Neutrophil reactive oxygen formation, bacterial infections and mortality in malnourished hemodialysis patients: Evaluation of clinical outcomes

Introduction: Patients with end stage kidney disease undergoing maintenance hemodialysis (MHD) are prone to malnutrition and infections. Objective: The objective of this study was to evaluate the effect of polymorphonuclear (PMN) cell dysfunction on clinical outcomes of MHD patients, in association with nutritional status. Methods: This prospective study investigated 39 MHD patients by evaluating the oxidative activity of their PMN cells using Phorbol 12-Myristate-13-Acetate (PMA) stimulation. Blood samples were taken from each participant at dialysis initiation. Demographics, laboratory data, and clinical outcomes during a 24-month follow-up period were obtained from electronic medical records. Results: Phagocytic activity was described in percentiles of mean fluorescence intensity (MFI) of PMA levels. There were no differences in comorbidities between patients with low or high MFI-PMA percentiles. Patients in the lowest (25th) MFI-PMA percentile ( N = 10) had poorer nutritional status and


| INTRODUCTION
Patients with end stage kidney disease (ESKD) undergoing maintenance hemodialysis (MHD) are highly susceptible to severe infections and related complications. Due to their weakened immune system, infections are a leading cause of morbidity among MHD patients and represent a significant contributor to mortality in this population, second only to cardiovascular events. 1 A combination of several factors, including uremic toxins, malnutrition, medications, anemia, and hemodialysis itself, contribute to alterations in both innate and adaptive immunity in MHD patients. 2,3 Accumulation of uremic toxins triggers several inflammatory mechanisms, leading to chronic oxidative stress and impaired neutrophil functioning, which are the front-line component in the initial response to bacterial invasion. 1,4,5 Previous studies confirmed that neutrophils are over-activated in hemodialysis patients. 6,7 Reactive oxygen molecules (ROM) are generated and released by these highly activated neutrophils, resulting in high oxidative stress, which contributes fundamentally to the pathogenicity of systemic inflammation. Furthermore, disturbances in neutrophil function in patients undergoing MHD manifest as decreased phagocytic ability compared to healthy people. [8][9][10] The resultant chronic, systemic inflammation, and impaired immune function lead to high morbidity and mortality. [11][12][13] Several inflammatory biomarkers have been identified as potential indicators for the degree of chronic inflammation and as possible predictors of adverse clinical outcomes in MHD patients. Plasma C-reactive protein (CRP), [14][15][16] pentraxin-3 (PTX3), 17 chemokine ligand-16 (CXCL16), 18 and neutrophil-lymphocyte ratio (NLR) 19 are some biomarkers that might predict morbidly and all-cause mortality among hemodialysis patients.
However, most are not clinically useful prognostic factors.
Malnutrition is another negative prognostic factor, with an estimated prevalence of 50% among MHD patients. 20 Previous studies reported correlations between years of dialysis and deterioration in nutritional status. 21,22 Malnutrition among MHD patients is multifactorial, resulting from several metabolic changes associated with ESKD, the uremic milieu, and dialysis itself. 23 It is well-established that poor nutritional status is a risk-factor for reduced immunity, including impaired phagocytosis. [24][25][26] We recently reported an association between high malnutritioninflammation status of MHD patients, 27 which predicts morbidity and mortality, 24,28 and polymorphonuclear (PMN) leukocyte dysfunction. We showed that MHD patients have impaired NADPH activity, as demonstrated by reduced stimulation of PMN by Escherichia coli and PMA (Phorbol 12-Myristate-13-Acetate). Reduced mean fluorescence intensity (MFI) levels among PMA-stimulated PMN were more pronounced in malnourished MHD patients. As a high percentage of MHD patients are malnourished 24,25 and nutritional status is potentially reversible using directed interventions, it is important to characterize the impact of low MFI-PMA levels on the clinical outcomes of MHD patients with high malnutrition-inflammation status. 29 The correlation between malnutrition-inflammation and reduced MFI levels among PMA-stimulated PMN, which was demonstrated in our previous study, encouraged us to further investigate associations between impaired phagocytic function and clinical outcomes in these patients. The objective of the current study was to evaluate the rates of severe infections, hospitalizations, and all-cause mortality among MHD patients with PMN dysfunction, with a focus on the associations between MFI-PMA levels and clinical outcomes.

