INTRODUCTION
Sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection (1). Mounting evidence indicates that the inflammatory response in sepsis is highly variable, ranging from severe hyperinflammation (HI) to profound immunosuppression (2,3). These differential states, or “immunological endotypes,” can be classified by various biomarkers, of which the most commonly used are increased serum ferritin concentrations, associated with HI (4), and decreased monocytic (m)HLA-DR expression, related to immunoparalysis (IP) (5,6).
Several patient-related factors might influence the immunological endotype of a patient with sepsis. In light of the increasing prevalence of obesity in the past decades, a higher proportion of sepsis patients has obesity. It is well known that obesity is associated with an increased risk of pneumonia and acute respiratory distress syndrome (7,8). Perhaps counterintuitively, obesity has also been associated with a survival benefit once a patient becomes critically ill, including those suffering from pneumonia and sepsis, and even when correcting for demographic baseline differences and disease severity (7–9). This phenomenon is known as the “obesity paradox,” and one potential underlying mechanism is the secretion of immunomodulatory mediators from the adipose tissue. These mediators are known as adipocytokines and have numerous regulatory effects, including involvement in inflammatory processes. Obesity results in increased concentrations of leptin, whereas adiponectin concentrations are decreased. Leptin has been shown to exert proinflammatory effects by promoting the phagocytotic activity of macrophages and production of proinflammatory cytokines (10). Conversely, an anti-inflammatory role of adiponectin has been proposed (11,12). Therefore, obesity may shift the balance toward a more proinflammatory endotype, causing a chronic low-grade inflammatory state (13). Resistin also is an adipocytokine with distinct proinflammatory effects; for instance, it induces increased expression of proinflammatory cytokines and leukocyte recruitment (14,15). Studies in patients with sepsis indicate increased concentrations of leptin and resistin and decreased concentrations of adiponectin compared with healthy controls (16–18). However, the role of these adipocytokines in modulation of the immune response and their relationship with clinical outcomes in sepsis patients with obesity are incompletely understood.
In the current study, we investigated the influence of obesity on the inflammatory response and clinical outcomes in sepsis patients and assessed to what extent adipocytokines play a role in this process. To this end, we leveraged a cohort of sepsis patients with distinct immunological endotypes.
METHODS
Study design and participants
In this secondary analysis of a prospective study, we used data and biomaterial from a Greek cohort of hospitalized patients with sepsis (19). All patients (18 years or older) with sepsis admitted to 1 of the 14 recruiting study sites in Greece were screened for eligibility. Informed consent was provided by the patients or by one's first-degree relative before enrollment (approval by the National Ethics Committee of Greece 78/17). All patients were clinically diagnosed with sepsis or septic shock, according to the Sepsis-3 criteria (1), due to community-acquired pneumonia, health care–associated pneumonia, ventilator-associated pneumonia, acute cholangitis, or primary bloodstream infection. Exclusion criteria included the following: “do no resuscitate” policy, pregnancy or lactation, other causes of infections, any stage IV malignancy, active tuberculosis, human immunodeficiency virus infection, primary immunodeficiency, oral or intravenously administered corticosteroids at a daily dose ≥0.4 mg/kg or equivalent prednisone for the last 15 days, treatment with any anticytokine biological during the last month, medical history of systemic or multiple sclerosis, or any other demyelinating disorder. Patients were classified into immunological endotypes by measuring serum ferritin concentrations and mHLA-DR expression. HI was defined as a serum ferritin concentration of 4,420 ng/mL or greater (4), regardless of mHLA-DR expression. IP was defined as the expression of mHLA-DR of 5,000 antibodies bound per cell (Ab/cell) or less (6), provided that ferritin concentration was lower than 4,420 ng/mL. Patients who did not fulfill the HI or IP criteria were categorized as “unclassified” (UC).
Data collection and analyses
Baseline patient characteristics, comorbidities, and clinical data during hospital admission were collected from medical files. Inflammatory parameters, measured for routine clinical assessment, were extracted from medical and laboratory files. In addition, the Acute Physiology and Chronic Health Evaluation (APACHE II) and Sequential Organ Failure Assessment (SOFA) score were calculated to assess disease severity.
