Rescuing lung transplant candidates with rapidly progressive interstitial lung disease who fail V-V ECMO: A case series of dual veno-arterial and veno-venous extracorporeal membrane oxygenation circuits

Abstract

Background

Providing adequate, awake and ambulatory mechanical circulatory support to patients with rapidly progressive advanced intersitial lung disease (ILD) remains challenging. In a subset of ILD patients with refractory hypoxemia or hemodynamic instability, despite optimal veno-venous (V-V) extracorporeal membrane oxygenation (ECMO) support, the addition of a veno-arterial (V-A) ECMO circuit may avoid the need for mechanical ventilation and protect against right ventricular dysfunction and subsequent end-organ dysfunction.

Methods

We herein report the first case series of three patients with ILD who received dual VV-VA ECMO support as a bridge to transplantation.

Results

All patients survived until lung transplantation 2 to 8 days after V-A ECMO initiation and had an unremarkable post-transplant recovery.

Conclusion

Although the addition a second ECMO circuit is a complex and resource-intensive strategy, it is a feasible approach to stabilze a subset of ILD patients who are indequately supported with V-V ECMO.

Background

Extracorporeal membrane oxygenation (ECMO) has been used to optimize the physiologic and functional state of lung transplant candidates with advanced respiratory failure.1, 2 However, some patients may experience life-threatening progression of respiratory failure and/or right ventricular (RV) dysfunction despite the use of ECMO. A recent international survey study of ECMO-bridge to transplantation (BTT) practices among patients with interstitial lung disease (ILD) demonstrated that greater than 90% of centers select veno-venous (V-V) ECMO as their initial configuration.³ In patients with significant secondary pulmonary hypertension, veno-arteriovenous (V-AV), upper body V-A ECMO (with an axillary or innominate arterial return cannula), or veno-pulmonary artery (V-PA) configurations are sometimes employed.4, 5, 6 Each of these configurations has certain limitations and can therefore fail to adequately support patients with rapidly progressing ILD.7, 8, 9 In such scenarios, the physiological demands of the patient may exceed the support provided by a single ECMO configuration, especially in a bridge to transplant context where preservation of end-organ function and avoidance of sedation are paramount.

To address this problem, we introduced a dual V-A ECMO circuit in selected patients with ILD who are insufficiently supported with initial V-V ECMO. The addition of a second circuit substantially increased the total blood flow rate and achieved hemodynamic stability. In doing so, we facilitated awakening and preserved lung transplant candidacy. This approach offers a novel alternative configuration strategy for ECMO-BTT in ILD patients inadequately supported with V-V ECMO.

Case 1

A 40-year-old man with familial ILD developed progressive dyspnea 3 months before admission. He was hospitalized locally for respiratory failure necessitating high fraction of inspired oxygen (FiO₂) via high flow nasal cannula (HFNC). Computed tomography imaging of the chest revealed acute bilateral diffuse ground-glass lung opacities on top of fibrosis. Due to progressive hypoxemia, he received V-V ECMO via a dual-lumen 30-Fr crescent cannula inserted into the right internal jugular (RIJ) vein at the outside hospital and was transferred to Vanderbilt University Medical Center (VUMC) for an expedited lung transplant evaluation.

Upon arrival at VUMC, the patient rapidly declined despite maximal ECMO blood flow and sweep gas flow, HFNC, and noninvasive positive pressure ventilation. In addition, the patient developed shock requiring vasopressor support and was unsafe to transport to the operating room for reconfiguration. Additional clinical characteristics are shown in Table 1. To address the patient’s severely impaired gas exchange and hemodynamic instability, we added a V-A ECMO circuit at the bedside via a 25-Fr right femoral vein (RFV) drainage cannula, a 15-Fr left common femoral artery (CFA) return cannula with a 6-Fr distal perfusion catheter (DPC) at the left superficial femoral artery (SFA) (Figure 1A). V-A ECMO blood flow rate was maintained between 2.5 and 3.0 liters per minute (LPM) and sweep gas flow rate of 2 LPM. V-V ECMO blood flow rate was maintained between 4.0 and 4.5 LPM (Table 2). Following the addition of V-A ECMO, the shock resolved, and gas exchange normalized. He remained awake on nasal cannula, participated in physical therapy, and was ultimately listed for lung transplantation.

