Establishment of a Veno-venous Extracorporeal Membrane Oxygenation in a Rat Model of acute Respiratory Distress Syndrome

Background: Venovenous extracorporeal membrane oxygenation (VV-ECMO) is now considered a reasonable option to salvage acute respiratory distress syndrome (ARDS). However, we lack a rodent model for experimental studies. This study was undertaken to establish an animal model of VV-ECMO in ARDS rats. Methods: Fifteen Sprague-Dawley (SD) rats (350 ± 50 g) were used in this study. Using a rat model of oleic acid (OA)-induced ARDS, VV-ECMO was established through bi-caval cannulation of the right jugular vein for venous drainage and venous reinfusion with a specially designed three-cavity catheter. Continuous arterial pressure monitoring was implemented by using a catheter through cannulation of the right femoral artery. The central temperature was monitored with a rectal probe. Arterial blood gas monitoring was implemented by a blood gas analyzer at three-time points: at baseline, 1-hour (after OA modeling), and 3.5-hour (after VV-ECMO support). Lung tissue and bronchoalveolar lavage uid were harvested respectively for protein concentration and pulmonary histologic evaluation to conrm the alleviation of lung injury during VV-ECMO. Results: Following ARDS induced by OA, ten rats were successfully established on VV-ECMO without failure, and survived from the ECMO procedure. VV-ECMO alleviated lung injury and restored adequate circulation for the return of lung function and oxygen. VV-ECMO was associated with decreased lung injury score, wet/dry weight ratio, and ﬂ uid leakage into airspaces. Conclusion: We have established a reliable, economical, and functioning ARDS rat model of VV-ECMO.


Background
Extracorporeal membrane oxygenation (ECMO) is a life support technique that provides cardiorespiratory support in patients with severe respiratory and cardiac failure [1]. There are two main con gurations for an ECMO circuit: veno-arterial and veno-venous [1,2]. The veno-venous ECMO (VV-ECMO) con guration is preferred in cases of respiratory failure with preserved cardiac function [3]. However, VV-ECMO is thought to play a key role in acute respiratory distress syndrome (ARDS), but its exact role is unclear and suitable animal models are needed to answer this question. Here, we describe a novel VV-ECMO technique using bi-caval cannulation in an oleic acid (OA)-induced rat model of ARDS. This rat model restores respiratory function and offers a tool to study physiological and molecular changes in lung disease during VV-ECMO.

Animals
Fifteen Sprague-Dawley (SD) rats (350 ± 50 g) were used for all the experiments and randomly divided into 2 groups: OA group (n = 3) and VV-ECMO group (n = 12). All animals received humane care in compliance with the 'Principles of laboratory animal care' formulated by the National Society for Medical Research and the 'Guide for the care and use of laboratory animal resources' published by the US National Institute of Health (NIH publication No. 85-23, revised 1996 Induction of lung injury ARDS and impaired gas exchange were modeled in rats by intravenous administration of 99% pure OA (Sigma-Aldrich) as previously described [4,5]. OA and 9mL physiologic saline solution (negative control) were given at a nal concentration of 25mg/mL. 100mg/kg of OA was administered via the femoral vein at an infusion rate of 20mg/kg/min over 5 mins to avoid sudden death due to massive pulmonary embolism [6].

Surgical procedure
Rats were anesthetized using 2% sevo urane to achieve stable anesthesia during the entire operation. Initial ventilatory parameters included 70-75 breaths/min, a tidal volume of 6 mL/kg, a 1:2 I:E ratio, and a 3 cmH 2 O positive end-expiratory pressure (PEEP). A 24-gauge catheter was inserted in the right femoral artery for continuous arterial pressure monitoring. Heparin was administered at 300 U/kg. The right jugular vein was exposed and cannulated with a 5.5 F specially designed three-cavity catheter. Central temperature was monitored using a rectal probe and maintained at 36 ± 0.5℃ using a heating lamp. Venous return was exclusively drained using a roller pump. ECMO was established at a ow rate of 80-90 mL/kg/min. Upon reaching optimal ow rate, it was adjusted to a level that could maintain desired arterial pressure. Gas ow (90% O 2 ) was initiated at around 80-100mL/min and adjusted in terms of sweep rate and FiO 2 to maintain blood gasses within the physiological range. 1mL of sodium bicarbonate is usually added to the circuit during ECMO to maintain acid-base balance. All volume substituting or buffering solutions were given via the femoral artery. After 3 h, the animals were weaned off ECMO and the remaining priming volume re-infused (Additional le video 1).
Blood gas analysis, protein concentration in bronchoalveolar lavage, and histology To assess the oxygenation and metabolic states of the animals during VV-ECMO (EG7 + , iStar, Abbott Co. Ltd.), blood gas analysis (BGA) was done on blood samples from the femoral artery. Femoral arterial blood gas measurements were done at 0 (T0, baseline), 1 h (T1, after OA modeling), and 3.5 h (T2, after VV-ECMO support). At the end of the experiment, bronchoalveolar lavage uid (BALF) was harvested by lavaging the lungs with 2.5mL PBS 1X and BALF protein concentration determined using BCA analysis. The lung was collected and xed in 4% paraformaldehyde for 24 h at 4 °C. Four micrometer para n sections were stained with hematoxylin and eosin (H&E), and organ damaged evaluated histologically. Lung injury score was evaluated using a modi ed, previously described scoring system [6]. Various degrees of lung injury score were designated degree 0, 1, 2, and 3 for mild, moderate, and severe edema, respectively. Similar scoring was used for in ammatory cell in ltration, with degree 0, 1, 2, and 3, indicating none, mild, moderate, and severe cellular in ltration, respectively. Histopathological assessment was done by several blinded laboratory assistants with each giving scores for edema and cell in ltration. The individual edema and cell ltration scores were then summed to obtain a nal score ranging from 0-6.

