Underlying Disorders and Their Impact on the Host Response to Infection
- 1Intensive Care and Emergency Department, Cochin University Hospital, René Descartes University, Paris, France
- 2Eli Lilly, Lilly Research Centre, Erl Wood Manor, United Kingdom
- 3Lilly Research Laboratories, Eli Lilly, Indianapolis, Indiana
- Reprints or correspondence: Dr. Jean-François Dhainaut, l'Université René Descartes, 12 rue de l'Ecole de Médecine, 75006 Paris, France(dhainaut{at}univ-paris5.fr).
Abstract
Underlying disorders, especially those that chronically impair immune host response (e.g., cancers and hematologic malignancies) but also those that acutely impair this response (e.g., major surgery and multiple trauma), increase the incidence of infection and alter the outcome of patients with sepsis. As a part of innate immunity, inflammatory and coagulation responses are lower in patients with underlying disorders than in patients without such disorders, whereas the need for vasopressors and mechanical ventilation is more frequent. Although these patients are older, age-related defects do not appear to be responsible for this lower response, because innate immunity is usually up-regulated in the elderly. Innate immunity seems to be negligibly affected by the direct consequences of underlying disorders, but underlying disorder—related chronic organ insufficiency certainly participates in the observed organ dysfunction, overestimating the infectious insult by itself. Although innate immunity seems not to be actually blunted in patients with underlying disorders, the underlying disorder itself contributes to the severity of the physiological response to sepsis, thereby resulting in a worse outcome.
Sepsis is defined as the systemic host response to infection. To the clinician, the word conjures up a life-threatening clinical syndrome that arises through the innate response to infection [1]. However, sepsis is not a single disease but is an intricate and heterogeneous process expressed through the interaction of a complex network of biochemical mediators and amplification cascades. Its clinical expression is variable, and its severity is influenced by the nature of the infection, the genetic background of the patient [2], the time to clinical intervention, and the supportive care provided by the physician. This inherently complex process, reflecting the dynamic interaction of an acute, life-threatening infection with the adaptive protective mechanisms of the host and its environment, is frequently modified, often in an unpredictable manner, by the effects of advancing age, sex, chronic alcohol abuse, and/or acute and chronic underlying disorders [1].
There has been a substantial increase in the incidence of sepsis during the past 2 decades, with an increasing number of deaths due to sepsis, despite a decline in overall in-hospital mortality [3]. Possible reasons for the increase in the incidence of sepsis include a more frequent aggressive management of severe underlying disorders with increasing use of invasive procedures and immunosuppressive drugs, chemotherapy, and transplantation; the emergence of HIV infection; and increasing microbial resistance [3]. Severe underlying disorders are more and more frequently reported in patients with sepsis, and Angus et al. [4] recently found that 55% of 192,980 patients admitted because of sepsis had underlying disorders. Here, we discuss the definition of underlying disorders, the prognostic impact of underlying disorders on outcome of sepsis, and by which mechanisms underlying disorders interfere with the incidence and severity of sepsis.
Definition Of Acute and Chronic Underlying Disorders
Underlying disorders have been first described as a comorbidity or chronic underlying disease in addition to the disease or condition designated as the principal diagnosis. This underlying disorder should be considered an active problem (excluding, among other things, a history of pneumonia and of cholecystectomy) that is expected to impair a patient' long-term survival [5–7]. More recently, the impact of an acute underlying condition, including major surgery, multiple trauma, or major cardiac event, has been emphasized by data showing an important role in determining the risk of developing septicemia and premature mortality [8]. As a result, we define an underlying disorder as an acute or chronic comorbidity or condition not related to sepsis that alters short- and/or long-term survival of infectious diseases.
Impact Of Chronic Underlying Disorders On Prognosis Of Sepsis
The incidence of infectious episodes is clearly increased in many chronic underlying disorders leading to immunosuppression, including hematologic malignancies [9], neoplasm [10–12], and diabetes mellitus—related organ damage [8, 13]. A high incidence of infection has also accompanied the use of immunosuppressive therapy, and the risk of developing infection increased with increasing the severity and duration of granulocytopenia [9].