Sample processing and measurements
Within 24 h after sepsis diagnosis, blood was collected in serum and ethylenediaminetetraacetic acid tubes and centrifuged at 3,500 relative centrifugal force at room temperature for 10 min. All samples were aliquoted and stored at −80°C before analysis. Concentrations of ferritin were measured in serum samples by an enzyme immunoassay (ORGENTEC Diagnostika GmbH, Mainz, Germany) with a lower detection limit of 75 ng/mL. For measurement of mHLA-DR expression, whole blood was incubated for 15 min in the dark with the Quantibrite anti–HLA-DR/antimonocyte PerCP-Cy5 antibodies (Becton Dickinson, Cockeysville, MD). After incubation, cells were analyzed on a flow cytometer (FC500; Beckman Coulter, Miami, FL). The expression of HLA-DR molecules on CD14/CD45 monocytes was reported as the number of Ab/cell. Plasma concentrations of leptin, adiponectin, and resistin were measured using enzyme-linked immunosorbent assays (DuoSet ELISA; R&D systems, Minneapolis, MN) according to the manufacturer's protocol, with lower detection limits of 0.78 ng/mL, 0.16 μg/mL, and 0.78 ng/mL, respectively.
Statistical analysis
Data were analyzed using SPSS software version 25.0 (IBM, Armonk, NY) and GraphPad Prism software version 8.0 (GraphPad Software, Inc., San Diego, CA). Differences between groups were assessed by Mann-Whitney U tests or Kruskal-Wallis tests with Dunn multiple comparisons tests for continuous variables and by χ2 tests or Fisher exact tests for categorical variables. Bonferroni correction for multiple testing was applied where appropriate. The SOFArank score was calculated as described before (20). First, a daily delta SOFA score was calculated by subtracting the baseline SOFA score from the daily SOFA score during the first 7 days. The daily SOFA score was set at zero after patients were discharged, whereas it was set at the maximum score of 24 after patients had deceased. Subsequently, the SOFArank score was defined by the sum of the daily delta SOFA scores. A negative score indicated a decline in the SOFA score and was classified as clinical improvement. A positive score indicated an increase in the SOFA score and was classified as clinical deterioration. Receiver operating characteristic analyses were performed to assess the prognostic performance of resistin for 28-day survival by calculating the area under the curve (AUC). The optimal cutoff value was defined based on the maximal Youden J index and used to assess differences in 28-day survival for high versus low resistin concentrations. Differences in 28-day survival between groups were assessed by survival analyses with the hazard ratio based on the log-rank (Mantel-Cox) test. Binary logistic regression was used to assess the effect of adipocytokine concentration, age, body mass index (BMI), and the APACHE II score on 28-day mortality. Multivariate analyses were performed with age, BMI, and APACHE II score as covariates. Correlations between BMI, adipocytokine concentrations, and inflammatory parameters were assessed by Spearman rank correlation tests. A P value <0.05 (two-tailed) was considered statistically significant.
RESULTS
Patient characteristics
Adipocytokine concentrations were measured in plasma of 167 sepsis patients with available plasma samples at sepsis diagnosis and available data on BMI. To assess the relationships between BMI and adipocytokines as well as inflammation, the cohort was divided in BMI categories according to the World Health Organization definition: patients with normal weight (BMI, 18.5–24.9 kg/m2; n = 67), overweight (BMI, 25.0–29.9 kg/m2; n = 56), and obesity (BMI ≥30 kg/m2; n = 42; Supplementary Figure 1, https://links.lww.com/SHK/B589). Two underweight patients (BMI <18.4 kg/m2) were excluded from this analysis. Table 1 shows that patients with overweight and obesity were younger compared with patients with normal weight (75 [62–81] and 67 [61–78] years vs. 82 [70–88] years, P = 0.003 and P < 0.001 between subgroups, respectively). Moreover, patients with obesity had a higher SOFA score (12 [10–15] vs. 10 [6–12], P < 0.001) and a higher incidence of septic shock (88% vs. 55%, P < 0.001) compared with patients with normal weight, whereas no other between-group differences were observed for the other parameters.
