Secondary Logo

Journal Logo

Gas Exchange

Percutaneous Venovenous CO2 Removal With Regional Anticoagulation in an Ovine Model

Cardenas, Victor J. Jr*; Miller, Lucinda*; Lynch, James E.; Anderson, Michael J.; Zwischenberger, Joseph B.

Author Information
doi: 10.1097/01.mat.0000227743.07743.5d
  • Free


In the past decade, lung-protective strategies in mechanical ventilation for acute respiratory distress syndrome (ARDS) have resulted in improved survival.1,2 However, in the patient with severe ARDS, limitation of airway pressures and tidal volumes may result in significant hypoventilation, acute respiratory acidosis, and deterioration of oxygenation. Extracorporeal CO2 removal (ECCO2R) coupled with low tidal volume mechanical ventilation has been demonstrated to reduce airway pressure while allowing adequate gas exchange and maintenance of acid-base homeostasis in the adult patient with ARDS.3–9 The acceptance of extracorporeal gas exchange devices as support for acute respiratory failure has been limited to specialized centers and selected patient populations.

Traditionally, extracorporeal gas exchange has been expensive, labor-intensive, and complex.7,9–13 Extracorporeal systems such as venoarterial extracorporeal membrane oxygenation (VA ECMO) and arteriovenous CO2 removal (AVCO2R) require arterial cannulation, whereas venovenous systems (VV ECMO, ECCO2R) require two large-bore venous catheters. All extracorporeal systems require systemic heparin administration at varying doses.

We designed a system that eliminates the risk of arterial cannulation and heparin anticoagulation/heparin-induced thrombocytopenia by using regional citrate anticoagulation similar to that used in hemodialysis.14–17 We eliminated the need for two cannulation sites with the use of a single, percutaneously placed double-lumen venous catheter and used a low-resistance, hollow-fiber oxygenator and a roller pump to complete the circuit. We tested the circuit for CO2 transfer capabilities and evaluated the efficacy of regional anticoagulation in an ovine model for up to 24 hours.

Materials and Methods

This study was conducted with the approval of The Institutional Animal Care and Use Committee (IACUC) of the University of Texas Medical Branch, Galveston, Texas. An independent member of the animal care facility monitored the experiment for compliance with IACUC guidelines. All animals used in the study received humane care in compliance with the Institute of Laboratory Animal Resources and the National Research Council. The care of the animals is outlined in the Guide for the Care and Use of Laboratory Animals, prepared by and published by the National Academy Press, revised 1996.

Adult sheep (n = 5; weight, 28 to 39 kg) were sedated and induced with ketamine (10 mg/kg IM, 5 mg/kg IV, respectively) and then intubated. Anesthesia was maintained with 1% to 2.5% isoflurane delivered with an anesthesia ventilator (Ohmeda 7000; BOC Health Care, Liberty Corner, NJ). A tracheostomy was performed. Femoral arterial, femoral venous, and pulmonary artery catheters were placed. Volume control mechanical ventilation (Servo 900C; Siemens-Elema, Sweden), tidal volume (VT) of 6 to 9 ml/kg, respiratory rate (RR) of 10 to 12 breaths/min (bpm), and a positive end-expiratory pressure (PEEP) of 5 cm H2O was initiated. Fio2 was initiated and maintained at 1.0 throughout the experiment.

A single 15-cm, 18F, double-lumen extracorporeal membrane oxygenation (ECMO) cannula (Origen, Austin, TX) was inserted percutaneously into the internal jugular vein connecting the sheep to our modified extracorporeal circuit consisting of quarter-inch tubing and a roller pump (Polystan Värlöse, Denmark) interposed with a hollow-fiber oxygenator (Affinity, Avecor Cardiovascular, Plymouth, MN) (Figure 1). Normothermia was maintained by using a surface warmer. One subject required use of the integrated heat exchanger in the oxygenator. The circuit was primed with 400 to 500 ml of 0.9% NaCl. No heparin was used. Regional anticoagulation was initiated by using a hypertonic trisodium citrate solution. The citrate solution (200 g/L in sterile water) was infused into the inflow port of the circuit and titrated to achieve a post-oxygenator ionized calcium level of 0.25 mmol/L. Calcium chloride solution (100 mg/ml) was infused into a central vein and adjusted to maintain systemic ionized calcium levels within the normal range. Systemic and post-oxygenator ionized calcium levels were drawn hourly. The extracorporeal circuit was visibly inspected for evidence of clot formation.

Figure 1.
Figure 1.:
Schematic representation of the extracorporeal circuit and ovine model.

