Aspects of circulatory failure in respiratory extracorporeal membrane oxygenation

Abstract

Extracorporeal Membrane Oxygenation (ECMO) was developed in the seventies for the intended use of supporting respiratory failure. Today ECMO has become a well-established treatment for patients with both respiratory and circulatory failure where conventional intensive care is inadequate. The conventional way of treating adults with respiratory failure has been venovenous ECMO (VV ECMO) and for cardiogenic failure venoarterial (VA ECMO). However, since the respiratory system is an intrinsic part of the cardiocirculatory system and these systems exist in conjunction with each other, a respiratory failure may also inflict impairment on the circulation. Furthermore, a distributive shock differs from a cardiogenic shock. Therefore, it also remains to be clarified if, and to what extent vasoplegic (distributive) circulatory failure in conjunction with respiratory failure benefits from ECMO support, and which mode (VV or VA) should be preferred. Furthermore, if VA ECMO is instituted in a patient with a respiratory failure there are several issues that needs to be addressed that differ from the VA patient with single organ cardiogenic failure.

AIMS: The first aim of this thesis was to describe if ECMO and more specifically VA ECMO has a positive effect on survival in adult patients with septic shock (Study 1). In Study 2 we investigated the incidence, indication and outcome in patients who were converted from VV to VA ECMO to clarify whether conversion has an impact on mortality. Furthermore, since patients on peripheral VA ECMO will have parallel circulations with ensuing differential hypoxemia (DH), we went on with investigating patients with signs of DH. Thus, in Study 3 we investigated the impact on oxygen saturation in the upper body by change of drainage position from the inferior vena cava (IVC) to the superior vena cava (SVC). Since septic shock on ECMO can lead to prolonged ECMO with significant lung parenchymal damage we continued in study 4 to investigate if pulmonary blood flow (PBF) measured with echocardiography may assist in assessment of the extent of pulmonary damage, and if echocardiography and CT findings were associated with patient outcome.

METHODS: All studies are retrospective, originating from a high-volume ECMO centre. Patients who were not treated at our unit during the whole ECMO run, and patients with ongoing cardiopulmonary resuscitation (CPR) at the time of ECMO initiation were excluded in all studies. In Study 1 all patients treated for septic shock between 2012 and 2017 with an age >18 years, fulfilling septic shock criteria according to Sepsis-3, and a vasoactive- inotropic support equivalent to a Vasoactive inotropic score (VIS) >50 to reach a mean arterial pressure >65 mmHg despite adequate fluid resuscitation, were included. In Study 2 all patients >18 years old who were commenced on VV ECMO between 2005 and 2018 were included. Patients who were converted to VA ECMO within the first six hours after ECMO treatment was commenced were excluded. In Study 3 all patients from the age of 15 years between 2009 and 2020 identified with differential hypoxemia were included. Patients were included if there had been a state of fulminant differential hypoxemia (FDH) leading to a repositioning or change of the drainage cannula. FDH was defined as a higher saturation in the lower part of the body compared to the upper part of the body or a saturation of the upper body below or equal to 60%. In Study 4, all patients from the age of 15 between 2011 and 2017 were screened. Patients with septic shock (according to Sepsis 2) originating from pneumonia and treated for >28 days were eligible for inclusion.

RESULTS: In Study 1, thirty-seven patients were included. Twenty-seven patients were submitted to VA and 10 patients to VV ECMO. Hospital survival was 90% in septic shock with left ventricular failure, and 65% in patients with distributive vasoplegic shock. In Study 2, 219 VV ECMO patients were evaluated, 21% (n=46) were converted to VA ECMO. The two main reasons for conversion were right ventricular failure (RVF) or cardiogenic shock. In the converted patients, there was a significant increase in Sequential Organ Failure Assessment (SOFA) scores between admission 12 (IQR 9-13) and conversion 15 (IQR 13-17), p<0.001. The converted patients had a higher mortality rate compared to the non-converted patients (62 vs. 16%, p<0.001). These patients also scored lower at admission on the Respiratory Extracorporeal Membrane Oxygenation Survival Prediction (RESP) score (2 (0–4) versus −2.5 (−4–1), p<0.001). Mortality among RVF patients was 67% compared to 50% in converted patients with circulatory shock. In Study 3, 472 patients were screened and seven were identified with FDH. The mean peripheral capillary saturation increased from 54(±6.6) to 86(±6.6) %, (p=<0.001) after repositioning of the cannula from the IVC to the SVC. Pre-oxygenator saturation increased from 62(±8.9) % to 74(±3.7) %, (p=0.016) after repositioning. In Study 4, CT failed to indicate any differences in viable lung parenchyma between survivors and non-survivors at any time over the course of ECMO treatment. A mixed effects model with time, survivors and non-survivors and the interaction between time and the two groups as independent variables, showed that the interaction was significant (p=0.004) with different coefficient slopes between the two groups regarding PBF.

CONCLUSIONS: Study 1 supported the use of VA ECMO for distributive septic shock. Study 2 indicated that VA ECMO should be considered as the first mode of choice in patients with respiratory failure combined with a compromised circulation. Study 3 elucidated DH in a clinical patient setting which has never been presented previously, showing that moving the drainage zone into the upper part of the body had a marked positive effect on upper body saturation. Finally in Study 4 we presented results demonstrating that CT was supported as a prognostic tool in prolonged respiratory ECMO. However, we found that PBF may possibly assist in the prediction of pulmonary recovery.