The severity of infectious episodes is also higher in patients with chronic underlying disorders. In 1962, McCabe and Jackson [14] used an arbitrary classification to determine that 173 fatalities resulting from gram-negative bacteremia, documented during an 8-year period, markedly depended on the severity of chronic underlying disorders of the patients. Indeed, the hospital mortality of patients with sepsis with acute leukemia (or blastic relapse of chronic leukemia) classified as having “rapidly fatal” disease was 91%, whereas that of patients having “ultimately fatal” disease (other hematologic malignancies, metastatic cancers, or end-stage liver or renal diseases) was 66%, and that of patients having “nonfatal” disease was 11%. Six years later, Freid and Vosti [15], using the same classification, confirmed the previous findings in 270 patients with gram-negative bacteremia: the fatality rates among patients with rapidly, ultimately, and nonfatal chronic underlying disorders were 86%, 46%, and 16%, respectively. Soriano et al. [16] also found that rapidly or ultimately fatal chronic underlying disorder was an independent predictor for mortality among 225 episodes of bacteremia due to infection with methicillin-resistant Staphylococcus aureus. Rapidly and ultimately fatal underlying disorder was also reported as an independent predictor of mortality in patients in intensive care units with community-acquired pneumonia [17] and nosocomial pneumonia [18]. Consequently, such severe chronic underlying disorders have been integrated into scoring systems deriving from high-quality databases. For instance, the mathematical model that allowed the development of SAPS II [19] specifically highlighted the major impact of hematologic malignancies, metastatic neoplasms, and AIDS, 3 underlying disorders clearly that are linked with impairment of the immune system and also are highly associated with infectious diseases.
More recently, Brun-Buisson et al. [20] examined the risk factors of severe sepsis in a cohort study from a 2-month prospective survey of 11,828 consecutive admissions to 170 adult intensive care units in France. The authors found that a rapidly or ultimately fatal chronic underlying disorder (P ∼ .001) and a preexisting liver or cardiovascular insufficiency (P = .002) were associated with secondary deaths (occurring >3 days after admission). On the basis of a 1995 chart review of discharge records from 847 nonfederal hospitals in 7 US states, Angus et al. [4] reported that the hospital mortality rate for 192,980 patients with severe sepsis with chronic underlying disorders was higher than that for patients without chronic underlying disorders. The overall hospital mortality was 28.6% and was higher in patients with sepsis with metastatic neoplasm (43%), nonmetastatic neoplasm (37%), chronic liver or renal disease (37%), HIV disease (34%), and chronic obstructive pulmonary disease (32%). Alberti et al. [21] examined risk factors for hospital mortality for 3608 patients with severe sepsis in intensive care units and found that, among chronic underlying disorders, liver cirrhosis, immunosuppression, and chronic heart failure were associated with a poor prognosis. Chronic severe renal failure was also been reported as a prognostic factor in patients with sepsis [21, 22]. Finally, chronic liver failure was frequently reported as an underlying disorder [23] that markedly influences the outcome for patients with sepsis [20, 21] and has been integrated in the APACHE scoring system [24]. This scoring system, which was developed by Knaus et al. [24], is based on a physiology score reflecting the degree of acute illness, the age of the patient, and the patient' preadmission health status. Recently, the last part of APACHE II was used to assess chronic organ dysfunctions [8] in patients with bacteremia and was found to be an independent predictor of mortality in intensive care units, as well as from rapidly and ultimately fatal underlying disorders.
