Patent foramen ovale (PFO), usually a benign and silent lesion (Fig. 1), can cause hypoxemia and embolic phenomena under circumstances when right atrial (RA) pressure exceeds left atrial (LA) pressure or when preferential flow from the inferior vena cava (IVC) toward the PFO persists, as in the prenatal circulation (1–3). These circumstances may occur during the perioperative period as a result of the effect of mechanical ventilation, thromboembolism or air embolism, the altered anatomic relationship between the IVC and interatrial septum, or an increase in intraabdominal pressure. Although not extensively studied, numerous case reports implicate PFO as a causative factor for hypoxemia and systemic thromboembolism in the perioperative period. In the following review, we describe the epidemiology, pathogenesis, modern methods of detection, and general clinical importance of PFO. In particular, we highlight the significance of PFO in the perioperative period.
Development, Epidemiology, Pathogenesis, Detection, and Management Development
During prenatal life, PFO is a fenestration in the septum secundum, bordered on the left side by the septum primum, which allows blood flow from the RA to the LA. After birth, when LA pressure exceeds RA pressure, the PFO closes functionally with permanent closure caused by adhesion occurring within the first year of life (4). Although a large PFO is sometimes a functional atrial septal defect (ASD), this should not be confused with true ASD, particularly of secundum type (Table 1).
A PFO is a common finding in approximately 25% of postmortem examinations. For example, Hagen et al. (5) identified PFOs in 263 of 965 autopsies. PFOs were found to range from 1 to 19 mm, with the size increasing with age, whereas prevalence declined. There was no significant difference in the incidence and size of PFO between men and women. In another study, Thompson (6) found a “pencil patent” PFO (0.6–1 cm) in 6% of cases and a “probe patent” PFO (0.2–0.6 cm) in 29% of cases among 1000 random autopsies.
The presence of a PFO is highly associated with atrial septal aneurysm. This association was demonstrated in the multicenter study of Mugge et al. (7), in which 54% of patients with atrial septal aneurysm (protruding >10 mm beyond the plane of the intraatrial septum) showed intraatrial shunting, in the majority of cases through a PFO.
Three factors are important in determining the likelihood that a PFO will be of clinical significance: (a) its size, (b) the pressure gradient between the RA and LA, and (c) the direction of IVC flow. A large PFO in combination with a transient or persistent right-to-left pressure gradient or IVC flow directed through the PFO will allow unwelcome material, such as deoxygenated blood, thrombi, or air, to enter the systemic circulation. The size of a PFO probably does not change over time, with the exception of smaller lesions, which can undergo spontaneous closure (5). The RA pressure can increase as part of everyday life, e.g., coughing, retching, or hypoxic pulmonary vasoconstriction at high altitude (8), or it can result from specific clinical situations, as will be described later. In addition, the direction of the blood flow in the RA may change in situations such as right lung resection when, as a result of postoperative shift of the mediastinum, the angle between the IVC flow and the intraatrial septum may be altered and preferential flow through a PFO may develop (3).
Modern methods of PFO detection include transesophageal echocardiography (TEE), transthoracic echocardiography (TTE), and transcranial Doppler imaging. TEE is now considered to be the “gold standard” on the basis of comparison with postmortem studies and with other methods (Table 2) (9–11). The resolution of two-dimensional ultrasound technology is not yet high enough, however, to identify an actual defect except where a PFO is very large. Therefore, passage of contrast material or color flow through the defect should be demonstrated to establish the diagnosis. Under normal circumstances, LA pressure is higher than RA pressure; hence, a provocative procedure that temporarily reverses this relationship needs to be applied to demonstrate the passage of the contrast (usually agitated saline) through the interatrial septum. Provocative procedures include a Valsalva maneuver or coughing. Either maneuver results in a reduction of blood flow into the thorax because of an increase in intrathoracic pressure. Upon sudden release of the intrathoracic pressure, blood rapidly enters the thoracic cavity and fills the RA, creating a temporary transatrial pressure gradient (RA > LA), which remains until flow equilibrates. Agitated saline or blood injected IV at this time will “cross” the PFO from right to left, demonstrating its presence and confirming the diagnosis (Fig. 2). Contrast-based methods may lead to false-negative results when LA pressure far exceeds RA pressure and a pressure gradient cannot be reversed with any provocative maneuver. Assessment of flow through the PFO with color Doppler imaging (Fig. 3) avoids this limitation, but it assumes continuous left-to-right flow (unless a provocative maneuver is used). Continuous left-to-right flow is not seen in PFOs unless they are both anatomically and functionally open.
With respect to the TEE diagnosis of PFO, Schneider et al. (9) validated the use of TEE for PFO detection by comparison with autopsy finding in 35 consecutive patients. All PFOs were correctly identified by the color Doppler imaging, but only eight of nine were diagnosed by means of a contrast study. This corresponds to sensitivities of 100% and 89%, respectively. The specificity of both methods was 100%. In contrast, Konstadt et al. (12), by using TEE for PFO detection in 50 patients undergoing elective cardiac surgery, were able to identify PFO in 11 patients (22%). Color flow through the shunt was seen in only three patients. However, in two of these patients, contrast studies failed to identify a PFO. Luotolahti et al. (10) found similar results in patients who had experienced cryptogenic stroke. Contrast TEE identified a PFO in 21 of 28 patients, whereas color Doppler imaging was positive in only 17 of these 28 patients. The combination of both methods allowed the identification of a PFO in 24 of 28 patients with cryptogenic stroke. Di Tullio et al. (11) compared TTE and transcranial Doppler imaging with TEE for PFO detection and found that transcranial Doppler imaging was more sensitive than TTE (68% and 47%, respectively), but both methods were 100% specific.