Table 1 -
Patients characteristics
|
All patients (n = 167) |
Normal weight (n = 67) |
Overweight (n = 56) |
Obese (n = 42) |
P* |
| Age (yr) |
75 (65–85) |
82 (70–88) |
75 (62–81) |
67 (61–78) |
<0.001
|
| Sex |
|
|
|
|
|
| Male |
96 (57) |
35 (52) |
16 (29) |
19 (45) |
0.02
|
| Female |
71 (43) |
32 (48) |
40 (71) |
23 (55) |
|
| BMI (kg/m2) |
26.2 (22.5–30.1) |
22.0 (20.8–23.1) |
27.7 (26.1–29.0) |
33.1 (31.1–38.2) |
<0.001
|
| Comorbidities |
|
|
|
|
|
| Diabetes mellitus |
49 (29) |
14 (21) |
17 (30) |
17 (41) |
0.09 |
| Heart failure |
44 (26) |
17 (25) |
18 (32) |
7 (17) |
0.22 |
| Coronary heart disease |
35 (21) |
7 (10) |
16 (29) |
11 (26) |
0.03
|
| Chronic renal disease |
18 (11) |
7 (10) |
6 (11) |
5 (12) |
0.97 |
| Chronic obstructive pulmonary disease |
40 (24) |
13 (19) |
17 (30) |
10 (24) |
0.37 |
| History of malignancy |
13 (8) |
5 (8) |
3 (5) |
5 (12) |
0.49 |
| Source of infection |
|
|
|
|
|
| CAP |
86 (51) |
35 (52) |
28 (50) |
21 (50) |
0.96 |
| HAP |
41 (25) |
18 (27) |
22 (20) |
12 (29) |
0.53 |
| VAP |
16 (10) |
5 (8) |
6 (11) |
5 (12) |
0.71 |
| Acute cholangitis |
10 (6) |
3 (5) |
5 (9) |
2 (5) |
0.54 |
| Primary bacteremia |
14 (8) |
7 (10) |
5 (9) |
2 (5) |
0.58 |
| Laboratory values |
|
|
|
|
|
| Leukocyte count (×109/L) |
14.2 (10.0–19.7) |
14.4 (10.1–19.7) |
14.3 (9.8–20.5) |
14.0 (9.9–19.7) |
0.89 |
| Ferritin (ng/mL) |
1,092 (363–3,511) |
843 (348–2,571) |
1,032 (346–4,721) |
1,697 (556–4,462) |
0.41 |
| mHLA-DR (Ab/cell) |
3,990 (2,546–7,099) |
3,778 (2,445–6,778) |
4,809 (2,539–7,232) |
4,470 (2,610–7,599) |
0.71 |
| APACHE II score |
23 (17–31) |
22 (15–28) |
22 (18–32) |
26 (18–32) |
0.16 |
| SOFA score |
11 (8–14) |
10 (6–12) |
11 (7–15) |
12 (10–15) |
0.002
|
| Septic shock |
116 (69) |
37 (55) |
40 (71) |
37 (88) |
0.001
|
| 28-d mortality |
96 (57) |
38 (57) |
31 (55) |
25 (60) |
0.92 |
Bold values represent statistical significance.
Data are presented as median (interquartile range) or n (%). Normal weight is defined as BMI of 18.5 to 24.9 kg/m2, overweight as BMI of 25.0 to 29.9 kg/m2, and obesity as BMI ≥30 kg/m2. Underweight patients (BMI <18.4; n = 2) are not separately presented.
*Normal weight versus overweight versus obese by Kruskal-Wallis or χ2 tests.
APACHE, Acute Physiology and Chronic Health Evaluation; BMI, body mass index; CAP, community-acquired pneumonia; HAP, health care–associated pneumonia; SOFA, Sequential Organ Failure Assessment; VAP, ventilator-associated pneumonia.
Influence of BMI on adipocytokines and inflammatory parameters
Leptin plasma concentrations were 2.3-fold higher in patients with obesity compared with patients with normal weight and overweight (P = 0.008 and P = 0.02, respectively; Fig. 1A), whereas no significant differences were observed for adiponectin and resistin plasma concentrations (P = 0.08 and P = 0.85, respectively; Fig. 1, B and C). This was also reflected in a positive correlation between leptin and BMI (r = 0.25, P = 0.005), whereas adiponectin and resistin did not significantly correlate with BMI (r = −0.12, P = 0.55, and r = 0.01, P > 0.99, respectively; Supplementary Fig. 2, A–C; https://links.lww.com/SHK/B589). The inflammatory markers ferritin and mHLA-DR expression were not significantly different between patients with normal weight, overweight, and obesity (P = 0.41 and P = 0.71, respectively; Fig. 1, D and E). In accordance, BMI was not correlated with ferritin concentrations and mHLA-DR expression (r = 0.07, P > 0.99, and r = 0.09, P > 0.99, respectively; Supplemental Fig. 2, D and E; https://links.lww.com/SHK/B589).
;)
Fig. 1: Circulating adipocytokine concentrations and inflammatory markers at sepsis diagnosis in patients with and without obesity. Violin plots with plasma concentrations of (A) leptin, (B) adiponectin, (C) resistin, (D) ferritin, and (E) HLA-DR expression on monocytes in sepsis patients with normal weight (n = 67), overweight (n = 56), and obesity (n = 42). Normal weight was defined as BMI of 18.5 to 24.9 kg/m2, overweight was defined as BMI or 25.0 to 29.9 kg/m2, and obesity was defined as BMI ≥30 kg/m2. Data are presented as median with interquartile range. The P values in the panels were calculated with Kruskal-Wallis tests. Dunn multiple comparisons test revealed the following P values for leptin: normal weight versus overweight, P > 0.99; normal weight versus obesity, P = 0.008; and overweight versus obesity, P = 0.02.
