The Immunoadsorption Effect of Veno-arterial Extracorporeal Membrane Oxygenation in Refractory Septic Shock, Ventilator-associated Pneumonia, and Acute Respiratory Distress Syndrome Following Severe Pulmonary Contusions

: Background


INTRODUCTION
The application of extracorporeal membrane oxygenation (ECMO) in adult patients with septic shock remains a matter of inotropes and vasopressors [1]. Recent evidence has emerged regarding the effectiveness of VA-ECMO in adult patients experiencing septic shock, as it is a reliable circulatory and ventilatory support method [1,2] with immunoadsorption effects on bacteremia and toxemia. This treatment option has been shown to improve short-and long-term outcomes in patients with bacterial septicemia, severe sepsis, and septic shock. Although its effectiveness in reducing mortality rates in neonatal and pediatric patients with septic shock has already been established [3 -5], its effect on adult patients with septic shock remains controversial, despite the publication of successful cases. Further research is required to establish a final protocol for the use of VA-ECMO in adult patients with severe sepsis, septic shock, ARDS, and ventilator-associated pneumonia (VAP) [6,7]. VA-ECMO was successfully employed during the influenza A (H1N1) outbreaks, significantly reducing morbidity and mortality rates. Nonetheless, further research is necessary to determine its efficacy in non-influenza patients with ARDS [8,9].

Objective
The objective of this study is to assess the potential VA-ECMO as an immunoadsorption therapy in improving the clinical outcomes in adult patients with severe sepsis complicated by septic shock, including ARDS and VAP, weaning from mechanical ventilation, the length of intensive care unit (ICU) stay, and mortality rates in patients with concurrent severe pulmonary contusions, septic shock, and respiratory failure attributed to ARDS and VAP.

