Abstract
Diaspirin crosslinked hemoglobin (DCHb) is a new blood substitute manufactured from human blood. To evaluate its microvascular filtration properties, we infused DCLHb into unanesthetized sheep (10%, 20 ml/kg) and measured the flow and composition of lung and soft tissue lymph. For comparison, we also infused human serum albumin (HSA; 10%, 20 ml/kg). DCLHb raised systemic and pulmonary arterial pressures from baseline values of 83 ± 7 and 13 ± 2 mm Hg, respectively, to peak values of 113 ± 9 and 26 ± 3 mm Hg (p < 0.05 versus baseline). These increases were significantly greater than those associated with HSA, which raised systemic and pulmonary arterial pressures from baseline values of 86 ± 4 and 13 ± 2 mm Hg, respectively, to peak values of 97 ± 3 and 21 ± 7 mm Hg (p ⩽ 0.05 versus baseline and versus DCLHb). These differences reflect the known pressor properties of DCLHb. Accordingly, DCLHb raised lung and soft tissue lymph flows to peak values of 12.2 ± 3.8 and 1.6 ± 0.7 ml/30 min, respectively, while HSA raised lung and soft tissue lymph flows to peak values of 7.5 ± 4.8 and 4.6 ± 1.9 ml/30 min, respectively (p ⩽ 0.05 versus DCLHb). The half-times of DCLHb equilibration from plasma into lung and soft tissue lymph of 1.0 ± 0.3 and 2.1 ± 1.1 h, respectively, were significantly faster than HSA equilibration half-times of 3.1 ± 0.2 and 3.8 ± 0.9 h. Filtration differences between DCLHb and HSA appear to be due to the pressor properties DCLHb.
Diaspirin crosslinked hemoglobin (DCLHb; Baxter Healthcare, Round Lake, IL) is a new blood substitute composed of stroma-free, chemically modified, human hemoglobin. It has oxygen binding characteristics similar to those of whole blood, is free of viral contamination, and can be infused without cross-matching (1-7).
With a molecular mass of 64 kilodaltons (kD), DCLHb should have osmotic properties similar to those of albumin (66 kD). Thus, DCLHb might be thought of as a colloid intravenous volume expander with the added ability of being able to transport oxygen (8-9). However, DCLHb has pharmacologic pressor properties that also distinguish it from other colloid solutions such as albumin and hetastarch (10-12).
The oxygen transporting properties of DCLHb have been extensively characterized (1-6), but its properties as a colloid volume expander have not. For example, it is not known how readily DCLHb filters into lung and soft tissue lymph, nor is it known how its filtration characteristics compare with those of established colloid solutions such as albumin. Filtration properties are important because the theoretical advantage of colloids over crystalloid solutions such as Ringer's lactate is that the colloid macromolecules are retained within the circulation, minimizing the risk of pulmonary and systemic edema. However, the pressor properties of DCLHb might augment filtration, and result in less plasma retention than a molecule of comparable mass such as albumin. DCLHb filtration might also provide a mechanism for oxygen transport out of the circulation. Thus, characterizing the filtration properties of DCLHb is important.
The goal of the studies we report here was to address this. We infused DCLHb into unanesthetized sheep, measured pulmonary and systemic intravascular osmotic and hydrostatic pressures, and also measured lung and soft tissue lymph flows. We used these data to determine the forces responsible for DCLHb filtration, and the rates at which this filtration occurred. For comparison, we made similar measurements after infusing albumin into a second set of sheep. Although solutions of DCLHb and albumin had similar osmotic pressures, we found the DCLHb entered lymph two to three times faster.
We prepared 12 sheep (45.2 ± 5.3 kg) with chronic lung and soft tissue lymph fistulas according to the methods described previously (13). Six of these sheep received DCLHb, and six received human serum albumin (HSA). Anesthesia was induced with sodium thiopental (10 ml/ kg, intravenously) and maintained by ventilation with a mixture of 1% halothane in oxygen. To collect lung lymph, we performed a right thoracotomy, and cannulated the efferent duct of the caudal, mediastinal lymph node using a silastic catheter (602-0150; Dow Corning, Midland, MI) inserted into the proximal end of the node. The catheter was routed through the chest wall for lymph collection. We cauterized diaphragmatic surface lymphatics and ligated the distal end of the node to prevent contamination of the lung lymph (14). To cannulate the prefemoral node, we made a small flank incision, identified the node, and ligated the efferent lymph vessel as originally described by Demling and coworkers (15). We inserted a silastic catheter into the vessel between the node and the ligature, and exteriorized the catheter through the skin. Vascular catheters were placed into the carotid artery to sample blood, and to measure mean aortic pressure (Pa) and heart rate (HR). An additional catheter was inserted into the jugular vein for infusion. After surgery, the sheep were housed in study cages where they had free access to food and water, and were allowed to recover for 4 d. One day prior to study, a Swan-Ganz catheter (7F; Edward Laboratories, Irvine, CA) was placed in the pulmonary artery for measurement of pulmonary arterial pressure (Ppa), pulmonary arterial wedge pressure (Ppaw), central venous pressure (Pven), cardiac output (CO), and temperature (T). All pressures were measured with transducers referenced to the height of the left atrium, and recorded with a physiograph (SpaceLabs, Redmond, WA).
Five days after surgery, we gathered baseline data for 3 h (Day 0). At 30-min intervals we collected plasma and timed lymph samples, and also measured Pa, Ppa, Ppaw, CO, Pven, and HR. The next morning (Day 1), plasma, lymph, and hemodynamic data were collected every 30 min for 1 h to confirm the baseline values measured the previous day. The animals were then infused (20 ml/kg) over 1 h with either DCLHb (n = 6) or HSA (n = 6). Hemodynamics and lymph dynamics were measured every 30 min for the next 6 h. The next day (Day 2) hourly samples were collected for 8 h (h 21 to 29).
