131 Sepsis and Multiple Organ System Failure in Children

Definitions of Sepsis, Severe Sepsis, Septic Shock, and Multiple Organ Failure

The 2001 International Sepsis Definitions Conference ¹ centered discussion on whether sepsis should continue to be defined as systemic inflammatory response syndrome plus infection or infection plus systemic inflammatory response syndrome plus signs of organ dysfunction. It was agreed that the definitions of severe sepsis remain intact. Most pediatric literature defines inclusion criteria for sepsis as hyperthermia or hypothermia, tachycardia (may be absent in the hypothermic patient), evidence of infection, and at least one of the following signs of new-onset organ dysfunction: altered mental status, hypoxemia, bounding pulses, or increased lactate. Severe sepsis is uniformly defined as sepsis and organ failure determined by various organ failure scores.²^–⁵ Septic shock has been defined as infection with hypothermia or hyperthermia, tachycardia (may be absent with hypothermia), and altered mental status in the presence of at least one, but usually more than one, of the following: decreased peripheral pulses compared with central pulses prolonged greater than 2 seconds (cold shock) or flash capillary refill (warm shock), mottled or cool extremities (cold shock), and decreased urine output (<1 mL/kg/h). Hypotension is observed in late decompensated shock.⁶

The American College of Critical Care Medicine ⁶ further defines shock according to response to therapy as fluid-refractory/dopamine-resistant, catecholamine-resistant, and refractory shock. Multiple organ failure is defined as more than one organ failure. The greater the number of concomitant organ failures, the greater the risk of mortality. Multiple organ failure generally is observed in septic shock patients who receive delayed resuscitation or inadequate source control therapies (inadequate nidus removal or ineffective antibiotic regimen). Multiple organ failure also is observed in patients with septic shock who have an underlying primary or acquired immunodeficiency that prevents timely eradication of infection and resolution of inflammation.

Changing Outcomes and Epidemiology

The mortality rate in neonatal and pediatric severe sepsis has improved from 97% in 1963 to 9% in 1999, to 4% in 2003.⁷^–¹³ Previously healthy children have better outcomes than children with chronic illness. The randomized controlled trial of bactericidal permeability-increasing protein¹⁴ for children with purpura fulminans/presumed meningococcal septic shock showed 10% mortality rates in the placebo groups. The reported outcomes in children with septic shock when using therapeutic approaches similar to those recommended in the 2002 American College of Critical Care Medicine Clinical Practice Parameters for Hemodynamic Support of Pediatric and Neonatal Patients in Septic Shock⁶ show a decreasing tendency. In children with meningococcal septic shock in the United Kingdom, a 5% mortality rate was reported,¹⁵ and in the Netherlands a decreasing mortality was shown in the same patient group.¹⁶ A single-center study in the United States reported a 10% mortality rate.¹⁷ The investigators observed 0% mortality in previously healthy children but a 15% mortality rate in children with chronic illness (for the most part cancer patients). All of these children died with multiple organ failure. Ngo and colleagues¹⁸ observed a 0% mortality rate in a randomized Dengue shock fluid resuscitation trial. The US KIDS database showed a 4.2% severe sepsis mortality overall, with 2% in the previously healthy and 8% in the chronically ill child.¹³

Although outcomes are improving, the burden of newborn and pediatric sepsis is increasing in the United States. More children die with severe sepsis than die with cancer, with an estimated yearly healthcare cost of $4 billion in the United States for patients with this condition.¹² Half are newborns, with most of these having low birth weight.⁹ Half of children with severe sepsis have underlying chronic illness. Neurologic and cardiovascular chronic illness is most common in infants with severe sepsis and cancer, whereas immune deficiency is most common in children with severe sepsis. Medical advances have affected etiology and epidemiology. In 1990, Jacobs and coworkers¹⁹ reported that the most common causes of septic shock in children were, in descending order, Haemophilus influenzae b, Neisseria meningitidis, and Streptococcus pneumoniae. The 1995 and 1999 U.S. estimates suggest a change. H. influenzae type b is all but nonexistent, N. meningitidis is prevalent in only a few regions of the United States, and group B Streptococcus is decreasing. The more recent use of S. pneumoniae vaccine is reducing the incidence of this infection. The Canadian government has implemented nationwide immunization in children younger than age 2 years for N. meningitidis serotype C.²⁰ The most prevalent causes of severe sepsis and septic shock in the United States now seem to be staphylococcal and fungal infections.¹² Methicillin-resistant Staphylococcus aureus (MRSA) is an emerging disease. Influenza vaccines are now universal for both endemic and pandemic forms (H1N1).

