin, normally 36 to 48 h, is reduced to 8 h or less in sepsis-associated coagulopathy, resulting in a rapid depletion of antithrombin and a shift towards a procoagulant microvascular environment. Despite data showing that low or falling antithrombin levels were associated with a worse outcome in severe sepsis, administration of exogenous antithrombin failed to improve the prognosis in these patients.
Endothelial cells synthesize and secrete TFPI, which inhibits the TF:VIIa complex and activated factor X. This results in suppression of the extrinsic pathway of coagulation and decreases the formation of thrombin and fibrin clots. Sepsis results in a truncated form of TFPI, with reduced anticoagulant properties. Although overall TFPI levels are elevated in septic patients, their reduced anticoagulant effect still favours thrombin formation and fibrin deposition at the endothelium. Unfortunately, administration of a recombinant TFPI also failed to produce clinical benefit in a large trial of patients with severe sepsis.
Protein C is synthesized in the liver through a vitamin K-dependent pathway and is secreted into the blood as an inactive zymogen. In the presence of endothelial protein C receptor (EPCR) and a properly functioning endothelium, protein C is converted to its active form by complexing with thrombin and endothelial cell thrombomodulin. APC inhibits PAI-1 and inactivates clotting factors Va and VIIIa (Fig. 188.8.131.52). These profibrinolytic and antithrombotic properties restore and maintain, respectively, microcirculatory blood flow which help re-establish coagulation homeostasis and preserve the microcirculation. APC also has anti-inflammatory properties, such as limiting the inflammatory response induced by thrombin and inhibiting TNF production (Fig. 184.108.40.206). Unfortunately, severe sepsis down-regulates EPCR causing it to be sloughed from the cell surface, resulting in endothelial dysfunction. Sepsis also decreases plasma concentrations of thrombomodulin. These changes impair the conversion of protein C to its active form, resulting in a deficiency of APC. Most patients with severe sepsis have low levels of protein C and lower levels are associated with poorer outcomes.
Under normal homeostatic circumstances, dissolution of fibrin clots (fibrinolysis) serves as an additional counter-regulatory measure to the coagulation cascade. Plasminogen activators convert plasminogen to active plasmin which then mediates clot lysis by degrading the cross-linked fibrin present in the clot. This fibrinolytic response is controlled primarily by plasminogen activator inhibitors, predominantly PAI-1. The production of PAI-1 is stimulated by proinflammatory cytokines, resulting in elevated levels which decrease plasminogen and plasmin levels over time (Fig. 220.127.116.11). Some plasmin-induced fibrinolysis still occurs in this environment, but is often insufficient to maintain coagulation homeostasis.
The incidence of sepsis is about 240 cases per 100 000 people per year. It is the second most common cause of death in critically ill patients, after cardiovascular events, resulting in the deaths of 225 000 out of almost 750 000 patients with sepsis each year in the United States of America alone where the overall cost of caring for these patients exceeds $17 billion annually. In recent years, the incidence of sepsis has risen by almost 9% each year and will continue to increase with the ageing population, growing antibiotic resistance, increasing immunocompromised state of patients, and the expanding use of invasive procedures.
Although many patients afflicted with severe sepsis have obvious risk factors, the syndrome is not limited to the old, debilitated, or immunocompromised. Sepsis can affect anyone, including healthy young people, and often has devastating consequences. Men are more commonly affected than women and nonwhite people are affected almost twice as often as white people. Respiratory infections are the most common cause, but genitourinary sources predominate in women. Gram-positive bacteria account for over one-half of all infections, Gram-negative bacteria for about one-third, and fungi for 5%.
Clinical features and criteria for diagnosis
Definitions of systemic inflammatory response syndrome (SIRS), sepsis, and severe sepsis are frequently confused. SIRS, which can occur as a result of an infectious or noninfectious insult, is defined by demonstrating at least two of the following four criteria: (1) hyperthermia or hypothermia (temperature ≥38°C or ≤36°C), (2) tachycardia (heart rate >90 beats/min), (3) tachypnoea or hyperventilation (> 30 respirations/min or PaCO2 <32 mmHg), and (4) leucocytosis, leucopenia, or left shift (≥12 or ≤4 × 109 white blood cells/litre or >10% immature neutrophils). Sepsis is defined as a suspected or confirmed infection plus at least two of the above SIRS criteria. Severe sepsis is used to describe sepsis resulting in the acute dysfunction of at least one organ system. When the infection results in hypotension despite adequate fluid resuscitation, the patient has septic shock.
