| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
We measured lead concentrations in three hemoglobin-based oxygen carriers (HBOCs; Oxyglobin®, Hemopure®, and HemolinkTM) and compared them with lead concentrations from blood-bank blood. Oxyhemoglobin dissociation was measured with large concentrations of lead in bovine HBOC, with or without bovine blood, and in bovine blood. Samples of each were prepared by combining one with normal saline (control), the second with small lead concentrations (22 µg/dL), and the third with toxic lead concentrations (70 µg/dL). They were blended in 2 tonometers at oxygen concentrations (2.5%, 5%, 8%, 10%, 21%, and 95%) with 5% CO2 and the remainder nitrogen for 5 min per sample after a 15-min wash-in with each level of oxygen and were measured with co-oximetry. Oxygen saturation was plotted against PO2, fitting fourth-order polynomial nonlinear regression to the data. The lead concentrations of the three HBOCs were 0.51, 0.22, 0.40 µg/dL. There were no clinically important differences of the oxyhemoglobin dissociation curves as a function of lead concentration. The lead concentrations of the three tested HBOCs were small and no larger than the average for blood-bank blood. The presence of increasing concentrations of lead in either concentrated solution of bovine HBOC or a 1:1 mixture of bovine HBOC and native bovine blood does not appear to affect hemoglobin oxygenation in an acute in vitro model of increased lead concentrations. IMPLICATIONS: Gunshot wounds rapidly increase circulating lead concentrations. Lead concentrations are small in three hemoglobin-based oxygen carriers (HBOCs), and HBOCs and/or bovine blood do not appear to be affected by lead concentrations in terms of immediate oxygen on-loading and off-loading. HBOCs may be useful in patients with gunshot wounds.
Hemoglobin-based oxygen carriers (HBOC) have been studied for >50 yr; two are approved for either veterinary or human use, and one is in active Phase III trials (1,2). HBOCs have been studied extensively in animal models (35). It would be advantageous if the HBOCs were approved for use in human trauma patients. King/Drew Medical Center (KDMC), one of the major trauma centers in the nation, cares for a large number of trauma victims, a large percentage of whom have gunshot wounds (GSW) (6). One recent study from KDMC documented that in a sample of 1000 packed red blood cell bags from KDMC, 2 had significantly increased concentrations of lead (7). This leads to an additional concern with blood-bank blood, especially because blood is not routinely checked for lead concentrations (7). A pilot study demonstrated acute interference in coagulation with increased lead concentrations; therefore, we hypothesized that there may be interference with oxygen loading and off-loading (8). New evidence reveals that in patients shot with lead bullets, lead concentrations may increase rapidly, within 24 h, and reach toxic levels (9). If increased lead concentrations interfere acutely with loading and off-loading oxygen in vertebrate hemoglobin, then HBOCs may not be of clinical value in this setting. To test the hypothesis that rapidly increasing lead concentrations interfere with oxygen loading and off-loading in hemoglobin with and without HBOCs, we performed three separate, but related, studies. Three HBOCs were tested for baseline lead concentrations and compared with historical controls from packed red blood cells (7). Bovine blood was tested with and without varying concentrations of added lead, and oxyhemoglobin dissociation curves were generated. With an HBOC, with and without bovine blood, with varying concentrations of lead added, oxyhemoglobin dissociation curves were also generated. Our model assumed that the increased concentrations in lead would immediately affect oxygen loading and off-loading. If there was no interference from rapidly increased lead concentrations with oxygen loading and off-loading in HBOCs with vertebrate blood, then HBOCs may be a source of oxygen therapeutics for the acute resuscitation of gunshot victims, as well as any victim of massive trauma needing transfusion. Therefore, it is reasonable to surmise that oxygen unloading may also be affected acutely, because lead, a heavy metal, may interfere with the hemoglobin moiety. However, this is the hypothesis of the study that has been tested in a pilot series of experiments.