| METHODS
This prospective study was conducted from January to December 2018 in the dialysis unit of Meir Medical Center, Kfar Saba, Israel, a secondary medical center with 66,000 admissions per year, serving a population of 600,000 individuals.
The study was approved by the local Ethics Committee (approval no. 0207-17-MMC). All participants provided signed informed consent prior to enrollment.

| Study population
Dialysis unit patients >18 years of age, on MHD for at least 3 months prior to enrollment were recruited to the study. They were followed for 2 years, until January 2021.

| Data collection
Electronic medical records were reviewed. Selected demographic and clinical data were obtained, including age, sex, comorbidities, bacterial infection events, hospital admissions, and dates of death.
Blood samples were taken from each participant at dialysis initiation for analysis of oxidative activity in the Laboratory for Leukocyte Functions at our medical center. Baseline laboratory data taken at dialysis initiation as part of routine follow-up of MHD patients were analyzed.

| Oxidative activity
FagoFlow Ex commercial kit (Exbio Diagnostics, Vestec, Czech Republic) was used to determine the phagocytic activity of neutrophils after stimulation with E. coli bacteria and PMA, with some modifications to the kit protocol, as previously described by our group. 27 PMA induces a respiratory burst of PMN by phosphorylating the intracellular enzyme protein kinase C (PKC).
To assess potential alterations in signal transduction of superoxide generation, PMN were stimulated with both E. coli bacteria and PMA. In brief, hydrogen peroxide release by activated PMNs was evaluated using the dihydrorhodamine (DHR) 123 assay. DHR is freely permeable, localized in the mitochondria, and after oxidation by H 2 O 2 to rhodamine 123, emits a bright fluorescent signal upon excitation. This assay determines the phagocytic activity of neutrophils after stimulation with E. coli bacteria, which activate NADPH oxidase. This produces reactive oxidative respiratory burst and positive control with PMA, which activates a respiratory burst of PMN by phosphorylation of the intercellular enzyme PKC. Phagocytic activity was measured using MFI-PMA and MFI E. coli levels. The study groups were categorized based on the neutrophil phagocytic activity of each patient.

| Clinical outcomes measured
Severe infections were defined as those that led to hospitalization.
We also recorded multiple emergent hospitalizations (defined as three or more hospitalizations) and all-cause mortality during 24 months of follow-up.

| Statistical analysis
Variables between study groups were compared using t-test, Mann-Whitney test, Fisher's exact test, or chi-square test, according to the scale of measured variables. The Student's t-test (or Mann-Whitney U-test for nonparametric conditions) was used to compare continuous variables.
Phagocytic activity was described by percentile of MFI-PMA levels.
Patients with the lowest MFI-PMA levels (25th percentile) were compared to patients with higher percentile group. Pearson or Spearman tests were analyzed to assess possible correlations between continuous variables, each when appropriate. A multivariate logistic regression model was applied to estimate odds ratios (ORs) for all-cause mortality in the cohort. Kaplan-Meier curves were used to show survival rates between study groups. p-Values <0.05 were considered statistically significant.

| Patient characteristics
A total of 39 MHD patients were included in the study. Relevant clinical data and baseline laboratory records at study initiation are shown in Table 1. Mean age was 65.6 ± 14.4 years and 24 (62%) were men.
Most patients (82%) were treated using high-flux high-efficiency membranes, with dialysis regimen of 12.5 ± 2.1 h/week. Of the 39 patients, the average MFI-PMA level was 63.7 ± 35.3 (range 1.62-118). Based on these levels, patients were categorized into two groups according to percentiles. The Low PMA group consisted of patients with MFI-PMA levels below the 25th percentile, or <23 (n = 10), while those with MFI-PMA levels above this threshold were included in the high MFI-PMA group, above 23 (n = 29). Figure 1 shows a histogram with a normality curve of MFI-PMA levels.
There were no differences in comorbidities between patients in the 25th percentile group and those with higher percentiles. However, their serum albumin, hemoglobin, total cholesterol, folic acid, and transferrin saturation index (TSI) levels were lower, and their CRP was higher than those with higher percentiles (Table 1).