Pathophysiology and Developmental Effects

Molecular Pathogenesis

Controlled Inflammation with Eradication of Infection

Endotoxin, mannose, and other glycoprotein moieties on the cell walls of yeast and fungi, superantigens, toxins associated with some gram-positive bacteria, mycobacteria, and viruses, also called pathogen-associated molecular patterns, activate the innate immune system after recognition by pathogen recognition receptors. The innate immune system comprises polymorphonuclear neutrophils, monocytes, and macrophages, in part through Toll-like receptors, CD14 receptors (endotoxin), and other costimulatory molecules. These innate immune cells internalize microorganisms and kill them. Monocytes and macrophages present processed antigens from these killed microorganisms to circulating T lymphocytes and coordinate the adaptive immune response. This second wave of immune response includes B-cell activation and antibody production and generation of cytotoxic T cells and natural killer cells (particularly in viral and fungal infection). Opsonization with antibodies allows more efficient recognition, killing, and clearing of microorganisms by resident macrophages in the reticuloendothelial system.^21,²²

The activated inflammatory cells also initiate a series of biochemical cascades that result in phospholipase A₂, platelet-activating factor, cyclooxygenase, complement, and cytokine release that orchestrate an efficient and controlled inflammatory/immune response. The cytokines, tumor necrosis factor (TNF) and interleukin (IL)-1β, synergistically interact to promote positive feedback cascades that result in fever and vasodilation. These cytokines stimulate the production of many important effector molecules, including proinflammatory cytokines (e.g., IL-6, IL-8, and interferon-[IFN]-γ), which promote immune cell-mediated killing and antiinflammatory cytokines (e.g., soluble TNF receptor, IL-1 receptor antagonist protein, IL-4, and IL-10), which turn off the immune response when the infection has been cleared. These cytokines also stimulate nitric oxide (NO) production, which leads to vasodilation. NO also combines with superoxide radicals to form peroxynitrite radicals (ONOO^–), which participate in intracellular killing of microorganisms. Cytokines also increase expression of endothelial-derived adhesion molecules, including E-selectin, which facilitates white blood cell rolling, and intercellular adhesion molecule and vascular adhesion molecule, which facilitate white blood cell adhesion and diapedesis. This activity guides activated inflammatory cells to the site of infection. The cytokines also induce a change in the endothelium to a prothrombotic and antifibrinolytic state. Expression of thrombomodulin is possibly decreased, and expression of the prothrombotic molecule tissue factor and the antifibrinolytic molecule plasminogen activator inhibitor-1 (PAI-1) is increased. The ensuing thrombus “walls off” the infection and allows vascular remodeling until antiinflammatory cytokines turn off the proinflammatory cytokine response and restore the antithrombotic profibrinolytic milieu after infection is cleared.

Uncontrolled Inflammation and Persistent Infection Lead to Septic Shock and Multiple Organ Failure

If the controlled activated immune cell response is ineffective in killing the infectious agent and clearing antigen, inflammation is uncontrolled, and systemic organ injury ensues. Increased TNF and NO production in cardiac cells and circulating myocardial depressant substances can lead to cardiac dysfunction and cardiovascular collapse. Peroxynitrite can cause DNA damage, and subsequent polyadenosyl ribose synthase (PARS) activation depletes cells of oxidized nicotinamide adenine dinucleotide and adenosine triphosphate (ATP), leading to secondary energy failure. Thrombosis and antifibrinolysis becomes systemic. Antithrombotic molecules, including protein C and antithrombin III, are consumed, and ongoing systemic release of tissue factor and PAI-1 results in unremitting thrombosis. At some point, consumption of procoagulant factors leads to a precarious state in which thrombosis is accompanied by bleeding because there are insufficient clotting factors. The antiinflammatory response also becomes deleterious. IL-10 induces a T_H2 response and reduces the ability of monocytes/macrophages to kill infection. Overactivated immune cells also release Fas and Fas ligand. Circulating Fas prevents activated immune cell apoptosis and ensures ongoing inflammation, and Fas ligand can induce liver injury. In patients with natural killer (NK) cell dysfunction, activated immune cell death is further hampered. Ineffective and unresolving inflammation leads to systemic organ failure.

