Hypercapnia Prevents Jugular Bulb Desaturation during Rewarming from Hypothermic Cardiopulmonary Bypass 

After institutional review board approval and informed patient consent were received, we enrolled 20 patients (American Society of Anesthesiologists physical status III) who were undergoing coronary artery bypass graft in the study. Patients with preexisting neurological dysfunction were not included. All patients were pre-medicated with flurazepam (1 mg/kg given orally) on the evening before surgery. Sixty minutes before induction of anesthesia, a premedication using dehydrobenzperidol (0.07 mg/kg) and fentanyl (1.5 [micro sign]g/kg) were given by subcutaneous injection. Anesthesia was induced with fentanyl (10–25 [micro sign]g/kg given intravenously), etomidate (50 [micro sign]g/kg given intravenously), and pancuronium (100 [micro sign]g/kg given intravenously) for neuromuscular blockade. After endotracheal intubation, all patients were normoventilated using oxygen in air (FIO²: 0.3–0.5). Anesthesia was maintained with fentanyl (<30 [micro sign]g/kg intravenously) and midazolam (150 [micro sign]g/kg intravenously) discontinuously as clinically required and with continuous infusion of etomidate (7 [micro sign]g [middle dot] kg^-1[middle dot] min^-1intravenously). During CPB the etomidate infusion was continued. Catheters were placed into one radial artery and one internal jugular vein for simultaneous measurements of mean arterial blood pressure, drug infusion, and blood sampling.

The jugular bulb venous oxygen saturation (SjO²) percentage and jugular bulb temperature (JBT) were measured using a fiberoptic thermodilution catheter (Opticath[R] F 5.5; Abbott Critical Care Systems, Mountain View, CA) placed into the right jugular bulb via the right internal jugular vein. Appropriate catheter position was confirmed by x-ray control before the measurements. The catheter was calibrated in vitro before insertion. Immediately after insertion and during the observation period, accurate fiberoptic saturation values were verified by drawing blood samples from the catheter and by measuring the oxygen saturation using a co-oximeter (OSM 3, Radiometer, Copenhagen, Denmark) every 5 min.

Cardiopulmonary bypass was performed using a membrane oxygenator (Monolyth[R] Sorin, Biomedica, Saluggia, Italy) and a roller pump (Stockert Instrumente, Munich, Germany) with an arterial line filter (40 [micro sign]m). Non-pulsatile perfusion (Q') through an ascending aortic cannula of 2–3 1 [middle dot] min^-1[middle dot] m^-2was maintained during the study period. The pump was primed with crystalloid solution, and a hematocrit of 0.22 +/- 0.02 was maintained during extracorporal circulation. Cardiopulmonary bypass was managed according to alpha-stat acid-base conditions under moderate hypothermia (27 +/- 1 [degree sign]C). During CPB arterial perfusate temperature (APT) was measured continuously at the oxygenator outlet.

After baseline measurements at the end of hypothermia (27 [degree sign]C) were made, hemoglobin concentration, the partial tension of oxygen (Pa^O(2)), mean arterial blood pressure, pump flow rate (Q'), JBT, ABT, and SjO²were monitored continuously and recorded every 5 min during rewarming of CPB. In the normocapnic group (n = 10), Pa^CO(2) was maintained within a range of 36–40 mmHg during rewarming. In the hypercapnic group (n = 10), Pa^CO(2) was increased with start of rewarming and was maintained within a range of 45–50 mmHg. The Pa^CO(2) in both groups was maintained by adjusting gas flow to the membrane oxygenator and measured at 37 [degree sign]C uncorrected to the patient's temperature (alpha-stat management). At the beginning of rewarming, the regulator of the heater was increased to 40–42 [degree sign]C and the APT was measured continuously at any time point.

Normally Pa^CO(2) was increased in the hypercapnic group by reducing the gas flow of the CPB to near 600 ml/min (i.e., about 60% of the gas flow before the start of rewarming) for 5–10 min until the Pa^CO(2) was increased to >45 mmHg, then for 5 min to 900 ml/min, and after 10 min to 1,200 ml/min to keep the Pa^CO(2) between 45–50 mmHg. To prevent hypoxemia, the FIO²was increased to 70% during hypercapnic rewarming. Patients were rewarmed until their nasopharyngeal temperature was >36 [degree sign]C and bladder temperature was >35 [degree sign]C. Norepinephrine was used to keep mean arterial blood pressure >40 mmHg during hypothermia and >55 mmHg during normothermia.

Physiologic variables are given as means +/- SD. Physiological variables at given time intervals were tested for Gaussian distribution using the Kolmogorow-Smirnow test. Analysis of variance for repeated measurements was used to test for significant differences between and within groups. Post hoc data were analyzed using paired or unpaired t tests when appropriate, with Bonferroni corrections for multiple comparisons. A probability value <0.05 was considered to indicate a significant difference.

