- Email: [email protected]
Management of air embolism in blunt and penetrating thoracic trauma
J
THoRAc CARDIOVASC SURG
85:661-668, 1983
Management of air embolism in blunt and penetrating thoracic trauma The charts of 61 patients treated from 1970 through 1981 were reviewed to determine the clinical outcome after treatment of air embolism from blunt (15 patients) and penetrating (21 gunshot and 25 stabbing) thoracic injuries. The diagnosis of air embolism was confirmed by the presence of air in the coronary arteries (57%), air aspirated from the heart (30%) or major artery (10%), or Doppler findings (3%). All patients were in shock or cardiac arrest, and in 36% of these patients there were early signs of hemoptysis or unexpected arrest after intubation and positive-pressure ventilation. Successful management included (1) early thoracotomy for diagnosis as well as for specific treatment, (2) hilar cross-clamping for control of bronchovenous communication, (3) maintenance of perfusion pressures with fluids, vasopressors, or aortic cross-clamping, and (4) prompt correction of the embolic source, usually a lung resection. The overall survival rate was 44 %, which correlated with the mechanism of injury, with associated nonthoracic injuries, and with the occurrence of arrest in a controlled setting. We conclude that (1) air embolism can insidiously occur even in blunt trauma; (2) suspicion should be high with hemoptysis or unexpected arrest; and (3) successful treatment includes immediate thoracotomy for diagnosis, resuscitation, and prompt control of the bronchovenous communication.
Edward S. Vee, M.D. (by invitation), Edward D. Verrier, M.D. (by invitation), and Arthur N. Thomas, M.D., San Francisco, Calif.
Entrance of air into the vascular system can be a serious and often fatal condition even in controlled situations such as cardiac operations.':' Fortunately, such fatal events are extremely rare.v" Small amounts of systemic air embolism during cardiopulmonary bypass are well tolerated.s ' Air embolism following trauma has not been universally recognized.r" The reason, in part, is that air embolism associated with trauma is more insidious and the diagnosis is usually not appreciated unless special circumstances allow discovery of air in a major vessel or coronary artery, such as during emergency resuscitative thoracotomy." Postmortem diagnosis can be difficult.":" Since the original reports of three successfully managed cases of traumatic air embolism at San Francisco General Hospital almost 10 years ago,Il,14 subsequent reports have either confirmed traumatic air embolism to be universally lethal"? or have questioned the existence
From the Department of Surgery 3-A. San Francisco General Hospital and University of California, San Francisco, Calif. Read at the Sixty-second Annual Meeting of The American Association for Thoracic Surgery, Phoenix, Ariz., May 3-5, 1982. Address for reprints: Dr. Yee, Department of Surgery, 488 Moffitt Hospital, University of California, San Francisco, Calif. 94143.
of this clinical entity.'?" The present retrospective study was undertaken to examine the characteristics and successful management of documented air embolism following thoracic trauma at San Francisco General Hospital.
Material and methods The charts of 447 patients who were treated for major thoracic trauma at San Francisco General Hospital from January, 1970, through December, 1981, were reviewed. Adequate and complete data for documented diagnosis of air embolism were found in 61 patients. These patients were found to have air in the coronary vessels, air aspirated from the heart or a major vessel during resuscitation, or Doppler fmdings (at time of craniotomy) consistent with air embolism. Numerous variables were tabulated and analyzed multifactorially in order to determine factors which influenced the eventual outcome. These variables included age, status at the time of admission, the mechanism of injury, the site of lung and heart injury, the means of diagnosing and managing air embolism, the associated thoracic and nonthoracic injuries, operative outcome and postoperative complications, adjunctive drug therapy, and eventual outcome. Operative reports were scrutinized to ascertain the type of incision used, 661
The Journal of Thoracic and Cardiovascular Surgery
662 Yee, Verrier, Thomas
Table I. Site of documented traumatic air embolism
Table ID. Lung and associated thoracic injuries
Patients No. Coronary arteries Left anterior descending Left circumflex Right coronary Heart Left ventricle Right ventricle Right atrium Arteries Femoral Aorta Retinal Cerebral (Doppler) Totals
Survivors/total (%) (%)
35
57
31 2 2 18
30
14 2 2 8
13
3 2 I
Admission status Shock Arrest During first hospital day Shock Arrest Mechanism of injury Penetrating Blunt. Associated nonthoracic injuries Present Absent
27/61 (44) 12/31 15/30 8/16(50) 3/6 3/4 1/4 1/2 4/8 (50) 1/1 2/3 1/4 0/1
(0)
2 61
100
Table II. Survival statistics Variables
Lung Right Left Cardiac Left ventricle Right ventricle Left atrium Right atrium Great vessels Pulmonary artery Pulmonary vein Cava-innominate Esophageal
Survivors/total (%)
9/37 (24)
P Value
< 0.001
8/22 (36) 1/15 (7) 18/24 (75) 5/8 (63) 13/16 (81) <0.05 24/46 (52) 3/15 (20) <0.01 3/18 (17) 24/43 (56)
the need for either hilar cross-clamping or lung clamping to control the bronchovenous communications, the use of aortic cross-clamping to maintain systemic arterial pressures, and the final operative techniques for the treatment of the bronchovenous communication. Data analysis included a contingency table of qualitative variables. Chi square statistical analysis was used unless there were fewer than two factors, in which case Fisher's exact test was used. Comparison of quantitative variables with respect to the events of hospital death were made by means of the unpaired t test. Results The diagnosis of air embolism was confirmed by the reported presence of air in the coronary arteries, air aspirated from the heart or a major artery, or Doppler findings as summarized in Table I. Age did not influence the survival rates. The average age was 29 years (2 to 65 years). Shock (30 patients) or cardiac arrest (31 patients) developed in all patients,
either by the time of admission (37 patients) or later during the first day of hospitalization (24 patients) (Table 11). As would be expected, the survival rate correlated inversely with the severity of shock or cardiovascular collapse at the time of admission. In contrast, survival was improved if the precipitating events of air embolism occurred during the first hospital day, usually during the first 12 hours, rather than when the patient presented in the emergency room (75% versus 24%, p < 0.(01). Of note, 13 of 16 (Table II) who had cardiac arrest on the first hospital day survived. The better prognosis and survival rates in this group were partially explained by the presence of more clinical signs, such as hemoptysis or bleeding from the endotracheal tube (10/61 patients, 16%) or unexpected arrest after intubation and administration of positive-pressure ventilation (12/61 patients, 20%). Forty-six of the patients had penetrating injuries (21 gunshot wounds and 25 stab wounds), with an overall survival rate of 52%. The mechanism of penetrating injury did not affect survival in this selective series. Fifteen patients had blunt trauma (six from falls and nine from motor vehicle accidents). Only three patients of whom survived (20%). Associated lung injury was due to multiple rib fractures in 10 of the patients having blunt injury, whereas in the other five patients there was major disruption of the hilum. The poorer prognosis in the blunt trauma group was partially explained by the greater severity of associated nonthoracic injuries, with a mean of 1.9 ± 0.8 organs injured (versus 0.2 ± 0.2 for penetrating trauma, p < 0.(01). The survival rates correlated with the outcome of the nonthoracic injuries (p < 0.01, Table 11). There were no survivors among five patients with associated direct traumatic neurologic damage, only one survivor among three patients with splenic trauma, only one survivor among seven with severe liver injuries, and
Volume 85
Air embolism
Number 5 May. 1983
663
Table IV. Operative treatment for traumatic air embolism Survivors/total (%)
p Value
14/23 7/15 3/18 3/5
Compared
Incision Left thoracotomy Right thoracotomy Bilateral Sternotomy Hilar or pulmonary cross-clamping Performed Not used Aortic cross-clamping Performed Not used Vasoprcssors Used Not given Lung resections Pneumonectomy
Lobectomy Wcdgc No lung resections
(61) (47) (17) (30)
NS
0.025
NS
0.005 19/28 (63) 8/33 (24)
NS
13/31 (42) 14/30 (47)
NS 21/46 (46) 6/15 (40) 12/17 (75)
0.01
15/44 (33)
Compared
5/10 (50) 5/5 (100) 2/2 (100)
only one survivor among three with multiple pelvic and long bone fractures. Surprisingly, the survival rates were not significantly different in patients with other associated thoracic trauma, the majority being cardiac and great vessel injuries (Table III). All 16 cardiac injuries resulted from penetrating trauma. Three of the four deaths associated with great vessel injuries were the result of blunt trauma. Successful operative techniques among the survivors are summarized in Table IV. The most successful incision used was the left anterior thoracotomy in the fourth or fifth intercostal space (61% survivors). Patients receiving a bilateral thoracotomy had the lowest survival rate because the contralateral thoracotomy was usually performed as the last effort and probably is a reflection of delayed recognition of the embolic source. Nevertheless, it was successful in 17% of the cases. The best predictor of successful operative survival is the adequacy of early control of the source of air embolus (19/28, 63%; P < 0.005, Table IV). Bronchovenous communications usually were controlled either by proximal hilar cross-clamping (21/28, 75%) or by isolating the lung injury with clamps (7/28,25%) (Fig. I). These techniques were initially unsuccessfully attempted in two patients with left hilar injuries, but control was aided by advancing the endotracheal tube into the right main-stem bronchus. Hilar cross-clamping was universally successful when cardiac arrest occurred in the operating room. The technique of thoracic aortic cross-clamping (Fig.
