44. Risk of Gas Embolism with CO2 and Other Gases

44. Risk of Gas Embolism with CO ₂ and Other Gases

David I. Watson, M.B., B.S., M.D., F.R.A.C.S.

At present, carbon dioxide (CO2) is almost universally used to establish and maintain the pneumoperitoneum that provides exposure during laparoscopic procedures. CO2has properties that make it a very suitable agent; it is cheap, noncombustible, and colorless. Furthermore, it is excreted by the lungs during respiration, and it is highly soluble in water, which reduces the risk that gas which finds its way into the venous system will impair cardiac function.

More recently, debate has arisen about the metabolic and oncologic conse- quences of insufflating the peritoneal cavity with CO2, and this has led to a reevaluation of the effects of CO2pneumoperitoneum and interest in the use of other gases for insufflation, in particular inert gases such as helium and argon [1, 2]. Although the use of these alternative gases has been predominantly in the setting of experimental studies, helium, in particular, is also being evaluated in clinical trials [3]. Among their other qualities and attributes, it is important to consider the relative risks of each type of gas in regard to venous embolism.

A. Gas Embolism

Gas embolism is one of the most serious complications of laparoscopy, and it is associated with a mortality rate of approximately 50% [4]. Fortunately, clin- ically significant gas embolism is rare. However, studies that utilized trans- esophageal echocardiography to detect gas in the cardiac chambers suggest that small asymptomatic gas embolism events are actually common during proce- dures such as laparoscopic cholecystectomy [5].

Pathophysiology of Gas Embolism

In the setting of laparoscopy, a gas embolism occurs when insufflated gas finds its way from the peritoneal cavity into the systemic venous circulation via an open vessel [6]. The embolus flows with the blood into the right heart, after which it is pumped into the lungs where the gas embolism is dispersed and becomes trapped as bubbles within the pulmonary capillary bed. A blocked cap- illary impairs both blood flow and gas exchange and results in physiologic shunt- ing due to alterations in pulmonary vascular resistance. As the physiologic deadspace increases, these alterations are manifested by a fall in the end-tidal

PCO2levels. Given a large enough single embolus or a series of low-volume gas emboli, outflow obstruction develops and cardiovascular collapse follows. This is further aggravated by the trapping of gas within the right ventricle. As gas is compressible, cardiac work is wasted on compression of the gas within the ventricle rather than the expulsion of noncompressible blood. In addition to these pulmonary and cardiovascular alterations, gas embolism can also incite an inflammatory reaction via activation of Hageman factor, which, in turn, activates the complement cascade [7]. For these reasons, the end result of a sufficient volume of gas entering the systemic venous circulation is cardiovascular and cir- culatory collapse.

The pulmonary and hemodynamic effects of gas embolism are ultimately determined by the volume of gas entering the venous circulation, the solubility coefficient of the gas in blood, and the diffusibility of the gas in tissue. CO2is highly soluble in water (0.5 mL gas/mL blood); therefore, a considerable per- centage of a venous embolus will be rapidly absorbed. The high solubility of CO2is one of the principal reasons why this gas is so widely utilized for laparo- scopic procedures. What of the other gases that might be used to provide expo- sure during minimally invasive procedures? Helium is the best alternative presently. Helium is attractive because it is highly diffusible in tissue. It has a diffusibility constant that is three times greater than CO2, and therefore helium which has diffused into the patient’s tissues during a case is more quickly elim- inated from the body after the procedure than CO2. However, helium is poorly soluble in blood (0.008 mL gas/mL blood), and thus in the setting of a venous embolus, unlike CO2, only a small volume is absorbed [8]. The poor solubility of helium may be a major impediment to its routine clinical use because the con- sequences of venous gas embolus are potentially far more serious than those relating to retention of gas in the patient’s tissues.

Diagnosis

The early diagnosis of venous gas embolism can be difficult. Classical signs include cardiac dysrhythmia, elevation of pulmonary arterial pressure, elevation of central venous pressure, decrease in end-tidal CO2partial pressure, a fall in systemic blood pressure, and hypoxia. As the volume of CO2in the circulation increases, systolic blood pressure and central venous pressure initially increase, however, systemic hypotension develops later and is a preterminal event.

A number of studies have questioned the reliability of end-tidal CO2in the detection of venous gas embolism [6, 9, 10]. The results of a recent experimental study from Yau et al. [6] suggest that a persistently elevated central venous pressure, combined with unexplained persistent tachycardia and elevated arter- ial PCO2levels, are more reliable indicators of CO2gas embolism. Early signs of helium gas embolism detected using the same porcine model included a fall in systolic and diastolic blood pressure, a fall in arterial PO2, and an elevation in central venous pressure and arterial PCO2[6].

