- Email: [email protected]
Pulmonary embolectomy: A 25 year experience
J
THoRAc CARDIOVASC SURG
92:442-445, 1986
Pulmonary embolectomy: A 25 year experience For the past 25 years an emergency pulmonary embolectomy service has been offered to the hospitals serving a conurbation of 1.5 million. Fifty-five of these procedures have been performed during a short period of normothermic circulatory standstill produced by clamping the superior and inferior venae cavae. Of 36 patients who underwent pulmonary embolectomy without an episode of asystole or ventricular fibrillation, 35 survived the operation (97.2% ~ However, there were seven dea~ during the postoperative period, three related to pulmonary embolism and four to other causes (morta6ty 20%). Conversely, in a group of 19 patients who had an episode of cardiac arrest, 14 died during or after the operation of pulmonary embolism and two of unrelated causes (mortanty 73.7% ~ In properly selected patients this technique achieves a satisfactory measure of success. It can beused in hospitals that do not have cardiac surgical facilities and, because of its simp6city, it can be performed during the early period after pulmonary embolism when the risk of death is greatest.
D. B. Clarke, F.R.C.S., and L. D. Abrams, ER.C.S., Birmingham, Eng/and
Massive pulmonary embolism is a significant cause of death in the hospital, despite improvements in patient management developed in recent years. For the past 25 years an emergency pulmonary embolectomy service has been offered to the hospitals serving a population of approximately 1.5 million in the greater Birmingham area. In the majority of these operations normothermic venous inflow occlusion has been used, as described by Lewis, I rather than cardiopulmonary bypass. Our decision to use this technique was initially determined by the limited facilities available for cardiac operations, but more recently we have come to regard inflow occlusion circulatory arrest as a superior method. It is rapid and simple, and it can be employed in hospitals without equipment for cardiac surgery by relatively inexperienced surgeons with results equalling those reported using more elaborate techniques. Methods Pulmonary embolectomy has been offered to patients having a pulmonary embolus sufficiently massive to produce profound hemodynamic disturbance with a reduction of the systolic pressure to 100 mm Hg or less. Diagnosis has been mainly based upon history and From Queen Elizabeth Hospital, Birmingham, England. Received for publication Oct. 22, 1985. Accepted for publication Nov. 26, 1985. Address for reprints: D. B. Clarke, ER.C.S., Cardiothoracic Surgical Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham BI5 2TH, United Kingdom.
442
physical examination supplemented by the chest radiograph and electrocardiogram. Pulmonary angiography has been used to confirm the diagnosis when facilities were available and when it was considered that imminent death was unlikely. Inevitably, there have been errors in diagnosis as a result of this policy. These will be detailed later. The necessary instruments are carried in a sterile pack. These consist of a Gigli saw and a 12 inch forceps for passing the saw behind the sternum, two vascular clamps for occluding the venae cavae, and a Satinskytype clamp to be applied to the pulmonary artery. Operating room staff may be unfamiliar with cardiac operations and time spent in explaining the steps of the operation is never wasted. The anesthetist is also warned that cardiac arrest may follow the vasodilation that occurs on the induction of anesthesia and that inotropic drugs must be instantly available. Division of the sternum, opening of the pericardium, and application of the vascular clamps to the venae cavae are performed without delaying to secure hemostasis, and a 2 em incision is made in the pulmonary artery. Right-angled Desjardins common bile duct forceps are used to explore the main pulmonary artery and its branches and to remove emboli. A sucker is also introduced, which often succeeds in extracting further small emboli. We consider that these hypoxic patients should not be subjected to more than 3 minutes of circulatory arrest. After 2V2 minutes of circulatory standstill, air is flushed from the heart by releasing the superior caval clamp, the side clamp is applied to the
Volume 92 Number 3, Part 1 September, 1986
Pulmonary embolectomy
Table I. Systolic blood pressure before embolectomy and mortality
Table II. Underlying condition before pulmonary embolism No.