Table V. Postoperative findings and complications No. Thoracic Left ventricular failure/death Myocardial infarction Pneumonia Rcoperation for bleeding
2
2 2 2
Central nervous system
Neurologic death
2
Paralysis/paresis
3 4
Confusion (wk)
2) to maintain perfusion pressure to the heart and brain was not statistically helpful. This surgical technique was done predominantly in conjunction with the emergency room thoracotomy (29 patients with seven survivors). Among the patients in whom the aorta was crossclamped in the operating room, four had had prior control of a bronchovenous communication and two had had prior evacuation of air from the left ventricle (six survivors among seven patients, p < 0.025). Adjunctive drug therapy was given in 75% of the 61 patients for short-term maintenance of perfusion pressures. Epinephrine was used as a cardiorespiratory stimulant in all 31 patients who had cardiac arrest. In another 15 patients additional drugs, dopamine or isoproterenol (lsuprel), were given to treat persistent hypotension and/or bradycardia. Survival rates were not statistically different (Table IV). If the resuscitation and treatment of patients with associated thoracic and nonthoracic injuries were successful, then a lung resection was performed. A majority
664 Yee, Verrier, Thomas
The Journal of Thoracic and Cardiovascular Surgery
Fig. 2. Simultaneous control of bronchovenous communication in a peripheral lung injury and descending aortic crossclamping for proximal hypertension and resuscitation of coronary air embolism. Fig. 1. Emergency hilar cross-clamping for proximal control of bronchovenous communication.
of the proximal or hilar lung injuries (17/27 patients) necessitated pulmonary resection. Five of the 10 patients having pneumonectomy died because of associated severe trauma to the head and abdomen. In the other 15 survivors,' the lung injury was peripheral and the bronchovenous communication was simply corrected by direct local ligation. The overall mortality was 56% (34/61 patients). The causes of the 30 operative deaths were refractory arrest in 19, exsanquination in eight, and severe brain trauma in three. Four late deaths were due to pump failure (two patients) and brain damage from multiple cerebral infarcts (two patients). Surprisingly, a majority of the patients with diagnosed air embolism to a coronary artery (31/35 patients or 89%) had normal electrocardiograms and serum cardiac enzymes. Seven surviving patients (26%) had less severe clinical neurologic findings (paresis, paralysis, and prolonged confusion) with complete resolution by the time of discharge (Table V).
Discussion Classically, air embolism has been divided into two clinical categories according to whether the right or left side of the circulation is involved.'? A relatively large amount of air can be tolerated on the right, but a very small volume on the left (as little as 0.5 rnl into the left anterior descending artery) can be lethal. l 8•23 The clini-
cal significance of air embolism following trauma is almost universally systemic air embolism as manifested in the coronary or cerebral circulation. The clinical diagnosis can be overlooked because (1) the nature of air embolism is insidious, (2) the signs and symptoms are not specific and not different from hypovolemic shock or arrest, and (3) documentation of the air embolism is most often fortuitous. Strict criteria for clinical diagnosis of traumatic air embolism was used in this retrospective review. Many cases in which air embolism was suspected as the cause of cardiovascular and cerebral insults were not included in this study if the criteria outlined for definitive diagnosis were not met. Documentation is facilitated by early thoracotomy, which is a part of the established resuscitative measures in the emergency and operating rooms at the San Francisco General Hospital. With the pericardium routinely opened, coronary arteries can be inspected and ventricular cavities aspirated, which was the method of diagnosis in almost 85% of the cases. The most common site of coronary artery air embolus was the left anterior descending artery, which is probably due to the anterior (superior) distribution of air flow as well as to the most common type of incision used in making the diagnosis of air embolism (the left anterior thoracotomy). Other traumatic lesions masked the diagnosis of air embolism because the clinical signs and symptoms were the same as hypovolemic shock and arrest. Two important subtle signs were noted in 36% of the patientseither hemoptysis or cardiac arrest after intubation and institution of positive-pressure ventilation. Hemoptysis
Volume 85 Number 5 May. 1983
was the first, and sometimes the only, clinical sign in patients with blunt thoracic trauma. These clinical signs prompted an earlier diagnosis and quicker treatment with emergency thoracotomy which, in part, explained the better outcome as compared to thoracotomy resuscitation in the emergency room. Endotracheal intubation promoted the recognition of air embolus since extremely high airway pressures occurred during resuscitative efforts (as documented by others). 1I. 12.24. 25 This mechanism was probably the precipitating event for further dissemination of air. Patients with low pulmonary venous pressures were at an additional risk, as may often be the case, when the operation was done primarily for treatment of shock.": 14 Air embolism associated with trauma is more insidious than its occurrence during other controlled operative settings. The operating team's attention is often directed elsewhere, to intrathoracic (41%) or extrathoracic (30%) injuries in this retrospective series. This was especially true in cases of blunt trauma, which involved more severe and a larger number of injuries. These additive factors collectively contributed to the lower survival rates for patients having air embolism after blunt trauma. The current series suggested that the likelihood of successful treatment was greatly increased if air embolism occurred in a controlled setting. This fact is commonly recognized during elective surgical procedures, such as cardiopulmonary bypass with the occurrence of systemic embolism. 1-6 Under these circumstances, with early diagnosis, specific treatment to control the source of air embolus can be quickly instituted and vital cardiovascular systems can be maintained.!" Similarly, the prognosis of air embolism occurring during operative treatment for trauma is expected to be better because therapy can be instituted immediately. As previously mentioned, early thoracotomy aided in diagnosis. It also was the first of a series of operative maneuvers. Resuscitative thoracotomy allowed early hilar cross-damping or isolation of lung injury by clamps. This operative technique was the key therapeutic step affecting survival and clinical outcome. High intra bronchial pressures were created during the resuscitative efforts, and without proximal control, more air was probably introduced, as seen experimentally.t":":" Cooperative attempts with intrathoracic guidance and advancement of the endotracheal tube can be extremely helpful for proximal hilar injuries, since cross-clamping can be difficult for proximal left-sided hilar lesions. Peripheral bronchovenous communications could be identified during positive-pressure ventilation, which
Air embolism
665