MATERIALS AND METHODS
This prospective cohort study enrolled 100 patients aged 18 -65 years who had severe pulmonary contusions due to severe chest trauma and who showed respiratory failure with severe ARDS as determined by the Murray score (> 3 points) from King Abdul-Aziz Specialist Hospital between September 2020 and February 2022. The Research Ethics Committee of the hospital approved this study. All patients received controlled mechanical ventilation (CMV) for ten days with a respiratory rate of 12/min, a positive end-expiratory pressure (PEEP) of 5 cm/H 2 O, and the fraction of inspired oxygen (FIO 2 ) adjusted to maintain arterial oxygen saturation above 90%. They were also administered sedatives, including fentanyl and midazolam intravenous infusion, to achieve a Richmond Agitation-Sedation Scale (RASS) score of -2--3 points ( Table 1) [10]. In addition, all patients were administered broad-spectrum antibiotics in the form of 1 gm meropenem via slow intravenous infusion at eight-hour intervals. Subsequent to three days of ventilation, a qualitative sputum culture was obtained from all patients, and the antibiotic regimen was adjusted in accordance with the culture results. On the second day of ventilation, enteral feeding was initiated using a feeding pump. The feeding rate was established at 70 mL of Ensure plus (Abbot) with a caloric density of 1.47 kcal/mL, designed to provide patients with an estimated 2500 kcal within 24 hours, based on an approximation of 35 kcal/kg [20]. Furthermore, all patients underwent percutaneous tracheostomy on the 7 th day of ventilation. To prevent VAP, a five-point bundle was implemented on all patients. This included raising the head of the bed by 30º to 45º, conducting a daily assessment for potential extubation, unitizing an endotracheal tube with subglottic secretion drainage, administering oral care with oral antiseptics, and initiating safe enteral nutrition within 24-48 hours of admission to the ICU and ventilation.
The sample size of this study was restricted to 100 patients. Inclusion criteria were set as follows: 1) Patients who exhibited no improvement in respiratory function despite ten days of ventilation, and 2) those who continued to experience severe ARDS with a Murray score of > 3 points (Table 2) [11 -20]. Exclusion criteria involved: 1) Patients who developed severe sepsis with a SOFA score of > 12 points (Table 3) [12,13], 2) those who developed VAP with a CPIS score of > 6 points ( Table 4) [14], or 3) those who required nor-adrenaline infusion > 2 mg/kg/min as circulatory support due to circulatory failure (septic shock).
The parameters of the 100 patients mentioned earlier indicated the need for VA-ECMO. However, given the controversial nature of VA-ECMO as a treatment modality, the study subjects were randomly assigned to a control group (group A, n = 50) and an intervention group (group B, n = 50). Group A received standard conventional management, while group B received VA-ECMO. The study spanned two weeks, during which all patients in both groups were assessed at regular intervals, specifically at the end of the first five days, the end of the second five days, and the end of the final four days.    Throughout the study period, both groups of patients were closely monitored using the Acute Physiology and Chronic Health Evaluation II (APACHE II) scoring system [15 -20], as well as the scores for severe sepsis (SOFA) and ARDS (CPIS). The findings of these measurements were compared between the two groups [17 -20].
To evaluate lung tissue recovery from ARDS and VAP, laboratory indicators such as lactate dehydrogenase (LDH) and C-reactive protein (CRP) were utilized. These indicators were collected concurrently and compared between groups A and B. Additionally, improvements in lung mechanics, including compliance measurements and the response of patients to lung recruitment maneuvers, were used to assess the restoration of lung ventilatory functions [17 -20]. A lung recruitment maneuver is a clinical procedure used to assess lung compliance in patients with ARDS or other conditions that cause difficulty in breathing. It involves increasing the peak inspiratory pressure to 40 cm/H 2 O for 40 seconds, followed by observing oxygen saturation (SpO 2 ) levels. A positive response to the maneuver is typically defined as an improvement in SpO 2 levels to more than 95% [16 -20].
This study involved recording and comparing the number of patients in both groups who were successfully weaned from mechanical ventilation according to our established protocol. Specifically, extubated patients were closely observed in the ICU for 24 hours before being discharged, and the number of discharged patients was recorded and compared between the two groups. With the aim of validating bacteriological cure, three sputum cultures were collected from all patients during the study period: one at the end of the first five days, another at the end of the second five days, and a final one at the end of the study period. Patients who had not shown any improvement in their SOFA and CPIS measurements by the end of the study period were classified as cases of morbidity. The number of such cases was documented and compared between the two groups. Similarly, failure to be weaned from mechanical ventilation during the study period was also an indicator of morbidity [17 -20]. The number of patients with this condition was also documented and compared between the two groups. Finally, the number of patients discharged from the ICU during the study period was recorded and compared between the two groups.
The process of connecting patients to VA-ECMO involved accessing their venous blood from a large central vein, particularly the right femoral vein, through an "access line." The blood was then directed through an oxygenator and back into the arterial system near the right atrium via a "return line." In severe respiratory and circulatory failure cases where the flow through a single access cannula was not enough to sustain the high ECMO flow rate needed, a second venous access cannula was used, usually the right internal jugular. However, in these specific cases, the left femoral artery was used instead of the right femoral vein to direct the blood back into the arterial system. Once VA-ECMO was fully connected and the flow established, the mechanical ventilation was discontinued, and a PEEP was set to 3 cm H 2 O, while FIO 2 was kept at 100%. For up to 72 h after ECMO initiation, cytokine adsorption filters (CytoSorbents Europe, Berlin, Germany) were used, with regular replacement every 24 hours. A team of experts was responsible for applying and maintaining ECMO equipment. At the end of the study, all patients with a Glasgow Coma Scale (GCS) [19,20] score < 9 points underwent a computerized axial tomography (CAT) scan to rule out any potential organic brain damage in either group.

Statistical Analysis of the Data
Data were fed to the computer using IBM SPSS software package version 21.0.
Qualitative data were described using numbers and percentages. Comparison between different groups regarding categorical variables was tested using Chi-square test [17 -20].

Chi-square Test
It tests the association between qualitative nominal variables, it is performed mainly on frequencies. It determines whether the observed frequencies differ significantly from the expected frequencies [17 -20].