DCLHb, supplied to us by Baxter Healthcare, is prepared from units of voluntarily donated, outdated human blood. The stroma-free hemoglobin is reacted with bis(3,5-dibromosalicyl) fumarate to covalently link the lysine99 moieties of the alpha subunits (1-3). This stabilizes the hemoglobin molecule as a tetramer and prevents degradation into nephrotoxic alpha-beta dimers that do not transport oxygen. The extent of crosslinking is greater than 99% (10). The solution is heat pasteurized, and ultrafiltered to remove any potential red cell or viral fragments. It is mixed with balanced electrolytes to produce a solution with a final hemoglobin concentration of 10%, oxygen half-saturation pressure of hemoglobin (P-50) of 32 mm Hg, an oxygen carrying capacity of 1.39 ml/g, and a pH 37° C of 7.4 (10). The methemoglobin concentration is less than 10%, and the endotoxin activity, as measured by Limulus amebocyte lysate assay, is less than 0.1 EU/ml. The solution is supplied frozen, and can be stored for up to 1 yr. We warmed it to room temperature before infusion.
We obtained units of HSA (10%) from our hospital pharmacy and added Evans blue dye (Sigma) to produce a final Evans blue concentration of 0.005%. This resulted in an Evans blue–albumin molar ratio of 1 to 30.
Evans blue–HSA concentrations in plasma and lymph were measured by spectrophotometry (610 nm). Hemoglobin concentrations were measured by reacting the samples with Drabkin's reagent (Sigma), then measuring the concentration of the resulting cyanmethemoglobin solution by spectrophotometry (540 nm). These data were used to calculate the rates of disappearance of HSA and DCLHb from the plasma, and their rates of equilibration with lymph.
We also used the concentration data to estimate the osmotic pressures contributed by DCLHb and HSA to plasma and lymph. To perform these calculations we first measured osmotic pressures of DCLHb and HSA in samples of known concentration, then fit least-squares linear regression equations through the DCLHb and HSA data sets (Figure 1) (16). The DCLHb and HSA concentrations measured in each sample of plasma and lymph were applied to these equations to estimate the fractional osmotic pressures contributed by DCLHb and HSA to each sample. Note that plasma and lymph are mixed macromolecular solutions. Molecular interactions in such solutions produce total osmotic pressures that are less than the sum of individual pressures of each component (17). For this reason the pressures we calculated using the regression equations should be considered to be an estimate of the fractional osmotic pressure contributed by DCLHb or HSA to each plasma and lymph sample.
Fig. 1. Osmotic pressures (π, mm Hg) produced by 0.5 to 5.0% solutions of DCLHb and HSA. After measuring concentrations of DCLHb or HSA in samples of plasma and lymph, the regression equations were used to calculate the osmotic pressures contributed by each solution.
[More] [Minimize]In plasma and lymph samples collected from animals infused with DCLHb, we used a co-oximeter (Instrumentation Laboratory) to measure methemoglobin concentrations. Methemoglobin does not transport oxygen, and we expected that some of the DCLHb might be converted to methemoglobin. Therefore, we made these measurements to evaluate the effective oxygen transporting characteristics of DCLHb in both plasma and lymph. Methemoglobin concentrations are expressed as a percentage of the total hemoglobin concentration. The co-oximeter also reported total hemoglobin concentrations; we found that these values equaled those measured using Drabkin's reagent.
We measured the macromolecular osmotic pressures of plasma and lymph samples using an osmometer (Wescor 4420, Logan, UT) equipped with a cellophane membrane that allowed the passage of molecules smaller than 10,000 daltons. The osmometer was calibrated with a bovine serum albumin standard (Osmocoll II; Wescor).
To estimate the rates of disappearance of DCLHb and HSA from plasma, and rates of entry into lymph, we plotted plasma and lymph concentrations, as well as the plasma–lymph concentration difference. When plotted with time on the horizontal axis, and the natural logarithm of the concentration on the vertical axis, the slopes of the plasma concentration, and plasma–lymph concentration difference are, respectively, the rate of disappearance from plasma, and the rate of equilibration with lymph. The slope of each plot is equal to the rate constant (k), expressed by the relationship:
| Equation 1 |
where C0 is the initial concentration and C is the concentration at any time t (16). We used the intercept of each plot with the vertical axis as C0. To complete the calculations, we fit a least squares linear regression equation through each data set and calculated k from the slope (Figure 2) (18). We then calculated t from the relationship t = 1/k. We expressed the rate of equilibration as a half-time (T1/2), calculated as T1/2 = 0.692t (16).
Fig. 2. Half-times of disappearance of DCLHb and HSA from plasma (T1/2, P), and equilibration into lung and soft tissue lymph (T1/2, P-L ). Half-times are calculated from the slopes of the least-squares regressions plotted through, respectively, the plasma and plasma-minus-lymph data. The regressions were fitted to the data that followed the maximum value in each set (16). The confidence intervals represent the 95% confidence limits of the slopes.
[More] [Minimize]Results are expressed as mean ± SD. For each solution (DCLHb or HSA) results were compared by one-way analysis of variance among baseline, infusion (0–1 h), postinfusion (1–2 h), end of Day 1 (3–5 h), and Day 2 (22–28 h). Within each of these periods, results were also compared between DCLHb and HSA by unpaired t test. Differences were considered significant at p ≦ 0.05. The comparisons were performed using a computer-based statistics program (18).
All sheep tolerated the DCLHb and HSA infusions without noticeable discomfort.
