Patients suffering from chronic, thromboembolic pulmonary hypertension are severely limited during exercise, and sometimes even at rest, mainly because of a ventilation/perfusion mismatch and pulmonary hypertension leading to chronic right heart failure (RHF) (1). Therefore, pulmonary artery thromboendarterectomy (PTE) has been regarded as a promising, potentially curative surgical procedure since its introduction approximately 30 yr ago (2–4). However, PTE is associated with specific postoperative complications, such as reperfusion pulmonary edema (RPE) and RHF, leading to a considerable mortality of 7% to 24% (5–7). RPE is characterized by sustained arterial hypoxemia caused by focal pulmonary infiltrates in regions distal to vessels subjected to endarterectomy. RPE after PTE was first described by Utley et al. (8) and requires prolonged postoperative mechanical ventilation and intensive care treatment with its potential risks and side effects (9–10).
Although PTE has been performed for approximately 30 yr, little progress has been reported in improving the postoperative management. No systematic postoperative treatment has been developed and examined with respect to its value in preventing postoperative complications and prolonged postopera- tive mechanical ventilation. Positive inotropic catecholamines and vasodilators are often used for termination of cardiopulmonary bypass (CPB) and thereafter to prevent and/or treat RHF. Furthermore, high peak inspiratory pressures (peak Paw) and large tidal volumes (VT) are often required during the postoperative period to maintain adequate PaO2 (11). However, the use of positive inotropic catecholamines and vasodilators may lead to excessive cardiac output (CO). Excessive CO in combination with surgical endothelial injury with activation of the complement, coagulation, and kinin systems and the release of a variety of mediators may lead to altered vascular permeability (12,13) and pulmonary edema formation. Furthermore, it has been shown in an animal model that excessive lung perfusion increases lung water in canine permeability pulmonary edema (14). Thereafter, the use of high peak Paw and large VT to maintain PaO2 may lead to further lung injury (15).
Based on these pathophysiological considerations, we hypothesized that it might be beneficial to minimize CO and, thus, “overflow” of desobliterated pulmonary vessels. Furthermore, the use of low peak Paw and small VT during mechanical ventilation might minimize pulmonary barotrauma.
To test our hypothesis, we evaluated two treatment strategies with respect to postoperative complications, such as RHF and RPE, duration of mechanical ventilation, morbidity and mortality. In Group I, positive inotropic catecholamines and vasodilators were avoided, and mechanical ventilation was performed with VT < 8 mL/kg and peak Paw < 18 cm H2O in 33 consecutive patients undergoing PTE at the University of Vienna (Vienna, Austria). In Group II, catecholamines and vasodilators were regularly used, and mechanical ventilation was performed with VT between 10 and 15 mL/kg and peak Paw up to 50 cm H2O to maintain oxygenation in 14 patients undergoing PTE at the University of Maryland (Baltimore, MD). The radiographic development of RPE, duration of mechanical ventilation, duration of intensive care unit stay, and hemodynamics were recorded in all patients and were compared between groups.
Thirty-three consecutive patients with chronic thromboembolic pulmonary hypertension underwent PTE at the University of Vienna, and 14 consecutive patients underwent PTE at the University of Maryland. Thus, 47 patients were included in this study. Recently, data from the University of Maryland have been partially published (16). The study was approved by the local review boards of both universities, and all patients gave written, informed consent. All patients were selected for PTE according to international standards (17). Critically obese subjects, as defined by body mass index > 150%, were excluded. Preoperative patient characteristics of both groups are displayed in Table 1.
Standard monitoring consisted of invasive blood pressure, Swan-Ganz Right Ventricular Ejection Fraction Oxymetry TD Catheter (Baxter, Irvine, CA) and transesophageal echocardiography (TEE; Sonos 1500, Hewlett Packard, Endover, MA) in both centers.
The induction of anesthesia was standardized in all subjects with midazolam (0.1–0.2 mg/kg), fentanyl (3–4 μg/kg), etomidate (0.12–0.24 mg/kg), and pancuronium (0.15 mg/kg). Anesthesia was maintained with repetitive bolus doses of midazolam (0.08–0.12 mg/kg) and fentanyl (0.002–0.008 mg/kg). Additionally, patients with body weight > 50 kg received 500 mg methylprednisolone and patients < 50 kg body weight received 250 mg methylprednisolone before deep hypothermic circulatory arrest (DHCA) in both groups.