The transcranial Doppler technique to detect PFO is based on the transient enhancement of an ultrasound signal reflected from air bubbles passing through cerebral arteries. It is a blind technique and, as with contrast echocardiography, it also relies on the creation of a temporary pressure gradient between the RA and LA with a provocative maneuver.
An indicator dilution technique and pulse oximetry have also been used in combination with a provocative maneuver to detect a transient left-to-right shunt (13). Karttunen et al. (14) have recently validated these methods in 103 cryptogenic stroke patients. They compared pulse oximetry and indicator dilution technique with contrast TEE and found that pulse oximetry had a sensitivity and specificity of 85% and 100%, whereas with indicator dilution technique these were 76% and 100%, respectively. In contrast, Sukernik et al. (15) evaluated pulse oximetry in 42 patients before coronary artery bypass grafting (CABG) and found a sensitivity of 27% and a specificity of 83%. Sensitivity for large PFO (more than 20 bubbles crossed) was slightly higher (40%). The difference in these results can probably be partly attributed to differences in the target populations, difference in pulse oximetry probe placement (ear versus finger), and the sophistication of the computer algorithm used by Karttunen et al. (14). More work needs to be done before this promising technique can be recommended as a screening method for PFO detection.
Ay et al. (16) reported that an M-shaped bifid notch, an electrocardiographic sign of ostium secundum ASD, can be seen in patients with PFO. The sensitivity and specificity of this method were 36% and 91%, respectively.
Asymptomatic PFO does not usually require any intervention. However, if hypoxemia develops or a patient experiences paradoxical embolism, treatment is usually indicated. This may include measures directed toward decreasing RA pressure (inotropes, nitric oxide, or thrombolysis) (17–19), prevention of thrombus formation (anticoagulation), or PFO closure. PFO closure may be accomplished at open heart surgery or as a percutaneous procedure (20) (Fig. 4).
Pathologic Consequences of a PFO in Patients with Normal RA Pressure
In certain situations, despite normal RA pressure, right-to-left shunt may develop. These situations may be caused by transient, rather than mean, interatrial right-to-left pressure gradients resulting from a Valsalva-like activity, such as coughing, or they may be caused by contraction of the RA against a stiff right ventricle. Further preferential flow from the IVC toward the fossa ovalis may be the mechanism responsible for right-to-left shunt in some cases (1). For example, Gallaher et al. (21) described three patients with a PFO, normal RA pressure, and a large eustachian valve and postulated that the eustachian valve was responsible for directing IVC flow toward the PFO, as is the case in prenatal life. Several other case reports describe hypoxemia occurring in association with a PFO in otherwise healthy subjects. Strunk et al. (1) described four patients with PFO who experienced decreased oxygen saturation and dyspnea. One patient had mild chronic obstructive pulmonary disease, and the others had no associated pulmonary or cardiac involvement. Contrast echocardiography revealed a right-to-left intraatrial shunt in all four subjects. Three of these four patients underwent surgical closure of the PFO, with a resultant increase in oxygen saturation and symptomatic improvement. Godart et al. (22) reported 11 cases of hypoxemia in the presence of normal RA pressure in patients with PFO or a small ASD. All of these patients were treated by percutaneous closure of the defect. There were no episodes of desaturation in the 30-day follow-up period in any of these patients.
PFO was found almost twice as often in young patients who experienced cryptogenic embolic stroke when compared with the general population (23,24). This observation suggests an underlying mechanism of paradoxical embolism. Lechat et al. (23) used TEE to study the prevalence of PFO in 60 patients under 55 yr old who had an ischemic stroke and a normal cardiac examination. They found a PFO in 40% of these patients and in only 10% of controls. In addition, Van Camp et al. (24) found a significant relationship between stroke and early and massive passage of contrast into the LA on TEE evaluation. Kozelj et al. (25) published a case report of a stroke in the postpartum period in a patient with a PFO and no other significant risk factors.
Migraine with an aura, transient amnesia, and retinal thrombosis are other manifestations of possible paradoxical embolism that have been described. Anzola et al. (26) compared the prevalence of PFO in 113 patients who had migraine with aura with 53 patients who had migraine without aura. The prevalence of PFO was 48% vs 23%, respectively. This finding suggests that paradoxical microembolization in the terminal vertebrobasilar territory might cause some cases of migraine with aura. Klotzsch et al. (27) evaluated 53 consecutive patients with transient global amnesia by use of transcranial Doppler imaging. They found a PFO in 55% of the patients and in 27% of 100 controls. Furthermore, a study of Steuber et al. (28) revealed a possible association between PFO and arterial occlusion of the retina and optic nerve.