Sample Size Calculation
The sample size should depend on the research context, including the researcher's objectives and proposed analyses.
The following formula was used to calculate the required sample size in this study; Where n is the sample size, Z is the statistic corresponding to the level of confidence, P is expected prevalence, and d is precision (corresponding to effect size). The level of confidence was 95%. By using this equation the sample size was 100 cases (i.e. 50 cases in each group). A number of expired patients from the two groups were collected and compared, then the new number of patients was recalculated after the expired patients were subtracted from the total sample size giving us a new sample size. This was done 3 times in the study at the end of the first 5 days, at the end of the second 5 days, and at the end of the last four days. X 2 is the value of the chi-square test is used to calculate the p-value. The P value is considered highly significant if less than 0.05 and marked by (*) beside the significant number in our study [17 -20]. Table 5 represent the demographic data of patients in both groups and shows no significant difference between the two groups as regard age and sex. Table 6 compares the mortality rate in both groups all over the studied duration and shows a non-significant difference between the two groups in all studied duration as (2, 2 and 3 patients) died at the end of 5,10,14 days consecutively from group A while (2, 2 and 2 patients) died from group B at the same duration of the study. Table 7 compares the APACH II score of patients in both groups all over the studied duration and shows significant improvement in that score in patients of group B in all studied duration as 19, 31 and 39 patients had scoreless <15 at the end of 5,10,14 days consecutively, while 9, 14 and 20 patients in group A achieved a same score at the same duration of the study. Table 8 compares the GCS score of patients in both groups all over the studied duration and shows significant improvement in that score in patients of group B in all studied duration as 21, 32 and 40 patients had scores>12 at the end of 5, 10, 14 days consecutively, while 11,16 and 21 patients in group A achieved the same score at the same duration of the study. Table 9 compares the MAP of patients in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 18, 30 and 36 patients had MAP >95 mmHg without inotropic support at the end of 5, 10, 14 days consecutively, while 7,18 and 20 patients in group A achieved same MAP at the same duration of the study. Table 10 compares the bilirubin level of patients in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 20, 33 and 41 patients had bilirubin <1.9 mg/dL at the end of 5,10,14 days consecutively, while 9, 19 and 22 patients in group A achieved same bilirubin level at the same duration of the study. Table 11 compares the creatinine level of patients in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 19, 28 and 38 patients had creatinine <1.9 mg/dL at the end of 5,10,14 days consecutively, while 9,16 and 18 patients in group A achieved same creatinine level at the same duration of the study. Table 12 compares the platelets count of patients in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 18,25 and 34 patients had platelets > 100.000 at the end of 5,10,14 days consecutively, while 10,17 and 20 patients in group A achieved same platelets count at the same duration of the study. Table 13 compares the nature of the tracheal secretion of patients in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 20, 30 and 42 patients had normal tracheal secretion at the end of 5,10,14 days consecutively, while (12, 16 and 20 patients) in group A achieved same tracheal secretion at the same duration of the study. Table 13 compares parenchymatous lung infiltrate on the chest X-ray of patients in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 23, 32 and 43 patients had almost normal chest X-rays at the end of 5,10,14 days consecutively, while 12, 18 and 21 patients in group A achieved same chest Xray at the same duration of the study. Table 13 compares the core temperature of patients in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 21, 31 and 41 patients had core temperatures> 36. 5 and < 38.4 at the end of 5,10,14 days consecutively, while 10, 20 and 22 patients in group A achieved same core temperature at the same duration of the study. Table 13 compares the total leucocytic count of patients in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 23, 33 and 40 patients had leucocytic count > 4000and < 11000 thousand/mL at the end of 5,10,14 days consecutively, while 10, 21 and 24 patients in group A achieved same leucocytic count at the same duration of the study. Table 13 compares the hypoxic index(PaO 2 /FIO 2 ) of patients in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 28, 38 and 44 patients had hypoxic index >240 at the end of 5,10,14 days consecutively, while 19, 24 and 33 patients in group A achieved same hypoxic index at the same duration of the study. Table 13 compares the microbiological results of patients in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 29, 35 and 43 patients had negative qualitative BAL culture and at the end of 5,10,14 days consecutively, while 20, 25 and 30 patients in group A had negative qualitative sputum culture at the same duration of the study. Table 14 compares lung compliance in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 19, 28 and 35 patients had measured lung compliance of > 80 ml/ 1 cm H 2 O at the end of 5,10,14 days consecutively, while 10,13 and 20 patients in group A achieved same measured lung compliance at the same duration of the study. Table 15 compares the number of patients who respond to the recruitment maneuver in both groups all over the studied duration and shows a significantly higher number of patients who respond to this maneuver in group B in all studied duration as 21, 31 and 40 patients were responder at the end of 5,10,14 days consecutively, while 12, 18 and 22 patients in group A were responder at the same duration of the study. Table 15 compares the number of patients who failed to be weaned in both groups at the end of the studied duration and shows a significant lower number of patients in group B (only 3 patients out of 44 patients) compared to group A (16 patients out of 43 patients).