Most of the studies described above used the classification developed by McCabe and Jackson [14], which is arbitrary and depends on the investigator' judgment of the prognosis of the underlying disorder. This classification system also does not take into account the real seriousness of each underlying disorder and the effects of the combination of underlying disorders. An alternative approach that uses a scheme of taxonomy for classifying underlying disorders and assessing the prognostic value of the classifications has been developed. The most interesting method of classifying prognostic underlying disorders in longitudinal studies has been developed by Charlson et al. [5] in a general population admitted with medical diagnoses. Underlying disorders were coded as either “absent” (score, 0) or “present” (score, 1), and a weighted index was developed that takes into account the number and the seriousness of underlying disorders ( table 1), as well as severity of illness at the time of admission. The reasons for admission were grouped according to whether the patient had a high or low risk of mortality during hospitalization, by use of an illness severity scoring system essentially based on clinical judgment. Among such patients, the weighted index of underlying disorders, illness severity, and reasons for admission (low or high risk) were all significant predictors of mortality at 1 year after admission (P ∼ .0001). The 1-year mortality rate is shown in table 2. There is a stepwise increase in the observed mortality with a higher underlying disorder index within each of the severity groups. Recently, Lesens et al. [25] demonstrated the effectiveness of the Charlson weighted index of underlying disorders for controlling underlying disorders in patients with bacteremia due to S. aureus and confirmed that a high weighted index of underlying disorders was an independently predictive factor of mortality among patients with S. aureus bacteremia.
Figure 1
Time course (from day [d] 0 to d7) of protein C activity (PC Act) and concentrations of D-dimer (expressed in ¼g/mL) and IL-6 (expressed in pg/mL) in patients with severe sepsis who received placebo in the PROWESS trial, comparing those with and without underlying disorders (UD). Changes in percentage of IL-6 concentrations are also illustrated in inset in IL-6 panel. *P ∼ .05.
Table 1
Weighted index of underlying disorders, as described by Charlson et al. [5].
Table 2
One-year mortality according to severity at admission and scores from the weighted index of underlying disorders.
Impact Of Acute Underlying Disorders On Prognosis Of Sepsis
However, these indexes described above may not pick up some severe acute underlying disorders that may be closely associated with death, such as trauma, burns, transplant rejection, ischemia-reperfusion injury, pancreatitis, or major surgical intervention, as well as a concomitant clinical event (myocardial infarction or pulmonary embolus). For instance, patients with trauma, especially those with neurological impairment, have high risk of developing pneumonia in intensive care units [26]. Pittet et al. [8] determined admission characteristics associated with the outcome of bacteremia in 173 critically ill surgical patients; they also assessed the prognostic value of acute and chronic underlying disorders commonly reported in the patient' past medical history and nonobjectively measured with the usual severity scoring systems in the intensive care unit, including the APACHE II score. They found that the 2 independent predictors of mortality from bacteremia were APACHE II score and underlying disorder score, which included major surgery within 2 months before admission, splenectomy before admission, previous antibiotic therapy within 2 months before admission and lasting for ⩾2 weeks, previous cardiogenic shock, or cardiopulmonary resuscitation before admission.
In addition, the Clinical Evaluation Committee of the H1-A1 trial [27] and the PROWESS trial [28] tried to address this issue by using a large number of both acute and chronic conditions in the definition of underlying disorder. The definition of underlying disorder used by Sprung et al. was a significant associated noninfectious life-threatening disorder: “the presence of an acute or chronic disease (excluding sepsis) that had a significant chance of causing death within the next 12 months, or a severe underlying disorder that might increase the chances of a decision to forego life-sustaining treatments” [27, p. 384].
This classification has also its limitations, because the classification of patients is arbitrary, is retrospective, and depends on 2 reviewers' judgment of the prognosis of the underlying disorder from the data included in the case report form and the investigator' summary. However, with the study of Pittet et al. [8], it remains the only approach that includes acute and chronic underlying disorders. The 2 clinical evaluation committees reported a higher mortality among patients with sepsis with underlying disorders than among those without underlying disorders. H1-A1 and PROWESS were randomized, double-blind, placebo-controlled trials evaluating the efficacy of human monoclonal antibody against endotoxin and of recombinant human activated protein C, respectively, in patients with severe sepsis.
