Abstract
Introduction
The dysregulated host immune response in sepsis is orchestrated by peripheral blood leukocytes. This study explored the associations of the peripheral blood leukocyte subpopulations with early clinical deterioration and mortality in sepsis.
Methods
We performed a prospective observational single-center study enrolling adult subjects with sepsis within 48 h of hospital admission. Peripheral blood flow cytometry was performed for the patients at enrolment and after 5 days. The primary outcome was to explore the association between various leukocyte subpopulations at enrolment and early clinical deterioration [defined as an increase in the sequential organ failure assessment (SOFA) score between enrolment and day 5, or death before day 5]. Other pre-specified outcomes explored associations of leukocyte subpopulations at enrolment and on day 5 with in-hospital mortality.
Results
A total of 100 patients, including 47 with septic shock were enrolled. The mean (SD) age of the patients was 53.99 (14.93) years. Among them, 26 patients had early clinical deterioration, whereas 41 died during hospitalization. There was no significant association between the leukocyte subpopulations at enrolment and early clinical deterioration on day 5. On multivariate logistic regression, a reduced percentage of CD8+CD25+ T-cells at enrolment was associated with in-hospital mortality [odds ratio (OR), 0.82 (0.70-0.97); p-value = 0.02]. A reduced lymphocyte percentage on day 5 was associated with in-hospital mortality [OR, 0.28 (0.11-0.69); p-value = 0.01]. In a post-hoc analysis, patients with “very early” deterioration within 48 h had an increased granulocyte CD64 median fluorescent intensity (MFI) [OR, 1.07 (1.01-1.14); p-value = 0.02] and a reduced granulocyte CD16 MFI [OR, 0.97 (0.95-1.00); p-value = 0.04] at enrolment.
Conclusions
None of the leukocyte subpopulations showed an association with early clinical deterioration at day 5. Impaired lymphocyte activation and lymphocytopenia indicative of adaptive immune dysfunction may be associated with in-hospital mortality.
Keywords: sepsis, flow cytometry, prognosis, leukocyte subpopulations
Introduction
Sepsis is a syndrome characterized by dysregulated host immune response to infection which may lead to organ dysfunction. It is a common cause of intensive care unit (ICU) admission globally with increasing incidence due to aging and immunosuppression.1,2 It is associated with significant morbidity and mortality. Patients with septic shock have in-hospital mortality rates exceeding 50%.3
Over the past few decades, our understanding of sepsis immunopathogenesis has evolved rapidly. Earlier accounts described sepsis as a hyperinflammatory state, known as the systemic inflammatory response syndrome (SIRS) secondary to infection.4 However, further data suggested that sepsis involves dysregulated pro-inflammatory and anti-inflammatory responses, the latter being referred to as the compensatory anti-inflammatory response syndrome (CARS).5 The current definition of sepsis (sepsis-3) has replaced SIRS with the term “dysregulated host response to infection”.6
These immune responses in sepsis are orchestrated by the cells of the innate and adaptive immune system and the mediators released by them. Flow cytometry (FCM) enables us to quantify the subpopulations of leukocytes involved in this immune response and to check their activation markers. It provides us with an invaluable means for profiling the immune status of patients with sepsis and thereby has potential applications in diagnosis, prognosis, and tailoring therapies for personalized medicine.7
The use of flow cytometry for early prognostication of sepsis may be of immense use for triage and appropriate management of these patients. Guerin et al and Daix et al have reported that an increase in immature granulocytes with low expressions of CD10 and CD16 at admission may be associated with early deterioration within 48 h in sepsis.8,9 On the other hand, Shankar-Hari et al have found that among patients who attend the emergency for a suspected infection, a combination of increased neutrophil CD24, increased neutrophil PD-1, and decreased monocyte HLA-DR may fore-shadow the development of sepsis within 72 h.10 Notably, while the HLA-DR levels on monocytes have been studied for sepsis prognostication, HLA-DR levels on lymphocytes have not. Both HLA-DR and CD25 have been recognized as activation markers on T-lymphocytes which have an important role in the adaptive immune response against infection.11,12
As summarized above, only a few studies have been performed in developed countries on the role of flow cytometry for early prognostication of sepsis. Hence, we performed this study among patients admitted with sepsis in a developing country to explore the association of a large number of peripheral blood leukocyte subpopulations with early clinical deterioration and in-hospital mortality. We also studied novel markers including the T-lymphocyte activation markers, namely HLA-DR and CD25 in this setting.
Methodology
Study Design and Settings
We conducted this prospective, observational study in the Pulmonary, Critical Care & Sleep Medicine department at a tertiary care teaching hospital in North India between March 2018 and September 2019. Informed consent was obtained from the subjects or the next of kin (if the patient was unable to give consent). The study was approved by the Institute Ethics Committee of the All India Institute of Medical Sciences, New Delhi (IECPG-413/28.09.2017, RT-44/21.03.2018).
Participants
All consecutive patients aged 18 years or above admitted to the respiratory ICU with a diagnosis of sepsis were screened for inclusion within 48 h of presentation to the hospital. Sepsis was defined as per the Sepsis-3 definition,6 ie, the presence of suspected or documented infection with organ dysfunction determined by an acute increase of sequential organ failure assessment (SOFA) score by 2 or more points.6 Pregnant women, and those suffering from malignancy, chronic renal or liver failure were excluded. Patients with immunosuppression (human immunodeficiency virus infection, steroid use ≥ 5 mg prednisolone equivalent per day, immunosuppressant drug use, and solid organ or hematopoietic transplantation) were excluded. Patients who were transferred from another hospital were also excluded. Apparently healthy volunteers aged 18 years or above who were attending the out-patient department for routine check-ups and were not suffering from ongoing infectious, neoplastic, or immunological diseases were recruited as controls.
Study Flow
Included participants had their demographic characteristics and comorbid illnesses recorded at enrolment. For the patients, the SOFA score was calculated at the time of enrolment and then daily for 5 days. All patients were investigated and treated as per the standard ICU management protocols. Microbiological cultures of blood, urine and respiratory (sputum or endotracheal aspirate) specimens were sent at the time of ICU admission. The clinical team managing the patients was blinded to the results of the flow cytometric assessment. Patients were followed up till hospital discharge or death for in-hospital outcomes. Hospital-acquired infections (HAIs) were diagnosed by a combination of clinical features and microbiological cultures using standard definitions.
Blood Sampling and Immunophenotyping of Leukocytes Subsets
Peripheral venous blood samples (2 mL) were drawn in an EDTA vial for immunophenotyping of leukocytes at the time of enrolment for controls. The sample was drawn twice from the patients; first at the time of enrolment and second 5 days later. The samples were processed within 6 h by the standard stain-lyse-wash method. The respective combinations of anti-bodies were added to 1 × 106 cells and incubated for 20 min in a dark room at ambient temperature. The red cells were lysed with an ammonium chloride-based solution (in-house preparation) and were washed twice with phosphate-buffered saline (PBS). The cells were then suspended in 500 μL of PBS solution. For flow cytometric analysis, at least 20,000 events were acquired, and the data was stored as list mode files. For all specimens, multiparameter FCM was performed on the Coulter Gallios instrument [Beckman Coulter (BC), Hialeah, FL, USA]. The antibody panel used was:
Tube 1: CD3FITC/X/CD19ECD/CD16 +CD56PC5/CD45PC7
Tube 2: CD25FITC/HLA-DRPE/CD3ECD/CD8PC5/CD4PC7
Tube 3: CD45FITC/CD24PE//CD10ECD/CD64PC5/CD16PC7
The gating strategy used for the identification of leukocyte sub-populations and activation markers is illustrated in Supplementary Figure 1. The results of FCM were reported as percentages of lymphocytes, T cells, B cells, CD4+cells, CD8+cells, and NK cells. The percentages of CD4+ and CD8+cells bearing the activation factors HLA-DR and CD25 were recorded. Finally, the granulocyte expressions in median fluorescent intensity (MFI) of CD64, CD24, CD16, and CD10 were noted. While CD64 is an activation marker on granulocytes which is raised in infection and sepsis, the markers CD16 and CD10 are expressed on mature granulocytes but not on band forms.13,14
Pre-Specified Outcomes
We first assessed the difference in the baseline peripheral blood leukocyte subpopulations between the control group, patients with sepsis, and patients with septic shock. The primary outcome of the study was to explore an association between the expression of peripheral blood leukocyte subsets at admission and early clinical outcome (at day 5) among patients with sepsis. For assessing the early clinical outcome, the SOFA score was calculated on day 5 of enrolment and was compared with that at baseline. An increase in the SOFA score on day 5 or mortality before day 5 was defined as “early deterioration”, whereas a decrease or no change in the score was denoted as “improvement/stability”. The other pre-specified outcomes were to explore the associations between the leukocyte subpopulations at enrolment or on day 5 after enrolment and the in-hospital mortality.
