Introduction

Extracorporeal membrane oxygenation (ECMO) is actively used for life support in patients with severe heart and/or respiratory failure. It was previously reported that hemodynamic features, in particular, decrease of pulsatile flow during venoarterial (VA) ECMO reduce bowel perfusion [1]. This could potentially mean that enteral nutrition is not safe and/or will not be well tolerated. In veno-venous (VV) ECMO, natural pulsatile blood flow is almost intact, but implantation of ECMO system may exacerbate gastrointestinal dysfunction through activation of systemic inflammatory response. In this regard, there is a risk of intestinal barrier disruption and translocation of bacterial flora [2]. There are few data on nutritional support in ECMO, and there are no targeted studies on the impact of VA and VV ECMO on gastrointestinal function. As a rule, nutritional support is not mentioned in the studies devoted to ECMO [3]. Apparently, it is erroneous to assume that patients received adequate nutrition [4]. Prior to the COVID-19 pandemic, nutritional protocols focused on early enteral (EN) rather than parenteral nutrition (PN). However, it is unclear to what extent this has been successfully implemented in clinical practice.

In winter of 2009, a novel influenza A (H1N1) epidemic increased the need for VV ECMO in the Southern Hemisphere [6]. A retrospective analysis of nutritional support in 86 adults undergoing ECMO showed advisability of early EN and its reasonable tolerability using standard EN protocols in patients undergoing VA and VV ECMO.

Intestinal oxygen therapy, known for more than a hundred years, is applied for various indications (from the treatment of helminthic invasion to improvement of systemic oxygenation). This case report does not pursue a detailed description of the entire intensive care, but focuses on the first experience of long-term use of intestinal oxygenation in combination with EN in a patient with severe hypoxemia undergoing ECMO.

A 66-year-old man admitted to the intensive care unit No. 7 of the Almazov National Medical Research Centre after establishing VV ECMO on 11/02/2020. The following diagnosis was established: COVID-19 (U07.01), (virus identified on 09/28/2020). Mild bilateral pneumonia (mild course dated 09.2020 (CT-1), progression dated 10/21/2020 (CT-3). Concomitant diagnosis: essential hypertension. Obesity class I. Complications: respiratory failure stage III. Acute severe respiratory distress syndrome. Critical illness polyneuropathy. Combined encephalopathy. Steroid-induced diabetes mellitus. Anemia stage 1. Toxic hepatitis stage 3.

The patient considered himself ill since September 17, 2020, when fever, progressive general weakness and anosmia occurred. Chest CT revealed mild lesion of lung parenchyma (CT-1) that was typical for COVID-19 associated pneumonia. Outpatient symptomatic therapy was prescribed. After 10 days, the patient was hospitalized to infectious disease hospital due to clinical deterioration.

Respiratory failure (RF) requiring oxygen therapy persisted throughout the entire hospital-stay (3 weeks). After a month (October 21, 2020), PaO₂/FiO₂ index decreased to 122 mm Hg. CT revealed bilateral pneumonia (CT 3-4). Considering previous three negative smears for COVID-19, we transferred the patient to the Almazov National Medical Research Centre on the same day.

Severe clinical status at admission was due to respiratory failure and severe asthenia. Low-flow oxygen insufflation was replaced with high-flow therapy (FiO₂ 0.9, F 60 l/min). Non-invasive ventilation was periodically used (StarMed CaStar R helmet, PEEP 7.5 cm H₂O, PS 13 cm H₂O, FiO₂ 0.65). Under therapy, SpO₂ was 94-96% within 10 days. However, minimal physical activity was followed by severe desaturation up to 70-80%.

Despite the ongoing therapy, there was progressive damage to lung tissue up to its total consolidation. By the moment of ECMO establishment (November 2, 2020), the volume of lung damage reached 100% (Fig. 1).

Fig. 1. Lung CT data in dynamics.

