2.3 Refractory ARDS: Prone Positioning, Neuromuscular Blockade & VV-ECMO

Tiered Escalation in Refractory ARDS

Refractory ARDS is defined clinically by persistent, severe hypoxemia ($\text{PaO}_2/\text{FiO}_2 < 150\text{ mmHg}$, and particularly $< 100\text{ mmHg}$) or unmanageable hypercapnic respiratory acidosis ($\text{pH} < 7.15$) despite the optimization of lung-protective mechanical ventilation ($V_T\ 6\text{ mL/kg}$ PBW, $P_{\text{plat}} \le 30\text{ cmH}_2\text{O}$, driving pressure $< 14\text{ cmH}_2\text{O}$, and titrated high PEEP). Management requires a rapid, structured, and tiered escalation of evidence-based interventions:

[Step 1: Baseline Lung-Protective Ventilation]
  - Volume Assist-Control 6 mL/kg PBW (titrate down to 4 mL/kg if Pplat > 30 cmH2O)
  - Titrate PEEP (Higher PEEP / Lower FiO2 strategy) | Target Delta P < 14 cmH2O
  - Conservative Fluid Management (FACTT protocol: CVP < 4 mmHg or PAWP < 8 mmHg)
                         │
                         ▼ (If PaO2/FiO2 < 150 on FiO2 >= 0.60, PEEP >= 5)
[Step 2: Early Prone Positioning + Targeted Neuromuscular Blockade]
  - Prone positioning >= 16 consecutive hours/day (PROSEVA protocol)
  - Deep sedation; if refractory patient-ventilator dyssynchrony or elevated Delta P persists:
    Initiate 48-hour continuous Cisatracurium infusion (ACURASYS)
                         │
                         ▼ (If PaO2/FiO2 remains < 100 or severe RV failure ensues)
[Step 3: Rescue Physiological Adjuncts]
  - Inhaled Pulmonary Vasodilators (Inhaled Epoprostenol 10-50 ng/kg/min or iNO 5-20 ppm)
  - Optimize RV afterload; treat acute cor pulmonale
                         │
                         ▼ (EOLIA criteria: P/F < 50 for > 3 h, P/F < 80 for > 6 h, or pH < 7.25 + PaCO2 >= 60 for > 6 h)
[Step 4: Venovenous Extracorporeal Membrane Oxygenation (VV-ECMO)]
  - Emergent cannulation at or transfer to high-volume ECMO center
  - Ultra-protective lung rest (Vt 2-4 mL/kg PBW, PEEP 10-15 cmH2O, RR 4-10/min, Delta P < 10)

---

Prone Positioning and the PROSEVA Trial

Prone positioning is one of the only interventions in critical care medicine with a proven, robust mortality reduction in severe ARDS.

Physiological Mechanisms of Prone Positioning

1. Homogenization of Transpulmonary Pressure: In the supine position, gravitational hydrostatic pressure, cardiac mass, and cephalad displacement of the diaphragm by abdominal contents compress dependent dorsal lung regions, generating highly negative regional transpulmonary pressures and causing extensive dorsal atelectasis. In the prone position, the heart rests on the anterior sternum, and the dorsal lung is relieved of compressive forces. Transpulmonary pressure ($P_{\text{tp}} = P_{\text{alv}} - P_{\text{pl}}$) becomes remarkably uniform from ventral to dorsal surfaces.
2. Preservation of Perfusion and Shunt Reduction: Pulmonary blood flow is governed by intrinsic vascular branching architecture and remains preferentially directed to dorsal lung regions, regardless of body position. When the patient is placed prone, the dorsal lung units re-expand while dorsal blood flow remains high, dramatically restoring ventilation-perfusion ($V/Q$) matching and eliminating right-to-left intrapulmonary shunt.
3. Reduction in Cyclic Strain and VILI: Homogeneous distribution of tidal volume prevents excessive regional hyperinflation of ventral units and cyclic opening/closing of dorsal units, directly reducing atelectrauma and biotrauma.
4. Enhanced Secretion Clearance: Dependent drainage of secretions from the tracheobronchial tree is facilitated.

The PROSEVA Landmark Trial (Guérin et al., NEJM 2013)

The PROSEVA trial randomized 466 patients with severe, confirmed ARDS ($\text{PaO}_2/\text{FiO}_2 < 150\text{ mmHg}$ on $\text{FiO}_2 \ge 0.60$, $\text{PEEP} \ge 5\text{ cmH}_2\text{O}$, and $V_T$ about $6\text{ mL/kg}$ PBW, confirmed after a 12-to-24-hour stabilization period) to either prone positioning or supine positioning. Prone sessions were mandated for at least 16 consecutive hours per day.

The findings demonstrated an exceptional mortality reduction:

• 28-Day Mortality: 16.0% in the prone group vs. 32.8% in the supine group ($p < 0.001$, hazard ratio 0.39, 95% CI 0.25-0.63).
• 90-Day Mortality: 23.6% in the prone group vs. 41.0% in the supine group ($p < 0.001$).
• Number Needed to Treat (NNT): Approximately 6 patients to save one life.
• Discontinuation Criteria: Prone therapy was stopped when $\text{PaO}_2/\text{FiO}_2 \ge 150\text{ mmHg}$ was sustained with $\text{PEEP} \le 10\text{ cmH}_2\text{O}$ and $\text{FiO}_2 \le 0.60$ for at least 4 hours after returning to the supine position.

