- Email: [email protected]
DOES ETOMIDATE CAUSE HAEMOLYSIS?
British Journal of Anaesthesia 1992; 69: 58-60
DOES ETOMIDATE CAUSE HAEMOLYSIS? A. E. NEBAUER, A. DOENICKE, R. HOERNECKE, R. ANGSTER AND M.MAYER
SUMMARY
KEY WORDS Anaesthetics, intravenous: etomidate. Complications: haemolysis.
Administration of etomidate is associated frequently with pain on injection and thrombosis or thrombophlebitis. The commercially available formulation of etomidate contains the organic solvent propylene glycol (PG). Earlier studies have shown that PG produces a high osmolality (4965 mosmol kg"1) in etomidate [1]. It contributes to cell damage and causes vascular tissue inflammation with subsequent intravascular thrombosis [2] and haemolysis [3]. Replacement of the solvent propylene glycol with lipid emulsion has significantly reduced the painful side effects, indicating that PG plays a significant role in this context [4]. This is true not only for etomidate, but also for diazepam [5]. In a previous study on healthy volunteers who received etomidate in two different solvents, we observed a brighter red colouration of the sera of some blood samples, indicating haemolysis; all these samples were from the group that received etomidate in propylene glycol (EtoPG). The purpose of this study was to investigate if etomidate causes haemolysis in man and to assess the role of the solvent in this process. SUBJECTS AND METHODS
Ethics Committee approval and written informed
ALEXANDER E. NEBAUER; ALFRED DOENICKE, M.D.; RAINER HOERNECKE, M . D . ; ROBERT ANGSTER, M . D . ; MICHAEL MAYER,
M.D.; Institut fur Anaesthesiologie, Ludwig-MaximiliansUniversitat Munchen, Pettenkoferstr. 8a, 8000 Munchen 2, Germany. Accepted for Publication: February 6, 1992. Correspondence to A.D.
Downloaded from http://bja.oxfordjournals.org/ at Emory University on April 18, 2015
Etomidate is currently presented as a solution with propy/ene glycol as solvent. This organic solvent has an extremely high osmolality and is probably responsible for some of the side effects of this drug. In order to detect haemolysis, an indication for cell damage, we have measured serum haptoglobin concentrations in 12 healthy male volunteers after administration of etomidate 0.3 mg kg~1. Six subjects received etomidate in propy/ene glycol (EtoPG) with an osmolality of 4965 mosmol kg~1 and six received etomidate in lipid emulsion (EtoLip, 400 mosmol kg'1). Haptoglobin concentrations in the EtoPG group decreased by 44% and 43 % from baseline values at 2 and 4 h after administration, respectively, and were significantly smaller than after administration of EtoLip. After 24 h, haptoglobin concentrations had not reached baseline values.
consent were obtained. We studied 12 healthy, male volunteers (ages 19-27 yr; weights 65-87 kg) allocated randomly to two groups of six each. One group received etomidate 0.3 mg kg"1 in lipid emulsion (EtoLip) (Eto-Lipuro, B. Braun, Melsungen, Germany); the other received the same dose of etomidate in propylene glycol (EtoPG) (Janssen, Neuss, Germany). EtoLip is a new formulation containing, as solvent, Lipofundin MCT 20% (B. Braun, Melsungen, Germany)—a lipid emulsion consisting of soyabean oil, glycerol, egg phosphatides and medium chain triglycerides. It was under clinical investigation at the time of this study and has been approved recently by the Bundesgesundheitsamt (German Health Administration). EtoPG (Hypnomidate, Amidate) is the commercially available formulation containing 35 vol % propylene glycol as solvent. Etomidate was administered within 60 s via a 16gauge cannula (Viggo Venflon), placed in an antecubital vein. Patency of the cannula was maintained over the following 4 h by a continuous infusion of 0.9% sodium chloride solution 250 ml. Samples were obtained by allowing blood to drip into tubes, in order to prevent haemolysis. Arterial pressure and ECG were monitored continuously for 30 min. No other medication was given. Blood was obtained for measurement of PCV and haemoglobin (Hb) and haptoglobin (Hp) concentrations 10 min before injection and 5, 10, 30, 60, 120, 240 min and 24 h after injection. Laboratory investigators were blinded. Hb and PCV were measured in a Coulter Counter T660. Blood for measurement of Hp was sampled into sodium citrated Sarstedt tubes, spun in a chilled centrifuge at 3500 £ for 10 min and deep frozen at - 2 0 °C. Further laboratory processing was performed at the end of the study, which was completed in 4 days. Hp was determined by laser nephelometry (Behring BNA Nephelometer). Because of the small number of subjects, data were analysed with non-parametric Mann—Whitney U test for comparison between groups. Significance was denned at P < 0.05.