Patients with PFO and a history of paradoxical embolism are at risk for recurrent embolic events (29,30). Major stroke recurrence in patients with a PFO may be as frequent as 1%–2% per year. Anticoagulation, antithrombotic therapy, and PFO closure have been suggested for secondary stroke prevention in these patients (31). Nendaz et al. (32), by using a decision-making model, suggested that if the stroke recurrence rate exceeds 0.8%, which is the case in paradoxical embolism, anticoagulation or closure of the defect is a preferred strategy in comparison with therapeutic abstinence. Although the clinical values of these therapeutic strategies are yet to be established, percutaneous PFO closure holds promise. Windecker et al. (20) followed 80 patients with a history of at least one episode of paradoxical embolism after percutaneous closure of PFO for 1.6 yr (sd, 1.4 yr). The annual incidence for cerebrovascular accident was 0%, for transient ischemic attack was 2.5%, and for peripheral emboli was 0.9%. Postprocedural shunt was a predictor of recurrent paradoxical embolism.
Divers represent another group in which a PFO can result in an increased likelihood of pathologic events. The current literature suggests that divers with a PFO are at increased risk of decompression sickness (33–35). Nitrogen bubbles that develop during decompression and are trapped in the pulmonary microvasculature may cause an increase in pulmonary artery pressure. These changes, in combination with a Valsalva maneuver, performed by divers to equalize their middle-ear pressure, can induce right-to-left intraatrial shunting and systemic gaseous embolism. Bove (35) determined the risk of decompression sickness on the basis of the analysis of three studies and concluded that PFO increases the odds ratio 2.5 times for this phenomenon.
PFO may also be important in the pathogenesis of high-altitude hypoxia (8,36). Levine et al. (8) performed TTE on a 42-yr-old woman who developed hypoxemia after climbing to 4200 m on Mount McKinley. They demonstrated significant pulmonary hypertension in this patient with continuous right-to-left flow through a PFO. Six months after the patient returned to sea level, a TTE was repeated and showed no pulmonary hypertension and no evidence of right-to-left shunting. They suggested that hypoxic pulmonary vasoconstriction in patients with PFO may trigger a vicious cycle and result in severe hypoxemia.
PFO in Conditions with Increased RA Pressure
Continuous right-to-left shunt across a PFO with subsequent hypoxemia has been described in cardiac tamponade (37), tricuspid endocarditis with significant tricuspid regurgitation (38), and adult respiratory distress syndrome (39), and it has been associated with right ventricular myocardial infarction (40). Konstantinides et al. (41) found that PFO was associated with adverse outcome in patients with major pulmonary embolism. Mortality, stroke rate, and peripheral embolism rates were substantially more frequent among those patients found to have a PFO with an overall 5.2 times increased risk of such complications. PFO presents in approximately two-thirds of patients with severe chronic pulmonary disease and obstructive sleep apnea and may contribute to hypoxemia and the progression of pulmonary hypertension secondary to intracardiac shunting (42,43). Kluger et al. (17) described a case of hypoxemia that developed after blunt chest injury in a previously healthy patient with PFO. A detailed workup revealed reactive pulmonary hypertension and right ventricular contusion. The patient was successfully treated with an amrinone infusion. Immediately after the administration of an initial dose of amrinone, the mean pulmonary arterial pressure decreased from 50 to 18 mm Hg, and arterial oxygen saturation improved from 85% to 93%, suggesting that right-to-left shunting of blood decreased. Boon et al. (44) described a case of severe asthma complicated by respiratory failure and circulatory arrest. The patient died despite intensive cardiopulmonary resuscitation. At postmortem examination a large PFO was found, suggesting that the cause of death was a severe right-to-left shunt from acute pulmonary hypertension. Herry et al. (45) published a case of severe respiratory failure and hypoxemia in a 73-yr-old woman with kyphoscoliosis and a PFO. She underwent percutaneous closure of her PFO with subsequent resolution of hypoxemia.
PFO in the Perioperative Period
No controlled study has been performed to address the possible pathologic significance of PFO in the perioperative period, although an abundance of case reports suggests an association between PFO and perioperative hypoxemia and embolism (19,46). Perioperative conditions with a potential to provoke right-to-left shunt through the PFO are summarized in Table 3.
Central Venous Catheterization
Venous air embolism resulting in paradoxical embolism during placement or withdrawal of internal jugular or subclavian central venous catheters has been described in several case reports (47–49). According to Heckmann et al. (48), although passage of the air through the pulmonary circulation is a possible route in many cases, PFO was implicated in 40% of reported cases. Hyperbaric oxygenation has been occasionally used to treat such conditions.
Right-to-left shunting of blood has been described during the emergence from general anesthesia when reaction to the endotracheal tube may predispose to a temporary positive pressure gradient between the RA and the LA. Moorthy et al. (50) published an article about a case of bilateral inguinal herniorrhaphy in a 2-mo-old patient. At the conclusion of surgery, despite the use of 100% oxygen, desaturation from 100% to 50% was noticed during periods of breath holding. Anesthesia was deepened, and the oxygen saturation improved. TTE with color flow Doppler imaging revealed a small left-to-right shunt. The anesthetic was discontinued, and the patient again reacted to the endotracheal tube by breath holding. During this second episode, oxygen saturation decreased again, and color flow Doppler imaging demonstrated reversal of the small left-to-right shunt to a marked right-to-left shunt. Controlled ventilation with 100% oxygen resulted in a return of oxygen saturation to 100%, and the patient was successfully extubated.