Contraindications and Complications

• Absolute Contraindications: Unstable spinal fractures, severe open facial or pelvic trauma, open abdomen with active packing, refractory ventricular arrhythmias, and acute massive hemoptysis.
• Relative Contraindications: Acute elevated intracranial pressure (ICP), pregnancy, severe hemodynamic instability requiring high-dose vasopressors, recent sternotomy, and deep venous thrombosis treated $< 48\text{ hours}$.
• Complications: Endotracheal tube extubation or kinking, central venous and arterial line dislodgement, facial, thoracic, and knee pressure ulcers, brachial plexus neuropraxia, and transient facial/airway edema.

---

Neuromuscular Blockade: ACURASYS vs. ROSE Trials

Neuromuscular blockade eliminates active patient breathing efforts that generate high transpulmonary pressures, eliminates patient-ventilator dyssynchrony (e.g., double-triggering, reverse triggering, breath-stacking), decreases patient-self-inflicted lung injury (P-SILI) and Pendelluft (intra-pulmonary gas redistribution between lung regions during uncoordinated diaphragmatic contraction), and decreases total body metabolic oxygen consumption and $\text{CO}_2$ production.

ACURASYS (Papazian et al., NEJM 2010)

The French ACURASYS trial evaluated 340 patients with early, severe ARDS ($\text{PaO}_2/\text{FiO}_2 < 150\text{ mmHg}$ within 48 hours of onset). Patients received a continuous infusion of cisatracurium besylate for 48 hours versus placebo. Both groups received deep sedation. Crude 90-day mortality was 31.6% vs. 40.7% ($p = 0.08$); after adjustment for baseline PaO2/FiO2, plateau pressure, and SAPS II, the hazard ratio for death was 0.68 (95% CI 0.48-0.98, $p = 0.04$). Cisatracurium also increased ventilator-free days and reduced barotrauma without increasing ICU-acquired weakness.

PETAL ROSE Trial (Moss et al., NEJM 2019)

The multicenter PETAL ROSE trial evaluated 1,006 patients with moderate-to-severe ARDS ($\text{PaO}_2/\text{FiO}_2 < 150\text{ mmHg}$ on $\text{PEEP} \ge 8\text{ cmH}_2\text{O}$) comparing early continuous cisatracurium infusion with deep sedation against a contemporary control strategy of light sedation targets (RASS -1 to 0) without routine neuromuscular blockade.

ROSE revealed no difference in 90-day in-hospital mortality (42.5% in the neuromuscular blockade group vs. 42.8% in the control group, $p = 0.93$). Furthermore, patients in the neuromuscular blockade group experienced significantly more serious cardiovascular adverse events (including bradycardia and hypotension) and delayed physical mobilization.

Evidence Synthesis for the Boards

Routine continuous neuromuscular blockade for every patient with $\text{PaO}_2/\text{FiO}_2 < 150\text{ mmHg}$ is not supported by ROSE. The 2024 ATS guideline makes a conditional recommendation for neuromuscular blockade in early severe ARDS, and the Surviving Sepsis Campaign (2021, reaffirmed 2026) suggests intermittent boluses over a continuous infusion. A practical synthesis:

1. Initial management targeting light-to-moderate analgosedation with lung-protective ventilation.
2. Reserving continuous neuromuscular blockade (for up to 48 hours) strictly for patients who exhibit refractory patient-ventilator dyssynchrony, severe breath-stacking generating high driving pressures, persistent shivering, or severe refractory hypoxemia despite deep sedation and prone positioning.

---

Inhaled Pulmonary Vasodilators

Inhaled vasodilators include inhaled nitric oxide (iNO, 5 to 20 ppm) and inhaled epoprostenol / prostacyclin ($\text{PGI}_2$, 10 to 50 ng/kg/min).

Mechanism of Action

Unlike intravenously administered vasodilators, which indiscriminately dilate both ventilated and non-ventilated pulmonary vascular beds (worsening shunt and systemic hypotension), inhaled vasodilators are delivered selectively to well-ventilated alveolar units. They induce local cyclic GMP (iNO) or cyclic AMP (epoprostenol) mediated relaxation of adjacent vascular smooth muscle, redistributing pulmonary arterial blood flow away from non-ventilated consolidated units to aerated units. This significantly improves $V/Q$ matching, reduces intrapulmonary shunt, and reduces right ventricular afterload.

Clinical Efficacy & Trial Outcomes

Multiple randomized clinical trials and systematic Cochrane reviews demonstrate that while inhaled pulmonary vasodilators produce a rapid, statistically significant improvement in arterial oxygenation (increasing $\text{PaO}_2/\text{FiO}_2$ by 20% to 30%), they confer no improvement in mortality, ventilator-free days, or ICU duration. Consequently, guidelines recommend against routine use; they are reserved strictly as a temporary rescue bridge to maintain oxygenation while mobilizing for prone positioning, VV-ECMO cannulation, or the resolution of acute right heart failure.

