Anesth Analg 2003;96:46-50
© 2003 International Anesthesia Research Society
CARDIOVASCULAR ANESTHESIA
Validity of Arterial and Mixed Venous Oxygen Saturation Measurements in a Canine Hemorrhage Model After Resuscitation with Varying Concentrations of Hemoglobin-Based Oxygen Carrier
Fedor Lurie, MD, PhD*,
Bernd Driessen, DVM, PhD
,
Jonathan S. Jahr, MD
,
Rashell Reynoso
, and
Robert A. Gunther, PhD||
*John A. Burns School of Medicine, University of Hawaii, Honolulu, Hawaii; School of Veterinary Medicine, University of Pennsylvania, Philadelphia, Pennsylvania; Department of Anesthesiology, UCLA, and King/Drew Medical Center, Los Angeles, California; University of California-Davis College of Letters and Science, Davis, California; and ||Department of Surgery, University of California-Davis, Davis, California
Address correspondence and reprint requests to J. S. Jahr, MD, UCLA Department of Anesthesiology, Box 951778, Los Angeles, CA 90095-1778. Address e-mail to jsjahr{at}mednet.ucla.edu
 |
Abstract
|
|---|
In this study, we evaluated the validity of saturation measurements in mixed venous and arterial blood during posthemorrhagic anemia and resuscitation with varying levels of hemoglobin-based oxygen carrier (Hemoglobin glutamer-200 [bovine]; Oxyglobin® [Hb-200]). Nineteen anesthetized, splenectomized, mixed-breed dogs were anesthetized (two were excluded from the data because they did not survive the exsanguination, supporting the validity of the model). Their pulmonary arteries were cannulated with the Abbott QVUE Oximetrix 3 catheter. An 18-gauge catheter was placed in the femoral artery, and a reusable Nellcor probe was applied to the tongue. Mixed venous and arterial samples were drawn at baseline, after 40% hemorrhage (to keep arterial pressure at 50 mm Hg), and postresuscitation with 30 mL/kg of 6% hetastarch in lactated Ringer’s solution (n = 4), 10 mL/kg of Hb-200, 20 mL/kg of hetastarch (n = 6), 20 mL/kg of Hb-200, and 10 mL/kg of hetastarch (n = 7). Samples were compared with oxygen content from the LEXO2CON-K oxygen analyzer, and oxygen content was calculated for all values from the monitors. Results were compared by using analysis of variance. There was good correlation (0.97 
r 0.92) for the measured versus calculated hemoglobin oxygen saturation values at baseline. After resuscitation, the correlation between calculated and measured values of oxygen content was significantly smaller for all tested instruments. The values of oxygen content calculated from the oxygen saturation monitor and from the oximetric pulmonary artery can deviate by as much as 20% from directly measured values. We conclude that the administration of this oxygen therapeutic may interfere with the values of some monitors.
IMPLICATIONS: This study evaluated oxygen saturation monitors in a canine model of acute blood loss and resuscitation with a blood substitute and found that these may interfere with the monitors’ results in a dose-dependent way.
|
Introduction
|
|---|
Hemoglobin (Hb)-based oxygen carriers (HBOCs) may be ideal candidates for fluid resuscitation (1). Three are undergoing or have completed Food and Drug Administration (FDA) Phase III clinical trials for use in humans, and one of the HBOCs (Hemoglobin glutamer-200 [bovine], Hb-200, Oxyglobin®; Biopure Corp., Cambridge, MA) has been approved by the FDA for the treatment of anemia in dogs (2). A related product (Hb glutamer-250 [bovine], HBOC-201, Hemopure®; Biopure Corp.) has been approved for human use in South Africa. The clinical indication for HBOCs is restoration of an adequate oxygen-carrying capacity of the blood, particularly in shock patients with acute posthemorrhagic anemia subsequent to resuscitations with nonoxygen-carrying solutions.