After blood flow through the circuit was established, the respiratory rate was decreased from 12 bpm to 5 bpm and the tidal volume decreased from 6 to 9 ml/kg to 3 to 4 ml/kg in one step. Ventilator settings remained unchanged after the initiation of extracorporeal blood flow. Sedation was maintained with ketamine and/or pentobarbital to control respirations and prevent cannula dislodgment. The sheep were monitored for bleeding, fluid status, and hemodynamic stability. Heart rate and arterial blood pressure were monitored continuously and recorded every 4 hours. Pulmonary artery pressure and cardiac output were measured and recorded every 4 hours.

After the reduction in minute ventilation, 100% O2 gas flow was initiated through the hollow-fiber oxygenator. CO2 transfer rates across the oxygenator were calculated by measuring the percent carbon dioxide exiting the oxygenator with a capnograph (Cosmo, Novametrix, Wallingford, CT) and multiplying the value by the gas flow to obtain CO2 transfer. CO2 transfer was measured under several combinations of circuit blood flows and gas flows. Blood flow was initially set at 500 ml/min, and gas flow was set at 2 l/min. CO2 transfer rate was measured after 1 hour on those settings to allow a steady state to develop. Gas flow was then increased in steps to 4, 6, 10, and 15 l/min with CO2 transfer rate measured 1 hour after each step change. After values for the five steps had been obtained, blood flow was increased to 800 ml/min and the process was repeated; the same process was also performed at the 1000 ml/min level. Arterial and mixed venous blood gases, as well as pre-oxygenator and post-oxygenator blood gases were obtained every hour, along with CO2 transfer rates.

The circuit was maintained for 24 hours, after which time the sheep were killed. Plasma hemoglobin was drawn before the end of the experiment to determine the presence of hemolysis. At the termination of the study, all components of the extracorporeal circuit were disassembled, flushed with saline, and visibly inspected for thrombus formation.

Statistical Analysis

All tests were derived with the use of software package SigmaStat version 2.03, SPSS, Inc. Table data are presented as mean value ± standard deviation. Comparison between groups and analysis of interactions between blood and gas flow were analyzed by two-way ANOVA. Linear regression equations used the least-squares method, with a p value ≤ 0.5.


Venovenous carbon dioxide removal with regional citrate anticoagulation did not adversely affect sheep hemodynamic stability. Heart rate (HR), mean arterial pressure (MAP), and cardiac output (CO) were within normal range at all blood flows studied. The mean values of all measurements at each blood flow are presented in (Table 1). The pressure drop across the oxygenator was only 10 mm Hg at 1000 ml/min blood flow.

Table 1
Table 1:
Hemodynamic Parameters

Carbon dioxide transfer across the oxygenator increased with increments in blood flow and gas flow (Figure 2). Maximum CO2 transfer of 150 ml/min occurred at a blood flow of 1000 ml/min and a gas flow of 15 l/min. Changes in gas flow resulted in significant differences between groups independent of blood flow (p < 0.001). Similarly, changes in blood flow resulted in significant differences between groups independent of gas flow (p < 0.018). There was no significant interaction between blood flow (BF) or gas flow (GF) on CO2 transfer by two-way ANOVA (p = 0.917). Multiple linear regression resulted in the predictive equation: CO2 transfer (ml/min) = 2.808 + (.003503 * GF) + (0.0619 * BF); R = 0.714, p < 0.001.

Figure 2.
Figure 2.:
Mean CO2 removal with standard deviation is displayed as function of blood flow through the circuit. Blood flow groups are further subdivided by gas flow. CO2 removal increases as blood flow increases. For a given blood flow, increments in gas flow also increase CO2 removal. No definite plateau in carbon dioxide transfer was seen at these blood and gas flows.

Extracorporeal CO2 transfer allowed up to a 75% reduction in minute ventilation while maintaining normocapnia. Arterial Pao2 remained >1 00 mm Hg in all the animals throughout the study. Although post-oxygenator pH was >7.8, we did not detect any overt impact on our subjects. Alkalemia was rapidly buffered in the central venous blood, resulting in a normal mixed venous pH (Table 2).

Table 2
Table 2:
Blood Gas Values

Citrate was used as a regional anticoagulant for our extracorporeal circuit but at higher blood flows than have been previously described.17,18 The ionized calcium levels in the circuit were persistently higher than our target level of 0.25 mmol/l. A progressive decline in the calcium chelation was observed with increasing blood flow (Figure 3). Inspection of the circuit tubing revealed no evidence of thrombus up to 24 hours, and plasma hemoglobin remained in the normal range for all animals. Clinical hemorrhage was not observed in any of the animals.

Figure 3.
Figure 3.:
Maximum arterial minus circuit ionized calcium concentration values obtained for each blood flow group are shown (mean ± SEM). A progressive decline was noted with increasing blood flow: [Ca2+] gradient = 0.332 – (0.156 * 10–3 * blood flow); R = 0.340, p = 0.032.


