Sage Journals: Discover world-class research

Abstract

Background: Veno-venous ECMO (VV ECMO) is a life-sustaining technique indicated in the most severe forms of acute respiratory distress syndrome (ARDS). VV ECMO remains associated with significant morbidity and mortality. In this context, the identification of prognostic factors during the first week following ECMO implantation seems of high interest. However, little is known about the prognostic values of the kinetics of clinical and biological parameters in this timeframe. Methods: Observational, retrospective, multicenter study conducted in three centers in France and Belgium. We enrolled patients aged 18 years or older who underwent VV ECMO for severe ARDS between March 2020 and January 2023, with at least one endocan assay within 7 days of VV ECMO implantation. Kinetics of endocan and of a panel of clinical and biological markers were compared between surviving and deceased patients. Results: Forty-nine patients were included in our study among whom 30 patients (61%) were deceased on discharge from intensive care. Compared with patients who died, surviving patients had lower membrane fraction of oxygen at D3, lower fresh gas flow at D7, higher tidal volumes at D1 and D7, higher lung compliance and mechanical power at D7, and lower SOFA score at D7. Compared with deceased patients, survivors had a higher pH and lower endocan concentration on D3, and lower fibrinogen on D3 and D7. Conclusion: One week after initiation of VV ECMO support for severe ARDS, there was a significant decrease in the level of support simultaneously with respiratory and overall improvement in surviving patients. Compared to patients discharged alive from ICU, non-survivors were characterized by lower concentrations of endocan concentration at D3.

Keywords

ARDS ECMO endocan prognosis inflammation biomarkers

Background

Despite its potential beneficial impact^1,2 in the treatment of severe acute respiratory distress syndrome (ARDS), veno-venous ECMO (VV ECMO) remains associated with high cost and mortality, with a high incidence of complications.¹ In this context, the early identification of patients on VV ECMO with a high probability of survival represents a major challenge. Numerous predictive scores have been developed to assess the risk of mortality under ECMO before its implantation.^3,4 However, few studies have investigated the dynamic impact on mortality of clinical ventilatory parameters, organ failure and biological markers during the first days after initiation of VV ECMO.

Endocan is a proteoglycan secreted by pulmonary endothelial cells in humans, described as an inflammatory marker of the lung, whose increase in concentration seems to be linked to the severity of ARDS.⁵ In a recent pilot study of ARDS patients receiving VV ECMO support for SARS-CoV-2 pneumonia, endocan concentration at D7 and the increase in endocan levels between D0 and D7 were significantly higher in non-surviving patients.⁶ However, the conclusions to be inferred from these results were limited due to the small sample size of the study.

Given the lack of data concerning the identification of dynamic prognostic factors in patients undergoing VV ECMO in the setting of severe ARDS, we conducted this retrospective multicenter study to compare, between surviving and deceased patients, the kinetics of levels of respiratory support, organ failure, and various biological markers, during the first week following VV ECMO implantation.

Methods

Study design

This retrospective, multicenter, observational study was conducted in the Department of Intensive Care of the Lille University Hospital (France), the Intensive Care Unit of the Foch Hospital in Suresnes (France) and the Intensive Care Unit of the Erasme Hospital in Brussels (Belgium). The patient recruitment period ran from March 2020 to January 2023.

Ethics statement

The collection of analyzed data was validated by the “Comité de Protection des Personnes Ile-de-France IV” (approval number: 2020/30) for patients from the Lille University Hospital, by the local ethics committee of the Foch Hospital (approval number: 20-07-15) for patients from the Foch Hospital in Suresnes, and by the local ethics committee of the Erasme Hospital (approval number: P2017/443) for patients from the Erasme Hospital in Brussels. Data collection and processing for this study have been declared to the French “Commission Nationale Informatique et Libertés”. Written informed consent to participate was obtained from the patients or from their relatives or trusted person in case written informed consent could not be obtained from study participants.

Inclusion and exclusion criteria

We included all patients admitted to intensive care in participating centers between March 2020 and January 2023, with the following inclusion criteria: age ≥18 years, VV ECMO support in the management of severe ARDS, and at least one endocan blood test within 7 days of VV ECMO placement.

Exclusion criteria included the need for veno-arterial ECMO support, and patients deceased within 24 h of VV ECMO implantation.

