«INFLAMMATORY PROTEINS, GENETIC VARIATION, AND ENVIRONMENTAL INFLUENCES ON HEALTH CARE ASSOCIATED INFECTION DEVELOPMENT IN SEPSIS A Dissertation ...»
• The hazard ratio to predict development of HAIs for each 10 point increase in anti-inflammatory cytokine based on plasma IL-10 is e0.0002799x10 = 1.002. For a 10 point increase in IL-10 there is a 0.3% increase in developing a HAI.
• The hazard ratio to predict development of HAIs for each 10 point increase in ratio of pro- to anti-inflammatory cytokine is e-0.00890x10 = 0.915. For a 10 point increase in IL-6:IL10 ratio there is a 8.5% lower risk of developing a HAI.
The multivariate model included several predictors associated with HAI development. All variable tested in univariate cox regression with a p 0.25 were included in the final model. These variables included: age, race, APACHE II score, invasive device score at first HAI, received insulin during ICU stay, baseline C-reactive protein, received antibiotics prior to ICU, IL-6 genotype, and IL-10 genotype. Only variables with of probability of 0.05 were included in the final model. As some IL-6 genotypes were not significant, a dichotomous variable for genotype was created to compared the CG and CC to the GG alleles. The final regression model is shown in Table 4-17.
CHAPTER 5. DISCUSSION
Baseline Infections The most common infection responsible for sepsis at ICU admission in this study was pneumonia, followed by UTI, skin and soft tissue infection, and gastrointestinal or intrabdominal infection. The pattern of infection (site and organism) differs across ICU populations as described in the discussion. This study focuses on medical ICU patients.
Our finding of pneumonia as the primary cause of sepsis causing ICU admission is in agreement with some studies43,92 but urinary tract infections may precede pneumonia as the primary cause.42 We found a significantly higher percentage of patients with high IL-6 levels versus lower levels who developed arrhythmias during their ICU stay (52.6% vs. 18.6%, p = 0.007). The median IL-6 difference was 303.3 vs. 98.1 (p = 0.0008) for those who developed arrhythmias versus those who did not. This is consistent with emerging literature. Aviles, et al. have shown that atrial fibrillation is associated with elevated CPR levels, 117 and Boos provides a summary of four studies that found associations of elevated CRP and IL-6 levels with atrial fibrillation and one study that did not.118 This population was composed primarily of males (97.9%), thus limiting conclusion about females. It is well documented that males have a higher rate of infections;119 however; this study included only 4 women. One female (25%) and 16 males (21.6%) developed HAI. This sample does not include enough females for a valid comparison. There were no differences in the rate of HAI developed between white and black patients in this study; however, we did find a difference in genotype among white and black patients with sepsis who did and did not develop HAI. The literature reports a higher incidence of infections in blacks as compared with whites.120
First Health Care Associate Infection
The percentage of HAI caused by Candida (64%) was a surprising finding in this study. Typically, the expected leading organisms associated with HAI are gram positive (such as Coagulate-negative Staphylococcus, Staphylococcus aureus, and Enterococcus spp.).43 This is a shift away from gram negative organisms in the past, and now Candida species have emerged as the fourth most common cause of health care associated blood stream infections.121,122 There were three species of Candida identified (Candida albicans, Candida glabrata, Candida tropicalis) in addition to unidentified Candida species. Colonization of the urinary tract with Candida is becoming more common and the line between colonization and urinary tract infection is not easily distinguished.123 Among this sample, there were only two urinary tract infections (Candida species not identified). There were two Candida albicans pneumonias, and all of the remaining Candida infections were bloodstream infections (3 Candida albicans, 2 Candida glabrata, 2 Candida tropicalis). The incidence of fungemia in the ICU had been increasing and there have also been changes in the species pattern of Candida. Candida albicans was the primary species in the 1980s, but now almost half of Candida infections are non-albicans.124 The mortality associated with Candida glabrata is higher than Candida albicans, 121,124 and Candida glabrata has been reported as the second most frequent cause of Candidemia in US hospitals.121 Two primary risk factors associated with Candida infections are colonization of the skin and mucous membranes and the presence of invasive devices.122 Candida colonization is not always treated and the urinary tract and upper airway may be colonized with Candida. Three patients with Candida infections in this study were not treated. One patient had Candiduria which was not treated and developed Candidemia three days later; however, the blood culture results were not complete until after the patient had expired. Another patient with Candidemia was not treated because care had been withdrawn before results were received, and another patient with Candidemia expired before results were received.