To better illustrate this approach, the characteristics of the 840 placebo patients of the PROWESS trial, who had previously undergone assessment by the Clinical Evaluation Committee as to the presence or absence of underlying disorders, were retrospectively analyzed. Table 3 summarizes the characteristics of the placebo population with underlying disorders (n = 165), compared with those without underlying disorders (n = 675). Patients with underlying disorders were older and had similar acute physiological scores (as the first part of the APACHE II scoring system) but needed more frequently mechanical ventilation and vasoactive support at baseline than did patients without underlying disorders (P ∼ .001), although organ dysfunctions assessed by the Sequential Organ Failure Assessment score was not statistically different. The mortality rate was markedly higher in patients with underlying disorders than in those without (55.8% vs. 24.7%; P ∼ .0001) [28]. It is of note that the mortality rates of the early period (until day 5), middle period (days 5–14), and late period (days 14–28) were all higher for patients with underlying disorders (P ∼ .0001).
Table 3
Severity of illness measures of patients with versus without underlying disorders.
Toward The Predisposition-Infectionresponse- Organ Dysfunction (Piro) Concept
The outcome of sepsis is related to a network of determinants described above, including the microorganism responsible for the infection, the systemic inflammatory response to the infective agent, the organ dysfunction related to the systemic stress, and the host' physiological status, with an evident overlap between these different entities. Genetic factors also play a role in mortality due to sepsis, as do other conditions, including premorbid health status and reversibility of concomitant diseases, the role of which is more difficult to establish. On the basis of data described above, there appears to be an important burden of age and comorbidities in the prognosis of patients with sepsis.
The need for a new classification of response to infection was driven by the lack of usefulness of the previous guidelines [29] to treat patients. In 1946, Pierre Denoix proposed classifying tumors on the basis of their primary histology and site of development (T), regional lymph nodes (N), and distant metastases (M) [30]. By analogy with this TNM stratification, Marshall et al. [1] proposed that sepsis should be considered a multistage disorder and should be classified by taking into account the main determinants of outcome. The PIRO classification emerged from the various meetings about sepsis since 2001 [1, 31, 32]. PIRO is an acronym in which P stands for predisposition, I for infection, R for response, and O for organ dysfunction. The P is for predisposing factors that would make someone more likely to develop infection and/or organ failure. The I categorizes the type of infection and its specific burden. R highlights the heterogeneity of the necessary but sometime deleterious systemic inflammatory response, and O incorporates the organ dysfunctions and sequelae related to the interaction between host and infective agent [32].
To better illustrate the concept, the “I” was recently subject to an extensive review of 55,854 patients with sepsis from 510 published articles over a period of 30 years [33]. The authors evaluated the burden of identifiable infection, source (localized or disseminated), microorganisms responsible for the sepsis, and the use of appropriate or inappropriate initial antimicrobial therapy. “I” was indexed as 1 if the infection was not bacteriologically proven, 2 if it was proven, and 3 if bacteremia was present. Mortality rates associated with each bacterial species were classified into 4 groups: level 1, ⩽5%; level 2, 6%–15%; level 3, 16%–30%; and level 4, >30%. This assessment allowed a stratification of death related to microorganism for each primary focus of infection. The relevance of the data for each infection site was also categorized as follows: A, >100 patients and >5 studies; B, >100 patients or >5 studies; C, >25 patients in ⩾1 study; D, small studies of ∼25 patients; and E, case reports (insufficient evidence). For example, a subject who developed pneumococcal pneumonia with bacteriologically confirmed bacteremia should be scored I2-2A in the PIRO model. According to the results of the study, the overall mortality of the population corresponding to this patient' condition should be 6%–15%.
The “P” is a more general item and includes genetics, underlying disorders, and environmental and social factors (e.g., consumption of alcohol). In addition, the skills of the physician or team in charge of the patient could be considered [31]. It is obvious that a different weight should be attributed to each predisposition factor. Future studies based on large high-quality databases will be necessary to address the “P.” The weighted index of Charlson et al. [5] could be an interesting preliminary approach that should be revisited, because the prognosis of many cancers and AIDS has improved with new therapeutic regimens since it was proposed in 1987. The index should also be validated in a large population of patients with sepsis.