Post-hoc Outcomes
A post-hoc analysis assessed the association between the incidence of hospital-acquired infection (HAI) till day 5 of enrolment and the early clinical outcome at day 5. Another post-hoc analysis explored the associations between leukocyte subpopulations at enrolment and “very early” clinical deterioration (ie, within 48 h of enrolment). We also performed post-hoc subgroup analyses of the associations of leukocyte subpopulations at enrolment with early clinical outcome at day 5 and in-hospital mortality in patients with suspected or confirmed bacterial sepsis. Herein, culture negative sepsis due to community-acquired pneumonia was considered as a suspected bacterial infection.
Statistical Analysis
Statistical analysis was performed using STATA (version 12) software. Categorical variables were expressed as numbers and percentages and continuous variables were expressed as mean and standard deviation (SD) or median and interquartile range (IQR) as appropriate. For the comparison of leukocyte subpopulations and activation markers between control, sepsis, and septic shock groups, the Kruskal–Wallis test was used. Intergroup comparisons between any two of these groups were made using the Mann–Whitney U-test. To assess the association between the early clinical outcome and hospital-acquired infections on day 5, we used the chi-squared test. To explore the associations between the peripheral blood leukocyte subsets and the clinical outcomes, we performed univariate logistic regression. For those parameters which had a p-value of less than 0.10 on univariate analysis, we performed multivariate logistic regression. A p-value of less than 0.05 was considered statistically significant. As this was an exploratory study of various peripheral blood leukocyte subpopulations, a sample size of 100 was chosen on pragmatic grounds and not based on statistical calculation. Furthermore, due to the limited sample size, we did not attempt to derive a prediction model from the study.
Results
A total of 188 patients with suspected sepsis were screened during the study period, out of which 100 patients with a mean (SD) age of 53.99 (14.93) years were included in the final analysis (Figure 1). During the same period, a total of 48 healthy controls were also recruited. The baseline characteristics of the controls and the patients along with the frequency of various outcomes are presented in Table 1. A total of 88 patients had suspected or confirmed bacterial sepsis, including 73 culture negative and 15 culture positive (respiratory specimen in 12 and blood culture in 3) subjects. The final microbiological profile of the enrolled subjects is listed in Table 1. Out of the 100 patients, 47 had septic shock at the time of enrolment.
Figure 1. Participant flow diagram.
Table 1. Baseline Clinical Characteristics and in-Hospital Outcomes.
| Characteristic | Patients (n = 100) |
Sepsis (n = 53) |
Septic Shock (n = 47) |
Control (n = 48) |
|---|---|---|---|---|
| Age (years), mean ± SD | 54.0 ± 14.9 | 52.7 ± 14.7 | 55.4 ± 15.2 | 48.3 ± 9.9 |
| Sex – male, n (%) | 64 (64.0) | 31 (58.5) | 33 (70.2) | 32 (66.7) |
| Comorbid conditions, n (%) | ||||
| - Hypertension | 30 (30.0) | 21 (39.6) | 9 (19.2) | 9 (18.7) |
| - Diabetes Mellitus | 18 (18.0) | 14 (26.4) | 4 (8.5) | 3 (6.3) |
| - Chronic Respiratory Disease | 73 (73.0) | 37 (69.8) | 36 (76.6) | 3 (6.3) |
| Microbiological Diagnosis, n (%) | - | |||
| - Community-acquired pneumonia (culture negative) | 73 (73.0) | 41 (75.5) | 32 (68.1) | |
| - Bacterial pneumonia | 15 (15.0) | 5 (11.3) | 10 (21.3) | |
| ○ Klebsiella | 5 (5.0) | 1 (1.9) | 4 (8.5) | |
| ○ Staphylococcus | 2 (2.0) | 1 (1.9) | 1 (2.1) | |
| ○ Pseudomonas | 2 (2.0) | 1 (1.9) | 1 (2.1) | |
| ○ Escherichia coli | 2 (2.0) | 1 (1.9) | 1 (2.1) | |
| ○ Others | 4 (4.0) | 1 (1.9) | 3 (6.4) | |
| - Fungal pneumonia | 3 (3.0) | 2 (3.8) | 1 (2.1) | |
| - Tuberculosis | 5 (5.0) | 3 (5.7) | 2 (4.3) | |
| - Influenza | 2 (2.0) | 2 (3.8) | 0 (0.0) | |
| - Brucellosis | 1 (1.0) | 0 | 1 (2.1) | |
| - Malaria | 1 (1.0) | 0 | 1 (2.1) | |
| SOFA Score (Enrolment), median (IQR) | 6 (4-9) | 5 (3-6) | 9 (6-10) | - |
| SOFA Score (Day 5), median (IQR) | 3 (2-6) | 3 (2-4) | 3 (2-8) | - |
| Invasive Mechanical Ventilation, n (%) | 74 (74) | 36 (67.9) | 38 (80.8) | - |
| Organ system failure | - | |||
| - Circulatory | 47 (47.0) | 0 | 47 (100.0) | |
| - Renal | 61 (61.0) | 31 (58.5) | 30 (63.8) | |
| - Respiratory | 93 (93.0) | 52 (98.1) | 41 (87.2) | |
| - Coagulation | 30 (30.0) | 20 (37.7) | 10 (21.3) | |
| - Hepatic | 36 (36.0) | 16 (30.2) | 20 (42.6) | |
| - Neurologic | 41 (41.0) | 16 (30.2) | 25 (53.2) | |
| In-Hospital Outcomes, n (%): | - | |||
| - Very-early clinical deterioration or death (at 48 h) | 22 (22.0) | 10 (18.9) | 12 (25.5) | |
| - Early clinical deterioration or death (at Day 5) | 26 (26.0) | 12 (22.6) | 14 (29.8) | |
| - Early hospital-acquired infections (at day 5) | 22 (22.0) | 10 (18.9) | 12 (25.5) | |
| ○ HAP/VAP | 18 (18.0) | 7 (13.2) | 11 (23.4) | |
| ○ UTI | 1 (1.0) | 1 (1.9) | 0 (0.0) | |
| ○ CRBSI | 2 (2.0) | 1 (1.9) | 1 (2.1) | |
| ○ IAI | 1 (1.0) | 1 (1.9) | 0 (0.0) | |
| - In-Hospital Mortality | 41 (41.0) | 14 (26.4) | 27 (57.5) | |
| ICU Stay (Days), median (IQR) | 7 (4.5-11) | 6 (4-8) | 8 (5-14) | - |
| Hospital Stay (Days), median (IQR) | 11 (7-16) | 10 (7-14) | 12 (7-21) | - |
CRBSI, catheter-related bloodstream infection; HAP/VAP, hospital-acquired pneumonia/ventilator-associated pneumonia; SD, standard deviation; SOFA, sequential organ failure assessment; IQR, interquartile range; IAI, intra-abdominal infection; ICU, intensive care unit; UTI, urinary tract infection.