Data collection

Data collected and analyzed included: patients characteristics on ICU admission, treatments used prior to VV ECMO implantation, characteristics of the stay in ICU, VV ECMO parameters, respiratory parameters, SAPS II,⁷ SOFA,⁸ PRESERVE³ and RESP⁹ scores and biological data (see Supplemental Appendix 1) on the day of ECMO implantation (D0), at 24 h (D1), 72 h (D3) and on day 7 following VV ECMO implantation (D7). Oxygen and ventilation parameters were measured using routine monitoring on ventilators and ECMO device (see definitions in Supplemental Appendix 2). The settings of these parameters were not standardized and at the discretion of the physicians in charge of the patients. Endocan was measured using the JDIYEK® ELISA Kit (Biothelis, Lille, France) in all participant centers, following the manufacturer’s instructions. Techniques used for other biomarkers were not standardized across centers.

Objectives

The primary objective of the study was to compare, between patients discharged alive from the ICU (survivors) and those who were deceased on ICU discharge (non-survivors), the evolution between D0 and D7 of VV ECMO implantation for endocan and the following parameters: level of respiratory support, organ failure and biological markers.

Statistical analysis

Continuous variables were expressed as median and interquartile range. Categorical variables were expressed as numbers and percentages.

To compare the kinetics of levels of severity, respiratory support and biological markers between survivors and non-survivors, we used a linear mixed model with survival on ICU discharge as a fixed effect and time from VV ECMO implantation as a random effect. Comparisons between two quantitative variables at a given time were performed using the Mann-Whitney test. Categorical variables were compared by Chi2 or Fischer tests, as appropriate.

We did not perform any missing value imputation for continuous variables in this study.

All statistical tests were two-tailed, and p values <0.05 were considered statistically significant. Statistical analyses were performed using R version 4.1.2 (R foundation for statistical computing, Vienna, Austria) and GraphPad Prism version 9.5.2 (GraphPad Software, Boston, MA, USA).

Results

Patient characteristics

A total of 49 patients were included in the study, of whom 30 (61%) died in intensive care and 19 (39%) survived. For all patients’ characteristics on baseline, no significant difference was found between survivors and non-survivors (Table 1).

Table 1.

Patients’ characteristics before ECMO implantation.

	Survivors (n = 19)	Non-survivors (n = 30)	p
Age (years)	54 (43; 57)	52 (44; 60)	0.58
Gender (male)	15 (79)	25 (83)	0.99
Height (cm)	175 (169; 183)	173 (168; 176)	0.28
Weight (kg)	90 (83; 110)	95 (82; 100)	0.92
BMI (kg/m²)	30 (27; 36)	33 (28; 38)	0.59
Comorbidities
Hypertension	4 (25)	10 (33)	0.80
Diabetes	6 (38)	7 (23)	0.50
Dyslipidemia	7 (44)	5 (17)	0.10
Tobacco use	1 (6)	7 (23)	0.30
Obesity	9 (47)	19 (63)	0.42
Coronary disease	2 (13)	3 (10)	1.00
Peripheral arterial disease	1 (6)	1 (3)	1.00
COPD or asthma	2 (13)	5 (17)	1.00
Chronic respiratory failure	0 (0)	1 (3)	1.00
Cancer	1 (6)	1 (3)	1.00
Immunosuppression	1 (6)	5 (17)	0.59
Time from ICU admission to ECMO implantation (days)	5 (2; 15)	7 (4; 13)	0.68
Time from intubation to ECMO implantation (days)	2 (0.5; 8)	5 (1; 9)	0.28
Severity before VV ECMO implantation
Compliance (mL/cmH₂0)	22 (20; 24)	23 (17; 27)	0.62
PaO₂/FiO₂	66 (57; 79)	73 (66; 85)	0.27
PaCO₂ (mmHg)	54 (46; 57)	62 (50; 71)	0.07
Severity scores
SAPS II	56 (43; 65)	56 (49; 65)	0.84
SOFA	10 (8; 12)	10 (8; 12)	0.95
PRESERVE score	3 (0,5; 4.5)	2 (1; 5.5)	0.86
RESP score	0 (−5; 1.5)	−3 (−4; 0)	0.43
Etiology of ARDS
Non SARS-CoV-2 viral pneumonia	0 (0)	1 (3)	1
Bacterial pneumonia	2 (12)	4 (13)	1
SARS-CoV-2 pneumonia	14 (82)	27 (90)	0.76
Other	2 (11)	2 (7)	1

The results of quantitative variables are expressed as median and interquartile range. Categorical variables are expressed by their count and percentage of the total count excluding missing data. ARDS: Acute Respiratory Distress Syndrome, BMI: Body Mass Index, COPD: Chronic Obstructive Pulmonary Disease, ECMO: Extracorporeal Membrane Oxygenation, ICU: Intensive Care Unit.