Association between immunological endotypes, inflammatory parameters, and adipocytokines
To assess the relationships between adipocytokines and immunological endotypes, the cohort was divided according to immunological endotype (HI [n = 40], UC [n = 55], and IP [n = 62]) based on ferritin concentrations and mHLA-DR expression following the design of the original study (19). The circulating concentrations of the adipocytokines leptin and adiponectin were not significantly different between the immunological endotypes (P = 0.25 and P = 0.16, respectively; Fig. 2, A and B). Furthermore, no correlations of adiponectin with ferritin concentrations and mHLA-DR expression were observed (r = −0.11, P = 0.90, and r = 0.05, P > 0.99, respectively), whereas leptin tended to correlate with ferritin concentrations and mHLA-DR expression, although this did not reach statistical significance (r = −0.19, P = 0.08, and r = 0.21, P = 0.053, respectively; Supplementary Fig. 3, https://links.lww.com/SHK/B589). In contrast, HI patients displayed higher resistin concentrations compared with IP and UC patients (P = 0.009 and P < 0.001, respectively; Fig. 2C). Of note, IP patients also displayed significantly higher concentrations of resistin compared with patients in the UC group (P = 0.013). This association was also reflected by the positive correlation between resistin and ferritin concentrations (r = 0.36, P < 0.001; Fig. 2D) and the negative correlation between resistin concentrations and mHLA-DR expression (r = −0.32, P = 0.003; Fig. 2E).
Fig. 2: Circulating adipocytokine concentrations at sepsis diagnosis in patients with hyperinflammation, unclassified inflammatory state, and immunoparalysis. Violin plots with plasma concentrations of (A) leptin, (B) adiponectin, and (C) resistin in HI (n = 40), UC (n = 55), and IP (n = 62) sepsis patients. Data are presented as median with interquartile range. The P values in the panels were calculated with Kruskal-Wallis tests. Dunn multiple comparisons test revealed the following P values for resistin: HI versus IP, P = 0.009; HI versus UC, P < 0.001; and IP versus UC, P = 0.013. Scatterplots of the correlations between resistin and (D) ferritin and (E) mHLA-DR expression. Correlation coefficients and P values are presented in the panels and were calculated with Spearman rank correlation tests with Bonferroni correction for multiple testing. HI indicates hyperinflammation; IP, immunoparalysis; UC, unclassified.
Moreover, we performed correlation analyses of resistin concentrations with white blood cell counts and ratios (Supplementary Fig. 4, https://links.lww.com/SHK/B589). Resistin concentrations correlated positively with neutrophil percentages (r = 0.29, P < 0.001) and negatively with lymphocyte percentages (r = −0.26, P < 0.001), whereas the correlation with absolute neutrophil counts seemed less pronounced (r = 0.21, P = 0.03) and no significant correlation with absolute lymphocyte counts was present (r = −0.14, P = 0.28). Furthermore, a significant negative correlation between resistin concentrations and the neutrophil-to-lymphocyte ratio was observed (r = −0.30, P < 0.001).
Clinical outcome
SOFArank scores were calculated to identify patients who clinically improved or deteriorated during the first week after sepsis diagnosis. A negative score indicated improvement based on a decline in the SOFA score, and a positive score indicated deterioration based on an increase in the SOFA score relative to diagnosis. Patients with a stable SOFArank score (n = 11) were excluded from this analysis to maintain a dichotomous outcome variable. Leptin and adiponectin plasma concentrations at diagnosis were not different between patients who improved and those who deteriorated (P = 0.65 and P = 0.75, respectively; Fig. 3, A and B). Significant higher plasma concentrations of resistin were observed in patients who deteriorated compared with patients who improved (P < 0.001; Fig. 3C). Moreover, Spearman rank correlation analyses yielded similar findings, showing a positive correlation between resistin and the SOFArank score (r = 0.25, P = 0.001), whereas leptin and adiponectin did not correlate with the SOFArank score (r = −0.092, P = 0.24, and r = −0.010, P = 0.90, respectively).
Fig. 3: Circulating adipocytokine concentrations at sepsis diagnosis in patients that improved and deteriorated in the first week of sepsis. Violin plots with plasma concentrations of (A) leptin, (B) adiponectin, and (C) resistin in patients with bacterial sepsis that improved (n = 77) and deteriorated (n = 79) based on the SOFArank score until day 7. A negative score indicated a decline in the SOFA score and was indicated as improvement. A positive score indicated an increase in the SOFA score and was indicated as deterioration. Data are presented as median with interquartile range. The P values in the panels were calculated with Mann-Whitney U tests. SOFA indicates Sequential Organ Failure Assessment.