DCLHb infusion caused a significant increase in Pa from 83.1 ± 6.5 mm Hg at baseline to 113.2 ± 8.7 mm Hg at 1 h (p ⩽ 0.05), but Pa gradually returned toward baseline over the next 4 h (Figure 3). HSA infusion caused a transient increase in Pa from 85.8 ± 4.0 mm Hg at baseline to 97.2 ± 2.5 mm Hg at 1 h (p ⩽ 0.05), but Pa quickly returned toward baseline once the HSA infusion was complete. At 1 h, Pa in DCLHb treated animals was significantly greater than with HSA.
Fig. 3. Systemic arterial pressures (Pa) and central venous pressures (Pven) following infusion (shaded areas) of DCLHb and HSA (mean ± SD). Baseline data are 2-h averages. *Significantly different from baseline; †significantly different from HSA (p ⩽ 0.05).
[More] [Minimize]DCLHb and HSA infusions raised Pven by an average of 6.5 ± 4.5 mm Hg above baseline (p ⩽ 0.05 compared with baseline; NS, DCLHb versus HSA) (Figure 3). Pven returned to baseline within 4 h.
DCLHb infusion caused a significant increase in Ppa from 13.0 ± 1.7 mm Hg at baseline to 26.4 ± 3.3 mm Hg at 1 h (p ⩽ 0.05); Ppa returned to 15 ± 1.4 mm Hg by 5 h (NS) (Figure 4). HSA infusion increased Ppa from baseline (13.2 ± 1.8 mm Hg) to 20.7 ± 7.2 mm Hg at 1 h (p ⩽ 0.05); this was significantly less than the rise in Ppa after DCLHb. At 5 h after HSA infusion Ppa was still significantly elevated above baseline (19.0 ± 8.1 mm Hg; p ⩽ 0.05), but Ppa returned to baseline values by 24 h.
Fig. 4. Pulmonary arterial (Ppa) and pulmonary arterial wedge pressures (Ppaw) (mm Hg) following infusion (shaded areas) of DCLHb and HSA (mean ± SD). Baseline data are 2-h averages. *Significantly different from baseline; †significantly different from HSA (p ⩽ 0.05).
[More] [Minimize]DCLHb infusion raised Ppaw from baseline (4.0 ± 1.8 mm Hg) to 11.7 ± 8.5 mm Hg at 1 h (p ⩽ 0.05) (Figure 4), and Ppaw was still slightly elevated at 5 h (4.7 ± 1.7 mm Hg; p ⩽ 0.05). HSA increased Ppaw from baseline values (3.8 ± 1.3 mm Hg) to 8.8 ± 4.9 mm Hg at 1 h (p ⩽ 0.05; NS versus DCLHb), but Ppaw returned to baseline by 5 h.
DCLHb infusion decreased CO significantly from baseline (4.6 ± 0.7 L/min) to 3.8 ± 1.0 L/min at 1 h and to 3.4 ± 0.8 L/ min by 5 h (p ⩽ 0.05) (Figure 5). CO returned to baseline by 24 h. HSA infusion increased CO slightly above baseline (5.4 ± 0.6 L/min) to 6.2 ± 0.5 L/min at 1 h and to 5.8 ± 0.4 at 5 h (NS). At 1 and 5 h, COs with DCLHb and HSA were significantly different.
Fig. 5. Cardiac outputs and heart rates in sheep infused with DCLHb or HSA. *Significantly different from HSA (p ⩽ 0.05).
[More] [Minimize]Infusion of DCLHb caused a modest decrease in heart rate from 114 ± 24 bpm at baseline to 100 ± 23 bpm at 1 h, and to 101 ± 18 bpm at 5 h. The pooled heart rate data for h 1 to 5 after DCLHb infusion were significantly lower than the baseline data (p = 0.04). HSA infusion caused no significant change in heart rate.
In animals infused with DCLHb, body temperature rose from 39.8 ± 0.3° C at baseline to 40.7 ± 0.7° C by 5 h (p ⩽ 0.05) (Table 1). HSA infusion produced no change in body temperature.
| DCLHb | Baseline | Infusion | Postinfusion | End Day 1 | ||||
|---|---|---|---|---|---|---|---|---|
| Hematocrit | 27.6 ± 4.0 | 23.6 ± 3.0* | 24.1 ± 2.9* | 25.2 ± 3.8* | ||||
| Arterial Po 2, mm Hg | 89.6 ± 6.3† | 91.3 ± 5.1 | 92.7 ± 11.5 | 95.2 ± 2.6*,† | ||||
| pH | 7.45 ± 0.04 | 7.45 ± 0.04 | 7.44 ± 0.06 | 7.46 ± 0.03 | ||||
| Temperature, °C | 39.8 ± 0.3 | 39.8 ± 0.3 | 40.4 ± 0.6*,† | 40.7 ± 0.7*,† | ||||
| HSA | ||||||||
| Hematocrit | 28.0 ± 1.7 | 24.6 ± 3.5* | 25.5 ± 3.0* | 25.8 ± 3.0 | ||||
| Arterial Po 2, mm Hg | 104.7 ± 16.7 | 98.9 ± 9.4 | 104.3 ± 3.4 | 107.5 ± 2.5 | ||||
| pH | 7.46 ± 0.03 | 7.43 ± 0.03 | 7.42 ± 0.03 | 7.43 ± 0.03 | ||||
| Temperature, °C | 39.7 ± 0.4 | 39.7 ± 0.4 | 39.8 ± 0.5 | 39.8 ± 0.6 |
DCLHb infusion caused the hematocrit to decrease from baseline (27.6 ± 4.0) to 23.1 ± 2.8 by 1 h (p ⩽ 0.05), although it returned to 25.2 ± 3.7 by 5 h (NS) (Table 1). HSA infusion lowered the hematocrit from 28 ± 1.7 at baseline, to 23.0 ± 3.6 at 1 h (p ⩽ 0.05). The hematocrit was still significantly below baseline at 5 h (26 ± 2.9; p ⩽ 0.05). The hematocrits did not differ significantly between DCLHb and HSA at any time.