The median (IQR) SOFA score at enrolment was 6 (4-9) and that at day 5 was 3 (2-6) (Supplementary Figure 2). There was early clinical deterioration or death at day 5 in 26 patients. Early death (within 5 days of admission) occurred in 8 patients, whereas the total in-hospital mortality rate was 41%. FCM data on peripheral blood leukocyte subsets was available for all 100 included patients at enrolment and for 76 patients on day 5 thereafter.
Pre-Specified Outcomes
The comparison of the peripheral blood leukocyte subsets between controls, sepsis, and septic shock patients is presented in Figure 2 and Supplementary Table 1. Patients with sepsis had significantly lower percentages of lymphocytes, T-cells, and CD4 T-cells compared to controls. In contrast, they had a higher percentage of CD8 T-cells compared to controls. The percentages of the activated T-lymphocyte subsets, namely CD3+HLA-DR+, CD4+CD25+, and CD8+HLA-DR+, were increased in patients with sepsis compared to controls. In contrast, the percentage of CD8+CD25+cells was reduced among patients with septic shock compared to controls. Gradients of increasing MFI of granulocyte CD64 and granulocyte CD24 were observed from controls to sepsis to septic shock patients.
Figure 2. Comparisons of various peripheral blood leukocyte subpopulations between controls, sepsis, and septic shock groups at the time of enrolment.
(A) Lymphocyte percentage, (B) T-cell percentage, (C) CD4 T-cell percentage, (D) CD8 T-cell percentage, (E) CD3+HLA-DR+percentage, (F) CD4+CD25+percentage, (G) CD8+CD25+percentage, (H) CD8+HLA-DR+percentage, (I) Granulocyte CD64 MFI.
The associations between the peripheral blood leukocyte subpopulations at the time of enrolment and the early clinical deterioration at day 5 are presented in Table 2. We did not find any peripheral leukocyte subset to have a significant association with early deterioration on univariate or multivariate logistic regression.
Table 2. Association of Peripheral Blood Leukocyte Subpopulations and Activation Markers at Enrolment with Early Clinical Outcome at 5 Days.
| Peripheral Blood Subpopulation | Early deterioration (n = 26) |
Early Improvement/Stability (n = 74) | OR | 95% CI | p value | Multivariate OR | 95% CI | p value |
|---|---|---|---|---|---|---|---|---|
| Lymphocyte Percentage (%) | 2.27 (1.64-4.12) | 2.61 (1.18-4.82) | 0.94 | 0.82–1.07 | 0.32 | - | - | - |
| Lymphocyte Count (cells/μL) | 468.5 (320-754) | 392.5 (183-785) | 1.00 | 0.99–1.01 | 0.77 | - | - | - |
| T Cell Percentage (%) | 59.38 (45.43-71.48) | 64.76 (49.62-70.66) | 0.99 | 0.97–1.02 | 0.66 | - | - | - |
| T Cell Count (cells/μL) | 297 (149-382) | 244.5 (105-413) | 0.99 | 0.98–1.00 | 0.56 | - | - | - |
| B Cell Percentage (%) | 17.96 (9.40-24.25) | 16.40 (9.38-25.97) | 0.99 | 0.97–1.03 | 0.95 | - | - | - |
| B Cell Count (cells/μL) | 71.5 (21-140) | 62 (28-141) | 1.00 | 0.99–1.02 | 0.66 | - | - | - |
| NK Cell Percentage (%) | 9.26 (5.68-16.85) | 9.66 (5.50-17.46) | 1.00 | 0.95–1.05 | 0.91 | - | - | - |
| NK Cell Count (cells/μL) | 44 (21-83) | 40 (16-92) | 0.99 | 0.99–1.00 | 0.81 | - | - | - |
| CD4 T Cell Percentage (%) | 56.26 (48.7-61.42) | 49.22 (39.48-58.34) | 1.02 | 0.99–1.06 | 0.20 | - | - | - |
| CD4 T Cell Count (cells/μL) | 159.5 (63-263) | 111.5 (52-256) | 0.99 | 0.99–1.00 | 0.81 | - | - | - |
| CD8 T Cell Percentage (%) | 35.34 (27.82-44.92) | 44.67 (30.43-55.45) | 0.97 | 0.94–1.01 | 0.10 | - | - | - |
| CD8 T Cell Count (cells/μL) | 91 (59-164) | 96 (42-211) | 0.99 | 0.99–1.00 | 0.34 | - | - | - |
| CD3+CD25+ Cell Percentage (%) | 3.25 (1.36-5.04) | 2.09 (0.88-6.64) | 0.97 | 0.91–1.04 | 0.46 | - | - | - |
| CD3+HLA-DR+Cell Percentage (%) | 6.25 (2.32-12.05) | 8.76 (3.70-26.51) | 0.96 | 0.93–1.00 | 0.06 | 0.97 | 0.93–1.02 | 0.20 |
| CD4+CD25+ Cell Percentage (%) | 3.66 (1.65-7.72) | 2.58 (1.14-6.19) | 1.01 | 0.95–1.07 | 0.79 | - | - | - |
| CD4+HLA-DR+Cell Percentage (%) | 6.90 (2.02-11.46) | 9.92 (3.75-39.05) | 0.98 | 0.96–1.00 | 0.09 | 0.99 | 0.97–1.02 | 0.52 |
| CD8+CD25+ Cell Percentage (%) | 0.30 (0.00-0.99) | 0.44 (0.00-3.83) | 0.98 | 0.91–1.05 | 0.54 | - | - | - |
| CD8+HLA-DR+Cell Percentage (%) | 9.00 (4.25-18.07) | 12.34 (5.00-24.09) | 0.97 | 0.94–1.01 | 0.15 | - | - | - |
| Granulocyte CD64 MFI | 6.88 (4.00-8.52) | 4.50 (2.03-8.51) | 1.05 | 0.99–1.11 | 0.11 | - | - | - |
| Granulocyte CD16 MFI | 19.71 (11.61-27.37) | 24.77 (9.61-50.26) | 0.99 | 0.96–1.00 | 0.15 | - | - | - |
| Granulocyte CD10 MFI | 1.44 (1.24-1.78) | 2.81 (1.60-4.99) | 0.97 | 0.83–1.14 | 0.73 | - | - | - |
| Granulocyte CD24 MFI | 21.27 (10.77-32.76) | 17.44 (7.94-24.92) | 1.02 | 0.99–1.05 | 0.22 | - | - | - |
Data presented as Median (IQR).
The associations between the peripheral blood leukocyte subpopulations at the time of enrolment and in-hospital mortality are presented in Table 3. On univariate analysis, mortality was associated with reduced percentages of lymphocytes, CD3+HLA-DR+cells, CD4+HLA-DR +cells, and CD8 + CD25+cells. On multivariate analysis, the median (IQR) CD8+CD25+cell percentage was significantly lower among non-survivors than survivors [non-survivors, 0.29 (0.00-1.02) versus survivors, 0.54 (0.00-5.08); OR, 0.82; 95% CI, 0.70–0.97; p-value = 0.02]. Additionally, among the non-survivors, there was a trend towards lower median (IQR) percentage of CD3 + HLA-DR +cells [non-survivors, 7.36 (2.16-14.63) versus survivors, 8.94 (4.82-28.41); OR, 0.96; 95% CI, 0.92–1.00; p-value = 0.06].
