ABSTRACT: A milestone has been reached in the treatment of sepsis—the institution of protocolized management that starts in the emergency department. Early goal-directed therapy, with targeted fluid resuscitation and measures of oxygen delivery, has been shown to improve survival in patients with septic shock. Although initiating aggressive fluid resuscitation is the first priority, it is also essential to rapidly obtain cultures and infuse broad-spectrum antibiotics. Norepinephrine(Drug information on norepinephrine) is a more potent vasoconstrictor than dopamine(Drug information on dopamine) and may be more effective in treating hypotension in patients with septic shock. Vasopressin(Drug information on vasopressin) is an effective second-line agent. Treatment with recombinant human activated protein C at 24 µg/kg/h for 96 hours has been shown to reduce mortality in patients with sepsis; its benefit is greatest in the most acutely ill patients. (J Respir Dis. 2009;30[1-2])
For the diagnosis and management of sepsis, the past 10 years have brought significant changes. Initial treatment involving fluid resuscitation, antibiotics, and source control is similar to what it was a decade ago, but we now realize that the speed of diagnosis coupled with expedited management is critical.1 New ways to support failing organ systems2 and the introduction of a pharmacological agent to reduce mortality3 have shed light on this disease and have given clinicians new optimism. Slowly but surely, we are making the transition toward delivering a multitude of treatments simultaneously in an effort to improve clinical care.
In this 2-part article, we will review the latest strategies for managing sepsis.
THE SEPSIS CONTINUUM
Sepsis is a significant public health problem that occurs at the rate of approximately 3 cases per 1000 persons, affecting nearly 900,000 persons every year in the United States and nearly 20,000,000 worldwide.4 As the leading cause of death in noncoronary ICUs and the 10th leading cause of death overall in the United States,4-6 sepsis has a tremendous economic and social burden. Not only does sepsis mortality range from 20% to 80% depending on illness severity4,5,7-9 but survivors experience a significant increase in morbidity and reduced quality of life.10,11
Recent data from the United States show that the incidence of sepsis is increasing, with a growing number of deaths despite an overall decrease in proportionate mortality.4 With the aging population, increased use of immunosuppressive drugs, emergence of HIV, increasing microbial resistance, and expanding availability of health care and health care interventions, the incidence of sepsis will continue to increase globally.
DEFINITION AND CASE FINDING
The diagnosis of sepsis can still be perplexing despite a well-accepted consensus definition. In 1992, the American College of Chest Physicians/Society of Critical Care Medicine consensus conference arrived at the current definition of sepsis as a systemic inflammatory syndrome (defined by a change in 2 or more abnormal clinical findings: temperature, heart rate, respiration rate, and white blood cell count) with a concomitant pathological infection (Table 1).12
Sepsis severity was defined by the addition of acute organ dysfunction, hypoperfusion, or hypotension, based on criteria proposed by Marshall and associates13 or by criteria used for Sequential Organ Failure Assessment score.14 Septic shock refers to sepsis-induced hypotension that persists despite adequate fluid resuscitation.
Sepsis is difficult to diagnose, particularly in the earlier stages when the symptoms may be subtle. Although it would be clearly beneficial to have an accurate test that identifies sepsis, no diagnostic test exists and the recognition of early sepsis often requires an astute clinician who is knowledgeable about the sepsis syndrome.
Various biomarkers have been evaluated for diagnosis, risk stratification, and prognosis in sepsis, including procalcitonin, C-reactive protein (CRP), B-type natriuretic peptide,15,16 and protein C. Although the procalcitonin level has limited diagnostic value in patients with systemic inflammatory response syndrome from other causes,17 it appears to be a better marker for illness severity and prognosis than the CRP level.18-21 The directional change in protein C levels have been shown to correlate with outcomes in patients with severe sepsis22 and may prove to be useful tools in the future.
A soluble triggering receptor expressed on myeloid cells-1, a recently discovered receptor expressed on the surface of neutrophils, has been reported to trigger the synthesis of proinflammatory cytokines in presence of microbial products23 and has been found to predict outcome in patients with sepsis.24 However, none of the above-mentioned biomarkers has thus far proved clinically useful for the diagnosis of sepsis.
