Allosteric Modification of Oxygen Delivery by Hemoglobin : Anesthesia & Analgesia

Secondary Logo

Journal Logo


Allosteric Modification of Oxygen Delivery by Hemoglobin

Wahr, Joyce A. MD*,; Gerber, Michael MD†,; Venitz, Jürgen MD, PhD‡,; Baliga, Narayan MD*

Author Information
Anesthesia & Analgesia 92(3):p 615-620, March 2001. | DOI: 10.1213/00000539-200103000-00011
  • Free


Allosteric modification of the affinity of hemoglobin for oxygen occurs in a wide variety of physiologic conditions. Both acidosis and increased temperature cause a rightward shift of the oxygen dissociation curve, resulting in an enhanced release of oxygen to the tissues. Acute altitude acclimation and anemia result in increased production of 2,3-diphosphoglycerate, again resulting in increased tissue oxygenation. A rightward shift of the oxygen dissociation curve is indicative of decreased hemoglobin-oxygen-binding affinity and is assessed by the p50, the oxygen tension which results in 50% saturation of hemoglobin. Although there are variants of hemoglobin that have an altered p50, there are no pharmacologic means to control the release of oxygen to the tissues. Conditions in which such an approach could prove useful include local ischemia, such as stroke or myocardial infarction, or instances of global increases in oxygen demand (sepsis) or decreased oxygen delivery (low cardiac output).

In the 1980s, two antilipidemic drugs (clofibrate and bezofibrate) were found to decrease the affinity of hemoglobin for oxygen, thus increasing the release of oxygen to tissues (1,2). The demonstrated in vitro allosteric effects of these drugs, however, are inhibited in vivo by serum albumin, decreasing the potential for use in humans. With this background, efforts to design and synthesize a potent drug that would effect allosteric modification of hemoglobin without inhibition by albumin resulted in the development of the compound known as RSR13 (Figure 1) (3,4). Since its description in 1992, RSR13 has been tested in a variety of animal models with the following results: 1) RSR13 effects a reproducible, dose-dependent rightward shift of the oxygen dissociation curve (4); 2) RSR13 increases tissue Po2(5,6); 3) RSR13 reverses the cerebral arterial vasodilation associated with hypoxia (7); and 4) RSR13 increases brain oxygen tension and reduces infarct size in a cat stroke model (8).

Figure 1:
Chemical structure of RSR13.

This drug had not previously been investigated in patients undergoing general anesthesia or surgery, and this trial was undertaken to determine the pharmacodynamic and pharmacokinetic effects of RSR13 in humans in this setting, and to determine the effect of this drug on hemodynamic and laboratory variables.


With the approval of the University of Michigan IRB, patients who gave informed consent were entered into this randomized, double-blinded, placebo-controlled, single-dose, sequential, dose-escalating trial to investigate the safety, tolerance, pharmacokinetics, and pharmacodynamics of single IV doses of RSR13 in elective surgery patients. This dose-escalation trial was designed to determine the dose of RSR13 that would cause an increase in p50 of 10 mm Hg. Patients were eligible to participate if they were between 18 and 70 yr old, ASA physical status I–III, were undergoing elective surgery under general anesthesia, with an anticipated transfusion requirement of <3 units of packed red blood cells (RBCs), and were demonstrating adequate hematologic variables and pulmonary function (pulse oximetry > 90% breathing room air). Female patients of childbearing age were required to have a negative pregnancy test at screening. Patients were excluded from participating if they were New York Heart Association Class III or IV; had a history of any hemoglobinopathy or of drug or alcohol abuse; a screening creatinine, bilirubin, or liver transaminase of >1.5 times the upper limit of normal; or current use of or allergy to antilipemic medications.

Patients meeting the inclusion criteria and who gave informed consent were enrolled in this trial. The study drug was administered IV in 8 ascending doses of 10, 20, 30, 40, 50, 60, 75, and 100 mg/kg after the induction of anesthesia and concurrently with the commencement of the surgical procedure. RSR13, or a comparable volume of placebo, was infused for 30 min in the 10–60-mg/kg dose groups, and for 60 min for the 75- and 100-mg/kg dose groups. The placebo solution used in this study was sterile 0.9% saline. Patients were randomly allocated to RSR13 or placebo in the ratio of 2:1 for the first seven dose groups. The last dose group, 100 mg/kg, consisted of five patients, three who received RSR13, and two who received placebo.

Blood samples were taken for pharmacokinetic and pharmacodynamic determinations at preinfusion, 15, 30, 45, 90, and 105 min, and then at 2, 3, 4, 6, 8, 12, 24, and 48 h after the start of the infusion. Plasma and RBCs were separated and frozen for pharmacokinetic analysis. RSR13 concentrations were determined in plasma, in RBCs, and in urine at the specified time points. After the addition of the internal standard (RSR4), samples (plasma, RBCs, or urine) were deproteinized, hemolyzed, extracted into acidified chloroform, evaporated to dryness, and reconstituted in mobile phase. Quantitation was performed by using reverse-phase high-performance liquid chromatography with ultraviolet detection (254 nm). The limit of quantitation was 5 μg/mL by using a 0.1-mL aliquot of sample.

