Retrograde Pulmonary Perfusion Improves Results in Pulmonary Embolectomy for Massive Pulmonary Embolism

Abstract

Mortality rates for pulmonary embolectomy in patients with acute massive pulmonary embolism have decreased in recent years. However, they still range from 30% to 45% when the surgery is performed on critically ill patients, and the rates reach 60% in patients who have experienced cardiac arrest before the procedure. The causes of death in these patients are generally attributed to right heart failure due to persistent pulmonary hypertension, intractable pulmonary edema, and massive parenchymal and intrabronchial hemorrhage. Clinical and experimental findings indicate that venous air embolism causes severe or even lethal damage to the pulmonary microvasculature and the lung parenchyma consequent to the release of endothelium-derived cytokines. These findings are similar to those observed when severely compromised patients undergo pulmonary embolectomy—air entrapped in the pulmonary artery during embolectomy can lead to fatal outcomes.

Besides enabling the removal of residual thrombotic material from the peripheral branches of the pulmonary artery, retrograde pulmonary perfusion fills the pulmonary artery with blood and prevents pulmonary air embolism. In this retrospective study, we analyzed a series of 21 consecutive critically ill patients in whom we applied retrograde pulmonary perfusion while performing standard pulmonary embolectomy. No patient died or experienced major postoperative complications. We believe that the use of retrograde pulmonary perfusion decreases morbidity and mortality rates associated with pulmonary embolectomy in critically ill patients.

The recently observed improvement in the results of surgical pulmonary embolectomy for massive pulmonary embolism has largely been credited to judicious selection of patients, rapid diagnosis, and early surgery performed on hemodynamically stable patients.^1–5 However, reported mortality rates range from 30% to 45% when embolectomy for massive pulmonary embolism is performed on critically ill patients, reaching 60% when those patients have experienced cardiac arrest before the procedure.1,3,6–10 The causes of death are usually attributed to right ventricular failure, persistent pulmonary hypertension, pulmonary edema, or massive parenchymal and intrabronchial hemorrhage.6,7,11–16

Experimental^17–19 and clinical^20–23 evidence indicates that pulmonary air embolism, through the release of endothelium-derived cytokines, causes severe damage to the pulmonary microvasculature and the lung parenchyma, consequently leading to serious and often lethal effects. These clinical and pathologic findings are quite similar to those observed in critically ill patients who undergo pulmonary embolectomy. Therefore, pulmonary air embolism can be considered a contributor to the negative outcomes in these patients.

Retrograde pulmonary perfusion (RPP) facilitates the removal of residual thrombotic material from the peripheral branches of the pulmonary artery. Moreover, by filling the arterial tree with blood, RPP eliminates entrapped air and the negative effects thereof.

In this preliminary report, we describe RPP as an adjunct to pulmonary embolectomy. We applied this technique while treating 21 consecutive critically ill patients, none of whom died or experienced major postoperative complications.

Patients

From January 1985 through June 2005, 21 consecutive critically ill patients with massive pulmonary embolism underwent pulmonary embolectomy supplemented by RPP. The group, which was studied retrospectively, comprised 8 men and 13 women whose ages ranged from 35 to 75 years.

Eight patients had deep vein thrombophlebitis. Seven patients had undergone major elective orthopedic surgery; 4, abdominal surgery; and 2, orthopedic surgery for a traumatic lesion. Two patients who experienced cardiac arrest in the ward were taken to the operating room under external cardiac massage. Each underwent emergency surgery on the sole basis of a presumptive clinical diagnosis of pulmonary embolism, which was later confirmed at the operating table. The other 19 patients exhibited acute, severe hemodynamic and respiratory compromise that required inotropic support. These patients presented with either a contraindication or a failure of response to the use of thrombolytic agents and were referred for emergent surgical pulmonary embolectomy. In these 19 cases, the diagnosis was established by transthoracic echocardiography (TTE) and pulmonary angiography. The TTE showed right ventricular dysfunction in all patients, paradoxical motion of the interventicular septum in 8 patients, and right atrial thrombi in 1 patient. In all instances, pulmonary angiograms revealed obstruction greater than 50% in the pulmonary artery. Three patients belonging to this group of 19 experienced cardiac arrest during induction of anesthesia, and surgery was initiated while resuscitative maneuvers were undertaken. A caval filter was implanted in 1 patient during surgery.

There were no in-hospital deaths. Postoperative complications involved 1 instance of cardiac tamponade, 2 of pericardial effusion that required drainage, and 3 of transient atrial fibrillation. All patients were discharged from the hospital on anticoagulant medication by the 10th postoperative day.

Surgical Technique

The RPP technique uses standard normothermic cardiopulmonary bypass with bicaval cannulation. The arterial line is connected to a Y connector. One branch of the connector is joined to the arterial cannula, which is inserted into the ascending aorta. The other branch of the connector is joined to a 20F clamped plastic cannula, which is inserted into the left atrium through a purse-string suture placed on the right upper pulmonary vein. After institution of cardiopulmonary bypass, cross-clamping of the aorta, and infusion of the cardioplegic solution, a longitudinal incision is made in the pulmonary artery trunk distal to the pulmonary valve and is extended into the proximal right and left pulmonary artery branches. The thrombotic material is extracted by means of forceps and suction. The right atrium and ventricle are then explored, and all visible clots are removed. Then, while the pulmonary artery is still open, the clamp on the left atrial cannula is released. Blood fills the left atrium; after approximately 1 min, the blood begins flowing into the pulmonary artery in a retrograde fashion. The lungs are repeatedly inflated in order to mobilize any residual fragment of thrombotic material that may be lodged in the distal branches of the pulmonary artery and to facilitate the elimination of residual air bubbles. The clots are aspirated, and all air is progressively eliminated from the pulmonary circulation. The pulmonary arteriotomy is then closed, and the left atrial cannula is disconnected from the arterial line and is used as a vent. The aorta is declamped, and the patient is weaned from cardiopulmonary bypass by standard method.