Evolutive profile of respiratory support and organ failure under VV ECMO

There was a significant difference over the period D0 - D7 between survivors and non-survivors for kinetics of FmO2 and fresh gas flow (FGF) (Φ <0.05 for each). Compared to the non-survivors group, survivors had lower FmO2 at D3 (80 (60; 100) % vs. 100 (80; 100) % (p = 0.049)) and lower FGF at D7 (3^3,6 L/min vs 7^5,8 L/min (p 0.011)) (Figure 1).

Figure 1.

Evolution of VV ECMO parameters, respiratory parameters and SOFA score in survivors and non-survivors. Vt: tidal volume. Φ: comparison by linear mixed model. *: p < 0.05 by Mann-Whitney test.

In addition, there were significant differences in the kinetics between survivors and non-survivors over the period D0 - D7 for tidal volume (Vt), lung compliance and mechanical power (Φ <0.05 for each). Compared to non-survivors, survivors had higher Vt at D1 (240 (160; 300) mL vs. 180 (100; 235) mL (p = 0.031)) and D7 (300 (220; 408) mL vs. 170 (115; 260) mL (p < 10^-2)), higher lung compliance at D7 (25 (19; 34) mL/cm H₂O vs. 12 (9; 17) mL/cm H₂O (p = 10^-3)), and higher mechanical power at D7 (15 (8; 25) J/min vs. 6^4,10 J/min (p < 10^-2)) (Figure 1).

Significantly different kinetics were observed between the two groups for the SOFA score as well (Φ <0.05), which was found lower at D7 in survivors compared to non-survivors (8^7,11 vs 12^9,12 (p = 0.036)) (Figure 1).

Evolution of biological markers under VV ECMO

We found significant differences in the kinetics between survivors and non-survivors over the period D0 - D7 for endocan, pH, and fibrinogen (Φ <0.05). Compared to non-survivors, survivors had lower endocan levels at D3 (3.9 (3.5; 5.2) ng/mL vs. 10.8 (7.1; 14.7) ng/mL (p < 10^-3)), higher pH at D3 (7.46 (7.42; 7.49) vs. 7.4 (7.38; 7.44) (p = 0.017)), and lower fibrinogen at D3 (4.3^4,5 g/L vs 5.7 (4.4; 7.2) g/L (p = 0.027)) and D7 (4.3 (2.8; 4.7) g/L vs. 5 (4.3; 5.5) g/L (p = 0.026)) (Figure 2).

Figure 2.

Evolution of biological markers in survivors and non-survivors. Φ: comparison by linear mixed model. *: p < 0.05 by Mann-Whitney test.

Discussion

Our study compared the kinetics of the level of support, organ failures and biological markers up to D7 of VV ECMO implantation, between surviving and deceased ICU patients. Logically, survivors showed a rapid decline in the level of VV ECMO support, with FmO2 significantly lower at D3, and fresh gas flow significantly lower at D7, compared with non-survivors. At the same time, survivors showed respiratory improvement, with Vt, lung compliance and mechanical power significantly higher on D7; and an overall improvement in organ failure, as illustrated by the lower SOFA score, significantly lower on D7 compared with non-survivors.

A meta-analysis of observational studies of patients undergoing VV ECMO for refractory hypoxemia identified driving pressure during the first 3 days as the only ventilatory parameter independently associated with mortality.¹¹ Our study showed different results and did not find significantly different driving pressure kinetics between survivors and non-survivors between D0 and D7 of VV ECMO support, likely explained by the majoritarian use of APRV mode, based on the setting of a driving pressure <15 cm H₂O.

On the other hand, in our study, survivors showed a progressive improvement in pulmonary compliance from D3 onwards, with an increase in tidal volume and mechanical power. The improvement in compliance and the increase in Vt for a stable pressure regime in surviving patients, testifies to a favorable respiratory evolution and therefore appears consistent.

In non-survivors, a significantly lower level of compliance persisted at D7, compared with survivors. Furthermore, in deceased patients, the level of VV ECMO support at D7 was still maximal, and the FGF at D7 was significantly higher, compared with survivors. In view of these results, the question arises as to whether such invasive management should be continued in the event of a complete absence of respiratory improvement, with the need for maximum support, after 7 days of VV ECMO treatment.