No significant difference in 28-day mortality was observed for patients with normal weight, overweight, or obesity (57%, 55%, and 60%, respectively; P = 0.92). The mortality rate of patients with HI was 80% (n = 32), which was significantly higher compared with patients in the IP group (55% [n = 34], P = 0.011) and the UC group (40% [n = 22], P < 0.001). Receiver operating characteristic analyses for mortality based on adipocytokine concentrations indicated that leptin and adiponectin did not significantly discriminate between survivors and nonsurvivors (AUC of 0.56 [95% confidence interval {CI}, 0.47–0.65; P = 0.16] and AUC of 0.51 [95% CI, 0.42–0.60; P = 0.87], respectively; Fig. 4, A and B). Resistin had discriminatory power for mortality (AUC, 0.68 [95% CI, 0.59–0.76]; P < 0.001; Fig. 4C). Subsequently, survival analyses based on the optimal cutoff of 84 ng/mL indicated an increased risk for 28-day mortality in patients with high resistin concentrations (hazard ratio, 2.11 [95% CI, 1.42–3.16]; P < 0.001; Fig. 4D). Logistic regression analyses further confirmed that higher resistin concentrations are associated with increased risk of death at day 28 (adjusted odds ratio, 1.04 per 10 ng/mL [95% CI, 1.00–1.07]; P = 0.04; Table 2), independent of age, BMI, and APACHE II score.
Fig. 4: Adipocytokines and 28-day mortality in sepsis. ROC curves based on 28-day mortality for (A) leptin, (B) adiponectin, and (C) resistin plasma concentrations. D, Survival analysis for high versus low concentrations of resistin. The optimal cutoff value was based on the maximal Youden J index derived from the ROC curve. The HR was calculated with the log-rank (Mantel-Cox) test. AUC indicates area under the curve; HR, hazard ratio; ROC, receiver operating characteristic.
Table 2 -
Risk factors for 28-day mortality in sepsis patients
|
Unadjusted OR |
95% CI |
P
|
Adjusted OR |
95% CI |
P
|
| Age (yr) |
1.02 |
1.00–1.05 |
0.051 |
|
|
|
| BMI (kg/m2) |
1.02 |
0.97–1.07 |
0.49 |
|
|
|
| APACHE II score |
1.09 |
1.05–1.14 |
<0.001
|
|
|
|
| Leptin (per 10 ng/mL) |
0.93 |
0.82–1.05 |
0.25 |
|
|
|
| Adiponectin (per μg/mL) |
1.01 |
0.92–1.12 |
0.79 |
|
|
|
| Resistin (per 10 ng/mL) |
1.05 |
1.02–1.08 |
0.002
|
1.04 |
1.00–1.07 |
0.04
|
Bold values represent statistical significance.
Odds ratios were calculated using binary logistic regression. The adjusted odds ratio for resistin was calculated using age, BMI, and APACHE II as covariates.
BMI, body mass index; CI, confidence interval; OR, odds ratio.
DISCUSSION
This study argues against a direct association between obesity and the inflammatory response in sepsis. Moreover, our results imply that the roles of the different adipocytokines in this process are diverse. On the one hand, leptin plasma concentrations were higher in patients with obesity, but these were not associated with immunological endotype and clinical outcome. On the other hand, resistin concentrations are related to HI, as well as to clinical deterioration and mortality, but were not different between patients with and without obesity. These results argue against a role of a different inflammatory response in general and adipocytokines in particular, to explain the impact of obesity on sepsis. In line with this, adipocytokines are unlikely to contribute to clinical outcomes in sepsis patients with obesity.
New interest in the inflammatory effects of adipocytokines in infectious diseases has emerged with the coronavirus disease 2019 (COVID-19) pandemic, in which obesity was found to be a risk factor for developing severe disease (21). The findings reported here in sepsis patients are similar to what we and others observed in patients with COVID-19, where an association neither between BMI and inflammation (22,23) nor between BMI-related adipocytokines, inflammatory parameters, and outcome was observed (de Nooijer et al., submitted for publication) (24). This is an interesting observation because the inverse association between BMI and mortality in patients with bacterial sepsis was not present in COVID-19 patients (25). Moreover, although obesity is associated with chronic low-grade inflammation, this might be overwhelmed by the excessive inflammatory response in severe infections (13,26). This further argues that inflammation and adipocytokines do not play a major role in the relation of obesity and mortality in severe infections.