Table 3. Association of Peripheral Blood Leukocyte Subpopulations and Activation Markers at Enrolment with in-Hospital Mortality.
| Peripheral Blood Subpopulation | Non-Survivors (n = 41) |
Survivors (n = 59) |
Univariate OR | 95% CI | p Value | Multivariate OR | 95% CI | p value |
|---|---|---|---|---|---|---|---|---|
| Lymphocyte Percentage (%) | 1.78 (1.08-3.15) | 2.88 (1.80-5.24) | 0.85 | 0.73–0.98 | 0.03 | 0.82 | 0.62–1.08 | 0.15 |
| Lymphocyte Count (cells/μL) | 365 (181-639) | 467 (242-975) | 0.99 | 0.99–1.00 | 0.08 | 1.00 | 1.00–1.00 | 0.55 |
| T Cell Percentage (%) | 66.64 (45.72-71.80) | 63.78 (49.87-69.78) | 1.00 | 0.98–1.03 | 0.78 | - | - | - |
| T Cell Count (cells/μL) | 231 (105-378) | 274 (122-438) | 0.99 | 0.99–1.00 | 0.08 | 1.00 | 1.00–1.00 | 0.45 |
| B Cell Percentage (%) | 18.15 (9.40-24.70) | 15.84 (9.38-23.69) | 1.00 | 0.97–1.03 | 0.96 | - | - | - |
| B Cell Count (cells/μL) | 59 (21-122) | 69 (31-168) | 0.99 | 0.99–1.00 | 0.41 | - | - | - |
| NK Cell Percentage (%) | 7.94 (5.35-13.01) | 11.74 (6.46-17.86) | 0.98 | 0.94–1.03 | 0.43 | - | - | - |
| NK Cell Count (cells/μL) | 26 (14-60) | 56 (19-112) | 0.99 | 0.99–1.00 | 0.12 | - | - | - |
| CD4 T Cell Percentage (%) | 53.77 (41.55-61.42) | 50.40 (39.48-58.34) | 1.01 | 0.98–1.04 | 0.40 | - | - | - |
| CD4 T Cell Count (cells/μL) | 104 (53-194) | 132 (60-293) | 0.99 | 0.99–1.00 | 0.15 | - | - | - |
| CD8 T Cell Percentage (%) | 38.47 (27.85-48.39) | 44.17 (30.43-55.35) | 0.98 | 0.96–1.01 | 0.33 | - | - | - |
| CD8 T Cell Count (cells/μL) | 84.5 (40-137.5) | 101.5 (48.5-232) | 0.99 | 0.99–1.00 | 0.08 | 1.00 | 0.99–1.01 | 0.99 |
| CD3+CD25+ Cell Percentage (%) | 2.41 (1.28-5.26) | 2.06 (0.88-6.73) | 0.99 | 0.95–1.04 | 0.74 | - | - | - |
| CD3+HLA-DR+ Cell Percentage (%) | 7.36 (2.16-14.63) | 8.94 (4.82-28.41) | 0.97 | 0.94–0.99 | 0.03 | 0.96 | 0.92–1.00 | 0.06 |
| CD4+CD25+ Cell Percentage (%) | 3.53 (1.60-6.19) | 2.68 (0.95-7.48) | 0.99 | 0.93–1.05 | 0.66 | - | - | - |
| CD4+HLA-DR+ Cell Percentage (%) | 5.85 (2.24-11.46) | 11.69 (6.33-44.01) | 0.98 | 0.96–0.99 | 0.03 | 0.99 | 0.97–1.02 | 0.60 |
| CD8+CD25+ Cell Percentage (%) | 0.29 (0.00-1.02) | 0.54 (0.00-5.08) | 0.85 | 0.73–0.98 | 0.03 | 0.82 | 0.70–0.97 | 0.02 |
| CD8+HLA-DR+ Cell Percentage (%) | 9.77 (3.38-18.39) | 12.63 (5.61-24.09) | 0.98 | 0.95–1.01 | 0.16 | - | - | - |
| Granulocyte CD64 MFI | 6.74 (3.75-8.70) | 3.83 (1.97-8.11) | 1.03 | 0.98–1.09 | 0.26 | - | - | - |
| Granulocyte CD16 MFI | 19.15 (9.68-40.88) | 25.08 (12.86-49.81) | 0.99 | 0.97–1.00 | 0.11 | - | - | - |
| Granulocyte CD10 MFI | 1.78 (1.32-5.20) | 2.33 (1.48-4.40) | 1.00 | 0.88–1.14 | 0.96 | - | - | - |
| Granulocyte CD24 MFI | 19.42 (10.63-31.69) | 16.25 (7.34-25.37) | 1.02 | 0.99–1.05 | 0.28 | - | - | - |
Data presented as Median (IQR).
The associations between the peripheral blood leukocyte subpopulations on day 5 after enrolment and in-hospital mortality are presented in Table 4. On univariate analysis, mortality was associated with a lower percentage of lymphocytes and a lower MFI of granulocyte CD16 on day 5. Additionally, lower counts of lymphocytes, NK-cells, T-cells, CD4 T-cells, and CD8 T-cells were associated with mortality. However, on multivariate analysis, only a lower median (IQR) day 5 lymphocyte percentage was associated with poorer survival [non-survivors, 1.93 (1.41-2.58) versus survivors, 5.62 (3.82-9.09); OR, 0.28; 95% CI, 0.11–0.69; p-value = 0.01].
Table 4. Association of Peripheral Blood Leukocyte Subpopulations and Activation Markers on day 5 After Enrolment with in-Hospital Mortality.
| Peripheral Blood Subpopulation | Non-survivors (n = 22) | Survivors (n = 54) | Univariate OR | 95% CI | p Value | Multivariate OR | 95% CI | P value |
|---|---|---|---|---|---|---|---|---|
| Lymphocyte Percentage | 1.93 (1.41-2.58) | 5.62 (3.82-9.09) | 0.63 | 0.48–0.82 | 0.001 | 0.28 | 0.11–0.69 | 0.01 |
| Lymphocyte Count | 259.5 (143-524) | 581 (340-1009) | 0.99 | 0.99–0.99 | 0.01 | 1.01 | 1.00–1.02 | 0.09 |
| T Cell Percentage (%) | 60.35 (46.80-68.99) | 64.22 (54.56-72.53) | 0.98 | 0.95–1.01 | 0.19 | |||
| T Cell Count | 161.5 (75-245) | 356 (205-619) | 0.99 | 0.99–0.99 | 0.01 | 0.94 | 0.87–1.03 | 0.18 |
| B Cell Percentage (%) | 21.52 (13.85-29.67) | 17.14 (10.74-24.07) | 1.03 | 0.99–1.08 | 0.14 | - | - | - |
| B Cell Count | 44.5 (17-137) | 100 (44-160) | 0.99 | 0.99–1.00 | 0.29 | - | - | - |
| NK Cell Percentage (%) | 8.80 (6.65-13.04) | 8.66 (5.74-15.60) | 1.00 | 0.93–1.08 | 0.95 | - | - | - |
| NK Cell Count | 18.5 (10-40) | 48.5 (25-92) | 0.98 | 0.96–0.99 | 0.01 | 1.01 | 0.97–1.04 | 0.61 |
| CD4 T Cell Percentage (%) | 52.26 (38.52-62.78) | 52.66 (44.16-62.68) | 0.99 | 0.96–1.03 | 0.82 | - | - | - |
| CD4 T Cell Count | 87.5 (28-140.5) | 184 (93-325) | 0.99 | 0.98–0.99 | 0.01 | 1.05 | 0.97–1.13 | 0.22 |
| CD8 T Cell Percentage (%) | 40.16 (30.71-48.50) | 41.08 (29.97-48.78) | 1.00 | 0.97–1.03 | 0.99 | - | - | - |
| CD8 T Cell Count | 54 (31-98) | 135 (78-259) | 0.99 | 0.98–0.99 | 0.01 | 1.06 | 0.98–1.16 | 0.15 |
| CD3+CD25+ Cell Percentage (%) | 1.43 (0.39-2.74) | 2.86 (0.83-6.30) | 0.86 | 0.70–1.05 | 0.14 | - | - | - |
| CD3+HLA-DR+ Cell Percentage (%) | 6.06 (1.82-14.59) | 9.08 (3.67-20.65) | 0.98 | 0.94–1.02 | 0.39 | - | - | - |
| CD4+CD25+ Cell Percentage (%) | 1.69 (0.78-6.20) | 3.56 (1.14-6.97) | 0.91 | 0.79–1.05 | 0.20 | - | - | - |
| CD4+HLA-DR+ Cell Percentage (%) | 7.20 (3.23-23.56) | 9.47 (4.80-22.58) | 0.99 | 0.97–1.02 | 0.82 | - | - | - |
| CD8+CD25+ Cell Percentage (%) | 0.23 (0.00-0.70) | 0.45 (0.03-6.25) | 0.72 | 0.50–1.02 | 0.06 | 0.68 | 0.41–1.10 | 0.12 |
| HLA-DR+CD8+ Cell Percentage (%) | 10.82 (4.70-22.32) | 11.74 (5.63-20.42) | 0.99 | 0.96–1.04 | 0.94 | - | - | - |
| Granulocyte CD64 MFI | 3.45 (1.40-5.35) | 1.91 (1.21-3.62) | 1.12 | 0.95–1.32 | 0.17 | - | - | - |
| Granulocyte CD16 MFI | 15.74 (12.44-34.23) | 38.87 (20.30-52.25) | 0.97 | 0.94–0.99 | 0.03 | 0.99 | 0.96–1.03 | 0.67 |
| Granulocyte CD10 MFI | 2.98 (1.51-5.75) | 3.42 (2.04-5.71) | 1.03 | 0.89–1.20 | 0.67 | - | - | - |
| Granulocyte CD24 MFI | 12.43 (7.24-28.00) | 17.03 (5.91-24.29) | 1.00 | 0.96–1.05 | 0.67 | - | - | - |
Data presented as Median (IQR).