Incidence and Mechanism of Gas Embolism During Laparoscopy

The clinical effect of gas embolism, or the lack thereof, depends upon the amount of gas within the circulation. Small volumes of gas are of little, if any, clinical significance; however, large-volume emboli can lead to cardiac arrest and death. Thus, it is important to distinguish small asymptomatic emboli from larger embolic events that can be fatal. In experimental and clinical studies, ultra- sound has been used for the detection of small gas bubbles within the systemic circulation [5]. However, this method does not permit full characterization of the nature of the embolic material, that is, fat, thrombi, or gas. It is also not possi- ble to determine the volume of the embolus with ultrasound. Thus, ultrasound will detect the presence of emboli but cannot determine the source of the emboli, the size of the embolus, or the clinical significance of the event.

How often does venous gas embolism occur? Mann et al. [11] have reported that transesophageal ultrasound can detect volumes of argon gas as small as 0.1 mL/kg in a pig model with a sensitivity of 100%. However, this author was unable to detect any evidence of gas embolism occurring during routine pneu- moperitoneum at an insufflation pressure of 15 mmHg when using either argon or CO2gas. These findings are supported by another porcine study that also used transesophageal ultrasound. In this study it was demonstrated that CO2 pneu- moperitoneum at a pressure of up to 30 mmHg was not associated with gas enter- ing the intravascular compartment during routine laparoscopy [12]. However, this study did demonstrate that a small volume of CO2could be detected fol- lowing injury to major veins, especially after deflation of the pneumoperitoneum [12]. Similar methodology has been applied in the clinical setting: one study demonstrated that small-volume gas embolism occurred in 11 of 16 patients undergoing laparoscopic cholecystectomy [5]. These patients manifested no sig- nificant cardiorespiratory events, and thus all embolic events were minor and of no clinical significance. This study, in contrast to other reports [13], therefore, suggests that gas embolism might be common during laparoscopy. Regardless, all the studies support the position that clinically relevant embolism rarely occurs during laparoscopy.

Vascular injury when attempting to establish pneumoperitoneum with a Veress needle or when blindly placing the first port via a sharp trocar probably results in the majority of clinically significant gas emboli [14]. A sizable volume of gas may find its way into the bloodstream via such vascular injuries; in addi- tion, such injuries, obviously, may also result in a major hemorrhage. Hemor- rhage can lead to death from exsanguination, whereas embolism, due to direct insufflation of a large vein, can lead to cardiovascular collapse and also death.

Distinguishing death from gas embolism from death due to exsanguination alone is difficult, and for this reason the mortality rate associated with gas embolism alone at the commencement of laparoscopy is uncertain. Nevertheless, published evidence suggests that the incidence of major vascular injury when using a Veress needle for the establishment of pneumoperitoneum lies somewhere between 0.8 and 7.5 injuries per 10,000 laparoscopies [14, 15], and the mortal- ity associated with this is approximately 1 death per 10,000 procedures. It is vir- tually certain that the incidence of major gas embolism is less than or equal to

these relatively low rates. This is supported by studies that report an incidence of clinically significant gas embolism during CO2 pneumoperitoneum which varies between 0.001% and 0.6% [8]. Although not all these events are associ- ated with mortality, in a recent French laparoscopy series two of seven patients who developed venous carbon dioxide embolism died [16].

It should be appreciated that major vascular injuries have not been reported following open insertion of the first laparoscopic port, and because direct intravascular insufflation is not possible during open insertion of the first port, major gas embolism is highly unlikely if the simple precaution of using an open insertion technique for the first laparoscopic trocar is followed. Clinically sig- nificant gas embolism, that is, an event that threatens life, is dependent on the balance of the volume of gas entering the circulation and the amount of gas that is removed. It is unlikely that large volumes of gas enter the circulation during routine uncomplicated laparoscopic procedures. As argued above, it is most likely that, for life-threatening gas embolism to occur, direct venous puncture with a Veress needle or the first trocar, followed by direct venous insufflation, is required. In a recent study, Welch et al. [17] investigated the pressure within the intraabdominal inferior vena cava (IVC). They demonstrated that as insufflation pressure increased, intracaval pressure also increased, and this always exceeded intraperitoneal pressure, thereby maintaining a pressure gradient from the IVC to the peritoneal cavity. A recent study (unpublished data) from the University of Adelaide Department of Surgery investigated the effect of deliberately lacer- ating the IVC under laparoscopic conditions. This was not followed by gas embolism. Both Jacobi et al. [18] and Dion et al. [19] conducted similar studies in which the IVC was deliberately lacerated under laparoscopic conditions, and significant venous gas embolism did not occur. A further study has found that a small volume of intravascular CO2gas bubbles can be detected following laparo- scopic injury to major veins, especially after deflation of the pneumoperitoneum [12]. However, the volume of gas involved was again not clinically significant.