PATHOPHYSIOLOGY
Over the years, a considerable amount has changed in our thinking about sepsis pathophysiology. Initially considered a syndrome of exaggerated inflammation,12 sepsis is now recognized as a complex set of interactions between the inciting microbe, the host immune response, and the inflammatory and coagulation pathways.
Inflammation and immune response
An infectious insult, classically described as bacterial endotoxin, initiates a pathophysiological cascade involving pattern-recognition receptors called toll-like receptors.25 Binding of these receptors to microorganisms results in the release of a number of proinflammatory cytokines, especially tumor necrosis factor-α, which are important for host immune defense and resolution of the inflammatory response as they interact with invading pathogens.26,27 These pathways often yield further activation of other myeloid-derived and/or endothelial cells.28
Simultaneously, activation of anti-inflammatory pathways may lessen the inflammatory response.29
Coagulation imbalance
In recent years, it has been recognized that the coagulation system acts in concert with the inflammatory cascade in the pathophysiology of sepsis.30-33 These abnormalities range from subclinical prolongation of clotting times to fulminant disseminated intravascular coagulation, characterized by global microvascular thrombosis and bleeding.30 Amelioration of this coagulopathy appears to attenuate organ failure and, subsequently, survival.3,32
The protein C pathway serves as an anticoagulant system, promoting fibrinolysis by inhibiting thrombosis and inflammation.34 Thrombin binds to thrombomodulin at the endothelial protein C receptor (EPCR) on the endothelium, resulting in a complex that rapidly activates protein C, which binds to protein S, inactivating factors Va and VIIIa.35,36 EPCR deletion exaggerates the host responses to lipopolysaccharide,33 suggesting that EPCR is important in controlling endotoxin-induced coagulation and inflammatory responses.
Multiple organ dysfunction syndrome
Multiorgan system dysfunction in sepsis may be partially caused by relative host immunosuppression. Stimulation by an infectious insult may cause a shift to an anti-inflammatory milieu, resulting in production of anti-inflammatory cytokines.37,38
In addition, apoptosis of circulating immune, epithelial, and endothelial cells induced by endotoxin and proinflammatory cytokines inducing cytopathic hypoxia can contribute to this immunosuppression.39,40 These altered signaling pathways lead to tissue injury and multiorgan dysfunction.
MANAGEMENT
In the past decade, there have been significant developments demonstrating the importance of speed and accuracy in diagnosing sepsis and instituting appropriate care.
Initial resuscitation
• Early goal-directed therapy: A milestone has been reached in the treatment of sepsis—the institution of protocolized management that starts in the emergency department (ED). Early goal-directed therapy, with targeted fluid resuscitation and measures of oxygen delivery, has been shown to improve survival in patients with septic shock.1
This evidence comes from a randomized controlled single-center trial that assigned 263 patients with severe sepsis or septic shock to receive protocolized early goal-directed therapy during the first 6 hours in the ED or to receive standard therapy before ICU admission.1 Depending on whether patients had central venous catheters or pulmonary artery catheters, central venous oxygen saturation (ScvO2) or mixed venous oxygen saturation (SvO2) was continuously measured in those receiving early goal-directed therapy and dictated further therapy if goals were not achieved.
Resuscitation goals for the initial 6 hours included a central venous pressure of 8 to 12 mm Hg, mean arterial pressure (MAP) greater than 65 mm Hg, urinary output greater than 0.5 mL/kg/h, and an ScvO2 or SvO2 of greater than 70% or 65%, respectively. If ScvO2 was less than 70%, packed red blood cells (PRBCs) were transfused to achieve a hematocrit level of 30%. If central venous pressure, MAP, and hematocrit were optimized and ScvO2 remained less than 70%, dobutamine(Drug information on dobutamine) was added to increase cardiac output and oxygen delivery (Table 2).
