Mechanical ventilation strategy following Glenn and Fontan surgeries: On going challenge!

Abstract

The Glenn and Fontan operations put the pulmonary and systemic circulations in series. It has been shown that positive pressure ventilation (PPV) decreases pulmonary blood flow (PBF) and cardiac output (CO), and negative pressure ventilation (NPV) significantly improves PBF and CO. If early extubation is not achievable, the postoperative ventilator management strategy should aim at promoting PBF and CO by lowering pulmonary vascular resistance (PVR) and intrathoracic pressure. Multiple ventilator strategies have been evaluated to optimize this physiology, including high frequency ventilation, hyperventilation post Glenn, hypoventilation post Glenn with buffered pH, and the use of inhaled nitric oxide as an adjunct therapy for mechanical ventilation. In this review, the results of these studies will be reviewed and discussed.

1. Objectives

By the end of this review, reader(s) should be able to:

• 1.Describe cardiopulmonary interaction following cavopulmonary shunts.
• 2.Recognize effects of various mechanical ventilation strategies on hemodynamics post cavopulmonary shunts.
• 3.Analyze the evidences regarding mechanical ventilation effects on hemodynamics post cavopulmonary shunts.

2. Introduction

The Glenn and Fontan operations and their modifications are palliative operations for patients with a functionally single ventricle. The bi-directional Glenn operation consists of end-to-side anastomosis of the superior vena cava with the right pulmonary artery (Glenn, 1958). The Fontan operation baffles venous return from the lower half of the body to the pulmonary arteries (Fontan and Baudet, 1971). Both operations have a number of surgical modifications but all operations put the pulmonary and systemic circulations in series. Therefore, pulmonary blood flow (PBF) occurs passively, in the absence of a pulmonary ventricle. Cardiac output (CO) depends on passive PBF, and is particularly sensitive to changes in intrathoracic pressure and pulmonary vascular resistance (PVR). CO increases with lower intrathoracic pressure or lower PVR and decreases with higher intrathoracic pressure or higher PVR. While PVR may be increased by extremes of lung volume (Whittenberger et al., 1960; Hakim et al., 1982), acidosis (Malik and Kidd, 1973; Viles and Shepherd, 1968) and alveolar hypoxia (Fishman, 1976; Marshall and Marshall, 1980), it is increasingly apparent that thoracic pressure gradients have a profound impact on PBF in these circulations.

3. Cardio-pulmonary interactions

An early study by Williams et al. evaluated the hemodynamics of 13 postoperative Fontan patients (mean age 13.7 yr) in response to increasing levels of positive end-expiratory pressure (PEEP) starting at 0 then 3, 6, 9, 12 cm H₂O (Williams et al., 1984). Using dye dilution method, increasing PEEP resulted in a significant increase in PVR index and a significant decrease in cardiac index. It has been shown that positive pressure ventilation (PPV) decreases PBF and CO, and negative pressure ventilation (NPV) (spontaneous breathing or negative pressure ventilators) significantly improve PBF and CO (Redington et al., 1991; Penny and Redington, 1991; Kaulitz et al., 1999; Hsia et al., 2000; Shekerdemian et al., 1997a,b; Henneveld et al., 1999; Penny et al., 1991; Pierce et al., 1995). In a prospective study using Doppler echocardiographic assessment of PBF, Redington et al. (1991) and Penny and Redington (1991) in 3 (mean age 9.6 yr) and 16 (mean age 9.9 yr) patients post Fontan procedure, respectively, described an increase in pulmonary blood flow during spontaneous inspiration and a decrease in pulmonary blood flow during expiration or with a Valsalva maneuver.

Kaulitz et al. (1999) and Hsia et al. (2000) prospectively studied 15 and 48 Fontan patients respectively, and elucidated the contribution of hepatic venous flow by Doppler during the respiratory cycle. During spontaneous breathing, inspiratory flow in the hepatic vein increased 3–4 fold compared to expiration. In addition, Hsia documented the adverse effect of gravity on infradiaphragmatic venous flow. Inferior Vena Cava (IVC) flow increased in the supine compared to upright position in Fontan patients, emphasizing the sensitivity of passive PBF to pressure gradients.