Underlying Disorders As Confounding Factors In Clinical Trials Of Sepsis
In human beings, infectious diseases are usually associated with various confounding conditions, such as aging and underlying disorders, that can modulate the systemic response and outcome. As a result, observations in murine and other mammal models of sepsis may not be directly applicable to humans. First, most animal models use young healthy adults. However, sepsis in humans is more common in neonates and the elderly [3]. In the majority of animal models used to study sepsis, there are no underlying disorders such as diabetes, neoplasia, or cardiovascular disorders. In patients, underlying disorders are frequent and account for impairment of the immune system. Because the systemic inflammatory response differs with aging and significant underlying medical history, it is clear that the use of healthy animals in sepsis models may be a poor model of the usual human population experiencing severe infection (reviewed elsewhere [34]).
New therapeutic interventions, potentially able to modulate the host response to infection, are usually first tested in different animal models in which underlying disorders are not present. Underlying disorder is a major confounding factor in patients with severe sepsis and represents the reality of the intensive care unit setting. Underlying disorder not only makes the reproducibility of experimental results difficult in clinical trials, because of our inability to adjust risk appropriately to the heterogeneity of acute or chronic conditions, but also modifies the cause of deaths and alters practices of life-sustaining therapy.
For example, in the placebo group of the PROWESS trial, the percentage of nonseptic deaths was markedly higher in patients with underlying disorders than in patients without (27% vs. 13%; P ∼ .0001). A higher percentage of patients was recorded as foregoing life-sustaining therapy in the group with underlying disorders than in the group without underlying disorders in this trial (31% vs. 11%; P ∼ .0001) [28]. Among those patients who had forgone therapy in the HA-1A trial, a high percentage of patients with underlying disorders was observed (52% vs. 11% in the global population) [27]. In addition, inadequate antimicrobial therapy was more frequent in patients with underlying disorders in the PROWESS trial (13.3% vs. 7.7%; P ∼ .02). This increases the “background noise” against which a trial must operate to evaluate the efficacy of a new molecule in sepsis.
Mechanisms By Which Underlying Disorders Interfere With The Incidence and Severity Of Sepsis
As discussed above, the presence of several underlying disorders alters clinical outcome of patients with sepsis. However, an open question is whether these diseases influence the systemic response to the infection. Data from animal models, healthy human volunteers, and patients with sepsis sustain the hypothesis of an inadequate local and systemic inflammatory response to microorganisms in the presence of concomitant conditions. The most significant underlying disorders that influence the host response to infection are described below.
Because underlying disorders are more frequent in the elderly population, the role of age is crucial in delineating the true influence of underlying disorders on host response and susceptibility to infection. Many large epidemiological reports now demonstrate that age is related to the occurrence [3] and prognosis [4] of infection.
Aging is associated with dysfunction of T cell—mediated adaptive immunity. Thymic involution leads to a progressive loss of the T cell repertoire in response to neoantigens [35] that is partially compensated for by the expansion of memory CD8 cells [36, 37]. In addition, B cell humoral response decreases with aging because of poor CD4-induced costimulation [38]. Moreover, the decreased production of reactive oxygen species by neutrophils and monocytes contributes to diminish the clearance of pathogens [36, 37, 39].
As a part of innate immunity, the proinflammatory and coagulation response to infection is not diminished in the elderly population. First, it is noteworthy that the baseline level of inflammatory stress increases during aging [40]. The magnitude of the increase in the baseline concentration of IL-6 without any infectious episode is a reliable marker for functional disability and a predictor of disability and mortality in the elderly [41]. Indeed, Cohen et al. [42] reported that activation of the coagulation (D-dimer) and inflammatory (IL-6) pathways at baseline is associated with mortality and decline in function. Additionally, aging is associated with inadequate response to infections and sepsis-related stimuli. Monocytes from elderly patients undergoing surgery produced more TNF-α and expressed higher levels of CD11b/CD18 than did those from younger patients [43]. An age-related increase in TNF-α, IL-1β, and IL-6 concentrations has been found in supernatants of mitogen-stimulated mononuclear blood cell cultures from healthy elderly people [44]. After challenge with lipopolysaccharide in healthy young and elderly volunteers [45], the latter showed a more prolonged fever response than that of the younger controls, and levels of TNF-α and soluble TNF receptor I levels were higher in the elderly group. The elderly patients also had a more profound and prolonged increase in soluble TNF receptor I levels. This study suggests that aging was associated with an altered host response, with initial hyperreactivity and a sustained secondary anti-inflammatory response. Elderly persons with pneumococcal infection also show prolonged and exaggerated cytokine responses, compared with those of younger persons [46].