DCLHb and HSA infusions did not significantly alter arterial Po 2 or pH (Table 1).
DCLHb, HSA. Serial dilutions of DCLHb and HSA produced similar osmotic pressures (Figure 1). We used these data to calculate the osmotic pressure contributed by DCLHb or HSA to each plasma and lymph sample. After the concentration of DCLHb or HSA was measured in each sample, the fractional osmotic pressure was calculated from linear regression equations fit through the data in Figure 1. The resulting fractional osmotic pressures are shown in Figure 6. The concentrations are reported below.
Fig. 6. Macromolecular osmotic pressures in plasma, lung lymph, and soft tissue lymph after infusion (shaded area) of DCLHb and HSA. Total pressures and the pressures contributed by each infusate are shown (mean ± SD). *Significantly different from baseline; †significantly different from DCLHb (p ⩽ 0.05).
[More] [Minimize]Plasma. Following DCLHb infusion plasma osmotic pressure rose from baseline (16.7 ± 3.1 mm Hg) to 20.8 ± 5.0 mm Hg by 1 h (p ⩽ 0.05), and remained elevated at 5 h (19.4 ± 2.4 mm Hg; p ⩽ 0.05) (Figure 6). HSA increased plasma osmotic pressure from baseline (18.7 ± 3.1 mm Hg) to 27.1 ± 2.2 mm Hg by 1 h, and to 24.5 ± 2.1 mm Hg by 5 h (p ⩽ 0.05 versus baseline and DCLHb). On Day 2 after DCLHb and HSA infusions, plasma osmotic pressures were, respectively, 17.9 ± 2.5 and 22.5 ± 2.4 mm Hg (p ⩽ 0.05).
DCLHb contributed 8.4 ± 2.4 mm Hg to the plasma osmotic pressure at 1 h, and 5.0 ± 1.4 mm Hg at 5 h (Figure 6). HSA contributed 9.4 ± 1.2 mm Hg to the plasma osmotic pressure at 1 h and 7.0 ± 0.7 mm Hg at 5 h (p ⩽ 0.05 versus DCLHb). On Day 2, DCLHb and HSA contributed, respectively, 2.1 ± 1.5 and 4.0 ± 1.1 mm Hg to the plasma osmotic pressure (p ⩽ 0.05).
Lung lymph. After DCLHb infusion, lung lymph osmotic pressure fell from baseline (13.7 ± 1.7 mm Hg) to 10.3 ± 3.3 mm Hg at 1 h (p ⩽ 0.05) then returned to 13.1 ± 3.0 mm Hg by 5 h (NS) (Figure 6). HSA infusion caused lung lymph osmotic pressure to rise from baseline (14.9 ± 2.1 mm Hg) to 15.5 ± 2.2 at 1 h (p ⩽ 0.05 versus DCLHb) and to 18.7 ± 3.0 mm Hg at 5 h (p ⩽ 0.05 versus baseline and DCLHb).
The osmotic pressure contributed by DCLHb to lung lymph was 2.4 ± 0.5 mm Hg at 1 h and 2.8 ± 1.0 mm Hg at 5 h (Figure 6). The osmotic pressure contributed by HSA to lung lymph was 1.5 ± 1.0 mm Hg at 1 h and 4.5 ± 0.5 mm Hg at 5 h (p ⩽ 0.05 versus DCLHb). On Day 2, DCLHb and HSA contributed, respectively, 2.5 ± 1.4 and 2.8 ± 0.5 mm Hg to the lung lymph osmotic pressure (NS).
Soft tissue lymph. Infusion of DCLHb caused soft tissue osmotic pressure to increase slowly from baseline (9.8 ± 2.6 mm Hg) to 11.0 ± 4.1 mm Hg at 5 h (NS). HSA infusion caused soft tissue osmotic pressure to increase from 13.3 ± 2.2 mm Hg at baseline to 15.2 ± 1.3 mm Hg at 1 h; at 5 h it remained elevated at 16.9 ± 0.9 mm Hg (p ⩽ 0.05 versus baseline and DCLHb).
The osmotic pressure contributed by DCLHb to soft tissue lymph was 0.2 ± 0.9 mm Hg at 1 h and 1.6 ± 0.2 mm Hg at 5 h (Figure 6). The osmotic pressure contributed by HSA was 0.1 ± 0.3 mm Hg at 1 h and 2.1 ± 0.2 mm Hg at 5 h (p ⩽ 0.05 versus DCLHb). On Day 2, DCLHb and HSA contributed, respectively, 0.9 ± 0.3 and 2.1 ± 0.1 mm Hg to the soft tissue lymph osmotic pressure (p ⩽ 0.05).
Lung lymph. Following DCLHb infusion, lung lymph flow increased from baseline (4.2 ± 1.1 ml/30 min) to a peak of 12.2 ± 3.8 ml/30 min at 1.5 h (Figure 7), and returned to 7.3 ± 3.5 ml/ 30 min at 5 h (p ⩽ 0.05 versus baseline) (Figure 7). On Day 2, lung lymph flow averaged 5.1 ± 1.1 ml/30 min (NS). Following HSA infusion, lung lymph flow increased from baseline (2.9 ± 1.1 ml/30 min) to a peak of 7.5 ± 4.8 ml/30 min at 1.5 h, (p ⩽ 0.05 versus baseline and DCLHb) and was 4.8 ± 0.6 ml/30 min at 5 h (p ⩽ 0.05 versus baseline and DCLHb). On Day 2, lung lymph flows averaged 4.0 ± 0.2 ml/30 min (NS).