Table 1.

Baseline Patient Characteristics

Case	Patient 1	Patient 2	Patient 3
Age (years)	40	57	66
Sex	Male	Male	Female
Body mass index (kg/m2)	25.5	30.87	28.2
Height (cm)	180.3	180.3	152.4
Weight (kg)	85.5	100.4	65.4
Body surface area (m2)	2.16	2.11	1.69
Duration from ILD diagnosis to rapid progression requiring maximum HFNC + NRB (months)	3	24	3
Comorbidities
Diabetes	No	No	No
Hypertension	Yes	No	Yes
Pre-existing cardiac disease	No	Yes (CAD)	Yes (CAD)
Pre-existing renal insufficiency	No	No	No
Serum creatinine (mg/dl) and eGFR (ml/min/1.73m2) at admission	0.61, >90	0.67, >90	0.76, >87
Pre-existing pulmonary hypertension	No	Yes (Group 3)	Yes (Group 3)
Right ventricular function	Normal	Normal	Normal
Left ventricular function (EF%)	55%-65%	55%-65%	55%-65%
Mean pulmonary artery pressure (mm Hg)	N/A	28	30
Fick cardiac output (liters)	5.4	5.08	6.01
Baseline hemodynamics and gas exchange before V-V ECMO cannulation
Respiratory rate (breaths/min)	36	37	22
Heart rate	93	109	92
Blood pressure (mean arterial pressure)	90	71	87
Oxygen requirements	HFNC (70 liters, 100%) + NRB (15 liters)	HFNC (70 liters, 100%)—>BIPAP 100% upon cannulation	HFNC (45 liters, 100%) + NRB (15 liters)
Blood gas: pH, paO2, paCO2	7.42, 76, 47	N/A	N/A
Days from hospital admission to V-V ECMO initiation	9	2	4
Baseline hemodynamics and gas exchange before initiation of second V-A ECMO
Respiratory rate (breaths/min)	40-50	30-40	30-40
Heart rate (beats/min)	130-140	120-130	90-100
Blood pressure (mean arterial pressure) mm Hg	63	65	85
Vasopressor support	Yes, 6 mcg/min	Yes, 10 mcg/min	None
V-V ECMO flow (liters/min)	4.21	5.5	4.09
V-V ECMO sweep gas flow (liters)	3.5	6	4
V-V ECMO fractional delivered oxygen	100%	100%	100%
Oxygen saturation	86%	88%	86%
Blood gas: pH, paO2, paCO2	7.42, 64, 53	7.38, 65, 49	7.44, 61, 42
Duration from V-V initiation to addition of second V-A circuit (hours)	48	19	48
CAS Score at the time of BOLTx	43.12	39.58	44.33

Abbreviations: BOLTx, bilateral orthotopic lung transplant; CAD, coronary artery disease; CAS, composite allocation score; ECMO, extracorporeal membrane oxygenation; EF, ejection fraction; FiO₂, fraction of inspired oxygen; HFNC, high flow nasal cannula; ILD, Interstitial lung disease; NRB, nonroom breather mask; V-A, veno-arterial; V-V, veno-venous.

Figure 1.

(A) Illustration of dual VV-VA ECMO circuits for patient 1. Circuit 1, V-V ECMO with 30-Fr RIJ cannula. Circuit 2, V-A ECMO with 25-Fr LFV drainage and 15-Fr left CFA return cannulas. (B) Illustration of dual VV-VA ECMO configuration for patients 2 and 3. Circuit 1, V-V ECMO with RIJ return and RFV drainage cannulas. Circuit 2, V-A ECMO with an LFV drainage and a left CFA return cannula. CFA, common femoral artery; ECMO, extracorporeal membrane oxygenation; Fr, French size; LFV, left femoral vein; RFV, right femoral vein; RIJ, right internal jugular vein; V-A, veno-arterial; V-V, veno-venous.

Table 2.