Resuscitation to these goals within 6 hours reduced in-hospital mortality in patients with severe sepsis from 46.5% to 30.5% (P < .009) and reduced mortality at 28 days (P = .01) and at 60 days (P = .03).1 Although this evidence comes from a single-center trial, this study reliably showed that early goal-directed resuscitation improved survival for ED patients with septic shock, and these goals have now become part of the treatment protocol in patients with sepsis-induced shock in recent Surviving Sepsis Campaign (SSC) guidelines.41 In addition, a growing number of studies have supported the use of early protocolized resuscitation to improve outcomes.42-45
During the initial 6 hours, patients in the early goal-directed therapy group received significantly more fluids, PRBCs, and inotropic support than those in the standard therapy group, and they required less intensive hospital care and had less severe illness through 72 hours.1 Of note, in mechanically ventilated patients, a higher target central venous pressure of 12 to 15 mm Hg is generally recommended to account for the impediment to cardiac filling caused by higher intrathoracic pressures.46
The design of this trial does not allow one to assess the relative contributions of the various components, particularly PRBC transfusion and dobutamine. Thus, once a patient has been adequately resuscitated with fluid therapy, it is up to the clinician whether PRBC transfusion or dobutamine is the best initial choice to maximize oxygen delivery.41
• Colloids versus crystalloids: Meta-analyses of studies evaluating the use of crystalloids versus colloids for resuscitation have had conflicting results, such that recent SSC guidelines cannot recommend with good evidence one type of fluid over the other.47-50 Although some studies have indicated that certain colloids cause organ damage, the Saline Versus Albumin Fluid Evaluation (SAFE) study, a multicenter randomized trial that compared 4% albumin with 0.9% normal saline in patients requiring fluid resuscitation, indicated that albumin was as safe as crystalloid and equally effective.51
There was no difference in 28-day all-cause mortality rate, organ dysfunction, the duration of mechanical ventilation, or ICU or hospital length of stay.51 There was a nonsignificant decrease in mortality in the subgroup of patients with severe sepsis treated with albumin (P = .09) and an increase in mortality in albumin-treated patients who had traumatic brain injury.51,52
A few studies have evaluated the use of hydroxyethyl starch (HES) for fluid resuscitation.53,54 In animal models, HES has been shown to improve microcirculation during endotoxemia,55-58 but it also has been known to have adverse effects, such as coagulopathy and renal failure.59,60 In a multicenter randomized trial, Schortgen and coworkers54 found that use of HES was an independent risk factor for acute renal failure in patients with severe sepsis or septic shock.
A recent randomized controlled trial comparing the use of HES with that of Ringer lactate in patients with severe sepsis and septic shock was suspended after the first preplanned interim analysis because of an increased risk of renal failure and a trend toward increased mortality at 90 days in the HES group.61
Antibiotic therapy
Although initiating aggressive fluid resuscitation is the first priority when managing severe sepsis or septic shock, it is also important to rapidly obtain cultures of suspected body fluids or blood and promptly infuse broad-spectrum antibiotics. It is advisable to obtain cultures before starting antimicrobial therapy (since rapid sterilization of blood cultures can occur within a few hours), but antibiotic therapy should not be unduly delayed.
A large study of 18,209 Medicare patients hospitalized with community-acquired pneumonia showed that antibiotic use within 4 hours of arrival at the hospital was associated with decreased mortality and hospital length of stay.62 In addition, in the presence of septic shock, each hour delay in antimicrobial administration has been found to decrease survival.63
Initially, empirical antibiotic therapy should be broad enough to cover all possible pathogens. It is important for physicians to be aware of their hospital’s antibiotic profile, in addition to the virulence patterns of pathogens in the community. When selecting antibiotics, care providers must not only guess early, but they must also guess right. Failure to initiate adequate antimicrobial therapy correlates with increased morbidity and mortality in patients with sepsis admitted to the ICU.64-67
Once the causative pathogen has been identified, the antibiotic regimen should be narrowed. However, this is not an appropriate initial strategy, and the desire to minimize superinfections and resistance should not take precedence over adequately treating patients with severe sepsis and septic shock.
Although combination therapy has never been shown to significantly improve outcomes,68-70 multiple antibiotics may be useful in certain situations. Recent guidelines suggest that combination therapy be used for neutropenic patients and for patients with known or suspected Pseudomonas infections as a cause of severe sepsis.41 When used empirically, however, combination therapy should not be continued for longer than 3 to 5 days.
