Secondary Logo

Journal Logo

Subclavian Artery Cannulation for Venoarterial Extracorporeal Membrane Oxygenation

Javidfar, Jeffrey*; Brodie, Daniel; Costa, Joseph*; Miller, Joanna*; Jurrado, Julissa*; LaVelle, Matthew*; Newmark, Alexis*; Takayama, Hiroo*; Sonett, Joshua R.*; Bacchetta, Matthew*

doi: 10.1097/MAT.0b013e318268ea15
Clinical Cardiovascular/Cardiopulmonary Bypass

Femoral artery cannulation for venoarterial extracorporeal membrane oxygenation (ECMO) can be associated with ischemic and neurologic complications. The subclavian artery offers an alternative cannulation site, which is helpful in patients with peripheral vascular disease, in those who have sustained pelvic trauma, or when ambulation is anticipated. This is a single-institution review of 20 adults who were placed on venoarterial ECMO using subclavian arterial cannulation over a 2 year period. Technical success with subclavian venoarterial ECMO was 100%. Median ECMO time was 168 hours (2.4–720 hours). Sufficient flows (median 4.24 L/min), oxygenation (median postcannulation PaO2 315 mm Hg), and ventricular unloading confirmed with intraoperative transesophageal echocardiogram were achieved in all patients. Seventy-five percent of patients were decannulated, 50% were extubated, and 45% were discharged. Seven patients (35%) had an entirely upper body ECMO configuration with the internal jugular vein serving as the venous drainage site. Complications included arterial cannula site hematoma and infection, as well as ipsilateral arm swelling. Each required conversion to femoral artery cannulation. There were no ischemic or neurologic complications. Patients with acute cardiopulmonary failure can safely be placed on subclavian venoarterial ECMO for prolonged periods with full flows, adequate oxygenation, and sufficient ventricular unloading.

From the *Division of Cardiothoracic Surgery, Department of Surgery, Columbia University Medical Center, New York, NY; and Division of Pulmonary, Allergy and Critical Care Medicine, Department of Medicine, Columbia University Medical Center, New York, NY.

Presented at the Annual Meeting of ASAIO in June 2011, Washington, DC.

Disclosures: Drs. Brodie and Bacchetta previously did consulting work for Maquet Cardiovascular.

Reprint Requests: Matthew Bacchetta, MD, Columbia University Medical Center, Herbert Irving Pavilion, 161 Fort Washington Ave, Suite 301, New York City, NY 10032. Email:

Extracorporeal membrane oxygenation (ECMO) has evolved as a life-saving measure for patients requiring emergent support of respiratory and cardiac function. The femoral artery is the standard site for immediate vascular access when initiating adult venoarterial ECMO. However, femoral artery cannulation is not without significant risks such as femoral artery occlusion, distal limb ischemia, reperfusion injury, resulting in compartment syndrome, retroperitoneal hemorrhage, thrombosis, embolization, and, most importantly, brain and myocardial ischemia as a result of poor perfusion to the upper body.1 In patients with severe peripheral vascular disease, these risks are heightened and may be considered a relative contraindication to femoral artery cannulation.

Subclavian artery cannulation is a viable alternative method for preserving cerebral and cardiac blood flow, while limiting the risks associated with femoral artery cannulation. Specifically, in patients with severe atherosclerotic and aneurysmal disease, the subclavian and axillary arteries tend to be disease free, which minimizes the risks of cerebral embolization.2,3 Similarly, in patients who have sustained pelvic trauma and go on to require venoarterial ECMO, upper body cannulation may be preferable.