Cujec et al. (51) compared the effect of 10 cm of positive end-expiratory pressure (PEEP) on shunt fraction in 7 patients who had a PFO with 39 patients who did not have a PFO. They found that shunt fraction increased in the majority of patients with a PFO when PEEP was added, whereas it decreased in patients without a PFO. In a case report by Jaffe et al. (52), right-to-left shunt through a PFO, observed by TEE in the early inspiratory phase of mechanical ventilation, became continuous after the administration of 15 cm of PEEP. Although the mechanism of PEEP-induced right-to-left intracardiac shunt is not clear, these authors suggested that the PEEP-induced increase in pulmonary vascular resistance might be responsible for this effect. The previously mentioned studies suggest that hypoxemia should not be treated with PEEP in mechanically ventilated patients with PFO.
Posterior fossa surgery in the sitting position may subject patients with a PFO to enhanced risk from paradoxical air embolism. In these patients it is critical to know whether the patient has a PFO because the incidence of air embolism during posterior fossa surgery ranges from 25% to 100%, depending on the method of detection used (53,54). Mammoto et al. (54), by using TEE, found that paradoxical air embolism followed severe venous air embolism in 3 of 21 patients during neurosurgery in the sitting position. Four cases of paradoxical air embolism were described by Clayton et al. (55) during neurologic procedures in the sitting position, with symptoms related to possible occlusion of cerebral or coronary arteries. Papadopoulos et al. (56), by using preoperative TEE, identified a PFO in 9 of 63 patients scheduled for posterior fossa surgery or laminectomy in the sitting position. In this series, patients with proven PFO were excluded from having surgery in the sitting position. Nevertheless, 2 of 17 patients undergoing posterior fossa neurosurgery developed paradoxical air embolism. On the basis of the fact that venous air embolism was present in 76% of posterior fossa surgeries and in 25% of cervical laminectomies, the authors concluded that the preoperative search for a PFO should be performed for all planned surgeries in the sitting position, and if a PFO is detected, the sitting position should be avoided. They also stated that a residual risk of paradoxical air embolism persists despite preoperative exclusion by TEE.
Patients with a PFO undergoing a major orthopedic procedure are probably at increased risk for the development of perioperative paradoxical embolism compared with the general surgical population. This difference in risk may relate to the frequent incidence of deep venous thrombosis associated with this surgery and the possibility of fat embolism.
Weiss et al. (57) described a case of fatal paradoxical fat embolism after bilateral knee arthroplasty. Another fatal case of cerebral infarction resulting from paradoxical embolism was reported by Masini and Stulberg (58) in a patient recovering from the revision of a total hip arthroplasty. Della Valle et al. (59) described two cases in which pulmonary embolism accompanied cerebral infarction after orthopedic procedures. Both patients were diagnosed with PFO by TEE. These authors suggested that the increase of RA pressure as a result of acute increases in pulmonary vascular resistance might be an important predisposing factor for paradoxical embolism in patients with PFO because these may open a right-to-left shunt. They stated that the diagnosis of paradoxical embolism should be considered in the presence of venous thromboembolism, right-to-left shunt (PFO), and arterial embolism without evidence of left-sided source (59).
Several case reports have been published describing the development of oxygen desaturation after pneumonectomy or lobectomy in patients with PFO (46, 60–63). Hypoxemia in the upright position (orthodeoxia) is usually accompanied by shortness of breath when the patient is standing up (platypnea) (60). Right-to-left shunting through a PFO often develops even in the absence of a pressure gradient between the RA and LA, especially in the case of a right pneumonectomy secondary to the altered anatomic relationship, which results in preferential flow from the IVC through the PFO (3). PFO closure in these patients significantly improves oxygenation and symptoms of dyspnea (60–62). Although it has been suggested that platypnea and orthodeoxia are rare complications of thoracic surgery, the actual incidences of these complications are not known and have probably been underestimated because dyspnea and low oxygen saturation are often attributed to pulmonary embolism, heart failure, or residual lung disease rather than to PFO-related complications (3).
Rose et al. (64) reported three cases of the development or worsening of a right-to-left shunt in patients diagnosed with PFO after an otherwise uncomplicated CABG surgery. The postulated causes for this increase in shunting were the development of a right ventricular myocardial infarction, a small pulmonary embolus, transient right ventricular dysfunction, or a loculated pericardial effusion causing collapse of the right atrium. De Backer et al. (18) reported a case of hypoxemia caused by PFO opening in a 74-yr-old woman who developed right ventricular failure after surgical closure of an intraventricular communication complicating an anterior myocardial infarction. The patient’s oxygenation improved after the administration of nitric oxide. A case of pulmonary embolism that triggered right-to-left shunting through a PFO and severe arterial desaturation on the eighth postoperative day after CABG was described by Kollar et al. (19). The patient was intubated and treated with urokinase, with improvement of oxygenation within 12 h. The patient was subsequently extubated, started on coumadin, and discharged from the hospital.