Extracorporeal carbon dioxide removal coupled with low tidal volume mechanical ventilation has been used in the treatment of the adult patient with acute respiratory failure. Use of an extracorporeal circuit for gas exchange has been complicated by the need for multiple catheter sites and systemic anticoagulation.3,8,9,12,18 We demonstrated that CO2 removal can be safely accomplished with regional anticoagulation in a low-flow extracorporeal circuit. Using a double-lumen venous catheter interposed with a roller pump and a hollow fiber oxygenator, we were able to demonstrate CO2 removal up to 130 ml/min. Carbon dioxide removal allowed up to a 75% reduction in minute ventilation in normal sheep while maintaining normocapnia.

Pesenti et al.4 and Gattinoni et al.5,6 described the use of dual venovenous catheter extracorporeal CO2 removal in patients with severe respiratory failure. We used a double-lumen pediatric ECMO cannula placed percutaneously into the internal jugular vein to provide access for the extracorporeal circuit. Advantages of percutaneous catheter placement include simplicity and improved hemostasis compared with surgical cutdown. Percutaneous access of the internal jugular vein can be easily performed at the bedside by physicians experienced in central vein access and Seldinger technique. A double-lumen catheter eliminates the need for multiple catheter sites and may result in improved patient comfort with need for less sedation.

Brunston et al.,12 Alpard et al.,13 and Zwischenberger et al.19 simplified the circuit design for extracorporeal CO2 removal in patients with ARDS by using an arteriovenous loop (AVCO2R). In AVCO2R, the arterial pulse pressure provides sufficient blood flow through a low-resistance, hollow-fiber oxygenator to effectively remove CO2 without need for a pump. However, in patient populations such as the elderly, where significant peripheral vascular disease is more common, there are concerns about the thrombotic and occlusive complications associated with arterial access. A venovenous configuration avoids the risk of arterial occlusion and potential limb loss. The venovenous configuration also eliminates the risk of significant hemorrhage associated with arterial decannulation in a systemically anticoagulated patient.

Extracorporeal CO2 removal, as described in the literature to date, requires systemic anticoagulation with heparin. Bleeding at the catheter site and hemorrhage at distant sites are the most common complications reported with extracorporeal circuits.20,21 Regional citrate anticoagulation has been associated with decreased bleeding complications during hemodialysis.14–17 Citrate infused into the inflow port chelates calcium in the extracorporeal circulation. The low calcium environment inhibits calcium-mediated cell activation of coagulation. A calcium chloride solution given intravenously restores systemic ionized calcium and the coagulation pathway.22

Using a protocol similar to those described for continuous venovenous hemodialysis, we were unable to achieve the predicted calcium concentration in the circuit without inducing systemic hypernatremia or hypocalcemia.15,16 The reduction in [Ca2+] in the circuit reached a plateau value after which further increases in citrate infusion resulted in arterial hypocalcemia. Increases in systemic calcium infusion reversed the arterial hypocalcemia, but the circuit [Ca2+] would rise in parallel. The arterial to circuit [Ca2+] difference remained constant and was roughly correlated with circuit blood flow (Figure 3). The inability to reach the target values may be explained by our substantially higher blood flows when compared with continuous venovenous hemodialysis. Alternately, ovine coagulation may be more resistant to effects of calcium chelation.

Despite an inability to achieve target levels of ionized calcium, there was no gross thrombus formation noted in the tubing at the termination of the experiments. Circuit gas exchange and pre-oxygenator pressure remained stable throughout the experiments, which suggests an absence of widespread microthrombosis in the oxygenator. Our higher flow rates may prevent areas of stagnant circulation, requiring less anticoagulation. Also, the low resistance oxygenator may minimize excessive shear forces and reduce blood surface interactions (hemolysis, platelet activation, and so forth). We believe that citrate can provide short-term effective regional anticoagulation without the bleeding complications commonly reported with systemic heparin.

We demonstrated that for a given blood flow, incremental increase in gas flow up to 15 l/min increases CO2 transfer. Similarly, increases in blood flow at a given gas flow also resulted in increased CO2 removal. We limited blood flow to 1000 ml/min to minimize pulsation in the double-lumen catheter at the insertion site and prevent local injury. We limited gas flow to the 15 l/min achievable by flowmeters in common clinical use. At CO2 removal rates of 100 to 130 ml/min, we anticipate a reduction in alveolar ventilation requirements of approximately 50%.

Our study has several limitations. The small number of subjects limits the power of our statistical analysis. The amount of recirculation as a function of increasing blood flow was not measured. However, there did not appear to be a significant detrimental effect on CO2 transfer, given the linear relation of blood flow to CO2 by multiple linear regression. The duration of the study was only up to 24 hours. There was no evidence of gross thrombus formation in the circuit, nor was there deterioration of oxygenator function. Our subjects were normal sheep, therefore, the efficacy of anticoagulation in an injury model with coagulation abnormalities such as severe sepsis was not addressed.