Endocan is a biological marker of pulmonary stress.⁵ This biomarker has been studied for its prognostic value in situations of acute pulmonary inflammation.⁵ In the context of ARDS, two prospective observational studies have shown a link between a high initial endocan level and an unfavorable outcome in ICU. No patients were on VV ECMO in either study.^10,13 A third study on a cohort of patients hospitalized for SARS-CoV-2 pneumonia also found endocan on admission to be a factor associated with poor respiratory outcome.¹²

In a previous study, the authors reported significantly higher endocan concentration at 24 h from diagnosis in patients with a poor prognosis, defined either as death or a duration of mechanical ventilation of more than 10 days.¹⁰ In this study, endocan kinetics in the first 24 h were not associated with poor prognosis, and biomarker kinetics beyond 24 h were not assessed.

There thus appears to be an association between high endocan levels, and poor prognosis in ARDS patients.

Our study, focusing more specifically on biomarker kinetics, seemed to provide additional information, and suggested that the decrease of endocan at D3 of VV ECMO support was associated with a higher survival rate in the ICU. These results raise the question of the extent to which this marker of pulmonary endothelial stress could be considered as a clinical sensor of lung inflammation, whose early negativation could be an indicator of resolution of pulmonary injury. Lowering endocan levels could thus be a target to monitor in the most severe forms of ARDS, in order to guide the therapies undertaken. However, larger-scale studies confirming these results would appear to be necessary before considering the benefits of such a strategy.

Our study has several strong points. It is a multicenter study, spanning a relatively short period of less than 3 years, during which the frequency of VV ECMO use increased considerably as a result of the worldwide COVID-19 pandemic. The study thus provides a recent view of the type of patients benefiting from VV ECMO support in severe ARDS. The multicentric nature of this study also makes it more representative. Few studies have examined the evolution of respiratory and biological parameters following VV ECMO implantation. When such an assessment has been carried out, it has extended to a maximum of 72 h after VV ECMO implantation.¹¹ Our study, evaluating the kinetics of various parameters up to D7 of support initiation, provides additional information.

However, our work has a number of limitations. Firstly, selection bias and unmeasured confounders appear as key limitations inherent to the retrospective design of the study. Secondly, most of the patients in our study presented with ARDS secondary to SARS-CoV-2 infection. This viral infection has certain specific features. The generalizability of our results to patients with ARDS of other origins needs to be confirmed. Thirdly, the small size of our study may be the source of a sampling bias, and may affect the generalizability of our results, especially given the potential variability of VV-ECMO indications across centers. In addition, this prevented us from performing multivariate analyses that would have accounted for potential confounding factors. Furthermore, because of the limited sample size along with the exploratory nature of our study, the analysis did not include any corrections for multiple testing. However, the use of VV ECMO support remains relatively infrequent in intensive care units, making it difficult to conduct studies with a larger number of patients. Moreover, all the participating centers were tertiary care centers, which could be seen as a center-level bias that could influence the prognosis or management of study participants. Furthermore, no sample size calculation was performed for this study, which constitutes another limitation of this work. Another limitation relates to the fact that there were missing data, particularly concerning the endocan assay. Thus, the possibility that some endocan measurements were unavailable due to deaths occurring before the considered time points cannot be excluded. As an additional limitation, we must note that selecting the first week as the critical window for prognostic assessment of endocan as well as other parameters did not allow us to assess whether meaningful prognostic signals could occur beyond day 7. Further, the kinetics of respiratory mechanics and biomarkers (at the exception of endocan) were available to the physicians in charge of the patients, and may certainly have influenced the management of VV ECMO weaning, ventilator settings, limitations of life-sustaining, or at a greater scale, allocation of resources and decisions of closing care. We should mention as an important other limitation that, at the exception of endocan, biomarker assays were not standardized across centers, which might have altered the comparability of the results.

Conclusion

In patients undergoing VV ECMO for severe ARDS, a decrease in the level of VV ECMO assistance was observed in surviving patients during the first week. At the same time, survivors showed an overall improvement during the first week, with a progressive increase in pulmonary compliance up to D7 and a reduction in the SOFA score. In non-survivors, the level of support was still maximal and pulmonary compliance still collapsed at D7 of VV ECMO support. Survivors seemed to be characterized by a decrease in endocan blood levels at D3. These results need to be further validated in a large prospective multicenter cohort. Such a study should include additional markers of phenotype and severity of ARDS, including imaging patterns on lung CT, additional physiological data and biomarkers. If these exploratory results were to be reproduced and confirmed, repeated measurements of blood endocan could be considered as a way to monitor the response to treatments in subjects undergoing ECMO for severe ARDS. A strategy aimed at lowering endocan concentration at an early stage could be an avenue of research to improve the prognosis of ARDS patients undergoing VV ECMO.