However, 2 small studies have failed to demonstrated higher levels of proinflammatory cytokines in response to lipopolysaccharide [47] or bacterial pneumonia [48] in elderly patients than in younger control subjects. Despite these apparently contradictory results, most of the studies concluded that the coagulation and pro- and anti-inflammatory response to endotoxin or bacterial infection in elderly people is higher and more prolonged than in the younger population, increasing the susceptibility to nosocomial infection.
It remains difficult to analyze the respective role of age and underlying disorders, because, in addition to being a common feature of healthy elderly subjects, low-grade inflammatory activity at baseline is strongly associated with such age-related diseases as atherosclerosis, dementia, type 2 diabetes, sarcopenia, and osteoporosis. However, Walston et al. [49] found no difference in low-grade inflammatory activity in elderly people with or without diabetes or cardiovascular disease [49]. The role of aging is then important for the outcome of infection, but the multiple mechanisms that contribute to this observation remain controversial.
Diabetes mellitus modifies the immune response to microorganisms. There is now a large body of data showing impairment of innate and adaptive immunity in patients with diabetes. For example, the production of Th1-related cytokines after challenge with lipopolysaccharide was impaired in diabetic BB rats [50]. Expression of inducible nitric oxide synthase was inhibited in this setting and is responsible for a decreased production of nitric oxide, a key inflammatory mediator [51]. Finally, neutrophils submitted to hyperglycemic media had delayed apoptosis [52].
Cancers and hematologic malignancies are frequently responsible for an increased risk of bacterial, fungal, or viral infection [53]. Although infectious complications in patients with cancer have been well described [54], estimates of the true incidence of severe sepsis in this population remain unknown. In a study by Angus et al. [4], 1 in 6 patients with sepsis had an underlying malignancy. Of every 1000 patients with cancer in a given year, 16.4 will experience an episode of severe sepsis. Severe sepsis accounts for 8.5% of cancer deaths, with an estimated cost of $3.4 billion a year [55]. Many conditions associated with cancer contribute to the increased incidence of infection in these patients, including frequent underlying disorders, indwelling catheters, mucosal impairment [56], chemotherapy, or radiotherapy. Various mechanisms contribute to the risk of infections related to cancer treatment. The use of corticosteroids is responsible for altered phagocytosis and defective cellular immune response [57] that exposes patients to fungal, viral, and bacterial infection. Neutropenia is a common side effect of chemotherapy that obviously contributes to immunosuppression. Neutropenia favors the emergence of systemic bacterial infections from any source and provides an opportunity for fungi to develop [54]. Other cell populations are also directly impaired by cytotoxic drugs, especially methotrexate and purine analogues, that decrease the number of T cells and facilitate the occurrence of infections due to Pneumocystis and Mycobacterium species and herpesviridae [58]. Antipanlymphocyte antibodies, such as alemtuzumab, are responsible for deep and prolonged decreases in B and T cell counts, with increased risk of severe opportunistic infections [59]. Hematologic malignancies induce immunosuppression by the direct impairment of immune cells, including number and functions of T cells, in lymphoproliferative [60] and myeloid disorders [15]. Abnormalities of the innate and acquired immune system have also been recognized early in the course of solid neoplasia [61, 62].
Ischemia-reperfusion injury occurs because of hypotension related to sepsis and/or underlying cardiovascular condition. The macrophage response in liver to secondary endotoxin stimulation is decreased during the early phase of liver ischemia-reperfusion injury [63]. Clinical infection after surgery or trauma, however, usually occurs later in a patient' course, when the Kupffer cell—mediated phase of injury has passed and the neutrophil-mediated phase of reperfusion injury predominates. A relative tolerance to secondary injury has been observed if endotoxin challenge is delayed for at least 24 h after liver ischemia-reperfusion [64]. Moreover, as described in the 1980s, resuscitated cardiac arrest is an illustration of ischemia-reperfusion. A blunted inflammatory response in this setting could explain an increased susceptibility to infection. In addition, the postresuscitation syndrome is frequently followed by bacteremia whose mechanism may be multifactorial [65].
