The time to change from pre-VAE baseline to minimum end-expired CO2 (EtCO2) for each VAE episode was 11±10 minutes; range, 2 to 36 minutes in the setting of extreme VAE. EtCO2 reduction with extreme VAE was 11±8 mm Hg; range, 2 to 28 mm Hg. The relative reduction in EtCO2 was 38.4%±20.4%; range, 15% to 85%. The time to change from baseline to minimum MAP was 9.8±8.6 minutes; range, 2 to 32 minutes. MAP reduction magnitude with extreme VAE was 46±26 mm Hg; range, 10 to 97 mm Hg. The relative MAP reduction was 46.3%±26.6%; range, 10% to 97%. SpO2 minimum was 95%±2%; range, 91% to 97% and mean SpO2 during the entire anesthetics was 98%±2%; range, 95% to 99%. In all but one patient (patient G), EtCO2 recovered completely to baseline. Patient G had recovered EtCO2 from low measures of 25 to 33 from a baseline 37 mm Hg by the end of the case. The intraoperative and perioperative details of each extreme VAE event are described below and within Table 2.
We provide 3 illustrative cases:
The surgical team requested the seated position in this morbidly obese (164 kg, body mass index=53.6 kg/m2) 50 years male patient to optimize suboccipital exposure, respiratory mechanics and intracranial venous drainage. Following induction of anesthesia, an arterial catheter, TEE, LaCVC, and precordial Doppler were placed. Agitated saline injection demonstrated a patent foramen ovale (PFO) on TEE with positive end-expiratory pressure release. The neurosurgeon was informed and indicated a continued preference for the seated position. The anesthetic was maintained with isoflurane and 50% N2O. During cerebellar biopsy, VAE was detected on precordial Doppler and confirmed by TEE (Fig. 1). EtCO2 markedly decreased minutes before a decrease in MAP. N2O was discontinued and surgery continued without immediate identification of the source. Small amounts of air on TEE were initially confined to right heart chambers. The anesthesia team supported the surgeon’s request to continue “a bit longer” to remove a tissue sample from the fourth ventricular floor. Six minutes later, TEE demonstrated intra-aortic air as the initial VAE episode continued. Shortly thereafter, larger amounts of right-sided and intra-aortic air were observed. The patient became hypotensive, was treated with ephedrine, and then experienced a 5 to 10 seconds episode of ventricular tachycardia. A 30 to 40 mL of frothy blood was aspirated over several minutes from the LaCVC. The wound was covered with moist sponges and the patient was rapidly lowered into the supine and then left lateral decubitus position. Right-sided and intra-aortic air resolved over several minutes and the blood pressure recovered after 2 additional doses of ephedrine. Left ventricular contractility remained dynamic and symmetric. No further ectopy was observed. After a second sterile preparation and redraping, the surgeons inspected the biopsy site to ensure hemostasis and then closed the wound. The patient awakened with an unchanged gross sensorimotor neurological examination. He had no significant troponinemia 6 hours postoperatively. His perioperative course was remarkable for cerebrospinal fluid hypotension requiring several surgeries to attain an adequate dural graft closure.
This 27 years female patient suffered recurrent cerebellar astrocytoma refractory to chemotherapy and previous tumor resection. Following induction of anesthesia, arterial access was obtained; however, after 2 unsuccessful attempts at LaCVC placement, the case proceeded without central venous access despite a positive PFO test. After enucleation of the tumor mass, VAE was observed by precordial Doppler and TEE without paradoxical air embolism (PAE). EtCO2 precipitously decreased before hemodynamic instability (Fig. 2). The surgeons were able to stop the VAE with electrocautery, irrigation, and moistened Surgicel (Ethicon, Somerville, New Jersey) packing. Hypotension was treated with phenylephrine. Several minutes later, while attempting to close the overlying dura mater, VAE was again detected with air present in both the right and left-sided heart chambers. Progressive cardiovascular instability manifested as hypotension, new S-T changes on the ECG, and new second degree atrioventricular blockade. As soon as paradoxical VAE was evident, the patient was placed supine and then into the left lateral decubitus position with almost immediate resolution of hemodynamic instability, ECG changes, and VAE. After surgical closure, the patient was extubated in the OR with an unchanged gross neurological examination. Postoperatively, she remained hemodynamically stable with a stable neurological examination. An ECG performed several hours later showed normal sinus rhythm and with no ischemic changes.