Toxicities and Monitoring

• Inhaled Nitric Oxide (iNO):
  • Binds avidly to hemoglobin to form methemoglobin. Daily methemoglobin levels must be monitored; levels $> 3\%$ mandate dose reduction, and levels $> 5\%$ require cessation and potentially methylene blue therapy ($1\text{--}2\text{ mg/kg}$ IV).
  • Combines with ambient oxygen to form nitrogen dioxide ($\text{NO}_2$), an airway irritant and toxin (keep $\text{NO}_2 < 2\text{ ppm}$).
  • Causes dose-dependent renal tubular injury and increased risk of acute kidney injury.
• Inhaled Epoprostenol:
  • Glycine buffer and aerosolized drug can crystalize within the ventilator circuit, clogging expiratory filters. This risks acute expiratory valve obstruction, life-threatening dynamic hyperinflation, and tension barotrauma; expiratory filters must be changed every 2 to 4 hours.
• Rebound Hypoxemia: Abrupt cessation of either agent precipitates acute rebound pulmonary arterial vasoconstriction and catastrophic arterial hypoxemia. They must always be weaned gradually in step decrements.

---

Venovenous Extracorporeal Membrane Oxygenation (VV-ECMO)

Venovenous ECMO provides complete extracorporeal gas exchange (oxygenation and carbon dioxide removal), bypassing the pulmonary circulation without providing direct hemodynamic circulatory support.

Circuit Mechanics

Deoxygenated venous blood is extracted from the inferior vena cava via a large drainage cannula (21 to 25 Fr) placed through the femoral vein, driven through an extracorporeal centrifugal pump into a polymethylpentene hollow-fiber membrane lung, and returned oxygenated and decarboxylated into the right atrium via a reinfusion cannula (17 to 21 Fr) positioned in the right internal jugular vein (or via a single bicaval dual-lumen cannula).

• Oxygenation ($Q_b$): Governed primarily by extracorporeal blood flow rate ($L/\min$). To maintain adequate systemic oxygenation (target arterial saturation $\ge 88\%$), ECMO blood flow must typically supply at least 60% to 70% of the patient's cardiac output (usually 4 to 6 L/min).
• $\text{CO}_2$ Removal (Sweep Gas Flow): Governed by the sweep gas flow rate ($L/\min$). Increasing sweep gas rapidly clears dissolved $\text{CO}_2$ across the membrane lung, allowing immediate normalization of severe respiratory acidosis.
• "Ultra-Lung-Protective Rest" Ventilation: Once VV-ECMO is established, mechanical ventilator settings are reduced to resting parameters: $V_T\ 2\text{--}4\text{ mL/kg}$ PBW, respiratory rate 4 to 10 breaths/min, PEEP $10\text{--}15\text{ cmH}_2\text{O}$, and plateau pressure $\le 25\text{ cmH}_2\text{O}$ (driving pressure $< 10\text{ cmH}_2\text{O}$), completely resting the lung to promote parenchymal healing.

The EOLIA Trial (Combes et al., NEJM 2018)

The landmark EOLIA trial established the validated clinical criteria for initiating VV-ECMO in severe ARDS:

1. $\text{PaO}_2/\text{FiO}_2 < 50\text{ mmHg}$ for $> 3\text{ hours}$ despite $\text{FiO}_2 \ge 0.80$ and $\text{PEEP} \ge 10\text{ cmH}_2\text{O}$; OR
2. $\text{PaO}_2/\text{FiO}_2 < 80\text{ mmHg}$ for $> 6\text{ hours}$ despite $\text{FiO}_2 \ge 0.80$ and $\text{PEEP} \ge 10\text{ cmH}_2\text{O}$; OR
3. Arterial $\text{pH} < 7.25$ for $> 6\text{ hours}$ resulting from severe hypercapnia ($\text{PaCO}_2 \ge 60\text{ mmHg}$) with respiratory rate set to 35 breaths/min and $P_{\text{plat}} \le 32\text{ cmH}_2\text{O}$.

Anticoagulation and Complications

• Anticoagulation: Requires systemic unfractionated heparin infusion targeting an anti-factor Xa level of $0.3\text{--}0.5\text{ IU/mL}$ (or aPTT 50 to 70 seconds) to prevent circuit and membrane thrombosis.
• Hemorrhagic Complications: The leading cause of ECMO-related morbidity; intracranial hemorrhage occurs in 5% to 7% and is frequently fatal.
• Recirculation: Occurs when oxygenated blood returning from the reinfusion cannula is immediately drawn back into the drainage cannula without entering the systemic circulation. Manifests as high pre-membrane saturation ($S_{\text{pre}}\text{O}_2 > 80\%$) combined with persistent patient arterial hypoxemia. Managed by physically increasing the distance between cannula tips or repositioning under fluoroscopic/echocardiographic guidance.