Underlying Disorders and Innate Immunity Response Of Patients With Infectious Disease
The state of innate immunity of septic animals and patients with underlying disorders is poorly described in the literature. In the Clinical Evaluation Committee report of the PROWESS trial, Dhainaut et al. [28] reported a blunted response (lower IL-6 concentrations at baseline) in patients with underlying disorders, compared with responses in patients without underlying disorders.
To better understand the processes involved, a new retrospective analysis was performed, including the inflammatory and coagulation response of the 850 placebo recipients in the PROWESS trial. Compared with patients without underlying disorders, the 165 patients with underlying disorders experienced a weaker procoagulant state, including higher protein C activity, which was statistically significant until day 3, and lower D-dimer levels, which were statistically significant at day 1. A weaker inflammatory response also appeared, with statistically significantly lower IL-6 concentrations until day 2 in patients with underlying disorders than those without underlying disorders (figure 1).
Three main factors may be responsible for this apparent blunted response to infection in patients with underlying disorders. First, patients with underlying disorders are older than patients without underlying disorders. However, taking into account the immunosenescence described above, the innate immunity seems to be well preserved or negligibly affected and, in some cases, up-regulated. Consequently, aging does not appear to be the explanation for the blunted response observed in patients with underlying disorders. Second, this blunted response could be related to the underlying disorder itself. The direct consequences of individual underlying disorders on innate immunity have not been extensively investigated in humans. However, several studies reported an enhanced response to endotoxin in patients with unstable angina [66] or type 2 diabetes [67] (although the response seems more complex in animal models [50, 51]) and in mice with preexisting burn injury [68]. Finally, the most likely explanation is that the host response to infection in patients with underlying disorders is not actually blunted but remains adequately adapted to the true severity of infection. In such patients, underlying disorders contribute not only to the preexisting conditions but also probably into the physiological part of scoring systems. Consequently, the severity scoring systems overestimate the true severity related to the infectious insult.
Conclusion
The overall mortality of patients with sepsis has increased over the past decade in a population in which underlying disorders were common. Most of these underlying disorders are associated with an increased mortality rate and may alter the overall immune response, even though the coagulation and inflammatory response to infection is not altered. It remains unclear whether these underlying disorders are responsible for death in patients with sepsis because of altered immune response, previous organ dysfunction and poor physiological status, or withdrawal of life supportive care. Whatever the cause of the death, we assume that the influence of each of the categories of underlying disorders will be examined in future studies according to the “P” classification of the PIRO concept, helping physicians in the interpretation of interventional studies.
Acknowledgements
We are indebted to the hard work of all staff who contributed to the PROWESS trial at the 155 institutions around the world.
Financial support. Eli Lilly and Company supported the PROWESS trial.
Potential conflicts of interest. J.-F.D.: investigator in the PROWESS study and is a consultant for Eli Lilly. J.J., D.R.N.: employees and stockholders of Eli Lilly during these studies. Y.-E.C.: no conflicts.
- © 2005 by the Infectious Diseases Society of America
References
This Article
- AbstractFree
- » Full Text (HTML)Free
- Full Text (PDF)Free
Classifications
Services
Citing Articles
Google Scholar
PubMed
Related Content
Share
- Email this article
-
CiteULike
Connotea
Delicious
Facebook
Google+
Mendeley
Twitter
Published on behalf of
IDSA Members: For your free access to this journal, log in via the IDSA members area.
The Journal
Impact Factor: 8.186
5-Yr impact factor: 7.898
For The Media
For Authors
- Services for authors
- Self-archiving Policy
- IDSA Conflict of Interest policy
- Instructions for Authors
- ICMJE Form
- Submit Now!
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy
Corporate Services
Alerting Services
Related IDSA Publications