Other benefits of subclavian artery cannulation include “central” support with antegrade flow while avoiding the need for sternotomy and direct aortic cannulation frequently required for adult central ECMO perfusion.2 The shoulder girdle and its associated vessels benefit from rich collateral flow from the thyrocervical trunk to the suprascapular and transverse cervical arteries. This collateral flow helps to avert upper extremity ischemia in the event of total distal occlusion. The risk of distal limb ischemia can be further minimized by using a graft sewn in end-to-side fashion to the artery.4

In our center, use of the distal subclavian artery with a side graft is the preferred method for venoarterial ECMO cannulation in patients who are stable enough for transport to the operating room. The use of a side graft also allows for indirect pressure monitoring of cerebral perfusion. Svensson5 reported that a higher antegrade selective cerebral perfusion pressure was associated with greater risk of stroke and neurocognitive deficit. Side-graft cannulation avoids the undesired high-pressure oscillations that can occur as a consequence of constant antegrade selective cerebral flow during direct subclavian artery cannulation.6

We describe our experience with cannulation of an end-to-side graft sewn onto the distal subclavian artery for venoarterial ECMO in adults who require urgent or emergent mechanical circulatory support.

Back to Top | Article Outline

Patients and Methods

Patient Selection

The study, which was approved by the Columbia University Medical Center institutional review board, is a retrospective review of a single institution’s experience with a subclavian artery cannulation technique in urgent or emergent settings for adult patients on ECMO. Over a 2-year period ending in November 2011, 70 patients were placed on venoarterial ECMO. Subclavian artery cannulation was carried out by one of two cardiothoracic surgeons in 20 patients using the technique described below.

The decision to place patients on venoarterial ECMO was made by a team composed of cardiothoracic surgeons and critical care intensivists. The indications for venoarterial ECMO included persistent cardiogenic shock or refractory respiratory failure in the setting of severe pulmonary hypertension with right heart failure. In the case of persistent cardiogenic shock, the patient had experienced a cardiac arrest but was resuscitated with advanced cardiac life support. After a stable rhythm and perfusion pressure were restored, the patient was determined to be a candidate for mechanical circulatory support and was placed on venoarterial ECMO as a bridge to decision. The causes of respiratory failure included hypoxemic respiratory failure from acute respiratory distress syndrome or hypercarbic respiratory failure as seen in patients with cystic fibrosis awaiting lung transplantation. Patients with primary respiratory failure who were inadequately supported with our preferred approach of venovenous ECMO were placed on venoarterial ECMO using the subclavian artery.

Back to Top | Article Outline


Our protocol for subclavian venoarterial ECMO involves a standardized arterial cannulation with variant peripheral venous cannulation techniques. The distal right subclavian artery is exposed through a small subclavicular incision (Figure 1). An 8 mm Gelweave graft (Terumo, Leeds, UK) is sewn in “end-to-side” fashion onto the artery, distal to the branching of the thyrocervical trunk. This is consistent with standard catheterization techniques used in adult cardiac surgery. Pledgets and hemostatic agents are used as needed. A 24 French elongated one-piece arterial (EOPA) ECMO cannula (Medtronic, Brooklyn Park, MN) is then tunneled through a separate site and inserted into the graft (Figure 2). Depending on the size of the axillary artery and patient, a combination of a 6 mm graft and an 18 French EOPA ECMO cannula (Medtronic) can also be used to provide support. The cannula is in turn secured to the graft by a series of permanent ties.

Figure 1

Figure 1

Figure 2

Figure 2

The choice for venous access sites and cannulae used depend on the patient and the clinical scenario. As previously described, femoral venous cannulation is carried out percutaneously.7 If required for additional device inflow, a second venous cannula can be placed in the internal jugular vein using a Seldinger technique. Alternatively, an entirely upper body modification of this technique involves the cannulation of the right internal jugular vein using either a 23 French Arterial Biomedicus cannula (Medtronic) or the Avalon Elite bicaval dual lumen catheter (Avalon Laboratories LLC, Rancho Dominugez, CA) (Figure 2). This allows for excellent drainage. In patients who have a pre-existing dual lumen cannula, this also eases conversion from venovenous to venoarterial cannulation. As described in the literature, image guidance in the form of a transesophageal echocardiogram or fluoroscopy is required to ensure proper placement and orientation of this dual lumen cannula.8 The ECMO circuit is attached to the cannula ports, carefully avoiding air entrapment, and then ECMO is instituted. The circuit consisted of a Quadrox D oxygenator (Maquet Inc., Rastatt, Germany) and a centrifugal pump, which was either a Rotaflow (Maquet Inc.) or a Levitronix Centrimag (Levitronix, GmbH, Zurich, Switzerland).