Pressure (mm Hg)
443
No. of patients
> 100 lOG-50
< 50
AsystolejVF Total
No. of deaths
1
25
1
10 19
3 16
55
20
Legend: VF, Ventricular fibrillation.
pulmonary artery, and the circulation is restored by releasing the inferior caval clamp. Total time from skin incisionto the completion of the embolectomy is usually no longer than 10 minutes. Early in our experience, we2 narrowed or clipped the inferior vena cava in the abdomen as a protection against further pulmonary embolism. This was usually attended by a profound fall in cardiac output and as the benefits of the procedure were not convincingly proven, we have now abandoned it. The sternotomy is closed with drainage and anticoagulant therapy is commenced on the first postoperative day. Results Sixty-fivepatients with a diagnosis of massive pulmonary embolism had exploration of the pulmonary artery. Three of these exploratory operations were performed with cardiopulmonary bypass, and these three patients survived. In seven the diagnosis was in error. All were in extremis with unrecordable blood pressures but with circumstantial evidence strongly suggestive of a diagnosis of pulmonary embolism. Five died after the operation. The correct diagnoses were subsequently found to be myocardial infarction in three, terminal pulmonary hypertension, prosthetic valve thrombosis, biliary peritonitis, and hemoperitoneum. Fifty-five patients had pulmonary emboli removed during inflow occlusion circulatory arrest, and these form the subject of this communication. Twenty-fivepatients were male and 30 female. Their ages ranged from 16 to 72 years. Systolic blood pressures at operation are shown in Table I and conditions for which the patient had been admitted to the hospital in Table II (most of these conditions carried a good prognosis). The time from embolism to operation ranged from 20 minutes to 24 hours (mean 5 hours). The overall mortality was 44.4%. The patients fell naturally into two groups: Group A, who had had an episode of ventricular fibrillation or
Pelvic operation Lower limb operation or injury Abdominal operation Hip replacement Intracranial operation Cardiac operation Myocardial infarct Femoral vein thrombosis Laminectomy Miscellaneous
9 9 10* 4 4 2 2 2
2 11
• Malignant disease in three.
asystole before the operation and Group B, who had no such episodes. Nineteen patients were in Group A; 10 died on the operating table, four died during the postoperative period of right ventricular failure, and two died of causes unrelated to pulmonary embolism. Only three left the hospital alive. The mortality in this group was 73.7%. Group B consisted of 36 patients who reached the operating room without an episode of asystole or ventricular fibrillation. Thirty-five of these survived pulmonary embolectomy (97.2%), but there were three deaths during the postoperative period attributable to pulmonary embolism (total mortality 11.2%). Four other deaths were due to the diseases for which the patient had originally been admitted to the hospital (preexisting renal failure, septicemia (two patients), and massive gastrointestinal bleeding). If this latter group is excluded, massive pulmonary embolism was successfully treated by embolectomy in 32 of 36 patients (88.8%). Two patients who required a second pulmonary embolectomy for recurrent embolism at 7 and 10 days survived. Guided by this experience, our selection of patients in recent years has enabled us to achieve a more consistent measure of success. There has been only one death, in a 72-year-old man who had a transitory episode of asystole, in the 13 patients submitted to pulmonary embolectomy during the past 5 years. Postoperative complications included renal failure (two patients), atelectasis necessitating bronchoscopy (one patient), and evidence of cerebral injury in two patients who had a period of asystole before the operation. All recovered. Most patients were discharged from the hospital by the tenth postoperative day. There have been no late sequelae other than edema of the legs in three patients who underwent interruption of the inferior vena cava. Serial postoperative pulmonary isotope scans performed in four patients have demonstrated
The Journal of
44 4 Clarke and Abrams
progressive resolution of areas of impaired perfusion resulting from small residual pulmonary emboli. In one patient resolution failed to occur and the pulmonary artery was explored later at thoracotomy, but successful perfusion of the affected lung was not achieved. Twenty-eight of these procedures have been performed in the cardiac surgical unit with 13 deaths (46.4%), and 27 have been performed in 10 other hospitals with 11 deaths (40.7%). The mean interval between embolism and operation at our base hospital was 3.5 hours (range 15 minutes to 8 hours). In hospitals without cardiac surgery facilities the mean interval was 9.3 hours (range 1 to 24 hours). Embolectomy was performed within 2 hours of the onset of symptoms on 19 occasions. Discussion There is a certain irony in the