These studies have led to the use of negative pressure ventilation, a mode of augmenting PBF, as a preventive and therapeutic modality for low cardiac output syndrome in the early post operative period.

4. Negative pressure ventilation

Following Fontan surgery, Penny et al. (1991) (n = 2) and Henneveld et al. (1999) (n = 1), described improvements in echocardiographic indices of PBF during NPV via chest cuirass when compared to PPV. Improvement of the clinical course with continuous NPV has also been reported as a rescue tool in patients with a Glenn shunt and respiratory failure (Pierce et al., 1995).

Shekerdemian et al. (1997a) prospectively demonstrated the advantage of NPV over PPV in healthy children during cardiac catheterization for trans-catheter occlusion of a patent ductus arteriosus (PDA) (n = 9) and in children after simple cardiac surgery (n = 7). Using the direct Fick method, CO was increased significantly in postoperative patients (28.1%) compared to the healthy children (10.8%). The same investigators Shekerdemian et al. (1997b) performed a similar study in 18 patients post Fontan procedure using the direct Fick method to measure PBF (nine in the early post operative period and nine anesthetized patients undergoing cardiac catheterization in the convalescent phase). Both groups underwent a period of PPV followed by a period of NPV. PBF increased by 42% during a brief period of negative extra-thoracic pressure ventilation and increased by 54% if the NPV was extended. These improvements were lost after reinstitution of PPV.

Collectively, these studies demonstrate the benefit of NPV following Fontan and Glenn surgery. However, the provision of NPV by the use of a chest cuirass is technically difficult in patients following open-heart surgery. These children commonly have median sternotomies, chest drains and chest wall edema making its use challenging.

The most practical and practiced method of NPV in postoperative cardiac care consists of early extubation to spontaneous breathing and has produced improved outcomes (Lofland, 2001; Shekerdemian and Bohn, 1999). Lofland (2001) described the hemodynamic effects of spontaneous negative pressure breathing in 50 consecutive Glenn and Fontan patients. Extubation occurred in the operating room or in the ICU within one-hour post operatively and aggressive pain control was instituted. He showed a significant decrease in pulmonary artery pressure and significant increase in CO post extubation when compared to pre-extubation, without change in complication or mortality rate.

From a physiologic perspective, early extubation to negative pressure spontaneous breathing is advantageous. Paradoxically, it is most beneficial in patients who are the least likely to tolerate extubation, i.e. unstable patients with low CO, or in the presence of profound hypoxia or lung disease. Early extubation may lead to atelectasis and hypoxia-related increases in PVR, impaired pain control and increases in metabolic demand. The sicker the patient, the less likely early extubation will occur as aggressive medical intervention is escalated. This may include escalation of PPV to achieve alkalosis in order to decrease PVR.

If early extubation is not achievable, the postoperative ventilator management strategy aims at promoting PBF and CO by lowering PVR and intrathoracic pressure. Multiple ventilator strategies have been evaluated to optimize this physiology, including high frequency ventilation, hyperventilation post Glenn, hypoventilation post Glenn with buffered pH, independent lung ventilation and the use of inhaled nitric oxide as a selective pulmonary vasodilator (Kornecki et al., 2002; Levine et al., 2002; Almodovar et al., 2000; Knez et al., 1999; Gamillscheg et al., 1997; Goldman et al., 1996; Meliones et al., 1991; Wheller et al., 1979; Burrows et al., 1986; Morray et al., 1988; Morris et al., 2000; Bradley et al., 2003, 1998; Hoskote et al., 2004).

5. High frequency ventilation

Meliones et al. (1991) evaluated the effect of high frequency jet ventilation in a study of 13 patients post Fontan procedure. He demonstrated significant decreases in mean airway pressure and PVR resulting in a significant increase in cardiac index. In contrast, Kornecki et al. (2002) applied high frequency oscillation to five Fontan patients with no demonstrable hemodynamic improvement. The principle difference between these studies was a decrease in mean airway pressure, compared to conventional ventilation, achieved in the Melione study but not in the Kornecki study. The lack of hemodynamic response in the absence of significant decrease in mean airway pressure may indicate that the hemodynamic effect is mediated mainly through changes in intra-thoracic pressure.