Plasma concentration-time data were analyzed by using noncompartmental methods (9). Concentration versus time profiles (linear and semilog) were plotted for each patient at each dose level. Individual pharmacokinetic variables (peak concentration in plasma, time of the peak concentration in plasma, total area under the plasma concentration time curve, apparent total body clearance in plasma, renal clearance, nonrenal clearance, and terminal half-life in plasma [t1/2(p)]) were calculated for each patient at each dose level. RBC concentration time data were analyzed in a similar manner. Urine data were also analyzed by using noncompartmental methods. The total amount excreted unchanged in urine was determined as the cumulative amount of RSR13 excreted at 48 h after the start of the infusion.

Heparinized blood samples were taken for determination of pharmacodynamic effect (p50 determination) before the induction; at preinfusion; and at 30 and 60 min and 2, 4, 6, 8, 12, 24, and 48 h after the start of the infusion. These whole blood samples were kept on ice and sent to the sponsor the same day for p50 determination by three-point tonometry. Each sample was split into three, and the subsequent aliquots were equilibrated in a tonometer at a target Po2 of 20, 40, and 60 mm Hg. After tonometry, SO2 of the three samples was determined by using a cooximeter. The three data pairs were fitted by using the following equation: SO2 = 100 × Po2n/(p50n + Po2n). Oxygen affinity (p50) and the Hill coefficient (n) were estimated by using nonlinear regression by using Scientist for Windows (Version 2.1; Micromath, Salt Lake City, UT).

Blood samples were taken for hematology, coagulation variables (prothrombin time, partial thromboplastin time), and serum chemistries (electrolytes, glucose, bicarbonate, urea nitrogen, creatinine, total protein, albumin, calcium, phosphorous, total bilirubin, alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, creatine kinase, lactose dehydrogenase, serum iron, and total iron binding capacity) at screening and at 4, 12, 24, and 48 h after the start of the infusion. If values for either creatine kinase or lactose dehydrogenase were increased, isoenzymes were determined. Creatinine clearance (CrCl) was determined from pooled urine collections from 0–24 and 24–48 h after the start of the infusion.

Vital signs (heart rate, respiratory rate, blood pressure, temperature, and SaO2) were recorded at baseline, every 5 min for the duration of surgery, and then per surgical routine. Supplemental oxygen was continued postoperatively until the patient’s SaO2 was more than 90% while breathing room air.

Descriptive narrative and/or statistics of serial physical examinations, vital signs, laboratory measurements, pulse oximetry, and treatment-emergent adverse events were used to assess the safety of RSR13 over time. The type and total number of treatment-emergent events were tabulated. As a result of the small sample size involved, no statistical analyses were conducted to determine differences among treatment groups.

For each patient, plasma, RBC, and urine drug concentration data were plotted versus time for all doses. Noncompartmental pharmacokinetic variables were estimated for each patient.


Twenty-six patients who gave informed consent were enrolled in the study. All patients successfully completed the protocol. Of these, 17 received RSR13, and 9 received placebo. The number of subjects receiving each dose is shown in Table 1. Patient characteristics are shown in Table 2. Although the groups were similar for most patient characteristics, and despite valid randomization techniques, the distribution of surgical procedures was different between the treatment arms. Four patients underwent nephrectomies, and 3 underwent prostatectomies; all 7 received RSR13. All of the 9 patients receiving placebo underwent orthopedic procedures.

Table 1:
Mean Changes in p50 at End-Infusion by Dose Group
Table 2:
Patient Characteristics

The mean shift in p50 in mm Hg at end-infusion is shown by dose group in Table 1. The target endpoint of a shift in p50 of at least 10 mm Hg was achieved in the 100-mg/kg dose group, with a mean maximum increase in p50 of 11.8 mm Hg. The mean time-response curves for changes in p50 for the dose groups are shown in Figure 2, and maximal response was achieved within 15 min of end-infusion in all dose groups.

Figure 2:
Mean time-response curves for changes in p50 for surgical patients receiving placebo, 50, 60, 75, or 100 mg/kg RSR13. Means (± sd) are shown for the placebo and 100-mg/kg groups only.