Fig. 7. Lung and soft tissue lymph flows following infusion of DCLHb and HSA (mean ± SD). Baseline data are 2-h averages. *Significantly different from baseline (p ⩽ 0.05); †significantly different from HSA (p ⩽ 0.05).
[More] [Minimize]Soft tissue lymph. Infusion of DCLHb raised soft tissue lymph flow from baseline (1.0 ± 0.1 ml/30 min) to a peak of 1.6 ± 0.7 ml/30 min at 1.5 h (NS), but soft tissue lymph flow then fell to 0.8 ± 0.5 ml/30 min at 5 h (NS) (Figure 7). On Day 2, soft tissue lymph flows averaged 0.7 ± 0.6 ml/30 min (NS). HSA infusion increased soft tissue lymph flow from baseline (2.5 ± 0.5 ml/30 min) to a peak of 4.6 ± 1.9 ml/30 min at 1.5 h (p ⩽ 0.05 versus baseline and DCLHb) and to 4.7 ± 1.8 ml/30 min at 5 h (p ⩽ 0.05 versus baseline and DCLHb). Soft tissue lymph flows averaged 4.8 ± 1.7 ml/30 min on Day 2 (p ⩽ 0.05).
Plasma concentrations of DCLHb and HSA reached peak values of 2.5 ± 0.7 and 2.8 ± 0.5 g/dl, respectively (NS). The half-times (T1/2) for DCLHb and HSA disappearance from plasma were, respectively, 3.4 ± 1.2 and 9.7 ± 1.4 h (p ⩽ 0.05) (Figure 2). DCLHb and HSA equilibrated from plasma into lung lymph with half-times, respectively, of 1.0 ± 0.3 and 3.1 ± 0.2 h (p ⩽ 0.05). The half-times of DCLHb and HSA equilibration into soft tissue lymph from plasma were, respectively, 2.1 ± 1.1 and 3.8 ± 0.9 h (p ⩽ 0.05). Plasma concentrations of DCLHb and HSA on Day 2 were, respectively, 0.8 ± 0.4 and 1.2 ± 0.2 g/dl.
On Days 1 and 2, respectively, methemoglobin accounted for 22.9 ± 2.5% and 23.5 ± 7.7% of the total hemoglobin concentration in plasma (Table 2). In lung lymph, methemoglobin accounted for 20.0 ± 3.7% and 27.6 ± 12.5%, respectively, of the total hemoglobin concentration. Due to small sample volumes, we were unable to measure methemoglobin concentrations in soft tissue lymph.
| Hour | Plasma | Lung Lymph | ||
|---|---|---|---|---|
| 0–6 | 22.9 ± 2.5 | 20.0 ± 3.7 | ||
| 20–30 | 23.5 ± 7.7 | 27.6 ± 12.5 |
As might be expected for solutes of equal molecular mass, we found that DCLHb and HSA solutions had equal macromolecular osmotic pressures (Figure 1). DCLHb and HSA also had similar effects on Pven and Ppaw, suggesting that both solutions increased intravascular filling pressures equally. Thus, from the viewpoint of hydrostatic and osmotic volume expansion, both solutions demonstrated similar properties.
Both DCLHb and HSA filtered into lymph, contributing osmotic pressures that ranged from 21 to 24% of the total lung lymph osmotic pressure, and from 12 to15% of the total soft tissue lymph osmotic pressure (Figure 6). However, DCLHb entered lung lymph threefold faster than HSA, and entered soft tissue lymph nearly twofold faster (Figure 2). Furthermore, DCLHb increased lung lymph flow more than HSA (Figure 7). Thus, DCLHb and HSA had similar osmotic effects, but the rates of liquid and solute flux into lymph differed between the two solutions.
These liquid and solute filtration rate differences appear to be due to the pressor effects of DCLHb. DCLHb raised Pa significantly more than HSA throughout Day 1 (Figure 3). DCLHb also briefly raised Ppa more than HSA, although after the infusions were completed, Ppa values between the two solutions were similar.
Filtration results from a balance of osmotic and hydrostatic forces across the microvascular filtration barrier as expressed by the net filtration pressure: PNF = PMV − 0.8(Δπ), where PMV is the calculated microvascular hydrostatic pressure and Δπ is the plasma-to-lymph osmotic pressure difference. Using the data in Figures 3 and 4, we calculated PMV for the pulmonary circulation as 0.4(Ppa − Ppaw) + Ppaw (19), and for the systemic circulation as 0.1(Pa − Pven) + Pven (9).
After calculating PNF for the two solutions (Table 3), we found that DCLHb raised peak pulmonary PNF by nearly twofold above baseline, and raised systemic PNF by nearly threefold. HSA produced a slight decrease in pulmonary PNF, but raised systemic PNF by about half. These differences were due to the fact that DCLHb raised systemic and pulmonary arterial pressures more than HSA (Figures 3 and 4). Systemic PNF for DCLHb was still significantly elevated above baseline at 5 h, although pulmonary PNF was not.
