Journal of Neurosurgical Anesthesiology Vol. 8, No. 1, pp. 30-34 : ; ; © 1996 Lippincott-Raven Publishers, Philadelphia Case Report Cerebral Ischemia After Venous Air Embolism in the Absence of Intracardiac Defects Concezione Tommasino, Roberto Rizzardi, Luigi Beretta, Marco Venturino, and Susanna Piccoli University of Milano, Institute of Anesthesia and Intensive Care, Scientific Institute H San Raffaele, Milano, Italy Summary: Cerebral air embolism occurred in a patient undergoing posterior fossa surgery performed in the sitting position for acoustic neuroma removal. The patient experienced two episodes of venous air embolism, as evidenced by precordial Doppler, end-tidal carbon dioxide reduction, and oxygen desatura- tion. In both cases, air was aspirated from the central venous catheter; during the second episode there was arterial hypotension and electrocardiogram changes, and air bubbles were visualized in the cerebellar arteries. The patient did not regain consciousness after surgery and developed early tonic-clonic convulsions and electroencephalogram status epilepticus, which was treated with barbiturate coma. Intracardiac septal defects were not detected by trans- esophageal echocardiography, and computerized tomography of the brain demonstrated multifocal discrete ischemic areas in the cerebral hemispheres. The patient died 6 days after surgery without having regained consciousness. This case appears to represent the occurrence of transpulmonary passage of venous air embolism. Key Words: Complication—Cerebral ischemia— Embolism—Venous air embolism. One of the hazards of neurosurgical procedures in the sitting position is venous air embolism (VAE) (1). When the embolus is large, venous air causes pulmonary insufficiency and/or cardiovascular al- terations. In only very few cases do systemic em- bolization and organ injury occur, usually due to the presence of a patent foramen ovale or to an atrial septal defect (paradoxic air embolism) (2-5). We report a case of multifocal cerebral ischemia subsequent to VAE that occurred during an upright neurosurgical procedure in a patient in whom no Address correspondence and reprint requests to Dr. Concezione Tommasino, Anesthesiology, H San Raffaele, Via Olgettina, 60, 20132 Milan, Italy. 30 evidence of any intracardiac defect could be de- tected. CASE REPORT A 63-year-old white man was scheduled for elec- tive surgical removal of a left acoustic neuroma in the sitting position. Neurological examination re- vealed left hypoacusia and hypoesthesia in the terri- tory of the left ophthalmic and maxillary nerves. Preoperative biochemical and hematological pro- files, electrocardiogram, and chest x-ray films were all unremarkable. Before induction of anesthesia, a central venous catheter was inserted via the right internal jugular ISCHEMIA AFTER VAE WITHOUT CARDIAC DEFECT 31 yein for fluid administration and air removal. The position of the tip within the right atrium was con- firmed, in the final sitting position, on chest x-ray film. Intraoperative monitoring included electrocar- diogram (ECG), blood pressure monitoring via in- traarterial catheter zeroed at the external auditory meatus and noninvasive sphygmomanometer, pre- cordial Doppler (TC2-64B, EME, Germany) (6), end-tidal carbon dioxide (ETCO,), SpO,, end-tidal volatile anesthetic agent, arterial blood gases (ABGs), rectal temperature, and urine output. An- esthesia was induced with fentanyl (0.01 mg) and thiopental (4 mg/kg) and was maintained with iso- flurane (0.4-1% end tidal) in oxygen:air (FiO, = 0.45), and boluses administration of fentanyl (0.75 mg, total dosage). Muscle relaxation was induced and maintained with pancuronium bromide. The pe- ripheral SpO, was 98% or higher, and the ETCO, was 30-32 mm Hg. Thirty minutes after induction of anesthesia, basal ABG monitoring showed pH at 7.48, P,co, at 34 mm Hg, and P,o, at 218 mm Hg. During surgery there were two episodes of VAE. The first occurred during the muscular dissection, and it was detected by a change in the precordial Doppler, ETCO, reduction from 30 to 20 mm Hg, oxygen desaturation (SpO, from 100 to 92%), and moderate systemic hypotension; systolic blood pressure (BP) decreased from 120 to 76 mm Hg (lowest recorded value). This VAE resolved promptly, and there were no changes in the ECG during this episode. Air was aspirated through the atrial catheter (~7 cc air), and the subsequent ABG showed pH at 7.39, P,co, at 40 mm Hg, and P,o, at 172 mm Hg (FiO, = 1). The second episode of VAE occurred after acci- dental opening of the transverse sinus; it was con- firmed by the precordial Doppler, ETCO, reduction from 30 to 15 mm Hg, oxygen desaturation (from 98 to 90%), and severe hypotension; systolic BP de- creased from 110 to a low of 54 mm Hg. Phenyleph- rine was administered, and systolic BP remained <60 mm Hg for <10 min. Air was aspirated through the atrial catheter (~15 cc air), and the subsequent ABG showed pH at 7.34, P,co, at 46 mm Hg, and P,0, at 265 (FiO, = 1). The arterial BP gradually returned to the level prior to the VAE episode. The ECG showed alterations of the S-T segment that lasted almost 30 min. After the second VAE epi- sode (as well as in the 24 h following surgery) lab- oratory tests performed for serum electrolytes (Na*, K*, Ca?