Immunologic Dissonance: A Continuing Evolution in Our Understanding of the Systemic Inflammatory Response Syndrome (SIRS) and the Multiple Organ Dysfunction Syndrome (MODS)

Fifty years ago, multiple organ failure was unheard ofwe could not keep patients alive long enough for the sequential dysfunction of distal organs to develop. Patients with hemorrhagic shock died of blood loss or kidney failure. Severe infections were often fatal because antibiotic therapy was in its infancy. It was only about 30 years ago that the acute respiratory distress syndrome was first described by Ashbaugh and colleagues [1, 2]. We had learned enough to be able to manage hemorrhagic shock without prompting renal shutdown [3], but too many patients still died of severe pulmonary complications. I was in Vietnam at that time, and watching so many young men die of what we then called shock lung or Da Nang lung was one of the formative experiences of my early career. In the 1960s, numerous other papers drew attention to various nonpulmonary causes of severe respiratory dysfunction, including abdominal surgery [4], peritonitis [5], and distant infections [6]. It became clear that pulmonary compromise could result from processes far removed from the lungs. Slightly more than 20 years ago, the first descriptions of multiple organ failure appeared. In 1973, Tilney and associates [7] discussed three patients who had died of distal organ failure that followed ruptured aortic aneurysms. Baue [3] described multiple, progressive, or sequential systems organ failure. These reports were followed by other classic papers, many of which attributed multiple organ failure to uncontrolled infection, especially gram-negative sepsis [8-11]. Ten years ago, it seemed as if we were finally beginning to understand the problem presented by sepsis and multiple organ failure. We started to realize that shock or infection alone did not cause distal organ dysfunction but rather that shock, infection, or other severe insults could set in motion an underlying reaction that would lead to widespread endothelial damage, edema resulting from increased vascular permeability, and impaired availability of oxygen [12]. Great hope arose when Ziegler and coworkers [13] showed that antiserum to endotoxin improved survival in patients with gram-negative sepsis. Five years ago, we appeared to be entering a brave new world. Several of the mediators thought to cause the reaction underlying sepsis and multiple organ failure had been identified [14]. Clinical trials [15-29] were investigating various agents that, it was hoped, would downregulate these mediators. The ability to overcome sepsis and multiple organ failure seemed to be within our grasp. The results of these trials were uniformly disappointing. Numerous studies were done with the following agents: monoclonal antibodies to endotoxin, methylprednisolone, monoclonal antibodies to tumor necrosis factor, dimeric tumor necrosis factor receptors, recombinant interleukin-1 receptor antagonist, and platelet-activating factor antagonists [30]. Agent after agent was found to have no effector worse, was found to increase mortality rates. Today, multiple organ failure affects as many as 40% of critically ill patients [31] and remains the leading cause of death in intensive care units [31, 32]. It is not surprising, therefore, that many physicians believe multiple organ failure to be an intractable problem. Although I understand their pessimism, I do not share it. I believe that we have made great strides in elucidating the pathogenesis of sepsis and multiple organ failure. It is true that the underlying inflammatory reaction is vastly more complicated than we had thought, and it is unlikely that a magic bullet will be found soon. However, by reexamining what we have learned and rethinking our assumptions, we can better understand how organ dysfunction develops and how we may someday be able to prevent it. The Systemic Inflammatory Response Syndrome and the Multiple Organ Dysfunction Syndrome Five years ago, it became clear that the terms that we had been using to describe sepsis and organ dysfunction were no longer adequate. New terms were therefore developed by a consensus conference [33]. The term systemic inflammatory response syndrome (SIRS) is now used to describe the clinical syndrome previously called sepsis; the word sepsis is used only when the cause of SIRS is documented infection. Multiple organ dysfunction syndrome (MODS) replaces the term multiple organ failure because it stresses the continuum of organ dysfunction, not just its result. Problems with Previous Theories Without question, a massive inflammatory reaction underlies both SIRS and MODS, as I will show. However, this reaction has both pro- and anti-inflammatory components. The anti-inflammatory reaction is often as great as, and sometimes greater than, the proinflammatory response. To a large extent, the theories put forth to explain the development of SIRS (my own included) have not incorporated this anti-inflammatory reaction. Perhaps this