An observational study of patients with ventilator-associated pneumonia (VAP) showed that monotherapy was associated with inappropriate therapy, which, in turn, was associated with increased in-hospital mortality.71 This suggests that initial use of combination therapy reduces the likelihood of inappropriate therapy, thereby reducing the risk of death. Most recently, a randomized multicenter clinical trial showed that although there was no difference in 28-day mortality between patients with VAP who were treated with combination antibiotic therapy and those who were treated with monotherapy, the subgroup of patients at high risk for difficult-to-treat gram-negative bacterial infection had better microbiological and clinical outcomes with combination antibiotic therapy.68 The recommended duration of antimicrobial therapy is 7 to 10 days, although longer courses may be appropriate in some patients.41
Hemodynamic management—use of vasopressors
Despite fluid resuscitation, vasopressor therapy is occasionally required. Below a certain MAP, autoregulation of pressure in vascular beds can be lost and perfusion becomes linearly dependent on pressure.72,73 However, loss of auto-regulation can occur at different levels in different organs.
Titration of norepinephrine to an MAP as low as 65 mm Hg has been shown to preserve tissue perfusion in a small study of 10 patients,72 and increasing the vasopressor dose to maintain a higher MAP (85 mm Hg) does not significantly affect metabolic parameters or renal function.72,74
However, the patient’s baseline blood pressure should also be considered. Someone with chronic systemic arterial hypertension might require an MAP greater than 65 mm Hg to maintain tissue perfusion, while a lower MAP may be adequate in a patient with chronic hepatic failure. Therefore, health care providers must always supplement arterial pressure with other measures of global tissue perfusion, such as ScvO2, tissue oximetry, blood lactate levels, delayed capillary refill, and urinary output.41,72
There has been a long-standing debate about vasopressor superiority. Although these discussions are intellectually stimulating, remember that each catecholamine agent has variable receptor-mediated effects, so distinct clinical situations may require different vasopressors. For example, norepinephrine has potent α-adrenergic effects and less potent β-adrenergic effects, while dopamine’s receptor effects are dose-dependent.
Some studies suggest that norepinephrine or dopamine may have some advantages over other vasopressors, and recent guidelines recommend either norepinephrine or dopamine as a first-choice vasopressor agent to correct hypotension in septic shock.41 Norepinephrine is a more potent vasoconstrictor than dopamine and may be more effective in treating hypotension in patients with septic shock. In the only randomized trial comparing these vasopressors, the ability of dopamine and norepinephrine to reverse hemodynamic derangements associated with septic shock was evaluated in 32 patients.75 More patients were successfully treated with norepinephrine, including those who did not respond to dopamine.
In a larger, observational study, the use of norepinephrine as the vasopressor of choice was associated with lower hospital mortality in patients with septic shock.76 A prospective, double-blind, multicenter, randomized, controlled trial with 1600 patients is under way to compare the efficacy of dopamine with that of norepinephrine in the treatment of shock.77
Epinephrine has been suggested as the first-choice alternative in septic shock that is poorly responsive to norepinephrine or dopamine. Annane and colleagues78 conducted a randomized controlled trial of 330 patients with septic shock to compare the efficacy of norepinephrine plus dobutamine with that of epinephrine(Drug information on epinephrine). They found that there was no difference in efficacy or safety between the 2 groups. In the past, smaller studies have found that patients treated with epinephrine tend to have inadequate, albeit transient, splanchnic oxygen utilization, resulting in a higher incidence of gastric mucosal acidosis.79
Vasopressin, an endogenous hormone synthesized in the hypothalamus, has emerged as an adjunct to catecholamines for patients with septic shock. Vasopressin levels have been found to be lower than expected in patients with septic shock, suggesting a relative vasopressin deficiency state.80,81 In addition, vasopressin has been found to spare catecholamine use and have other beneficial physiological effects.80-84
Recently, a large, multicenter, randomized, double-blind trial of approximately 800 patients was conducted to determine whether norepinephrine and vasopressin at 0.03 U/min decreased mortality compared with norepinephrine alone.85 There was no difference in mortality, ICU and hospital length of stay, days alive and free of vasopressor use, use of corticosteroids, or organ dysfunction, but the dose of norepinephrine infusion was significantly lower in the group receiving vasopressin. Although there was no difference in the rates of adverse events overall, there was a trend toward a higher rate of cardiac arrest in the norepinephrine group and a trend toward a higher rate of digital ischemia in the vasopressin plus norepinephrine group.