ECMO Characteristics

Key ECMO variables and parameters	Patient 1	Patient 2	Patient 3
V-V ECMO configuration (drainage-reinfusion)	RIJ Crescent	RFV-RIJ	RFV- RIJ
Cannula size (drainage-reinfusion, Fr)	30	25-22	25-21
Oxygenator brand	Nautilus	Nautilus	EOS -> Nautilus
V-V ECMO settings 24 hours after cannulation
Blood flow rate (LPM)	3.88	4.4	4.21
FdO2 (%)	100	100	100
Sweep gas flow (LPM)	3.5	2	2
SVO2 (%)	N/A	68	74
V-V ECMO settings before second V-A circuit
Blood flow rate (LPM)	4.21	5.5	4.09
FdO2 (%)	100	100	100
Sweep gas flow (LPM)	3.5	6	4
V-V ECMO run before second circuit (hours)	48	12	48
SVO2 (%)	N/A	57	68
Oxygen requirements before second V-A circuit support
Mode	BIPAP	HFNC + NRB	HFNC + NRB
FiO2 (%)	100	100	100
Flow (liters)	NA	15-30	15-30
Arterial blood gas before second V-A circuit
pH	7.42	7.38	7.44
PaCO2 (mm Hg)	50	49	42
PaO2 (mm Hg)	60	65	61
V-A ECMO configuration (drainage-reinfusion)	RFV-LFA	LFV-LFA	LFV-LFA
Cannula sizes (drainage-reinfusion) (Fr)	25-15	23-17	21-15
DPC (6 Fr)	L SFA	L SFA	L SFA
Oxygenator brand for second V-A circuit	Nautilus	Nautilus	Nautilus
ECMO settings 4 hours on dual VV-VA
V-A - Flow	2.6	2.6	2.1
V-A - FdO2 (%)	100	100	100
V-A - Sweep gas flow (liters)	2	2	2
SVO2 (%) - V-A	N/A	87	60
V-V - Flow (LPM)	3.67	4.4	2.99
V-V - FdO2 (%)	100	100	100
V-V - Sweep gas flow (liters)	8	2	9
SVO2 (%) - V-V	N/A	73	68
Arterial blood gas 4 hours on dual VV-VA
pH	7.48	7.41	7.48
PaCO2 (mm Hg)	45	45	38
PaO2 (mm Hg)	106	94	34
ECMO settings 24 hours on dual VV-VA
V-A - Flow	2.9	2.3	2.2
V-A - FdO2 (%)	100	100	100
V-A - Sweep gas flow (liters)	1.5	3	1.5
SVO2 (%) - V-A	N/A	76	84
V-V - Flow (LPM)	3.2	4.32	4.53
V-V - FdO2 (%)	100	100	100
V-V - Sweep gas flow (liters)	7	4.5	4.5
SVO2 (%) - VV	N/A	72	82
Arterial blood gas 24 hours on dual VV-VA
pH	7.47	7.38	7.42
PaCO2 (mm Hg)	47	47	45
PaO2 (mm Hg)	111	81	70
ECMO settings 48 hours on dual VV-VA
V-A - Flow	N/A - in OR	2.5	2.2
V-A - FdO2 (%)	N/A - in OR	100	100
V-A - Sweep gas flow (liters)	N/A - in OR	3	1.5
SVO2 (%) - V-A	N/A	75	90
V-V - Flow (LPM)	N/A - in OR	4.3	3.52
V-V - FdO2 (%)	N/A - in OR	100	100
V-V - Sweep gas flow (liters)	N/A - in OR	5	5
SVO2 (%) - VV	N/A	76	86
Arterial blood gas 48 hours on dual VV-VA
pH	N/A - in OR	7.39	7.41
PaCO2 (mm Hg)	N/A - in OR	43	45
PaO2 (mm Hg)	N/A - in OR	73	84
ECMO settings on dual VV-VA before transplant
V-A - Flow	2.2	2.2	2.2
V-A - FdO2 (%)	100	100	100
V-A - Sweep gas flow (liters)	1	3	0.9
SVO2 (%) - V-A	N/A	81	97
V-V - Flow (LPM)	3.98	4.31	4.16
V-V - FdO2 (%)	100	100	100
V-V - Sweep gas flow (liters)	6	5	4
SVO2 (%) - VV	N/A	73	94
Arterial blood gas on dual VV-VA before transplant
pH	7.45	7.4	7.44
PaCO2 (mm Hg)	45	42	49
PaO2 (mm Hg)	144	66	69
Duration on dual VA-VV ECMO before transplantation (hours)	47	69	145
Anticoagulation therapy while on ECMO	Heparin -> Bival (40-60)	Bival (40-60)	Held-bleeding concerns