| Pulmonary Circulation | ||||||||
|---|---|---|---|---|---|---|---|---|
| Baseline | Peak | 5 h | Day 2 | |||||
| DCLHb | 5.2 ± 1.1 | 9.4 ± 2.9* | 4.7 ± 1.0 | 4.4 ± 0.7 | ||||
| HSA | 4.5 ± 0.8 | 4.1 ± 1.1 | 3.2 ± 0.6 | 3.6 ± 0.4 | ||||
| Systemic Circulation | ||||||||
| Baseline | Peak | 5 h | Day 2 | |||||
| DCLHb | 2.5 ± 0.2 | 6.9 ± 1.9* | 4.3 ± 0.6* | 2.4 ± 0.3 | ||||
| HSA | 2.5 ± 0.4 | 3.8 ± 1.6 | 2.3 ± 0.2 | 1.5 ± 0.2 | ||||
Pressor effects associated with DCLHb have been shown to be related to dose, although not linearly. Malcolm and colleagues infused 62.5 mg/kg of DCLHb into conscious rats and found that mean arterial pressure rose by 5 to 6 mm Hg (20). Doses of 125 to 500 mg/kg produced pressure rises of 30 to 40 mm Hg. However, doses of 1,000 to 4,000 mg/kg produced pressure rises similar to those at 500 mg/kg. Thus, the pressure increase reached a limit at 25 to 30% above baseline. This is consistent with our findings: we infused 2,000 mg/kg and found that mean arterial pressure rose by 30 mm Hg; Ppa rose by 13 mm Hg. Pressor effects are not unique to DCLHb, and appear to be a property of nearly all acellular hemoglobin solutions.
Our Ppa data are in striking contrast to those reported by Figueiredo and coworkers, who infused 400 mg/kg of an α–α diaspirin crosslinked hemoglobin solution into hemorrhaged pigs (21). They reported that Ppa rose by 28 mm Hg, to a peak value of 44 mm Hg. This represents a twofold greater rise in Ppa pressure than we measured in normal animals into which we infused a fivefold greater volume of solution. One possible explanation for this contrast between Figueiredo's results and ours is that pigs may respond to α–α diaspirin crosslinked hemoglobin differently than sheep or rats. This hypothesis is supported by results of Hess and colleagues who infused 2,475 mg/kg of α–α hemoglobin to resuscitate pigs from hemorrhagic shock. They reported that Pa and Ppa rose by 39 and 20 mm Hg above their respective values during hemorrhage (22). We recently reported on the effects of using DCLHb to resuscitate sheep from hemorrhagic shock. After infusing 3,500 mg/ kg we found that Pa rose by 25 mm Hg (23). Together, these results suggest that pigs may be more sensitive to hemoglobin solutions than sheep.
The pressor properties of DCLHb in the peripheral circulation have been associated with adrenergic mechanisms, endothelin, and nitric oxide (10). Evidence for the role of adrenergic mechanisms was reported by Gulati and colleagues who found that the pressor effect persisted in both cervically sectioned and adrenalectomized rats, suggesting that this effect was not mediated at the level of either the CNS or the adrenals, but at the level of the resistance vessels (24). They reported that DCLHb potentiated the pressor effect of norepinephrine, but found that this could be blocked by phenoxybenzamine. DCLHb also potentiated the pressor effects of phenylephrine and clonidine, but these were blocked by pretreatment with prazosin and yohimbine, respectively (24, 25). Gulati and colleagues concluded that DCLHb potentiated both α1 and α2 adrenoreceptors in the peripheral vascular system of rats.
Roles for endothelin and nitric oxide (NO) were reported by Schultz and colleagues who found that the pressor effects of DCLHb could be attenuated by pretreatment with phosphoramidon, a proendothelin inhibitor (26). They also found that the pressor effect could be reversed by infusion of l-arginine, the substrate for NO, and by nitroglycerine, an NO donor. They further reported that cyanomet DCLHb, which does not react with NO, had no pressor effect. These results suggest that, in addition to effects on the α1 and α2 adrenoreceptors, DCLHb pressor effects are also mediated by endothelin and inhibition of NO.
We found that Pven and Ppaw rose immediately after infusion of both solutions, as would be expected from the addition of 20 ml/kg of liquid to the circulation. However, cardiac output and heart rate decreased following infusion of DCLHb. We believe these may have been baroreceptor-mediated reflex responses to the pressor effects of the hemoglobin solution.
We believe the pressor properties of DCLHb are also responsible for its rapid clearance from the circulation, and rapid equilibration with lung and soft tissue lymph compared with HSA. However, the plasma clearance half-times we measured were significantly shorter than those reported previously. Hess and colleagues infused 7 ml/kg of DCLHb into rats, rabbits, and monkeys, and reported plasma clearance T1/2 values of, respectively, 4.4, 12.5, and 16 h (27). Variations in responses among species may account for the range of values they measured, as well as for differences between their data and ours.
The rise in lymph flows we measured following DCLHb infusion contrasts with reports that extracellular hemoglobin paralyzes lymphatic vessels (28). Elias and colleagues reported that red cell erythrolysate modulated lymphatic pumping. Using isolated, perfused sheep intestinal lymphatics they showed that erythrolysates with hemoglobin concentrations of 10−5 to 10−8 M reduced lymphatic pumping by 40 to 100%. They speculated that erythrolysate hemoglobin was responsible for this effect. In contrast, we found that plasma DCLHb concentrations of 10−4 M produced peak lung lymph flows that were 60% higher than those obtained with HSA. A possible explanation for the discrepancy between our observations and those of Elias and colleagues is related to purity. DCLHb is a highly purified solution produced under conditions consistent with human clinical application (6, 7). The red cell erythrolysate used by Elias and colleagues consisted principally of hemoglobin, but other components of this solution might also have been a factor in the modulation of lymphatic pumping they observed.