Post-hoc Analyses
A total of 22 (22%) patients had incident HAIs within 5 days of enrolment into the study (Table 1). The incidence of HAI was significantly higher among patients with early clinical deterioration at day 5 compared to those with early clinical improvement or stability (46.2% vs 13.5%, p<0.001; see Supplementary Table 2).
A total of 22 (22%) patients had “very early” clinical deterioration within 48 h of enrolment. The associations between peripheral blood leukocyte subpopulations at the time of enrolment and “very early” clinical deterioration are shown in Table 5. On univariate analysis, clinical deterioration at 48 h was associated with an increased granulocyte CD64 MFI and a reduced granulocyte CD16 MFI at enrolment. On multivariate analysis, clinical deterioration at 48 h was associated with a significantly higher median (IQR) granulocyte CD64 MFI [“very early deterioration”, 7.38 (3.23-19.18) versus “very early improvement/stability, 4.66 (2.19-8.08); OR, 1.07; 95% CI, 1.01–1.14; p-value = 0.02] and a significantly lower median (IQR) CD16 MFI [“very early deterioration”, 19.15 (10.51-25.11) versus “very early improvement/stability”, 24.77 (9.68-50.70); OR, 0.97; 95% CI, 0.95–1.00; p-value = 0.04].
Table 5. Association of Peripheral Blood Leukocyte Subpopulations and Activation Markers at Enrolment with “Very Early” Clinical Outcome (at 48 h After Enrolment).
| Peripheral Blood Subpopulation | Very Early deterioration (n = 22) | Early Improvement/Stability (n = 78) | OR | 95% CI | p value | Multivariate OR | 95% CI | p value |
|---|---|---|---|---|---|---|---|---|
| Lymphocyte Percentage (%) | 2.00 (1.24-4.12) | 2.63 (1.51-4.97) | 0.88 | 0.74–1.05 | 0.16 | - | - | - |
| Lymphocyte Count (cells/μL) | 461 (181-704) | 404 (210-806) | 1.00 | 1.00–1.00 | 0.32 | - | - | - |
| T Cell Percentage (%) | 65.57 (56.00-71.88) | 62.64 (47.51-70.62) | 1.02 | 0.99–1.05 | 0.32 | - | - | - |
| T Cell Count (cells/μL) | 260 (135-383) | 253 (111-413) | 1.00 | 1.00–1.00 | 0.35 | - | - | - |
| B Cell Percentage (%) | 16.84 (9.38-23.68) | 16.67 (10.15-25.97) | 0.99 | 0.95–1.03 | 0.56 | - | - | - |
| B Cell Count (cells/μL) | 68 (16-120) | 62 (29-155) | 1.00 | 0.99–1.00 | 0.45 | - | - | - |
| NK Cell Percentage (%) | 7.90 (5.68-11.71) | 9.93 (5.50-17.78) | 0.99 | 0.94–1.04 | 0.78 | - | - | - |
| NK Cell Count (cells/μL) | 37 (19-61) | 45 (16-99) | 1.00 | 0.99–1.00 | 0.58 | - | - | - |
| CD4 T Cell Percentage (%) | 55.02 (49.28-65.77) | 49.84 (39.48-58.34) | 1.03 | 0.99–1.06 | 0.12 | - | - | - |
| CD4 T Cell Count (cells/μL) | 139 (67-206) | 116 (53-261) | 1.00 | 1.00–1.00 | 0.59 | - | - | - |
| CD8 T Cell Percentage (%) | 33.65 (30.43-45.63) | 43.58 (29.66-55.24) | 0.98 | 0.94–1.01 | 0.18 | - | - | - |
| CD8 T Cell Count (cells/μL) | 87 (40-137) | 96 (44-214) | 1.00 | 0.99–1.00 | 0.24 | - | - | - |
| CD3+CD25+ Cell Percentage (%) | 2.28 (0.61-4.25) | 2.26 (1.01-6.60) | 0.93 | 0.83–1.05 | 0.25 | - | - | - |
| CD3+HLA-DR+ Cell Percentage (%) | 8.20 (2.69-17.95) | 8.26 (3.58-25.47) | 0.99 | 0.96–1.02 | 0.49 | - | - | - |
| CD4+CD25+ Cell Percentage (%) | 2.52 (0.68-5.61) | 3.41 (1.36-7.48) | 0.97 | 0.88–1.06 | 0.47 | - | - | - |
| CD4+HLA-DR+ Cell Percentage (%) | 9.43 (3.75-19.20) | 8.73 (3.04-35.71) | 0.99 | 0.98–1.01 | 0.56 | - | - | - |
| CD8+CD25+ Cell Percentage (%) | 0.40 (0–1.82) | 0.38 (0-2.51) | 0.95 | 0.86–1.06 | 0.38 | - | - | - |
| CD8+HLA-DR+ Cell Percentage (%) | 10.06 (5.75-21.29) | 11.48 (4.08-23.14) | 0.99 | 0.95–1.02 | 0.49 | - | - | - |
| Granulocyte CD64 MFI | 7.38 (3.23-19.18) | 4.66 (2.19-8.08) | 1.07 | 1.01–1.14 | 0.03 | 1.08 | 1.01–1.15 | 0.02 |
| Granulocyte CD16 MFI | 19.15 (10.51-25.11) | 24.77 (9.68-50.70) | 0.97 | 0.95–1.00 | 0.047 | 0.97 | 0.95–1.00 | 0.04 |
| Granulocyte CD10 MFI | 1.48 (1.15–1.78) | 2.76 (1.51-5.08) | 0.89 | 0.72–1.12 | 0.32 | - | - | - |
| Granulocyte CD24 MFI | 24.56 (9.00-34.87) | 17.44 (8.18-24.92) | 1.02 | 0.99–1.06 | 0.15 | - | - | - |
Data presented as Median (IQR).