Abbreviations: Bival, bivalirudin; ECMO, extracorporeal membrane oxygenation; FdO₂, fraction of delivered oxygen; Fr, French size; LFA, left femoral artery; LFV, left femoral vein; LPM, liters per minute; LV, left ventricular function; OR, operating room; RFA, right femoral artery; RFV, right femoral vein; RIJ, right internal jugular vein; SFA, superficial femoral artery; SVO₂, mixed venous oxygen saturation; V-A, veno-arterial; V-V, veno-venous.

Forty-eight hours later, the patient underwent bilateral orthotopic lung transplant (BOLTx). Intraoperatively, he was reconfigured to a central V-A ECMO using the crescent RIJ cannula and right femoral venous cannula for drainage, with return to a newly placed aortic cannula. Post-transplant, the patient was left on V-V ECMO via the dual lumen RIJ crescent cannula. He was weaned from ECMO on postoperative day (POD) 2, and from mechanical ventilation on POD 14. The patient was discharged home on POD 20 (Figure 2, Figure 3). At his 4-month follow-up visit, the patient walked 1,070 feet (ft) on a 6-minute walk test (MWT) on room air. His forced expiratory volume in 1 second (FEV₁) was 1.61 liters (38% predicted).

Figure 2.

Summary of hospital course for patients 1 to 3.

Figure 3.

Chest X-ray of patients before ECMO, after cannulation, and on the day of discharge post BOLTx. (A) CXR before ECMO, (B) CXR after dual VV-VA ECMO cannulation, (C) CXR on the day of discharge post BOLTx. BOLTx, bilateral orthotopic lung transplant; CXR, chest X-ray; ECMO, extracorporeal membrane oxygenation; V-A, veno-arterial; V-V, veno-venous.

Case 2

A 57-year-old man with coronary artery disease, with 80% stenosis of his mid-distal left anterior descending artery and fibrosing pneumonia ILD, was admitted to the medical intensive care unit in July 2025 with septic shock due to community-acquired pneumonia, and subsequent acute kidney injury (AKI). He was hypoxemic, requiring HFNC 60%/40 LPM. Over the next 48 hours, his shock and AKI resolved with fluid resuscitation and broad-spectrum antibiotics. After a multidisciplinary meeting with both cardiac surgery and advanced lung disease and lung surgical teams, he was deemed a candidate for simultaneous coronary artery bypass graft surgery and BOLTx.

The patient had progressive respiratory failure and was placed on V-V ECMO (25-Fr RFV drainage cannula and a 21-Fr RIJ reinfusion cannula). However, 12 hours later, the patient acutely decompensated, developing moderate shock and refractory hypoxemia despite maximal ECMO and noninvasive positive pressure ventilation support (Table 1). A second ECMO circuit was added (V-A configuration with a 23-Fr LFV drainage cannula to a 17-Fr left CFA return cannula, plus a 6-Fr left SFA DPC) (Figure 1B). V-A ECMO flows were maintained between 2.5 and 3.0 LPM, sweep of 1 and 2 LPM, while V-V ECMO flows were maintained between 4.0 and 4.5 LPM, sweep of 4 and 6 LPM, which normalized hemodynamics and gas exchange (Table 2). Three days later, the patient underwent simultaneous coronary artery bypass graft (with a left internal mammary artery to left anterior descending artery anastomosis) and BOLTx via median sternotomy. Intraoperatively, his V-V ECMO configuration was maintained throughout the case, and his peripheral V-A ECMO was converted to central V-A with an aortic return cannula. At the end of the operation, he was decannulated from both V-V and V-A ECMO.

The patient recovered quickly and was extubated on POD 1 and discharged home on POD 15 (Figure 2). At his 4-month follow-up appointment, the patient has been walking a mile daily on room air. His FEV₁ was 2.57 liters (75% predicted).