RESULTS
Table 15 compares the total number of patients discharged from the ICU at end of the studied period and shows a significantly higher number of patients discharged from group B (41 patients) compared to group A (27 patients).
Table 16 compares the LDH levels of patients in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 23, 32 and 42 patients had LDH < 200 U/L at the end of 5, 10, 14 days consecutively, while 8,19 and 25 patients in group A achieved same LDH level at the same duration of the study. Table 17 compares the C-reactive protein levels in both groups all over the studied duration and shows significant improvement in patients of group B in all studied duration as 20, 33 and 43 patients had C-reactive protein level <100 mg/L at the end of 5,10,14 days consecutively, while 8, 20 and 27 patients in group A achieved same C-reactive protein level at the same duration of the study. Table 18 compares the morbidity recorded in all patients between the two groups in the studied periods.

Effect of Veno-arterial Extracorporeal Membrane
The

DISCUSSION
In relation to enhancing the overall health condition of patients [17 -21], noteworthy improvements were observed in the APACHE II score, total leucocyte count, and core body temperature in patients of group B compared with those of group A. These improvements can be attributed to the use of ECMO, which provided better tissue oxygenation compared with non-functional, infected, and contused lungs. The better tissue oxygenation resulted in improved systemic oxygenation and improved the patient's cellular and humoral immunity in group B, which helped manage both bacteremia and toxemia caused by VAP. Consequently, this facilitated the management of all general signs of systemic inflammatory response syndrome (SIRS) and helped quickly control various parameters, such as CPIS score, SOFA score, and weaning from inotropic support and mechanical ventilation.
The improvement of pulmonary function can be classified into two different categories. The first category is clinical improvement, which refers to the significant improvement in clinical parameters such as oxygen saturation, hypoxic index, compliance, and response to recruitment maneuvers in patients of group B compared with those in patients of group A throughout the study period [17 -21]. The second category is a radiological improvement, which refers to the significant improvement in imaging findings of the lungs, such as reduced infiltrates or opacities on chest X-ray or CT scan in patients of group B compared with those in patients of group A over the same duration [17 -21]. This phenomenon may be attributed to the enhancement of local immunity of the lung induced by improved oxygenation. Physiologically, increased oxygenation of lung tissue causes pulmonary vasodilatation and increases local blood supply, resulting in a concomitant enhancement of local immunity of the lung. In addition, ECMO provides lung protection against the deleterious effect of mechanical ventilation (especially PEEP), thus facilitating the acceleration of the healing process of lung tissue from both septic inflammation (VAP) and traumatic inflammation (pulmonary contusions). Furthermore, the utilization of VA-ECMO in the delivery of tissue oxygenation appears to be superior to the use of inotropes in supporting the patient's circulation. This superiority can be attributed to the vasoconstrictive of all inotropes due to their alpha-agonist actions, which improve brain and vital organ tissue perfusion, while also potentially resulting in severe systemic tissue ischemia as a consequence of severe peripheral vasoconstriction.
In this study, bacteriological improvement [17 -21] was evaluated by analyzing the qualitative sputum cultures of all patients on days 10 and 14. The results revealed a statistically significant increase in positive results between groups A and B. This difference may be attributed to the superior immunity of patients in group B, including both local immunity of the lung and systemic immunity. Additionally, the enhanced oxygenation observed in group B could play a role in this outcome. Improved local tissue oxygenation in the lung and systemic oxygenation might result from the aforementioned factors contributing to the higher positive culture rates in group B.