usually resulted in bloody froth. Contralateral thoracotomy for isolation of the lung injury should always be considered if the source of air embolism is not immediately evident. Once the bronchovenous communication is controlled, efforts should be directed to safely evacuate the residual air that is commonly located in the left side of the heart and aorta.' Left ventricular aspiration with a needle was most effective and can be done in conjunction with intracardiac administration of drugs for cardiac arrest. Remedial measures for directing air to less critical organs should include placing the patient in the head-down position. Proximal hypertension to flush air embolus through the arteries can be induced by temporarily crossclamping the descending aorta, as first suggested and tried by Geoghegan and Lam. 20 This operative technique was significantly less successful than hilar crossclamping for numerous reasons. Aortic cross-damping was predominantly done as a reflex or an adjunctive maneuver to the emergency room thoracotomy after cardiac arrest had already occurred. Although difficult to prove directly in this review, aortic cross-damping can be deleterious in this setting, since the bronchovenous communication had not been controlled initially and aortic cross-clamping might have further disseminated air to the vital organs (heart and brain). The cases in which aortic cross-clamping was successful probably occurred when this measure was used prospectively after the source of the air embolism had been controlled. Systemic arterial pressure was maintained by adequate fluid resuscitation, and intracardiac or intravenous epinephrine was used only for patients in extremis (51%). Since most of these patients had normal myocardium, further inotropic and chronotropic support with vasopressors was not routinely needed «25%). Proximal bronchial and associated venous injuries were best managed by lung resections, either pneumonectomy or lobectomy. The more peripheral lesions were controlled by simple ligation of the bronchovenous communication or wedge resections. Postoperative complications were a further reflection of difficulties encountered during operative management. The two vital organs most directly affected by systemic air embolism were the heart and brain, and injury to those organs was the common cause of postoperative deaths and complications. The lack of enzymatic or electrocardiographic confirmation postoperatively in a majority of the patients with coronary air embolism is due to the reversibility of air embolisme' " or the lack of sensitivity of these tests to diagnose small focal infarcts." 23
The Journal of Thoracic and Cardiovascular Surgery
666 Yee, Verrier, Thomas
The postoperative neurologic problems of confusion were atypical and prolonged (usually greater than 2 weeks) but slowly resolved with time. Recently, specific adjunctive measures for cerebral protection, such as hypothermia, hyperbaric oxygenation, or intravascular carbon dioxide treatment during cardiopulmonary bypass,26-30 have been suggested for cerebral air embolism but were not tried in these trauma patients. These methods are probably of limited use because of the practical as well as technical problems of applying them in the patient with multisystem injuries. Only under optimal clinical settings can these suggested treatments be tried. Hypothermia promotes bleeding, hyperbaric oxygenation requires special facilities (further transportation of a critically ill patient), and intravascular carbon dioxide washout is difficult without cardiopulmonary bypass. Cerebral air embolism is even more difficult to diagnose prospectively than cardiac air insults.l':" All patients with suspected embolism should have their retinal arteries carefully examined" during operative or resuscitative measures. Although convalescence is prolonged, complete recovery is more common after nonfatal cerebral air injury than after embolism from clots. 6 The subtle occurrence of air embolism with traumatic thoracic injuries has been documented. This clinical entity can occur even with blunt trauma. The outcome is not universally lethal if cardiac arrest and associated problems are promptly and appropriately treated. Early signs of hemoptysis or unexpected arrest in the operating room are favorable because they prompt earlier diagnosis for successful management. We sincerely thank Miss Sheila Finnigan for her help in preparing this manuscript. REFERENCES
2
3
4 5
6
Justice C, Leach J, Edwards WS: The harmful effects and treatment of coronary air embolism during open-heart surgery. Ann Thorac Surg 14:47-53, 1972 Lawrence GH, McKay HA, Sherensky RT: Effective measures in the prevention of intraoperative aeroembolus. J THORAC CARDIOVASC SURG 62:731-735, 1971 Taber RE, Maraan BM, Tomatis L: Prevention of air embolism during open-heart surgery. A study of the role of trapped air in the left ventricle. Surgery 68:685-691, 1970 Nicks R: Arterial air embolism. Thorax 22:320-326, 1967 Spencer FC, Rossi NP, Yu SC, Koepke JA: The significance of air embolism during cardiopulmonary bypass. J THORAC CARDIOVASC SURG 49:615-634, 1965 Mills NL, Ochsner JL: Massive air embolism during
7
8
9 10
11
12
13 14
15
16 17 18 19 20
21
22
23
24
25
26
cardiopulmonary bypass. Causes, prevention, and management. J THORAC CARDIOVASC SURG 80:708-717, 1980 Graham JM, Beall AC, Mattox KL, Vaughan GD: Systemic air embolism following penetrating trauma to the lung. Chest 72:449-454, 1977 Donato AT, Arciniegas E, Lam CR: Fatal air embolism during thoracotomy for gunshot injury to the lung. J THORAC CARDIOVASC SURG 69:827-829, 1975 Ellis GR, Brown JR: Massive air embolism due to gunshot wound. JAMA 189:177-179,1964 Waldo WJ, Harlaftis NN, Symbas PN: Systemic air embolism. Does it occur after experimental penetrating lung injury? J THORAC CARDIOVASC SURG 71:96-101, 1976 Meier GH, Wood WJ, Symbas PN: Systemic air embolization from penetrating lung injury. Ann Thorac Surg 27:161-166,1979 Meier GH, Symbas PN: Systemic air embolization. Factors involved in its production following penetrating lung injury. Am Surg 12:765-771, 1978 Thomas AN: Air embolism following penetrating lung injuries. J THORAC CARDIOVASC SURG 66:533-540, 1973 Thomas AN, Stephens BG: Air embolism. A cause of morbidity and death after penetrating chest trauma. J Trauma 14:633-638, 1974 Erben J, Nadvornik F: The quantitative demonstration of air embolism in certain cases of fatal trauma. J Forensic Med 10:45-50, 1963 Taylor JD: Post-mortem diagnosis of air embolism by radiography. Br Med J 4:890-893, 1952 Durant TM, Long J, Oppenheimer MJ: Pulmonary (venous) air embolism. Am Heart J 33:269-281, 1947 Rukstinat G: Experimental air embolism of the coronary arteries. JAMA 96:26-28, 1931 Rukstinat GJ, LeCount ER: Air in the coronary arteries. JAMA 91:1776-1779, 1929 Geoghegan T, Lam CR: The mechanism of death from intracardiac air and its reversibility. Ann Surg 138:351359, 1933 James TN, Geoghegan T, Lam CR: Electrocardiographic manifestations of air in the coronary arteries of dying and resuscitated hearts. Am Heart J 46:215-228, 1953 Eguchi S, Bosher LH: Myocardial dysfunction resulting from coronary air embolism. Surgery 51:103-111, 1962 Goldfarb D, Bahnson HT: Early and late effects on the heart of small amounts of air in the coronary circulation. J THORAC CARDIOVASC SURG 46:368-378, 1963 Lenaghan R, Silva YJ, Walt AJ: Hemodynamic alterations associated with expansion rupture of the lung. Arch Surg 99:339-343, 1969 Chiu CJ, Golding MR, Linder JB, Fries CC: Pulmonary venous air embolism. A hemodynamic reappraisal. Surgery 61:816-819,1967 Tomatis L, Nemiroff M, Riahi M, Visser J, et al: Massive arterial air embolism due to rupture of pulsatile assist
