Friday 31 May 2013
Immunomodulation in community-acquired pneumonia
Promotor: Prof.dr D.H. Biesma
Defence: 31 May 2013
Community-acquired pneumonia (CAP) is a common disease, which causes considerable morbidity and mortality worldwide. Despite the availability of effective antibiotics, pneumonia remains the leading cause of death from infectious diseases and mortality from CAP has not decreased in the last decades. To ultimately improve outcome of lower respiratory tract infections, new (non-antibiotic) treatment strategies are therefore urgently needed. In this thesis, options for adjuvant treatment are described, with a focus on immunomodulation.
The immune response in community-acquired pneumonia
The host response to pneumonia is characterized by an acute inflammatory response. The nature of the inflammatory response to a degree is determined by the type of micro-organism that is encountered. Thus, intracellular bacteria require a different response from the immune system than extracellular bacteria. The existing knowledge on pathogen specific inflammatory response patterns in CAP however is limited, and therefore, further elucidation of pathogen-specific inflammatory profiles is needed.
In Chapter 2, we report the measurement of a comprehensive panel of systemic inflammatory markers in 469 patients with CAP, including prototype pro- and anti-inflammatory cytokines and chemokines. Compared to patients with CAP caused by an atypical pathogen or a virus, patients with pneumococcal pneumonia showed significantly higher leukocyte counts, procalcitonin (PCT), interleukin-6 (IL-6), interleukin-1 receptor-antagonist (IL-1ra), and monocyte-chemotactic-protein-1(MCP-1) concentrations. The PCT concentration was comparable between those with atypical pathogens and viruses. Patients with CAP caused by an atypical pathogen had significantly higher IL-17 and interferon-γ (IFN-γ) concentrations than did patients with CAP caused by other micro-organisms. Moreover, patients with an atypical pathogen had the lowest concentrations of IL-10 and leukocyte counts. Within the group of atypical pathogens, also differences in the inflammatory response were found. In particular, Legionella pneumophila elicited the highest concentrations of C-reactive protein (CRP), PCT, leukocytes, IL-6, tumor necrosis factor-α (TNF-α), IL-10, and IL-8. Antibiotic use prior to admission did not change inflammatory profiles. We conclude that the major causative micro-organisms in CAP trigger distinct inflammatory response profiles in the host.
While an inflammatory response as such is required to combat invading pathogens, an excessive inflammatory response contributes to adverse outcome. Modulation of the immune response could therefore offer promising treatment options in CAP. In this thesis, we focus on the immunomodulatory properties of corticosteroids, macrolides and vitamin D, and their potential role in prevention and treatment of CAP.
Immunomodulation by dexamethasone
Corticosteroids are potentially of use for adjuvant treatment in CAP, due to their dampening effect on the inflammatory response or reversal of an inadequate hypothalamic-pituitary-adrenal response. Adjuvant corticosteroids have already been proven beneficial in other infectious diseases, such as bacterial meningitis and septic shock. The effectiveness of corticosteroids in CAP is still a matter of debate, because prior studies in CAP have reported conflicting results.
In a randomised placebo-controlled trial, we studied the effect of adjuvant dexamethasone on length of hospital stay in non-immunocompromised adults with CAP. The results of this trial are shown in Chapter 3. On hospital admission with CAP, patients were randomly assigned to either a four-day course of adjuvant dexamethasone (5 mg once a day) or placebo. A total of 304 patients were included, of which 151 received dexamethasone and 153 received placebo. Compared with placebo, adjuvant dexamethasone reduced length of hospital stay by one day (median length of stay 6.5 days versus 7.5 days; p:0.048). Furthermore, levels of CRP and IL-6 decreased faster in the dexamethasone group than in the placebo group, while the decline in IL-10 levels did not differ between the two groups. Hyperglycaemia was more often detected in the dexamethasone group, but severe adverse events were rare and did not differ between the groups.
The influence of dexamethasone on the cytokine response in CAP is further explored in Chapter 4. In serum from the aforementioned 304 patients, an extensive panel of cytokines and chemokines was measured. In general, compared to placebo-treated patients, dexamethasone-treated patients had a faster decline of IL-6, IL-8, MCP-1 and TNF-α. Differences in dexamethasone effect were found for different microbial aetiologies. In patients with pneumococcal pneumonia, dexamethasone had little additional influence on cytokine concentrations, while in patients with CAP caused by an atypical pathogen, dexamethasone gave a faster decrease of the levels of IL-1ra, IL-6 and MCP-1.
The beneficial effect of exogenous corticosteroids suggests that in CAP the physiological regulatory systems for the inflammatory response are inadequate. Cortisol, the endogenous corticosteroid secreted by the adrenal cortex, is an important regulator of inflammation and exerts anti-inflammatory and immunosuppressive activities. Generally, severe illness and stress strongly activate the hypothalamic-pituitary-adrenal (HPA) axis, leading to an increased level of cortisol. However, many critically ill patients develop dysfunction of the HPA axis. This syndrome is referred to as critical illness-related corticosteroid insufficiency (CIRCI), defined as cortisol <10 μg/dL or a change in serum cortisol (delta) of <9 μg/dL after administration of 250 μg synthetic ACTH. CIRCI is associated with poor outcome, and corticosteroid replacement may be beneficial in these patients. In Chapter 5, we investigated whether subgroups can be defined that will benefit in particular from corticosteroid therapy, based on the cortisol response to infection. In patients with CAP, we showed that a low serum cortisol <10 μg/dL was not associated with adverse clinical outcome. Most of the patients with a serum cortisol <10 μg/dL had non-severe disease rather than corticosteroid insufficiency. This finding suggests that the current definition of CIRCI is not applicable in patients with CAP, as low cortisol responses merely reflect low disease severity in many patients. We show that patients with a high pro-inflammatory cytokine response but a discrepantly low cortisol benefited most from adjuvant dexamethasone. In the placebo group, 43% of the patients died or were admitted to the ICU, compared to 0% in the dexamethasone group (p:0.020). We conclude that in CAP patients, the relation between cytokine response and cortisol level might better reflect corticosteroid insufficiency, and might help to identify patients who may benefit in particular from adjuvant corticosteroid therapy.