DOES ETOMIDATE CAUSE HAEMOLYSIS? 2.0
T
£1.5 c u
§1.0 c !o
2
CD
g-0.5
0.0 -10
5
10
30 60 Time (min)
120
240
1440
RESULTS
One subject in each group had to be eliminated because Hp concentrations were beyond the reference range (0.5—2.0 g litre"1). Hp concentrations after injection of EtoPG decreased steadily from baseline, with maximal mean decreases of 44.1% (EtoLip: 10.5% decrease) and 43.1% (EtoLip: 4.6% decrease) after 2 and 4 h, respectively. The Hp concentration was still less than baseline (27% decrease) 24 h after administration, at which time the concentration in the EtoLip group had increased to 4.4% greater than baseline (fig. 1). Differences between the two groups were significant at 2 h (P = 0.009), 4 h (P = 0.009) and 24 h (P = 0.028) after injection. However, the numbers of subjects in each group were so small that statistical analysis must be interpreted very cautiously. Haemoglobin concentrations, PCV values and red blood cell counts did not change significantly within or between groups (data not shown). In both groups, PCV decreased slightly, probably because of dilution by the saline infusion. Free haemoglobin was not detected in any blood sample. DISCUSSION
Haptoglobin is a highly sensitive marker for haemolysis. It binds irreversibly free haemoglobin that appears in blood plasma. This otj-glycoprotein is synthesized in the liver and acts as an acute phase reaction marker. The normal plasma values range from about 0.5 to 2.0 g litre"1, depending on the individual phenotype. Because of its genetic polymorphism, interindividual plasma concentrations may vary considerably. Hp has a half-life of about 5 days in the circulation, but when it binds Hb to form the HpHb complex, its half-life is reduced to about 10 min. This complex is removed rapidly by the reticuloendothelial system, mainly by catabolism in the liver. Increasing concentrations of free Hb
therefore reduce Hp concentrations and, when total Hb-binding capacity is reached, Hp is eliminated from the plasma and haemoglobinaemia occurs. Depletion of this highly sensitive marker for free Hb does not induce compensatory synthesis, and Hp concentrations return to normal only after 3-6 days after cessation of haemolysis. The HpHb complex, being too large to be filtered through the renal glomeruli, thus acts as a primary renal threshold for Hb, conserving iron and preventing tubular damage [6, 7]. Only a very small number of volunteers were included in this pilot study. We believe that it would not be ethically justifiable to administer an active i.v. hypnotic to a large number of healthy subjects. At least 35 volunteers in each group would be needed to show a statistically significant difference, and in patients such a study would be influenced by too many factors related to surgery and the underlying disease. The decrease in Hp concentration in the volunteers who received EtoPG compared with those treated with EtoLip provides good evidence for our assumption that PG causes haemolysis of erythrocytes. Mean Hp concentrations decreased steadily after administration of EtoPG and at 2 and 4 h after injection, all values were less than those of the EtoLip group (fig. 1). We believe that haemolysis was caused by the extremely unphysiological osmolality of EtoPG. Our measurements confirm the value of 4965 mosmol kg"1 found by Bretschneider [1]. This is 16 times greater than blood osmolality. In comparison, EtoLip has a more physiological osmolality of 400 mosmol kg"1 (our measurements). We cannot explain the reason why Hp values were least at 2-4 h after injection, as one would expect a faster decrease. It is possible that haemolysis is protracted, or that haptoglobin is stored in some unknown pools. The degree of haemolysis was not clinically significant in these healthy, young men because the haemoglobin binding capacity of haptoglobin was not exceeded. Other indicators of haemolysis, such as free or total haemoglobin concentration were not abnormal, although the sera of the centrifugated plasma samples of the EtoPG group had a reddish colour compared with those of the EtoLip group. Many other drugs containing PG as solvent (e.g. nitroglycerin, diazepam or antibiotics) are often administered in larger volumes or infused over several hours and this may cause greater haemolysis. A recent study has confirmed the haemolytic activity of hypertonic solutions. Saline-Dextran infusion for resuscitation in haemorrhaged dogs induced significant haemolysis in relation to the osmolality of the solution and the route of administration. When given into peripheral veins, free haemoglobin values were greater than those after central infusion. In vitro incubation of human and canine blood with solutions of increasing osmolarity produced measurable haemolysis for concentrations > 900 mosmol litre"1. [8]. Acute intravascular haemolysis in man has been reported after infusion of nitroglycerin containing
Downloaded from http://bja.oxfordjournals.org/ at Emory University on April 18, 2015
FIG. 1. Haptoglobin concentrations of 10 male volunteers after injection of ctomidate 0.3 mg kg"1 in lipid emulsion (EtoLip) ( • ) (n = 5) and in propylene glycol CEtoPG) ( # ) (n = 5). Reference range of normal haptoglobin concentrations: 0.5—2.0 g litre"1.