Case 3

A 66-year-old woman with CAD (status post drug-eluting stent × 3 in 2023) and pulmonary fibrosis developed progressive dyspnea 3 months prior to admission, requiring home oxygen. She was admitted to VUMC with an acute exacerbation of ILD requiring 15 LPM via non-rebreather to maintain oxygenation. Her infectious workup was notable for Klebsiella on sputum culture, which was managed with broad-spectrum antibiotic therapy. On hospital day 5, she had worsening hypoxia with oxygen saturation in the 70% range with minimal exertion, despite maximal HFNC; V-V ECMO support was initiated with a 25-Fr RFV drainage cannula and a 21-Fr RIJ return cannula (ECMO flows of 3.5-4.0 LPM, sweep of 3 LPM, 6 LPM NC) and she was listed for lung transplantation. She initially demonstrated good functional recovery and was able to ambulate 300 ft on day 2 following V-V ECMO initiation. However, on day 3 after V-V ECMO initiation, her clinical status deteriorated despite maximal supportive therapy (Table 1). She developed refractory hypoxemia (SpO₂ <88%) with prolonged recovery periods, along with tachypnea (respiratory rate 30-40 breaths/min) and resting tachycardia (heart rate 100-110 bpm). She was unable to reposition or mobilize in bed without experiencing severe desaturation episodes, with SpO₂ dropping below 80%. Thus, a parallel V-A ECMO circuit was added to improve physiologic support (21-Fr LFV drainage, 15-Fr left CFA arterial return, and a 6-Fr left SFA DPC (Figure 1B)). Her V-A ECMO flows were maintained between 2.2 and 2.6 LPM (sweep of 1-2 LPM), while V-V ECMO flows were maintained between 4.0 and 4.5 LPM (sweep of 4-6 LPM) (Table 2).

During the first 48 hours post-dual VV-VA ECMO initiation, the patient developed diffuse oozing from her cannulation sites, mild hemoptysis, and significant flow variation on both ECMO circuits, resulting in worsening hypoxia. She initially stabilized after large volume resuscitation with crystalloid and colloid, though she developed worsening thrombocytopenia and hypofibrinogenemia, concerning for circuit-related disseminated intravascular coagulation (DIC). Following the exchange of both ECMO circuits, the patient’s coagulopathy and hemodynamics improved. Six days after initiation of dual VV-VA ECMO, she underwent BOLTx via bilateral thoracotomies. Intraoperatively, her V-V ECMO configuration was maintained during the case, and her peripheral V-A ECMO was converted to central V-A with an aortic return cannula. She was decannulated from both V-V and V-A ECMO at the end of the operation.

She was extubated to 4 LPM NC on POD 2 and was discharged to a rehabilitation facility on POD 17 (Figure 2). At her 4-month follow-up appointment, the patient had been discharged home from the rehab facility. She walked 1,060 ft on a 6MWT on room air and her FEV1 was 1.09 liters (55% predicted).

Discussion

ECMO-BTT is a life-saving therapy for selected patients with progressive advanced lung disease.3, 10 There is a well-recognized subset of ILD patients with rapid disease progression and hemodynamic instability, where a single ECMO circuit support is inadequate. We demonstrate that refractory hypoxemia and hemodynamic instability from RV dysfunction in this subset of ILD patients can be ameliorated with a second dual V-A ECMO circuit. This configuration may support an awake state, which is essential for lung transplant candidates to reduce delirium risk and enable active participation in rehabilitation and exercise. Outcomes of this approach were favorable for our 3 patients, all of them surviving to transplantation and hospital discharge (Table 3).

Table 3.