Tachypnoea is usually the first detectable clinical sign and is so common in severe sepsis that its absence should make the diagnosis suspect. The cause of the rapid breathing is often multifactorial. The lung is the most frequent site of infection and acute lung injury resulting from nonpulmonary sources of infection and respiratory compensation for metabolic acidosis also contribute. Tachycardia is virtually universal unless prevented by a cardiac conduction defect or pharmacotherapy (e.g. β-blockers). Increasing heart rate is an important compensatory mechanism to respond to the hypermetabolic state of sepsis as well as to maintain perfusion in response to intravascular volume deficits, reduced cardiac contractility, and vasodilation.
Differential diagnosis and clinical investigation
The diagnosis of sepsis may be obvious in some patients, but making the diagnosis in most requires a high level of suspicion and a fair amount of investigation. Noninfectious conditions such as pancreatitis, trauma, severe haemorrhage, myocardial infarction, drug overdose, and even heat stroke can mimic sepsis and often need to be ruled out before the diagnosis of sepsis syndrome can be established.
Laboratory abnormalities are often present, but are not specific for infection. Although suspicious, leucocytosis, bandaemia, and leucopenia are neither sensitive nor specific for the diagnosis. The hallmark of sepsis is a positive culture from a normally sterile body site, such as blood, urine, or cerebrospinal fluid, but culture results often take 24 to 48 h to return and are negative in up to one-third of cases. Clinical investigations, such as radiographs, urinalysis, or cerebrospinal fluid examination may demonstrate the site of infection. Other suggestive, but nonspecific laboratory results that may aid diagnosis include: elevated arterial lactate, metabolic acidosis with reduced serum bicarbonate, blood urea nitrogen, creatinine, glucose, bilirubin, alkaline phosphatase, and aminotransferase measurements. An elevated serum procalcitonin level may be a reasonable marker of sepsis in patients with SIRS, but is usually not readily available.
Treatment of sepsis
Early administration of properly chosen antibiotics reduces morbidity and mortality in patients with sepsis. In clinical trials, an appropriate antibiotic regimen is begun in a timely fashion in 85 to 95% of occasions, but failure to do so is associated with a 25% higher overall case fatality. Since culture results are not immediately available for most septic patients, clinicians must use antibiotics empirically. When initiating therapy, the suspected site of infection, patient’s immune status, recent antibiotic use, local resistance patterns, location in which the patient acquired the infection (i.e. nosocomial vs community), and Gram’s stain and culture results should all be considered. Unless the causative organism and susceptibility are known, initial antibiotic therapy should cover a broader spectrum of possibilities when the patient is critically ill. Once microbiological data are available, therapy should be tailored promptly to the narrowest spectrum, least toxic, and least expensive agent.
Searching for the site of infection should begin as soon as haemodynamic and respiratory status are stabilized. Effective management relies on eradicating the source of infection, as antibiotics without source control often fail. Any devitalized tissue or sizeable collection of pus should be resected or drained, either surgically or using less invasive percutaneous drainage techniques. All foreign bodies, including intravascular catheters, should be carefully examined and removed promptly if there is any suspicion of infection.
Early goal-directed resuscitation
The inflammatory cytokines released in sepsis decrease systemic resistance, reduce filling pressure, and depress myocardial contractility. Increased venous pooling, greater insensible losses from anorexia, sweating, vomiting and diarrhoea, and worsening microvascular permeability all contribute to decreased intravascular volume. Consequently, many septic patients rapidly develop cardiovascular insufficiency manifested as hypotension. Aggressive resuscitation, aimed at restoring and maintaining an adequate blood pressure, should be initiated early in the course of treatment. In adults of normal size, this resuscitation often requires large amounts of colloid or 6 to 10 litres isotonic crystalloid. Although the endpoint for discontinuing aggressive resuscitation and the optimal measure of adequate perfusion pressure remains ill-defined, early goal-directed resuscitation (EGDT) during the first 6 h has been shown to lower mortality in one single-centre study of patients with severe sepsis and lactic acidosis or septic shock. Intention-to-treat analysis demonstrated that patients receiving EGDT during the first 6 h of care had a 33% relative reduction and 16% absolute reduction in hospital mortality (30.5% vs 46.5%; P = 0.009), with improvement in mortality persisting to at least 60 days. The reduction in mortality was largely attributable to a decrease in late, sudden cardiovascular collapse in patients treated with EGDT.