59
BRITISH JOURNAL OF ANAESTHESIA
60
ACKNOWLEDGEMENT We thank Prof. A. Fateh-Moghadam from the Institute of Clinical Chemistry, LMU Munchen, for the measurements of haptoglobin concentrations. REFERENCES 1. Bretschneider H. Osmolalines of commercially supplied
drugs often used in anesthesia. Anesthesia and Analgesia 1987; 66: 361-362. 2. Graham CW, Pagano RR, Katz RL. Thrombophlebitis after intravenous diazepam—can it be prevented? Anesthesia and Analgesia 1977; 56: 409-413. 3. Demey HE, Daelemans RA, Verpooten GA, De Broe ME, Van Campenhout CHM, Lakiere FV, Schepens PJ, Bossaert LL. Propylene glycol-induced side effects during nitroglycerin therapy. Intensive Care Medicine 1988; 14: 221-226. 4. Doenicke A, Kugler A, Vollmann N, Suttmann H, Taeger K. Etomidat mit einem neuen Losungsvermittler. Klinischexperimentelle Untersuchungen zur VenenvertrSglichkeit und Bioverfugbarkeit. Anaesthesist 1990; 39: 475-480. 5. Von Dardel O, Mebius C, Mossberg T, Svensson B. Fat emulsion as a vehicle for diazepam. A study of 9492 patients. British Journal of Anaesthesia 1983; 55: 41-47. 6. Jandl JH. Physiology of red cells. In: Jandl JH, ed. Blood—Textbook of Hematology, 1st Edn. Boston: Little, Brown and Co, 1987; 87-89. 7. Thomas L. Haptoglobin/HSmopexin. In: Thomas L, ed. Labor und Diagnose, 3rd Edn. Marburg/Lahn: Die Medizinische Verlagsgesellschaft, 1988; 698-705. 8. Rocha e Silva M, Velasco IT, Porfirio MF. Hypertonic saline resuscitation: Saturated salt-Dextran solutions are equally effective, but induce hemolysis in dogs. Critical Care Medicine 1990; 18: 203-207. 9. Lehman AJ, Newman HW. Propylene glycol: Rate of metabolism, absorption, and excretion, with a method for estimation in body fluids. Journal of Pharmacology and Experimental Therapy 1937; 60: 312-322. 10. Randolph TG, Mallery OT. The effect in vitro of propylene glycol on erythrocytes. Journal of Laboratory and Clinical Medicine 1944; 29: 197-202. 11. Cadwallader DE. Behavior of erythrocytes in various solvent systems I. Water-glycerin and water-propylene glycol. Journal of Pharmaceutical Sciences 1963; 52: 1175-1180. 12. Potter BJ. Haemoglobinuria caused by propylene glycol in sheep. British Journal of Pharmacology 1958; 12: 385-938. 13. Aspelin P. Effect of ionic and non-ionic contrast media on morphology of human erythrocytes. Ada Radiologica [Diagnosis] 1978; 19: 675-687. 14. Aspelin P. Effect of ionic and non-ionic media on red cell deformability in vitro. Acta Radiologica [Diagnosis] 1979; 20: 1-12. 15. Findlay SR, Dvorak AM, Kagey-Sobotka A, Lichtenstein LM. Hyperosmolar triggering of histamine release from human basophils. Journal of Clinical Investigation 1981; 67: 1604-1613. 16. Denning DW, Webster DB. Detrimental effect of propylene glycol on natural killer cell and neutrophil function. Journal of Pharmacy and Pharmacology 1987; 39; 236-238. 17. Brazeau GA, Fung H-L. Mechanisms of creatine kinase release from isolated rat skeletal muscles damaged by propylene glycol and ethanol. Journal of Pharmaceutical Sciences 1990; 79: 393-397.