This study demonstrated that although vasopressin is an effective second-line agent, it is not more effective than using norepinephrine alone. Vasopressin, however, may be used at low doses (0.03 U/min), particularly for refractory hypotension, and should be reserved for a specific subset of patients (those without coronary or mesenteric ischemia or those who are unable to tolerate high doses of norepinephrine).
Recombinant human activated protein C
The PROtein C Worldwide Evaluation in Severe Sepsis (PROWESS) study randomized 1690 patients with severe sepsis to receive either recombinant human activated protein C (rhAPC) at 24 µg/kg/h for 96 hours or placebo.86 Treatment reduced absolute mortality by 6.1% and relative mortality by 19.4% (P = .005). The benefit was greatest—a decrease in absolute mortality of 13%—in the most acutely ill patients (those with APACHE II scores of 25 or higher).86
Subsequently, in the Administration of Drotrecogin Alfa [Activated] in Early Stage Severe Sepsis (ADDRESS) trial, 2613 patients assessed to have a low risk of death (generally, an APACHE II score of less than 25 or single-organ dysfunction) were randomized to receive rhAPC or placebo; this study found no difference in 28-day mortality.87 Importantly, the ADDRESS trial showed that the 28-day mortality rate was significantly higher in patients who had had recent surgery and who had single-organ dysfunction who were treated with rhAPC (20.7% vs 14.1%, P = .03).
Faced with critically ill patients with severe sepsis, intensivists must decide whom to treat with rhAPC based on the extensive list of exclusion criteria in the clinical studies and contraindications. Serious adverse events did not differ in the 2 studies except for serious bleeding during infusion, which was increased in the group that received rhAPC (PROWESS, 3.5% vs 2%, P = .06; and ADDRESS, 3.9% vs 2.2%, P < .01)3,87 and tended to occur in patients with a predisposition to bleeding (such as those with GI ulceration or coagulopathy and those undergoing procedures).
Additional safety information came from an open-label observational study, Extended Evaluation of Recombinant Human Activated Protein C (ENHANCE), which showed a 28-day all-cause mortality similar to that of PROWESS (25.3% vs 24.7%) but a 3.6% increase in serious bleeding during infusion and a 6.5% increase at 28 days.88 In addition, the rate of intracranial hemorrhage was increased in patients who received rhAPC in ENHANCE versus PROWESS (1.5% vs 0.2%).88
Recently, the use of prophylactic heparin(Drug information on heparin) was evaluated in patients who received rhAPC.89 There was no harmful interaction between rhAPC and heparin and, in fact, there was a nonsignificant reduction in mortality in patients who received heparin.89
Given all this information, critical care physicians must decide which patients will benefit from rhAPC and whether the benefit exceeds the risk in each patient. rhAPC is an expensive drug, but multiple analyses have reported its cost-effectiveness in patients who have severe sepsis and a high predicted mortality at baseline.90-92 Although the conclusions from PROWESS and ADDRESS are limited, there still is a probable mortality reduction in patients with sepsis-induced organ dysfunction associated with a high risk of death based on clinical assessment, most of whom will have an APACHE II score greater than 25 or multiorgan failure. Therefore, it is suggested that rhAPC be considered for these patients, presuming there are no contraindications.41
Corticosteroids
The past decade has seen a considerable debate and the emergence of new evidence regarding the use of corticosteroids in septic shock. In the past, randomized clinical trials and meta-analyses have shown that high-dose corticosteroid therapy is ineffective in patients with severe sepsis or septic shock.93-96 Until recently, there was 1 multicenter randomized controlled trial that suggested better shock reversal and a survival benefit in patients with vasopressor-unresponsive septic shock and relative adrenal insufficiency, defined as a post–adrenocorticotropic hormone (ACTH) cortisol level increase of 9 µg/dL or less.97 Two smaller single-center studies also suggested that there was a greater incidence of shock reversal with corticosteroids.98-99
A large European multicenter trial, Corticosteroid Therapy of Septic Shock (CORTICUS), randomized 499 patients with septic shock to receive either low-dose hydrocortisone(Drug information on hydrocortisone) therapy or placebo for 5 days.100 The authors concluded that at 28 days, there was no significant difference in mortality between patients in the 2 treatment groups, irrespective of any response to ACTH.100 While corticosteroids hastened the reversal of septic shock, they were also associated with a greater risk of nosocomial infections and recurrent sepsis.100 These results suggest that ACTH stimulation testing is not useful in predicting which patients with sepsis may benefit from corticosteroids and that corticosteroid therapy in general does not improve clinical outcomes in patients with septic shock.