Results
After a 1 h OA treatment, rats developed hypoxemia (PaO 2 /FiO 2 < 300 mmHg). Ten rats were successfully put on VV-ECMO, with all surviving the ECMO procedure. One rat died immediately after OA injection, probably from pulmonary embolism due to rapid OA administration. Another died from massive bleeding after ruptured jugular vein during catheter insertion. Figure 2 shows VV-ECMO's hemodynamic data.
Mean arterial pressure signi cantly reduced 1 h after OA administration relative to pre-operation, but remained stable during VV-ECMO and 30 min after weaning from ECMO. Table 1 summarizes the blood gas and biochemical parameters of the VV-ECMO procedure. PaO 2 , PaCO 2 and SaO 2 were stable during VV-ECMO, indicating attenuated OA-induced systemic hypoxemia. pH, base excess and HCO 3 − also remained stable throughout the experiment. Relative to pre-ECMO, Hct and Hb signi cantly reduced but remained stable during VV-ECMO. No excessive blood loss was noted, which was consistent with the relatively stable Hct and Hb values. These ndings were comparable to our previous data, and the lower values were due to hemodilution. To ensure proper membrane oxygenator function, blood sampling was done immediately before, and after oxygenation following VV-ECMO initiation (Fig. 3). Pre-oxygenator PO 2 (40 ± 7 mmHg) and post-oxygenator PO 2 (490 ± 48 mmHg) increased signi cantly (p = < 0.0001), indicating excellent oxygenation capacity of our membrane oxygenator, and minimal recirculation. There was a signi cant increase in PaO 2 /FiO 2 after 3 h of VV-ECMO. Histological analysis revealed severe diffuse alveolar damage, the pathologic hallmark of ARDS, with presence of hyaline membranes, alveolar wall thickening, and extensive in ltration by in ammatory cells (Fig. 4A). However, treating OA-injected rats with VV-ECMO 1 h after OA injection ameliorated virtually all major histopathologic changes induced by OA in the lungs (Fig. 4B). VV-ECMO was associated with decreased lung injury score (Fig. 4C), wet/dry weight ratio (Fig. 4D), and uid leakage into airspaces ( Fig. 4E), illustrating that VV-ECMO alleviates OA-induced ARDS.

Discussion
Based on our previously described rat model of CPB [7][8][9], we have developed a stable VV-ECMO method in a rat model of ARDS. After OA modeling, PaO 2 /FiO 2 rapidly reduced, indicating lung injury [6]. After a Despite advances in our understanding of ARDS and its treatment, its mortality rate remains high [10].
VV-ECMO supports whole blood oxygenation and carbon dioxide elimination and has the potential to rescue ARDS patients with refractory hypoxemia or those unable to tolerate conventional therapy [1,11].
To better study the effects of VV-ECMO, reliable animal models are needed [12]. Rats have multiple advantages as animal models of human conditions, including anatomical structures that are almost identical to human ones, small volume, low cost, and easy handling relative to large animal models [13].
Thus, we used rats to create the rst rodent model of VV-ECMO and used it for successful resuscitation after OA-induced lung injury. OA administration recapitulated basic ARDS characteristics, including hypoxemia [4,5]. VV-ECMO was offered with minimal recirculation, rapidly correcting hypoxemia and improving survival. This rodent model more closely copies real clinical situations relative to existing models.
Animal models of bi-caval ECMO are rare. To our knowledge, there are no reports on bi-caval double lumen ECMO using rat models, making this a novel approach. Venous drainage is crucial for successful establishment of VV-ECMO. Previous studies reported that drainage of venous blood by gravity achieves adequate perfusion [12]. Here, venous drainage was achieved through active roller-head assist. The specially designed venous cannula was then inserted into the jugular vein and moved towards the right atrium into the inferior vena cava, allowing a perfusion ow rate of 80-90 mL/kg/min. Normal rat cardiac output ranges from 160-180 mL/kg/min [13]. Thus, on average, we achieved a VV-ECMO ow of 50% rat cardiac output. Relative to VA-ECMO, VV-ECMO requires lower blood ow because excessive perfusion of right atrium causes heart failure due to excess blood pressure in right ventricle [14]. Clinically, a VV-ECMO ow of 50-75% of the cardiac output is su cient for adequate oxygenation in ARDS patients on mechanical ventilation [11]. Unnecessary increases in VV-ECMO ow rate may cause hemolysis damage, and unnecessary recirculation of the main portion of venous blood between IVC and SVC [15]. Furthermore, we observed that increasing ECMO ow rate caused excessive negative pressure, which could cause air suction into intubation sites.
Hemodynamic and blood gas analyses revealed that these parameters uctuate within an acceptable range. Relative to pre-ECMO, Hct and Hb decreased signi cantly, mainly due to dilution of the priming solution. However, compared to past studies [7][8][9], we reduced overall priming volume to 6 mL without blood to avoid adverse effects from transfusion. In our clinical experience, routine addition of sodium bicarbonate during bypass achieves satisfactory blood gas parameters. In general, our model of VV-ECMO had excellent results in ARDS rats. Our study offers a valuable model for further physiological research into potential organ protection and complications from VV-ECMO, and serve as a useful tool for future lung disease studies.

Conclusions
We have established a reliable and economical VV-ECMO approach using a rat model of ARDS. Biochemical analysis show that despite lung injury, our VV-ECMO circuit restored adequate circulation to restore lung function and oxygen. This protocol demonstrates the detailed techniques, physiological observations, and blood gas analysis required for successful VV-ECMO. Availability of data and materials

List Of Abbreviations
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.