Downloaded from http://bja.oxfordjournals.org/ at Emory University on April 18, 2015
96 vol % PG. In one case, haemolysis occurred during simultaneous transfusion of packed red cells through the same cannula [3]. In 1937, Lehman and Newman observed haemoglobinuria after injecting a 33.3-vol % solution of PG into dogs. They assumed that the blood cells had been destroyed osmotically [9]. Consequent in vitro studies with PG did not clarify the haemolytic properties of PG. Controversial results were published indicating that haemolysis did not always occur, but could be prevented by addition of saline to solutions containing up to 30-40% PG [10, 11]. Haemolysis and haemoglobinuria have been observed also after anaesthesia with PG containing pentobarbitone preparations in sheep. This species may be particularly sensitive to PG, because haemoglobinuria was not observed in rabbits [12]. Hyperosmolar solutions affect the integrity of cells and may trigger release of mediators. Hypertonic contrast media (1500-2400 mosmol kg"1) and hypertonic saline are known to lead to the formation of so-called desiccocytes—shrunken erythrocytes. Consequently, whole blood viscosity is increased because of reduced red cell deformability impairing blood flow through capillaries [13, 14]. In vitro histamine release from human basophils exposed to hyperosmolar mannitol, glucose and saline solutions and contrast media has been described [15]. It is possible that PG may also have direct toxic effects on blood and other tissue cells. Immunosuppressive effects on natural killer cells and neutrophils [16] and skeletal muscle damage [17] have been described. We believe that etomidate in propylene glycol causes haemolysis. However, further studies on a larger number of volunteers, receiving only the solvent, are necessary to confirm this hypothesis.
DOES ETOMIDATE CAUSE HAEMOLYSIS? A. E. NEBAUER, A. DOENICKE, R. HOERNECKE, R. ANGSTER AND M.MAYER
SUMMARY
KEY WORDS Anaesthetics, intravenous: etomidate. Complications: haemolysis.
Administration of etomidate is associated frequently with pain on injection and thrombosis or thrombophlebitis. The commercially available formulation of etomidate contains the organic solvent propylene glycol (PG). Earlier studies have shown that PG produces a high osmolality (4965 mosmol kg"1) in etomidate [1]. It contributes to cell damage and causes vascular tissue inflammation with subsequent intravascular thrombosis [2] and haemolysis [3]. Replacement of the solvent propylene glycol with lipid emulsion has significantly reduced the painful side effects, indicating that PG plays a significant role in this context [4]. This is true not only for etomidate, but also for diazepam [5]. In a previous study on healthy volunteers who received etomidate in two different solvents, we observed a brighter red colouration of the sera of some blood samples, indicating haemolysis; all these samples were from the group that received etomidate in propylene glycol (EtoPG). The purpose of this study was to investigate if etomidate causes haemolysis in man and to assess the role of the solvent in this process. SUBJECTS AND METHODS
Ethics Committee approval and written informed
ALEXANDER E. NEBAUER; ALFRED DOENICKE, M.D.; RAINER HOERNECKE, M . D . ; ROBERT ANGSTER, M . D . ; MICHAEL MAYER,
M.D.; Institut fur Anaesthesiologie, Ludwig-MaximiliansUniversitat Munchen, Pettenkoferstr. 8a, 8000 Munchen 2, Germany. Accepted for Publication: February 6, 1992. Correspondence to A.D.