*) and CPK were in the normal range. During the second VAE episode, air bubbles were visualized in the cerebellar arteries. In the recovery room, the patient did not awaken from anesthesia, and he was transferred to our neu- rosurgical intensive care unit (NICU). The patient was comatose, his pupils were small and equally reactive to light, and neurologic examination re- vealed abnormal flexion to painful stimuli. The P,o, was 81 mm Hg at an FiO, of 0.6, and the patient’s lungs were ventilated in the assisted mode with a positive end-expiratory pressure (PEEP) level of 7 cm H,0. Computed tomography (CT) brain scan performed soon after the surgical procedure re- vealed pneumocephalus. On the same day, the pa- tient underwent transesophageal bidimensional echocardiography (2-D TEE, Sonos 1000; Hewlett Packard) with contrast; 10 ml of agitated saline with microbubbles was injected through the right atrial catheter during continuous 2-D TEE visualization of the cardiac chambers. The examinations were performed at zero end-expiratory pressure and twice at a PEEP level of 10 cm H,0. The injection was performed just before the release of PEEP. This study did not demonstrate a patent foramen ovale, interatrial septal defects, or right-to-left shunt on color Doppler imaging. The remainder of the patient’s hospital course was characterized by tonic-clonic convulsions with electroencephalogram (EEG) status epilepticus, which occurred the day of surgery. Control of the seizure activity was achieved only after general an- esthetic levels of barbiturate (barbiturate coma) were achieved. Sensory evoked potentials per- formed | day after surgery revealed normal brain stem auditory evoked potentials (BAEPs) on the right side and absent BAEPs on the left side (left acoustic neuroma). Somatosensory evoked poten- tials with stimulation of the median nerve at the wrist revealed no peripheral or cortical response to right-sided stimulation. There was a small nonre- producible cortical response to stimulation on the left. A CT brain scan performed 3 days after surgery demonstrated bilateral, multifocal ischemic areas of recent origin and of variable sizes at the cortical and subcortical levels (Fig. 1). The patient died 6 days after the neurosurgical procedure without having regained consciousness. Autopsy was not per- formed. Journal of Neurosurgical Anesthesiology, Vol, 8, No. 1, 1996 32 C. TOMMASINO ET AL. FIG. 1. CT brain scan taken 3 days after the surgical procedure shows bilateral multifocal ischemic areas. DISCUSSION Cerebral emboli subsequent to venous air embo- lism in the sitting position are relatively rare (7), and in the majority of the cases, the cerebral emboli reported in the literature are of paradoxic origin (2- 5). In patients with no obvious congenital cardiac defect, the passage of embolized air from right to left circulation generally occurs through a patent foramen ovale; in such cases, the estimated risk of paradoxic air embolism in the sitting position is 6- 12% (4,7,8). Detection of intracardiac septal defects is possible by transesophageal two-dimensional contrast echocardiography, which can also be help- ful in selecting the neurosurgical operative position with a lower incidence of venous air embolism (8,9). The patient described in this case report experi- enced a paradoxic air embolus since air was seen in the cerebellar arteries, but he had no evidence of Journal of Neurosurgical Anesthesiology, Vol. 8, No. 1, 1996 atrial septal defect, patency of the foramen ovale, or right-to-left shunt as evaluated by 2-D TEE. Al- though we cannot exclude the presence of a very small patent foramen ovale, which can occur when right-sided circulatory pressure increases (8,10), the 2-D TEE study, performed under conditions meant to simulate VAE occurrence (10), did not evidence the presence of right-to-left shunt. In the absence of intracardiac defects permitting right-to-left shunting, the other possible explanation