A 60 year old male with Parkinson’s disease underwent a DBS placement with dexmedetomidine infusion for sedation in the seated position. His medical history was complicated by coronary artery disease and mild pulmonary hypertension. Precordial Doppler was placed for VAE monitoring. After burr hole and durotomy, the patient began coughing and complaining of chest pain with simultaneous detection of air on precordial Doppler and acutely decreased EtCO2 via nasal cannula (Fig. 3). The patient was rapidly placed in the supine position and hypotension was treated with ephedrine. The chest pain resolved and an emergent ECG displayed unchanged ST-T waves and normal sinus rhythm. After bone waxing, cessation of precordial Doppler VAE and continuing stable blood pressure, the patient was again placed in the seated position and the procedure was completed. Postoperative, serum troponin concentrations were not suggestive of myocardial ischemia. Aside from a reduction in tremor, his neurological status was unchanged through his hospital stay.
There was 1 patient death in this series, patient C, but the neurosurgical team attributed this to catastrophic disruption of cerebral venous vasculature. There was a VAE, but the fundamental problem was uncontrolled hemorrhage. The wound was packed and closed over open dura and the patient was admitted to the intensive care unit. Arterial blood gases and postoperative echocardiogram studies were normal.
In this case series of patients who sustained extreme VAE during seated neurosurgical procedures, emergent position change was motivated by progressive hemodynamic instability, evolving malignant arrhythmias, persistent or PAE or significant clinical change in respiratory or neurological symptoms in sedated patients. Similar to prior literature in smaller series,2,16 we show that extreme VAE is uncommon, occurring in ~1% of patients having intracranial neurosurgical procedures in the seated position. VAE occurred most commonly during posterior fossa craniotomy, but there was confidence overlap with the incidence in DBS implantation cases. Furthermore, despite hemodynamic instability or evidence of intracardiac PAE in those with extreme VAE, we report no new or unexpected postoperative neurological deficits and good overall postoperative outcomes without persistent cardiac dysfunction.
VAE-related hemodynamic instability and gas exchange abnormalities depend upon the volume of air entrained and the rate of air entrainment—neither of which are quantitatively measurable or well understood in surgical patients. When the gas embolic load is sufficiently high in the pulmonary circulation, increased alveolar dead space and reduced right cardiac output manifest as decreased EtCO2.17 This was observed in all but 1 patient with extreme VAE in this series. The reduction in EtCO2 was remarkably short-lived following position change, on the order of 20 to 30 minutes, the first time this has been reported in human subjects and is consistent with previous animal studies.18,19 Full recovery to pre-VAE EtCO2 occurred in all but one patient, in whom it substantially recovered. This rapid CO2 exchange recovery in patients after extreme VAE parallels microvascular bubble resorption kinetics, but may also reflect resolution of VAE-induced pulmonary arterial vasoconstriction.20–24 None of the patients in this series were hypoxemic either during or after the extreme VAE, indicating that oxygenation and O2 exchange remained adequate. The possibility that there was an increased alveolar-arterial O2 gradient was not assessed, but in the only 2 patients with an arterial blood gas shortly after VAE, this gradient was not large.
MAP reductions occurred after reductions in EtCO2 in all but patients C and F. Similar to VAE-related EtCO2 reductions, MAP reductions were rapid in onset and similar in magnitude. This is presumably on the basis of decreased systemic cardiac output due to decreased right ventricular cardiac output with pulmonary vascular obstruction and hypertension. EtCO2 reduction preceded MAP reduction in all cases with mechanical ventilation except patient C, wherein hypovolemic shock was occurring. Capnometric EtCO2 reduction occurred in 1 DBS case (Fig. 3), but did not significantly change in the other. It is unclear why EtCO2 did not decrease in the latter patient, but patients undergoing DBS surgery were awake and spontaneously ventilating with capnometry obtained from nasal cannula. Reductions in MAP rapidly responded to pressor administration often recovering before EtCO2 (Figs. 1–3). Before and following position change, these patients with extreme VAE were readily resuscitated.