Outcomes

Primary outcome	Patient 1	Patient 2	Patient 3
Hospital length of stay (days) at VUMC	23	22	29
ICU length of stay before transplant (days)	3	6	9
V-V ECMO duration before transplant (days)	4	4	8
Duration of dual VA-VV circuit support (days) before transplant	2	3	6
ECMO-associated complications
Bleeding complications	None	None	Epistaxis, hemoptysis, circuit-related DIC
Limb ischemia	No	No	No
Thrombocytopenia and severity	Nadir, 148K	Nadir, 140K	Severe, nadir 48K
ECMO flow variation and issues	No	No	Yes, smaller-sized IVC (<20 mm) and hypovolemia
Circuit-related issues requiring intervention	No	No	Yes, oxygenator-related DIC requiring exchange.
Blood product transfusion while on ECMO	2uPRBCs	1uPRBC	13 PRBCs, 6 plts, 6 FFP, 5 cryo
Requirement of renal replacement therapy during ECMO	No	No	No
Post ECMO decannulation thromboembolic complications
Deep vein thrombosis	RIJ cannula associated	No	No
Pulmonary embolism	No	No	No
Cerebrovascular complications	No	No	No
Tracheostomy post BOLTx	Yes	No	No
Survival to transplant	Yes	Yes	Yes
Survival to decannulation	Yes	Yes	Yes
Survival to discharge	Yes	Yes	Yes
Discharge disposition	Home	Local housing (relocated home after 10 weeks).	Rehab facility (discharged home after 16 weeks).
4 months follow-up with 6MWT (distance in feet and oxygen requirement)	1,580, room air	1,650 ft, room air	1,060, room air

Abbreviations: 6MWT, 6-minute walking test; BOLTx, bilateral orthotopic lung transplant; Cryo, cryoprecipitate; DIC, disseminated intravascular coagulation; ECMO, extracorporeal membrane oxygenation; FFP, fresh frozen plasma; ICU, intensive care unit; IVC, inferior vena cava; plts, platelets; PRBC, red blood cells; V-A, veno-arterial; VUMC, Vanderbilt University Medical Center; V-V, veno-venous.

This is the first case series reporting the use of dual ECMO circuits (VV-VA) to support a patient for lung transplantation. In our initial experience using dual VV-VA ECMO circuits as BTT, several features of ECMO management should be highlighted.

Alternatives to the addition of a parallel V-A circuit—conversion to V-AV or upper body V-A ECMO: The 3 patients presented herein showed.

• •Worsening tachypnea (respiratory rate [RR] >35) and hypoxemia (SpO₂ < 88%) at rest despite optimal V-V ECMO settings and 100% FiO₂.
• •Persistent tachycardia (heart rate [HR] >120 at rest or HR >140 with mild exertion) and/or hemodynamic instability (need for vasopressors, rising lactate).
• •Evidence of end-organ hypoperfusion (e.g., oliguric AKI, rising liver enzymes).
• •Refractory hypoxia with activity in bed, limiting safe mobilization.
• •Inability to safely transport to the operating room.

Our ECMO team excluded clinically significant recirculation through daily chest radiography and confirmation of a wide pre- and postoxygenator systemic venous oxygen saturation gradient (>20%),¹¹ before consideration of escalation of support. Although objective RV data (such as serial echocardiography and invasive hemodynamic assessment with Swan-Ganz catheter or right heart catheterization) were not feasible due to patient instability, when clinical indications for escalation of support are present, a multidisciplinary discussion involving the ECMO, intensive care unit (ICU), and advanced lung disease teams was undertaken to determine the most appropriate escalation strategy. In patients who are stable enough and have favorable upper-body vascular anatomy, and when operating room availability permits, our preferred approach is reconfiguration to an upper-body “sport model” V-A ECMO with subclavian arterial return. For patients with lower V-V flow requirements (<4 liters/min), a bedside reconfiguration to V-AV ECMO is considered. Although favorable outcomes with V-PA ECMO have been reported,⁶ percutaneous deployment—even at the bedside—typically requires fluoroscopic and transesophageal echocardiographic guidance.¹² Consequently, this strategy may be more appropriate for patients with sufficient clinical stability to tolerate these procedural requirements.