The surgical technique was as described by Jamieson et al. (17) for all patients undergoing PTE. After median sternotomy, preparation, aortic and bicaval cannulation, CPB was performed. Cooling was initiated simultaneously by instituting CPB. A vent was placed into the left atrium as well as into the pulmonary artery. Cardioplegic lines were placed in the aortic root and in the sinus coronarius for myocardial protection with cold blood cardioplegia. Afterwards, arteriotomy of the pulmonary artery and dissection to the peripheral segments was extended. For complete visualization of the peripheral segments, DHCA, at 18°C bladder temperature, was performed. After each period of DHCA, CPB was initiated until the mixed venous saturation reached 90%. The atrial septum was inspected to allow closure of a patent foramen ovale or atrial septum defect. Patients of each center were operated on by one experienced surgeon (WK in Vienna and AS in Baltimore).
Hemodynamic measurements, including heart rate, mean arterial pressure, mean pulmonary artery pressure (mPAP), pulmonary capillary wedge pressure, CO, pulmonary vascular resistance, and right ventricular ejection fraction were performed in both groups before skin incision and 20 min after termination of CPB. The measuring time point 20 min after termination of CPB was chosen because this is an especially critical time for hemodynamic instabilities usually treated with catecholamines and vasodilators. To avoid stress reactions, all patients were sedated for at least 24 h after surgery.
For termination of CPB, different strategies were used in the two groups (Table 2). Catecholamines with positive inotropic activity and vasodilators were avoided for termination of CPB and only administered in case of severe RHF in Group I (Table 2). Severe RHF was diagnosed if patients presented with a combination of an increase in pulmonary artery pressure above suprasystemic values, an increase in central venous pressure above 20 mm Hg, and echocardiographic hypokinesia of the right ventricle in both groups. Norepinephrine was only administered in cases with profound hypotension, defined as a mean arterial pressure < 50 mm Hg in Group I. In contrast, positive inotropic catecholamines and vasodilators were regularly used to prevent RHF in Group II during termination of CPB (Table 2).
Immediately after CPB, adequate volume status was assessed by means of central venous pressure and pulmonary capillary wedge pressure in both groups. Right heart function was monitored by means of TEE and pulmonary artery catheter (venous oxygen saturation) in all patients, to detect RHF as early as possible. RHF was treated similarly in both groups (Table 2).
For postoperative mechanical ventilation (Evita®; Dräger, Lübeck, Germany), airway pressure release ventilation-biphasic positive airway pressure, a pressure-controlled strategy with peak Paw as low as possible, duration of inspiration:duration of expiration = 3:1, and optimal positive end-expiratory pressure was used in Group I (Table 2). Optimal positive end-expiratory pressure was determined daily by using the approach described by Putensen et al. (18) in Group I. In contrast, synchronized intermittent mandatory ventilation-pressure support ventilation with characteristics as described in Table 2 was used in Group II.
Chest radiographs were taken at least once per day during the intensive care unit stay. RPE was diagnosed on the basis of chest radiographs taken during the first 3 days after surgery in all patients. This time span was chosen for systematic assessment of chest radiographs because RPE usually develops during the first 72 h after PTE (9).
Additionally, the incidence of severe complications such as RHF and RPE, the duration of mechanical ventilation, morbidity, and mortality were carefully recorded in both groups.
Statistical analysis was done with Student’s t-test for unpaired observations for intergroup comparison and Student’s t-test for paired observations for comparison within one group for different measuring time points. If values were not normally distributed, the median is given. Absolute changes in cardiac index (CI) and pulmonary vascular resistance index (PVRI) were compared between values measured before skin incision and values derived 20 min after termination of CPB. Results are expressed as mean ± SEM if not otherwise indicated. P < 0.05 was considered significant.
To guarantee uniformity of the data, absolute values of CI and PVRI were expressed as percentage of the value determined before skin incision (normalized CI and normalized PVRI, respectively). Furthermore, normalized CI > 150% 20 min after termination of CPB was defined as a significant increase in CI. A decrease in normalized PVRI < 60% 20 min after termination of CPB was defined as a substantial decrease in PVRI.