Of note, we observed that leptin concentrations tended to relate to the inflammatory response through a negative correlation with ferritin concentrations and a positive correlation with mHLA-DR expression, both with borderline significance. Although leptin concentrations are associated with BMI, both leptin and BMI are not associated with clinical outcomes in this study. Therefore, we consider it unlikely that the correlation of leptin with the inflammatory markers is clinically relevant.
In contrast, resistin is an adipocytokine that, independent from BMI, is related to inflammation and disease severity. Resistin was first described in 2001 as a hormone secreted by adipocytes in mice linking obesity to insulin resistance (27). However, in humans, resistin is predominantly produced by macrophages. Results from a “humanized resistin mouse” model, in which resistin was only produced by macrophages, indicated that human resistin leads to insulin resistance because of its proinflammatory properties (28). Resistin transcription was shown to be increased by proinflammatory cytokines (29), whereas resistin itself stimulates the secretion of proinflammatory cytokines in vitro (15). In previous studies with sepsis patients, resistin concentrations were positively correlated and formed a strong network with proinflammatory cytokines (16,30). Furthermore, associations of resistin with organ failure and mortality were observed (16). Our study further underlines the relationship of resistin with the proinflammatory response and clinical outcome in sepsis. Moreover, we demonstrate that circulating resistin concentrations are associated with an increased mortality risk that remained significant after correcting for age, BMI, and disease severity, indicating that it might serve as a prognostic biomarker. Importantly, resistin concentrations were not related to BMI, again illustrating the notion that BMI is not the driver of adipocytokine-mediated effects on clinical outcome.
Strikingly, in our study, resistin was not only associated with the hyperinflammatory endotype but also with the immunoparalytic endotype. A possible explanation for this observation could be that patients with sepsis-induced IP also have a concurrent hyperinflammatory profile (3). Another possible explanation could be that resistin itself also has inhibitory effects on immune function. In vitro studies have described that resistin reduced the production of reactive oxygen species and therefore inhibits bacterial killing by neutrophils (31,32). This suggests that high resistin concentrations are associated with a more complex immune dysregulation rather than only HI. In vivo studies with the humanized resistin mouse model revealed conflicting effects of human resistin on inflammation and survival. One study indicated a protective role of resistin in LPS-induced sepsis by inhibition of the Toll-like receptor 4 pathway resulting in a switch from proinflammatory to anti-inflammatory responses (33). In contrast, another study with LPS-induced sepsis showed an increased proinflammatory activation of neutrophils and more severe pulmonary disease in the presence of resistin (34), whereas a recent study observed no effects of resistin on inflammation and survival after sepsis induced by cecal ligation and puncture (35). Therefore, the exact role of resistin in the immune dysregulation in sepsis and whether the association with mortality is due to hyperinflammatory or immunosuppressive effects of resistin remain unclear.
The present study has several limitations. First, this study is a secondary analysis of a clinical trial using its baseline data and samples. Although the analyzed data in the current study were collected before any trial-related intervention, the preselection resulted in a specific cohort of sepsis patients. The most important distinctive characteristics of the current cohort are the large proportion of patients with pneumonia and the exclusion of patients using immunosuppressive medications (19). Consequently, this preselected population resulted in a more homogeneous population of sepsis patients and reduced potential immune-related confounders. Moreover, because this study is a secondary analysis, the sample size was not based on a power calculation for the end points investigated in the current work. Second, all data are observational and based on measurements within the first 24 h after sepsis diagnosis, limiting conclusions on causality. In the future, longitudinal measurements of adipocytokines and inflammatory parameters could provide further insight into possible effects of changing adipocytokine concentrations on disease course. Third, mortality was very high in our sepsis cohort (57%), whereas the World Health Organization reports a hospital mortality rate between 27% and 42% for sepsis patients treated in intensive care units (36). Therefore, caution is needed when extrapolating these data to other sepsis patient populations. Fourth, we did not observe a relationship between BMI and mortality and therefore did not confirm the presence of the obesity paradox in this study population, which is likely due to a relatively modest sample size and thus limited statistical power to show this association. Fifth, because of a relatively limited sample size, it was not possible to divide the cohort into different obesity classes. Therefore, potential effects of severe obesity could have been missed.
CONCLUSION
In this study, we found no association between obesity or BMI-related adipocytokines and several inflammatory markers in patients with bacterial sepsis. However, we identified a potentially important role of resistin in sepsis, which is independent from BMI. High resistin concentrations are associated with the inflammatory response and have predictive value for poor clinical outcome in patients with sepsis, independent of age and disease severity. Taken together, although there exists a strong association between inflammation and sepsis mortality, our results do not point toward a direct role for obesity or BMI-related adipocytokines in immune dysregulation in sepsis patients.