Considering previous predominantly negative outcomes in patients on invasive ventilation, we initiated VV ECMO under spontaneous breathing. ECMO was established using the following scheme: femoral vein (Medtronic 23 Fr inflow cannula) — internal jugular vein (Medtronic 21 Fr outflow cannula). A week later (11/08/21), the patient was intubated. Formally, we applied ultraprotective parameters of ventilation considering the ongoing ECMO. In fact, severe violation of mechanics of external respiration (compliance 3 ml / cm H₂O) determined "volumeless" ventilation.

In total, ECMO continued until January 13, 2021 (68 days), and its high performance was required to maintain adequate gas exchange. Duration of invasive ventilation was 78 days. Oxygenator was replaced 3 times throughout the entire period of ECMO. General course of disease from its onset to recovery and labor rehabilitation is schematically presented in Fig. 2.

Fig. 2. Course of disease from its onset to recovery and returning to work.

Changes of tidal volume, lactate and SpO₂ during ECMO are shown in Fig. 3.

Fig. 3. Changes of tidal volume, lactate and SpO₂ during ECMO.

Tidal volume was ≤100 ml throughout 30 days of invasive ventilation that corresponded to ventilation of dead space. These data were also confirmed by indirect calorimetry (FiO₂ = FeO₂).

Lon-term ECMO was accompanied by hyperdynamic circulation, and SpO₂ rarely exceeded 90%. It was due to significant venous admixture in lungs as a result of own cardiac output. We administered beta-blockers (esmolol) to reduce cardiac output and pulmonary bypass. Serum lactate did not exceed 3.8 mmol/l that indicated no severe hypoperfusion and oxygen debt.

The main stages and events of intensive care are summarized in Table.

Major stages and events during intensive care

By the time ECMO was started, the patient had severe protein-energy deficiency confirmed by extremely low levels of total protein, albumin, transferrin, severe lymphopenia (total protein 39 g/l, albumin 20 g/l, transferrin 0.81 g/l, lymphocytes 510 /µl). The main stages and methods of nutritional support (EN, PN + oral enteral nutrition), laboratory data (C-reactive protein, lactate, total protein, albumin, lymphocyte count), injected volumes of albumin and propofol are presented in Fig. 4.

Fig. 4. Changes of C-reactive protein, lactate, total protein, albumin and lymphocyte count during ECMO, invasive mechanical ventilation, intestinal oxygenation and nutritional support for the period from 11/2/20 to 01/23/21.

Note. EN volume before intestinal oxygen therapy — 300-1000 ml/day, PN volume — 1000 ml/day, frequent transfusions of albumin. As soon as intestinal oxygen therapy was initiated, the volume of EN gradually increased up to 1500 ml/day, PN volume decreased to almost zero, albumin transfusions were much less frequent or were not performed. Abbreviations: enteral mixture H — Hepa type, S — Standard type, E — Energy type. ECMO — extracorporeal membrane oxygenation, IO — intestinal oxygen therapy, NS — nutritional support.

While the patient breathed spontaneously, he underwent oral EN (sipping) with hypercaloric enteral mixtures (up to 400 ml/day). Invasive ventilation required EN through a nasogastric tube. Depending on clinical situation, we applied standard and specialized enteral mixtures such as Hepa, Diabetes, Energy Fiber. Additional PN was necessary due to insufficient amount of enteral diet or episodes of gastrointestinal bleeding.

Considering impossible indirect calorimetry (“volumeless” ventilation), the need for energy and macronutrients was analyzed empirically or using the Sheldon method (daily nitrogen loss).

PN was performed using a two-component container (glucose + amino acids) without a fat emulsion, since the patient received propofol up to 1000 ml/day for a long time. This recommendation is currently included in the FAR guidelines for nutritional support in COVID-19 patients. Propofol, being a fat emulsion, fully compensates deficiency of essential fatty acids and provides the required amount of non-protein calories together with glucose from the container.

A component of intensive care was intestinal oxygen therapy. The last one was carried out using the EMO 500 ventilator (“Zavod Elektromedoborudovanie” LLC, St. Petersburg) equipped with the function of low-flow oxygen insufflation and parallel monitoring of intestinal pressure (Mazurok V.A. Method for extrapulmonary blood oxygenation; Matus K.M., Mazurok V.A. Respiratory support device, patent No. 2535072) (Fig. 5).