Corticosteroids are not without adverse effects. These drugs are immunosuppressive, potentially leading to secondary infections and impaired wound healing, and can cause myopathy, hyperglycemia, and hypernatremia.98-100 Thus, corticosteroid therapy should be discontinued as early as possible.41 However, to date no study has compared a fixed duration of corticosteroid therapy followed by tapering over several days92,93 or abrupt discontinuation91 versus tapering therapy after shock resolution,99 so it remains uncertain whether outcome is affected by tapering of corticosteroids.
However, despite these controversies, corticosteroids are still suggested only for patients with septic shock, since no studies suggest a benefit in patients with less severe forms of sepsis whose blood pressure is poorly responsive to fluids and vasopressor therapy.41
Treatment of anemia in sepsis
Anemia is a common feature among critically ill patients with sepsis who often require transfusions of PRBCs.101 Hemoglobin concentrations typically decline during the first few days of an ICU stay, but while they tend to stabilize in patients without sepsis, hemoglobin concentrations continue to decline in patients with sepsis.102 To date, the optimum hemoglobin level in patients with sepsis has not been evaluated, but it certainly varies according to the phase (early vs late) of sepsis.
Two main randomized controlled trials have evaluated transfusion strategies in patients with sepsis.1,103 Early goal-directed therapy includes transfusion of PRBCs to a target hematocrit of 30% or higher during the first 6 hours of resuscitation for septic shock.1 Although not limited to patients with sepsis, the Transfusion Requirements in Critical Care (TRICC) trial compared 2 transfusion strategies in 838 euvolemic critically ill patients.103 After initial resuscitation, the patients were randomized to either a restrictive (maintenance of hemoglobin level at 7.0 to 9.0 g/dL with a threshold of 7.0 g/dL) or a liberal (maintenance of hemoglobin level of 10.0 to 12.0 g/dL with a threshold of 10.0 g/dL) transfusion strategy that was adhered to throughout the patient’s ICU stay.
There was no difference in 30-day all-cause mortality between the 2 groups overall, but there was improved survival in the younger and less severely ill patients.103 Sepsis was the primary diagnosis in only 5% of patients, limiting the interpretation of these results for this population. Together, these studies suggest that PRBC transfusion is valuable during the early stage of sepsis as opposed to later stages when patients are more likely to be euvolemic—transfusion to a hematocrit level of 30% or higher decreases mortality during the first 6 hours of resuscitation, while maintenance of hemoglobin levels of 7.0 to 9.0 g/dL is adequate after initial resuscitation.
Recombinant human erythropoietin(Drug information on erythropoietin) has also been evaluated as a therapy for critically ill anemic patients. Patients with sepsis exhibit inappropriately low levels of erythropoietin,104,105 and 2 previous trials involving critically ill patients showed that erythropoietin therapy reduced red cell transfusions but did not decrease mortality.106,107
In a large randomized trial of 1460 patients, of which 13% had sepsis, erythropoietin therapy did not reduce the incidence of PRBC transfusion among critically ill patients.108 While mortality was reduced in the trauma patients, the incidence of thrombotic events was increased in all patients who received erythropoietin therapy.108 As such, current guidelines do not recommend erythropoietin for the treatment of anemia associated with sepsis.41
[Editor’s note: In part 2, Drs Cribbs and Martin will continue their review of the management of sepsis.]