Hb oxygen saturation (SO2) is usually determined by cooximetry or blood gas analysis. It can also be monitored in real-time by using oximetric pulmonary artery catheters or pulse oximeters. The use of heterogeneous Hb (such as Hb-200) adds another dimension to the complexity of interpreting blood-gas and cooximetry data (3–6). As a result of the glutaraldehyde-polymeri-zation process, Hb-200 and HBOC-201 have a different oxyhemoglobin dissociation curve compared with native bovine or human Hb (7). This is a source of potential error in interpreting cooximeter analyzer data, especially in samples with low SO2, such as mixed venous blood (8,9). The optical properties of Hb-200 are also different from those of human Hb. Cooximeters, oximetric catheters, and pulse oximeters base their measurements on the optical spectrum of human Hb. The Instrumentation Laboratories (Lexington, MA) cooximeter applies a photometric method operating with light in the wavelength band of 500–670 nm (i.e., 535.0, 585.2, 594.5, and 626.6 nm) to measure concentrations of Hb fractions and SO2. The technology is based on the phenomenon that light absorption of oxygenated Hb (oxyhemoglobin) is different from deoxyhemoglobin. A similarly designed cooximeter, the Nova CO-Oximeter (Nova Biomedical Corp., Waltham, MA) accurately measures SO2 in canine blood after infusion of HBOCs such as Hb-200 (bovine) (3). The Instrumentation Laboratories cooximeter used in this study operates with similar technology to the Nova CO-Oximeter.
However, real-life physiologic conditions in situations of hemorrhage and during resuscitation are very different from those ex vivo. This study was performed to test the correlation between arterial and mixed venous SO2 measurements (compared with gold standard measurements) in mixed venous and arterial blood during posthemorrhagic anemia and resuscitation with Hb-200.
|
Methods
|
|---|
After receiving protocol approval from the University of California-Davis Animal Use and Care Committee, 19 anesthetized, splenectomized, mixed-breed dogs of both genders were studied (2 were removed because of inability to resuscitate, and 1 extra in the hetastarch group was added; therefore, there were initially 6 subjects in each group, with 2 mortalities in Group 1). After the induction of anesthesia with IV propofol and diazepam, the pulmonary artery was cannulated through the jugular vein, and the cephalic vein was cannulated (10,11). Temperature was maintained as normal for dogs, which is between 38°C and 39°C, by using various warming techniques. The position of the pulmonary artery catheter was documented by pulmonary artery pressure monitoring and by obtaining consistent pulmonary artery occlusion pressures with balloon inflation. Arterial saturations were obtained in real time by placing a reusable Nellcor pulse oximeter probe on the tongue and periodically changing its position on the tongue to ensure a lack of compression artifacts. Mixed venous blood samples were slowly and anaerobically collected (after removal of all flush solution in the tubing) from the distal port of the pulmonary artery catheter while systemic arterial blood samples were simultaneously withdrawn from a femoral arterial catheter. Blood samples were collected at 4 time points: baseline, 1 h after completion of a 30-min hemorrhage (approximately 40% of circulating blood volume, to keep the mean arterial blood pressure at 50 mm Hg), immediately, and then 3 h after resuscitation. In Group 1 (n = 4), resuscitation was accomplished with 30 mL/kg of 6% hetastarch in lactated Ringer’s solution (Hextend®; Abbott Laboratories, Chicago, IL). In Group 2 (n = 6), 10 mL/kg of Hb-200 (bovine) with 20 mL/kg of 6% hetastarch was used. In Group 3 (n = 7), 20 mL/kg of Hb-200 with 10 mL/kg of 6% hetastarch was used to resuscitate the subjects. Samples were immediately evaluated in triplicate on the LEXO2CON-K oxygen fuel cell (Hospex Fiberoptics, Chestnut Hill, MA) and averaged. Quantitative measurements of oxygen content (CO2) were performed with an oxygen-specific electrode, also called a fuel cell (LEXO2CON-K). This instrument measures the total of both dissolved and bound oxygen in blood, molecule by molecule, and therefore is accurate in all ranges with all blood of any species, regardless of whether or not the blood sample contains any stroma-free Hb in addition to red blood cell Hb. For this reason, the oxygen fuel cell is considered the gold standard technique for the quantification of O2ct (see Appendix).