Clinical Pathologic Correlates

On the basis of in vivo biochemical analyses and autopsy histology, several forms of multiple organ failure could be characterized.²³^–²⁶ Thrombocytopenia-associated multiple organ failure (platelet count <100,000/µL or a 50% decrease in platelet count from baseline) was attributable to purpura fulminans and disseminated intravascular coagulation (DIC) with increased tissue factor activity in vivo and fibrin thrombi at autopsy in only 20% of patients. Of these patients, 80% showed thrombotic thrombocytopenic purpura pathophysiology with increased thrombogenic ultra-large von Willebrand factor multimers, absent von Willebrand factor cleaving protease (ADAMTS 13), increased PAI-1 activity in vivo, and platelet/fibrin thrombi at autopsy.

Sequential or liver dysfunction–associated multiple organ failure (shock/acute respiratory distress syndrome followed sequentially by liver and renal failure) was associated with viral sepsis and lymphoproliferative disease. These patients were found to have unremitting Epstein-Barr virus infection, with lymphocyte Fas ligand–mediated destruction of liver and high circulating Fas and Fas ligand levels. This syndrome is also found in patients with defects in NK cell activity. Absent NK cell activity is found in primary hemophagocytic lymphohistiocytosis (HLH), and decreased NK cell activity in secondary HLH. NK cells are responsible for killing viruses and stopping lymphoproliferation.

Unresolving multiple organ failure with prolonged monocyte deactivation (monocyte HLA-DR expression <30% or ex vivo TNF response to lipopolysaccharide <200 pg/mL for >5 days) was associated with secondary bacterial, fungal, or herpesvirus family infection. These patients had elevated IL-10 and IL-6 levels. Patients who died had infection at autopsy.

Lymphoid depletion syndrome (lymphocyte depletion of lymph nodes and spleen) was found at autopsy. All of these children had fungal, bacterial, or herpesvirus family infection at the time of death. Risk factors (odds ratio >10) for this process included lymphocytopenia (<1000/mm³) or hypoprolactinemia or both for more than 7 days. Phagocytosis of these apoptotic bodies by monocytes/macrophages leads to immunoparalysis.

These clinical pathologic correlates support the following hypotheses: (1) uncontrolled inflammation contributes to organ failure after septic shock; (2) uncontrolled inflammation contributes to systemic thrombosis; (3) uncontrolled inflammation leads to adrenal dysfunction not only through thrombosis but also potentially through NO-mediated inhibition of cytochrome P450 activity; and (4) uncontrolled inflammation is commonly associated with uneradicated infection. It is likely that genetic and environmental factors can increase an individual patient’s risk for systemic thrombosis and uneradicated infection.

Coagulation System

As is generally accepted and explained in many reviews, coagulation and fibrinolysis are an integrative part of the immune system.²⁷ There are important physiologic differences in the hemostatic system in children compared with adults. The decreased levels of several crucial coagulants and increased levels of α2-macroglobulin may contribute in part to the lower risk of thrombotic events in childhood during physiologic conditions.^28,²⁹ In pathologic conditions, these physiologic differences might lead to an earlier exhaustion of coagulation factors and DIC in infants and young children.³⁰ ADAMTS 13 is also decreased in infancy, therefore there may be an increased susceptibility to systemic fibrin and platelet thrombosis The coagulation system is a marker of organ dysfunction in sepsis. It is associated with subsequent endothelium activation and systemic clotting and finally antifibrinolysis.