Volume 85
Air embolism
Number 5 May, 1983
27
28
29
30
3I
32
33
device. Successful treatment in the hyperbaric chamber. Ann Thorac Surg 32:604-608, 1981 Steward D, Williams WG, Freedom R: Hypothermia in conjunction with hyperbaric oxygenation in the treatment of massive air embolism during cardipulmonary bypass. Ann Thorac Surg 24:591-593, 1977 Roe BB: Prevention of air embolism with intravascular carbon dioxide washout. J THORAC CARDIOVASC SURG 71:628-630, 1976 Winter PM, Alvis HJ, Gage AA: Hyperbaric treatment of cerebral air embolism during cardiopulmonary bypass. JAMA 215:1786-1788,1971 Spampinato N, Stassano P, Gagliardi C, Tufano R, Iorio D: Massive air embolism during cardiopulmonary bypass. Successful treatment with immediate hypothermia and circulatory support. Ann Thorac Surg 32:602, 198I Bedell AJ: Clinical differentiation of emboli in the retinal arteries from endarteritis. Arch Ophthalmol 34:311-317, 1945 Gomes OM, Pereira SN, Castagna RC, Bittencourt D, Amaral RVG, Zerbini EJ: The importance of the different sites of air injection in the tolerance of arterial air embolism. J THORAC CARDIOVASC SURG 65:563-568, 1973 Gillen HW: Symptomatology of cerebral gas embolism. Neurology 18:507-512, 1968
667
ventilatory pressure becomes excessively high, over 25 to 30 mm Hg, the pressure gradient is reversed and air may be pushed into the pulmonary vein. Such air can get into the coronary arteries and can also go to the brain. A number of years ago I described a patient who suffered a so-called "pleural shock" during thoracentesis, with signs of cerebral air embolism. However, at that time a prominent surgeon who did not believe in the entity of traumatic air embolism dismissed this as just a fainting episode.' Now we have better technology to substantiate our suspicions. This particular young man had a gunshot wound of the left lung and had a sudden cardiac arrest during thoracotomy. The cardiac arrest occurred in a stable patient when the lung was being reexpanded with high airway pressure. Postoperatively, cognitive dysphasia developed and a computed axial tomographic (CAT) scan of the brain showed localized temporoparietal infarction. The lesson, of course, is not to use excessively high ventilatory pressure in such cases. Other measures described by the authors should also be useful. REFERENCE Symbas PN: Reply to Chiu CJ: Air embolism in penetrating lung injury. J THORAC CARDIOVASC SURG 72:815-816, 1976 DR. AGUSTIN ARBULU Detroit. Mich.
Discussion DR. RAY CHU-JENG CHIU Montreal. Quebec. Canada
I would like to make a few additional comments on the mechanism of this condition, since understanding the mechanism may help in maintaining a high index of suspicion. This is important, because the complications of traumatic air embolism are still underdiagnosed and the confirmation of this diagnosis is often very difficult. As mentioned by the authors, a small amount of hemoptysis preceding sudden cardiovascular collapse is not uncommon in traumatic air embolism. Such hemoptysis, in addition to being a telltale sign, may be an important mechanism in causing air embolism. In a conscious patient, blood entering the airway can induce vigorous cough reflex, build up a high intrathoracic pressure, and push the air into the left side of the heart through the bronchopulmonary venous fistula caused by trauma. The role of positive-pressureventilation in causing traumatic' air embolism was elucidated in our study, in which the pulmonary venous pressure and lower airway pressure were simultaneous monitored. The pulmonary venous pressure is higher in when the patient's chest is open and rises with increased ventilatory pressure, because of the transmission of the pressure. Under such circumstances, the blood runs into the bronchus through the venobronchial communication rather than in the reverse direction. Perhaps this is why we do not see air embolism every day. However, when the positive
In the early I970s, as Dr. Richard Lenaghan in our department of surgery was working in the experimental laboratory, he developed a model in dogs to produce air embolism in the venous and arterial systems. The model was very simple. As Dr. Chiu has expressed, increasing the ventilatory pressure to more than 30 mm Hg consistently produced this lesion in dogs. A review of the medical literature reveals that this observation was made early in the 19305 by Dr. Macklin and is well documented and published. I have a question for Dr. Vee. In some of these patients could hyperbaric chambers offer some help? DR. YEE (Closing) I would like to thank Drs. Chiu and Arbulu for their kind remarks. In reply to Dr. Chiu's statements, specific clinical signs of air embolism, such as hemoptysis or unexpected arrest after positive-pressure ventilation, occurred in only 36% of our patients. Hemoptysis occurred in only 16%, a very small number, and occurred only in association with the very proximal injuries. This fact highlighted the insidious nature of this clinical entity. For the other 20%, the early clue of unexpected arrest might be overlooked as simply a manifestation of the associated injuries. Our treatment for arrest at San Francisco General Hospital has been an emergency resuscitative thoracotomy, which permitted an earlier diagnosis with a direct cardiac examination. Systemic air embolism should be
668
Vee, Verrier, Thomas
suspected when cerebral or cardiac dysfunction is out of proportion to the clinical setting. To answer Dr. Arbulu's question, adjunctive therapy for cerebral protection might be helpful, but there are several practical and technical limitations. Preoperative manifestations were commonly overlooked, since they were not frequently localizing. Most of the diagnoses in our trauma patients were made later than would have been the case in patients having cardiopulmonary bypass, when the passage of air is noted during the procedure. Simultaneous cerebral CAT scans during chest or abdominal studies might be helpful, as illustrated by Dr. Chiu's case.
The Journal of Thoracic and Cardiovascular Surgery
Usage of hypothermia, hyperbaric oxygenation or carbon dioxide, and barbiturates has been of limited help in the injured patient when the control of hemorrhage, myocardial resuscitation, and immediate intensive care are mandatory. One anecdotal example, however, was a 25-year-old man who had both an overdose of barbiturates and cardiac arrest from a chest stab wound. Circumstantial hypothermia (90 0 F) from transfusion of 11 L of blood and fluids probably helped his complete neurologic recovery.
Bound volumes available to subscribers Bound volumes of THE JOURNAL OF THORAOC AND CARDIOVASCULAR SURGERY are available to subscribers (only) for the 1983 issues from the Publisher, at a cost of $58.50 ($73.00 international) for Vol. 85 (January-June) and Vol. 86 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertising is removed. Copies are shipped within 30 days after publication of the last issue of the volume. The binding is durable buckram with the JOURNAL name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mr. Deans Lynch at The C. V. Mosby Company, 11830 Westline Industrial Drive, St. Louis, Missouri 63141, USA. Subscriptions must be In force to qua6fy. Bound "oIumes an not a"ailable In pilla of a regular JOURNAL subscription.