Endothelial macromolecular permeability is affected by net solute charge as expressed by the isoelectric point (pI). Swanson and Kern measured a pI for native rabbit albumin of 4.4 to 5.1 (29). These investigators also showed that cationic albumin, with a pI of 7.2 to 8.0, had a permeability–surface area (PS) product in isolated rabbit lungs that was threefold greater than that of native albumin. DCLHb has pIs in the oxy and deoxy forms, respectively, of 7.00 and 7.04 (1). If the data of Swanson and Kern can be extended to intact sheep lungs, they suggest that the shorter T1/2 we measured for DCLHb may be partially attributable to differences in pI between DCLHb and albumin.
We labeled albumin with Evans blue to measure plasma– lymph albumin equilibration rates. However, the use of Evans blue to label albumin has been challanged by Dallal and Chang. They measured PS products for Evans blue labeled albumin in isolated rat lungs that were fivefold higher than PS products measured using 125I labeled albumin (30). Their explanation was that Evans blue had become bound to tissue proteins. The Evans blue concentration they used (Evans blue– albumin molar ratio, 1 to 1) was thirtyfold higher than that we used (Evans blue–albumin molar ratio, 1 to 30). An excess of Evans blue in their solutions might explain their findings. We cannot rule out the possibility that Evans blue became bound to tissue proteins in our studies. If such binding occurred, it would have retarded the entry of Evans blue into lymph and might explain our data showing that albumin entered lymph more slowly than DCLHb. However, Arakawa and colleagues labeled albumin with 125I and measured plasma-to-lung lymph equilibration half-times of 2.2 ± 0.6 h (31). This is similar to the value of 3.1 ± 0.2 h we measured, and suggests that our Evans blue labeling procedures were adequate.
Filtration of DCLHb out of the circulation suggests that this acellular hemoglobin solution might transport oxygen close to sites of mitochondrial metabolism in peripheral tissue. However, the magnitude of such transport is likely to be small. We measured peak DCLHb concentrations in soft tissue lymph of 0.6 g/dl (Figure 2), but because methemoglobin probably accounted for 20 to 30% of these concentrations (Table 3), the effective hemoglobin concentrations were lower. Soft tissue lymph flows reached a maximal value of 1.6 ml/30 min. If we assume that hemoglobin has an oxygen carrying capacity of 1.39 ml O2 per gram, then multiplying the effective hemoglobin concentration by the lymph flow results in an oxygen transport into soft tissue lymph of less than 0.01 ml/30 min. At the Po 2 of peripheral capillary blood, the transport would be less. Thus, convective transvascular oxygen transport associated with DCLHb filtration is likely to be insignificant.
However, diffusive oxygen transport to tissue might be facilitated by DCLHb. Hemoglobin solutions are known to augment O2 diffusion as a consequence of reversible binding of O2 to heme (32). This implies that the presence of DCLHb in the diffusion pathway between capillary blood and metabolizing cells might result in augmented oxygen diffusion to those cells. The ability of hemoglobin solutions to augment diffusion has been shown to be inversely proportional to Po 2 (32). Thus, at the low Po 2 values typically present in peripheral tissue, the presence of DCLHb might significantly facilitate oxygen diffusion.
In conclusion, DCLHb has oncotic properties similar to those of other macromolecular blood substitutes including HSA. However, unlike other blood substitiutes DCLHb has the pharmacologic properties of a pressor and the oxygen transporting properties of a hemoglobin solution. The oxygen transporting properties of DCLHb differ subtlely from those of whole blood because DCLHb readily crosses the microvascular filtration barrier. Furthermore, because it is acellular, DCLHb may have access to compressed or injured microvascular regions that exclude red cells. In summary, DCLHb appears to be a new blood substitute with unique properties that distinguish it from whole blood as well as from other previously available blood substitutes.
The authors thank Barbara Imm and Virginia Korent for their technical assistance.
Supported by grants from Baxter Healthcare, the Department of Veterans Affairs, and the National Institutes of Health (HL46236).
| 1. | Chatterjee R., Welty E. V., Walder R. Y., Pruitt S. L., Rogers P. H., Arnone A., Walder J. A.Isolation and characterization of a new hemoglobin derivative crosslinked between the a chains (lysine 99a1–lysine 99a2). J. Biol. Chem261198699299937 |
| 2. | Vandergriff K. D., Medina F., Marini M. M., Winslow R. M.Equilibrium oxygen binding to human hemoglobin crosslinked between the a chains by bis(3,5-dibromosalicyl) fumarate. J. Biol. Chem.26419891782417833 |
| 3. | Snyder S. R., Welty E. V., Walder R. Y., Williams L. A., Walder J. A.HbXL99a: a hemoglobin derivative that is cross-linked between the a subunits is useful as a blood substitute. Proc. Natl. Acad. Sci.84198772807284 |
| 4. | Bush S. L., Nelson D. J., Shelbourn T. E.New blood product: diaspirin crosslinked hemoglobin as a blood substitute (abstract). Transfusion32S19928S |
| 5. | Bush S., Knight R., Kray M., Moore E.DCLHb (diaspirin crosslinked hemoglobin) 10% and electrolyte injection: significant and consistent deliverable oxygen content (abstract). Artif. Cells Blood Substit. Immobil. Biotechnol241996313 |