Fig. 5. Ventilator with available low-flow oxygen insufflation and monitoring of intra-abdominal pressure for intestinal oxygenation.

Low-flow oxygen therapy at a rate of 3-15 ml/min was delivered through the nasointestinal probe under control of abdominal manometry. Nasointestinal probe was passed to the ligament of Treitz and connected to the nozzle of insufflator. The upper limit of intra-abdominal pressure was set at 20-25 cm H₂O, i.e. abdominal hypertension grade 1-2. As soon as threshold intra-abdominal pressure was achieved, oxygen supply by insufflator was stopped, and automatically resumed when the pressure decreased. The total daily volume of oxygen ranged from 4 to 12 liters.

Oxygen was administered continuously after the 13^th day of ECMO for more than 30 days. At the same time, EN was delivered in parallel through the same nasointestinal probe. The upper limit of intra-abdominal pressure on insufflator had to be increased up to 35-40 cm H₂O due to high viscosity of enteral mixtures and impossible "pumping" of oxygen through a thin nasointestinal probe in some cases. As a result, the total operating time of insufflator was more than 700 hours, and the total volume of enteral oxygenation was about 350 liters (maximum value for the entire existence of intestinal oxygen therapy).

In our opinion, combination of intestinal oxygen therapy and EN contributed to assimilation of necessary amount of EN and, accordingly, reducing the incidence of additional PN, maintaining the level of serum albumin and reducing the need for albumin transfusion (Fig. 4).

In the intensive care unit, the patient received antibacterial therapy, infusions, anticoagulation, hormonal (glucocorticoids) and biological (anticytokine) agents. We used morphine to reduce the proportion of spontaneous breathing and oxygen consumption. Antifibrotic drugs (Vargatef®, pirfenidone) were used; twice transfusions of mesenchymal stem cells were applied. As external respiration improved and pulmonary function recovered, the patient was weaned from ECMO, subsequently transferred to spontaneous breathing and finally decannulated.

Discussion

Our clinical demonstration presents only a part of intensive care provided to a patient with extremely severe course of viral respiratory failure. We emphasized metabolic status, nutritional support and enteral oxygen therapy accompanying ECMO. In our opinion, this is the first experience of long-term combination of nasointestinal EN and intestinal oxygen therapy in a patient with severe hypoxemia.

We, like many today, are faced with the problem of long-term EN in patients undergoing ECMO. Only the so-called trophic EN was previously recommended in these patients. There were almost no recommendations for nutritional support of patients undergoing long-term ECMO since this procedure was short-term in most cases. In this regard, certain questions arise. What should we do after 10, 20, 30, 60 ... days? How and how much mixtures to administer? In our case, we achieved a volume of >1000 ml/day by the 3^rd day after beginning of EN. In total, the patient received up to 1500 ml/day of enteral mixture for more than 50 days without any severe adverse reactions and complications. EN was administered for 15-16 hours a day only through a dosing device. Pauses in EN were required in case of gastrointestinal bleeding.

Scott L. et al. [2] analyzed 27 patients undergoing VV ECMO. Nine ones required complete or additional EN. Patients received about 68% of the target amount of food. Target energy requirements were 25 kcal/kg x day, protein requirements — 1.2—1.5 g/kg x day. Most patients received prokinetic agents (erythromycin) within 48 hours. None patient developed bowel ischemia, gastrointestinal bleeding or other complications associated with early EN.

Lukas G. et al. [6] reported 48 patients in an Australian hospital (VA ECMO — 35, VV ECMO — 13); 14 out of 48 patients needed for complete or additional EN. In general, patients undergoing ECMO received only 55% of their target volume.

A retrospective study of patients on ECMO revealed more patients who achieved nutritional goals compared to previous studies [5]. However, some of these differences may be due to energetic contribution of propofol used for long-term sedation. The authors also concluded that EN can be well tolerated by patients during ECMO. ECMO mode had no significant effect on successful enteral nutrition, although this conclusion should be interpreted with caution considering small number of cases of VA ECMO in this study [7].