The patients were managed on ECMO according to established low-dose anticoagulation and lung-protective ventilation protocols.9 Decannulation consisted of ligating the graft either using a vascular stapler or by oversewing the graft after cannula removal.

Back to Top | Article Outline

Statistical Methods

Overall survival was defined as time from placement on ECMO to death or last follow-up to November 30, 2011. Median values were provided with interquartile ranges (IQRs). To quantify the severity of organ dysfunction in the study population, initial (pre-ECMO), highest, and mean sequential organ failure assessment (SOFA) scores were determined for each patient. The Wilcoxon–Mann–Whitney rank sum test, Fischer’s exact test, and univariate logistical regression were used for comparisons and subgroup analysis. Primary outcomes were successful cannulation and weaning from ECMO. Secondary outcomes were liberation from mechanical ventilation, discharge from the hospital, and survival. Also noted were complications such as catheter-related infection, thrombosis, and cannulation site bleeding events. All calculations were conducted using Stata/IC 11.0 software (StataCorp, College Station, TX). For the purpose of assessing statistical significance, a conventional alpha of 0.05 was used.

Back to Top | Article Outline


Twenty patients were placed on venoarterial ECMO using the subclavian artery cannulation technique with 100% technical success. One patient was converted from venovenous to venoarterial ECMO and the venous cannula was kept in place. All patients in this cohort were emergently or urgently placed on subclavian venoarterial ECMO and were critically ill with severe organ dysfunction. The median initial, high, and mean SOFA scores for this cohort were 11, 13.5, and 11.5, respectively. There was a slight female predominance (55%) and the median age of this population was 54.5 years (IQR, 44–65 years). The remaining patient characteristics are listed in Table 1. The etiology of cardiac arrest in those 10 patients with refractory cardiogenic shock included acute myocardial infarction (n = 4), exacerbation of nonischemic cardiomyopathy (n = 5), and septic shock (n = 1).

Table 1

Table 1

Using the subclavian configuration, sufficient flows (median 4.24 L/min), adequate gas exchange (median postcannulation PaCO2 and PaO2 were 37 and 315 mm Hg, respectively), and ventricular unloading on transesophageal echocardiogram were confirmed in all patients. This was an improvement from the pre-ECMO PaCO2 and PaO2, which were 50 and 61 mm Hg, respectively (Tables 2 and 3).

Table 3

Table 3

Table 2

Table 2

There was no need for additional mechanical circulatory support in this study population. The mean total procedure time was 46 minutes, including patient positioning and confirmation of venous cannula location using fluoroscopy or endoscopic ultrasound. The mean time to initiation of ECMO was 33 minutes.

Seventy-five percent of patients were decannulated and 50% were liberated from mechanical ventilation. This cohort had a 45% survival and discharge rate from the hospital. Three patients were bridged to a biventricular assist device using the existing arterial cannula configuration. Two patients in this subgroup were extubated and one was discharged alive from the hospital (Table 4).

Table 4

Table 4

Three patients had adverse events related to the arterial cannulation technique, each requiring conversion to femoral artery cannulation. One patient developed arterial cannulae site infection. Another developed a hematoma at the arterial cannulation site that required evacuation. A third patient developed ipsilateral arm swelling without neurologic or ischemic symptoms, but was still converted to femoral artery venoarterial ECMO as a precaution (Table 4). Comparably, there was no significant difference in the rates of adverse events based on arterial cannulation site (p = 0.72). In the population who underwent femoral artery cannulation (n = 50) during the same time period, there were three incidences of vascular injury. Two patients had injuries to the common femoral artery or its bifurcation with resultant hematoma and one patient had a superficial femoral artery injury and developed an arteriovenous fistula. All three required surgical intervention. An additional three patients had a distal perfusion catheter placed to counter ischemic issues. The aggregate rate of adverse events for femoral artery cannulation was 12%. In the venoarterial ECMO population, during this time period, there no was significant survival advantage based on femoral versus subclavian artery cannulation (p = 0.65).