This study also demonstrated notable decreases in LDH and CRP levels in group B, indicating a significant improvement in controlling laboratory indicators related to tissue destruction compared with that in group A [17 -21]. The observed difference is likely attributed to the faster lung tissue healing and more efficient management of tissue destruction caused by trauma and sepsis in group B. The positive outcome is plausibly ascribed to the superior both lung and systemic oxygenation and immunity achieved with VA-ECMO relative to group A.
In patients receiving VA-ECMO (group B), immunoadsorption, a process of removing bacteria, toxins, antigens, and cytokines from the bloodstream, positively affected patient outcomes. These substances are known to potentiate the sepsis cascade, aggravate tissue destruction, and trigger a vicious circle of immune response and further tissue destruction, accompanied by circulatory failure. The immunoadsorption effect of VA-ECMO intervenes in this circle by removing or reducing the concentration of these harmful substances. This results in a rapid recovery from a low tissue perfusion state and restoration of the MAP with sufficient tissue perfusion pressure in all organs. As a result of this intervention, patients in group B experienced significant improvements in their GCS score, creatinine level, bilirubin level, and platelet count. These improvements were attributed to the restoration of proper perfusion pressure, which allowed organs such as the brain (recovery from septic encephalopathy), kidney, liver, and bone marrow to function more effectively. Additionally, a faster improvement in global tissue perfusion and reduction in the release of systemic toxins and pyrogens led to better control of other measures of patient health, such as the APACHE II score, leucocyte count, core temperature, CPIS score, SOFA score, and Murray score. This resulted in improved ventilatory function and quicker weaning from mechanical ventilation, further supporting the idea that immunoadsorption can have positive effects on patient outcomes. Several case reports [22 -27] have been published describing the use of cytokine adsorption techniques. One such technique is the Cytosorb hemoadsorption column (Linc Medical, Leicestershire, United Kingdom), which is installed in the ECMO circuit to stabilize septic shock patients more rapidly and improve their outcomes. These reports suggest that Cytosorb is effective in removing proinflammatory cytokines from the bloodstream, which can halt the injurious cascade of sepsis and manage tissue destruction. Furthermore, it has been shown to reduce vasopressor doses and serum inflammatory cytokines in septic patients.
In terms of morbidity and mortality rates [17 -21], several factors were observed to exert substantial impacts on patient outcomes in this study. These factors include an APACHE II score > 25 points, a GCS score < 6 points, MAP < 70 mmHg with/without inotropic support, a bilirubin level > 6 mg/dL, a creatinine level > 5 mg/dL, a platelet count < 5×10 4 /mL, hypoxic index < 100, chest X-ray showing infiltrates in all lung quadrants, lung compliance > 19 mL/cmH 2 O, no response to recruitment, core temperature > 39°C or < 36°C, abundant and purulent tracheal secretion, a leucocytic count of 2 on the CPIS score, an LDH level > 400 U/L, CRP > 200 mg/L, and positive sputum cultures after 10 d and 14 d of treatment, and failure to wean from mechanical ventilation at the end of the study period. Notably, these factors were significantly higher in group A than those in group B. However, patients in group B exhibited significantly higher incidence rates of cerebrovascular and local complications from cannulation than those in group A. This can be attributed to the use of anticoagulants in group B. Moreover, the two groups had no significant difference in the mortality rate.