reflection that it was the death of a woman from pulmonary embolism which spurred Gibbon to develop a life support system that eventually led to the birth of cardiopulmonary bypass surgery, yet there have been so few reported series of pulmonary embolectomies using the techniques he originated. Pulmonary embolism is a common emergency, yet nearly 80 years after Trendelenburg's first pioneering operation':" there is no clear concensus on the role of surgical therapy. To some extent the issue has been clouded by the introduction of thrombolytic enzymes, but experience with these agents has not been uniformly successful in the management of the patient with a life-threatening embolus.' Although spontaneous improvement may be observed after pulmonary embolism sufficiently massive to cause profound hypotension, the observation that two thirds of such patients die within 2 hours of the onset of symptoms'- 7 must surely dictate an active rather than an expectant approach to management. In our view, emergency pulmonary embolectomy should be offered to such patients. Between 60% and 70% of the pulmonary vascular bed is obstructed and survival cannot be anticipated. Sautter" commented, however, "The identification of an unequivocal set of circumstances or findings which indicates those patients who will die without embolectomy remains to be made." The first pulmonary embolectomy performed with cardiopulmonary bypass was reported by Sharp? in 1962. The few large series published since then have demonstrated that the highest success rates have been achieved using this method, supplemented by supportive femoro-femoral bypass when appropriate.v'"!' These results have been particularly impressive when the
Thoracic and Cardiovascular Surgery
patient had had an episode of asystole or had required external cardiac compression. In the series of 40 pulmonary embolectomies reported by Mattox and associates," a 50% survival rate was achieved in a group of patients, most of whom fell into this category.. In a review of 108 pulmonary embolectomies performed between 1962 and 1979 by Kieny and colleagues," 57 procedures were performed on cardiopulmonary bypass with 50 survivors; 18 of these patients had had episodes of asystole or had required cardiac massage. However, these procedures were performed on patients fortunate enough to be in hospitals where advanced technological facilities were readily and immediately available when the massive embolus occurred. Nevertheless, in Kieny's series only 18 of 57 operations were performed within 6 hours of the onset of symptoms. In the Brompton Hospital series of embolectomies on cardiopulmonary bypass, a 50% survival rate was achieved in patients having an episode of cardiac arrest, with an overall survival rate of 64% in the series.' These excellent results were achieved during operations performed a mean of 11.9 hours after the onset of symptoms. In the Birmingham series, the mean time to operation was 5.2 hours (3.5 hours in our base hospital). The technique of transluminal extraction of the embolus using suction applied through a modified cardiac catheter is reported to be successful, but the method has not yet been used widely." We have been concerned particularly with the care of patients who have pulmonary embolism in hospitals without facilities for open cardiac operations. Guided by the conviction that speed is essential if lives are to be saved during the first 2 critical and dangerous hours after embolism, we have elected to take the operation to the patient rather than accept the delay occasioned by transferring the patient to the cardiac surgery unit. In our view, normothermic venous inflow occlusion is ideal for such a peripatetic pulmonary embolectomy service. Mistakenly referred to as the Trendelenberg operation, which it resembles only in that extracorporeal circulation is not needed, the procedure was first described by Ivor Lewis' and later by Vossschulte, Stiller, and Eisenreich." Since the introduction of cardiopulmonary bypass there have been no reports of the technique being used in the United States although sporadic papers describing its use are to be found in the European literature. In Kieny's series, 51 such operations are described
Volume 92 Number 3, Part 1 September, 1986