SO2 values from the oximetric pulmonary artery catheter were obtained continuously during the experiments and recorded at the time of the blood sampling. The total Hb concentration was measured in all samples by using the cooximeter (Instrumentation Laboratories). All stroma-free Hb in plasma was considered to represent the bovine HBOC fraction (bHb), and its concentration was measured after centrifugation by using a HemoCue® hemoglobin photometer (HemoCue AB, Angelholm, Sweden) (6). Cellular (canine red blood cell) Hb (cHb) was calculated as the difference between total Hb and Hb in plasma. On the basis of the measurements obtained with the investigated instruments (Instrumentation Laboratories cooximeter, Nellcor pulse oximeter probe, and Abbott oximetric pulmonary artery catheter), CO2 was calculated for each sample by using the following equation (3): CO2 = (SO2) x (1.32bHb + 1.39cHb) + (PO2 x 0.003), where SO2 = hemoglobin oxygen saturation, bHb = HBOC (bovine hemoglobin) concentration, and cHb = canine hemoglobin concentration. The constants 1.32 and 1.36 represent the theoretical oxygen-binding capacities of 1 g of bHb and cHb calculated by molecular weight (68,000 and 64,458 Da, respectively). Results were compared with the CO2 from the LEXO2CON-K. Statistical analysis was performed with SPSS 10.1 software (SPSS Inc., Chicago, IL). Descriptive statistics, one-way analysis of variance, and linear correlation were used with a significance level of P 
0.05.
|
Results
|
|---|
Of 19 subjects studied, 2 in the hetastarch group did not tolerate the exsanguination and died before resuscitation, leaving only 4 in the hetastarch group to study. Table 1 lists pH and Hb levels during four study periods. In baseline blood samples, the CO2 calculated from the SO2 values was well correlated with values measured directly by the LEXO2CON instrument (0.92 r 0.97; P < 0.01; Table 2). The mean differences between measured values and calculated values of CO2 were 5.5% ± 1.9% for the pulmonary artery catheter, 0.4% ± 2.3% for the cooximeter, and 2.1% ± 0.7% for the pulse oximeter and were not significantly different.
View this table:
[in this window]
[in a new window]
|
Table 2. Correlation Coefficients Between Measured and Calculated Values of Arterial (Pulse Oximeter) and Mixed Venous (PA Catheter and Cooximeter) Oxygen Content
|
|
Immediately after the hemorrhage period, the correlation remained high (Table 2), and the difference between calculated and measured values of CO2 remained within the same range (4.0% ± 2.9% for the cooximeter, 4.9% ± 2.7% for the catheter, and 2.6% ± 0.94% for the pulse oximeter). One hour after resuscitation, a difference among the three treatment groups was detected. In animals resuscitated with 30 mL/kg of hetastarch (treatment Group 1), the differences between calculated and measured CO2 remained within the same range as at baseline (Tables 3–5) and were not significantly different. In the animals resuscitated with 10 mL/kg of Hb-200 and 20 mL of 6% hetastarch (Group 2), the mean difference between the pulmonary artery catheter and measured values reached 21%, and the mean difference between the pulse oximeter and measured values reached 9.1%. This difference was even larger in the group resuscitated with 20 mL/kg of Hb-200 and 10 mL/kg of hetastarch (Group 3). The correlation between measured and calculated values of CO2 was significantly less for all tested instruments (Table 2).
View this table:
[in this window]
[in a new window]
|
Table 3. Difference Between Measured and Calculated Values of Mixed Venous Oxygen Content (% Measured Value; mean ± se) for the Abbott Oximetric Pulmonary Artery Catheter
|
|
View this table:
[in this window]
[in a new window]
|
Table 4. Difference Between Measured and Calculated Values of Mixed Venous Oxygen Content (% Measured Value; mean ± se) for the Instrumentation Laboratories Cooximeter
|
|
View this table:
[in this window]
[in a new window]
|
Table 5. Difference Between Measured and Calculated Values of Arterial Oxygen Content (% Measured Value; mean ± se) for the Nellcor Oxygen Saturation Monitor
|
|
Three hours after resuscitation, there was no significant difference among the three treatment groups (with the exception of the co-oximeter), and correlation between the calculated and measured values of CO2 was high (Table 2). However, in cases of the Nellcor pulse oximeter and the Instrumentation Laboratories cooximeter, the differences between the calculated and measured values were significantly higher than at baseline (P < 0.05).