Values characterizing the surgical procedure such as mean CPB time and the mean DHCA are shown in Table 1. No significant difference between groups could be observed.
Positive inotropic catecholamines and vasodilators were avoided in all patients of Group I, except the three patients who developed severe RHF and one patient who developed multiple organ failure during the entire observation period. However, these drugs were regularly used in Group II (Table 2).
The mortality rate was 9.1% for Group I versus 21.4% for Group II (Table 3). RHF and RPE were less frequent in Group I compared with Group II (Table 3). RPE occurred rarely in Group I (6.1%) and could be dealt with successfully, whereas the incidence of RPE was higher in Group II (14.3%), and all patients in this group who developed RPE died. The duration of mechanical ventilation was shorter in Group I compared with Group II (Table 3). Mechanical ventilation was required for <5 days in 67% of patients in Group I, and two patients were ventilated for more than 10 days, with one extreme outliner ventilated for 42 days because of severe RHF. Mean mechanical ventilation time was 4.9 ± 1.3 days (median 3.0 days) in Group I, in contrast to 8.2 ± 3.4 (median 5.5 days) in Group II.
No significant change in CI could be observed in Group I after CPB, whereas CI increased significantly (P < 0.05) in Group II (Table 3). A significant (P < 0.05) decrease in mPAP and PVRI could be observed in both groups after CPB (Table 3). Nevertheless, no significant intergroup differences could be observed for CI, mPAP, and PVRI at all measuring time points (Table 3).
Normalized CI was > 150% in 4 of the patients in Group I after CPB (Figure 1). In two of these patients, mechanical ventilation of > 5 days was required (Figure 1). If normalized CI remained <150% (29 patients) after CPB, 9 of these patients required mechanical ventilation of >5 days in Group I. In Group II, normalized CI was >150% in two patients after CPB (Figure 1). In one of these patients, mechanical ventilation of >5 days was required (Figure 1). If normalized CI remained <150% (nine patients) in Group II, five of these patients required mechanical ventilation of >5 days. Thus, significantly increased CI after CPB did not result in prolonged mechanical ventilation in both groups. Furthermore, no significant correlation between duration of mechanical ventilation and increase in CI after CPB could be observed for Group I (r = −0.02;P = 0.92) or Group II (r = −0.48;P = 0,14).
Normalized PVRI was <60% in 16 patients in Group I after CPB. Normalized PVRI was >60% in 17 patients in Group I (Figure 2). In one of these patients with significantly decreased PVRI, mechanical ventilation of >5 days was required. In 7 of the 17 patients with postoperative PVRI, >60% of the preoperative value, prolonged mechanical ventilation >5 days was required in Group I (Figure 2). In Group II, normalized PVRI was <60% in six patients after CPB (Figure 2). In two of these patients, prolonged mechanical ventilation of >5 days was required (Figure 2). In five of the patients, PVRI was >60% of the preoperative value in Group II. In four of these patients, prolonged mechanical ventilation of >5 days was required in Group II.
Mechanical ventilation could be performed as defined in our study protocol and the methods section in Group I and Group II.
The purpose of this study was to test two different postoperative treatment strategies with respect to hemodynamic changes, specific complications such as RPE and RHF, and mortality after PTE. Treatment strategies varied substantially in the administration versus avoidance of positive inotropic catecholamines and vasodilators and in the mechanical ventilation applied (Table 2). If positive inotropic catecholamines and vasodilators were avoided and mechanical ventilation was performed with low peak Paw and low VT (Group I), the incidence of RPE and RHF, duration of mechanical ventilation, and mortality was lower. The change in CI after PTE was not associated with the development of RPE or with the duration of mechanical ventilation (Figure 1) in both groups. In contrast, the inability to lower PVRI by means of surgery resulted in prolonged duration of mechanical ventilation for subjects in both treatment groups (Figure 2). The use of positive inotropic catecholamines and vasodilators for prevention of RHF did not result in an advantage with respect to the incidence of RPE, RHF, and the mortality rate.
The 30-day mortality (9.1% for Group I and 21.4% for Group II) in both study populations was comparable with recent reports of other centers (Table 4). The mortality rate was substantially lower in Group I despite the avoidance of positive inotropic catecholamines and vasodilators for prevention of RHF compared with Group II in which these drugs were regularly administered. Thus, it can be concluded that, with respect to mortality as examination end-point, the treatment strategy of combining noninvasive mechanical ventilation with avoidance of positive inotropic catecholamines and vasodilators proved to be sufficient.