Position change with extreme VAE is an undesired event for patients undergoing neurosurgical procedures in the seated position. Position changes after extreme VAE ranged from supine to lateral decubitus to prone, with surgical logistics as the primary determinant of final position. The physiological rationale for position change in these cases is to collapse the pressure gradient between the VAE source and the central venous circulation sufficiently to terminate VAE and its sequelae. This maneuver was successful in terminating VAE in all patients with extreme VAE. Concerns associated with intraoperative position change include increased infectious risk, unrecognized evolving neurological deficits and interruption and postponement of surgery. There were no perioperative infections observed in patients with extreme VAE suggesting that intraoperative position change does not necessarily increase this risk. All patients but one (patient C) recovered from the VAE-induced hemodynamic and EtCO2 reductions following position change. Hemodynamic recovery was sufficient after position change such that the majority of cases were continued to completion either in a new position or resumption of the seated position with patients undergoing DBS placement without further incident. Surgery was terminated and rescheduled for a later date in 3 of the 8 patients and the subsequent surgeries were not performed in the seated position. The incidence of case postponement is <0.5%, for all intracranial surgeries and < 0.2% for all seated cases in the entire series of surgeries performed in the seated position, the latter less than half that reported in another series.2 This may be due to the satisfactory resolution of hemodynamic effects and the agreeability of the neurosurgical and anesthesia care teams to continue with the procedure on this basis.
VAE may cross intracardiac and intrapulmonary shunts,9 resulting in arterial air embolism. This exposes the coronary and cerebral arterial circulation to obstructive air embolism with resultant myocardial ischemia, arrhythmia, and possible cerebral ischemia and stroke. There were 2 instances of transient ST-segment changes and ventricular arrhythmia in patients with air appearing in the aortic root and in both instances this rapidly resolved. The absence of persistent myocardial dysfunction, ECG changes, arrhythmia or troponinemia indicates that air-related obstructive effects in the coronary arterial circulation were transient, presumably due to resorption before permanent damage occurred. Despite TEE demonstrated intra-aortic air in 2 cases, there were no neurological deficits not predicted on the basis of the preoperative examination and the surgical procedure. Since no patient in this series had postoperative magnetic resonance imaging, it is possible that small, ineloquent cerebral ischemic events occurred. Other systemic manifestations previously associated with VAE, including pulmonary edema and coagulopathy were not observed in these patients.25,26 Several of these cases were undertaken in the seated position despite a demonstrated PFO and there were no persistent adverse effects associate with extreme VAE in these patients. This result is similar to that reported in a series of semisitting cases with precordial Doppler and TEE monitoring.16 These results suggest PFO need not disqualify a patient from undergoing a procedure in the seated/semiseated position with appropriate monitoring. It is the practice at this institution to discuss a PFO with the neurosurgical team. This generally does not result in a change in preference for the seated position.
Removal of air from the right atrium with a multiorificed CVC was initially recommended as essential in the treatment of VAE.27–31 In contrast, in only one of the extreme VAE cases reported here was there any appreciable air aspirated from the LaCVC. Even in this instance, it was unclear whether the aspiration of air, the administration of pressors or temporary occlusion of the VAE site was the primary factor improving hemodynamics. The LaCVC tip was routinely observed on TEE and was at the recommended cavoatrial junction.29,32 The effectiveness of routine CVC placement in patients in anticipation of even extreme VAE in the seated position is, at best, questionable. This probably reflects the short residence time and small volume of entrained air bubbles at the cavoatrial junction en route to the pulmonary circulation, although this has not been documented in patients in the seated position.
Nitrous oxide was used in several of the cases, including those with a PFO. N2O decreases the LD50 for VAE in dogs33 and expands bubbles in fluid saturated with the gas.34 However, in a study in seated neurosurgical patients, 50% N2O did not increase the severity of VAE-related hemodynamic changes or have any impact on neurological outcome.10 In view of this, N2O use need not be avoided on this basis in seated neurosurgical cases. In the present series, the severity of the VAE most likely had more to do with surgical-anatomic issues than with N2O use as a similar number of patients experienced extreme VAE without as with N2O.
This investigation takes advantage of a busy neurosurgical practice at a single institution that continues to utilize the seated position on a routine basis. As such, variation between neurosurgeons and anesthesiologists is minimized. Moreover, electronic documentation of clinical events has been unchanged since 2002. Therefore, documentation and management of VAE are globally similar during this period. In addition, this is the largest VAE cohort ever reviewed. Our electronic database search methods were validated by previous studies in VAE at this institution over a similar period of time.3 In addition, this retrospective study relies on complications found in the electronic medical record, surgical operative report, and anesthetic charting record by a single, trained abstractor.