ACKNOWLEDGMENTS
The authors would like to thank the entire PROVIDE study team for conducting the trial. The authors would also like to thank Helga Dijkstra and Heidi Lemmers for their support in the measurement of the adipocytokines.
REFERENCES
1. Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, Bellomo R, Bernard GR, Chiche JD, Coopersmith CM, et al.: The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3).
JAMA 2016;315(8):801–10.
2. van der Poll T, van de Veerdonk FL, Scicluna BP, Netea MG: The immunopathology of sepsis and potential therapeutic targets.
Nat Rev Immunol 17(7):407–420, 2017.
3. Hotchkiss RS, Monneret G, Payen D: Sepsis-induced immunosuppression: from cellular dysfunctions to immunotherapy.
Nat Rev Immunol 13(12):862–874, 2013.
4. Kyriazopoulou E, Leventogiannis K, Norrby-Teglund A, Dimopoulos G, Pantazi A, Orfanos SE, Rovina N, Tsangaris I, Gkavogianni T, Botsa E, et al.: Macrophage activation-like syndrome: an immunological entity associated with rapid progression to death in sepsis.
BMC Med 15(1):172, 2017.
5. Quadrini KJ, Patti-Diaz L, Maghsoudlou J, Cuomo J, Hedrick MN, McCloskey TW: A flow cytometric assay for HLA-DR expression on monocytes validated as a biomarker for enrollment in sepsis clinical trials.
Cytometry B Clin Cytom 100(1):103–114, 2021.
6. Docke WD, Hoflich C, Davis KA, Röttgers K, Meisel C, Kiefer P, Weber SU, Hedwig-Geissing M, Kreuzfelder E, Tschentscher P, et al.: Monitoring temporary immunodepression by flow cytometric measurement of monocytic HLA-DR expression: a multicenter standardized study.
Clin Chem 51(12):2341–2347, 2005.
7. Zhi G, Xin W, Ying W, Guohong X, Shuying L: “Obesity paradox” in acute respiratory distress syndrome: asystematic review and meta-analysis.
PLoS One 11(9):e0163677, 2016.
8. Nie W, Zhang Y, Jee SH, Jung KJ, Li B, Xiu Q: Obesity survival paradox in pneumonia: a meta-analysis.
BMC Med 12:61, 2014.
9. Pepper DJ, Sun J, Welsh J, Cui X, Suffredini AF, Eichacker PQ: Increased body mass index and adjusted mortality in ICU patients with sepsis or septic shock: a systematic review and meta-analysis.
Crit Care 20(1):181, 2016.
10. La Cava A: Leptin in inflammation and autoimmunity.
Cytokine 98:51–58, 2017.
11. Zelechowska P, Kozlowska E, Pastwinska J, Agier J, Brzezinska-Blaszczyk E: Adipocytokine involvement in innate immune mechanisms.
J Interferon Cytokine Res 38(12):527–538, 2018.
12. Ouchi N, Walsh K: Adiponectin as an anti-inflammatory factor.
Clin Chim Acta 380(1–2):24–30, 2007.
13. Khanna D, Khanna S, Khanna P, Kahar P, Patel BM: Obesity: a chronic low-grade inflammation and its markers.
Cureus 14(2):e22711, 2022.
14. Hsu WY, Chao YW, Tsai YL, Lien CC, Chang CF, Deng MC, Ho LT, Kwok CF, Juan CC: Resistin induces monocyte-endothelial cell adhesion by increasing ICAM-1 and VCAM-1 expression in endothelial cells via p38MAPK-dependent pathway.
J Cell Physiol 226(8):2181–2188, 2011.
15. Bokarewa M, Nagaev I, Dahlberg L, Smith U, Tarkowski A: Resistin, an adipokine with potent proinflammatory properties.
J Immunol 174(9):5789–5795, 2005.
16. Koch A, Gressner OA, Sanson E, Tacke F, Trautwein C: Serum resistin levels in critically ill patients are associated with inflammation, organ dysfunction and metabolism and may predict survival of non-septic patients.
Crit Care 13(3):R95, 2009.
17. Chen M, Wang B, Xu Y, Deng Z, Xue H, Wang L, He L: Diagnostic value of serum leptin and a promising novel diagnostic model for sepsis.
Exp Ther Med 7(4):881–886, 2014.
18. Venkatesh B, Hickman I, Nisbet J, Cohen J, Prins J: Changes in serum adiponectin concentrations in critical illness: a preliminary investigation.
Crit Care 13(4):R105, 2009.
19. Leventogiannis K, Kyriazopoulou E, Antonakos N, Kotsaki A, Tsangaris I, Markopoulou D, Grondman I, Rovina N, Theodorou V, Antoniadou E, et al.: Toward personalized immunotherapy in sepsis: the PROVIDE randomized clinical trial.