| Staphylococcus aureus and Gram-negative bacteria infections
Gram-negative bacterial infections occurred in six patients due to catheter-related bloodstream infections, diabetic foot ulcers, or urinary tract infections. We did not observe increased mortality among patients with Gram-negative infections compared to other patients.
No association between Gram-negative bacteria infections and MFI-PMA levels was found (  The mortality rate was significantly higher in patients in the low MFI-PMA group than in the high MFI-PMA group (80% versus 31%, p = 0.007, OR for mortality was 8.85 with 95%CI [1.56-50], Table 2).
Survival curves of patients with low versus high percentile MFI-PMA levels is shown in Figure 2 ( p = 0.002).
In a multivariate analysis, the MFI-PMA percentile and ischemic heart disease were the strongest predictors of mortality (Table 3). We recently reported that phagocyte activity in MHD patients is impaired, especially among malnourished patients. 27 All 44 MHD patients included in that study had reduced MFI-E. coli levels;

| DISCUSSION
T A B L E 1 Clinical characteristics of study population at enrollment. although, malnourished patients demonstrated reduced MFI-PMA levels compared to patients with better nutritional status. 27 It is wellestablished that poor nutritional status among MHD patients is associated with disturbances in the oxidative burst of neutrophils and adverse clinical outcomes. 24,28 However, to the best of our knowledge, the specific impact of nutritional status on MFI-PMA levels has not been previously reported. Furthermore, malnutrition is an independent risk-factor for mortality in MHD patients. 30  Furthermore, our study revealed a positive correlation between MFI-PMA and both serum albumin and creatinine levels, which was in line with previous findings and was anticipated. 32 Serum creatinine concentration has been linked to nutritional status, reflecting somatic protein stores, muscle mass, and dietary protein intake, and has also been shown to be a strong predictor of future mortality in dialysis patients, as demonstrated by the DOPPS study. 32 The chemical agent PMA is frequently used to activate NADPH oxidase of PMN, 33 mostly as part of the molecular diagnosis of chronic granulomatous disease (CGD). 34 Interestingly, the low mean MFI-PMA level (63.7 ± 35.3) found in this cohort of MHD patients is similar to that found among patients with CGD. 35 This supports the finding that MHD patients are considered immunocompromised and therefore, predisposed to severe infections with high mortality rates. 36 The findings of this study demonstrate the potential utility of MFI Our results also showed that low MFI-PMA levels were not associated with a risk for Gram-negative infections.
The mortality rate was significantly higher in patients with the lowest phagocytic activity (the lowest percentile of MFI-PMA levels) versus all others. Multivariate analysis reinforced the importance of impaired phagocyte function as a possible prognostic factor for mortality, by highlighting low MFI-PMA levels and ischemic heart disease as the strongest predictors of mortality.
Correlations between low MFI-PMA levels and recurrent hospitalizations for any cause were found. While assessing hospitalization Previous studies highlighted several biomarkers as potential prognostic factors among MHD patients. 16,17,19,38 The role of CRP, which was higher among patients in the low MFI-PMA group, as a marker of inflammation is well-established. Approximately 30% to 50% of chronic kidney disease patients have noticeably elevated CRP levels. 18 Kawaguchi et al. 15 reported associations of CRP levels with mortality in Japanese hemodialysis patients (hazard ratio of death was 1.6-to 2.4-fold higher for CRP levels >3 mg/L versus <1.0 mg/L, p < 0.05).
Korevaar et al. 14 showed that elevated CRP levels after a hemodialysis session is an independent risk-factor for mortality, as an increase of only 1 mg/L in CRP was associated with a 9% risk for increased mor-

| CONCLUSION AND PRACTICAL APPLICATION
Oxidative activity of neutrophils might serve as a prognostic biomarker of adverse clinical outcomes, such as severe infections and mortality among MHD patients. The all-cause mortality rate of MHD patients with the lowest MFI-PMA levels was 8.8 times higher than that of patients with higher MFI-PMA levels during a 24-month study.

ACKNOWLEDGMENTS
We thank X, MS for editing the manuscript. She is an employee of Meir Medical Center. Informed consent was obtained from all participants prior to study entry. The study was performed in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines.

CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.

ETHICS STATEMENT
The study was approved by the Ethics Committee and Institutional