We have demonstrated significant extracorporeal CO2 removal by using a compact circuit and regional anticoagulation in an ovine model. We used clinically available components and developed procedures that can be performed in most intensive care units. VVCO2R may be a viable option or adjunct to mechanical ventilation for acute respiratory failure.


1. Polkey M: Muscle metabolism and exercise tolerance in COPD. Chest 5: 1315–1355, 2002.
2. Menzies R, Gibbson W, Goldberg P: Determinants of weaning and survival among patients with COPD who require mechanical ventilation for acute respiratory failure. Chest 95: 398–405, 1998.
3. Brunet F, Mira JP, Dhainaut JF, et al: Efficacy of low frequency positive pressure ventilation-extracorporeal CO2 removal. Am J Respir Crit Care Med 151: 1269–1270, 1994.
4. Pesenti A, Kolobow T, Gattinoni L: Extracorporeal respiratory support in the adult. ASAIO Trans 34: 1006–1008, 1988.
5. Gattinoni L, Pesenti A, Masheroni D, et al: Low-frequency positive pressure ventilation with extracorporeal CO2 removal in severe acute respiratory failure. JAMA 256: 881–886, 1986.
6. Gattinoni L, Kolobow T, Damia G: Extracorporeal carbon dioxide removal (ECCO2R): a new form of respiratory assistance. Int J Artif Organs 2: 183–185, 1979.
7. Pesenti A, Rossi G, Pelosi P, et al: Percutaneous extracorporeal CO2 removal in a patient with bullous emphysema with recurrent bilateral pneumothoraces and respiratory failure. Anesthesiology 72: 571–573, 1990.
8. Pranikoff T, Hirschl R, Remenapp R, et al: Venovenous extracorporeal life support via percutaneous cannulation in 94 patients. Chest 115: 818–822, 1999.
9. Alpard S, Zwischenberger J, Tao W, et al: Reduced ventilator pressure and improved P/F ratio during percutaneous arteriovenous carbon dioxide removal for severe respiratory failure. Ann Surg 230: 215–224, 1999.
10. Deslauriers J, Awad JA: Is extracorporeal CO2 removal an option in the treatment of adult respiratory distress syndrome? Ann Thorac Surg 64: 1581–1582, 1997.
11. Pesenti A, Romagnoli G, Fox U: Saphenous-saphenous cannulation for LFPPV-ECCO2R. Paper presented at: 10th Congress of European Society of Artificial Organs; Bologna, Italy. 1983.
12. Brunston R, Zwischenberger J, Tao W, et al: Total arteriovenous CO2 removal: simplifying extracorporeal support for respiratory failure. Ann Thorac Surg 64: 1599–1605, 1997.
13. Alpard S, Zwischenberger J, Tao W, et al: Reduced ventilator pressure and improved P/F ratio during percutaneous arteriovenous carbon dioxide removal for severe respiratory failure. Ann Surg 230: 215–224, 1999.
14. Meier-Kriesche HU, Finkel KW, Gitomer JJ, et al: Unexpected severe hypocalcaemia during continuous venovenous hemodialysis with regional citrate anticoagulation. Am J Kidney Dis 33: 8, 1999.
15. Kutsogiannis DJ, Mayers I, Chen WD, et al: Regional citrate anticoagulation in continuous venovenous hemodiafiltration. Am J Kidney Dis 35: 802–811, 2000.
16. Apsner R, Buchmayer H, Lang T, et al: Simplified citrate anticoagulation for high flux hemodialysis. Am J Kidney Dis 38: 979–987, 2001.
17. Abramson S, Niles JL: Anticoagulation in continuous renal replacement therapy. Curr Opin Nephrol Hypertens. 8: 701–707, 1999.
18. Morris AH, Menlove RL, Rollins RJ, et al: A controlled clinical trial of a new 3-step therapy that includes extracorporeal CO2 removal for ARDS. ASAIO Trans 34: 48–53, 1988.
19. Zwischenberger J, Alpard S, Tao W, et al: Percutaneous extracorporeal arteriovenous carbon dioxide removal improves survival in respiratory distress syndrome: a prospective randomized outcomes study in adult sheep. J Thorac Cardiovasc Surg 121: 542–551, 2001.
20. McCoy-Pardington D, Judd WJ, Knafl P, et al: Blood use during extracorporeal membrane oxygenation. Transfusion 30: 307–309, 1990.
21. Uziel L, Cugno M, Fabrizi I, et al: Physiopathology and management of coagulation during long-term extracorporeal respiratory assistance. Int J Artif Organs 13: 288–296, 1990.
22. Seaton RD, Duncan KA, Pinnick RV, et al: Regional citrate anticoagulation in chronic hemodialysis patients. Trans Am Soc Artif Intern Organs 29: 414–188, 1983.
Copyright © 2006 by the American Society for Artificial Internal Organs