A high cortisol level on admission with CAP is known to be associated with adverse clinical outcome. Whether this also holds true for cortisol levels that remain increased during the course of disease is unknown. Moreover, a potential risk of synthetic corticosteroids is secondary adrenal insufficiency. The time to recovery of the HPA axis after a short course of dexamethasone during infection is unclear. In Chapter 6, we report serial cortisol measurements throughout the course of CAP. It is shown that persisting high cortisol levels are also associated with poor outcome. In addition to cortisol level on admission, which may serve as prognostic biomarker in CAP, the change in cortisol from day zero to day two or four could be another meaningful biomarker in CAP. Additionally, we report that the endogenous cortisol production was almost completely suppressed by dexamethasone therapy after the first dose, but was fully recovered on day 30.
Immunomodulation by macrolides
Macrolides are known to possess immunomodulatory properties, next to their antimicrobial effects. The mmunomodulatory effects of macrolides are beneficial in chronic pulmonary inflammatory syndromes, such as diffuse panbronchiolitis, cystic fibrosis, asthma and bronchiectasis. Whether macrolides also exert favourable munomodulatory effects during acute inflammatory conditions, such as CAP, is less clear. In Chapter 7, we provide an overview of the existing evidence from in vitro and in vivo studies on the immunomodulatory effects of macrolides in CAP. Macrolides were shown tochange the nature of the immune response during acute inflammation in three ways: by suppression of the cytokine response, by changing the behaviour of inflammatory cells to a more anti-inflammatory nature, and by affecting structural cells of the respiratory tract. Current evidence is restricted by the heterogeneity of the studies, with regard to methods and experimental model systems. Furthermore, experimental and clinical studies on the immunomodulatory effects of macrolides when given adjuvant to β-lactam antibiotics are lacking.
In order to further elucidate the mechanisms of immunomodulation by macrolides during acute inflammation, in particular when given in combination with β-lactam antibiotics, we designed an in vitro model of acute infection with Streptococcus pneumoniae. In Chapter 8, we report that macrolides, alone or adjuvant to β-lactam antibiotics, attenuated the pro-inflammatory cytokine response in whole blood stimulated with heat-killed S. pneumoniae. This suggests an immunomodulatory effect of macrolides. This effect was not observed in experiments with viable S. pneumoniae (either macrolide-susceptible or -resistant). Stimulation of whole blood with viable pneumococci induced an overwhelming cytokine response, but no difference in cytokine response was detected between a macrolide-containing antibiotic regimen and β-lactams alone.
Immunomodulation by vitamin D
Vitamin D has pleiotropic immunomodulatory properties, apart from its function in calcium and bone homeostasis. Vitamin D deficiency, which is common worldwide, is associated with an increased susceptibility to respiratory tract infections. In patients hospitalized with CAP, vitamin D deficiency has been associated with adverse clinical
outcome, but the prognostic value of vitamin D status is unknown. In Chapter 9, it is reported that in our CAP cohort, 53% of the patients was vitamin D deficient. We confirmed that vitamin D deficiency is associated with adverse clinical outcome in CAP. Median 25-hydroxyvitamin D levels were significantly lower in patients who were admitted to the intensive care unit (ICU), compared to patients without ICU admission (34.9 nmol/L versus 48.3 nmol/L; p:0.04). Patients who died within 30 days had significantly lower 25-hydroxyvitamin D levels, compared with patients who survived (25.8 nmol/L versus 48.8 nmol/L; p<0.01). Vitamin D status at the time of hospital admission appeared to be an independent predictor for 30-day mortality, adding prognostic value to other biomarkers and prognostic scores. With an area under the curve (AUC) of 0.83, the prognostic accuracy of the combination of vitamin D status and PSI score was significantly superior to the accuracy of other predictors of 30-day mortality, or the PSI score alone.
Whether the association between vitamin D deficiency and increased susceptibility to Infections, and the association between vitamin D deficiency and poor outcome in CAP, are based on a causal relationship is unknown. In that case, vitamin D supplementation to high risk populations, or in the acute care of CAP, would be beneficial. Therefore, in three large independent case control studies described in Chapter 10, we studied the association between vitamin D supplementation and the risk of pneumonia. In hospitalised patients in study 1 and 2, after adjustment for confounding, vitamin D supplementation was not associated with a lower risk of developing pneumonia. In a primary care setting in study 3, after adjustment for confounding, an even increased risk of pneumonia was observed among vitamin D users. Interestingly, the use of bisphosphonates and/or calcium or oral corticosteroids significantly modified the association, suggesting that effects of vitamin D supplementation depend on indication and/or underlying illnesses. In patients using vitamin D supplementation without concomitant use of bisphosphonates, calcium or oral corticosteroids, the risk of pneumonia was increased. In conclusion, based on the current available data, (standard) vitamin D supplementation cannot be recommended for prevention of CAP.