These studies reinforce the view that the Veress needle insertion technique is the means by which the majority of clinically relevant gas embolism events occur.

B. The Relative Safety of Alternative Insufflation Gases

Potential oncologic and immunologic advantages demonstrated by experi- mental studies of inert insufflation gases have encouraged investigators to conduct experimental studies to evaluate the relative safety of these gases, in particular the consequences of iatrogenic gas embolism events. Concern has been expressed about the safety of helium insufflation if gas embolism occurs.

There is conflicting experimental evidence concerning the relative dangers of embolism when using helium insufflation compared to CO2. Porcine studies by Rudston-Brown et al. [20], Yau et al. [6], and Wolf et al. [8] have demon- strated that helium gas embolism has a greater deleterious effect than CO2

embolism on survival following the intravenous injection of gas. When com-

pared to results following direct intravascular injection of CO2, a smaller volume of helium is required to bring about cardiac arrest. This experimental model mimics the clinical situation of inadvertent insertion of a Veress needle into a vein when attempting to establish pneumoperitoneum. Jacobi et al. [18] used a different model to assess the frequency of gas embolism from a venous injury that occurs in the midst of a laparoscopic procedure, after the successful estab- lishment of pneumoperitoneum. During these procedures the IVC was lacerated and allowed to bleed for a period of time, after which the injury was repaired.

No embolic events were noted in any of the animals; these results suggested that the incidence of gas embolism during laparoscopy from accidental injury of a large vessel is likely to be low. In their dog model, Dion et al. [19] reported a similar outcome.

Helium has been used clinically for laparoscopic procedures; thus far, no complications specific to its use have been reported. Although thus far there are no reports of gas embolism, only a limited number of centers are using helium, most often in the setting of a study. As it is most likely that clinically significant and life-threatening embolism problems result from the direct injection of a large volume of gas into a large vein via a Veress needle at induction of the pneu- moperitoneum, the problem of gas embolism with inert gases is potentially avoidable if care is taken to always use an open cannulation technique to com- mence laparoscopy.

Argon, like helium, is inert, nonflammable, and available [21]. It is also rel- atively insoluble in blood and, therefore, the consequences of argon gas embolism are likely to be, similar to helium, greater than for CO2. Argon insuf- flation has been investigated by Mann et al. [11], also using a porcine model.

During a stable pneumoperitoneum of either argon or CO2gas, no episodes of gas embolism were detected. However, when gas was directly injected into the venous circulation, the cardiovascular effects of gas embolism were greater with argon than for CO2. These results mirrored those of the helium experiments men- tioned above.

Nitrous oxide was used commonly for insufflation in the 1970s and 1980s [22]. Concerns about the combustibility of this gas when electrocautery is in use led to a decline in the utilization of nitrous oxide. More recently, nitrogen has been proposed for insufflation [23]. However, there is currently little evidence to support its use. The clinical implications of nitrous oxide and nitrogen gas embolism have not been investigated thus far.

C. Conclusions

Venous gas embolism is a rare, but potentially fatal, complication of laparoscopy. It is largely because of concern about this risk that CO2, which is more rapidly eliminated from the blood than other gases, has become the stan- dard insufflation agent for laparoscopic surgery. Life-threatening gas embolism is most likely to occur following the inadvertent introduction of a Veress needle into a large vein, followed by direct intravenous insufflation. Once the pneu-

moperitoneum has been established and the first port placed, clinically relevant gas embolic events are exceptionally rare. Therefore, use of an open “cut-down”

cannulation technique to commence laparoscopy should drastically reduce the incidence of a major gas embolism.

The use of alternative insufflation gases, in particular the inert gas helium, is supported by experimental evidence that suggests that by utilizing an alter- native gas the adverse metabolic and oncologic consequences associated with CO2can largely be avoided. However, because of their low solubility in blood, inert insufflation gases are inherently more dangerous than CO2from the point of view of embolism because cardiovascular compromise occurs following a much smaller volume of intravascular inert gas than of CO2. However, there may be a legitimate clinical role for inert gases, provided pneumoperitoneum is not established with a Veress needle, because gas embolism is so unlikely to occur after the pneumoperitoneum has been safely established.

D. References

1. Jacobi CA, Wenger F, Sabat R, Volk T, Ordemann J, Muller JM. The impact of laparoscopy with carbon dioxide versus helium on immunologic function and tumor growth in a rat model. Dig Surg 1998;15:110–116.