The lead concentrations in the HBOCs were analyzed with a PerkinElmer (Norwalk, CT) 4100ZL Zeeman atomic absorption spectrometer, which was equipped with a graphite furnace, correcting for background interference (7). Two IL tonometers (Model 237; Instrumentation Laboratory, Lexington, MA) were used to equilibrate the PO2 and the PCO2 in the samples with a gas mixture of oxygen, CO2, and nitrogen (balance) in the tank. For the procedure, a cuvette is filled with the sample and set into an equilibration chamber. The equilibration chamber is surrounded by a thermostat-controlled water bath with a preset temperature. During the equilibration process, prewarmed humidified gas flows through the chamber. At the same time, the cuvette undergoes a constant two-stage cycle, creating a thin film of blood in the cuvette that allows more blood to be exposed to the gas. Each tonometer was connected to a gas tank (Puritan Medical Products, Overland Park, KS) containing oxygen in concentrations of 2.5%, 5%, 8%, 10%, 21%, and 95%. Each tank contained 5% CO2 and nitrogen for balance. The tonometry gases have certified standards from the vendor with a guaranteed error of ±0.03%. Co-oximetry is a photometric method used to determine the hemoglobin oxygen saturation. The principle is based on the phenomenon of light absorption at different wavelengths for oxygenated hemoglobin and deoxygenated hemoglobin. Concentrations of hemoglobin fractions, saturation, and oxygen content were measured with the Radiometer OSM co-oximeter (on "4" cattle mode; Radiometer Medical A/S, Copenhagen, Denmark). Prior work by our group has validated a similar co-oximeter with HBOC (10). The LEXO2CON-K oxygen analyzer (Hospex Fiberoptics, Chestnut Hill, MA) measured the oxygen content (11). For the first experiment, three HBOCs were used: hemoglobin glutamer-200 (bovine); HBOC-200, Oxyglobin® (Biopure Corp., Cambridge, MA); hemoglobin glutamer-250 (bovine); HBOC-201, Hemopure® (Biopure Corp.); and hemoglobin raffimer, HemolinkTM (Hemosol Inc., Toronto, ON). The Oxyglobin was purchased from a veterinary pharmacy, and the Hemopure and Hemolink were given to us for laboratory validation (12,13); they were refrigerated aseptically and anaerobically at 10°C and tested for colloid oncotic pressure with a Wescor Inc. (Logan, UT) onco- meter by using a 10,000-µm pore membrane. The colloid oncotic pressures were 25 and 26 mm Hg, which are similar to the published values, and allowed for the assumption that the samples had not undergone protein or other degradation (13). The bovine blood for the second and third set of experiments was from a healthy cow (No. 130) from the laboratory of the University of California-Davis Veterinary Medical Teaching Hospital. The blood was drawn during a routine blood-work analysis. The remainder of the sample was given to us on the day of the experiment, drawn in gray top tubes (VacutainerTM, containing potassium oxalate and sodium fluoride; Becton Dickinson, Franklin Lakes, NJ), placed in ice immediately, and studied the same day. In the first set of experiments, we measured lead concentrations in three HBOCs. Each sample was analyzed in triplicate, and the results were averaged. Historical data were used to compare them with 1000 U of blood-bank blood for lead concentrations (7). In the second set of experiments, remainder samples of single bovine healthy donor blood were obtained, and baseline lead concentrations were measured. Three sets of samples were prepared at each oxygen level: 2 mL of bovine blood with 100 µL of normal saline (control), 2 mL of bovine blood with large lead concentrations (70.4 µg/dL) added at a 100-µL volume, and 2 mL of bovine blood with small lead concentrations (23.5 µg/dL) at a 100-µL volume. The samples were placed in a cuvette, and after a 15-min equilibration period at 37°C, an aliquot of blood was drawn into a syringe anaerobically. The sample was immediately analyzed in the LEXO2CON-K co-oximeter at 37°C. The second part of this experiment was performed by using the same protocol with undiluted HBOC (hemoglobin glutamer-200) and no, small, and large lead concentrations, as in the above experiment. Only one oxygen content measurement was performed at each lead concentration (pilot data). The pH of the lead nitrate was 1.5, but the small volume use would not be expected to change the pH or buffering capacity of the tested solution.