Downloaded from http://bja.oxfordjournals.org/ at Emory University on April 18, 2015
Etomidate is currently presented as a solution with propy/ene glycol as solvent. This organic solvent has an extremely high osmolality and is probably responsible for some of the side effects of this drug. In order to detect haemolysis, an indication for cell damage, we have measured serum haptoglobin concentrations in 12 healthy male volunteers after administration of etomidate 0.3 mg kg~1. Six subjects received etomidate in propy/ene glycol (EtoPG) with an osmolality of 4965 mosmol kg~1 and six received etomidate in lipid emulsion (EtoLip, 400 mosmol kg'1). Haptoglobin concentrations in the EtoPG group decreased by 44% and 43 % from baseline values at 2 and 4 h after administration, respectively, and were significantly smaller than after administration of EtoLip. After 24 h, haptoglobin concentrations had not reached baseline values.
consent were obtained. We studied 12 healthy, male volunteers (ages 19-27 yr; weights 65-87 kg) allocated randomly to two groups of six each. One group received etomidate 0.3 mg kg"1 in lipid emulsion (EtoLip) (Eto-Lipuro, B. Braun, Melsungen, Germany); the other received the same dose of etomidate in propylene glycol (EtoPG) (Janssen, Neuss, Germany). EtoLip is a new formulation containing, as solvent, Lipofundin MCT 20% (B. Braun, Melsungen, Germany)—a lipid emulsion consisting of soyabean oil, glycerol, egg phosphatides and medium chain triglycerides. It was under clinical investigation at the time of this study and has been approved recently by the Bundesgesundheitsamt (German Health Administration). EtoPG (Hypnomidate, Amidate) is the commercially available formulation containing 35 vol % propylene glycol as solvent. Etomidate was administered within 60 s via a 16gauge cannula (Viggo Venflon), placed in an antecubital vein. Patency of the cannula was maintained over the following 4 h by a continuous infusion of 0.9% sodium chloride solution 250 ml. Samples were obtained by allowing blood to drip into tubes, in order to prevent haemolysis. Arterial pressure and ECG were monitored continuously for 30 min. No other medication was given. Blood was obtained for measurement of PCV and haemoglobin (Hb) and haptoglobin (Hp) concentrations 10 min before injection and 5, 10, 30, 60, 120, 240 min and 24 h after injection. Laboratory investigators were blinded. Hb and PCV were measured in a Coulter Counter T660. Blood for measurement of Hp was sampled into sodium citrated Sarstedt tubes, spun in a chilled centrifuge at 3500 £ for 10 min and deep frozen at - 2 0 °C. Further laboratory processing was performed at the end of the study, which was completed in 4 days. Hp was determined by laser nephelometry (Behring BNA Nephelometer). Because of the small number of subjects, data were analysed with non-parametric Mann—Whitney U test for comparison between groups. Significance was denned at P < 0.05.
DOES ETOMIDATE CAUSE HAEMOLYSIS? 2.0
T
£1.5 c u
§1.0 c !o
2
CD
g-0.5
0.0 -10
5
10
30 60 Time (min)
120
240
1440
RESULTS
One subject in each group had to be eliminated because Hp concentrations were beyond the reference range (0.5—2.0 g litre"1). Hp concentrations after injection of EtoPG decreased steadily from baseline, with maximal mean decreases of 44.1% (EtoLip: 10.5% decrease) and 43.1% (EtoLip: 4.6% decrease) after 2 and 4 h, respectively. The Hp concentration was still less than baseline (27% decrease) 24 h after administration, at which time the concentration in the EtoLip group had increased to 4.4% greater than baseline (fig. 1). Differences between the two groups were significant at 2 h (P = 0.009), 4 h (P = 0.009) and 24 h (P = 0.028) after injection. However, the numbers of subjects in each group were so small that statistical analysis must be interpreted very cautiously. Haemoglobin concentrations, PCV values and red blood cell counts did not change significantly within or between groups (data not shown). In both groups, PCV decreased slightly, probably because of dilution by the saline infusion. Free haemoglobin was not detected in any blood sample. DISCUSSION