Among the subgroup of 88 patients with suspected or confirmed bacterial sepsis, 22 (25%) patients had early clinical deterioration at day 5 and 36 (40.9%) died in hospital. The post-hoc analysis of the associations of leukocyte subpopulations at enrolment with early clinical deterioration at day 5 in those with suspected or confirmed bacterial sepsis is presented in Table 6. On multivariate analysis, early clinical deterioration was associated with a lower median (IQR) percentage of CD3 + HLA-DR +cells [early deterioration, 5.72 (1.99-12.05) versus early improvement/stability, 8.26 (3.58-25.90); OR, 0.88; 95% CI, 0.80–0.96; p-value = 0.01] and a higher median (IQR) granulocyte CD24 MFI [early deterioration, 27.22 (10.77-34.48) versus early improvement/stability, 17.44 (7.94-24.45); OR, 1.07; 95% CI, 1.01–1.14; p-value = 0.02]. The associations of leukocyte subpopulations at enrolment with in-hospital mortality in the subgroup with suspected or confirmed bacterial sepsis is presented in Supplementary Table 3. On multivariate analysis, in-hospital mortality was associated with a lower median (IQR) percentage of CD3 + HLA-DR +cells [non-survivors, 5.92 (2.00-14.59) versus survivors, 8.26 (4.80-26.67); OR, 0.93; 95% CI, 0.87–0.99; p-value = 0.04].
Table 6. Association of Peripheral Blood Leukocyte Subpopulations and Activation Markers at Enrolment with Early Clinical Outcome at 5 Days in the Subset with Suspected or Confirmed Bacterial Sepsis.
| Peripheral Blood Subpopulation | Early deterioration (n = 22) | Early Improvement/Stability (n = 66) | OR | 95% CI | p value | Multivariate OR | 95% CI | p value |
|---|---|---|---|---|---|---|---|---|
| Lymphocyte Percentage (%) | 1.86 (1.54-3.64) | 2.55 (1.12-4.29) | 0.93 | 0.79–1.10 | 0.43 | - | - | - |
| Lymphocyte Count (cells/μL) | 468 (315-754) | 388 (181-681) | 1.00 | 1.00–1.00 | 0.83 | - | - | - |
| T Cell Percentage (%) | 59.37 (45.43-71.11) | 63.85 (49.62-69.79) | 0.99 | 0.96–1.02 | 0.63 | - | - | - |
| T Cell Count (cells/μL) | 275 (135-382) | 234 (105-365) | 1.00 | 1.00–1.00 | 0.88 | - | - | - |
| B Cell Percentage (%) | 15.70 (7.86-24.25) | 16.13 (10.15-25.97) | 1.00 | 0.96–1.04 | 0.88 | - | - | - |
| B Cell Count (cells/μL) | 57 (19-120) | 57 (27-122) | 1.00 | 1.00–1.00 | 0.36 | - | - | - |
| NK Cell Percentage (%) | 9.26 (5.68-16.85) | 9.66 (5.50-17.75) | 1.00 | 0.96–1.05 | 0.80 | - | - | - |
| NK Cell Count (cells/μL) | 45 (21-83) | 34 (15-79) | 1.00 | 1.00–1.00 | 0.89 | - | - | - |
| CD4 T Cell Percentage (%) | 56.26 (51.67-61.42) | 50.47 (39.63-59.83) | 1.02 | 0.99–1.06 | 0.22 | - | - | - |
| CD4 T Cell Count (cells/μL) | 159 (63-263) | 106 (46-191) | 1.00 | 1.00–1.00 | 0.94 | - | - | - |
| CD8 T Cell Percentage (%) | 33.82 (27.82-44.16) | 43.75 (30.43-55.45) | 0.97 | 0.94–1.01 | 0.13 | - | - | - |
| CD8 T Cell Count (cells/μL) | 87 (44-163) | 85 (42-175) | 1.00 | 1.00–1.00 | 0.70 | - | - | - |
| CD3+CD25+ Cell Percentage (%) | 3.44 (1.66-5.04) | 2.30 (0.88-6.73) | 0.98 | 0.91–1.06 | 0.60 | - | - | - |
| CD3+HLA-DR+ Cell Percentage (%) | 5.72 (1.99-12.05) | 8.26 (3.58-25.90) | 0.96 | 0.92–1.00 | 0.08 | 0.88 | 0.80–0.96 | 0.01 |
| CD4+CD25+ Cell Percentage (%) | 3.30 (1.65-7.72) | 2.58 (1.14-6.12) | 1.04 | 0.95–1.14 | 0.39 | - | - | - |
| CD4+HLA-DR+ Cell Percentage (%) | 6.91 (2.02-11.46) | 9.20 (3.75-36.87) | 0.98 | 0.96–1.01 | 0.17 | - | - | - |
| CD8+CD25+ Cell Percentage (%) | 0.34 (0-1.73) | 0.44 (0-3.83) | 1.00 | 0.92–1.08 | 0.92 | - | - | - |
| CD8+HLA-DR+ Cell Percentage (%) | 9.87 (4.08-18.07) | 12.34 (4.30-23.43) | 0.97 | 0.93–1.02 | 0.23 | - | - | - |
| Granulocyte CD64 MFI | 6.90 (3.45-11.83) | 4.29 (2.03-7.87) | 1.10 | 1.01–1.18 | 0.02 | 1.08 | 0.98–1.20 | 0.10 |
| Granulocyte CD16 MFI | 21.56 (10.96-30.72) | 24.77 (9.39-49.33) | 0.99 | 0.97–1.01 | 0.27 | - | - | - |
| Granulocyte CD10 MFI | 1.49 (1.24-1.99) | 3.03 (1.80-5.08) | 1.01 | 0.86–1.18 | 0.90 | - | - | - |
| Granulocyte CD24 MFI | 27.22 (10.77-34.48) | 17.44 (7.94-24.45) | 1.03 | 1.00–1.07 | 0.09 | 1.07 | 1.01–1.14 | 0.02 |
Data presented as Median (IQR).
Discussion
In this single-center pilot observational study conducted among patients admitted with sepsis in a respiratory ICU within 48 h of hospital admission, we have used flow cytometry to explore the associations of an extensive array of peripheral blood leukocyte subpopulations and activation markers with early clinical course and in-hospital outcomes. We did not find any leukocyte subpopulation to be significantly associated with early clinical deterioration at day 5 of enrolment.
The time cut-offs to define “early outcomes” in sepsis have varied widely in literature between 1,15 28,9 and 10 days16 after sepsis onset. Our choice of day 5 for determination of early outcomes in sepsis was guided by the dichotomous nature of immune responses in early (day 1-3) and late (day 6 or later) sepsis-related deaths noted by Hotchkiss et al5 These authors have theorized that the early deaths occur due to a hyperinflammatory “cytokine storm”. They suggest that the immune system recovers among survivors by day 6 after sepsis onset. In contrast, those who die later are hypothesized to have persistent suppression of the adaptive immune response. These findings are supported by Daviaud et al, who noted a trough in the bimodal distribution of sepsis-related deaths on days 4 and 5 of sepsis onset.17 Based on these findings, we hypothesized that outcome assessment at day 5 could identify all patients with early deterioration.
We found that 22% of our patients developed HAIs till day 5 of enrolment indicating presence of immunosuppression at enrolment. Further, there was a significant association between incident HAIs and early deterioration on day 5 of enrolment. To eliminate the impact of HAIs on our “early” outcome, we performed a post-hoc analysis to assess the association of the baseline peripheral blood leukocyte subpopulations with “very early” clinical deterioration at 48 h after enrolment. We found a hyperinflammatory state (increased granulocyte CD64 MFI) with a marker of immature granulocytes (reduced granulocyte CD16 MFI) among patients with very early clinical deterioration at 48 h after enrolment.