with 11 survivors. These rather disappointing figures reflect the general experience that has led to the abandonment of pulmonary embolectomy without the use of extracorporeal circulation. However, of nine patients treated in recent years, there were six survivors. Our experience has been similar in this regard and has been reported in earlier papers. 16• 17 Confidence in the operative techniques, improved diagnosis, and appropriate patient selection have enabled us to perform successfulembolectomies in 12 of the 13 patients treated during the past 5 years. We acknowledge that we are open to criticism for the small numbers of patients who underwent pulmonary angiography. Certainly, the diagnostic errors referred to would have been avoided if this investigation had been performed. However, in most cases facilities for immediate angiography were not available or the delay was believed to be unacceptable. Sixteen of Kieny's patients did not have pulmonary angiography before operation, and there were two episodes of cardiac arrest and six instances of circulatory deterioration in those submitted to this investigation. Twenty-five years' experience has enabled us to crystallize our views on the management of massive pulmonary embolism. Profound circulatory disturbances caused by pulmonary embolism are attended by poor prognosis,and death can occur at any time. A watching policy and the uncertain effects of thrombolytic enzyme therapy are not acceptable in such patients. In an ideal situation, and particularly when the patient has had an episode of cardiac arrest, pulmonary embolectomy performed with extracorporeal circulation provides the best conditions for survival, but these advantages are negated if removal of the embolus is unduly delayed by the rather complex preparations required. Pulmonary embolectomy by normothermic venous inflow occlusion can be performed in any operating room with the assistance of staff untrained in cardiac surgery. The operation can be mastered by any surgeon with a basic understanding of cardiac surgical techniques. The interval between diagnosis and operation is minimal. This operation offers an excellent chance of survival to any patient having a massive pulmonary embolus in any hospital provided that an episode of cardiac arrest has not occurred, with results which compare favorably with those achieved with cardiopulmonary bypass.
Pulmonary embolectomy
2 3 4
5
6 7 8 9
10
II
12
13
14
15
16 17
445
REFERENCES Lewis I: Problems in diagnosis and management of pulmonary embolism, Modern Trends in Cardiac Surgery, HRS Harley, ed., London, 1960, Butterworth & Co., Ltd., P 64 Clarke DB: Pulmonary embolectomy using normothermic venous inflow occlusion. Thorax 23: 131-135, 1968 Trendelenberg F: Ueber die operative behandlung der embolie der lungenarterie. Arch Klin Chir 86:686, 1908 Sabiston D: Trendelenberg's classic work on the operative treatment of pulmonary embolism. Ann Thorac Surg 35:570-574, 1983 Miller GAH, Hall RJC, Paneth M: Pulmonary embolectomy, heparin and streptokinase. Their place in the treatment of acute massive pulmonary embolism. Am Heart J 93:568-574, 1977 Gorham LW: A study of pulmonary embolism. Arch Intern Med 108:8-22, 1961 Soloff LA, Rodman T: Acute pulmonary embolism. II. Clinical. Am Heart J 74:829-847, 1967 Sautter RD: Treatment of massive pulmonary embolism (editorial). Chest 71:127-128, 1977 Sharp EH: Pulmonary embolectomy. Successful removal of a massive pulmonary embolus with the support of cardiopulmonary bypass. Case report. Ann Surg 156: 1-4, 1962 Cross FS, Mowlem A: A survey of the current status of pulmonary embolectomy for massive pulmonary embolus. Circulation 35:Suppl 1:86-91, 1967 Heimbacher RO, Keon WJ, Richards KV: Massive pulmonary embolism. A new look at surgical management. Arch Surg 107:740-746, 1973 Mattox KL, Feldtman RW, Beall AC, DeBakey ME: Pulmonary embolectomy for acute massive pulmonary embolism. Ann Surg 195:726-729, 1982 Kieny R, Eisemann JG, Kretz JG, Levy MT, Heitz A: Traitement chirurgical de l'embolie pulmonaire massive. A propos d'une sene de 62 embolectomies effectuees avec succes. Chirurgie 106:370-373, 1980 Greenfield LJ, Zocco JJ: Intraluminal management of acute massive pulmonary embolism. J THORAC CARDIoVASC SURG 77:402-410, 1979 Vossschuite K, Stiller H, Eisenreich F: Emergency embolectomy by the transstemal approach in acute pulmonary embolism. Surgery 58:317-323, 1965 Clarke DB, Abrams LD: Pulmonary embolectomy with venous inflow occlusion. Lancet 1:767-769, 1972 Clarke DB: Pulmonary embolectomy re-evaluated. Ann R Coli Surg Engl 63:18-24, 1981
THoRAc CARDIOVASC SURG
92:442-445, 1986
Pulmonary embolectomy: A 25 year experience For the past 25 years an emergency pulmonary embolectomy service has been offered to the hospitals serving a conurbation of 1.5 million. Fifty-five of these procedures have been performed during a short period of normothermic circulatory standstill produced by clamping the superior and inferior venae cavae. Of 36 patients who underwent pulmonary embolectomy without an episode of asystole or ventricular fibrillation, 35 survived the operation (97.2% ~ However, there were seven dea~ during the postoperative period, three related to pulmonary embolism and four to other causes (morta6ty 20%). Conversely, in a group of 19 patients who had an episode of cardiac arrest, 14 died during or after the operation of pulmonary embolism and two of unrelated causes (mortanty 73.7% ~ In properly selected patients this technique achieves a satisfactory measure of success. It can beused in hospitals that do not have cardiac surgical facilities and, because of its simp6city, it can be performed during the early period after pulmonary embolism when the risk of death is greatest.