for VAE is that the air in the venous system crossed to the systemic circulation through the pulmonary vascular bed, via physiologic arteriovenous anasto- moses or through the filtering capability of the lung (11-13). Under normal conditions, the lungs are an impor- tant physiologic filter for venous air emboli. Some factors, however, might compromise the capability of the lungs to trap venous bubbles, including large volumes of venous air, the use of vasodilators, and anesthetic agents (13-18). When the total volume of venous air infused is large, the filter capability of the lung is impaired and transpulmonary passage of venous air emboli occurs (13,14,17). Butler and Hills demonstrated in dogs that when air infusion is increased to 0.4 ml/kg/min, transpul- monary passage of air emboli occur in 71% of ani- mals (13). However, the threshold doses for the transpulmonary passage of bolus injection of air into the pulmonary artery in the dog can be as small as 0.05 mi/kg (15,18). The bubble breakthrough phe- nomenon may involve a degree of species specific- ity (17), and although the occurrence of cerebral air embolism via the pulmonary vasculature is contro- versial (19), cerebral air embolism has been re- ported in patients with no intracardiac defects at autopsy (20-22). Direct documentation of transpul- monary passage of air in humans is lacking; how- ever, a recent case report demonstrated that this phenomenon can indeed occur. In a patient oper- ated on in the sitting position and with no evidence of intracardiac right-to-left shunt, intraoperative 2D-TEE clearly documented the occurrence of transpulmonary passage of venous air embolism (23). Pulmonary vasodilators can enhance transpulmo- nary passage of air (13), and the filtering capacity of the lungs is impaired by volatile anesthetics (16). In a dog model during halothane-induced anesthesia, the threshold for transpulmonary passage of air was ISCHEMIA AFTER VAE WITHOUT CARDIAC DEFECT 33 lower than with i.v. anesthetic agents (15,18). The mechanisms by which anesthetics influence the threshold for transpulmonary passage of air are un- known, but passage may be mediated through ef- fects on pulmonary blood flow at the level of arte- rioles, shunt vessels, or capillaries. Volatile anes- thetic agents have a pulmonary vasodilating effect (24), and the possibility that they can enhance the transpulmonary passage of air in humans is a matter of interest (25), especially because this phenome- non has been reported in experimental settings (15,18). In our patient, the use of isoflurane and/or the episodes of intraoperative hypotension could have played a role in enhancing transpulmonary passage of venous air. The ischemic areas on the CT brain scan were not mainly localized at the level of watershed zones, and the intraoperative short-lasting systemic hy- potension subsequent to the VAE episodes seems unlikely to be totally responsible for the occurrence of the cerebral ischemia, although it could have worsened it. With the patient in the sitting position, air crossing to the systemic arterial circulation mi- grates up to the carotid and vertebral arteries and to the cerebral-cerebellar hemispheres; in our case this was confirmed by visualization of air bubbles in the cerebellar arteries. The air bubbles in the arter- ies produce an immediate but transient block, which is responsible for the ischemic damage. The localization of the air emboli and the extent of the cerebral ischemia determine the degree of neurolog- ical deficit. Distribution of cerebral air emboli is more common at the level of supratentorial struc- tures, and the neurologic condition of the patient (postoperative coma and status epilepticus) and the tomographic appearance of the ischemic areas in the brain likely confirm such a distribution. The passage of air from the venous to the arterial circulation, regardless of the pathway involved, must be considered a serious intraoperative event. Clinicians should attempt all necessary maneuvers to reduce the complications from the VAE. Intra- operatively, the anesthesiologist should support the cardiovascular system so that the ischemic injury can be reduced and/or avoided, and the surgeon should flood the field with saline to submerge the area of air entry. Jugular pressure of the anterior neck can be applied, assuring carotid flow. In the postoperative period in the presence of focal or gen- eralized neurologic deficits, the patient should be considered at risk of cerebral ischemia and ade- quately treated. In conclusion, the present case illustrates the oc- currence of cerebral emboli subsequent to severe episodes of venous air embolism in a patient oper- ated on in the sitting position. The patient did not have intracardiac defects on bidimensional trans- esophageal contrast echocardiography. He failed to emerge from anesthesia. In the absence of right-to- left shunting at the atrial level, we speculate that the cerebral emboli were due to transpulmonary move- ment of VAE into the left side of the heart. REFERENCES 1. Michenfelder JD, Miller RH, Gronert GA. Evaluation of an ultrasonic device (Doppler) for the diagnosis of venous air embolism. Anesthesiology 1972;36: 164-7. 