Among all subjects developing infections, less than half (47%) developed fever or hypothermia, in contrast to 74.4% experiencing fever or hypothermia at initial presentation. Fever in the ICU is one of the triggers that lead to a careful clinical assessment for its cause and possible laboratory assessments.54 Patients receiving corticosteroids have a blunted febrile response due to down-regulation of inflammatory mediators (IL-1β, TNF-α, IL-6) that cause fever. The threshold for laboratory assessment (blood culture, BAL, and others cultures) is lower in patients receiving corticosteroids,125 which may explain why half of the infections were detected in the absence of fever.
Meduri et al. identified 60% of infections occurring after ICU day seven in patients receiving corticosteroids in the absence of fever.126 The presentation of HAI required less fluid resuscitation and fewer vasopressors (17% vs. 32%), than their initial ICU presentation with sepsis.
A significantly higher number of patients who received corticosteroids developed HAI compared to those who did not (82.4% vs. 42.6%, p = 0.005); however, when controlling for other factors (shown in AIM3) this was not significant. Overall, patients who developed HAI had more invasive devices, more days of mechanical ventilation (11.1 vs. 5.9, p = 0.03) and more days of ICU stay (15.3 vs. 6.4, p 0.0001) when compared to those who did not develop HAI. This is in agreement with literature which consistently shows high correlation of these variables with development of HAI.42
The primary goal of aim one was to investigate whether baseline protein expression levels of pro-inflammatory cytokines, anti-inflammatory cytokines, or their ratios influence the development of subsequent HAI in patients with sepsis.
There were no significant differences in levels of pro-inflammatory cytokine, anti-inflammatory cytokine, or their ratio among subjects who did and did not develop at least one HAI during their ICU stay. Specifically, an exaggerated pro-inflammatory response was present in 6 (31.6%) compared to 11 (18.6%) participants without an exaggerated pro-inflammatory response who developed subsequent HAI. This difference was not significant (p = 0.55). Likewise, an exaggerated anti-inflammatory response was present in 5 (26.3%) compared to 12 (20.3) participants without an exaggerated antiinflammatory response who developed subsequent HAI. This difference was also not significant (p = 0.43). There was also no significant difference in the log of proinflammatory to anti-inflammatory ratios among those who did and did not develop subsequent HAI.
One study investigated a similar relationship. Ramirez et al. performed a prospective observational study to examine the relationship between systemic inflammatory response and development of ventilator-associated pneumonia (VAP). 127 They included patients on mechanical ventilation who were expected to remain on mechanical ventilation for more than 48 hours. They excluded patients who developed other HAIs. They measured several cytokine (IL-6, Il-10, and others) at baseline and then every 96 hours. Their sample included 44 patients, among which 9 developed VAP.
Findings included higher baseline IL-6 pg/ml among patients who subsequently developed VAP. Higher median IL-6 pg/ml were reported for confirmed VAP cases (235 (188-620)) as compared to non-suspected (92 (43 - 167)), suspected (120 (112 - 161)) VAP cases (p = 0.02). They found no significant difference in median IL-10 pg/ml for non-suspected (0 (0 - 4)), suspected (0 (0 - 4)), and confirmed (6 (3 - 12)) cases of VAP (p = 0.16). They reported IL-10 detection limits of 1pg/ml.