Oxygen saturations were calculated from the oxygen content with the following formula:
The Hüfner factor of bovine blood is 1.32. Oxygen saturations were plotted against PO2, and oxyhemoglobin dissociation curves were derived by using a fourth-order polynomial nonlinear regression to the data with Kelmans model (14). The HBOC samples were also tested on the Radiometer co-oximeter. The co-oximeter measured the hemoglobin, oxyhemoglobin saturation, carboxyhemoglobin, methemoglobin, and oxygen content. In the third set of experiments, a newly opened bag of hemoglobin glutamer-200 (bovine) showed methemoglobin concentrations to be small (0.9%1.1%). With the same protocol, three samples with 2 mL of HBOC-200 with no, small, and large lead concentrations were analyzed with a Radiometer co-oximeter. The second group of samples used 1 mL of HBOC-200 and 1 mL of bovine blood with no, small, and large lead concentrations by using the same protocol. Tonometry on each of these samples was performed at six different oxygen concentrations as in the previous experiment (one sample each).
The atmospheric pressure on the day of the experiment was recorded at 763 torr. The three sets of samples were tested at each of the six oxygen concentrations. The actual PO2 in the samples was calculated as
where Patm is the atmospheric pressure, PH2O is the water vapor pressure at 37°C, and %O2 is the oxygen concentration of the gas mixture in the gas tank. Oxygen saturations were obtained from the co-oximeter. In the second and the third set of experiments, internal validation was performed by two methods:
Oxyhemoglobin dissociation curves were generated by plotting oxygen saturation against PO2, fitting the fourth-order polynomial nonlinear regression to the data modeled after the Kelman model (14). Because the data collected were single-point data, the P50 (the partial pressure of oxygen at which 50% of hemoglobin is saturated) was interpolated and compared with published values. This provided a guide to the reliability of the data generated. Because of single-set data collection, the oxyhemoglobin dissociation curves were evaluated from a clinically important standpoint without providing statistical feedback on the curves. P50 values from each of the derived curves were interpolated and compared with values in the literature to help validate the data generated. The comparisons of the hemoglobin dissociation curves were based on the polynomial equations derived for the data sets, and the curves were compared by using the polynomial equations for differences.
Baseline lead concentrations were measured: hemoglobin glutamer-200 (bovine) was 0.51 µg/dL, hemoglobin glutamer-250 (bovine) was 0.22 µg/dL, and hemoglobin raffimer was 0.40 µg/dL (Table 1). In 0.5% of samples of blood-bank blood tested, lead concentrations were found to be more than 10 µg/dL, and in two samples they exceeded 30 µg/dL (7). Had those two units been transfused to premature or neonate patients, the patients receiving those units might have been at additional high risk of lead toxicity.
Oxygen saturations in bovine blood with varying lead concentrations are documented in Table 2. The oxyhemoglobin curve is shown in Figure 1. No clinical differences were found among the three sets of data, on the basis of the comparison of the polynomial equations.
In the third experiment with HBOC-200, with or without bovine blood, the oxygen saturations are documented in Table 2. The figures are as shown (Figs. 2 and 3).
Results of validations are as follows.
We interpolated the P50 values (PO2 at 50% saturation) for bovine blood, HBOC-200, and HBOC-200/bovine blood from the oxyhemoglobin dissociation curves. The following are the calculated PO2 pressures at 50% saturation (P50):
The published value of bovine P50 is 26 mm Hg (15). We feel confident that our analysis was valid because of the tight fit of this value to our interpolated value. The manufacturers P50 value for HBOC-200 is 35 mm Hg (13), which fits well with the 50:50 HBOC-200/bovine blood. However, the value we determined for pure HBOC-200 is larger. Others have also documented larger P50 values (16,17), including Alayash et al. (16), with levels of 46 mm Hg (as compared with our value of 58 mm Hg). One explanation for this deviation might be the autooxidation and possible formation of methemoglobin with the larger oxygen concentrations. We measured levels of methemoglobin in the tonometer with 21% oxygen (Table 3).
