Cerebral ischemia caused by obstructed superior vena cava cannula is detected by near-infrared spectroscopy

OBJECTIVE: Bicaval venous cannulation is being used with increasing frequency in neonates and infants to avoid circulatory arrest. However, superior vena cava (SVC) cannula obstruction may result in cerebral ischemia with no change in blood pressure or mixed venous O2 saturation. The authors hypothesized that near-infrared spectroscopy (NIRS) would allow noninvasive detection of SVC cannula obstruction. METHODS: Fifteen Yorkshire piglets (9.07 +/- 0.20 kg) underwent total cardiopulmonary bypass (CPB) (100 mL/kg/min, pH-stat strategy, hematocrit of 20%) with ascending aortic and bicaval cannulations. Femoral arterial and SVC pressure were monitored as well as mixed venous O2 saturation. NIRS monitoring of tissue oxygenation index (TOI) as well as oxyhemoglobin and deoxyhemoglobin (HHb) was undertaken. Animals were cooled to an esophageal temperature of 25 degrees C over 20 minutes. CPB flow was reduced to 50 mL/kg/min for 20 minutes. Animals then underwent a 60-minute study period of continuous CPB at 50 mL/kg/min with manipulation of the SVC cannula: group 1, open; group 2, partial occlusion; and group 3, complete occlusion. Animals were rewarmed to 37 degrees C at full flow with the SVC cannula open. Cerebral blood flow was assessed at onset of CPB, at end of cooling, at end of low flow, at end of SVC manipulation period, and at end of rewarming using radioactive microspheres. RESULTS: CBF decreased to 27.9 +/- 1.5 mL/min/100 g with complete occlusion (p < 0.01 v group 1: 39.7 +/- 1.9, group 2, 38.3 +/- 2.0 mL/min/100 g) with no change in arterial pressure or mixed venous saturation. There were also significant differences in cerebral oxygen delivery between group 3 and other groups (p < 0.01). SVC pressure increased to 19.5 +/- 4.5 and 32.5 +/- 3.1mmHg with partial and complete occlusion. NIRS indicated significant cerebral ischemia with a decrease in TOI (p < 0.05; group 3 v group 1 and 2) and an increase in HHb (p < 0.05; group 3 v group 1). At the end of the study, significant acidosis was found in group 3 compared with group 1 (p < 0.05). CONCLUSION: SVC cannula obstruction causes cerebral ischemia with no change in blood pressure or venous oxygen saturation. In view of the difficulties and risks of CVP monitoring in babies, it is recommended to use other monitoring modalities such as NIRS to assess adequacy of cerebral perfusion if bicaval cannulation is used in neonates and infants.