Treatment of hypotension
The EGDT algorithm utilizes crystalloid infusions to maintain a central venous pressure of 8 to 12 mmHg. If mean arterial pressure remains below 65 mmHg, vasopressors are initiated. During resuscitation, central venous oxygen saturation (ScvO2) is monitored continuously using a central venous catheter. If the patient has a central venous pressure and mean arterial pressure within the target ranges, but the ScvO2 measurement remains below 70%, packed red blood cells are transfused to achieve a haematocrit of 30%. If the haematocrit is 30% or higher and the ScvO2 is still below 70%, dobutamine is begun. Vasopressin, a stress hormone released in response to hypotension, stimulates a family of receptors: AVPR1a, AVPR1b, AVPR2, oxytocin receptors, and purinergic receptors. In septic shock, vasopressin secretion is inadequate. Low-dose vasopressin infusion improves blood pressure, decreases requirements for noradrenaline, improves renal function and, in the Vasopressin and Septic Shock Trial (VASST), low-dose vasopressin decreased mortality in patients with less severe septic shock (26% vs 36%, respectively, P = 0.04). Patients with ‘less severe shock’ were defined as those on modest doses of noradrenaline (5 to 14 μg/minute) at randomization and with lower serum lactate concentrations. However, in the whole group of patients, there was no difference in 28-day mortality between those treated with vasopressin and with noradrenaline (35% vs 39%, respectively). Low-dose vasopressin infusion plus corticosteroids significantly decreased 28-day mortality compared with corticosteroids plus noradrenaline (44% vs 35%, respectively, P = 0.03; P = 0.008 interaction statistic).
In one study, 54% of patients with septic shock who had adrenal insufficiency defined by basal plasma cortisol levels above 34 μg/dL, and cortisol response to short corticotrophin stimulation test below 9 μg/dL, had a higher mortality. Impairement of hypothalamo–pituitary–adrenal responses during critical illness may result from decreased production of corticotropin-releasing hormone (CRH), ACTH, and cortisol and dysfunction of their receptors such as CRH receptor 1 (CRHR1) and AVPR1b. Although corticosteroids possess a variety of anti-inflammatory actions, numerous studies have demonstrated that high doses, aimed at suppressing the inflammatory response to infection, confer no benefit to patients with severe sepsis or septic shock. Meta-analyses of these studies confirm the lack of efficacy and even suggest a trend towards harm. More recent data suggest that lower, more physiological doses, administered over a longer period of time, may benefit a subset of patients with septic shock and relative adrenal insufficiency. Unfortunately, a multinational double-blind placebo-controlled study investigating this ‘replacement dose’ strategy found no difference in 28-day all-cause mortality or shock-free days. Given the numerous negative studies, the routine use of corticosteroids in patients with severe sepsis or septic shock cannot be recommended at this time. However, a recent systematic review suggested that low dose corticosteroid might improve survival in patients with severe sepsis.
Eritoran tetrasodium, anti-Toll-like receptor (TLR)-4
Unfortunately, the attempt to manipulate TLR4 signalling with the new anti-Toll-like receptor (TLR)-4 compound eritoran tetrasodium (Eisai) failed to demonstrate improvement in 28-day all-cause mortality in a cohort of 2000 patients with severe sepsis.
Drotrecogin alfa (activated) or recombinant human activated protein C
In late 2001, recombinant human activated protein C (rhAPC), or drotrecogin alfa (activated) (DAA), became the first drug approved for use in patients with a high risk of death from severe sepsis. A large phase III randomized blind placebo-controlled multinational trial was discontinued early, after 1690 of the planned 2280 patients were enrolled, because the predefined stopping boundary for efficacy was surpassed. rhAPC, administered as a continuous intravenous infusion of 24 µg/kg per h for 96 h, resulted in a 6% absolute and 19% relative reduction in 28-day all-cause mortality compared to placebo (24.7% vs 30.8%; P = 0.005). The survival difference became apparent shortly after initiation of the infusion, continued to increase for the duration of the study period, and persisted out to at least 1 year of long-term follow-up. Subsequent prospective data suggested that early treatment with rhAPC (i.e. within 24 h of organ dysfunction onset) resulted in better outcomes than delayed initiation. Although the criteria for high risk of death are hotly debated, patients with Acute Physiology and Chronic Health Evaluation II (APACHE II) scores of more than 25 or requiring vasopressors or mechanical ventilation are widely considered appropriate candidates for the drug.