D. B. Clarke, F.R.C.S., and L. D. Abrams, ER.C.S., Birmingham, Eng/and
Massive pulmonary embolism is a significant cause of death in the hospital, despite improvements in patient management developed in recent years. For the past 25 years an emergency pulmonary embolectomy service has been offered to the hospitals serving a population of approximately 1.5 million in the greater Birmingham area. In the majority of these operations normothermic venous inflow occlusion has been used, as described by Lewis, I rather than cardiopulmonary bypass. Our decision to use this technique was initially determined by the limited facilities available for cardiac operations, but more recently we have come to regard inflow occlusion circulatory arrest as a superior method. It is rapid and simple, and it can be employed in hospitals without equipment for cardiac surgery by relatively inexperienced surgeons with results equalling those reported using more elaborate techniques. Methods Pulmonary embolectomy has been offered to patients having a pulmonary embolus sufficiently massive to produce profound hemodynamic disturbance with a reduction of the systolic pressure to 100 mm Hg or less. Diagnosis has been mainly based upon history and From Queen Elizabeth Hospital, Birmingham, England. Received for publication Oct. 22, 1985. Accepted for publication Nov. 26, 1985. Address for reprints: D. B. Clarke, ER.C.S., Cardiothoracic Surgical Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham BI5 2TH, United Kingdom.
442
physical examination supplemented by the chest radiograph and electrocardiogram. Pulmonary angiography has been used to confirm the diagnosis when facilities were available and when it was considered that imminent death was unlikely. Inevitably, there have been errors in diagnosis as a result of this policy. These will be detailed later. The necessary instruments are carried in a sterile pack. These consist of a Gigli saw and a 12 inch forceps for passing the saw behind the sternum, two vascular clamps for occluding the venae cavae, and a Satinskytype clamp to be applied to the pulmonary artery. Operating room staff may be unfamiliar with cardiac operations and time spent in explaining the steps of the operation is never wasted. The anesthetist is also warned that cardiac arrest may follow the vasodilation that occurs on the induction of anesthesia and that inotropic drugs must be instantly available. Division of the sternum, opening of the pericardium, and application of the vascular clamps to the venae cavae are performed without delaying to secure hemostasis, and a 2 em incision is made in the pulmonary artery. Right-angled Desjardins common bile duct forceps are used to explore the main pulmonary artery and its branches and to remove emboli. A sucker is also introduced, which often succeeds in extracting further small emboli. We consider that these hypoxic patients should not be subjected to more than 3 minutes of circulatory arrest. After 2V2 minutes of circulatory standstill, air is flushed from the heart by releasing the superior caval clamp, the side clamp is applied to the
Volume 92 Number 3, Part 1 September, 1986
Pulmonary embolectomy
Table I. Systolic blood pressure before embolectomy and mortality
Table II. Underlying condition before pulmonary embolism No.