Since all 3 patients were clinically too unstable for transport to the operating room and had exhausted their V-V ECMO flows, we favored the addition of a second V-A ECMO circuit (Figure 4). The addition of a second V-A circuit over alternative cannulation strategies was considered for 2 primary reasons:

• •First, hybrid configurations such as V-AV ECMO, which utilize a single drainage cannula with split venous and arterial return, may provide insufficient flow to ensure adequate end-organ perfusion or upper-body oxygenation.¹³ Similarly, V-PA ECMO is inherently flow-limited by venous drainage capacity, cannula size, pump performance, and pulmonary vascular resistance, as all return flow must transverse the pulmonary circulation. In patients with severe parenchymal lung disease, pulmonary hypertension, or rapidly progressive ILD, elevated pulmonary vascular resistance can significantly impair effective oxygen delivery. Additionally, cannula size constraints—most commonly 29-Fr or 31-Fr ProtekDuo dual-lumen cannulas in the United States—further limit achievable flow rates, which are typically lower than those attainable with conventional V-V or V-A ECMO.8, 9 Consequently, V-PA and V-AV configurations may fail to provide sufficient respiratory and hemodynamic support in this high-risk population.
• •Second, even brief clamping of the V-V ECMO circuit during reconfiguration carries an increased risk of cardiac arrest; this risk may be mitigated by the addition of a second V-A ECMO circuit, which avoids the need for VV circuit clamping.

Figure 4.

Flowchart for decision-making regarding the addition of a dual V-A ECMO circuit. BTT, bridge to transplantation; ECMO, extracorporeal membrane oxygenation; ILD, Interstitial lung disease; LPM, liters per minute; V-A, veno-arterial; V-AV, veno-arteriovenous; V-V, veno-venous.

Ensuring adequate ECMO flows without excessive negative drainage pressures from 2 drainage cannulas

Awake ILD patients can have vigorous coughing fits, which result in increased intrathoracic and intra-abdominal pressure. This can lead to significant ECMO flow variation, especially in patients with a smaller IVC. In the setting of 2 IVC draining cannulas, it is therefore important to optimize volume status. Other adjuncts, such as intermittent use of inhaled lidocaine or low-dose IV dilaudid, can help decrease patient’s coughing fits and ensure more stable ECMO flows.

Thrombotic and bleeding complications with dual VV-VA circuits

One of our patients developed a cannula-associated deep venous thrombosis, requiring systemic anticoagulation therapy post-transplant. There was 1 major ECMO-associated bleeding complication in patient 3 secondary to oxygenator-related DIC, which resulted in extensive blood product transfusion. Pretransplant blood transfusion can increase a patient’s Panel Reactive Antibody levels, which can significantly lower their chances of finding a compatible organ and can lengthen their waitlist time.¹³ Thus, in these patients, selecting optimal ECMO pumps and oxygenators with low pressure gradients is essential.

Management of ECMO flows and sweep gas of the 2 ECMO circuits

Rapidly progressing ILD is usually due to worsening native lung gas exchange that leads to progressive hypoxemia despite maximum V-V ECMO support. Additionally, in this subset of patients, their worsening group 3 potential of hydrogen contributes to RV dysfunction and hemodynamic instability that threatens end-organ function. It is therefore necessary to ensure hemodynamic stability with V-A ECMO support. With regards to flow distribution between the 2 circuits, the goal is to ensure adequate coronary and upper body oxygenation with V-V ECMO and decrease RV preload to promote visceral organ perfusion with the peripheral V-A ECMO. Providing an ECMO flow of about 60% of the patient’s estimated cardiac output through the V-V circuit allowed for adequate blood oxygenation and resulted in significant relief from hypoxia.¹⁴ V-A ECMO flows were titrated to minimize the need for vasopressors and to achieve lactate clearance. V-A ECMO flows of 2 to 3.0 LPM achieved these goals. Finding an ideal pH goal can be challenging on this platform since the arterial blood gas drawn from the right radial artery usually reflects gas exchange driven by the V-V ECMO circuit given their hyperdynamic LV function. Manipulating the sweep gas flow primarily on the V-V ECMO enabled us to achieve our pH goal and provide patients with symptomatic relief from their respiratory distress. Sweep gas flow was kept below 2 LPM on the V-A circuit.