Supplemental Material

Supplemental material - Assessment of endocan along with routine clinical and biological markers to characterize the evolutive profile of patients with severe ARDS under veno-venous ECMO

Supplemental material for Assessment of endocan along with routine clinical and biological markers to characterize the evolutive profile of patients with severe ARDS under veno-venous ECMO by Camille Prouteau, Julien Poissy, Tiffany Pascreau, Olivier Lheureux, Sylvain Dubucquoi, Victoria Dubar, Benjamin Zuber, Marc Vasse, Sylvain Normandin, Nicolas Dognon, Clémentine Levy, Thibault Duburcq, Alexandre Gaudet in European Journal of Inflammation.

Footnotes

ORCID iD

Alexandre Gaudet

Ethical considerations

Consent to participate

Written informed consent to participate was obtained from the patients or from their relatives or trusted person in case written informed consent could not be obtained from study participants.

Author contributions

CP, TP, OL, VD, BZ, SN, ND and CL collected the data. TP, SD and MV conducted the biological analysis. CP and AG performed the statistical analysis and drafted the manuscript. JP, TD and AG designed the study and supervised the project. All authors participated in reviewing. All authors read and approved the final version of the manuscript.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

AG received financial sponsorship from Pfizer for clinical research, and had attendance fees covered by MSD for a medical conference. The other authors have no conflicts of interest to disclose.

Data Availability Statement

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.*

Supplemental Material

Supplemental material for this article is available online.

Appendix

References

Combes

Hajage

Capellier

, et al. Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. N Engl J Med 2018; 378(21): 1965–1975.

Goligher

Tomlinson

Hajage

, et al. Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome and posterior probability of mortality benefit in a post hoc bayesian analysis of a randomized clinical trial. JAMA. 4 déc 2018; 320(21): 2251–2259.

Schmidt

Zogheib

Rozé

, et al. The PRESERVE mortality risk score and analysis of long-term outcomes after extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. Intensive Care Med 2013; 39(10): 1704–1713.

Rozencwajg

Pilcher

Combes

, et al. Outcomes and survival prediction models for severe adult acute respiratory distress syndrome treated with extracorporeal membrane oxygenation. Crit Care 2016; 20(1): 392.

De Freitas Caires

Gaudet

Portier

, et al. Endocan, sepsis, pneumonia, and acute respiratory distress syndrome. Crit Care 2018; 22(1): 280.

Levy

Dognon

Normandin

Duburcq

Gaudet

Lille Intensive Care COVID-19 group , et al. Assessment of plasma endocan for the prediction of mortality in COVID-19 patients undergoing veno-venous ECMO: a pilot study. Med Intensiva 2023; 47(1): 51–54.

Le Gall

Lemeshow

Saulnier

. A new simplified acute physiology score (SAPS II) based on a european/north American multicenter study. JAMA J Am Med Assoc. 22 déc 1993; 270(24): 2957–2963.

Vincent

Moreno

Takala

, et al. The SOFA (Sepsis-related organ failure assessment) score to describe organ dysfunction/failure. On behalf of the working group on sepsis-related problems of the european society of intensive care medicine. Intensive Care Med 1996; 22(7): 707–710.

Schmidt

Bailey

Sheldrake

, et al. Predicting survival after extracorporeal membrane oxygenation for severe acute respiratory failure. The respiratory extracorporeal membrane oxygenation survival prediction (RESP) score. Am J Respir Crit Care Med 1 juin 2014; 189(11): 1374–1382.

10.

Orbegozo

Rahmania

Irazabal

, et al. Endocan as an early biomarker of severity in patients with acute respiratory distress syndrome. Ann Intensive Care 2017; 7(1): 93.

11.

Serpa Neto

Schmidt

Azevedo

LCP

, et al. Associations between ventilator settings during extracorporeal membrane oxygenation for refractory hypoxemia and outcome in patients with acute respiratory distress syndrome: a pooled individual patient data analysis: mechanical ventilation during ECMO. Intensive Care Med 2016; 42(11): 1672–1684.

12.

Gaudet

Ghozlan

Dupont

, et al. Derivation and validation of a predictive score for respiratory failure worsening leading to secondary intubation in COVID-19: the CERES score. J Clin Med 13 avr 2022; 11(8): 2172.

13.

Tang

Zhao

Wang

, et al. Endocan levels in peripheral blood predict outcomes of acute respiratory distress syndrome. Mediat Inflamm 2014; 2014: 625180–625189.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.49 MB