Pressure (mm Hg)
443
No. of patients
> 100 lOG-50
< 50
AsystolejVF Total
No. of deaths
1
25
1
10 19
3 16
55
20
Legend: VF, Ventricular fibrillation.
pulmonary artery, and the circulation is restored by releasing the inferior caval clamp. Total time from skin incisionto the completion of the embolectomy is usually no longer than 10 minutes. Early in our experience, we2 narrowed or clipped the inferior vena cava in the abdomen as a protection against further pulmonary embolism. This was usually attended by a profound fall in cardiac output and as the benefits of the procedure were not convincingly proven, we have now abandoned it. The sternotomy is closed with drainage and anticoagulant therapy is commenced on the first postoperative day. Results Sixty-fivepatients with a diagnosis of massive pulmonary embolism had exploration of the pulmonary artery. Three of these exploratory operations were performed with cardiopulmonary bypass, and these three patients survived. In seven the diagnosis was in error. All were in extremis with unrecordable blood pressures but with circumstantial evidence strongly suggestive of a diagnosis of pulmonary embolism. Five died after the operation. The correct diagnoses were subsequently found to be myocardial infarction in three, terminal pulmonary hypertension, prosthetic valve thrombosis, biliary peritonitis, and hemoperitoneum. Fifty-five patients had pulmonary emboli removed during inflow occlusion circulatory arrest, and these form the subject of this communication. Twenty-fivepatients were male and 30 female. Their ages ranged from 16 to 72 years. Systolic blood pressures at operation are shown in Table I and conditions for which the patient had been admitted to the hospital in Table II (most of these conditions carried a good prognosis). The time from embolism to operation ranged from 20 minutes to 24 hours (mean 5 hours). The overall mortality was 44.4%. The patients fell naturally into two groups: Group A, who had had an episode of ventricular fibrillation or
Pelvic operation Lower limb operation or injury Abdominal operation Hip replacement Intracranial operation Cardiac operation Myocardial infarct Femoral vein thrombosis Laminectomy Miscellaneous
9 9 10* 4 4 2 2 2
2 11
• Malignant disease in three.
asystole before the operation and Group B, who had no such episodes. Nineteen patients were in Group A; 10 died on the operating table, four died during the postoperative period of right ventricular failure, and two died of causes unrelated to pulmonary embolism. Only three left the hospital alive. The mortality in this group was 73.7%. Group B consisted of 36 patients who reached the operating room without an episode of asystole or ventricular fibrillation. Thirty-five of these survived pulmonary embolectomy (97.2%), but there were three deaths during the postoperative period attributable to pulmonary embolism (total mortality 11.2%). Four other deaths were due to the diseases for which the patient had originally been admitted to the hospital (preexisting renal failure, septicemia (two patients), and massive gastrointestinal bleeding). If this latter group is excluded, massive pulmonary embolism was successfully treated by embolectomy in 32 of 36 patients (88.8%). Two patients who required a second pulmonary embolectomy for recurrent embolism at 7 and 10 days survived. Guided by this experience, our selection of patients in recent years has enabled us to achieve a more consistent measure of success. There has been only one death, in a 72-year-old man who had a transitory episode of asystole, in the 13 patients submitted to pulmonary embolectomy during the past 5 years. Postoperative complications included renal failure (two patients), atelectasis necessitating bronchoscopy (one patient), and evidence of cerebral injury in two patients who had a period of asystole before the operation. All recovered. Most patients were discharged from the hospital by the tenth postoperative day. There have been no late sequelae other than edema of the legs in three patients who underwent interruption of the inferior vena cava. Serial postoperative pulmonary isotope scans performed in four patients have demonstrated
The Journal of
44 4 Clarke and Abrams