Anderson et al. [8] reported 24 adults with polytrauma complicated by respiratory failure undergoing either VV or VA ECMO. In these patients, EN was accompanied by enteral mixture intolerance. Kolla et al. [9] reported 100 adults with acute respiratory failure undergoing ECMO. Enteral nutrition was a first-line nutritional support. Unfortunately, the authors did not describe tolerability of EN, adverse reactions and complications, as well as achievement of target energy and protein requirements. Tolerability of early EN in adults undergoing VA ECMO was analyzed in a prospective observational study conducted at the cardiac intensive care unit. Nutritional support was carried out in accordance with the local protocol. Energy requirements were calculated as 25 kcal/kg x day with achievement within 4 days. Nutritional tolerance was defined as the ratio of effective nutrition delivered to target calories. Over 70% nutritional tolerance was achieved within the first week. All patients received only EN and no clinically significant side effects occurred [10, 11].

A distinctive feature of our case is the use of intestinal oxygen therapy in intensive care. We applied this approach for alternative oxygen delivery due to limited (fixed) systemic oxygen delivery (no pulmonary gas exchange and periodic difficulties with the function of ECMO) and resulting severe systemic hypoxemia. The second important goal of enteral oxygenation was maintaining the functional state of enterocytes and integrity of intestinal barrier, as well as prevention of bacterial translocation.

To date, multiple experimental and clinical data have already been published in favor of enteral oxygen therapy for hypoxemia and enteropathies of critical conditions [12-19]. A new wave of interest to this method was due to acute problem of intestinal failure in critically ill patients, as well as few treatment approaches (oxygen, mechanical ventilation, ECMO) for severe disorders of alveolar-capillary gas exchange. Intestinal oxygen therapy received additional scientific justification as soon as academician Ugolev A.M. et al. described the “Phenomena of bilateral respiration of mammalian enterocytes under normal conditions” (No. A-147 dated January 19, 1998; No. OT-12022 dated May 25, 1990). In particular, experimental studies conducted at the Pavlov Institute of Physiology (St. Petersburg) showed that supply of enterocytes is largely determined by oxygen and nutrients from intestinal lumen.

We can first of all talk about the likely contribution of intestinal oxygen therapy to recovery of our patient due to no symptoms of multiple organ failure throughout the entire period of critical illness. Potential explanation is maintenance of intact gastrointestinal function and more effective control of systemic infection. It is extremely difficult to make a definite (especially uncompromising) conclusion, since it is hardly possible to isolate the role of any measure.

Our own experience of intestinal oxygen therapy in critically ill patients with signs of intestinal failure allows us to make certain conclusions. First, the intestine absorbs oxygen well. Secondly, safety of this technology is ensured by mandatory parallel abdominal manometry. Thirdly, enteral administration of small doses of oxygen stimulates peristalsis, preserves the structure of intestinal epithelium, increases systemic oxygenation unpredictably and can have very wide indications.

A detailed description of idea and clinical application of enteral oxygen therapy is beyond the scope of this report. Therefore, we invite the reader to refer to thematic sources including those listed in bibliography of this manuscript for more detailed information.

Conclusion

Considering the above-mentioned data, it is worth highlighting the fundamental provisions characterizing specifics of nutritional support in intestinal oxygen therapy. Firstly, patients undergoing VV ECMO develops severe protein-energy deficiency due to critical illness. Secondly, several studies devoted to ECMO in "pre-Covid" era showed that enteral nutrition during ECMO is possible not only in "trophic mode", but also as complete EN. Finally, combination of intestinal oxygenation with enteral nutrition is well tolerated and, apparently, contributes to: a) absorption of macronutrients (proteins), b) preserving the integrity of gastrointestinal tract, c) prevention of bacterial translocation and, therefore, generalization of infection.

Author contribution:

Concept and design of the study — Leiderman I.N., Mazurok V.A.

Collection and analysis of data — Marichev A.O., Kanshaov N.Z.

Writing the text — Leiderman I.N., Mazurok V.A., Rzheutskaya R.E., Bautin A.E.

Editing — Leiderman I.N., Mazurok V.A., Rzheutskaya R.E., Bautin A.E.

The authors declare no conflicts of interest.