The findings of this study corroborate the results of previous research conducted by Peek et al. (2009) [28] in the United Kingdom. Peek's study examined 90 patients with ARDS who received VA-ECMO for 6-16 days. The study reported a mortality rate of 37%, while the remaining patients showed improved lung function and were discharged from the hospital earlier than those treated with conventional ventilation. In 2011, Noah et al. [8] conducted a study involving 69 patients with H1N1 influenza-induced ARDS who received ECMO for 6-16 d. The study showed a marked improvement in lung condition and a lower mortality rate of 24%. Similarly, Pham et al. (2013) [29] studied 123 patients with severe ARDS due to an H1N1 influenza epidemic and found a mortality rate of 36% with ECMO use for 8-22 days. Brechot et al. (2013) [30] examined 14 patients with refractory septic shock who received VA-ECMO within 24 hours of onset. The study reported an 86% weaning success rate and a 71% discharge rate with good quality of life after a median follow-up of 13 months . Schmidt et al. (2014) [31] developed a reliable survival prediction score for ECMO patients based on a study involving 2,355 patients with ARDS respiratory failure who received ECMO for 14 days. The study reported that 1,338 patients were discharged alive from the hospital in good condition, while the remaining patients had a mortality rate of 57%. Finally, Nesseler et al. (2015) [32] investigated the use of ECMO in patients with intra-abdominal sepsis-induced ARDS. The study found a significantly larger number of patients being weaned from both inotropic support and mechanical ventilation compared with that in the control group, although the ECMO group consisted of only 40 patients who received ECMO for two weeks due to ARDS-induced respiratory failure.
Several studies have investigated the efficacy of VA-ECMO in adult patients with septic shock and respiratory failure associated with ARDS. These studies reveal that the VA-ECMO group does not significantly differ from the control group concerning the length of ICU stay and mortality rate. A prospective study by Huang et al. (2013) [33] included 52 adult patients who had refractory septic shock and required VA-ECMO for circulatory support. The study found no significant difference between the VA-ECMO group and the control group in terms of outcomes, and only 15% of included patients were discharged home, suggesting poor outcomes. Similarly, a recent multicenter study by Takauji et al. (2017) [34] enrolled 3195 patients from 42 Japanese ICUs. Out of 570 patients who suffered from severe respiratory failure, 285 experienced respiratory failure induced by lung infections. Overall, the study included 40 patients in the ECMO group (supported by ECMO) and 150 patients in the control group. The two groups had comparable SOFA scores (12 points vs. 13 points). In addition, there were no significant differences in the mortality rate and length of ICU stay between the ECMO group and the control group. Although the studies above suggest that further multicenter research is needed to prove the efficacy of VA-ECMO in patients with sepsis-induced circulatory and respiratory failure, the therapy has demonstrated favorable outcomes in our center.
This study exhibited limitations, namely, a small sample size and the inclusion of only ARDS cases induced by infected contused lungs. Therefore, large-scale investigations are imperative, encompassing all etiologies of ARDS, such as trauma, lung infection, and other systemic causes, to assess the efficacy of VA-ECMO in all types of ARDS with septic shock.

CONCLUSION
VA-ECMO notably impedes the progression of sepsis, shortens ICU stay, and expedites the weaning from inotropic support and mechanical ventilation. However, it has no impact on the mortality rate of adult patients with septic shock.

HUMAN AND ANIMAL RIGHTS
No animals were used for studies that are the basis of this research. All the humans were used in accordance with the ethical standards of the committee responsible for human experimentation (institutional and national), and with the Helsinki Declaration of 1975, as revised in 2013 (http://ethics.iit.edu/ecodes/node/3931

CONSENT FOR PUBLICATION
Written consent from all patients was taken from the firstdegree relatives of the patients.

STANDARDS OF REPORTING
STROBE guidelines were followed.

FUNDING
None.

CONFLICT OF INTEREST
The authors declare no conflict of interest, financial or otherwise.