THoRAc CARDIOVASC SURG
85:661-668, 1983
Management of air embolism in blunt and penetrating thoracic trauma The charts of 61 patients treated from 1970 through 1981 were reviewed to determine the clinical outcome after treatment of air embolism from blunt (15 patients) and penetrating (21 gunshot and 25 stabbing) thoracic injuries. The diagnosis of air embolism was confirmed by the presence of air in the coronary arteries (57%), air aspirated from the heart (30%) or major artery (10%), or Doppler findings (3%). All patients were in shock or cardiac arrest, and in 36% of these patients there were early signs of hemoptysis or unexpected arrest after intubation and positive-pressure ventilation. Successful management included (1) early thoracotomy for diagnosis as well as for specific treatment, (2) hilar cross-clamping for control of bronchovenous communication, (3) maintenance of perfusion pressures with fluids, vasopressors, or aortic cross-clamping, and (4) prompt correction of the embolic source, usually a lung resection. The overall survival rate was 44 %, which correlated with the mechanism of injury, with associated nonthoracic injuries, and with the occurrence of arrest in a controlled setting. We conclude that (1) air embolism can insidiously occur even in blunt trauma; (2) suspicion should be high with hemoptysis or unexpected arrest; and (3) successful treatment includes immediate thoracotomy for diagnosis, resuscitation, and prompt control of the bronchovenous communication.
Edward S. Vee, M.D. (by invitation), Edward D. Verrier, M.D. (by invitation), and Arthur N. Thomas, M.D., San Francisco, Calif.
Entrance of air into the vascular system can be a serious and often fatal condition even in controlled situations such as cardiac operations.':' Fortunately, such fatal events are extremely rare.v" Small amounts of systemic air embolism during cardiopulmonary bypass are well tolerated.s ' Air embolism following trauma has not been universally recognized.r" The reason, in part, is that air embolism associated with trauma is more insidious and the diagnosis is usually not appreciated unless special circumstances allow discovery of air in a major vessel or coronary artery, such as during emergency resuscitative thoracotomy." Postmortem diagnosis can be difficult.":" Since the original reports of three successfully managed cases of traumatic air embolism at San Francisco General Hospital almost 10 years ago,Il,14 subsequent reports have either confirmed traumatic air embolism to be universally lethal"? or have questioned the existence
From the Department of Surgery 3-A. San Francisco General Hospital and University of California, San Francisco, Calif. Read at the Sixty-second Annual Meeting of The American Association for Thoracic Surgery, Phoenix, Ariz., May 3-5, 1982. Address for reprints: Dr. Yee, Department of Surgery, 488 Moffitt Hospital, University of California, San Francisco, Calif. 94143.
of this clinical entity.'?" The present retrospective study was undertaken to examine the characteristics and successful management of documented air embolism following thoracic trauma at San Francisco General Hospital.
Material and methods The charts of 447 patients who were treated for major thoracic trauma at San Francisco General Hospital from January, 1970, through December, 1981, were reviewed. Adequate and complete data for documented diagnosis of air embolism were found in 61 patients. These patients were found to have air in the coronary vessels, air aspirated from the heart or a major vessel during resuscitation, or Doppler fmdings (at time of craniotomy) consistent with air embolism. Numerous variables were tabulated and analyzed multifactorially in order to determine factors which influenced the eventual outcome. These variables included age, status at the time of admission, the mechanism of injury, the site of lung and heart injury, the means of diagnosing and managing air embolism, the associated thoracic and nonthoracic injuries, operative outcome and postoperative complications, adjunctive drug therapy, and eventual outcome. Operative reports were scrutinized to ascertain the type of incision used, 661
The Journal of Thoracic and Cardiovascular Surgery
662 Yee, Verrier, Thomas
Table I. Site of documented traumatic air embolism
Table ID. Lung and associated thoracic injuries
Patients No. Coronary arteries Left anterior descending Left circumflex Right coronary Heart Left ventricle Right ventricle Right atrium Arteries Femoral Aorta Retinal Cerebral (Doppler) Totals
Survivors/total (%) (%)
35
57
31 2 2 18
30
14 2 2 8
13
3 2 I
Admission status Shock Arrest During first hospital day Shock Arrest Mechanism of injury Penetrating Blunt. Associated nonthoracic injuries Present Absent
27/61 (44) 12/31 15/30 8/16(50) 3/6 3/4 1/4 1/2 4/8 (50) 1/1 2/3 1/4 0/1
(0)
2 61
100
Table II. Survival statistics Variables
Lung Right Left Cardiac Left ventricle Right ventricle Left atrium Right atrium Great vessels Pulmonary artery Pulmonary vein Cava-innominate Esophageal
Survivors/total (%)
9/37 (24)
P Value
< 0.001
8/22 (36) 1/15 (7) 18/24 (75) 5/8 (63) 13/16 (81) <0.05 24/46 (52) 3/15 (20) <0.01 3/18 (17) 24/43 (56)
the need for either hilar cross-clamping or lung clamping to control the bronchovenous communications, the use of aortic cross-clamping to maintain systemic arterial pressures, and the final operative techniques for the treatment of the bronchovenous communication. Data analysis included a contingency table of qualitative variables. Chi square statistical analysis was used unless there were fewer than two factors, in which case Fisher's exact test was used. Comparison of quantitative variables with respect to the events of hospital death were made by means of the unpaired t test. Results The diagnosis of air embolism was confirmed by the reported presence of air in the coronary arteries, air aspirated from the heart or a major artery, or Doppler findings as summarized in Table I. Age did not influence the survival rates. The average age was 29 years (2 to 65 years). Shock (30 patients) or cardiac arrest (31 patients) developed in all patients,
either by the time of admission (37 patients) or later during the first day of hospitalization (24 patients) (Table 11). As would be expected, the survival rate correlated inversely with the severity of shock or cardiovascular collapse at the time of admission. In contrast, survival was improved if the precipitating events of air embolism occurred during the first hospital day, usually during the first 12 hours, rather than when the patient presented in the emergency room (75% versus 24%, p < 0.(01). Of note, 13 of 16 (Table II) who had cardiac arrest on the first hospital day survived. The better prognosis and survival rates in this group were partially explained by the presence of more clinical signs, such as hemoptysis or bleeding from the endotracheal tube (10/61 patients, 16%) or unexpected arrest after intubation and administration of positive-pressure ventilation (12/61 patients, 20%). Forty-six of the patients had penetrating injuries (21 gunshot wounds and 25 stab wounds), with an overall survival rate of 52%. The mechanism of penetrating injury did not affect survival in this selective series. Fifteen patients had blunt trauma (six from falls and nine from motor vehicle accidents). Only three patients of whom survived (20%). Associated lung injury was due to multiple rib fractures in 10 of the patients having blunt injury, whereas in the other five patients there was major disruption of the hilum. The poorer prognosis in the blunt trauma group was partially explained by the greater severity of associated nonthoracic injuries, with a mean of 1.9 ± 0.8 organs injured (versus 0.2 ± 0.2 for penetrating trauma, p < 0.(01). The survival rates correlated with the outcome of the nonthoracic injuries (p < 0.01, Table 11). There were no survivors among five patients with associated direct traumatic neurologic damage, only one survivor among three patients with splenic trauma, only one survivor among seven with severe liver injuries, and
Volume 85
Air embolism
Number 5 May. 1983
663
Table IV. Operative treatment for traumatic air embolism Survivors/total (%)
p Value
14/23 7/15 3/18 3/5
Compared
Incision Left thoracotomy Right thoracotomy Bilateral Sternotomy Hilar or pulmonary cross-clamping Performed Not used Aortic cross-clamping Performed Not used Vasoprcssors Used Not given Lung resections Pneumonectomy
Lobectomy Wcdgc No lung resections
(61) (47) (17) (30)
NS
0.025
NS
0.005 19/28 (63) 8/33 (24)
NS
13/31 (42) 14/30 (47)
NS 21/46 (46) 6/15 (40) 12/17 (75)
0.01
15/44 (33)
Compared
5/10 (50) 5/5 (100) 2/2 (100)
only one survivor among three with multiple pelvic and long bone fractures. Surprisingly, the survival rates were not significantly different in patients with other associated thoracic trauma, the majority being cardiac and great vessel injuries (Table III). All 16 cardiac injuries resulted from penetrating trauma. Three of the four deaths associated with great vessel injuries were the result of blunt trauma. Successful operative techniques among the survivors are summarized in Table IV. The most successful incision used was the left anterior thoracotomy in the fourth or fifth intercostal space (61% survivors). Patients receiving a bilateral thoracotomy had the lowest survival rate because the contralateral thoracotomy was usually performed as the last effort and probably is a reflection of delayed recognition of the embolic source. Nevertheless, it was successful in 17% of the cases. The best predictor of successful operative survival is the adequacy of early control of the source of air embolus (19/28, 63%; P < 0.005, Table IV). Bronchovenous communications usually were controlled either by proximal hilar cross-clamping (21/28, 75%) or by isolating the lung injury with clamps (7/28,25%) (Fig. I). These techniques were initially unsuccessfully attempted in two patients with left hilar injuries, but control was aided by advancing the endotracheal tube into the right main-stem bronchus. Hilar cross-clamping was universally successful when cardiac arrest occurred in the operating room. The technique of thoracic aortic cross-clamping (Fig.