The causes of death in the cohort who was not decannulated included multiorgan system dysfunction (n = 4) and hemorrhagic shock (n = 1). In the subgroup that was decannulated, the causes of death included multiorgan system dysfunction (n = 3), septic shock (n = 1), cardiac arrest (n = 1), and withdrawal of life support secondary to failure to recover brain activity after initial arrest (n = 1). This included one patient who was extubated, but expired from multiorgan system dysfunction while in the intensive care unit.

Back to Top | Article Outline


The adult ECMO program at our center uses subclavian artery cannulation as an alternative site for providing venoarterial ECMO to patients, without the need for median sternotomy or without incurring the complications associated with femoral artery cannulation. Tunneled cannulation of an end-to-side graft that is sewn onto the distal subclavian artery can provide the benefits of central cannulation with the relative technical ease and lower complication profile of peripheral cannulation. The makeup of the subgroups was comparable and evenly split in their indication for mechanical support between a primarily acute cardiac event and respiratory failure with concomitant need for ventricular decompression. Both subgroups demonstrated adequate oxygenation and gas exchange using the subclavian cannulation configuration for venoarterial ECMO.

As previously demonstrated by Svensson et al.,3 upper body cannulation via the axillary or subclavian arteries may be superior to other sites of cannulation. The majority of the work examining the safety and efficacy of peripheral upper body cannulation involves the axillary artery. Few reports are available describing the role of the subclavian artery. Our technique involves cannulating the subclavian artery distal to the takeoff of the thyrocervical trunk and near the axillary–subclavian junction. Therefore, the axillary data are applicable.

According to the literature, axillary and subclavian artery cannulation injury rates range between 0% and 2% with some centers reporting as high as 5%.10 Comparably, complications associated with femoral artery cannulation range from 3% to 8.6%.11 We achieved 100% technical success, but three patients went on to experience minor adverse events directly associated with the subclavian arterial cannulation site. The incidence was similar, but less severe than the adverse events seen in the cohort that underwent femoral artery cannulation during the same time period. The low rates of bleeding from the subclavian artery site were like due to meticulous surgical hemostasis and use of a low-dose heparin anticoagulation regimen.

The risks of cannulation site infection are at the least comparable between the femoral and subclavian artery sites. With this technique, the risks may be less in the latter site due in part to using a tunneled cannula, performing insertions only in a sterile operating room (versus at the bedside), and having a meticulous cannula site care and dressing change protocol.12

The entirely upper body modification has the added advantage of allowing adults on venoarterial ECMO the opportunity to participate in physical therapy. Patients may be able to get out of bed to a chair and even ambulate with a mobile ECMO circuit. The seven patients with complete upper body cannulation sat up and participated in physical therapy. Two of the seven patients were extubated, rode an upright stationary bike, and ambulated with assistance around the intensive care unit. Participation in physical therapy while on mechanical circulatory support affords the patient the opportunity to avoid the deconditioning associated with a prolonged intensive care unit stay for cardiac or pulmonary failure.13

The survival figures in the study population of venoarterial ECMO were comparable to national trends for venoarterial ECMO. The cohort consists of patients with severe concomitant comorbidities such as pulmonary hypertension. As one would expect from patients who require emergent or urgent venoarterial ECMO support, the pre-ECMO SOFA scores in the study population were high. The SOFA score is an objective and verified predictor of mortality in the critically ill patients; it takes into account factors such as P:F ration, vasopressor requirements, and select laboratory values. Based on literature, the elevated SOFA score for this study population correlates at best with a 15% to 20% survival rate.14 However, in this cohort supported on ECMO, the survival was 55%.