progressive resolution of areas of impaired perfusion resulting from small residual pulmonary emboli. In one patient resolution failed to occur and the pulmonary artery was explored later at thoracotomy, but successful perfusion of the affected lung was not achieved. Twenty-eight of these procedures have been performed in the cardiac surgical unit with 13 deaths (46.4%), and 27 have been performed in 10 other hospitals with 11 deaths (40.7%). The mean interval between embolism and operation at our base hospital was 3.5 hours (range 15 minutes to 8 hours). In hospitals without cardiac surgery facilities the mean interval was 9.3 hours (range 1 to 24 hours). Embolectomy was performed within 2 hours of the onset of symptoms on 19 occasions. Discussion There is a certain irony in the reflection that it was the death of a woman from pulmonary embolism which spurred Gibbon to develop a life support system that eventually led to the birth of cardiopulmonary bypass surgery, yet there have been so few reported series of pulmonary embolectomies using the techniques he originated. Pulmonary embolism is a common emergency, yet nearly 80 years after Trendelenburg's first pioneering operation':" there is no clear concensus on the role of surgical therapy. To some extent the issue has been clouded by the introduction of thrombolytic enzymes, but experience with these agents has not been uniformly successful in the management of the patient with a life-threatening embolus.' Although spontaneous improvement may be observed after pulmonary embolism sufficiently massive to cause profound hypotension, the observation that two thirds of such patients die within 2 hours of the onset of symptoms'- 7 must surely dictate an active rather than an expectant approach to management. In our view, emergency pulmonary embolectomy should be offered to such patients. Between 60% and 70% of the pulmonary vascular bed is obstructed and survival cannot be anticipated. Sautter" commented, however, "The identification of an unequivocal set of circumstances or findings which indicates those patients who will die without embolectomy remains to be made." The first pulmonary embolectomy performed with cardiopulmonary bypass was reported by Sharp? in 1962. The few large series published since then have demonstrated that the highest success rates have been achieved using this method, supplemented by supportive femoro-femoral bypass when appropriate.v'"!' These results have been particularly impressive when the
Thoracic and Cardiovascular Surgery
patient had had an episode of asystole or had required external cardiac compression. In the series of 40 pulmonary embolectomies reported by Mattox and associates," a 50% survival rate was achieved in a group of patients, most of whom fell into this category.. In a review of 108 pulmonary embolectomies performed between 1962 and 1979 by Kieny and colleagues," 57 procedures were performed on cardiopulmonary bypass with 50 survivors; 18 of these patients had had episodes of asystole or had required cardiac massage. However, these procedures were performed on patients fortunate enough to be in hospitals where advanced technological facilities were readily and immediately available when the massive embolus occurred. Nevertheless, in Kieny's series only 18 of 57 operations were performed within 6 hours of the onset of symptoms. In the Brompton Hospital series of embolectomies on cardiopulmonary bypass, a 50% survival rate was achieved in patients having an episode of cardiac arrest, with an overall survival rate of 64% in the series.' These excellent results were achieved during operations performed a mean of 11.9 hours after the onset of symptoms. In the Birmingham series, the mean time to operation was 5.2 hours (3.5 hours in our base hospital). The technique of transluminal extraction of the embolus using suction applied through a modified cardiac catheter is reported to be successful, but the method has not yet been used widely." We have been concerned particularly with the care of patients who have pulmonary embolism in hospitals without facilities for open cardiac operations. Guided by the conviction that speed is essential if lives are to be saved during the first 2 critical and dangerous hours after embolism, we have elected to take the operation to the patient rather than accept the delay occasioned by transferring the patient to the cardiac surgery unit. In our view, normothermic venous inflow occlusion is ideal for such a peripatetic pulmonary embolectomy service. Mistakenly referred to as the Trendelenberg operation, which it resembles only in that extracorporeal circulation is not needed, the procedure was first described by Ivor Lewis' and later by Vossschulte, Stiller, and Eisenreich." Since the introduction of cardiopulmonary bypass there have been no reports of the technique being used in the United States although sporadic papers describing its use are to be found in the European literature. In Kieny's series, 51 such operations are described