Discussion

Recently published reports state that results of open pulmonary embolectomy have improved, with mortality rates ranging from 8% to 27%.^1–3 However, these series consisted primarily of patients who were not in critical condition at the time of surgery.1,2,4,5 Mortality rates for open pulmonary embolectomy remain high—ranging from 30% to 45%—when surgery is performed on critically ill patients who have massive pulmonary embolism.1,6,8,9 It has been suggested that this high death rate is the consequence of reserving pulmonary embolectomy as a last-resort treatment for patients in severe shock and for those patients in whom less invasive forms of treatment have failed.^2,3,7

Mortality rates for pulmonary embolectomy in patients who experienced cardiac arrest before surgery remain around 60%.^6–8 Dauphine and Omari³ reported that 3 of their 4 patients who required resuscitation died. Caleb¹⁰ stated that none of his 5 patients who experienced preoperative hemodynamic collapse survived surgery.

The causes of death in patients who undergo pulmonary embolectomy have been attributed to right heart failure secondary to persistent pulmonary hypertension,6,7,12,13 intra-alveolar and interstitial pulmonary edema with normal left-sided pressures, and massive parenchymal and intrabronchial hemorrhage.11,14–16,24 The common pathologic finding is pulmonary hemorrhagic infarction.¹¹

Incomplete removal of thrombotic material lodged in the distal pulmonary arterial tree is considered an important cause of persistent pulmonary hypertension.6,7,12,13 The extraction of clots from the distal branches of the pulmonary artery is commonly performed through an extended pulmonary arteriotomy by suction and the use of standard or gallbladder-stone forceps, Fogarty catheters, or similar instruments, along with manual compression of the lungs as was advocated in the original report by Cooley and colleagues.²⁵ Mechanical injury to the pulmonary arterial wall by these means is thought to be responsible for the parenchymal and endobronchial bleeding.^4,15,16 The danger of injury to distal vessels has prompted recommendations to avoid blind instrumentation and to limit extraction to visible clots.⁴

Scant attention has been devoted to the role of air embolism in causing these adverse, often fatal effects during pulmonary embolectomy. In fact, abundant experimental^18,19 and clinical^20–23 evidence indicates that pulmonary air embolism releases endothelium-derived cytokines, damaging and occluding the microvasculature, with consequent pulmonary hypertension, pulmonary edema, and injury to the lung parenchyma. These findings are strikingly similar to those presented by critically ill patients who undergo pulmonary embolectomy. We believe that air entrapped in the pulmonary artery during pulmonary embolectomy is likely a major cause of the negative outcomes in these patients.

Distal air entrapment is enhanced by the structure of the peripheral pulmonary arterial branches, which are held open by the elastic parenchyma of the lung itself. Upon discontinuation of cardiopulmonary bypass and restoration of normal blood flow, the entrapped air is driven into the peripheral arterial branches, where it forms microbubbles that contribute to the obstruction of the circulation. Blood flow is obstructed further by the persistence of peripheral thrombi between the air bubbles and the alveolocapillary barrier. This thrombotic obstruction, in turn, impedes the air from reaching the alveoli, where it could be dissipated.

The severity of the obstructive syndrome depends upon the quantity of entrapped air, which, in turn, depends upon the ratio between the volume of air that has entered the pulmonary arterial system and the capability of the lungs to dissipate it through the alveoli. This renders more vulnerable those patients in whom there is a combination of negative factors: massive air embolism, peripheral migration of thrombotic material, right ventricular failure, and critical preoperative condition.

Other surgical procedures in which the pulmonary artery is opened widely, such as the correction of congenital cyanotic heart disease, do not generally present a threat of air embolism of the pulmonary artery because of the copious blood return from the bronchial circulation, which fills the pulmonary artery and expels the air.

Retrograde pulmonary perfusion performed as an adjunct to pulmonary embolectomy, as we have adopted in the present series, appears to confer 2 benefits: it helps to flush out residual thrombotic material lodged in the distal pulmonary arterial branches, and—more important, in our opinion—it prevents air embolism within the pulmonary artery, thus helping to prevent the associated detrimental effects.

Since the initial report in 1966 of the clinical use of RPP for pulmonary embolism, in which the authors described the use of retrograde injection of a fibrinolysin solution into the pulmonary veins in 3 patients with 1 long-term survivor,²⁶ RPP has been successfully used as an aid to treat acute pulmonary embolism in a few isolated cases.^27,28 In an additional instance, RPP was used to flush a cyanoacrylate glue obstruction from the distal pulmonary artery after its embolization from a cerebral arteriovenous malformation in a 3-year-old boy.²⁹

Our use of RPP during open pulmonary embolectomy led to no deaths or major postoperative complications in 21 consecutive gravely ill patients.

To our knowledge, this is the 1st report in which RPP was used as an adjunct to standard pulmonary embolectomy in a consecutive series of patients. The technique is simple, and it appears effective in reducing the morbidity and death that have accompanied pulmonary embolectomy. Should further clinical investigation confirm that RPP improves the outcomes of surgical pulmonary embolectomy, wider implementation of the procedure would be warranted.