Not unexpectedly given its anticoagulant properties, rhAPC increases the risk of bleeding. Although uncommon in both arms of the study, severe bleeding episodes were also almost twice as high in those receiving rhAPC compared to placebo (3.5% vs 2.0%; P = 0.06). These bleeding rates are similar to those seen with other forms of full systemic anticoagulation, such as full-dose heparin. Although patients at high risk of bleeding were excluded from the landmark study, those with either traumatic injury of a highly vascular organ or major blood vessel, ulcerations in the gastrointestinal system, meningitis, or markedly abnormal coagulation parameters (platelet counts <30 000/ml, INR >3, or activated partial thromboplastin time >120 s) had more bleeding.
Subsequent studies in adult patients with low risk of death (i.e. APACHE II <25, not on vasopressors, not in the intensive care unit) and children were stopped early because of inefficacy and side effects. Analysis of these data found that children with sepsis and postoperative patients with single-organ failure from sepsis did not derive benefit from administration of rhAPC, and both groups experienced higher bleeding rates. The drug is therefore not recommended for use in these populations. Unfortunately, in October 2011, Eli Lilly announced withdrawal of drotrecogin alfa (Xigris) because of the failure of its PROWESS shock trial to demonstrate improved outcome. This was the only drug approved specifically for the treatment of severe sepsis.
Volume-limited and pressure-limited mechanical ventilation strategy
The vast majority of patients with severe sepsis require mechanical ventilation, many from sepsis-induced acute lung injury. Preventing undue distension of normally compliant segments of the injured lung, and subsequent release of inflammatory cytokines by using a volume-limited and pressure-limited ventilation strategy in patients with acute lung injury has been shown to decrease mortality and increase time alive and off the ventilator. To accomplish this, patients should be ventilated using volume-limited ventilation strategies with tidal volumes set at 6 ml/kg of predicted body weight. These tidal volumes should be titrated downwards as needed to maintain plateau pressure less than 30 cm H2O. Ventilator weaning that is protocol-driven and employs daily spontaneous breathing trials results in earlier successful extubation.
Preventing complications and nosocomial infections
Since patients with severe sepsis possess multiple risk factors for deep venous thrombosis, prophylaxis for this complication should be nearly routine. Patients with low bleeding risk should receive a known effective dose of low molecular weight or unfractionated heparin, while intermittent compression devices should be used in patients at significant risk of bleeding. H2-receptor antagonists have been shown to be superior to treatment with sucralfate in prevention of gastrointestinal bleeding for mechanically ventilated patients. Therefore, current therapy should almost always include an H2-receptor blocker or a proton pump inhibitor.
The compensatory anti-inflammatory response and impaired host defence, along with numerous invasive procedures and broad-spectrum antibiotic exposure, predisposes patients with severe sepsis to nosocomial infections. Hand washing between patients and using barrier precautions (gloves and gowns) when examining patients colonized with resistant organisms should be universally employed. Likewise, inserting central lines using full barrier precautions, limiting the number of catheter manipulations, and utilizing closed infusion systems with minimal tubing changes can minimize vascular catheter infections. The risk of nosocomial pneumonia can be reduced by raising the head of the bed to 30 to 45 degrees in mechanically ventilated patients, especially those receiving enteral feedings. Draining the condensate from ventilator tubing and minimizing tubing changes also decreases the incidence of nosocomial pneumonia. If possible, all feeding and endotracheal tubes should be orally placed to reduce the incidence of sinusitis. If a tube must be placed through the nose, it should be small-bore and flexible to decrease the degree of sinus ostial obstruction.
Sepsis without organ dysfunction is a very serious condition, with in-hospital mortality rates ranging from 5 to 10%. However, once organ dysfunction is present, even with modern advances in the care of the critically ill, one-third to one-half of all patients with severe sepsis die before being discharged from hospital, and patients with septic shock have hospital mortality rates of 50 to 80%. Patients who survive their initial encounter with severe sepsis continue to have higher rates of death throughout the first year after hospital discharge for unknown reasons. The principal prognostic factors include age, severity of underlying diseases, number of organ system dysfunctions, severity of illness scores, hypothermia, neutropenia, thrombocytopenia, lactic acidosis, multisource of infection, positive blood culture, type of infecting organism and blood concentrations of endotoxin and cytokines.
Areas of uncertainty
As with all diseases, areas of uncertainty exist in treating patients with severe sepsis.