|
Discussion
|
|---|
The use of heterogeneous Hb solutions in resuscitation after acute hemorrhage presents a new challenge to the monitoring of important physiological values such as the SO2 of mixed venous blood. The equipment usually used for this purpose is based on the estimation of the optical properties of blood, assuming the spectrum of human Hb as a standard. Introduction of Hb solutions with different spectral characteristics causes a potential measurement error.
In our ex vivo studies, mixtures of canine blood with HBOC in different concentrations and with different levels of oxygenation were tested to validate laboratory equipment, including the cooximeter. Those experiments did not detect any significant measurement error caused by the introduction of polymerized bovine Hb (3,4). Recent work by Ali et al. (9) demonstrated that in multiple cooximeters and with multiple HBOCs (including Hb-200), instruments gave less accurate but clinically useful measurements in the presence of HBOCs.
In this study, however, we detected that when a HBOC is used for resuscitation after acute hemorrhage in dogs, the values of CO2 calculated from the cooximeter, oximetric pulmonary artery catheter, and pulse oximeter can deviate from directly measured values. This difference increases with an increasing dose of infused HBOC and is present during the first hour after infusion. The difference was detected immediately after resuscitation and then declined over time. This seems to parallel the dramatic changes in vital physiologic variables associated with hemorrhage and resuscitation, such as changes in circulating volumes, electrolytes, and acid-based balance (10,11). From Table 5, it is noteworthy that there is still a 10% difference in CO2 between the measured and calculated CO2, on the basis of the Nellcor oxygen saturation reading. This has been noted clinically and has obvious clinical implications for physicians and nurses caring for patients receiving HBOCs. In conclusion, the behavior of polymerized Hb solutions in these conditions has not been sufficiently studied and deserves further investigation.
Appendix: Oxygen Fuel Cell Technology
According to the LEXO2CON manual, the classical method by which all other oxygen content measurements are judged, is the Van Slyke method (12–15). The Van Slyke method involves an apparatus, which extracts all the gas from a blood sample in a sealed chamber of fixed volume. The partial pressure is measured. The oxygen is absorbed by a chemical combination and the reduced partial pressure is measured. The difference between the two pressures yields the quantity of absorbed oxygen, and hence the oxygen content of the sample. This method is capable of extreme accuracy. However, the technique is very cumbersome and can only process 10 samples in a day.
The LEXO2CON utilizes a galvanic cell that produces an electric current in proportion to oxygen molecules present. Using a carrier gas of nitrogen with hydrogen that is completely free of oxygen, the sample is mixed with the carrier gas and comes into equilibrium with the carrier gas. This gas passes into the fuel cell, and this cell combines with the oxygen and releases electrons. These electrons form the electric current that is proportional to the oxygen in the cell. The sum of these electrons is proportional to the total oxygen content in the sample. Eventually all of the oxygen contained in the blood sample is removed and passed to the cell. The cell and integrating circuits count all the oxygen molecules released.
There are some papers listed comparing the LEXO2CON to the Van Slyke Analyzer (13–15). It appears that the LEXO2CON is a close second in terms of accuracy in measuring oxygen contents. Below is a description of samples in duplicate from our laboratory, demonstrating the tight correlation as well as the manufacturer’s accuracy ranges.
LEXO2CON-K specifications are (12):
Accuracy 2% ± 0.1 Volumes %
resolution 0.1 Volumes %
Reproducibility 2% ± 0.1 volumes %
Linearity 2% ± 0.1 Volumes % from 0 to 99.9
Volumes %
|
References
|
|---|
- Dietz NM, Joyner MJ, Warner MA. Blood Substitutes: fluid, drugs or miracle solutions? Anesth Analg 1996; 82: 390–405.[Abstract]
- Food and Drug Administration: 21CFR522.1125 hemoglobin glutamer-200 (bovine). 63 Federal Register 11598, March 10, 1998.