Volume 92 Number 3, Part 1 September, 1986
with 11 survivors. These rather disappointing figures reflect the general experience that has led to the abandonment of pulmonary embolectomy without the use of extracorporeal circulation. However, of nine patients treated in recent years, there were six survivors. Our experience has been similar in this regard and has been reported in earlier papers. 16• 17 Confidence in the operative techniques, improved diagnosis, and appropriate patient selection have enabled us to perform successfulembolectomies in 12 of the 13 patients treated during the past 5 years. We acknowledge that we are open to criticism for the small numbers of patients who underwent pulmonary angiography. Certainly, the diagnostic errors referred to would have been avoided if this investigation had been performed. However, in most cases facilities for immediate angiography were not available or the delay was believed to be unacceptable. Sixteen of Kieny's patients did not have pulmonary angiography before operation, and there were two episodes of cardiac arrest and six instances of circulatory deterioration in those submitted to this investigation. Twenty-five years' experience has enabled us to crystallize our views on the management of massive pulmonary embolism. Profound circulatory disturbances caused by pulmonary embolism are attended by poor prognosis,and death can occur at any time. A watching policy and the uncertain effects of thrombolytic enzyme therapy are not acceptable in such patients. In an ideal situation, and particularly when the patient has had an episode of cardiac arrest, pulmonary embolectomy performed with extracorporeal circulation provides the best conditions for survival, but these advantages are negated if removal of the embolus is unduly delayed by the rather complex preparations required. Pulmonary embolectomy by normothermic venous inflow occlusion can be performed in any operating room with the assistance of staff untrained in cardiac surgery. The operation can be mastered by any surgeon with a basic understanding of cardiac surgical techniques. The interval between diagnosis and operation is minimal. This operation offers an excellent chance of survival to any patient having a massive pulmonary embolus in any hospital provided that an episode of cardiac arrest has not occurred, with results which compare favorably with those achieved with cardiopulmonary bypass.
Pulmonary embolectomy
2 3 4
5
6 7 8 9
10
II
12
13
14
15
16 17
445
REFERENCES Lewis I: Problems in diagnosis and management of pulmonary embolism, Modern Trends in Cardiac Surgery, HRS Harley, ed., London, 1960, Butterworth & Co., Ltd., P 64 Clarke DB: Pulmonary embolectomy using normothermic venous inflow occlusion. Thorax 23: 131-135, 1968 Trendelenberg F: Ueber die operative behandlung der embolie der lungenarterie. Arch Klin Chir 86:686, 1908 Sabiston D: Trendelenberg's classic work on the operative treatment of pulmonary embolism. Ann Thorac Surg 35:570-574, 1983 Miller GAH, Hall RJC, Paneth M: Pulmonary embolectomy, heparin and streptokinase. Their place in the treatment of acute massive pulmonary embolism. Am Heart J 93:568-574, 1977 Gorham LW: A study of pulmonary embolism. Arch Intern Med 108:8-22, 1961 Soloff LA, Rodman T: Acute pulmonary embolism. II. Clinical. Am Heart J 74:829-847, 1967 Sautter RD: Treatment of massive pulmonary embolism (editorial). Chest 71:127-128, 1977 Sharp EH: Pulmonary embolectomy. Successful removal of a massive pulmonary embolus with the support of cardiopulmonary bypass. Case report. Ann Surg 156: 1-4, 1962 Cross FS, Mowlem A: A survey of the current status of pulmonary embolectomy for massive pulmonary embolus. Circulation 35:Suppl 1:86-91, 1967 Heimbacher RO, Keon WJ, Richards KV: Massive pulmonary embolism. A new look at surgical management. Arch Surg 107:740-746, 1973 Mattox KL, Feldtman RW, Beall AC, DeBakey ME: Pulmonary embolectomy for acute massive pulmonary embolism. Ann Surg 195:726-729, 1982 Kieny R, Eisemann JG, Kretz JG, Levy MT, Heitz A: Traitement chirurgical de l'embolie pulmonaire massive. A propos d'une sene de 62 embolectomies effectuees avec succes. Chirurgie 106:370-373, 1980 Greenfield LJ, Zocco JJ: Intraluminal management of acute massive pulmonary embolism. J THORAC CARDIoVASC SURG 77:402-410, 1979 Vossschuite K, Stiller H, Eisenreich F: Emergency embolectomy by the transstemal approach in acute pulmonary embolism. Surgery 58:317-323, 1965 Clarke DB, Abrams LD: Pulmonary embolectomy with venous inflow occlusion. Lancet 1:767-769, 1972 Clarke DB: Pulmonary embolectomy re-evaluated. Ann R Coli Surg Engl 63:18-24, 1981