Systemic Inflammatory Response Syndrome After Extracorporeal Membrane Oxygenation Decannulation in COVID-19 Patients

Introduction: Systemic inflammatory response syndrome (SIRS) is frequently observed following decannulation from extracorporeal membrane oxygenation (ECMO). Differentiating cytokine release due to infection from endothelial injury from cannula removal and/or discontinuation from the ECMO circuit has been shown to impact treatment and outcomes. This response, however, may be complicated in COVID-19 patients due to prevalent glucocorticoid and immune modulator use. It remains unclear whether COVID-19 infection and/or associated immune modulator use impact the incidence of SIRS following decannulation. Objectives: The aim of this study is to investigate the incidence of the SIRS phenomenon and associated outcomes in patients with COVID-19 after ECMO decannulation. Methods: An IRB-approved retrospective chart review of all patients who survived ECMO between June 31, 2010 and July 7, 2021 was done to identify patients who experienced SIRS within 48 hours of decannulation from ECMO support. Patients with COVID-19 were confirmed by a positive reverse transcription polymerase chain reaction (RT-PCR) assay for SARS-CoV2. SIRS was confirmed when two out of three of the following criteria were met: fever, leukocytosis, and/or initiation/escalation of vasopressors. Patients who developed post-ECMO SIRS were then distinguished based on the presence of infection. Infection was defined by the presence of either a new or positive culture following decannulation. We compared the incidence of SIRS and infection within 48 hours of decannulation in patients with and without COVID-19. Results: We identified 227 eligible patients who survived ECMO. Twenty-eight patients (12%) had COVID-19. Of these patients, ten patients with COVID-19 (36%) experienced post-ECMO SIRS, including those with true SIRS (n=3) and associated infections (n=7). Five of the ten patients with COVID-19 who experienced post-ECMO SIRS were exposed to immune modulators within two weeks of decannulation. Ninety-five (42%) patients without COVID-19 developed post-ECMO SIRS. Thirty-day survival in COVID patients who experienced post-ECMO SIRS compared to COVID patients who did not experience post-ECMO SIRS was 73% vs. 94%. (p=0.11). Conclusion: Post-ECMO SIRS is common. The incidence of SIRS following decannulation was similar when historically compared to non-COVID patients who survived ECMO in a previously reported cohort from our institution. Immune-modulation exposure within two weeks of decannulation did not affect the incidence of SIRS in patients with COVID-19.


Introduction
Extracorporeal membrane oxygenation (ECMO) is a technology most often used to support patients with acute severe cardiac and/or respiratory failure and high mortality risk despite the use of maximal conventional therapy [1][2][3]. In the midst of a pandemic, ECMO was increasingly being used in patients worldwide with coronavirus disease 2019 (COVID-19) [4][5][6]. ECMO is intended to augment oxygenation and gas exchange with or without hemodynamic support to allow time for end-organ recovery. Patients are ultimately decannulated from ECMO once criteria for adequate tissue perfusion and organ recovery are met [6]. However, there are numerous complications that may occur after weaning from extracorporeal support, which may hinder subsequent recovery [6]. Development of systemic inflammatory response syndrome (SIRS) while on ECMO can occur due to a complex and multi-faceted innate inflammatory response to the foreign extracorporeal circuit [5][6][7][8]. Systemic inflammation can also occur due to the underlying disease process or secondary insults related to infection, prolonged ventilator support, aspiration, or extensive vascular injury [6][7][8]. SIRS has also been observed and, more recently, described in the literature following ECMO decannulation. Poorer outcomes are associated with those who have evidence of true infection and associated sepsis following decannulation [4]. It remains unclear whether COVID-19 or associated treatment with immune-modulating therapies affect the incidence of post-decannulation SIRS or infection. This retrospective chart review was conducted in order to identify and compare the incidence, risk factors, and associated outcomes of SIRS and infection following decannulation from ECMO in patients with and without COVID-19. This article was previously presented as a meeting abstract at the 2022 American Thoracic Society International Conference on May 15, 2022.