Coronary revascularization without cardiopulmonary bypass (off-pump CABG) is becoming a more popular approach. Off-pump CABG can be especially challenging in patients with PFO because elevation of the heart may lead to an increase of RA pressure, with the development of hypoxemia. Akhter and Lajos (65) published a case report describing a patient with unstable angina and triple-vessel coronary artery disease in whom off-pump CABG was attempted. When the heart was elevated by placing a wet lap sponge under the diaphragmatic surface, oxygen saturation decreased to 80%, systolic arterial pressure decreased to 80 mm Hg, and the RA pressure increased from 8 to 16 mm Hg. Removal of the sponge reversed the hypoxemia and hypotension. TEE was performed and showed no defect. However, when the sponge was placed under the diaphragmatic surface again, RA pressure increased and a significant right-to-left shunt was demonstrated through the PFO by TEE. Cardiopulmonary bypass was initiated, revascularization was completed, and the PFO was closed. Yasu et al. (66) followed two patients with right-to-left shunt from a PFO who presented after mitral valve surgery. The shunt disappeared in one patient and significantly decreased in another after 1 and 2 yr of follow-up, respectively. No embolic complication was noted in these patients.
All patients undergoing left ventricular assist device placement should be screened for the presence of PFO. Any PFO should be closed in this setting because the underloaded left heart will create a permanent right-to-left shunt through the PFO, with subsequent hypoxemia (67).
Hypoxemia resulting from a previously silent PFO was described after heart transplantation (68). Because the transplanted heart usually faces relatively high pulmonary vascular resistance in the recipient, and this resistance can provoke right-to-left shunting, a meticulous search for a PFO needs to be performed and, if found, the PFO needs to be closed. Venous air embolism with subsequent paradoxical embolism has also been reported in liver transplantation, suggesting that the exclusion of PFO in these patients may also be important (69,70).
Iwase et al. (71) reported a case of right-to-left shunt through a PFO during pneumoperitoneum for laparoscopic cholecystectomy detected by TEE. The patient had a history of mitral valve replacement with residual tricuspid regurgitation. They suggested that an intraabdominal pressure of 12 mm Hg might have contributed to intraatrial shunting. Although the exact mechanism for this phenomenon is not clear, they suggested that an increase of intrathoracic pressure secondary to diaphragm elevation might be a contributing factor.
PFO Awareness in the Perioperative Period
Although the prevalence of PFO in patients undergoing surgery is very frequent, the presence of PFO is probably of little pathologic significance in most patients. However, the possibility of a PFO should always be kept in mind in the differential diagnosis of perioperative hypoxemia, especially when hypoxemia is out of proportion to pulmonary signs or radiographic findings and cannot easily be corrected by increasing inspired oxygen concentration or applying PEEP. Hypoxia worsening with PEEP administration may be a clue to the presence of PFO. Arterial embolism without an obvious left-sided source should raise the question of a possible PFO presence.
Preoperative Screening for PFO
On the basis of the current literature, we believe that preoperative screening for PFO is justified only in those situations in which its consequences may be devastating and a preventable strategy is feasible, e.g., posterior fossa surgery. Although TEE offers the best sensitivity, it is semiinvasive and an expensive tool. The procedure is uncomfortable if performed on awake patients. TTE and transcranial Doppler are noninvasive, less expensive, well established alternatives and may also be used for the screening purposes with the understanding that some, apparently smaller, PFOs could be missed. Pulse oximetry is an inexpensive and easy-to-perform test, which holds some promise as a screening tool but needs further evaluation.
In cardiac surgery when TEE is used, a search for a PFO should be performed. The fossa ovalis area of the intraarterial septum can be visualized best from the midesophageal level, usually in a longitudinal (90° angle) view. Color flow mapping of this area should be first performed without a Valsalva-like maneuver to search for a left-to-right-shunt. Then a sustained positive pressure of 20 cm H2O should be delivered to the lungs and abruptly released while color flow mapping is continued. Transient bulging of the intraatrial septum toward the LA after airway pressure release confirms the development of a right-to-left pressure gradient necessary to help identify a right-to-left shunt. A contrast study should complete the search for a right-to-left shunt. This is performed by the injection of 10 mL of agitated saline and positive airway pressure release as soon as the contrast material is visualized in the RA.
Preoperative Prevention of PFO-Related Complications
In patients with a history of paradoxical embolism associated with PFO, preoperative PFO closure or perioperative anticoagulation should be considered, especially when an operation with a frequent incidence of postoperative deep venous thrombosis is planned. Definitive recommendations for the exact method of secondary prevention of paradoxical embolism have not yet been developed, pending the results of clinical trials. Until these results are available, a decision on whether anticoagulation or PFO closure is the preferable method of treatment should be individualized. Because this decision has long-term implications, this choice should be made by a multidisciplinary team, including the patients’ cardiologist, neurologist, and primary care physician.
Perioperative Management of PFO-Related Complications
The reversal of a right-to-left shunt can be accomplished by the administration of positive inotropic drugs, nitric oxide, or both in the setting of right ventricular failure or pulmonary hypertension, the removal of pericardial fluid or thrombi in cardiac tamponade, thrombolysis in pulmonary embolism (if not contraindicated), or, if other methods fail, PFO closure. Particularly, PFO closure should be strongly considered when hypoxemia does not improve with time, as in platypnea-orthodeoxia syndrome after pneumonectomy. In cardiac surgery, we believe that PFO should be closed if the intraatrial septum is exposed for other reasons, such as for mitral or tricuspid valve surgery, or when the chances of PFO-related hypoxemia are very high, as in left ventricular assist device placement. In other cardiac operations, such as coronary revascularization or aortic valve operation, the decision to close a PFO needs to be individualized. During heart transplantation, the donor’s heart should be inspected for a PFO and, if found, the PFO should be closed before the heart is transplanted. In the event of paradoxical embolism of air, hyperbaric oxygenation therapy can be considered to decrease the size of the air bubbles and deliver a large oxygen concentration to ischemic tissue (72).