At least two HBOCs, hemoglobin glutamer-250 (bovine) (Hemopure) and hemoglobin raffimer (Hemolink), are either completed or are currently undergoing Food and Drug Administration (FDA) Phase III trials as oxygen therapeutics in humans (2). Hemopure has completed a multinational, multicenter pivotal Phase III trial (1820). Hemopure has been approved for human use in South Africa. Hemoglobin glutamer-200 (bovine) (Oxyglobin) has been approved by the FDA for canine anemia since 1998. Hemolink has been studied in Phase III trials in Canada and the United Kingdom in coronary artery bypass graft surgery patients (2). The results confirm that it is effective and safe (2). Our concern is with acutely increased lead concentrations in patients who might receive HBOCs for resuscitation and increasing oxygen-carrying capacity. Coagulation appears to be acutely interfered with increased lead concentration (8). This led us to wonder whether other acute changes in hematologic function also occur. To determine this, we undertook three related studies. In the first set of experiments, we measured lead concentrations in the three HBOCs. In the second set, single bovine healthy donor blood was measured for baseline lead concentrations, and samples with varying concentrations of lead were added. Oxygen contents were plotted against PO2 to create oxyhemoglobin dissociation curves. In the third set of experiments, an HBOC, with or without bovine blood and with varying lead concentrations, oxygen content was plotted against PO2 to create oxyhemoglobin dissociation curves. The experiments were performed with tonometry and co-oximetry.
We measured very small concentrations of lead in all the HBOC products (<1 µg/dL) as compared with the present average national blood lead concentrations ( All of our determinations of oxygen content of HBOC with added lead were undertaken within 10 minutes of preparation of the sample. We did not measure oxygen content hours after adding lead and are unable to state that lead in HBOC, either initially present in the product or introduced by lead from the patient, will interfere with oxygen carrying of the product over the estimated 18- to 24-hour duration of action of present HBOC products. We used an older method to derive oxyhemoglobin dissociation curves. A newer, more elegant system has been described by Vandegriff et al. (24). Here, protocatechuic acid is used to deoxygenate hemoglobin solutions enzymatically. If we had access to this technology, we could derive oxygen curves more accurately, because there is inherent error in transferring blood/hemoglobin samples from the tonometer to the LEXO2CON or co-oximeter. There was only one data point for each sample. Although lack of replicate measurements for these oxygen curves makes statistical comparison among the different lead conditions more challenging, comparison of the small differences among the curves (Fig. 1) reveals that they do not differ by more than the inherent variability of the measurements, as revealed by the replicate validation measurements at one oxygen level (Table 4). The data presented may be viewed only as preliminary and require corroboration, perhaps with the above model to confirm. Gunshot victims may have rapidly increasing blood lead concentrations. The concentrations vary depending on the area exposed, the amount of lead in the bullet, the length of time after wounding, and so on. Our preliminary findings indicate no systematic effect of lead on the oxygen-carrying capacity of HBOCs. If this proves true in human patients, HBOCs may be valuable in the treatment of hemorrhagic shock from GSWs without a risk of impaired acute oxygen-binding capacity from increased lead concentrations.
The authors wish to thank Katja Herrman, DVM, for technical support with the LEXO2CON; Richard Louie, Jessica Davis, Mario Manalo, and Rashell Reynoso for laboratory support; and Sara Faulds for her help with manuscript preparation. The authors would like to thank Hemosol, Inc., for the sample of HemolinkTM, and Biopure Corp. for the sample of Hemopure® donated for laboratory testing. In Memoriam This article is dedicated to Professor Ahsanul Karim Khan, MD, who recently died while presenting a seminar at the Institute of Postgraduate Medicine and Research, Dhaka, Bangladesh, his home country. His colleagues will miss him greatly.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|