6. The effect of CO₂ (hyperventilation and hypoventilation)

Hyperventilation induced alkalosis relaxes the pulmonary vascular bed and may promote PBF (Meliones et al., 1991; Wheller et al., 1979; Burrows et al., 1986), It is unclear if this benefit is outweighed by the cost of ventilation-induced increases in intrathoracic pressure, which restrict PBF in the Fontan and Glenn circulations. On the other hand, hypocarbia associated with hyperventilation may result in decreased cerebral blood flow, which in turn reduces PBF (Morris et al., 2000; Bradley et al., 2003, 1998; Hoskote et al., 2004). Bradly et al. tried to answer this question post Glenn procedure in two prospective studies, involving 12 patients in the hyperventilation study and 15 patients in the hypoventilation with buffered PH study (Morris et al., 2000; Bradley et al., 1998). Oxygenation indexes, pulmonary artery pressure and cerebral blood flow were the main outcome measures. He described impaired oxygenation and cerebral blood flow with hyperventilation; but improved oxygenation, and better cerebral blood flow despite the increase in pulmonary artery pressure in hypoventilation and buffered PH study. More recently, Hoskote et al. in nine patients where CO₂ was added to the inspired gas of mechanical ventilation, demonstrated that after Glenn procedure, systemic oxygenation, PBF and cerebral blood flow increased while systemic vascular resistance index decreased at CO₂ tensions of 45 and 55 mm Hg compared with 35 mm Hg (Hoskote et al., 2004). This may indicate that the contribution of CBF to PBF, mediated through CO_2, may overweigh the negative impact of mild acidosis on PVR within the limits of above mentioned studies.

7. Independent synchronized lung ventilation

For unilateral lung disease, independent synchronized lung ventilation is a rarely used rescue strategy in single ventricle patients (Malik and Kidd, 1973; Levine et al., 2002). The rational for its use relies on minimizing mean airway pressure in the healthy lung and avoiding its adverse effect on PBF. On the other hand, it allows higher ventilator pressure to be used in ventilating the more abnormal lung on the other side. However, there is inadequate data and lack of experience among many cardiac intensivest to support this modality.

8. Inhaled nitric oxide

The adjunctive use of a selective pulmonary vasodilator such as inhaled nitric oxide (iNO) has been reported to be effective in augmenting PBF and improving outcomes, possibly by counteracting hypoxemia-induced pulmonary vasoconstriction and/or pulmonary endothelial dysfunction after cardiopulmonary bypass (Knez et al., 1999; Gamillscheg et al., 1997; Goldman et al., 1996). Morris et al. (2000) showed that both hyperventilation and iNO were effective in decreasing PAP and PVR children following cardiac surgery. However, it is difficult to separate the effect of iNO from other changes induced by hyperventilation, and if the changes in PAP or PVR will be translated to meaningful clinical outcome such oxygenation indexes. Gamillscheg et al. (1997) evaluated the effects of iNO in 13 patients with low pulmonary perfusion after Fontan and bidirectional Glenn procedures. iNO decreased central venous pressure and transpulmonary pressure gradients and increased mean systemic arterial and left atrial pressures. Arterial and venous oxygen saturations improved. These improvements were not compared to baseline state, but rather were made in comparison to the state after abrupt withdrawal of iNO, which may have exaggerated the benefit simply from the suppression of endogenous nitric oxide production by inhalation therapy. Goldman et al. (1996) showed that children with fenestrated Fontans only respond to iNO if they have baseline hypoxemia (SaO₂ < 85%) leading to improved oxygenation and a reduction in transpulmonary gradient (TPG).

Knez et al. (1999) described a single center experience regarding the influence of different strategies on clinical outcome in patients undergoing cavopulmonary connection and suggested a positive correlation between the use of iNO and improved survival rate. However, it is difficult to isolate the use of iNO from other operative or post-operative factors that may have contributed to the improved survival.

In summary, various strategies may improve PBF in mechanically ventilated patients with Fontan and Glenn circulation, including NPV, ventilator strategy that consists of adequate minute ventilation, low mean airway pressure, acceptance of mild increase in CO₂ with the option to buffer the acidosis. iNO could be beneficial in patients with poor saturation. There is lack of a studies that evaluates the use of alternate approaches to conventional PPV. Improvements in conventional strategies are important as they can be applied to the sickest patients who may derive the most benefit.