Table V. Postoperative findings and complications No. Thoracic Left ventricular failure/death Myocardial infarction Pneumonia Rcoperation for bleeding
2
2 2 2
Central nervous system
Neurologic death
2
Paralysis/paresis
3 4
Confusion (wk)
2) to maintain perfusion pressure to the heart and brain was not statistically helpful. This surgical technique was done predominantly in conjunction with the emergency room thoracotomy (29 patients with seven survivors). Among the patients in whom the aorta was crossclamped in the operating room, four had had prior control of a bronchovenous communication and two had had prior evacuation of air from the left ventricle (six survivors among seven patients, p < 0.025). Adjunctive drug therapy was given in 75% of the 61 patients for short-term maintenance of perfusion pressures. Epinephrine was used as a cardiorespiratory stimulant in all 31 patients who had cardiac arrest. In another 15 patients additional drugs, dopamine or isoproterenol (lsuprel), were given to treat persistent hypotension and/or bradycardia. Survival rates were not statistically different (Table IV). If the resuscitation and treatment of patients with associated thoracic and nonthoracic injuries were successful, then a lung resection was performed. A majority
664 Yee, Verrier, Thomas
The Journal of Thoracic and Cardiovascular Surgery
Fig. 2. Simultaneous control of bronchovenous communication in a peripheral lung injury and descending aortic crossclamping for proximal hypertension and resuscitation of coronary air embolism. Fig. 1. Emergency hilar cross-clamping for proximal control of bronchovenous communication.
of the proximal or hilar lung injuries (17/27 patients) necessitated pulmonary resection. Five of the 10 patients having pneumonectomy died because of associated severe trauma to the head and abdomen. In the other 15 survivors,' the lung injury was peripheral and the bronchovenous communication was simply corrected by direct local ligation. The overall mortality was 56% (34/61 patients). The causes of the 30 operative deaths were refractory arrest in 19, exsanquination in eight, and severe brain trauma in three. Four late deaths were due to pump failure (two patients) and brain damage from multiple cerebral infarcts (two patients). Surprisingly, a majority of the patients with diagnosed air embolism to a coronary artery (31/35 patients or 89%) had normal electrocardiograms and serum cardiac enzymes. Seven surviving patients (26%) had less severe clinical neurologic findings (paresis, paralysis, and prolonged confusion) with complete resolution by the time of discharge (Table V).
Discussion Classically, air embolism has been divided into two clinical categories according to whether the right or left side of the circulation is involved.'? A relatively large amount of air can be tolerated on the right, but a very small volume on the left (as little as 0.5 rnl into the left anterior descending artery) can be lethal. l 8•23 The clini-
cal significance of air embolism following trauma is almost universally systemic air embolism as manifested in the coronary or cerebral circulation. The clinical diagnosis can be overlooked because (1) the nature of air embolism is insidious, (2) the signs and symptoms are not specific and not different from hypovolemic shock or arrest, and (3) documentation of the air embolism is most often fortuitous. Strict criteria for clinical diagnosis of traumatic air embolism was used in this retrospective review. Many cases in which air embolism was suspected as the cause of cardiovascular and cerebral insults were not included in this study if the criteria outlined for definitive diagnosis were not met. Documentation is facilitated by early thoracotomy, which is a part of the established resuscitative measures in the emergency and operating rooms at the San Francisco General Hospital. With the pericardium routinely opened, coronary arteries can be inspected and ventricular cavities aspirated, which was the method of diagnosis in almost 85% of the cases. The most common site of coronary artery air embolus was the left anterior descending artery, which is probably due to the anterior (superior) distribution of air flow as well as to the most common type of incision used in making the diagnosis of air embolism (the left anterior thoracotomy). Other traumatic lesions masked the diagnosis of air embolism because the clinical signs and symptoms were the same as hypovolemic shock and arrest. Two important subtle signs were noted in 36% of the patientseither hemoptysis or cardiac arrest after intubation and institution of positive-pressure ventilation. Hemoptysis
Volume 85 Number 5 May. 1983
was the first, and sometimes the only, clinical sign in patients with blunt thoracic trauma. These clinical signs prompted an earlier diagnosis and quicker treatment with emergency thoracotomy which, in part, explained the better outcome as compared to thoracotomy resuscitation in the emergency room. Endotracheal intubation promoted the recognition of air embolus since extremely high airway pressures occurred during resuscitative efforts (as documented by others). 1I. 12.24. 25 This mechanism was probably the precipitating event for further dissemination of air. Patients with low pulmonary venous pressures were at an additional risk, as may often be the case, when the operation was done primarily for treatment of shock.": 14 Air embolism associated with trauma is more insidious than its occurrence during other controlled operative settings. The operating team's attention is often directed elsewhere, to intrathoracic (41%) or extrathoracic (30%) injuries in this retrospective series. This was especially true in cases of blunt trauma, which involved more severe and a larger number of injuries. These additive factors collectively contributed to the lower survival rates for patients having air embolism after blunt trauma. The current series suggested that the likelihood of successful treatment was greatly increased if air embolism occurred in a controlled setting. This fact is commonly recognized during elective surgical procedures, such as cardiopulmonary bypass with the occurrence of systemic embolism. 1-6 Under these circumstances, with early diagnosis, specific treatment to control the source of air embolus can be quickly instituted and vital cardiovascular systems can be maintained.!" Similarly, the prognosis of air embolism occurring during operative treatment for trauma is expected to be better because therapy can be instituted immediately. As previously mentioned, early thoracotomy aided in diagnosis. It also was the first of a series of operative maneuvers. Resuscitative thoracotomy allowed early hilar cross-damping or isolation of lung injury by clamps. This operative technique was the key therapeutic step affecting survival and clinical outcome. High intra bronchial pressures were created during the resuscitative efforts, and without proximal control, more air was probably introduced, as seen experimentally.t":":" Cooperative attempts with intrathoracic guidance and advancement of the endotracheal tube can be extremely helpful for proximal hilar injuries, since cross-clamping can be difficult for proximal left-sided hilar lesions. Peripheral bronchovenous communications could be identified during positive-pressure ventilation, which
Air embolism
665