This is a retrospective case series, although prospective studies of this rare event are challenging to complete in a reasonable timeframe. Nonetheless, its validity is dependent on the possibility of missing clinical documentation by anesthesia and neurosurgical providers in the anesthetic chart, operative reports, and hospital summaries. We acknowledge the potential of underreporting extreme VAE and misdocumentation of VAE management. However, we found complete consistency with the automated data search and the manual search. The outcomes reported here should not be taken to suggest that these are exclusively expected outcomes. However, they do suggest that neurological or cardiopulmonary morbidity after even extreme VAE do not necessarily follow. Lastly, our outcomes represent data from a single institution with a group of neurosurgeons, neurosurgical nursing room, and technical staff, and neuroanesthesiologists with expertise in the seated position, including VAE vigilant monitoring, rapid detection and management. These results may not be translatable to institutions with less experience with this practice, a potentially major precaution against extending these results to facilities with less experience or in the development of a new practice. Familiarity with and attentive use of VAE monitoring for early detection, including TEE, precordial Doppler and capnometry, remains essential in all cases utilizing the sitting position.
In summary, the incidence of extreme VAE in this series of 1698 patients undergoing neurosurgical procedures is 0.47% and was highest (1.5%) for suboccipital craniotomies. No persistent pulmonary, cardiac or neurological adverse effects were observed. LaCVC placement and aspiration during extreme VAE played no demonstrable role in patient outcome with no appreciable air aspirated in the majority of cases. An intraoperative position change is sufficient treatment when the VAE site is not rapidly identified and may permit resumption of the procedure in the seated position, a new position or acute closure and rescheduling of the case in a level position. Extreme VAE-related hemodynamic and respiratory changes resolve over a relatively short period of time following position change from the seated to a level position. The seated position carries a risk of VAE-related position change, but this was not associated with adverse neurological or cardiovascular outcome. The frequency and consequences of such extreme VAE-related events is placed in better perspective by these results.
We would like to thank Tim Weister and colleagues at the Mayo Clinic Datamart and Dr. Darrell Schroeder, Department Biostatistics for providing the CI estimates.
1. Schafer ST, Sandalcioglu IE, Stegen B, et al. Venous air embolism
during semi-sitting craniotomy evokes thrombocytopenia. Anaesthesia. 2011;66:25–30.
2. Ganslandt O, Merkel A, Schmitt H, et al. The sitting position in neurosurgery: indications, complications and results. a single institution experience of 600 cases. Acta Neurochir (Wien). 2013;155:1887–1893.
3. Himes BT, Mallory GW, Abcejo AS, et al. Contemporary analysis of the intraoperative and perioperative complications of neurosurgical procedures performed in the sitting position. J Neurosurg. 2017;127:182–188.
4. Schaffranietz L, Grothe A, Olthoff D. Use of the sitting position in neurosurgery. Results of a 1998 survey in Germany. Anaesthesist. 2000;49:269–274.
5. Grand View Research. Brain implants market size report by product (deep brain stimulator, vagus nerve stimulator), by application (chronic pain, parkinson’s, epilepsy, depression, alzheimer’s), and segment forecasts, 2018-2025. 2018. Available at: www.grandviewresearch.com/industry-analysis/brain-implants-market
. Accessed March 3, 2018.
6. Standefer M, Bay JW, Trusso R. The sitting position in neurosurgery: a retrospective analysis of 488 cases. Neurosurgery. 1984;14:649–658.
7. Matjasko J, Petrozza P, Cohen M, et al. Anesthesia and surgery
in the seated position
: analysis of 554 cases. Neurosurgery. 1985;17:695–702.
8. Black S, Ockert DB, Oliver WC Jr, et al. Outcome following posterior fossa
craniectomy in patients in the sitting or horizontal positions. Anesthesiology. 1988;69:49–56.
9. Mirski MA, Lele AV, Fitzsimmons L, et al. Diagnosis and treatment of vascular air embolism. Anesthesiology. 2007;106:164–177.
10. Losasso TJ, Muzzi DA, Dietz NM, et al. Fifty percent nitrous oxide does not increase the risk of venous air embolism
in neurosurgical patients operated upon in the sitting position. Anesthesiology. 1992;77:21–30.
11. Giri J, Li M, Pickering B, et al. Validation of computerized sniffer for monitoring perioperative normothermia. Stud Health Technol Inform. 2013;192:943.