To develop definitive recommendations on the perioperative management of PFO-related complications, observational studies followed (if feasible) by clinical trials still need to be performed. These studies should include patients with preexistent right-to-left shunt or history of paradoxical embolism or those who undergo surgery in the sitting position, lung resection, or open-heart surgery.
The results of current continuing clinical trials on secondary prevention of paradoxical embolism (PFO and cryptogenic embolism trial, PFO in cryptogenic stroke study) should shed light on the questions of preferable methods of paradoxical embolism prophylaxis and PFO closure.
1. Strunk BL, Cheitlin MD, Stulbarg MS, Schiller NB. Right-to-left interatrial shunting through a patent foramen ovale despite normal intracardiac pressures. Am J Cardiol 1987; 60: 413–5.
2. Nazzal SB, Bansal RC, Fitzmorris SJ, Schmidt CA. Platypnea-orthodeoxia as a cause of unexplained hypoxemia in an 82-yr-old female. Cathet Cardiovasc Diagn 1990; 19: 242–5.
3. Smeenk FW, Postmus PE. Interatrial right-to-left shunting developing after pulmonary resection in the absence of elevated right-sided heart pressures: review of the literature. Chest 1993; 103: 528–31.
4. Patten B. The closure of the foramen ovale. Am J Anat 1931; 48: 19–44.
5. Hagen PT, Scholz DG, Edwards WD. Incidence and size of patent foramen ovale during the first 10 decades of life: an autopsy study of 965 normal hearts. Mayo Clin Proc 1984; 59: 17–20.
6. Thompson TEW. Paradoxical embolism. Q J Med 1930; 23: 135–50.
7. Mugge A, Daniel WG, Angermann C, et al. Atrial septal aneurysm in adult patients: a multicenter study using transthoracic and transesophageal echocardiography. Circulation 1995; 91: 2785–92.
8. Levine BD, Grayburn PA, Voyles WF, et al. Intracardiac shunting across a patent foramen ovale may exacerbate hypoxemia in high-altitude pulmonary edema. Ann Intern Med 1991; 114: 569–70.
9. Schneider B, Zienkiewicz T, Jansen V, et al. Diagnosis of patent foramen ovale by transesophageal echocardiography and correlation with autopsy findings. Am J Cardiol 1996; 77: 1202–9.
10. Luotolahti M, Saraste M, Hartiala J. Saline contrast and colour Doppler transoesophageal echocardiography in detecting a patent foramen ovale and right-to-left shunts in stroke patients. Clin Physiol 1995; 15: 265–73.
11. Di Tullio M, Sacco RL, Venketasubramanian N, et al. Comparison of diagnostic techniques for the detection of a patent foramen ovale in stroke patients. Stroke 1993; 24: 1020–4.
12. Konstadt SN, Louie EK, Black S, et al. Intraoperative detection of patent foramen ovale by transesophageal echocardiography. Anesthesiology 1991; 74: 212–6.
13. Karttunen V, Ventila M, Hillbom M, et al. Dye dilution and oximetry for detection of patent foramen ovale. Acta Neurol Scand 1998; 97: 231–6.
14. Karttunen V, Ventila M, Ikaheimo M, et al. Ear oximetry: a noninvasive method for detection of patent foramen ovale—a study comparing dye dilution method and oximetry with contrast transesophageal echocardiography. Stroke 2001; 32: 448–53.
15. Sukernik MR, Kachulis B, Mets B, Bennett-Guerrero E. Detection of patent foramen ovale through pulse oximetry with provocative maneuvers [abstract]. Anesthesiology 2000; 93: A575.
16. Ay H, Buonanno FS, Abraham SA, et al. An electrocardiographic criterion for diagnosis of patent foramen ovale associated with ischemic stroke. Stroke 1998; 29: 1393–7.
17. Kluger Y, Morgan JH, Tullius CR, et al. Hypoxemia after blunt chest injury: right-to-left intracardiac shunt managed non-operatively. Thorac Cardiovasc Surg 1993; 41: 121–3.
18. De Backer D, Moures JM, Vachiery JL, et al. Oxygenation improvement with nitric oxide in right-to-left shunt without significant effects on pulmonary arterial pressure. Chest 1996; 110: 1361–3.
19. Kollar A, Reames MK, Coyle JP. Patent foramen ovale and pulmonary embolism: an underestimated co-morbidity following cardiac surgery. J Cardiovasc Surg (Torino) 1998; 39: 355–8.
20. Windecker S, Wahl A, Chatterjee T, et al. Percutaneous closure of patent foramen ovale in patients with paradoxical embolism: long-term risk of recurrent thromboembolic events. Circulation 2000; 101: 893–8.
21. Gallaher MESD, Gwinn JL, Meyer BW, Fyler DC. Functional drainage of inferior vena cava into the left atrium: three cases. Am J Cardiol 1963; 11: 561–6.