usually resulted in bloody froth. Contralateral thoracotomy for isolation of the lung injury should always be considered if the source of air embolism is not immediately evident. Once the bronchovenous communication is controlled, efforts should be directed to safely evacuate the residual air that is commonly located in the left side of the heart and aorta.' Left ventricular aspiration with a needle was most effective and can be done in conjunction with intracardiac administration of drugs for cardiac arrest. Remedial measures for directing air to less critical organs should include placing the patient in the head-down position. Proximal hypertension to flush air embolus through the arteries can be induced by temporarily crossclamping the descending aorta, as first suggested and tried by Geoghegan and Lam. 20 This operative technique was significantly less successful than hilar crossclamping for numerous reasons. Aortic cross-damping was predominantly done as a reflex or an adjunctive maneuver to the emergency room thoracotomy after cardiac arrest had already occurred. Although difficult to prove directly in this review, aortic cross-damping can be deleterious in this setting, since the bronchovenous communication had not been controlled initially and aortic cross-clamping might have further disseminated air to the vital organs (heart and brain). The cases in which aortic cross-clamping was successful probably occurred when this measure was used prospectively after the source of the air embolism had been controlled. Systemic arterial pressure was maintained by adequate fluid resuscitation, and intracardiac or intravenous epinephrine was used only for patients in extremis (51%). Since most of these patients had normal myocardium, further inotropic and chronotropic support with vasopressors was not routinely needed «25%). Proximal bronchial and associated venous injuries were best managed by lung resections, either pneumonectomy or lobectomy. The more peripheral lesions were controlled by simple ligation of the bronchovenous communication or wedge resections. Postoperative complications were a further reflection of difficulties encountered during operative management. The two vital organs most directly affected by systemic air embolism were the heart and brain, and injury to those organs was the common cause of postoperative deaths and complications. The lack of enzymatic or electrocardiographic confirmation postoperatively in a majority of the patients with coronary air embolism is due to the reversibility of air embolisme' " or the lack of sensitivity of these tests to diagnose small focal infarcts." 23
The Journal of Thoracic and Cardiovascular Surgery
666 Yee, Verrier, Thomas
The postoperative neurologic problems of confusion were atypical and prolonged (usually greater than 2 weeks) but slowly resolved with time. Recently, specific adjunctive measures for cerebral protection, such as hypothermia, hyperbaric oxygenation, or intravascular carbon dioxide treatment during cardiopulmonary bypass,26-30 have been suggested for cerebral air embolism but were not tried in these trauma patients. These methods are probably of limited use because of the practical as well as technical problems of applying them in the patient with multisystem injuries. Only under optimal clinical settings can these suggested treatments be tried. Hypothermia promotes bleeding, hyperbaric oxygenation requires special facilities (further transportation of a critically ill patient), and intravascular carbon dioxide washout is difficult without cardiopulmonary bypass. Cerebral air embolism is even more difficult to diagnose prospectively than cardiac air insults.l':" All patients with suspected embolism should have their retinal arteries carefully examined" during operative or resuscitative measures. Although convalescence is prolonged, complete recovery is more common after nonfatal cerebral air injury than after embolism from clots. 6 The subtle occurrence of air embolism with traumatic thoracic injuries has been documented. This clinical entity can occur even with blunt trauma. The outcome is not universally lethal if cardiac arrest and associated problems are promptly and appropriately treated. Early signs of hemoptysis or unexpected arrest in the operating room are favorable because they prompt earlier diagnosis for successful management. We sincerely thank Miss Sheila Finnigan for her help in preparing this manuscript. REFERENCES
2
3
4 5
6
Justice C, Leach J, Edwards WS: The harmful effects and treatment of coronary air embolism during open-heart surgery. Ann Thorac Surg 14:47-53, 1972 Lawrence GH, McKay HA, Sherensky RT: Effective measures in the prevention of intraoperative aeroembolus. J THORAC CARDIOVASC SURG 62:731-735, 1971 Taber RE, Maraan BM, Tomatis L: Prevention of air embolism during open-heart surgery. A study of the role of trapped air in the left ventricle. Surgery 68:685-691, 1970 Nicks R: Arterial air embolism. Thorax 22:320-326, 1967 Spencer FC, Rossi NP, Yu SC, Koepke JA: The significance of air embolism during cardiopulmonary bypass. J THORAC CARDIOVASC SURG 49:615-634, 1965 Mills NL, Ochsner JL: Massive air embolism during
7
8
9 10
11
12
13 14
15
16 17 18 19 20
21
22
23
24
25
26
cardiopulmonary bypass. Causes, prevention, and management. J THORAC CARDIOVASC SURG 80:708-717, 1980 Graham JM, Beall AC, Mattox KL, Vaughan GD: Systemic air embolism following penetrating trauma to the lung. Chest 72:449-454, 1977 Donato AT, Arciniegas E, Lam CR: Fatal air embolism during thoracotomy for gunshot injury to the lung. J THORAC CARDIOVASC SURG 69:827-829, 1975 Ellis GR, Brown JR: Massive air embolism due to gunshot wound. JAMA 189:177-179,1964 Waldo WJ, Harlaftis NN, Symbas PN: Systemic air embolism. Does it occur after experimental penetrating lung injury? J THORAC CARDIOVASC SURG 71:96-101, 1976 Meier GH, Wood WJ, Symbas PN: Systemic air embolization from penetrating lung injury. Ann Thorac Surg 27:161-166,1979 Meier GH, Symbas PN: Systemic air embolization. Factors involved in its production following penetrating lung injury. Am Surg 12:765-771, 1978 Thomas AN: Air embolism following penetrating lung injuries. J THORAC CARDIOVASC SURG 66:533-540, 1973 Thomas AN, Stephens BG: Air embolism. A cause of morbidity and death after penetrating chest trauma. J Trauma 14:633-638, 1974 Erben J, Nadvornik F: The quantitative demonstration of air embolism in certain cases of fatal trauma. J Forensic Med 10:45-50, 1963 Taylor JD: Post-mortem diagnosis of air embolism by radiography. Br Med J 4:890-893, 1952 Durant TM, Long J, Oppenheimer MJ: Pulmonary (venous) air embolism. Am Heart J 33:269-281, 1947 Rukstinat G: Experimental air embolism of the coronary arteries. JAMA 96:26-28, 1931 Rukstinat GJ, LeCount ER: Air in the coronary arteries. JAMA 91:1776-1779, 1929 Geoghegan T, Lam CR: The mechanism of death from intracardiac air and its reversibility. Ann Surg 138:351359, 1933 James TN, Geoghegan T, Lam CR: Electrocardiographic manifestations of air in the coronary arteries of dying and resuscitated hearts. Am Heart J 46:215-228, 1953 Eguchi S, Bosher LH: Myocardial dysfunction resulting from coronary air embolism. Surgery 51:103-111, 1962 Goldfarb D, Bahnson HT: Early and late effects on the heart of small amounts of air in the coronary circulation. J THORAC CARDIOVASC SURG 46:368-378, 1963 Lenaghan R, Silva YJ, Walt AJ: Hemodynamic alterations associated with expansion rupture of the lung. Arch Surg 99:339-343, 1969 Chiu CJ, Golding MR, Linder JB, Fries CC: Pulmonary venous air embolism. A hemodynamic reappraisal. Surgery 61:816-819,1967 Tomatis L, Nemiroff M, Riahi M, Visser J, et al: Massive arterial air embolism due to rupture of pulsatile assist