Admittedly, unlike percutaneous femoral cannulation, subclavian artery cannulation can be technically challenging and time-consuming. It involves a requisite trip to the operating room, which mandates that the patient be stable enough for transport and its upper body location prohibits using the technique during the active administration of chest compressions. Further limitations of this cannulation technique include a mandatory closure of the graft conduit at decannulation. However, it should be noted that the site of femoral artery cannulation occasionally requires primary surgical repair. Limitations of the study include its single-institution, retrospective analysis of nonrandomized patients.

Subclavian artery cannulation provides a safe and perhaps improved means for providing venoarterial ECMO support in a limited segment of patients who are stable enough to be transported to the operating room. We anticipate routinely using entirely upper body cannulation as a means of minimizing deconditioning in patients with right heart dysfunction and pulmonary hypertension who are being bridged to transplant or recovery from acute insults by allowing them to participate more readily in physical therapy. Additional studies will be required to quantify potential benefits seen in these patients compared to those who are bridged to transplant with mechanical circulatory support using more conventional cannulation strategies.

Back to Top | Article Outline


1. Stulak JM, Dearani JA, Burkhart HM, Barnes RD, Scott PD, Schears GJ. ECMO cannulation controversies and complications. Semin Cardiothorac Vasc Anesth. 2009;58:176–182
2. Navia JL, Atik FA, Beyer EA, Ruda Vega P. Extracorporeal membrane oxygenation with right axillary artery perfusion. Ann Thorac Surg. 2005;79:2163–2165
3. Svensson LG, Blackstone EH, Rajeswaran J, et al. Does the arterial cannulation site for circulatory arrest influence stroke risk? Ann Thorac Surg.;78:1274–1284 discussion 1274, 2004
4. Kokotsakit J, Lazopoulos G, Milonakis M, et al. Right axillary artery cannulation for surgical management of the hostile ascending aorta. Tex Heart Inst J.;32:189–193 discussion 185, 2005
5. Svensson LG. Antegrade perfusion during suspended animation? J Thorac Cardiovasc Surg. 2002;124:1068–1070
6. Yilik L, Emrecan B, Kestelli M, et al. Direct versus side-graft cannulation of the right axillary artery for antegrade cerebral perfusion. Tex Heart Inst J. 2006;33:310–315
7. Javidfar J, Brodie D, Takayama H, et al. Safe transport of critically ill adult patients on extracorporeal membrane oxygenation support to a regional extracorporeal membrane oxygenation center. ASAIO J. 2011;57:421–425
8. Javidfar J, Wang D, Zwischenberger JB, et al. Insertion of bicaval dual lumen extracorporeal membrane oxygenation catheter with image guidance. ASAIO J. 2011;57:203–205
9. Brodie D, Bacchetta M. Extracorporeal membrane oxygenation for ARDS in adults. N Engl J Med. 2011;365:1905–1914
10. Schachner T, Nagiller J, Zimmer A, Laufer G, Bonatti J. Technical problems and complications of axillary artery cannulation. Eur J Cardiothorac Surg. 2005;27:634–637
11. Tiwari KK, Murzi M, Bevilacqua S, Glauber M. Which cannulation (ascending aortic cannulation or peripheral arterial cannulation) is better for acute type A aortic dissection surgery? Interact Cardiovasc Thorac Surg. 2010;10:797–802
12. Javidfar J, Brodie D, Wang D, et al. Use of bicaval dual-lumen catheter for adult venovenous extracorporeal membrane oxygenation. Ann Thorac Surg.;91:1763–8 discussion 1769, 2011
13. Turner DA, Cheifetz IM, Rehder KJ, et al. Active rehabilitation and physical therapy during extracorporeal membrane oxygenation while awaiting lung transplantation: A practical approach. Crit Care Med. 2011;39:2593–2598
14. Ferreira FL, Bota DP, Bross A, Mélot C, Vincent JL. Serial evaluation of the SOFA score to predict outcome in critically ill patients. JAMA. 2001;286:1754–1758
Copyright © 2012 by the American Society for Artificial Internal Organs