The findings of Ramirez provide limited evidence that a clinical relationship between exaggerated pro-inflammatory response and development of infections exists.127 In our study, we were unable to identify a difference. Our findings may be limited by the timing of cytokine measurement and other factors. Baseline cytokine measurements were collected at enrollment within 72 hours of developing sepsis. IL-6 levels usually peak after IL-1β and TNF-α, around eight hours following an insult. The variation in timing for each subject (0 – 72 hours) and use of corticosteroids ( 1 mg/kg) may have limited these findings.
The goal of aim two was to investigate the variance in cytokine genes to determine if they influence levels of protein expression or development of HAI.
Genotye and Cytokine Level We examined cytokine levels by genotype and compared subjects grouped by degree of inflammatory response. In Chi-square analysis comparing subjects with an exaggerated inflammatory response to those without an exaggerated response, we found no difference for rs1800795 or rs1800896. There were no significant differences in median or log IL-6 level based on genotype rs1800795. We did find a significant difference in IL-10 level based on genotype rs1800896. Participants who were homozygous AA or GG had significantly higher log IL-10 levels than heterozygous GA genotypes (0.02).
While promoter polymorphism may increase or decrease transcription levels, there are also epigenetic and other factors that influence gene regulation. Taudorf et al.
studied the cytokine response to an endotoxin challenge in 200 young healthy men. They found up to a 6 fold increase in cytokine levels following endotoxin injection but they did not find significant differences in cytokine levels among many commonly studies cytokine SNPs, including rs1800896 and rs1800795.100 Among all comparisons involving genotype and cytokine level, we only found one difference. Other studies have shown varied results.
Our findings, or lack of findings, may be limited by several factors. First, as noted in aim one, timing of cytokine measurement may be a factor. An underlying assumption of this research was that cytokine levels would be persistently elevated. This has been well documented in the literature, particularly for non-survivors.87-89 This assumption could not be evaluated in this study because plasma samples were only collected at baseline and not over time. Second, variability in cytokine measures may be a factor.
This variability ranges from technician technique to the method selected for cytokine measurement to the various proprietary antibodies used by different biotech companies for each cytokine. Lastly, the findings of Taudorf suggest that either there may be no difference or, more likely, there are several other factors that may be involved such that these differences are not expressed in a healthier or less acutely ill population.
Genotype and Development of HAI
There were no statistically significant differences among genotypes or haplotypes for development of HAI. There were no CC_AA, CC_GA, or GG_AA haplotypes among the 17 participants who developed HAI. The CC_GG haplotype was not observed in another study.128 The most common haplotype was the heterozygous CG_GA. Baier, et al., in a retrospective review of 293 low birth weight infants, investigated the IL-6 rs1800795 and IL-10 rs1800896 SNPs and sepsis outcomes. In their study, the rs1800795 C Allele was associated with late blood stream infections in African American but not Caucasian infants. The incidence of blood stream infections was 69% in African American infants with the C Allele compared to 46% in African American infants with the GG genotype (p = 0.02). Racial differences have not been investigated in this dissertation study and will be explored in future analyses. Baier et al. also found an increased association of the rs1800896 A Allele with increased incidence of late blood stream infections.128 The rs1800795 CC genotype has been associated with fungal blood stream infections in Caucasians. This study population includes a large number of fungal blood stream infections. Among the 7 participants with the rs1800795 CC genotype only one developed a HAI, which was fungemia caused by Candida albicans.
The goal of aim three was to investigate the effects of protein expression levels, genetic variation, and environment on development of HAI. A series of Cox regression analyses were performed among those who did and did not develop HAIs during ICU stay. Although the univariate models provided important risk ratios, the multivariate model provides the variables most predictive for the risk ratio and can be used to predict the risk of HAI development. We found no difference in protein expression levels, one genetic variation, and two environmental factors that strongly predict risk of HAI development. The final model included APACHE II, invasive devise score, and antibiotics prior to ICU, and IL6 rs1800795.