Haptoglobin is a highly sensitive marker for haemolysis. It binds irreversibly free haemoglobin that appears in blood plasma. This otj-glycoprotein is synthesized in the liver and acts as an acute phase reaction marker. The normal plasma values range from about 0.5 to 2.0 g litre"1, depending on the individual phenotype. Because of its genetic polymorphism, interindividual plasma concentrations may vary considerably. Hp has a half-life of about 5 days in the circulation, but when it binds Hb to form the HpHb complex, its half-life is reduced to about 10 min. This complex is removed rapidly by the reticuloendothelial system, mainly by catabolism in the liver. Increasing concentrations of free Hb
therefore reduce Hp concentrations and, when total Hb-binding capacity is reached, Hp is eliminated from the plasma and haemoglobinaemia occurs. Depletion of this highly sensitive marker for free Hb does not induce compensatory synthesis, and Hp concentrations return to normal only after 3-6 days after cessation of haemolysis. The HpHb complex, being too large to be filtered through the renal glomeruli, thus acts as a primary renal threshold for Hb, conserving iron and preventing tubular damage [6, 7]. Only a very small number of volunteers were included in this pilot study. We believe that it would not be ethically justifiable to administer an active i.v. hypnotic to a large number of healthy subjects. At least 35 volunteers in each group would be needed to show a statistically significant difference, and in patients such a study would be influenced by too many factors related to surgery and the underlying disease. The decrease in Hp concentration in the volunteers who received EtoPG compared with those treated with EtoLip provides good evidence for our assumption that PG causes haemolysis of erythrocytes. Mean Hp concentrations decreased steadily after administration of EtoPG and at 2 and 4 h after injection, all values were less than those of the EtoLip group (fig. 1). We believe that haemolysis was caused by the extremely unphysiological osmolality of EtoPG. Our measurements confirm the value of 4965 mosmol kg"1 found by Bretschneider [1]. This is 16 times greater than blood osmolality. In comparison, EtoLip has a more physiological osmolality of 400 mosmol kg"1 (our measurements). We cannot explain the reason why Hp values were least at 2-4 h after injection, as one would expect a faster decrease. It is possible that haemolysis is protracted, or that haptoglobin is stored in some unknown pools. The degree of haemolysis was not clinically significant in these healthy, young men because the haemoglobin binding capacity of haptoglobin was not exceeded. Other indicators of haemolysis, such as free or total haemoglobin concentration were not abnormal, although the sera of the centrifugated plasma samples of the EtoPG group had a reddish colour compared with those of the EtoLip group. Many other drugs containing PG as solvent (e.g. nitroglycerin, diazepam or antibiotics) are often administered in larger volumes or infused over several hours and this may cause greater haemolysis. A recent study has confirmed the haemolytic activity of hypertonic solutions. Saline-Dextran infusion for resuscitation in haemorrhaged dogs induced significant haemolysis in relation to the osmolality of the solution and the route of administration. When given into peripheral veins, free haemoglobin values were greater than those after central infusion. In vitro incubation of human and canine blood with solutions of increasing osmolarity produced measurable haemolysis for concentrations > 900 mosmol litre"1. [8]. Acute intravascular haemolysis in man has been reported after infusion of nitroglycerin containing
Downloaded from http://bja.oxfordjournals.org/ at Emory University on April 18, 2015
FIG. 1. Haptoglobin concentrations of 10 male volunteers after injection of ctomidate 0.3 mg kg"1 in lipid emulsion (EtoLip) ( • ) (n = 5) and in propylene glycol CEtoPG) ( # ) (n = 5). Reference range of normal haptoglobin concentrations: 0.5—2.0 g litre"1.