Materials And Methods
We conducted an IRB-approved retrospective chart review of all patients who survived ECMO at Thomas Jefferson University Hospital between June 31, 2010 and July 7, 2021. Patients who died within 24 hours of ECMO cannulation were excluded. Patients who died within 48 hours of decannulation were also excluded since these deaths were presumably attributed to preceding organ failure or complications rather than the SIRS response. Patients with COVID-19 were confirmed by a positive reverse transcription polymerase chain reaction (RT-PCR) assay for SARS-CoV2. Per institutional protocol, all patients were confirmed to be afebrile without evidence of uncontrolled or active infection at the time of ECMO decannulation. ECMO decannulation was initially exclusively performed in the operating room with direct vessel repair; however, during COVID, bedside decannulation was increasingly performed to optimize resource utilization. All patients received perioperative antibiotics post-decannulation for 24 hours unless a predetermined course for controlled infection was being completed. SIRS was confirmed when two out of three of the following criteria (fever [>101.5 ℉], leukocytosis [white blood cell > 12,000 or 25% increase from prior baseline], and/or initiation/escalation of vasopressors) were met within 48 hours of decannulation. Conventional SIRS criteria, including tachycardia and tachypnea, were not used due to potential confounding effects related to concurrent analgesia and sedation, vasopressors, inotropes, neuromuscular blocking agents, deconditioning, or specific ventilator settings. Patients who developed post-ECMO SIRS were then distinguished based on the presence of infection. Infection was defined by either (a) a previously recognized infection during ECMO that was appropriately controlled at the time of decannulation or (b) new positive culture data from blood, sputum, urine, or stool assays following decannulation. Patients who developed SIRS criteria without infection were further characterized as having "true SIRS." Administration of immune modulating therapies, including tocilizumab, baricitinib, hydroxychloroquine, and glucocorticoids, during the treatment course for COVID-19 infection was also noted. Veno-arterial ECMO (VA-ECMO) via femoral cannulation was primarily used for patients with cardiogenic shock, whereas the majority of patients with respiratory failure underwent veno-venous ECMO (VV-ECMO) using either a single-site, dual-lumen right internal jugular (IJ) venous cannulation or traditional dual-site cannulation via the femoral and right IJ venous cannulations. A minority of patients with predominant respiratory failure received VA-ECMO due to concurrent heart failure, hypotension, or anatomical or technical limitations. ECMO circuits included a Rotaflow centrifugal pump (Maquet, Rastatt, Germany) and a Quadrox-D diffusion membrane hollow-fiber oxygenator (Maquet, Rastatt, Germany).

Statistical analysis
Data were expressed as a number paired with the percentage, mean, or median (quantile), as appropriate. The two groups were compared using chi-squared tests for categorical variables and standard t-tests for continuous variables. Statistical significance was accepted at a level of p = 0.05.

Discussion
This is the first study to further describe the incidence and clinical implications of SIRS following decannulation from ECMO in patients with COVID-19. The clinical correlation between ECMO and systemic inflammation has been well described in the literature [8]. The introduction of blood into an extracorporeal system is known to incite a robust innate humoral and cellular cascade that can closely resemble SIRS. Leukocyte function, complement activation, and endothelial and thrombotic dysregulation are among the many established downstream effects [6][7][8][9][10]. This interplay between coagulation, complement, and endothelial systems initiates and perpetuates a pro-inflammatory cellular and cytokine milieu, which, if unchecked, can lead to vascular and end-organ injury [10][11][12][13][14][15][16].
This study demonstrated that SIRS was commonly observed following decannulation from ECMO support, with an overall incidence of approximately 44% in this cohort. The incidence was similar when comparing those with and without concomitant COVID-19 infection. Despite COVID-19 patients being older and enduring longer durations of ECMO support, neither of these variables significantly impacted the incidence of SIRS response or relevant outcomes. It was initially hypothesized that prolonged cannulation would result in endothelial activation, which might sustain pro-inflammatory activation. It was additionally considered that the severe inflammatory state associated with COVID-19 might predispose to more frequent SIRS following decannulation [8]. These data do not confirm either of these hypotheses. It is unclear whether these results could have been mitigated by other variables.
Among those who developed SIRS, patients with COVID-19 tended to have lower creatinine, hepatic aminotransferase, and lactate levels prior to ECMO cannulation. However, there was no sustained difference in these markers prior to decannulation, which may have contributed to comparable rates of SIRS and subsequent 30-day mortality.
When evaluating patients with COVID-19 more closely, we did not identify any pertinent pre-ECMO or predecannulation variable, other than age, which was associated with an increased risk of developing post-ECMO SIRS. Advanced age, in contrast to relevant studies in cardiac surgery, was associated with an increased risk of an inflammatory response and resultant SIRS. Despite an older cohort, SIRS was not associated with worse outcomes, which is consistent with a previous study of non-COVID patients at our institution [1]. There was similar use of adjunctive therapies (i.e., tocilizumab, baricitinib, hydroxychloroquine, and glucocorticoids) within two weeks of decannulation in all patients with COVID-19, suggesting no demonstrable impact of these therapies on SIRS in this cohort.
When evaluating the SIRS response further, the majority of patients with COVID-19 developed SIRS in the context of an associated infection, whereas only three patients developed "true SIRS." Unlike our previous experience with post-ECMO SIRS in patients without COVID-19, the development of post-ECMO SIRS with infection was not associated with worse 30-day survival.

Limitations
This study has several limitations owing to its retrospective methodology, small sample size, and singleinstitution experience. Furthermore, our definition of SIRS included the initiation/escalation of vasopressors, which varies from conventional definitions and excludes patients with associated tachycardia or tachypnea.