22. Godart F, Rey C, Prat A, et al. Atrial right-to-left shunting causing severe hypoxaemia despite normal right-sided pressures: report of 11 consecutive cases corrected by percutaneous closure. Eur Heart J 2000; 21: 483–9.
23. Lechat P, Mas JL, Lascault G, et al. Prevalence of patent foramen ovale in patients with stroke. N Engl J Med 1988; 318: 1148–52.
24. Van Camp G, Schulze D, Cosyns B, Vandenbossche JL. Relation between patent foramen ovale and unexplained stroke. Am J Cardiol 1993; 71: 596–8.
25. Kozelj M, Novak-Antolic Z, Grad A, Peternel P. Patent foramen ovale as a potential cause of paradoxical embolism in the postpartum period. Eur J Obstet Gynecol Reprod Biol 1999; 84: 55–7.
26. Anzola GP, Magoni M, Guindani M, et al. Potential source of cerebral embolism in migraine with aura: a transcranial Doppler study. Neurology 1999; 52: 1622–5.
27. Klotzsch C, Sliwka U, Berlit P, Noth J. An increased frequency of patent foramen ovale in patients with transient global amnesia: analysis of 53 consecutive patients. Arch Neurol 1996; 53: 504–8.
28. Steuber C, Panzner B, Steuber T, et al. [Open foramen ovale in patients with arterial vascular occlusions of the retina and optic nerve]. Ophthalmologe 1997; 94: 871–6.
29. Bogousslavsky J, Garazi S, Jeanrenaud X, et al. Stroke recurrence in patients with patent foramen ovale: the Lausanne Study—Lausanne Stroke with Paradoxal Embolism Study Group. Neurology 1996; 46: 1301–5.
30. Mas JL, Zuber M. Recurrent cerebrovascular events in patients with patent foramen ovale, atrial septal aneurysm, or both and cryptogenic stroke or transient ischemic attack: French Study Group on Patent Foramen Ovale and Atrial Septal Aneurysm. Am Heart J 1995; 130: 1083–8.
31. Nendaz M, Sarasin FP, Bogousslavsky J. How to prevent stroke recurrence in patients with patent foramen ovale: anticoagulants, antiaggregants, foramen closure, or nothing? Eur Neurol 1997; 37: 199–204.
32. Nendaz MR, Sarasin FP, Junod AF, Bogousslavsky J. Preventing stroke recurrence in patients with patent foramen ovale: antithrombotic therapy, foramen closure, or therapeutic abstention? A decision analytic perspective. Am Heart J 1998; 135: 532–41.
33. Knauth M, Ries S, Pohimann S, et al. Cohort study of multiple brain lesions in sport divers: role of a patent foramen ovale. BMJ 1997; 314: 701–5.
34. Moon RE, Camporesi EM, Kisslo JA. Patent foramen ovale and decompression sickness in divers. Lancet 1989; 1: 513–4.
35. Bove AA. Risk of decompression sickness with patent foramen ovale. Undersea Hyperb Med 1998; 25: 175–8.
36. Pilmanis AA, Meissner FW, Olson RM. Left ventricular gas emboli in six cases of altitude-induced decompression sickness. Aviat Space Environ Med 1996; 67: 1092–6.
37. Thompson RC, Finck SJ, Leventhal JP, Safford RE. Right-to-left shunt across a patent foramen ovale caused by cardiac tamponade: diagnosis by transesophageal echocardiography. Mayo Clin Proc 1991; 66: 391–4.
38. Morelon P, Eicher JC, Chavanet P, et al. [Tricuspid endocarditis with right-left auricular shunt through a patent foramen ovale]. Ann Cardiol Angeiol (Paris) 1987; 36: 23–6.
39. Dewan NA, Gayasaddin M, Angelillo VA, et al. Persistent hypoxemia due to patent foramen ovale in a patient with adult respiratory distress syndrome. Chest 1986; 89: 611–3.
40. Fedullo AJ, Swinburne AJ, Mathew TM, et al. Hypoxemia from right to left shunting through patent foramen ovale. Am J Med Sci 1985; 289: 164–6.
41. Konstantinides S, Geibel A, Kasper W, et al. Patent foramen ovale is an important predictor of adverse outcome in patients with major pulmonary embolism. Circulation 1998; 97: 1946–51.
42. Soliman A, Shanoudy H, Liu J, et al. Increased prevalence of patent foramen ovale in patients with severe chronic obstructive pulmonary disease. J Am Soc Echocardiogr 1999; 12: 99–105.
43. Shanoudy H, Soliman A, Raggi P, et al. Prevalence of patent foramen ovale and its contribution to hypoxemia in patients with obstructive sleep apnea. Chest 1998; 113: 91–6.
44. Boon ES, Maesen BL, Valcke Y, Koudstaal J. Fatal asthma attributed to patent foramen ovale. Eur Respir J 1993; 6: 1567–8.
45. Herry I, Iung B, Piechaud JY, et al. Cardiac cause of hypoxaemia in a kyphoscoliotic patient. Eur Respir J 1999; 14: 1433–4.
46. Dlabal PW, Stutts BS, Jenkins DW, et al. Cyanosis following right pneumonectomy: importance of patent foramen ovale. Chest 1982; 81: 370–2.
47. Zuha R, Price T, Powles R, Treleaven J. Paradoxical emboli after central venous catheter removal. Ann Oncol 2000; 11: 885–6.