The incidence of RPE was lower in Group I (6.06% for Group I versus 14.3% for Group II). RPE is known to be a severe complication after PTE, and deaths caused by postoperative RPE are generally reported in the literature (Table 4). Hartz et al. (6) reported recently an incidence of 30% for RPE after PTE. Daily et al. (5), Levinson et al. (9), and Jamieson et al. (17) reported an incidence ranging from 30% to almost 100%, and similar data are available (Table 4).
One of our goals was to test the hypothesis that an increase in CO can lead to an overflow of the pulmonary circulation and, thus, contribute substantially to RPE after PTE. This hypothesis is based on an animal model showing that the combination of excessive lung perfusion with high inspiratory peak Paw worsens lung edema formation significantly (19). However, no significant correlation between the increase in CI and the duration of mechanical ventilation could be shown for both groups after PTE (Figure 1). This might be because of the fact that CI is a crude variable estimating overall pulmonary perfusion but not indicating exactly the amount of blood passing through the desobliterated pulmonary vessels. Thus, the theory of overflow of desobliterated pulmonary vessels leading to RPE could not be confirmed by our data. However, our data indicate that the use of catecholamines and vasodilators is not necessary for achieving a low mortality or morbidity rate after PTE, because the mortality and morbidity was lower in the group in which these drugs were avoided. Nevertheless, it remains possible that the use of positive inotropic catecholamines and vasodilators per se leads to an increased permeability of desobliterated pulmonary vessels.
RHF occurred in 9.1% of patients in Group 1 versus 21.4% of patients in Group II. This is especially striking, because the use of positive inotropic catecholamines and vasodilators (Table 2) could not reduce the incidence of RHF after PTE in Group II. These drugs are generally used in the presence of RHF to improve right ventricular function. Volume status was managed similarly in both study groups (Table 2). Therefore, the high incidence of RHF in Group II cannot be attributed to hypervolemia or hypovolemia. Absolute values of preoperative PVRI and postoperative PVRI were comparable in both groups (Table 3). Thus, the high incidence of RHF in Group II does not seem to be caused by differences in PVRI.
It is likely that PTE leads to surgical endothelial injury with activation of the complement, coagulation, and kinin systems and concomitant changes in endothelial permeability (11,20,21). Several authors compared RPE with the adult respiratory distress syndrome (ARDS), because it is similar in radiographic and clinical appearance (11). Since the results of Dreyfuss et al. (22,23), it became obvious that even healthy lung tissue can be injured merely by mechanical ventilation with large VT and/or high Paw. Low VT (<8 mL/kg) and inflation pressures <18 cm H2O were used in Group I. The incidence of RPE (6.06%) was very low in this group. In contrast, high VT between 10 and 15 mL/kg and inflation pressures <50 cm H2O were used in Group II. The incidence of RPE (14.3%) was considerably higher in Group II as was the duration of mechanical ventilation. Thus, taking into account the findings of several authors (22–25), it seems likely that RPE, which has been repeatedly compared with ARDS (11), might have been prevented or was sufficiently treated by using a specific ventilatory strategy usually applied in patients with ARDS.
In summary, the avoidance of positive inotropic catecholamines and vasodilators in combination with low VT and inspiratory Paw (Group I) was associated with a low incidence of RPE and RHF after PTE. Furthermore, mortality was lower with this postoperative treatment strategy. Nevertheless, the complexity of pathophysiological changes after PTE is overwhelming, and thus, postoperative treatment after PTE is performed on an empirical level.
1. Moser KM, Spragg RG, Utley J, Daily PO. Chronic thromboembolic obstruction of major pulmonary arteries. Ann Intern Med 1983; 99:299–305.
2. Hurwitt ES, Schein CJ, Rifkin H, Lebendiger A. A surgical approach to the problem of chronic pulmonary artery obstruction due to thrombosis or stenosis. Ann Surg 1958; 147:157–65.
3. Daily PO, Johnston GG, Simmons C, Moser KM. Surgical management of chronic pulmonary embolism. J Thorac Cardiovasc Surg 1980; 79:523–31.
