ADDITION OF VENO-VENOUS CO2 REMOVAL TO A DIALYSIS CIRCUIT IN A CHILD WITH A SINGLE VENTRICLE

John L. Morris, BS,^a David A. Rosen, MD,^b,^d Kelly S. Calvert, BSN, CCP, ^c Robert A. Gustafson, MD, ^c, ^d Robert J. Steelman, MD, ^d Kathleen R. Rosen, MD ^b, ^d and Dianne G. Muchant, MD ^d

From the School of Medicine,^a Departments of Anesthesiology, ^b Surgery, ^c and Pediatrics, ^d West Virginia University, Morgantown, WV., USA

Corresponding Author:
David A. Rosen, MD
West Virginia University School of Medicine, Department of Anesthesiology
3618 Health Sciences Center South, PO Box 9134
Morgantown, WV, USA
E-mail:

Management of the patient with single ventricle physiology places the patient at risk for increased venous pressures, which can lead to the development of ascites and pleural effusions. The increased fluid in the chest and abdomen can cause lung and hepatorenal compromise. In Fontan patients, techniques that lower mean airway pressure and lower PaCO2 are advantageous as flow through the lung is dependent on pulmonary resistance. Managing the patients with spontaneous ventilation is advantageous because the negative pressures will help to lower the central venous pressures when compared to positive pressure ventilation. Carbon dioxide removal can be impaired if the fluid in the abdomen and chest persists. A membrane oxygenator on the venous side may reduce the carbon dioxide to which the pulmonary system is exposed, thus optimizing flow through the pulmonary circuit. We report the addition of a membrane oxygenator in series with a veno-venous hemodialysis circuit in a pediatric patient.

CLINICAL SUMMARY

At birth this 6 year old (19.2 kg, 0.78 m2 BSA, CO = 1.87 L/min) boy was diagnosed with total anomalous pulmonary venous return, dextrocardia with asplenia, pulmonary stenosis, left superior vena cava, and double outlet right ventricle. He survived numerous operations to undergo final palliation via a modified Fontan on this admission. The patient's post-operative course was complicated by chronic pleural effusions and ascites, as well as Acinetobacter baumani sepsis leading to MODS (including hepatorenal syndrome and respiratory failure). His cardiac function as measured by ECHO remained adequate. Hemodialysis through a 7 French dual lumen femoral venous line was instituted for a rise in BUN, creatinine and fluid management secondary to nonoliguric renal failure. The blood filtration rate through the dialysis circuit was 100 cc/min (5.3% of patient's cardiac output) limited by catheter size. After 22 hrs of continuous veno-venous hemodialysis (CVVH-D) an oxygenator (CB Minimax or Lilliput I) was cut into the dialysis circuit (Figure 1). This required the use of a 1/4" to 3/16" adapter. The oxygenator was placed after the roller pump but before the dialysis filter, which allowed a standard dialysis bubble trap to be utilized. A heat source was not added and normothermia was maintained by external warming (blankets and heat lamps). For the first 3 hrs the oxygenator was run at room air with a 1:1 gas to blood flow ratio. Subsequently the FiO2 was increased to 60% with a 1:1 gas to blood flow ratio. PO2 and PCO2 before addition of an oxygenator, during veno-venous CO2 removal (VVCO2R), and after increasing blood flow (upsizing venous catheters) are shown in Table 1.

Tabla 1. Patient Respiratory Parameters.

	Before Oxygenator	After Oxygenator	After Increasing Catheter Size
PCO2 (mmHg)	54,7 (48,3-63,7)	45,2 (41,8-48,8)	43,3 (39,5-47,4)
PO2 (mmHg)	41,5 (31-46)	47,2 (39-56)	53 (50-60)
Saturations (%)	84	91	92
Mean Airway Pressure (cm H2O)	19 (19)	14 (12-18)	14 (12-15)
Peak Airway Pressure (cm H2O)	43 (43)	36 (35-38)	35 (33-38)
FiO2	0,85	0,81	0,79
Ventilatory Rate (bpm)	40	36	33

After 24 hours on VV CO2R, the femoral venous catheters were upsized to a 11.5 French double lumen catheter which allowed the blood filtration rate to be increased to 150 ml/min (8.0% of cardiac output). During the 96 hours of VVCO2R the oxygenator was changed five times due to protein leak from the gas flow exit port. No oxygenators failed. No additional heparin was used in addition to that for dialysis. Although the VVCO2R was effective, a decision was made to increase the patient's support by placing him on VA ECMO. The patient deteriorated on this mode of support and after 36 hours the parents chose to discontinue life support.

DISCUSSION

This report documents the first attempt to facilitate CO2 removal using a hemofiltration dialysis machine. Adding the oxygenator to the veno-venous dialysis circuit produced no deleterious effects and required no additional work on the part of the dialysis team who was able to continue to manage the system on a shift basis, eliminating the support crew usually needed to maintain a patient on ECMO. Throughout the VVCO2R the hemofiltration functioned as it did prior to addition of the oxygenator. The CO2 removal was effective in reducing the PaCO2 in the patient.

Nonventilatory modes of improving oxygenation and CO2 removal have been the subject of much study, beginning with the use of intravascular oxygenators (IVOX). This modality had a high incidence of adverse events (24.5% of trial patients) in one large series [1]. However, AVCO2R, a newer technique, has demonstrated promise in animal as well as limited human studies [2-4]. Conrad et al has shown that support at flows between 10-15% of CO, with a device diffusing capacity of 0.5 ml/min/torr/kg and a gas to blood flow ratio of 5 or greater, is adequate for complete CO2 removal [5]. We were able to produce an 18% reduction in PaCO2 with only 5% of the patient's cardiac output. AVCO2R was not used in this patient because it would have required additional arterial access and an arteriovenous shunt, which may have had adverse consequences on hemodynamics and vital organ perfusion.

REFERENCES