The use of intensive insulin therapy, aimed at maintaining serum blood glucose levels between 80 and 110 mg/dl, also remains controversial. An initial study demonstrated that this tight glycaemic control improved survival, decreased bacteraemia, and reduced renal failure in cardiac surgery patients, regardless of a history of diabetes. A subsequent study in critically ill medical patients was unable to replicate the results, although some benefit was seen in patients requiring intensive care for longer than 72 h. Although hypoglycaemia is uncommon, patients who develop it have significantly higher mortality rates. Further research into this area will need to be undertaken to identify the optimal patient population and target glucose levels for intensive insulin treatment in patients with sepsis.
Aird WC (2001). Vascular bed-specific hemostasis: role of endothelium in sepsis pathogenesis. Crit Care Med, 29 Suppl 7, 28–35.Find this resource:
Angus DC (2011). The search for effective therapy for sepsis: back to the drawing board? JAMA, 306 2614–5.Find this resource:
Annane D, et al. (2000). A 3-level prognostic classification in septic shock based on cortisol levels and cortisol response to corticotropin. JAMA, 283, 1038–45.Find this resource:
Annane D, et al. (2002). Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock. JAMA, 288, 862–71.Find this resource:
Annane D, et al. (2009). Corticosteroids in the treatment of severe sepsis and septic shock in adults: a systematic review. JAMA, 301, 2362–75.Find this resource:
Bernard GR, et al. (2001). Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med, 344, 699–709.Find this resource:
Bone RC (1996). Sir Isaac Newton, sepsis, SIRS and CARS. Crit Care Med, 24, 1125–8.Find this resource:
Boomer JS, et al. (2011). Immunosuppression in patients who die of sepsis and multiple organ failure. JAMA, 306, 2594–605.Find this resource:
Cook DJ, et al. (1998). A comparison of sucralfate and ranitidine for the prevention of upper gastrointestinal bleeding in patients requiring mechanical ventilation. N Engl J Med, 338, 791–7.Find this resource:
Dellinger RP, et al. (2004). Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock. Crit Care Med, 32, 858–73.Find this resource:
Hotchkiss RS, Karl IE (2003). The pathophysiology and treatment of sepsis. N Engl J Med, 348, 138–50.Find this resource:
Kollef MH, et al. (1999). Inadequate antimicrobial treatment of infections: a risk factor for hospital mortality among critically ill patients. Chest, 115, 462–74.Find this resource:
Levi M, et al. (1993). Pathogenesis of disseminated intravascular coagulopathy in sepsis. JAMA, 270, 975–9.Find this resource:
MacIntyre NR, et al. (2001). Evidence-based guidelines for weaning and discontinuing ventilatory support: a collective task force facilitated by the American College of Chest Physicians; the American Association for Respiratory Care; and the American College of Critical Care Medicine. Chest, 120 Suppl 6, S375–95.Find this resource:
Martin GS, et al. (2003). The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med, 348, 1546–54.Find this resource:
Mitka M (2011). Drug for severe sepsis is withdrawn from market, fails to reduce mortality. JAMA, 306, 2439–40.Find this resource:
Rivers E, et al. (2001). Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med, 345, 1368–77.Find this resource:
Russell JA (2011). Bench-to-bedside review: Vasopressin in the management of septic shock. Crit Care, 15, 226.Find this resource:
Russell JA, et al. (2008). Vasopressin versus norepinephrine infusion in patients with septic shock. N Engl J Med, 358, 877–87.Find this resource:
The Acute Respiratory Distress Syndrome Network (2000). Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med, 342, 1301–8.Find this resource:
Ulloa L, et al. (2002). Ethyl pyruvate prevents lethality in mice with established lethal sepsis and systemic inflammation. Proc Natl Acad Sci U S A, 99, 12 351–6.Find this resource:
van den Berghe G, et al. (2001). Intensive insulin therapy in critically ill patients. N Engl J Med, 345, 1359–67.Find this resource:
van den Berghe G, et al. (2006). Intensive insulin therapy in the medical ICU. N Engl J Med, 354, 449–61.Find this resource:
van Deventer SJH, et al. (1990). Experimental endotoxemia in humans: analysis of cytokine release and coagulation, fibrinolytic and complement pathways. Blood, 76, 2520–6.Find this resource:
Wang H, et al. (1999). HMG-1 as a late mediator of endotoxin lethality in mice. Science, 285, 248–51.Find this resource:
Wiersinga WJ (2011). Current insights in sepsis: from pathogenesis to new treatment targets. Curr Opin Crit Care, 17, 480–6.Find this resource:
Yang H, et al. (2004). Reversing established sepsis with antagonists of endogenous HMGB1. Proc Natl Acad Sci U S A, 101, 296–301.Find this resource: