Leukocyte fate in adult patients supported with extracorporeal membrane oxygenation: A retrospective cohort analysis

doi:10.21203/rs.3.rs-4778424/v1

Download PDF

Article

Leukocyte fate in adult patients supported with extracorporeal membrane oxygenation: A retrospective cohort analysis

https://doi.org/10.21203/rs.3.rs-4778424/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Introduction

Extracorporeal membrane oxygenation (ECMO) is known to influence all blood components. Reduction in leukocyte numbers during ECMO and their slow recovery ECMO have been associated with poorer prognosis. However, few studies on leukocyte fate have been conducted on adult patients on ECMO and are predominately cardiogenic shock-specific cohorts. Here, we attempt to examine the leukocyte profiles of ECMO-supported adult patients with both heart and/or lung failure and their associations with mortality and morbidity.

Methods

This multicenter, retrospective study included adult patients with refractory cardiac and/or respiratory failure supported by veno-arterial (VA) and veno-venous (VV) ECMO between 2016 and 2017. Data were collected from intensive care units of five ECMO centers in Australia, Italy, Japan, Hong Kong, and Germany. The primary outcome was the temporal trend of differential peripheral blood leukocyte numbers pre, during and post ECMO cannulation and survival in patients receiving venovenous and/or venoarterial ECMO. In addition, we evaluated the associations between leukocyte numbers and bleeding, infection, and organ dysfunction.

Results

Among 164 ECMO patients, mean age was 51 ± 16 years, and 67.7% of patients were male. 58.5% were placed on VA-ECMO, 39% on VV-ECMO, and 2.4% on VA/VV ECMO. Sixty-six patients who underwent ECMO (40.2%) did not survive hospitalization, and 96.9% of deaths occurred during ICU stay. In univariate analysis, a lower monocyte count (HR 0.45, 95% CI 0.21–0.93, p = 0.032), lower platelet count (HR 0.99, 95% CI 0.99-1.00, p = 0.009), higher lymphocyte count (HR 1.10, 95% CI 1.007–1.19, p = 0.033) and higher International Normalised Ratio (HR 3.98, 95% CI 2.64–5.99, p < 0.001) peri-ECMO were associated with increased risk of death. An elevated neutrophil count (HR 1.19, 95% CI 1.04–1.36, p = 0.013), age and lactate dehydrogenase were associated with mortality in multivariate analysis. There were no correlations between leukocyte variables and the development of infectious or bleeding complications. Integrated Discrimination Improvement index showed that SAPS II score with the addition of peri-ECMO lymphocyte (p = 0.001) or monocyte (p < 0.001) numbers have a better predictive value for death in ICU than SAPS II score alone.

Conclusions

Assessment of ECMO-related monocyte and lymphocyte numeric changes may be useful outcome prognosticators when used in conjunction with SAPS II score. Further investigation with larger patient cohorts will be required.

Biological sciences/Cell biology

Biological sciences/Immunology

Biological sciences/Physiology

Health sciences/Biomarkers

Health sciences/Cardiology

Health sciences/Diseases

Health sciences/Health care

Health sciences/Medical research

Health sciences/Pathogenesis

Health sciences/Risk factors

Extracorporeal membrane oxygenation

leukocytes

cardiac failure

respiratory failure

prognosis

Extracorporeal membrane oxygenation (ECMO) is a lifesaving intervention for treating critically ill patients with severe cardiac and/or pulmonary dysfunction^1–3. However, it is associated with an overall hospital mortality of 42% and high adverse event rates despite advances in management and technology^4–6. These complications include the development of bleeding, systemic inflammatory response syndrome (SIRS), hospital-acquired infections, sepsis and other deleterious complications (e.g., multiorgan dysfunction (MOD))^4–6. While extensive research has demonstrated a correlation between modulated inflammatory response and poor patient prognosis, few inflammatory biomarkers are utilized in routine clinical practice. Identification and quantification of standard-of-care factors such as leukocyte numbers (fate) may be useful as an early predictive tool for risk stratification of ECMO-related mortality and morbidity.

Leukocytes are central to innate and adaptive immunity in the regulation and stimulation of the inflammatory reaction that protects against illness⁷. In critically ill patients with acute respiratory distress syndrome (ARDS), cardiovascular diseases and cancer, changes in the neutrophil‒lymphocyte ratio have been proposed as useful inflammatory markers and prognostic tools for mortality^8–11. Yost and colleagues (2018) reported similar outcomes in ECMO patients with cardiogenic shock¹². In addition, an association between reduction in T lymphocyte count during ECMO and a slow recovery of T lymphocyte and monocyte numbers following intervention with increased patient mortality has also been reported in a clinical setting^13–15.

Although studies have evaluated the effect of ECMO on the fate of different leukocyte populations, there has been limited investigation in adults^12–15 as most studies have been conducted in the pediatric population^16–22. Amongst the current literature on leukocytes in adult ECMO patients, there is a large focus on its association with cardiogenic shock and mortality as a primary outcome. Importantly, veno-venous (VV) ECMO tends to be more commonly used in infectious conditions, leading to specific accompanying changes, whereas veno-arterial (VA) ECMO typically leans toward noninfective conditions. In this study, we aimed to further examine the numeric changes in leukocytes during ECMO and how they are correlated with mortality and complications in adult patients who underwent veno-arterial (VA) and/or veno-venous (VV) ECMO.

Study Design

This multicenter retrospective cohort study included adult patients who underwent ECMO for refractory cardiac and respiratory failure. Data were collected from intensive care units (Wang, #11) of The Prince Charles Hospital (TPCH, Queensland, Australia), Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlino (Milan, Italy), Teine Keijinkai Hospital (Sapporo, Japan), Queen Mary Hospital (Kowloon, Hong Kong) and University of Regensburg (Regensburg, Germany). The study protocol was approved by the Metro North Ethics Committee (HREC/17/QPCH/425), the University of Queensland Human Ethics Research Committee (2018000099), Institutional Review Board of the University of Hong Kong and Hospital Authority Hong Kong West Cluster (UW 20–259) and Teine Keijinkai Institutional Review Board (2–019240–00). Ethics approval was waived for Milano Policlinico Hospital and University of Regensburg. Given the retrospective nature of the study, it was deemed a neglible risk, and patient consent was waived at all participant centers.

Inclusion and exclusion criteria

Patients included in the study were ≥ 18 years of age and underwent VA and/or VV ECMO support between January 2016 and December 2017. Patients who did not have ≥ 3 days of differential leukocyte data during ECMO were excluded from this study.

Data collection

Data were collected retrospectively from electronic medical records. These included patient demographics (sex, age, body mass index (BMI) and comorbidities), the Simplified Acute Physiology (SAPS) II score, the Sequential Organ Failure Assessment (SOFA) score and Acute Physiology And Chronic Health Evaluation (APACHE) II scores on the day of ICU admission, ECMO settings and durations, and laboratory data pre-, peri- and post-ECMO. The laboratory data of particular interest included the differential leukocyte (total leukocyte, neutrophil, lymphocyte and monocyte) numbers collected before ECMO and during and after ECMO (immediately after ECMO, every day for 7 days and before discharge). In addition, data associated with clinical outcomes during intensive care unit (ICU) stays and at time of hospital discharge were collected, including survival rate, bleeding events, infection rate developed post-ECMO cannulation, organ dysfunction, and length of ICU/hospital stays.

Outcomes

The primary outcome was the temporal trend of differential peripheral blood leukocyte numbers pre, during and post ECMO cannulation, and survival in patients receiving venovenous and/or venoarterial ECMO. Secondary outcomes were the association of differential leukocyte numbers against complications and patient demographics.

Statistical analysis

Cox proportional hazard analysis was used to compute hazard ratios (HRs) and 95% confidence intervals (CIs) for predictors of death in ICU and/or in-hospital (including leukocytes), all adjusted for age. Significant parameters in univariate analysis were assessed via multivariate analysis. Association between leukocyte numbers and the clinical outcomes of infection or bleeding as complications of ECMO was assessed via univariate and multivariate logistic regression analyses. Outcomes were presented as odds ratios (ORs) and 95% CIs. To further evaluate the capacity of leukocyte numbers to predict patient survival in ICU and hospital, we used Integrated Discrimination Improvement (Dalton, #26) analysis, comparing the SAPS II score as the standard model with the SAPS II score and lymphocytes or monocytes during ECMO as the new model²³.

Clinical characteristics

A total of 533 patients who received ECMO support were identified from January 2016 to December 2017. Among the identified patients, 164 were over the age of 18 years, had more than 3 days of leukocyte data collected and, therefore, were included in our final analysis (Fig. 1). Mean age was of 50.8 ± 15.7 years, and 67.7% of the patients were male (Table 1). Sixty-six patients (40.2%) had prior cardiovascular disease, and 25 patients (15.2%) had chronic pulmonary diseases such as COPD, asthma, and cystic fibrosis (CF). Indicators for ECMO were predominantly associated with cardiogenic shock (17%), cardiopulmonary resuscitation (16.5%), postcardiac surgery (12.6%), respiratory failure (16.5%) and ARDS (8.4%) (Table 1 and Supplemental Tables S1 and S2).

Table 1

Characteristics and outcomes of survivors and nonsurvivors.
Variables	All patients (n = 164)	Survivors (n = 98)	Nonsurvivors (n = 66)
Baseline characteristics ^b Gender – Male no. (%)	111 (67.7%)	64 (65.3%)		47 (71.2%)
Age (year)	50.87 ± 15.72	47.81 ± 15.89		55.41 ± 14.43
BMI (kg/m)	26.35 ± 6.26	26.61 ± 7.09		25.95 ± 4.76
SAPS II score @ ICU admission	50.15 ± 19.24	46.65 ± 17.78		55.35 ± 20.26
APACHE II score @ ICU admission	26.2 ± 8.34	24.27 ± 7.11		29.27 ± 9.04
APACHE III score @ ICU admission	86.83 ± 25.88	83.73 ± 23.97		93.59 ± 29.25
SOFA score @ ICU admission	9.96 ± 3.61	9.95 ± 3.37		9.97 ± 3.99
Comorbidities – no. (%) ^a
Cardiovascular disease	66 (40.2%)	38 (38.8%)	28 (42.4%)
Chronic pulmonary disease	25 (15.2%)	18 (18.4%)	7 (10.6%)
Immunosuppression	11 (6.7%)	5 (5.1%)	6 (9.1%)
Chronic liver disease	5 (3%)	4 (4.1%)	1 (1.5%)
Chronic renal failure	5 (3%)	3 (3.1%)	2 (3.0%)
ICU admission diagnosis – no. (%) ^a Cardiac Respiratory Liver Sepsis Unknown^c	92 (56.1%) 63 (38.4%) 3 (1.8%) 5 (3.0%) 1 (0.6%)	43 (43.9%) 50 (51%) 2 (2%) 2 (2%) 1 (1%)	49 (74.2%) 13 (19.7%) 1 (1.5%) 3 (4.5%) N/A
ECMO mode – no. (%) ^a VA VV VA/VV Peripheral Central Peripheral/Central	96 (58.5%) 64 (39%) 4 (2.4%) 150 (91.5%) 10 (6.1%) 4 (2.4%)	46 (46.9%) 50 (51%) 2 (2%) 92 (93.9%) 4 (4.1%) 2 (2%)	50 (75.8%) 14 (21.2%) 2 (3.0%) 58 (87.9%) 6 (9.1%) 2 (3.0%)
ECMO duration (days) ^b VA VV Outcomes^b	11.05 ± 13.49 6.59 ± 5.15 17.29 ± 18.69	10.74 ± 11.13 6.50 ± 4.23 14.56 ± 14.07	11.52 ± 16.47 6.67 ± 5.95 25.8 ± 27.65
Length of ICU stays (days)	21.82 ± 22.50	26.43 ± 22.35	14.98 ± 20.26
Length of hospital stays (days)	34.09 ± 38.20	44.26 ± 39.96	18.98 ± 29.81
ICU mortality rate In-hospital mortality rate	64 (96.97%) 66 (100%)	N/A N/A	64 (96.97%) 66 (100%)
^ano. (%) values are the frequencies and percentages.
^bPlus-minus values are presented as means ± standard deviations.
^aCorresponding centers did not record indications for the specific patients during data collection.

Abbreviations: BMI, body mass index; ICU, intensive care unit; SAPS II, simplified acute physiology score; APACHE, Acute Physiology and Chronic Health Evaluation; SOFA, Sequential Organ Failure Assessment; N/A, not applicable; VA-ECMO, veno-arterial extracorporeal membrane oxygenation; V-V ECMO, veno-venous extracorporeal membrane oxygenation.

In our study, peripheral VA ECMO remained the most common configuration (Table 1). Fifty-two percent of patients who underwent VA ECMO died, whereas 21.8% of patients who underwent V-V ECMO died. Fifty percent of patients who underwent VA‒VV ECMO died, although only 4 patients were included in this study. Nonsurvivors were slightly older, with generally higher SAPS II, APACHE II, and III scores; however, no statistically significant differences were observed between the ECMO groups. All nonsurvivors died during their in-hospital stays, with 96.97% occurring in the ICU and 3.03% following transfer from ICU. Most deaths occurred by day 15 of ICU stay.

Monocyte, lymphocyte, neutrophil and platelet numbers during ECMO associated with mortality

Survivorship was associated with a higher number of monocytes during and post-ECMO (Fig. 2B) compared to nonsurvivors. Platelets were higher in the survivor group than in the nonsurvivor group pre-ECMO, and this difference increased during and post-ECMO (Fig. 2E). Patients who survived also had lower lymphocyte numbers at time of cannulation with a slow rise peri-ECMO in comparison to those who did not (Fig. 2D). No significant differences were observed in the total leukocyte population (Fig. 2A) or neutrophils (Fig. 2C) between survivors and nonsurvivors over time.

In the univariate Cox proportional hazard analysis, higher monocyte counts and higher platelet counts peri-ECMO were independently associated with decreased mortality (HR 0.45; 95% CI 0.21–0.93, p = 0.032 and HR 0.99; 95% CI 0.99–1.00, p = 0.009, respectively). A lower lymphocyte count was associated with increased mortality (HR 1.095 (95% CI = 1.007–1.19, p = 0.033)) (Table 2). No obvious changes were observed in the total leukocyte population or neutrophils between survivors over time. In addition, several independent variables were significant in determining patient mortality risk (Table 3). These included patient age, ICU admission scores (SAPS II and APACHE II scores), SOFA scores following ECMO cannulation, fraction of inspired oxygen (FiO₂), partial pressure of oxygen (PO₂) and lactate dehydrogenase (LDH) levels peri-ECMO. While bicarbonate was reported with an HR of 0.828, the variable was considered significant (95% CI 0.78–0.88, p < 0.001). In the multivariate analysis, the HRs of age and LDH peri-ECMO for mortality were 1.079 (95% CI = 1.024–1.14, p = 0.004) and 1.001 (95% CI = 1.000–1.002, p = 0.043), respectively (Table 2).

Table 2

Univariable and multivariable Cox proportional hazard analyses for survival in the ICU and hospital.
	Univariate					Multivariate
	HR	95% CI for HR			P value	HR	95% CI for HR		P value
		Lower	Upper				Lower	Upper
Age ^a	1.03	1.008	1.043	0.003		1.079	1.024	1.14	0.004
SAPS II Score (ICU Admission) ^a	1.02	1.004	1.033	0.013		0.99	0.94	1.051	0.82
APACHE II Score (ICU Admission) ^a	1.07	1.035	1.109	< 0.001		0.99	0.88	1.13	0.92
SOFA Score (ECMO Day) ^a	1.13	1.035	1.236	0.007		1.095	0.85	1.41	0.49
ECMO Days ^b	1.00	1.00	1.02	0.85
FiO2	1.019	1.005	1.032	0.006		1.010	0.98	1.045	0.55
PO₂	1.007	1.005	1.010	< 0.001		1.00	0.99	1.007	0.69
Bicarbonate	0.83	0.78	0.88	< 0.001		0.91	0.76	1.094	0.31
Lactate	1.009	1.000	1.018	0.059
LDH	1.001	1.001	1.001	< 0.001		1.001	1.000	1.002	0.043
Hemoglobin	0.99	0.98	1.002	0.12
Total Leukocytes	0.98	0.93	1.021	0.29
Monocytes	0.45	0.21	0.93	0.032		2.45	0.39	15.52	0.34
Neutrophils	0.97	0.93	1.019	0.23
Lymphocytes	1.10	1.007	1.19	0.033		1.23	0.89	1.70	0.20
Platelets	0.99	0.99	1.00	0.009		1.003	0.99	1.02	0.63
All parameters were analyzed via mean values from peri-ECMO, unless otherwise stated.
^a Data were collected at ICU admission or on the day of ECMO cannulation.
^b Data represent the total number of days patients were on ECMO.
Abbreviations: CI, confidence interval; HR, hazard ratio; APACHE II, Acute Physiology And Chronic Health Evaluation II; ECMO, extracorporeal membrane oxygenation; FiO2, fraction of acquired oxygen; ICU, intensive care unit; INR, international normalization ratio; LDH, lactate dehydrogenase; PO2, partial pressure of oxygen; SAPS II, simplified acute physiology score I; SOFA, Sequential Organ Failure Assessment.

Table 3

Univariable and multivariable Logistic Regression analysis for infectious complications.
	Univariate					Multivariate
	OR	95% CI for OR		P value		OR	95% CI for OR		P value
		Lower	Upper				Lower	Upper
Age ^a	0.986	0.967	1.006		0.163
SAPS II Score (ICU Admission) ^a	1.020	1.003	1.037		0.018	1.036	0.993	1.081	0.10
APACHE II Score (ICU Admission) ^a	0.997	0.954	1.042		0.888
SOFA Score (ICU Admission) ^a	0.926	0.847	1.013		0.092
SOFA Score (ECMO Day) ^a	0.936	0.827	1.059		0.282
Immunosuppressant Use (Post-ICU Admission) ^b	1.285	0.562	0.294		0.552
Antibiotics Use (Post-ICU Admission) ^b	0.503	0.184	1.370		0.179
ECMO Days ^c	1.115	1.052	1.183		< .001	1.279	1.006	1.626	0.04
FiO2	1.018	1.003	1.032		0.016	0.996	0.966	1.027	0.81
ECMO Flow Rates	1.438	1.009	2.048		0.044	0.966	0.646	1.444	0.87
MAP	1.024	0.059	0.999		1.050
HR	1.032	1.008	1.056		0.008	1.033	0.978	1.090	0.24
PO₂	1.001	0.997	1.004		0.740
Bicarbonate	1.007	0.939	1.080		0.838
Lactate peri ECMO	0.866	0.816	0.920		< .001	1.070	0.901	1.272	0.44
LDH	1.000	1.000	1.000		0.749
INR	0.747	0.390	1.431		0.379
Hemoglobin	0.999	0.984	1.015		0.941
Hemoglobin post ECMO	0.979	0.959	1.000		0.049	0.987	0.960	1.014	0.33
Total Leukocytes	1.011	0.960	1.064		0.679
Total Leukocytes post ECMO	0.981	0.917	1.049		0.570
Monocytes	0.794	0.369	1.710		0.555
Monocytes post ECMO	1.258	0.529	2.990		0.604
Neutrophils	1.010	0.959	1.065		0.697
Neutrophils post ECMO	0.993	0.918	1.074		0.860
Lymphocytes	0.927	0.827	1.039		0.195
Lymphocytes post ECMO	1.012	0.964	1.061		0.640
Platelets	0.998	0.993	1.003		0.443
Platelets post ECMO	1.005	1.001	1.008		0.014	1.002	0.996	1.008	0.54
All parameters were analyzed via mean values from peri-ECMO, unless otherwise stated.
^a Data were collected at ICU admission or on the day of ECMO cannulation.
^b Nominal data were collected post-ICU admission.
^c Data represent the total number of days patients were on ECMO.
Abbreviations: OR, odds ratio; CI, confidence interval; APACHE II, Acute Physiology and Chronic Health Evaluation II; ECMO, extracorporeal membrane oxygenation; FiO2, fraction of acquired oxygen; ICU, intensive care unit; INR, international normalization ratio; LDH, lactate dehydrogenase; PO2, partial pressure of oxygen; SAPS II, simplified acute physiology score I; SOFA, Sequential Organ Failure Assessment.

Factors associated with infectious complications

Among the 164 patients, 88 (53.65%) developed infection during ECMO. Leukocyte number was not a significant independent variable for infectious complications in patients undergoing ECMO. As shown in Table 3, the risk for infectious complications increased with higher SAPS II at ICU admission (OR 1.020, 95% CI 1.003–1.037; p = 0.018). Univariate analysis also revealed that ECMO duration (OR 1.152, 95% CI 1.052–1.183, p < 0.001), FiO₂ requirement (OR 1.018, 95% CI 1.003–1.032, p < 0.016), ECMO flow (OR 1.438, 95% CI 1.009–2.048, p < 0.044), peri-ECMO lactate level (OR 0.86, 95% CI 0.816–0.920, p < 0.001), peri-ECMO HR (OR 1.032, 95% CI 1.008–1.056, p < 0.008), post-ECMO hemoglobin (OR 0.979, 95% CI 0.959–1.000, p < 0.049), and post-ECMO platelets (OR 1.005, 95% CI 1.001–1.008, p < 0.014) were correlated with infectious complications. Only the number of ECMO days remained significant after multivariate analysis.

Factors associated with bleeding complications

Among the 164 patients, 52 (31.70%) developed bleeding complications during ECMO. Leukocyte number did not correlate with bleeding complications in patients undergoing ECMO. As shown in Table 4, the only predictor for bleeding complications was a lower platelet count peri-ECMO.

Table 4

Univariable logistic regression analysis for bleeding complications
	Univariate
	OR	95% CI for OR	P value
		Lower	Upper
Age ^a	0.997	0.976	1.018	0.764
SAPS II Score (ICU Admission) ^a	1.007	0.990	1.025	0.411
APACHE II Score (ICU Admission) ^a	1.012	0.969	1.057	0.596
SOFA Score (ICU Admission) ^a	0.976	0.889	1.072	0.612
SOFA Score (ECMO Day) ^a	1.070	0.948	1.208	0.272
Immunosuppressant Use (Post-ICU Admission) ^b	0.503	0.216	1.170	0.111
Antibiotics Use (Post-ICU Admission) ^b	0.577	0.180	1.849	0.336
ECMO Days ^c	1.018	0.994	1.044	0.143
FiO2	1.001	0.987	1.016	0.851
ECMO Flow Rates	1.050	0.950	1.161	0.336
MAP	1.005	0.858	1.459	0.408
HR	1.017	0.995	1.040	0.133
PO₂	1.002	0.998	1.005	0.354
Bicarbonate	0.982	0.911	1.058	0.636
Lactate ECMO	0.977	0.950	1.004	0.098
LDH	1.000	1.000	1.000	0.989
INR	0.842	0.411	1.724	0.638
Hemoglobin	0.995	0.979	1.011	0.554
Hemoglobin post ECMO	1.008	0.986	1.030	0.473
Total Leukocytes	1.038	0.984	1.095	0.176
Total Leukocytes post ECMO	1.016	0.943	1.093	0.683
Monocytes	0.918	0.400	2.105	0.840
Monocytes post ECMO	0.363	0.116	1.138	0.082
Neutrophils	1.047	0.991	1.106	0.105
Neutrophils post ECMO	1.052	0.965	1.148	0.252
Lymphocytes	0.782	0.525	1.166	0.228
Lymphocytes post ECMO	0.905	0.706	1.159	0.429
Platelets	0.993	0.987	0.999	0.023
Platelets post ECMO	0.999	0.995	1.003	0.563
All parameters were analyzed via mean values from peri-ECMO, unless otherwise stated.
^a Data were collected at ICU admission or on the day of ECMO cannulation.
^b Nominal data were collected post-ICU admission.
^c Data represent the total number of days patients were on ECMO.
Abbreviations: OR, odds ratio; CI, confidence interval; APACHE II, Acute Physiology and Chronic Health Evaluation II; ECMO, extracorporeal membrane oxygenation; FiO2, fraction of acquired oxygen; ICU, intensive care unit; INR, international normalization ratio; LDH, lactate dehydrogenase; PO2, partial pressure of oxygen; SAPS II, simplified acute physiology score I; SOFA, Sequential Organ Failure Assessment.

Monocyte and lymphocyte counts improved the ability of the SAPS II score to predict death via integrated discrimination improvement (IDI)

IDI analysis was used to evaluate the capacity of SAPS II score versus SAPS II score in conjunction with different leukocyte counts to predict survival in ICU and/or in-hospital. Compared with the SAPS score alone, lymphocyte number during ECMO in conjunction with SAPS II score yielded an IDI of 0.132 (p = 0.0013). Similar outcomes were observed when SAPS II score was compared against SAPS II in conjunction with monocytes during ECMO with an IDI of 0.727 (p < 0.001).

Exposure of blood to foreign surfaces and the nonphysiological shear stress of ECMO cause blood damage. Reduction in leukocyte numbers during ECMO and their slow recovery post-ECMO have been associated with poorer prognosis^{8–11,13−15}. However, studies reporting on leukocyte fate in adults are rare and have focused largely on cardiogenic shock-specific cohorts. Our study reported an increased risk for mortality in patients who independently exhibited differences in monocyte, lymphocyte and platelet counts during ECMO. Notably, our study is the first to suggest that SAPS II score combined with monocyte or lymphocyte numbers during ECMO has better predictive value for death in ICU than SAPS II score alone.

Among the leukocyte subsets, monocyte and lymphocyte numbers during ECMO have been associated with mortality in both present and past studies^13–15. We report an increased risk for mortality in both cardiac and respiratory failure patients when they independently exhibited a decrease in monocyte numbers, as well as platelet numbers, during ECMO. A higher lymphocyte number was also observed in nonsurvivors than in survivors. In addition to the cellular numeric changes, high HRs were also revealed with significance for age, SAPS II score, APACHE II score, SOFA score, FiO2, PO2, bicarbonate and LDH. This finding is in accordance with reports that slow or delayed recovery of monocytes is correlated with poorer outcomes and mortality in both VA and VV ECMO patients¹⁵. Specifically, low number of cluster differentiation (CD)14⁺CD16⁺ and CD14⁺ toll-like receptor (TLR)4⁺ monocyte subsets has been linked to increased mortality¹⁵. Monocytes are important innate immune cells that initiate rapid immune response and help direct downstream adaptive immunity. Therefore, reduced number of monocytes may alter the rapid immune response of patients and their downstream communication with the adaptive immune system, resulting further complications that increase the risk of death. Despite it being well-recognised that veno-venous (VV) ECMO tends to be more commonly used in infectious conditions, resulting in specific accompanying changes, and veno-arterial (VA) ECMO typically leans toward noninfective conditions, it is interesting to note that monocyte numbers do not seem to be directly associated with infection. Instead, they may indicate a correlation with the patient's specific response to the inflammatory excess induced by ECMO or, alternatively, with their “immunological reserve”.

Additionally, in contrast to Hong et al. (2015), we did not observe a reduction in overall lymphocytes during ECMO. We also observed a trend toward higher lymphocyte levels in patients who did not survive. As described by Hong and colleagues, the overall lymphocyte count was significantly lower after 2 hours of ECMO than when ECMO was cannulated. The authors further subcategorized the changes into reduction in CD4 + T (helper) cells after 2 h, T helper lymphocytes at 6 h, and overall T lymphocytes at 24 h with increased mortality, alongside a number of other markers¹³. The variation in the results may be due to the timing in which the data are recorded, where we reported changes every 24 hours rather than the earlier time points within the 24 hours of ECMO cannulation. Therefore, the significant drop in lymphocyte numbers may have been missed in the present study. The decreasing trend and slow recovery of lymphocyte numbers following ECMO withdrawal were also associated with mortality¹⁴. A sharp decline in lymphocyte numbers in the recovery phase post-ECMO in nonsurvivors was also observed in the present study. However, no statistical significance was reported, as the data were likely underpowered. As lymphocytes are essential to the adaptive immune system, if they become dysregulated, the capacity of the immune system to distinguish self from foreign and provide specialized immune defense can be hindered, potentially attributed to mortality. In conjunction with monocyte and lymphocyte changes, older patients; higher SAPS II, APACHE II and SOFA scores at the time of ICU admission; and higher FiO2, PO_2, LDH, and INR during ECMO also contributed to increased mortality. Interestingly, in our study population, differences in lactate levels were not associated with mortality, which is a frequent marker of end-organ perfusion used in the ICU.

We further investigated predictive capacity of SAPS II alone and SAPS II in conjunction with monocytes or lymphocytes to predict death in ECMO patients. Given that SAPS II has long been used to estimate mortality in ICU, specifically for parameters such as age, heart rate, systolic BP, temperature, PaO₂/FiO₂, total leukocyte numbers and more. Our IDI index analysis showed that the association of SAPS II with peri-ECMO monocyte or lymphocyte was able to better predict death in ICU than SAPS II alone. This data suggests that the addition of monocyte and lymphocyte numbers into the physiology score using IDI could potentially be incorporated as an easy-to-measure and inexpensive parameter for mortality prediction.

In addition to mortality, we propose that dysregulation of these closely regulated leukocyte subsets could play a crucial role in predicting infection and bleeding complications, which are frequently reported in ECMO patients^24–26. The reason is that leukocytes are central to the intricate communication between innate and adaptive immunity²⁷ as well as contributing to coagulation^28–32. However, we did not observe significant differences in ORs for either leukocytes or platelets between patients who developed infection and/or bleeding complications and those who did not. One possible cofounder is that some patients are placed on ECMO several days post initiation of infection, such as viral and bacterial pneumonia, and potential changes in their leukocyte profile could be secondary to the infection several days prior and not to ECMO. Similar to our results, Santiago-Lozano and colleagues (2018) were unable to find significant correlation between the overall leukocyte number alone and the number of patients who developed infections following VA ECMO cannulation³³. While direct correlation between leukocytes and bleeding complications has not been explored previously to warrant comparison, our study was unable to find any different correlation between these two variables. Of note, the potential underlying importance of leukocytes in ECMO-acquired immunoparalysis and hemostatic derangement should not be discounted. As shown by several studies, further phenotypic and functional leukocyte subset studies can provide more comprehensive additional information regarding the state of immune dysregulation in the absence of numeric modulations³⁴.

This study has several limitations. There was a relatively small sample size for subgroup analysis (VA ECMO versus VV ECMO) and tracking of leukocyte recovery, despite access to more than 500 patients. This was due to differential full blood counts not being a standard routine practice across all participating centers, resulting in n = 164 satisfying the selection criteria. As seen in previous publications, this study remained underpowered. Another important limitation was the large interval of data collection, especially within the first 24 hours of cannulation, which possibly hinders differences in lymphocyte analysis. The retrospective design of the study was also a limitation. There was significant heterogeneity in the way the different centers worked up ECMO complications, as there are no standardized criteria for doing so. For similar reasons, there was also heterogeneity in the different antimicrobial therapies that were chosen, which may have also been attributed in part to the differing microbial ecologies of the different centers. Finally, the use of the SAPS instead of ECMO-specific predictor scores such as the RESP or SAVE may have weakened the accuracy of our survival predictions.

Our study revealed that ECMO-related monocyte and lymphocyte numeric changes may be useful indicators for predicting death when used in conjunction with the SAPS II score. However, further investigation is required to access a larger patient leukocyte database and to carry out subanalyses between this heterogeneous VA and VV ECMO population.

Acknowledgment

The authors would like to thank The Prince Charles Hospital Adult Intensive Care Unit data management team for their assistance in the study.

Declarations of conflicts of interest

The authors declare that they have no conflicts of interest related to the authorship and/or publication of this article.

Data Availability Statement

Data is provided within the manuscript or supplementary information files.

Author Contributions

LH and KW contributed equally to this manuscript as co-first authors. KKK, KW, AC, MVM, JYS and JFF designed the study. LH and KKK drafted the manuscript. SV, PYN, TA, KM, MVM, IL, SMC, FS, MDF, TL, SH, SK, NB, NK, ED, JR, CA, JSJ, AG and JL acquired access to data and/or collected and collated the data. LH, KW and KKK analyzed and interpreted the data. The listed authors reviewed the manuscript and rewrote individual sections. All the authors read and approved the final manuscript.

Contributors

Fang Shu, MD⁶; Marianna Di Feliciantonio, MD⁹; Teresa Liew, MD¹; India Lye, RN, BN¹; Sam Huth, MD¹; Sarah King, MD¹; Nina Kaeser, MD⁸, Esther Dreier, MD⁸; Jessica Riordan¹; Carmen Ainola, RN^1,3; Jae-Seung Jung, MD, PhD^1,3; Ashlen Garret, BSc(Hons)¹; Jayshree Lavana, MD³.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Fraser JF, Shekar K, Diab S, et al. ECMO – the clinician’s view. ISBT Science Series. 2012;7(1):82–88.
Makdisi G, Wang Iw. Extra Corporeal Membrane Oxygenation (ECMO) review of a lifesaving technology. Journal of Thoracic Disease. 2015;7(7):E166-E176.
Millar JE, Gregory SD, C. SM, Bartlett RH, Fraser JF. The future of mechanical circulatory and respiratory support. In: Gregory SD, C. SM, Fraser JF, eds. Mechanical Circulatory and Respiratory Support Academic Press; 2017:1–36.
Marasco SF, Lukas G, McDonald M, McMillan J, Ihle B. Review of ECMO (extra corporeal membrane oxygenation) support in critically ill adult patients. Heart Lung Circ. 2008;17 Suppl 4:S41-47.
Thiagarajan RR, Barbaro RP, Rycus PT, et al. Extracorporeal Life Support Organization Registry International Report 2016. Asaio J. 2017;63(1):60–67.
Mosier JM, Kelsey M, Raz Y, et al. Extracorporeal membrane oxygenation (ECMO) for critically ill adults in the emergency department: history, current applications, and future directions. Critical Care. 2015;19:431.
Goldman AP, BS. Immunology Overview. In: Baron S, ed. Medical Microbiology (4th edition). 4th edition ed. Galveston: University of Texas Medical Branch; 1996.
Dell'Aquila E, Cremolini C, Zeppola T, et al. Prognostic and predictive role of neutrophil/lymphocytes ratio in metastatic colorectal cancer: a retrospective analysis of the TRIBE study by GONO. Ann Oncol. 2018;29(4):924–930.
Demirci NS, Erdem GU. Prognostic role of the neutrophil-to-lymphocyte ratio (NLR) in patients with operable ampullary carcinoma. Bosn J Basic Med Sci. 2017.
Li W, Ai X, Ni Y, Ye Z, Liang Z. The Association Between the Neutrophil-to-Lymphocyte Ratio and Mortality in Patients with Acute Respiratory Distress Syndrome: A Retrospective Cohort Study. Shock. 2018;in press.
Wang Y, Ju M, Chen C, et al. Neutrophil-to-lymphocyte ratio as a prognostic marker in acute respiratory distress syndrome patients: a retrospective study. J Thorac Dis. 2018;10(1):273–282.
Yost G, Bhat G, Pappas P, Tatooles A. The neutrophil to lymphocyte ratio in patients supported with extracorporeal membrane oxygenation. Perfusion.33(7):562–567.
Hong TH, Hu FC, Kuo SW, et al. Predicting outcome in patients under extracorporeal membrane oxygenation due to cardiogenic shock through dynamic change of lymphocytes and interleukins. IJC Metab Endocr 2015;7:36–44.
Kawahito K, Kobayashi E, Misawa Y, et al. Recovery from lymphocytopenia and prognosis after adult extracorporeal membrane oxygenation. Arch Surg. 1998;133(2):216–217.
Liu CH, Kuo SW, Ko WJ, et al. Early measurement of IL-10 predicts the outcomes of patients with acute respiratory distress syndrome receiving extracorporeal membrane oxygenation. Sci Rep. 2017;7(1):1021.
Hocker JR, Wellhausen SR, Ward RA, Simpson PM, Cook LN. Effect of extracorporeal membrane oxygenation on leukocyte function in neonates. Artif Organs. 1991;15(1):23–28.