- Jahr JS, Driessen B, Lurie F, et al. Oxygen saturation mea-surements in canine blood containing hemoglobin glutamer-200 (bovine): in-vitro validation of the NOVA CO-Oximeter. Vet Clin Pathol 2001; 30: 39–45.[Web of Science][Medline]
- Jahr JS, Lurie F, Driessen B, et al. Validation of oxygen saturation measurements in a canine model of hemoglobin-based oxygen carrier (HBOC) infusion. Clin Lab Sci 2000; 13: 173–9.
- Jahr JS, Lurie F, Gosselin R, et al. Effects of Hemoglobin glutamer-250 (bovine) (HBOC-201, Hemopure) on coagulation testing. Am J Ther 2002; 9: 431–6.[Medline]
- Jahr JS, Lurie F, Driessen B, et al. The HemoCue®, a point of care B-hemoglobin photometer, accurately measures hemoglobin levels when mixed in vitro with canine plasma and three hemoglobin-based oxygen carriers (HBOC). Can J Anaesth 2002; 49: 243–8.[Web of Science][Medline]
- Hughes GS, Antal EJ, Locker PK, et al. Physiology and pharmacokinetics of a novel hemoglobin-based oxygen carrier in humans. Crit Care Med 1996; 24: 756–64.[Web of Science][Medline]
- Nierman DM, Schechter CB. Mixed venous O2 saturation: measured by co-oximetry versus calculated from PVO2. J Clin Monit 1994; 10: 39–44.[Medline]
- Ali AA, Ali GS, Steinke JM, Shepard AP. Co-oximetry interference by hemoglobin-based blood substitutes. Anesth Analg 2001; 92: 863–9.[Abstract/Free Full Text]
- Cheung ATW, Jahr JS, Driessen B, et al. The effects of hemoglobin glutamer-200 (bovine) on the microcirculation in a canine hypovolemia model: a noninvasive computer-assisted intravital microscopy study. Anesth Analg 2001; 93: 832–8.[Abstract/Free Full Text]
- Driessen B, Jahr JS, Lurie F, et al. Effects of hemoglobin-based oxygen carrier hemoglobin glutamer-200 (bovine) on intestinal perfusion and oxygenation in a canine hypovolemia model. Br J Anaesth 2001; 86: 683–92.[Abstract/Free Full Text]
- Hoxpex. LEXO2CON-K operating instructions and maintenance manual. Chestnut Hill, MA: 1980.
- Valeri CR, Zaroulis CG, Marchionni L, Patti KJ. A simple method for measuring oxygen content in blood. J Lab Clin Med 1972; 79: 1035–40.[Medline]
- Selman BJ, White YS, Tait AR. An evaluation of the LexO2Con content analyzer. Anesthesia 1975; 30: 206–11.
- Kusumi F, Butts WC, Nuff WL. Superior analytical performance by electrolytic cell analysis of blood O2 content. J Appl Physiol 1973; 35: 299–300.[Free Full Text]
Accepted for publication September 25, 2002.
This article has been cited by other articles:
|
|
|
|
 
T. O. Morgan
Cost, quality, and risk: Measuring and stopping the hidden costs of coronary artery bypass graft surgery
Am. J. Health Syst. Pharm.,
September 15, 2005;
62(18_Supplement_4):
S2 - S5.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. K. Khan, J. S. Jahr, S. Nesargi, S. J. Rothenberg, Z. Tang, A. Cheung, R. A. Gunther, G. J. Kost, and B. Driessen
Does Lead Interfere with Hemoglobin-Based Oxygen Carrier (HBOC) Function? A Pilot Study of Lead Concentrations in Three Approved or Tested HBOCs and Oxyhemoglobin Dissociation with HBOCs and/or Bovine Blood with Varying Lead Concentrations
Anesth. Analg.,
June 1, 2003;
96(6):
1813 - 1820.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