59
BRITISH JOURNAL OF ANAESTHESIA
60
ACKNOWLEDGEMENT We thank Prof. A. Fateh-Moghadam from the Institute of Clinical Chemistry, LMU Munchen, for the measurements of haptoglobin concentrations. REFERENCES 1. Bretschneider H. Osmolalines of commercially supplied
drugs often used in anesthesia. Anesthesia and Analgesia 1987; 66: 361-362. 2. Graham CW, Pagano RR, Katz RL. Thrombophlebitis after intravenous diazepam—can it be prevented? Anesthesia and Analgesia 1977; 56: 409-413. 3. Demey HE, Daelemans RA, Verpooten GA, De Broe ME, Van Campenhout CHM, Lakiere FV, Schepens PJ, Bossaert LL. Propylene glycol-induced side effects during nitroglycerin therapy. Intensive Care Medicine 1988; 14: 221-226. 4. Doenicke A, Kugler A, Vollmann N, Suttmann H, Taeger K. Etomidat mit einem neuen Losungsvermittler. Klinischexperimentelle Untersuchungen zur VenenvertrSglichkeit und Bioverfugbarkeit. Anaesthesist 1990; 39: 475-480. 5. Von Dardel O, Mebius C, Mossberg T, Svensson B. Fat emulsion as a vehicle for diazepam. A study of 9492 patients. British Journal of Anaesthesia 1983; 55: 41-47. 6. Jandl JH. Physiology of red cells. In: Jandl JH, ed. Blood—Textbook of Hematology, 1st Edn. Boston: Little, Brown and Co, 1987; 87-89. 7. Thomas L. Haptoglobin/HSmopexin. In: Thomas L, ed. Labor und Diagnose, 3rd Edn. Marburg/Lahn: Die Medizinische Verlagsgesellschaft, 1988; 698-705. 8. Rocha e Silva M, Velasco IT, Porfirio MF. Hypertonic saline resuscitation: Saturated salt-Dextran solutions are equally effective, but induce hemolysis in dogs. Critical Care Medicine 1990; 18: 203-207. 9. Lehman AJ, Newman HW. Propylene glycol: Rate of metabolism, absorption, and excretion, with a method for estimation in body fluids. Journal of Pharmacology and Experimental Therapy 1937; 60: 312-322. 10. Randolph TG, Mallery OT. The effect in vitro of propylene glycol on erythrocytes. Journal of Laboratory and Clinical Medicine 1944; 29: 197-202. 11. Cadwallader DE. Behavior of erythrocytes in various solvent systems I. Water-glycerin and water-propylene glycol. Journal of Pharmaceutical Sciences 1963; 52: 1175-1180. 12. Potter BJ. Haemoglobinuria caused by propylene glycol in sheep. British Journal of Pharmacology 1958; 12: 385-938. 13. Aspelin P. Effect of ionic and non-ionic contrast media on morphology of human erythrocytes. Ada Radiologica [Diagnosis] 1978; 19: 675-687. 14. Aspelin P. Effect of ionic and non-ionic media on red cell deformability in vitro. Acta Radiologica [Diagnosis] 1979; 20: 1-12. 15. Findlay SR, Dvorak AM, Kagey-Sobotka A, Lichtenstein LM. Hyperosmolar triggering of histamine release from human basophils. Journal of Clinical Investigation 1981; 67: 1604-1613. 16. Denning DW, Webster DB. Detrimental effect of propylene glycol on natural killer cell and neutrophil function. Journal of Pharmacy and Pharmacology 1987; 39; 236-238. 17. Brazeau GA, Fung H-L. Mechanisms of creatine kinase release from isolated rat skeletal muscles damaged by propylene glycol and ethanol. Journal of Pharmaceutical Sciences 1990; 79: 393-397.
Downloaded from http://bja.oxfordjournals.org/ at Emory University on April 18, 2015
96 vol % PG. In one case, haemolysis occurred during simultaneous transfusion of packed red cells through the same cannula [3]. In 1937, Lehman and Newman observed haemoglobinuria after injecting a 33.3-vol % solution of PG into dogs. They assumed that the blood cells had been destroyed osmotically [9]. Consequent in vitro studies with PG did not clarify the haemolytic properties of PG. Controversial results were published indicating that haemolysis did not always occur, but could be prevented by addition of saline to solutions containing up to 30-40% PG [10, 11]. Haemolysis and haemoglobinuria have been observed also after anaesthesia with PG containing pentobarbitone preparations in sheep. This species may be particularly sensitive to PG, because haemoglobinuria was not observed in rabbits [12]. Hyperosmolar solutions affect the integrity of cells and may trigger release of mediators. Hypertonic contrast media (1500-2400 mosmol kg"1) and hypertonic saline are known to lead to the formation of so-called desiccocytes—shrunken erythrocytes. Consequently, whole blood viscosity is increased because of reduced red cell deformability impairing blood flow through capillaries [13, 14]. In vitro histamine release from human basophils exposed to hyperosmolar mannitol, glucose and saline solutions and contrast media has been described [15]. It is possible that PG may also have direct toxic effects on blood and other tissue cells. Immunosuppressive effects on natural killer cells and neutrophils [16] and skeletal muscle damage [17] have been described. We believe that etomidate in propylene glycol causes haemolysis. However, further studies on a larger number of volunteers, receiving only the solvent, are necessary to confirm this hypothesis.