| 6. | Nelson D., Azari M., Brown R., Burhop K., Bush S., Catarello J., Chuang H., Downing C., Estep T., Lowen A., McClure K., McDaniel A., Michalek E., Mozier N., Rohn K., Spicuzza J., Zieske P., Zimmerman G.Preparation and characterization of diaspirin cross-linked hemoglobin solutions for preclinical studies. Biomat. Artif. Cells Immobil. Biotech.201992423427 |
| 7. | Azari M., Catarello J., Burhop K., Camacho T., Ebeling A., Estep T., Guzder S., Krause K., Marshall T., Kohn K., Sarajari R.Validation of the heat treatment step used in the production of diaspirin crosslinked hemoglobin (DCLHb) for viral inactivation-effect of crosslinking (abstract). Artif. Cells Blood Substit. Immobil. Biotechnol241996303 |
| 8. | Arakawa M., Jerome E. H., Enzan K., Grady M., Staub N. C.Effects of Dextran 70 on hemodynamics and lung liquid and protein exchange in awake sheep. Circ. Res.671990852861 |
| 9. | Conhaim, R. L., D. J. Rosenfeld, M. A. Schreiber, D. M. Baaske, and B. A. Harms. 1993. Effects of intravenous pentafraction on lung and soft tissue liquid exchange in hypoproteinemic sheep. Am. J. Physiol. 265(Heart Circ. Physiol. 34):H1536–H1543. |
| 10. | Przybelski, R. J., and E. K. Daily. 1994. The pressor/perfusion effect of diaspirin cross-linked hemoglobin (DCLHb). Yearbook Intensive Care Emerg. Med. 252–263. |
| 11. | Amberson W. R., Jennings J. J., Rhode M.Clinical experience with hemoglobin-saline solutions. J. Appl. Physiol.11949469489 |
| 12. | Katsuyama S. S., Cole D. J., Drummond J. C., Bradley K.Nitric oxide mediates the hypertensive response to a modified hemoglobin solution (DCLHb) in rats. Artif. Cells Blood Substit. Immobil. Biotechnol22199417 |
| 13. | Staub N. C., Bland R. D., Brigham K. L., Demling R. H., Erdmann A. J., Woolverton W. C.Preparation of chronic lung lymph fistulas in sheep. J. Surg. Res191975315320 |
| 14. | Drake R., Adair T., Traber D., Gabel J.Contamination of caudal mediastinal node efferent lymph in sheep. Am. J. Physiol.2411981H354H357 |
| 15. | Demling R. H., Smith M., Gunther R., Wandzilak T., Pederson N. C.Use of a chronic prefemoral fistula for monitoring systemic capillary integrity in unanesthetized sheep. J. Surg. Res.311981144146 |
| 16. | Glantz, S. A. 1979. Mathematics for Biomedical Applications. University of California Press. 19–23. |
| 17. | Michel, C. C. 1984. Fluid movements through capillary walls. In E. M. Renkin and C. C. Michel, editors. Handbook of Physiology, Section 2: The Cardiovascular System, Vol. IV, Part 1, Chapter 9, p. 382. American Physiological Society, Bethesda, MD. |
| 18. | Feldman, D. S., J. Gagnon, R. Hofmann, and J. Simpson. 1991. Statview II, version 1.03. Abacus Concepts, Berkeley, CA. |
| 19. | Gaar K. A., Taylor A. E., Owens L. J., Guyton A. C.Pulmonary capillary pressure and filtration coefficient in the isolated perfused lung. Am. J. Physiol.2131967910914 |
| 20. | Malcolm, D. S., I. N. Hamilton, S. C. Schultz, F. Cole, and K. Burhop. 1994. Characterization of the hemodynamic response to intravenous diaspiring crosslinked hemoglobin solution in rats. Art. Cells, Blood Subs., Immob. Biotech. 22:91–107. |
| 21. | de Figueiredo L. F. P., Mathru M., Solanki D., Macdonald V. M., Hess J. R., Kramer G. C.Pulmonary hypertension and systemic vasoconstriction may offset the benefits of acellular hemoglobin blood substitutes. J. Trauma421997847854 |
| 22. | Hess J. R., Macdonald V. W., Brinkley W. W.Systemic and pulmonary hypertension after resuscitation with cell-free hemoglobin. J. Appl. Physiol.74199317691778 |
| 23. | DeAngeles D. A., Scott A. M., McGrath A. M., Korent V. A., Rodenkirch L. A., Conhaim R. L., Harms B. A.Resuscitation from hemorrhagic shock with diaspirin cross-linked hemoglobin in unanesthetized sheep. J. Trauma421997406414 |
| 24. | Gulati A., Rebello S.Role of adrenergic mechanisms in the pressor effect of diaspirin cross-linked hemoglobin. J. Lab. Clin. Med.1241994125133 |
| 25. | Gulati A., Sharma A. C.Prazosin blocks the pressor but not the regional circulatory effects of diaspirin crosslinked hemoglobin. Life Sci.551994121130 |
| 26. | Schultz S. C., Grady B., Cole F., Hamilton I., Burhop K., Malcolm D. S.A role for endothelin and nitric oxide in the pressor response to diaspirin cross-linked hemoglobin. J. Lab. Clin. Med.1221993301308 |
| 27. | Hess J. R., Fadare S. O., Tolentino L. S. L., Bangal N. R., Winslow R. M.The intravascular persistance of crosslinked human hemoglobin. Prog. Clin. Biol. Res.3191989351360 |
| 28. | Elias R. M., Wandolo G., Ranadive N. S., Eisenhoffer J., Johnston M. G.Lymphatic pumping in response to changes in transmural pressure is modulated by erythrolysate/hemoglobin. Circ. Res.67199010971106 |
| 29. | Swanson, J. A., and D. F. Kern. 1994. Characterization of the pulmonary endothelial charge barrier. Am. J. Physiol. 266(4, Pt. 2):H1300–H1303. |
| 30. | Dallal M. M., Chang S. W.Evans blue dye in the assessment of PS area product in perfused rat lungs. J. Appl. Physiol.77199410301035 |
| 31. | Arakawa M., Jerome E. H., Enzan K., Grady M., Staub N. C.Effects of dextran 70 on hemodynamics and lung liquid and protein exchange in awake sheep. Circ. Res.671990852861 |
| 32. | Scholander P. F.Oxygen transport through hemoglobin solutions. Science1211960585590 |