4. Moser KM, Braunwald NS. Successful surgical intervention in severe chronic thromboembolic pulmonary hypertension. Chest 1973; 64:29–35.
5. Daily PO, Dembitsky WP, Iversen S, et al. Risk factors for pulmonary thromboendarterectomy. J Thorac Cardiovasc Surg 1990; 99:670–8.
6. Hartz RS, Byrne JG, Levitsky S, Park J, Rich S. Predictors of mortality in pulmonary thromboendarterectomy. Ann Thorac Surg 1996; 62:1255–60.
7. Mayer E, Dahm M, Hake U, et al. Mid-term results of pulmonary thromboendarterectomy for chronic thromboembolic pulmonary hypertension. Ann Thorac Surg 1996; 61:1788–92.
8. Utley JR, Spragg RG, Long WB, Moser KM. Pulmonary endarterectomy for chronic thromboembolic obstruction: recent surgical experience. Surgery 1982; 92:1096–102.
9. Levinson RM, Shure D, Moser KM. Reperfusion pulmonary edema after pulmonary artery thromboendarterectomy. Am Rev Respir Dis 1986; 134:1241–5.
10. Moser KM, Daily PO, Peterson KL, et al. Thromboendarterectomy for chronic, major vessel thromboembolic pulmonary hypertension: immediate and long-term results in 42 patients. Ann Intern Med 1987; 107:560–5.
11. Moser KM, Auger WR, Fedullo SW, Jamieson SW. Chronic thromboembolic pulmonary hypertension: clinical picture and surgical treatment. Eur Respir J 1992; 5:334–42.
12. Shure D. Thromboendarterectomy: some unanswered questions. Ann Thorac Surg 1996; 62:1253–4.
13. Barie PS, Hakim TS, Malik AB. Effect of pulmonary arteria occlusion and reperfusion on extravascular fluid accumulation. J Appl Physiol 1981; 50:102–6.
14. Hasinoff I, Ducas J, Prewitt RM. Increased cardiac output increases lung water in canine permeability pulmonary edema. J Crit Care 1988; 3:225–31.
15. Corbridge TC, Wood LDH, Crawford GP, et al. Adverse effects of large tidal volume and low PEEP in canine acid aspiration. Am Rev Respir Dis 1990; 142:311–5.
16. Gilbert TB, Gaine SP, Rubin LJ, Sequeira AJ. Short-term outcome and predictors of adverse events following pulmonary thromboendarterectomy. World J Surg 1998; 22:1029–33.
17. Jamieson SW, Auger WR, Fedullo PF, et al. Experience results with 150 pulmonary thromboendarterectomy operations over a 29-month period. J Thorac Cardiovasc Surg 1993; 106:116–27.
18. Putensen C, Baum M, Hormann C. Selecting ventilator settings according to variables derived from the quasi-static pressure/volume relationship in patients with acute lung injury. Anesth Analg 1993; 77:436–47.
19. Broccard AF, Hotchkiss JR, Kuwayama N, et al. Consequences of vascular flow on lung injury induced by mechanical ventilation. Am J Respir Crit Care Med 1998; 157:1935–42.
20. Morgan TE, Edmunds LH. Pulmonary artery occlusion. III. Biochemical alterations. J Appl Physiol 1967; 22:1012–6.
21. Bulkley GB. The role of oxygen free radicals in human disease processes. Surgery 1983; 94:407–11.
22. Dreyfuss D, Basset G, Soler P, Saumon G. Intermittent positive-pressure hyperventilation with high inflation pressures produces pulmonary microvascular injury in rats. Am Rev Respir Dis 1985; 132:880–4.
23. Dreyfuss D, Soler P, Basset G, Saumon G. High inflation pressure pulmonary edema. Am Rev Respir Dis 1988; 137:1159–64.
24. Dreyfuss D, Saumon G. Role of tidal volume, FRC, and end-expiratory volume in the development of pulmonary edema following mechanical ventilation. Am Rev Respir Dis 1993; 148:1194–03.
25. Gattinoni L, Pesenti A, Avalli L, et al. Pressure-volume curve of total respiratory system in acute respiratory failure: computed tomography scan study. Am Rev Respir Dis 1987; 136:730–6.