2. Buckland RW, Manners JM. Venous air embolism during neurosurgery: a comparison of various methods of detection in man. Anaesthesia 1976;31:633—43. 3. Root B, Levy MN, Pollack S, Mortimer L, Kalindi P. Gas embolism death after laparoscopy delayed by “trapping” in portal circulation. Anesth Analg 1978:57:232-7. 4. Gronert GA, Messick JM, Cucchiara RF, Michenfelder JA. Paradoxical air embolism from a patent foramen ovale. An- esthesiolagy 1979;50:548-9. 5. Liu S, Holley HH, Stulberg SD, Cohen B, Failure to awaken after general anaesthesia secondary to paradoxical venous embolus. Can J Anaesth 1991;38:335~7. 6. Tommasino C, Beretta L, Dell'Acqua A, Rizzardi R, Ferrari da Passano C. Doppler transcranico per il monitoraggio dell’embolia gassosa negli interventi neurochirurgici in po- sizione seduta. Minerva Anestesiol 1992;58:1311-4. 7. Matjasko J, Petrozza P, Cohen M, Steinberg P. Anesthesia and surgery in the seated position: analysis of 554 cases. Neurosurgery 1985;17:695-702. 8. Cucchiara RF, Seward JB, Nishimura RA, Nugent M, Faust RJ. Identification of patent foramen ovale during sitting po- sition craniotomy by transesophageal echocardiography with positive airway pressure. Anesthesiology 1985;63: 107-9. 9. Guggiari M, Lechat Ph, Garen-Colonne C, Fusciardi J, Viars P. Early detection of patent foramen ovale by two- dimensional contrast echocardiography for prevention of. paradoxical air embolism during sitting position. Anesth An- alg 1988;67:192-4. 10. Cucchiara RF, Nishimura RA, Black S. Failure of preoper- ative echo testing to prevent paradoxical air embolism: re- port of two cases. Anesthesiology 1989;71:604-7. 11. Emerson LV, Hempleman HV, Lentle RG. The passage of gaseous emboli through the pulmonary circulation. Resp Physiol 1967;3:213-9. 12. Baskin SE, Wozniak RF. Air embolism occurring during he- modialysis. N Engl J Med 1975;293:184-5. 13. Butler BD, Hills BA. The lung as a filter for microbubbles. J App! Physiol 1979:47:537-43. 14. Butler BD, Hills BA. Transpulmonary passage of venous air emboli. J Appl Physiol 1985;59:543-7. 15. Yahagi N, Furuya H. The effects of halothane and pento- Journal of Neurosurgical Anesthesiology, Vol. 8, No. 1, 1996 34 16. 17. 18. 19. 20. C. TOMMASINO ET AL. barbital on the threshold of transpulmonary passage of ve- nous air emboli in dogs. Anesthesiology 1987;67:905-9. Katz J, Leiman BC, Butler BD. Effects of inhalational anaesthetics on filtration of venous gas emboli by the pul- monary vasculature. Br J Anaesth 1988;61:200-5. Vik A, Brubakk AO, Hennessy TR, Jenssen BM, Ekker M, Slordahl SA. Venous air embolism in swine: transport of gas bubbles through the pulmonary circulation. J App! Physiol 1990;69:237-44. Yahagi N, Furuya H, Sai Y, Amakata Y. Effect of halo- thane, fentanyl, and ketamine on the threshold for transpul- monary passage of venous air emboli in dogs. Anesth Analg 1992;75:720-3. Black S, Cucchiara RF, Nishimura RA, Michenfelder JD. Parameters affecting the occurrence of paradoxical air em- bolism. Anesthesiology 1989;71:235-41. Marquez J, Sladen A, Gendell H, Boehnke M, Mendelow H. Journal of Neurosurgical Anesthesiology, Vol. 8, No. 1, 1996 21. 22. 23. 24. 25, Paradoxical cerebral air embolism without an intracardiac septal defect. J Neurosurg 1981;55:997—1000. Ahmat KP, Riley RH, Sims C, House AK, Cooke OT. Fatal air embolism following anesthesia for insertion of a perito- neovenous shunt. Anesthesiology 1989;70:702—-4. Albin MS, Ritter RR, Pruett CE, Kalff K. Venous air em- bolism during lumbar laminectomy in the prone position: report of three cases. Anesth Analg 1991;73:346-9. Bedell EA, Berge Kh, Losasso TJ. Paradoxic air embolism during venous air embolism: transesophageal echocardio- graphic evidence of transpulmonary air passage. Anesthesi- ology 1994;80:947-50. Mathers J, Benumof J, Wahrenbrock EA. General anesthet- ics and regional hypoxic pulmonary vasoconstriction. Anes- thesiology 1977;46:111-4. Drummond JC, Of venous air embolism, aminophylline, and volatile agents [Letter]. Anesthesiology 1984;60:389.