The findings of our post-hoc analysis are similar to those of Guérin et al8 and Daix et al,9 who implicated the role of CD10dim and CD16dim immature granulocytes in early clinical worsening at 48 h. These immature granulocytes are equivalent to the band forms observed on the peripheral blood smear, which are known to indicate a worse prognosis in sepsis.18 Interestingly, contrary to some studies,8,9 and akin to the study by Hanna et al,19 we did not find differences in granulocyte CD10 and CD16 levels between controls and patients with sepsis at baseline. The heterogeneity of our sepsis population which was inclusive of bacterial, viral, fungal, parasitic, and culture-negative sepsis may have led to this discordance. Inter-laboratory variability and methodological differences in the estimation of CD10 and CD16 expression on granulocytes may also have contributed to these differences. Furthermore, while we compared the flow cytometric parameters of controls with those of patients within 48 h of admission, the nadir of CD10 and CD16 expressions may occur between day 3 and day 8 after the onset of sepsis.20
On evaluating the associations of leukocyte subpopulations at enrolment with in-hospital mortality, we found a significantly lower percentage of activated CD8 T-cells with CD25 expression among those patients who died. Interestingly, we also found that patients with septic shock, but not those with sepsis had a lower percentage of CD8+CD25 +cells at enrolment compared with healthy controls. CD25 is an important T-lymphocyte activation marker that serves as the IL2 receptor and is necessary for lymphocyte proliferation, survival, and function. It has moderate basal expression, which increases within 24 h of immune activation.21,22 A lower level of CD25 expressing CD8 T-cells at baseline among those with septic shock, and its association with subsequent mortality, is indicative of adaptive immune dysfunction early in the course of severe or life-threatening illness. Akin to CD4+CD25+cells, a subset of CD8+CD25+cells with Foxp3 expression may serve as regulatory T-cells (Treg) which suppress the immune response. Chen et al have found that higher proportions of CD4+CD25+ Treg cells on day 7 after sepsis onset are associated with mortality.23 However, we did not specifically evaluate the CD8+CD25 +subset with Foxp3 expression functioning as Tregs in this study.
We also found a trend between a lower percentage of activated T-cells with HLA-DR expression at enrolment with the occurrence of in-hospital mortality. Additionally, in post-hoc subgroup analyses of patients with suspected or confirmed bacterial sepsis, a lower percentage of CD3+HLA-DR+T-cells was associated with early clinical deterioration at day 5 and in-hospital mortality. The presence of HLA-DR, a receptor of the major histocompatibility complex (MHC) class II on antigen-presenting cells is necessary for activating the T-cells to initiate the adaptive immune response. Sepsis-related reduction in monocyte HLA-DR expression is well described and results in adaptive immune dysfunction and increased mortality.24,25 However, there is a paucity of data on the prognostic role of HLA-DR expression on lymphocytes in sepsis mortality. Since the presence of HLA-DR on T-lymphocyte is a marker of the activation status of the latter,22 the associations of lower CD3 + HLA-DR+ T-cell percentage at enrolment with poorer sepsis-related outcomes in our study may further indicate an early onset of adaptive immune dysfunction in these patients. However, in contrast to the CD8 + CD25 + cells, we found an increasing percentage of CD3 + HLA-DR + cells from controls to sepsis and septic shock patients. Whether the inappropriately low expression of HLA-DR on T-cells in septic shock is a predictor of poor prognosis needs to be further explored.
We also found that in-hospital mortality was significantly associated with a reduced day 5 lymphocyte percentage. Lymphocytopenia has been previously described as a risk factor for mortality in patients with sepsis. This may be related to the suppressor activity of myeloid-derived suppressor cells,9 increased apoptosis of lymphocytes,26,27 or impaired lymphocyte maturation in sepsis.28 Furthermore, non-survivors of sepsis are unable to increase their T-lymphocyte counts on serial peripheral blood testing implying a persistently dysfunctional immune status.29
Our study had a few limitations. Although we attempted to recruit patients with community-acquired sepsis within 48 h of presentation to the hospital, patients with varying durations of incubating sepsis prior to the hospital admission may have been enrolled. This may not accurately shed light on the very early sepsis-related immune alterations, which occur before presenting to the hospital and is problematic in most current FCM studies of patients with sepsis.30 Secondly, our sepsis population was heterogenous with varying bacterial, fungal, and viral etiologies. This is because we enrolled patients with a diagnosis of sepsis within 48 h of admission and subsequent microbiological reports revealed organisms other than those conventionally associated with gram-positive and gramnegative sepsis in a small proportion (12%) of patients. Our post-hoc analysis in the subgroup with suspected or confirmed bacterial sepsis showed associations of early clinical deterioration with hyperinflammation (higher granulocyte CD24 MFI)and adaptive immune suppression (lower CD3+HLA-DR + T-cell percentage). Further exploration of the differences in flow-cytometric profiles in various types of infection is necessary. Thirdly, 73% of our population had culture negative sepsis. Nonetheless, our sample was representative of a real-world sepsis cohort because about half or more of all sepsis patients are culture negative and the in-hospital outcomes of culture positive and culture negative have been found to be similar.31,32 Finally, we have conducted an exploratory analysis of the prognostic impact of several leukocyte subpopulations in a relatively small number of patients with sepsis. Hence, the alteration of some leukocyte parameters revealed a prognostic trend that did not achieve statistical significance. A study with a larger sample size may clarify the validity of these associations. Larger studies are also needed to establish and validate optimal cut-off values and develop prediction models based on optimal combinations of these prognostic markers.
The strength of our study is that we included many candidate leukocyte subpopulations and activation markers using multi-parameter FCM, many of which have been sparsely studied in sepsis prognostication before. Further, this is one of the few studies from the developing world which has a disproportionately higher burden of sepsis. In addition to the leukocyte subpopulations included in this report, many others such as monocyte HLA-DR and PD-L1, and lymphocyte PD1 deserve further study.
To conclude, we performed flow cytometry within 48 h of hospital admission to explore associations of various leukocyte subpopulations with early clinical outcomes and in-hospital mortality among patients with sepsis. We did not find any of the leukocyte subpopulations included in this study to be associated with early clinical deterioration on day 5. However, in a post-hoc analysis, we found that patients with “very early” clinical deterioration at 48 h had granulocytes with increased CD64 MFI and reduced CD16 MFI, suggestive of a hyperinflammatory state in this subset.
In our pre-specified analysis exploring the association between peripheral blood leukocyte subpopulations with in-hospital mortality, we found a significantly reduced percentage of activated CD-25 expressing CD8 T-cells at the time of enrolment among non-survivors. We also found that mortality was associated with a reduced lymphocyte percentage on day 5 after enrollment. These findings suggest that adaptive immune dysfunction is established early and remains persistent during the course of lethal sepsis.
Supplementary Material
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the All India Institute of Medical Sciences, New Delhi, (grant number A-485).