Volume 85
Air embolism
Number 5 May, 1983
27
28
29
30
3I
32
33
device. Successful treatment in the hyperbaric chamber. Ann Thorac Surg 32:604-608, 1981 Steward D, Williams WG, Freedom R: Hypothermia in conjunction with hyperbaric oxygenation in the treatment of massive air embolism during cardipulmonary bypass. Ann Thorac Surg 24:591-593, 1977 Roe BB: Prevention of air embolism with intravascular carbon dioxide washout. J THORAC CARDIOVASC SURG 71:628-630, 1976 Winter PM, Alvis HJ, Gage AA: Hyperbaric treatment of cerebral air embolism during cardiopulmonary bypass. JAMA 215:1786-1788,1971 Spampinato N, Stassano P, Gagliardi C, Tufano R, Iorio D: Massive air embolism during cardiopulmonary bypass. Successful treatment with immediate hypothermia and circulatory support. Ann Thorac Surg 32:602, 198I Bedell AJ: Clinical differentiation of emboli in the retinal arteries from endarteritis. Arch Ophthalmol 34:311-317, 1945 Gomes OM, Pereira SN, Castagna RC, Bittencourt D, Amaral RVG, Zerbini EJ: The importance of the different sites of air injection in the tolerance of arterial air embolism. J THORAC CARDIOVASC SURG 65:563-568, 1973 Gillen HW: Symptomatology of cerebral gas embolism. Neurology 18:507-512, 1968
667
ventilatory pressure becomes excessively high, over 25 to 30 mm Hg, the pressure gradient is reversed and air may be pushed into the pulmonary vein. Such air can get into the coronary arteries and can also go to the brain. A number of years ago I described a patient who suffered a so-called "pleural shock" during thoracentesis, with signs of cerebral air embolism. However, at that time a prominent surgeon who did not believe in the entity of traumatic air embolism dismissed this as just a fainting episode.' Now we have better technology to substantiate our suspicions. This particular young man had a gunshot wound of the left lung and had a sudden cardiac arrest during thoracotomy. The cardiac arrest occurred in a stable patient when the lung was being reexpanded with high airway pressure. Postoperatively, cognitive dysphasia developed and a computed axial tomographic (CAT) scan of the brain showed localized temporoparietal infarction. The lesson, of course, is not to use excessively high ventilatory pressure in such cases. Other measures described by the authors should also be useful. REFERENCE Symbas PN: Reply to Chiu CJ: Air embolism in penetrating lung injury. J THORAC CARDIOVASC SURG 72:815-816, 1976 DR. AGUSTIN ARBULU Detroit. Mich.
Discussion DR. RAY CHU-JENG CHIU Montreal. Quebec. Canada
I would like to make a few additional comments on the mechanism of this condition, since understanding the mechanism may help in maintaining a high index of suspicion. This is important, because the complications of traumatic air embolism are still underdiagnosed and the confirmation of this diagnosis is often very difficult. As mentioned by the authors, a small amount of hemoptysis preceding sudden cardiovascular collapse is not uncommon in traumatic air embolism. Such hemoptysis, in addition to being a telltale sign, may be an important mechanism in causing air embolism. In a conscious patient, blood entering the airway can induce vigorous cough reflex, build up a high intrathoracic pressure, and push the air into the left side of the heart through the bronchopulmonary venous fistula caused by trauma. The role of positive-pressureventilation in causing traumatic' air embolism was elucidated in our study, in which the pulmonary venous pressure and lower airway pressure were simultaneous monitored. The pulmonary venous pressure is higher in when the patient's chest is open and rises with increased ventilatory pressure, because of the transmission of the pressure. Under such circumstances, the blood runs into the bronchus through the venobronchial communication rather than in the reverse direction. Perhaps this is why we do not see air embolism every day. However, when the positive
In the early I970s, as Dr. Richard Lenaghan in our department of surgery was working in the experimental laboratory, he developed a model in dogs to produce air embolism in the venous and arterial systems. The model was very simple. As Dr. Chiu has expressed, increasing the ventilatory pressure to more than 30 mm Hg consistently produced this lesion in dogs. A review of the medical literature reveals that this observation was made early in the 19305 by Dr. Macklin and is well documented and published. I have a question for Dr. Vee. In some of these patients could hyperbaric chambers offer some help? DR. YEE (Closing) I would like to thank Drs. Chiu and Arbulu for their kind remarks. In reply to Dr. Chiu's statements, specific clinical signs of air embolism, such as hemoptysis or unexpected arrest after positive-pressure ventilation, occurred in only 36% of our patients. Hemoptysis occurred in only 16%, a very small number, and occurred only in association with the very proximal injuries. This fact highlighted the insidious nature of this clinical entity. For the other 20%, the early clue of unexpected arrest might be overlooked as simply a manifestation of the associated injuries. Our treatment for arrest at San Francisco General Hospital has been an emergency resuscitative thoracotomy, which permitted an earlier diagnosis with a direct cardiac examination. Systemic air embolism should be
668
Vee, Verrier, Thomas
suspected when cerebral or cardiac dysfunction is out of proportion to the clinical setting. To answer Dr. Arbulu's question, adjunctive therapy for cerebral protection might be helpful, but there are several practical and technical limitations. Preoperative manifestations were commonly overlooked, since they were not frequently localizing. Most of the diagnoses in our trauma patients were made later than would have been the case in patients having cardiopulmonary bypass, when the passage of air is noted during the procedure. Simultaneous cerebral CAT scans during chest or abdominal studies might be helpful, as illustrated by Dr. Chiu's case.
The Journal of Thoracic and Cardiovascular Surgery
Usage of hypothermia, hyperbaric oxygenation or carbon dioxide, and barbiturates has been of limited help in the injured patient when the control of hemorrhage, myocardial resuscitation, and immediate intensive care are mandatory. One anecdotal example, however, was a 25-year-old man who had both an overdose of barbiturates and cardiac arrest from a chest stab wound. Circumstantial hypothermia (90 0 F) from transfusion of 11 L of blood and fluids probably helped his complete neurologic recovery.
Bound volumes available to subscribers Bound volumes of THE JOURNAL OF THORAOC AND CARDIOVASCULAR SURGERY are available to subscribers (only) for the 1983 issues from the Publisher, at a cost of $58.50 ($73.00 international) for Vol. 85 (January-June) and Vol. 86 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertising is removed. Copies are shipped within 30 days after publication of the last issue of the volume. The binding is durable buckram with the JOURNAL name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mr. Deans Lynch at The C. V. Mosby Company, 11830 Westline Industrial Drive, St. Louis, Missouri 63141, USA. Subscriptions must be In force to qua6fy. Bound "oIumes an not a"ailable In pilla of a regular JOURNAL subscription.