Zach TL, Steinhorn RH, Georgieff MK, Mills MM, Green TP. Leukopenia associated with extracorporeal membrane oxygenation in newborn infants. J Pediatr. 1990;116(3):440–444.
DePalma L, Short BL, Meurs KV, Luban NLC. A flow cytometric analysis of lymphocyte subpopulations in neonates undergoing extracorporeal membrane oxygenation. J Pediatr. 1991;118(1):117–120.
Plotz FB, van Oeveren W, Bartlett RH, Wildevuur CR. Blood activation during neonatal extracorporeal life support. J Thorac Cardiovasc Surg. 1993;105(5):823–832.
DePuydt LE, Schuit KE, Smith SD. Effect of extracorporeal membrane oxygenation on neutrophil function in neonates. Crit Care Med. 1993;21(9):1324–1327.
Fortenberry JD, Bhardwaj V, Niemer P, Cornish JD, Wright JA, Bland L. Neutrophil and cytokine activation with neonatal extracorporeal membrane oxygenation. J Pediatr. 1996;128(5):670–678.
Ortega SB, Pandiyan P, Windsor J, et al. A Pilot Study Identifying Brain-Targeting Adaptive Immunity in Pediatric Extracorporeal Membrane Oxygenation Patients With Acquired Brain Injury. Crit Care Med. 2019;47(3):e206-e213.
Pencina MJ, D'Agostino RB, Sr., D'Agostino RB, Jr., Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med. 2008;27(2):157–172; discussion 207 – 112.
MacLaren G, Schlapbach LJ, Aiken AM. Nosocomial Infections During Extracorporeal Membrane Oxygenation in Neonatal, Pediatric, and Adult Patients: A Comprehensive Narrative Review. Pediatr Crit Care Med. 2020;21(3):283–290.
Extracorporeal Life Support Organization. ECLS Registry Report: Interntional Summary. Ann Arbor 2019.
Dalton HJ, Reeder R, Garcia-Filion P, et al. Factors Associated with Bleeding and Thrombosis in Children Receiving Extracorporeal Membrane Oxygenation. Am J Respir Crit Care Med. 2017;196(6):762–771.
Rich RR, Chaplin DD. 1 - The Human Immune Response. In: Rich RR, Fleisher TA, Shearer WT, Schroeder HW, Frew AJ, Weyand CM, eds. Clinical Immunology (Fifth Edition). London: Content Repository Only; 2019:3–17.e11.
Wang Y, Gao H, Shi C, et al. Leukocyte integrin Mac-1 regulates thrombosis via interaction with platelet GPIbα. Nat Commun. 2017;8:15559–15559.
Laurance S, Bertin F-R, Ebrahimian T, et al. Gas6 Promotes Inflammatory (CCR2 hi CX3CR1 lo) Monocyte Recruitment in Venous Thrombosis. Arterioscler Thromb Vasc Biol. 2017;37(7):1315–1322.
von Brühl M-L, Stark K, Steinhart A, et al. Monocytes, neutrophils, and platelets cooperate to initiate and propagate venous thrombosis in mice in vivo. J Exp Med. 2012;209(4):819–835.
Liu X, Xue Y, Ding T, Sun J. Enhancement of proinflammatory and procoagulant responses to silica particles by monocyte-endothelial cell interactions. Part. Fibre Toxicol. 2012;9(1):36.
Michelson AD, Barnard MR, Krueger LA, Valeri CR, Furman MI. Circulating monocyte-platelet aggregates are a more sensitive marker of in vivo platelet activation than platelet surface P-selectin: studies in baboons, human coronary intervention, and human acute myocardial infarction. Circulation. 2001;104(13):1533–1537.
Santiago-Lozano MJ, Barquín-Conde ML, Fuentes-Moreno L, et al. Infectious complications in paediatric patients treated with extracorporeal membrane oxygenation. Enferm Infecc Microbiol Clin. 2018;36(9):563–567.
Ki KK, Millar JE, Langguth D, et al. Current Understanding of Leukocyte Phenotypic and Functional Modulation During Extracorporeal Membrane Oxygenation: A Narrative Review. 2021;11(3396).

No competing interests reported.

HelmsetalSupplementaryTable8.1.docx

Download PDF

Reviews received at journal
06 Nov, 2024
Reviews received at journal
04 Nov, 2024
Reviewers agreed at journal
26 Oct, 2024
Reviewers agreed at journal
26 Oct, 2024
Reviewers invited by journal
24 Oct, 2024
Editor assigned by journal
23 Oct, 2024
Editor invited by journal
08 Aug, 2024
Submission checks completed at journal
07 Aug, 2024
First submitted to journal
21 Jul, 2024

You are reading this latest preprint version

Leukocyte fate in adult patients supported with extracorporeal membrane oxygenation: A retrospective cohort analysis

Status:

Version 1

Abstract

Figures

Introduction

Methods

Study Design

Inclusion and exclusion criteria

Data collection

Outcomes

Statistical analysis

Results

Clinical characteristics

Monocyte, lymphocyte, neutrophil and platelet numbers during ECMO associated with mortality

Factors associated with infectious complications

Factors associated with bleeding complications

Discussion

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1