Cell Rep Med 3(11):100817, 2022.
20. Gelissen H, de Grooth HJ, Smulders Y, Wils EJ, de Ruijter W, Vink R, Smit B, Röttgering J, Atmowihardjo L, Girbes A, et al.: Effect of low-normal vs high-normal oxygenation targets on organ dysfunction in critically ill patients: a randomized clinical trial.
JAMA 326(10):940–948, 2021.
21. Huang Y, Lu Y, Huang YM, Wang M, Ling W, Sui Y, Zhao HL: Obesity in patients with COVID-19: a systematic review and meta-analysis.
Metabolism 113:154378, 2020.
22. Kooistra EJ, de Nooijer AH, Claassen WJ, Grondman I, Janssen NAF, Netea MG, van de Veerdonk FL, van der Hoeven JG, Kox M, Pickkers P, et al.: A higher BMI is not associated with a different immune response and disease course in critically ill COVID-19 patients.
Int J Obes (Lond) 45(3):687–694, 2021.
23. Wolf M, Alladina J, Navarrete-Welton A, et al.: Obesity and critical illness in COVID-19: respiratory pathophysiology.
Obesity (Silver Spring) 29(5):870–878, 2021.
24. de Nooijer AH, Kooistra EJ, Grondman I, Janssen NAF, Joosten LAB, van de Veerdonk FL, et al.: Adipocytokine plasma concentrations reflect influence of inflammation but not body mass index (BMI) on clinical outcomes of COVID-19 patients: A prospective observational study from the Netherlands.
Clin Obes e12568, 2022.
25. Kooistra EJ, Brinkman S, van der Voort PHJ, de Keizer NF, Dongelmans DA, Kox M, Pickkers P: Body mass index and mortality in coronavirus disease 2019 and other diseases: a cohort study in 35,506 ICU patients.
Crit Care Med 50:e1–e10, 2021.
26. Kolyva AS, Zolota V, Mpatsoulis D, et al.: The role of obesity in the immune response during sepsis.
Nutr Diabetes 4:e137, 2014.
27. Steppan CM, Bailey ST, Bhat S, et al.: The hormone resistin links obesity to diabetes.
Nature 409(6818):307–312, 2001.
28. Qatanani M, Szwergold NR, Greaves DR, Ahima RS, Lazar MA: Macrophage-derived human resistin exacerbates adipose tissue inflammation and insulin resistance in mice.
J Clin Invest 119(3):531–539, 2009.
29. Kaser S, Kaser A, Sandhofer A, Ebenbichler CF, Tilg H, Patsch JR: Resistin messenger-RNA expression is increased by proinflammatory cytokines
in vitro.
Biochem Biophys Res Commun 309(2):286–290, 2003.
30. Ebihara T, Matsumoto H, Matsubara T, Matsuura H, Hirose T, Shimizu K, Ogura H, Kang S, Tanaka T, Shimazu T: Adipocytokine profile reveals resistin forming a prognostic-related cytokine network in the acute phase of sepsis.
Shock 56(5):718–726, 2021.
31. Cohen G, Ilic D, Raupachova J, Horl WH: Resistin inhibits essential functions of polymorphonuclear leukocytes.
J Immunol 181(6):3761–3768, 2008.
32. Miller L, Singbartl K, Chroneos ZC, Ruiz-Velasco V, Lang CH, Bonavia A: Resistin directly inhibits bacterial killing in neutrophils.
Intensive Care Med Exp 7(1):30, 2019.
33. Jang JC, Li J, Gambini L, Batugedara HM, Sati S, Lazar MA, Fan L, Pellecchia M, Nair MG: Human resistin protects against endotoxic shock by blocking LPS-TLR4 interaction.
Proc Natl Acad Sci U S A 114(48):E10399–E10408, 2017.
34. Jiang S, Park DW, Tadie JM, Gregoire M, Deshane J, Pittet JF, Abraham E, Zmijewski JW: Human resistin promotes neutrophil proinflammatory activation and neutrophil extracellular trap formation and increases severity of acute lung injury.
J Immunol 192(10):4795–4803, 2014.
35. Bonavia AS, Chroneos ZC, Ruiz-Velasco V, Lang CH: Resistin production does not affect outcomes in a mouse model of acute surgical sepsis.
PLoS One 17(3):e0265241, 2022.
36. Global Report on the Epidemiology and Burden of Sepsis: Current Evidence, Identifying Gaps and Future Directions. Geneva, Switzerland: World Health Organization, 2020. Licence: CC BY-NC-SA 3.0 IGO.