was inevitable; many of the anti-inflammatory mediators have been discovered only recently. However, these theories have also often overstated the dangers presented by proinflammatory mediators. Although excessive levels of these mediators can cause problems, lower levels are beneficialinflammation is required to combat pathogenic organisms and promote healing. The overemphasis placed on proinflammatory mediators may have resulted in part from the surrogate models we have used to study SIRS and MODS: studies done in animals; experiments in which endotoxin, tumor necrosis factor, or another mediator was injected into human volunteers; and analyses of serum levels of proinflammatory mediators in patients with sepsis, burns, or other severe injuries. These studies may not accurately reflect what happens in critically ill patients. For example, there is marked interspecies variation in cytokine release, which makes it difficult to extrapolate findings in animals to humans [34]. Experiments are done on healthy animals, and observation periods are generally short [35]. Human studies are done in healthy persons, the amount of stimulus injected is sublethal, and follow-up periods are brief [36]. In contrast, SIRS and MODS develop over time in severely ill or injured patients, who often have numerous preexisting disorders [14]. Analyses of serum levels are even more troublesome to interpret. Most immunoassays detect only circulating mediators, not mediators bound to cells or receptors [37-39]; thus, they may underestimate the effective amount of mediator acting at a cellular level. Bioassays, which measure the functional activity of cytokines, often lack specificity and may over-report amounts [40]. Furthermore, serum analyses are usually done sporadically (often once daily or less), yet mediator release is phasic. Most analyses have assumed that the presence of proinflammatory mediators is a direct result of the immediate insult (for example, a burn) and not a consequence of a preexisting condition. However, proinflammatory mediators are present in patients with widely varying disorders [41]. Range of Clinical Responses Patients who develop sepsis or have extensive burns, massive traumatic injuries, or other severe insults can be grouped into several categories. 1. At one extreme are patients who show little evidence of a systemic reaction. Although recovery may be protracted because of the severity of the underlying illness, organ dysfunction rarely develops. 2. Next are patients who develop a mild form of SIRS and show some evidence of organ dysfunction early in the clinical course of disease. Dysfunction is usually limited to one or two organs and resolves rapidlyoften within a day or two. 3. In other patients, a massive systemic inflammatory reaction develops rapidly after the initial insult. These patients often die of profound shock within a few days. 4. Still others have a less severe initial course of disease but deteriorate markedly several days or more after the original insult. Failure of one or more organs is common, and many of these patients die. Most of the research done on SIRS and MODS has focused on the last two groups of patients. Clinical trials usually exclude patients who have mild symptoms of organ dysfunction or symptoms that last for less than 48 hours. Thus, these trials have excluded all patients in the first category and many of those in the second. However necessary this exclusion may have been to the design and conduct of those trials, it has skewed our perception of SIRS and MODS. Some have suggested that evidence of organ dysfunction during the first 48 hours after insult reflects not the onset of MODS but rather the inciting event or incomplete resuscitation [42]. I believe that this distinction is artificial. Every severe insult to the body produces a response with pro- and anti-inflammatory components. But we must ask the following questions: Is the inflammatory response of an appropriate magnitude, and is it appropriately downregulated (as in the first two groups of patients)? Or does something go wrong (as in the second two groups)? By considering why SIRS and MODS do not develop in some patients, we can better understand why these disorders do arise in others. A New Theory The body is designed to compensate for any assault. Its defenses include macrophages and their products, such as tumor necrosis factor; interleukin-1, interleukin-6, and interleukin-8; neutrophils and the products of their degranulation; platelets and the coagulation factors formed on their surfaces; derivatives of arachidonic acid; T and B lymphocytes and their products; and many other substances [14]. How these agents interact is far from understood, but it is clear that they create a complex, often overlapping, network of interactions. Presented below is my explanation for how these agents work together to overcome a severe assault and, paradoxically, how they can cause SIRS and MODS (Figure 1). I admit that this explanation contains lacunae; too many facts remain unknown. However, I believe