12. Hyder JA, Kor DJ, Cima RR, et al. How to improve the performance of intraoperative risk models: an example with vital signs using the surgical apgar score. Anesth Analg. 2013;117:1338–1346.
13. Saladino A, Lamperti M, Mangraviti A, et al. The semisitting position: analysis of the risks and surgical outcomes in a contemporary series of 425 adult patients undergoing cranial surgery
. J Neurosurg. 2017;127:867–876.
14. McGrail KM, Muzzi DA, Losasso TJ, et al. Ventriculoatrial shunt distal catheter placement using transesophageal echocardiography: technical note. Neurosurgery. 1992;30:747–749.
15. Black S, Muzzi DA, Nishimura RA, et al. Preoperative and intraoperative echocardiography to detect right-to-left shunt in patients undergoing neurosurgical procedures in the sitting position. Anesthesiology. 1990;72:436–438.
16. Feigl GC, Decker K, Wurms M, et al. Neurosurgical procedures in the semisitting position: evaluation of the risk of paradoxical venous air embolism
in patients with a patent foramen ovale. World Neurosurg. 2014;81:159–164.
17. English JB, Westenskow D, Hodges MR, et al. Comparison of venous air embolism
monitoring methods in supine dogs. Anesthesiology. 1978;48:425–429.
18. Kytta J, Randell T, Tanskanen P, et al. Monitoring lung compliance and end-tidal oxygen content for the detection of venous air embolism
. Br J Anaesth. 1995;75:447–451.
19. Matjasko MJ, Hellman J, Mackenzie CF. Venous air embolism
, hypotension, and end-tidal nitrogen. Neurosurgery. 1987;21:378–382.
20. Branger AB, Eckmann DM. Theoretical and experimental intravascular gas embolism absorption dynamics. J Appl Physiol. 1999;87:1287–1295.
21. Eckmann DM, Lomivorotov VN. Microvascular gas embolization clearance following perfluorocarbon administration. J Appl Physiol. 2003;94:860–868.
22. Eckmann DM, Diamond SL. Surfactants attenuate gas embolism-induced thrombin production. Anesthesiology. 2004;100:77–84.
23. Eckmann DM, Armstead SC, Mardini F. Surfactants reduce platelet-bubble and platelet-platelet binding induced by in vitro air embolism. Anesthesiology. 2005;103:1204–1210.
24. Eckmann DM, Armstead SC. Influence of endothelial glycocalyx degradation and surfactants on air embolism adhesion. Anesthesiology. 2006;105:1220–1227.
25. Ishak BA, Seleny FL, Noah ZL. Venous air embolism
, a possible cause of acute pulmonary edema. Anesthesiology. 1976;45:453–455.
26. Moningi S, Kulkarni D, Bhattacharjee S. Coagulopathy following venous air embolism
: a disastrous consequence -a case report. Korean J Anesthesiol. 2013;65:349–352.
27. Michenfelder JD, Terry HR, Daw EF, et al. Air embolism during neurosurgery. Anesth Analg. 1966;45:390–395.
28. Michenfelder JD, Martin JT, Altenburg BM, et al. Air embolism during neurosurgery. JAMA. 1969;208:1353–1358.
29. Albin MS, Babinski M, Maroon JC, et al. Anesthetic management of posterior fossa surgery
in the sitting position. Acta Anaesthesiol Scand. 1976;20:117–128.
30. Albin MS, Carroll RG, Maroon JC. Clinical considerations concerning detection of venous air embolism
. Neurosurgery. 1978;3:380–384.
31. Albin MS, Ritter RR, Pruett CE, et al. Venous air embolism
during lumbar laminectomy in the prone position: report of three cases. Anesth Analg. 1991;73:346–349.
32. Artru AA. Modification of a new catheter for air retrieval and resuscitation from lethal venous air embolism
: effect of nitrous oxide on air retrieval. Anesth Analg. 1992;75:226–231.
33. Munson ES, Merrick HC. Effect of nitrous oxide on venous air embolism
. Anesthesiology. 1966;27:783–787.
34. Lockwood G. Expansion of air bubbles in aqueous solutions of nitrous oxide or xenon. Br J Anaesth. 2002;89:282–286.
Keywords:Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved
seated position; venous air embolism; VAE; central venous catheter; posterior fossa; surgery; cervical spine surgery; deep brain stimulator placement