Financial Support
This study was supported by an intramural grant from All India Institute of Medical Sciences, New Delhi (grant number A-485)
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
References
- 1.Fleischmann C, Scherag A, Adhikari NKJ, et al. Assessment of global incidence and mortality of hospital-treated sepsis. Current estimates and limitations. Am J Respir Crit Care Med. 2016;193(3):259–272. doi: 10.1164/rccm.201504-0781OC. [DOI] [PubMed] [Google Scholar]
- 2.Harrison DA, Welch CA, Eddleston JM. The epidemiology of severe sepsis in England, Wales and northern Ireland, 1996 to 2004: Secondary analysis of a high quality clinical database, the ICNARC case mix programme database. Crit Care. 2006;10(2):R42. doi: 10.1186/cc4854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kadri SS, Rhee C, Strich JR, et al. Estimating ten-year trends in septic shock incidence and mortality in United States academic medical centers using clinical data. Chest. 2017;151(2):278–285. doi: 10.1016/j.chest.2016.07.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest. 1992;101(6):1644–1655. doi: 10.1378/chest.101.6.1644. [DOI] [PubMed] [Google Scholar]
- 5.Hotchkiss RS, Monneret G, Payen D. Immunosuppression in sepsis: A novel understanding of the disorder and a new therapeutic approach. Lancet Infect Dis. 2013;13(3):260–268. doi: 10.1016/S1473-3099(13)70001-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Singer M, Deutschman CS, Seymour CW, et al. The third international consensus definitions for sepsis and septic shock (Sepsis-3) JAMA. 2016;315(8):801–810. doi: 10.1001/jama.2016.0287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Monneret G, Venet F. Sepsis-induced immune alterations monitoring by flow cytometry as a promising tool for individualized therapy. Cytometry B Clin Cytom. 2016;90(4):376–386. doi: 10.1002/cyto.b.21270. [DOI] [PubMed] [Google Scholar]
- 8.Guérin E, Orabona M, Raquil MA, et al. Circulating immature granulocytes with T-cell killing functions predict sepsis deterioration*. Crit Care Med. 2014;42(9):2007–2018. doi: 10.1097/CCM.0000000000000344. [DOI] [PubMed] [Google Scholar]
- 9.Daix T, Guerin E, Tavernier E, et al. Multicentric standardized flow cytometry routine assessment of patients with sepsis to predict clinical worsening. Chest. 2018;154(3):617–627. doi: 10.1016/j.chest.2018.03.058. [DOI] [PubMed] [Google Scholar]
- 10.Shankar-Hari M, Datta D, Wilson J, et al. Early PREdiction of sepsis using leukocyte surface biomarkers: The ExPRES-sepsis cohort study. Intensive Care Med. 2018;44(11):1836–1848. doi: 10.1007/s00134-018-5389-0. [DOI] [PubMed] [Google Scholar]
- 11.Caruso A, Licenziati S, Corulli M, et al. Flow cytometric analysis of activation markers on stimulated T cells and their correlation with cell proliferation. Cytometry. 1997;27(1):71–76. doi: 10.1002/(sici)1097-0320(19970101)27:1<71::aid-cyto9>3.0.co;2-o. [DOI] [PubMed] [Google Scholar]
- 12.Revenfeld ALS, Steffensen R, Pugholm LH, Jørgensen MM, Stensballe A, Varming K. Presence of HLA-DR molecules and HLA-DRB1 mRNA in circulating CD4(+) T cells. Scand J Immunol. 2016;84(4):211–221. doi: 10.1111/sji.12462. [DOI] [PubMed] [Google Scholar]
- 13.Hoffmann JJML. Neutrophil CD64: A diagnostic marker for infection and sepsis. Clin Chem Lab Med. 2009;47(8):903–916. doi: 10.1515/CCLM.2009.224. [DOI] [PubMed] [Google Scholar]
- 14.Elghetany MT. Surface antigen changes during normal neutrophilic development: A critical review. Blood Cells Mol Dis. 2002;28(2):260–274. doi: 10.1006/bcmd.2002.0513. [DOI] [PubMed] [Google Scholar]
- 15.Javed A, Guirgis FW, Sterling SA, et al. Clinical predictors of early death from sepsis. J Crit Care. 2017;42:30–34. doi: 10.1016/j.jcrc.2017.06.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Karakike E, Giamarellos-Bourboulis EJ. Macrophage activation-like syndrome: A distinct entity leading to early death in sepsis. Front Immunol. 2019;10:55. doi: 10.3389/fimmu.2019.00055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Daviaud F, Grimaldi D, Dechartres A, et al. Timing and causes of death in septic shock. Ann Intensive Care. 2015;5(1):16. doi: 10.1186/s13613-015-0058-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mare TA, Treacher DF, Shankar-Hari M, et al. The diagnostic and prognostic significance of monitoring blood levels of immature neutrophils in patients with systemic inflammation. Crit Care. 2015;19(1):57. doi: 10.1186/s13054-015-0778-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Hanna MOF, Abdelhameed AM, Abou-Elalla AA, Hassan RM, Kostandi I. Neutrophil and monocyte receptor expression in patients with sepsis: Implications for diagnosis and prognosis of sepsis. Pathog Dis. 2019;77(6):ftz055. doi: 10.1093/femspd/ftz055. [DOI] [PubMed] [Google Scholar]
- 20.Demaret J, Venet F, Friggeri A, et al. Marked alterations of neutrophil functions during sepsis-induced immunosuppression. J Leukoc Biol. 2015;98(6):1081–1090. doi: 10.1189/jlb.4A0415-168RR. [DOI] [PubMed] [Google Scholar]
- 21.Reddy M, Eirikis E, Davis C, Davis HM, Prabhakar U. Comparative analysis of lymphocyte activation marker expression and cytokine secretion profile in stimulated human peripheral blood mononuclear cell cultures: An in vitro model to monitor cellular immune function. J Immunol Methods. 2004;293(1-2):127–142. doi: 10.1016/j.jim.2004.07.006. [DOI] [PubMed] [Google Scholar]
- 22.Bajnok A, Ivanova M, Rigó J, Toldi G. The distribution of activation markers and selectins on peripheral T lymphocytes in preeclampsia. Mediators Inflamm. 2017;2017:8045161. doi: 10.1155/2017/8045161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Chen K, Zhou QX, Shan HW, Li WF, Lin ZF. Prognostic value of CD4 + CD25 + tregs as a valuable biomarker for patients with sepsis in ICU. World J Emerg Med. 2015;6(1):40. doi: 10.5847/wjem.j.1920-8642.2015.01.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Zhuang Y, Peng H, Chen Y, Zhou S, Chen Y. Dynamic monitoring of monocyte HLA-DR expression for the diagnosis, prognosis, and prediction of sepsis. Front Biosci (Landmark Ed) 2017;22(8):1344–1354. doi: 10.2741/4547. [DOI] [PubMed] [Google Scholar]
- 25.Cajander S, Rasmussen G, Tina E, et al. Dynamics of monocytic HLA-DR expression differs between bacterial etiologies during the course of bloodstream infection. PLoS ONE. 2018;13(2):e0192883. doi: 10.1371/journal.pone.0192883. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Brenneis M, Aghajaanpour R, Knape T, et al. Pparγ expression in T cells as a prognostic marker of sepsis. Shock. 2016;45(6):591–597. doi: 10.1097/SHK.0000000000000568. [DOI] [PubMed] [Google Scholar]
- 27.Le Tulzo Y, Pangault C, Gacouin A, et al. Early circulating lymphocyte apoptosis in human septic shock is associated with poor outcome. Shock. 2002;18(6):487. doi: 10.1097/00024382-200212000-00001. [DOI] [PubMed] [Google Scholar]
- 28.Duan S, Jiao Y, Wang J, et al. Impaired B-cell maturation contributes to reduced B cell numbers and poor prognosis in sepsis. SHOCK. 2020;54(1):70–77. doi: 10.1097/SHK.0000000000001478. [DOI] [PubMed] [Google Scholar]
- 29.Hohlstein P, Gussen H, Bartneck M, et al. Prognostic Relevance of Altered Lymphocyte Subpopulations in Critical Illness and Sepsis. J Clin Med. 2019;8(3):353. doi: 10.3390/jcm8030353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.van der Poll T, van de Veerdonk FL, Scicluna BP, Netea MG. The immunopathology of sepsis and potential therapeutic targets. Nat Rev Immunol. 2017;17(7):407–420. doi: 10.1038/nri.2017.36. [DOI] [PubMed] [Google Scholar]
- 31.Gupta S, Sakhuja A, Kumar G, McGrath E, Nanchal RS, Kashani KB. Culture-Negative severe sepsis: Nationwide trends and outcomes. CHEST. 2016;150(6):1251–1259. doi: 10.1016/j.chest.2016.08.1460. [DOI] [PubMed] [Google Scholar]
- 32.Li Y, Guo J, Yang H, Li H, Shen Y, Zhang D. Comparison of culture-negative and culture-positive sepsis or septic shock: A systematic review and meta-analysis. Crit Care. 2021;25(1):167. doi: 10.1186/s13054-021-03592-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.