2. Neuhaus SJ, Watson DI, Ellis T, et al. Wound metastasis following different insuffla- tion gases. Surgery 1998;123:579–583.

3. Leighton TA, Liu SY, Bongard FS. Comparative cardiopulmonary effects of carbon dioxide versus helium pneumoperitoneum. Surgery 1993;113:527–531.

4. Baxter JN, O’Dwyer PJ. Pathophysiology of laparoscopy. Br J Surg 1995;82:1–2.

5. Derouin M, Couture P, Boudreault D, Girard D, Gravel D. Detection of gas embolism by transesophageal echocardiography during laparoscopic cholecystectomy. Anesth Analg 1996;82:119–124.

6. Yau P, Watson DI, Lafullarde T, Jamieson GG. An experimental study of the effect of gas embolism using different laparoscopy insufflation gases. J Laparoendosc Adv Surg Tech 2000;10:211–216.

7. Neuberger T, Andrus C, Wittgen C, Wade T, Kaminski DL. Prospective comparison of helium versus carbon dioxide pneumoperitoneum. Gastrointest Endosc 1996;43:

38–41.

8. Wolf JS, Carrier S, Stoller M. Gas embolism: helium is more lethal than carbon dioxide. J Laparoendosc Surg 1994;4:173–177.

9. Mayer K, Ho HS, Mathiesen K, Wolfe B. Cardiopulmonary responses to experimental venous carbon dioxide embolism. Surg Endosc 1998;12:1025–1030.

10. Byrick RJ, Kay JC, Mullen JB. Capnography is not as sensitive as pulmonary artery pressure monitoring in detecting marrow microembolism; studies in a canine model.

Anesth Analg 1989;68:94–100.

11. Mann C, Boccara G, Grevy V, Navarro F, Fabre J, Colson P. Argon pneumoperitoneum is more dangerous than CO2pneumoperitoneum during venous gas embolism. Anesth Analg 1997;85:1367–1371.

12. Bazin JE, Gillart T, Rasson P, Conio N, Aigouy L, Schoeffler P. Haemodynamic con- ditions enhancing gas embolism after venous injury during laparoscopy: a study in pigs. Br J Anesth 1997;78:570–575.

13. Thio JM, Reichert C. Transesophageal echocardiographic assessment of venous carbon dioxide embolism during laparoscopic cholecystectomy. Anesthesiology 1994;

31:A112 [abstract].

14. Bonjer HJ, Hazebroek EJ, Kazemier G, Giuffrida MC, Meijer WS, Lange JF. Open versus closed establishment of pneumoperitoneum in laparoscopic surgery. Br J Surg 1997;84:599–602.

15. Hanney RM, Alle KM, Cregan PC. Major vascular injury and laparoscopy. Aust N Z J Surg 1995;65:533–535.

16. Cottin V, Delafosse B, Viale J. Gas embolism during laparoscopy: a report of seven cases in patients with pervious abdominal surgical history. Surg Endosc 1996;10:

166–169.

17. Welch LS, Urbach DR, Herron DM, Ludemann R, Swanstrom LL, Hansen PD. The relationship between pneumoperitoneum pressure and pressure within the intra- abdominal inferior vena cava in a pig model. Surg Endosc 2000;14:S227 [abstract].

18. Jacobi CA, Junghans T, Peter F, et al. Cardiopulmonary changes during laparoscopy and vessels injury: comparison of CO2and helium in an animal model. Langenbecks Arch Surg 2000;385:459–466.

19. Dion YM, Levesque C, Diollon CJ. Experimental carbon dioxide pulmonary emboli- sation after vena cava laceration under pneumoperitoneum. Surg Endosc 1995;9:

1065–1069.

20. Rudston-Brown B, Draper PN, Warriner B, Walley KR, Phang PT. Venous gas embolism: a comparison of carbon dioxide and helium in pigs. Can J Anaesth 1997;

44:1102–1107.

21. Eisenhauer DM, Saunders CJ, Ho HS, Wolfe BM. Hemodynamic effects of argon pneumoperitoneum. Surg Endosc 1994;8:315–320.

22. Aitola P, Airo I, Kaukinen S, Ylitalo P. Comparison of N2O and CO2pneumoperi- toneums during laparoscopic cholecystectomy with special reference to postoperative pain. Surg Laparosc Endosc 1998;8:140–144.

23. Aneman A, Svenson M, Stenqvist O, Dalenback J, Lonnroth H. Intestinal perfusion during pneumoperitoneum with carbon dioxide, nitrogen and nitric oxide during laparoscopic surgery. Eur J Surg 2000;166:70–76.