(Table 1) shows the demographic variables and CPB times. No significant differences occurred in any parameter between the groups. Table 2shows the physiologic variables in patients with normocapnia (group 1) or hypercapnia (group 2) during hypothermia and rewarming. The hemoglobin value and mean arterial blood pressure did not change during rewarming compared with baseline (i.e., at the end of hypothermic CPB) and were not different between the groups. The amount of vasoactive drugs used was not different between the groups. The JBT and APT were not significantly different between the groups. The Pa^CO(2) was kept constant in group 1 but was increased in group 2 according to the study protocol. The Pa^O(2) was not significantly different between groups. Q' (pump flow rate) was increased in both groups to a similar extent.

(Figure 1) shows SjO²at baseline, nadir, and at the end of rewarming. In group 2, SjO²did not change during rewarming from hypothermia, and no patient showed a SjO²<50%. However, in group 1 SjO (²) was reduced to a nadir of 45 +/- 4%. Fifteen minutes after the start of rewarming, a maximum reduction of SjO²was seen in group 1, when JBT and APT were 36.5 +/- 1.1 [degree sign]C and 37.5 +/- 0.5 [degree sign]C (Table 2).

Our data show that the degree of jugular bulb venous hemoglobin saturation during rewarming from CPB is related to the level of Pa^CO(2). During normocapnia jugular bulb venous hemoglobin desaturation (SjO²<50%) occurred, with the lowest values seen at 36 [degree sign]C JBT. This suggests a mismatch between global cerebral oxygen demand and CBF during the rewarming period. In contrast, desaturation did not occur with hypercapnic acid base management. This suggests that hypercapnic cerebrovascular dilation is effective in reversing the imbalance of oxygen demand and supply during the rewarming period. However, an increased carbon dioxide level could increase the embolic load to the brain.

Previous studies have shown that rewarming from hypothermic CPB is frequently associated with decreases in SjO²(SjO²<50%). [6,14–17]These may be related to postoperative cognitive dysfunction. [6]The imbalance may be due to differences in the speed of rewarming within the various tissues and compartments. Studies in rabbits [13]have shown that rewarming from hypothermic CPB decreases cerebral venous hemoglobin saturation as a result of increases in the cerebral metabolic rate for oxygen that are temporarily greater than the increase in CBF. Therefore oxygen extraction was increased and SjO²was decreased. This suggests that the rewarming process activates cerebral function despite the fact that the brain temperature is still within the hypothermic range and that cerebral venous hemoglobin desaturation is related to inadequate cerebrovascular dilation as neuronal functional activity increases.

Cerebral venous oxygenation is a function of the position of the oxygen - hemoglobin dissociation curve and the venous oxygen pressure, and decreases in SjO²are not necessarily related to pathological conditions. According to a theoretical model of oxygen transfer from the blood to the brain, oxygen transfer depends closely on hemoglobin oxygen affinity. [17]Warm blood with decreased oxygen affinity (increased P⁵⁰= the oxygen partial pressure at which hemoglobin oxygen saturation equals 0.5) facilitates oxygen transfer from “warm” hemoglobin to the cold brain. This results in desaturation and is expected to occur at the beginning of rewarming, when the temperature gradient between the blood and the brain is most pronounced. In contrast, desaturation in normocapnic patients occurred during minimal temperature differences (JBT: 36.5 +/- 1.1 [degree sign]C vs. APT: 37.3 +/- 0.7 [degree sign]C) between the blood and the brain (i.e., 15 min after the start of rewarming, just before or when normothermia is reached;Table 2). However, the JBT probably represents only the area of the brain near the vessels with high perfusion, and the temperature of areas that are less perfused are lower, indicating a temperature gradient in these areas. Twenty-five minutes after the start of rewarming, the JBT is higher than the APT (Table 2). This difference may be related to the fact that the two different temperature measurement devices were not independently calibrated.

The P⁵⁰theory cannot explain the association between cerebral venous hemoglobin desaturation and the neuropsychological impairment found in human studies in which the arterial-jugular-venous oxygen difference and jugular venous hemoglobin desaturation at the end of rewarming was associated with postoperative neuropsychological impairment. [6]Similarly, this model cannot explain the reversal of jugular venous hemoglobin desaturation with hypercapnia during rewarming (group 2).

The fact that mild hypercapnia (45–50 mmHg) during alpha-stat management reversed jugular bulb venous hemoglobin oxygen desaturation seen with the classical alpha-stat approach (36–40 mmHg) tends to support the vascular response issue largely because the pH shift produced with mild hypercapnia is not great enough to facilitate unloading. Thus the difference in SjO²between the two groups is unlikely related to differences in the cerebral metabolic rate for oxygen, which was not measured in the present study.

In conclusion, these data show that under normocapnia the cerebral vascular reserve to vasodilate may not be exhausted during the rewarming period from hypothermic CPB. Further, these data show that mild hypercapnia during rewarming from hypothermic CPB reverses jugular bulb venous hemoglobin desaturation, possibly suggesting a better balance between oxygen supply and demand.