Maintaining an ambulatory status is challenging on this configuration and can contribute to deconditioning

While patients maintained an awake state on this platform, they were not mobilized. Unlike other hybrid ECMO configurations, such as upper body arterial return cannula,¹⁵ that permit ambulation and enhanced rehabilitation while preserving adequate cardiopulmonary support, dual VV-VA configuration may not be feasible for ambulation or aggressive rehabilitation. Pasrija et al have demonstrated the safety of patient ambulation on peripheral V-A ECMO with femoral arterial return and venous drainage cannulas¹⁶; however, there is no data supporting the safety of ambulating patients with 3 femoral cannulas, although we have previously mobilized patients with dual V-V ECMO circuits. Given concerns about the safety of ambulation with triple femoral cannulation, patient mobility was restricted to in-bed activity and sitting at the edge of the bed. Our physical therapists ensured the patients received adequate in-bed active range of motion exercises with weight-bearing arm and leg exercises performed in bed. All 3 patients were able to perform upper body exercises and gentle leg raise exercises. Patient 3, who had a longer run on her dual VV-VA ECMO (6 days), was able to sit at the edge of the bed on days 5 and 6 before her transplantation. Two of the 3 patients (66%) were able to be discharged home with outpatient physical therapy, and 1 patient was discharged to inpatient rehabilitation facility for reconditioning. This highlights the need for novel hybrid configurations that can be quickly deployed in unstable ILD patients, preserve adequate cardiopulmonary support, and permit ambulation and enhanced rehabilitation while awaiting transplantation.

Resource-intensive therapy

ECMO utilization can be lifesaving, but it is costly and resource-intensive, with strain on both personnel and material resources.¹⁴ Unlike other ECMO configurations, such as upper body V-A ECMO, which often requires utilizing the operating room and anesthesia services, we are able to safely perform dual VV-VA ECMO in an awake state at the bedside (in the ICU). Nevertheless, dual VV-VA ECMO support remains highly resource-intensive, necessitating specialized staffing (including ECMO specialists, perfusionists, and dedicated one-to-one nursing), frequent circuit surveillance and coordination between bedside nurses and perfusionists, multidisciplinary collaboration among ICU teams and rehabilitation services, increased blood product utilization, and substantial equipment and consumable resources, including frequent laboratory monitoring.¹⁷ As a result, the resource demands of dual ECMO circuits as a BTT may limit feasibility and generalizability, particularly in low-volume or resource-constrained centers.

Although all patients survived to transplantation and hospital discharge, this case series has several important limitations. The small sample size limits generalizability, and the inability to obtain objective RV data—such as serial echocardiography or invasive hemodynamic monitoring with a pulmonary artery catheter—due to patient instability precluded more precise characterization of physiologic triggers for escalation of support. In addition, follow-up was limited to early post-transplant outcomes at 4 months; longer-term assessments of functional status and quality-of-life measures will be necessary to fully evaluate the impact of this strategy.

Conclusion

The use of dual VV-VA ECMO as a BTT may be considered a rescue strategy in highly selected patients with rapidly progressive ILD who fail conventional V-V ECMO support, facilitating physiologic stabilization and preservation of transplant eligibility. These findings should be interpreted as preliminary and hypothesis-generating. We hope this report highlights the need for future multicenter and registry-based studies to develop feasible approaches for obtaining objective RV and hemodynamic monitoring data in unstable patients, refine physiologic triggers for ECMO escalation, and validate patient selection, reproducibility, and long-term outcomes.

CRediT authorship contribution statement

Enock Adjei: Conceptualization, Methodology, Investigation, Data curation, Writing – original draft. Blaine Sklar: Data curation, Writing – review & editing. John W. Stokes: Writing – review & editing. Whitney D. Gannon: Writing – review & editing. Amir Teimouri Dereshgi: Writing – review & editing. Anil J. Trindade: Writing – review & editing. Caitlin T. Demarest: Writing – review & editing. Matthew Bacchetta: Conceptualization, Supervision, Writing – review & editing. Konrad Hoetzenecker: Conceptualization, Supervision, Project administration, Writing – review & editing. All authors approved the final manuscript and agree to be accountable for all aspects of the work.

Ethics approval and consent

Informed consent was obtained from all patients included in this case series. The study was reviewed by the Vanderbilt University Medical Center Institutional Review Board and deemed exempt from formal review.

Funding

This work was supported by Mrs. and Mr. Ragland Fund, Vanderbilt University Medical Center.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.