48. Heckmann JG, Lang CJ, Kindler K, et al. Neurologic manifestations of cerebral air embolism as a complication of central venous catheterization. Crit Care Med 2000; 28: 1621–5.
49. Schlotterbeck K, Tanzer H, Alber G, Muller P. [Cerebral air embolism after central venous catheter]. Anasthesiol Intensivmed Notfallmed Schmerzther 1997; 32: 458–62.
50. Moorthy SS, Haselby KA, Caldwell RL, et al. Transient right-left interatrial shunt during emergence from anesthesia: demonstration by color flow Doppler mapping. Anesth Analg 1989; 68: 820–2.
51. Cujec B, Polasek P, Mayers I, Johnson D. Positive end-expiratory pressure increases the right-to-left shunt in mechanically ventilated patients with patent foramen ovale. Ann Intern Med 1993; 119: 887–94.
52. Jaffe RA, Pinto FJ, Schnittger I, Brock-Utne JG. Intraoperative ventilator-induced right-to-left intracardiac shunt. Anesthesiology 1991; 75: 153–5.
53. Albin MSBM, Maroon JC, Jannetta PJ. Anaesthetic management of posterior fossa surgery in the sitting position. Acta Anaesthesiol Scand 1976; 20: 117–28.
54. Mammoto T, Hayashi Y, Ohnishi Y, Kuro M. Incidence of venous and paradoxical air embolism in neurosurgical patients in the sitting position: detection by transesophageal echocardiography. Acta Anaesthesiol Scand 1998; 42: 643–7.
55. Clayton DG, Evans P, Williams C, Thurlow AC. Paradoxical air embolism during neurosurgery. Anaesthesia 1985; 40: 981–9.
56. Papadopoulos G, Kuhly P, Brock M, et al. Venous and paradoxical air embolism in the sitting position: a prospective study with transoesophageal echocardiography. Acta Neurochir 1994; 126: 140–3.
57. Weiss SJ, Cheung AT, Stecker MM, et al. Fatal paradoxical cerebral embolization during bilateral knee arthroplasty. Anesthesiology 1996; 84: 721–3.
58. Masini MA, Stulberg SD. Paradoxical cerebral embolism leading to death after revision total hip arthroplasty. Orthopedics 1995; 18: 484–8.
59. Della Valle CJ, Jazrawi LM, Di Cesare PE, Steiger DJ. Paradoxical cerebral embolism complicating a major orthopaedic operation: a report of two cases. J Bone Joint Surg Am 1999; 81: 108–10.
60. Bakris NC, Siddiqi AJ, Fraser CD Jr, Mehta AC. Right-to-left interatrial shunt after pneumonectomy. Ann Thorac Surg 1997; 63: 198–201.
61. Godart F, Porte HL, Rey C, et al. Postpneumonectomy interatrial right-to-left shunt: successful percutaneous treatment. Ann Thorac Surg 1997; 64: 834–6.
62. Vincent L, Perol M, Bourlon F, et al. [Percutaneous closure of an right-left interatrial shunt acquired after pneumonectomy]. Rev Mal Respir 1998; 15: 103–5.
63. Vacek JL, Foster J, Quinton RR, Savage PJ. Right-to-left shunting after lobectomy through a patent foramen ovale. Ann Thorac Surg 1985; 39: 576–8.
64. Rose GC, Armstrong WF, Mahomed Y, Feigenbaum H. Atrial level right to left intracardiac shunt associated with postoperative hypoxemia: demonstration with contrast two-dimensional echocardiography. J Am Coll Cardiol 1985; 6: 920–2.
65. Akhter M, Lajos TZ. Pitfalls of undetected patent foramen ovale in off-pump cases. Ann Thorac Surg 1999; 67: 546–8.
66. Yasu T, Fujii M, Saito M. Long-term follow-up of an interatrial right-to-left shunt that appeared after cardiac surgery: evaluation by transesophageal Doppler echocardiography. Jpn Circ J 1996; 60: 70–4.
67. Baldwin RT, Duncan JM, Frazier OH, Wilansky S. Patent foramen ovale: a cause of hypoxemia in patients on left ventricular support. Ann Thorac Surg 1991; 52: 865–7.
68. Ouseph R, Stoddard MF, Lederer ED. Patent foramen ovale presenting as refractory hypoxemia after heart transplantation. J Am Soc Echocardiogr 1997; 10: 973–6.
69. Ellis JE, Lichtor JL, Feinstein SB, et al. Right heart dysfunction, pulmonary embolism, and paradoxical embolization during liver transplantation: a transesophageal two-dimensional echocardiographic study. Anesth Analg 1989; 68: 777–82.
70. Thiery G, Le Corre F, Kirstetter P, et al. Paradoxical air embolism during orthoptic liver transplantation: diagnosis by transoesophageal echocardiography. Eur J Anaesthesiol 1999; 16: 342–5.
71. Iwase K, Takao T, Watanabe H, et al. Right atrial to left atrial shunt through foramen ovale during pneumoperitoneum for laparoscopic cholecystectomy. Surg Endosc 1994; 8: 1110–2.
72. Murphy BP, Harford FJ, Cramer FS. Cerebral air embolism resulting from invasive medical procedures: treatment with hyperbaric oxygen. Ann Surg 1985; 201: 242–5.