Cardiovascular System

Ceneviva and associates ³¹ found that in contrast to adults, who predominantly have high-cardiac-output/low-vascular-resistance shock, children with fluid-refractory/inotropic-resistant shock have varied hemodynamic states, including low cardiac output/high systemic vascular resistance (60%), low cardiac output/low vascular resistance (20%), and high cardiac output/low vascular resistance (20%), which can change with time and depend on age. In contrast to adults, death from shock is most commonly associated with progressive cardiac failure, not vascular failure. Infants and children frequently are insensitive to dopamine or dobutamine and respond to epinephrine (cold shock) or norepinephrine (warm shock).³¹^–³³ Newborns are different as well. Adults can double their heart rate to improve cardiac output, but newborns cannot. Newborns, although tachycardic, depend on increased vascular tone to maintain blood pressure. Persistent pulmonary hypertension and right ventricular failure also complicate newborn septic shock.^34,³⁵

Predisposing Factors and Prevention Strategies

Environmental and genetic factors associated with reduced immune function predispose children to the development of sepsis and septic shock. These factors include age (prematurity, neonate, and age < 1 year), cancer and immunosuppressive chemotherapeutic agents, transplantation and immunosuppressive agents, primary immunodeficiency disorders (e.g., hypocomplementemia, hypogammaglobulinemia, chronic granulomatous disease), acquired immunodeficiency disorders (neutropenia, lymphocytopenia, monocyte deactivation), and malnutrition. Prolonged use of invasive catheters, muscle relaxants, and broad-spectrum antibiotics also predispose to infection.

Among the community-acquired causes of sepsis, N. meningitidis has a diverse clinical picture, ranging from a self-limiting bacteremia to meningitis to a severe rapidly fatal sepsis. After invasion of the bloodstream by the bacteria, three main cascade pathways are activated: the complement system, the inflammatory response, and the coagulation and fibrinolysis pathway. These pathways do not act independently but are able to interact with each other. Genetic polymorphisms among components of these pathways have been shown to be involved in the susceptibility, severity, and outcome of meningococcal disease. Knowledge of genetic variations associated with susceptibility to and severity of meningococcal infection has been reviewed.³⁶

Complement deficiencies and defects in sensing or opsonophagocytic pathways, such as the rare Toll-like receptor 4 single nucleotide polymorphisms and combinations of inefficient variants of Fcγ-receptors, seem to have the most important role in genetically established susceptibility. The most recent and largest study on susceptibility is a genome-wide analysis of DNA from 1600 children with meningococcal sepsis. This study showed the significant influence of genetic variants in the complement factor H in the susceptibility.³⁷ Effect on severity has repeatedly been reported for FcγRIIa and PAI-1 polymorphisms. Angiotensin-converting enzyme is associated with a proinflammatory response. The absence of a 284-base pair marker in the angiotensin-converting enzyme gene (D allele) is associated with higher circulating angiotensin-converting enzyme activity compared with the presence of this marker (I allele). The DD genotype is associated with increased disease severity, and although not significant, a twofold increase in mortality rate has been reported. Outcome effects have been confirmed for single nucleotide polymorphisms in properdin deficiencies, PAI-1 and combination of the −511C/T single nucleotide polymorphisms in IL-1β, and +2018C/T single nucleotide polymorphisms in IL RN. Conflicting results are reported for the effect of the −308G/A promoter polymorphism in TNF. These differences may reflect discrepancies in group definitions among studies or the influence of additional single nucleotide polymorphisms in the TNF promoter, which can form haplotypes representing different cytokine production capacity. For several single-nucleotide polymorphisms, the potential effect on susceptibility, severity, or outcome has not yet been confirmed in an independent study.

The hallmark of pediatric medicine is prevention. Public health programs that reduce prematurity could be expected to have the greatest impact on the incidence of sepsis. The use of group B streptococcal prophylaxis in at-risk mothers has reduced the incidence of septic shock in premature and term infants. Immunization programs for diphtheria, pertussis, tetanus, measles, mumps, rubella, H. influenzae type b, S. pneumoniae, N. meningitidis (type C for infants and type C, A, and Y for college students), and influenza all effectively reduce the incidence of sepsis in newborns and children. The primary immunodeficiency initiative is an important physician education program. Children with frequent pneumonia, sinus infections, or skin infections can benefit from early immunodeficiency workups, including quantitative immunoglobulins, complement levels, nitroblue toluene testing of polymorphonuclear neutrophil function, and antibody titer response to immunization. Early identification of these children can lead to use of therapies that reduce the incidence of sepsis.

< div class='tao-gold-member'>

Only gold members can continue reading. Log In or Register to continue