Pharmacokinetic data for plasma concentration are shown by dose group in Table 3. Because of the small sample sizes in the treatment groups, the range or actual values are presented rather than mean values. Comparison of plasma and RBC concentration-time profiles showed that the concentrations in RBC’s were smaller than those in plasma for all patients at most time points. Log plasma and RBC concentration-time profiles indicated that t1/2 (RBC) parallels t1/2 (plasma) for most patients. There was no apparent relationship between the plasma pharmacokinetic variables and the administered dose; however, the population variability across dose groups in total plasma clearance and volume of distribution of approximately 40% in these surgical patients was moderately large. There was no apparent gender pharmacokinetic difference; however, there were only 5 female patients compared with 12 male patients in the study. Plots of noncompartmental clearance (total, renal and nonrenal) versus the maximum change from baseline (during the first 24 h after dosing) in serum albumin concentration, hematocrit, and calculated CrCl suggested that RSR13 clearance (total, renal and nonrenal) was significantly influenced by renal function and potentially by hemodilution as a result of fluid administration during surgery. There were no obvious relationships between comedications and changes in RSR13 pharmacokinetics; however, the study was not designed to assess drug interactions with RSR13. No alterations in RSR13 pharmacokinetics or pharmacodynamics related to the concomitant use of anesthesia were observed in this study.

Table 3:
Range of RSR13 Pharmacokinetic Parameters from Plasma Concentrations

All Causality Adverse Events

Adverse events, although noted in virtually every patient, were rated as mild or moderate, and were comparable between RSR13 and placebo recipients. Fever was reported in more RSR13 patients (65%) than placebo patients (44%), but was transient, of only mild or moderate intensity, and may have been related to the concentration of more serious surgeries in the RSR13 treatment groups. The average number of adverse events per patient compared by treatment group did not reveal a dose-response relationship. Gastrointestinal, cardiovascular, and respiratory system events were comparable in both groups. There were three reports (18%) of hypoxia in RSR13 patients versus one (11%) in placebo patients, all of which resolved with supplemental oxygen and time.

Treatment-Related Adverse Events

Three patients experienced transient increases in serum creatinine. The first patient was a 59-yr-old man who received 60 mg/kg of RSR13 during a nephrectomy. The patient had been receiving ibuprofen, furosemide, and a calcium channel antagonist preoperatively, and received a cephalosporin postoperatively. Serum creatinine increased to 4.5 mg/dL at 48 h despite normal urine output. The renal dysfunction resolved with conservative treatment and did not prolong hospitalization, and the patient had a normal creatinine at the follow-up visit. The second patient was a 58 yr-old woman who received 75 mg/kg RSR13. She received vancomycin preoperatively, had deliberate hypotension during surgery with the administration of labetalol, and received ketorolac in the postoperative period. She experienced nonoliguric renal dysfunction after extensive spinal surgery, as demonstrated by an increase of creatinine from 0.8 mg/dL at screening to 2.2 mg/dL at 48 h. Conservative treatment resulted in return of normal serum creatinine on Day 7, and this event did not prolong hospitalization. One week later, the patient underwent additional spinal surgery without RSR13 administration and developed a comparable transient increase in serum creatinine. The third patient, a 53 yr-old man who received 100 mg/kg RSR13 underwent an anterior cervical corpectomy (C5-6), also experienced an increase of creatinine, from 0.9 mg/dL at screening to 3.5 mg/dL at 48 h. Preoperatively, he received lisinopril and a cephalosporin. He also had deliberate hypotension during surgery with the administration of labetalol. Urine output was normal, and conservative treatment resulted in normal creatinine measurement on Day 6. Once again, this event did not prolong hospitalization.

Laboratory Evaluations

Chemistry variables were generally unremarkable except for the aforementioned increases of creatinine. Three of the 17 patients receiving RSR13 developed clinically relevant transient increases in creatinine compared with none of the 9 patients receiving placebo. In 3 other patients, there was a mild transient increase in serum creatinine with maximum values ranging from 1.2 to 1.9 mg/dL. In all patients, the increases were transient, returned to clinically normal values with conservative therapy, and did not prolong hospitalization in any patient.


No pattern of clinically significant hemodynamic alterations was associated with RSR13 infusion, either at the time of infusion or any time in the follow-up.


This study provides conclusive evidence of the ability of RSR13 to cause a dose-dependent rightward shift of the oxygen dissociation curve in patients undergoing surgery and a general anesthetic. The primary objective of achieving an increase in p50 of 10 mm Hg was attained within the dose range of 75–100 mg/kg. In general, the RSR13 administration appeared to be well tolerated in this population, in that the incidence of adverse events was similar between treatment groups, and there was no evidence of a dose-response for the frequency or severity of adverse events. The data, however, suggest that RSR13 may be associated with a reversible nonoliguric renal dysfunction in patients who have other predisposing factors and concomitant medications.

The lipid lowering drugs clofibrate and bezofibrate bind to specific sites in the central water cavity of deoxyhemoglobin, shifting the allosteric equilibrium toward the low-affinity deoxy form (4). Testing of isomeric series of structurally related compounds identified RSR13 as a drug with significant allosteric activity that was also resistant to in vivo binding by albumin (4). This organic molecule readily crosses the red cell membrane, and exerts its allosteric effect by noncovalent interaction with three subunits of the deoxyhemoglobin tetramer. This interaction stabilizes deoxyhemoglobin by preventing narrowing of the central water cavity, thereby reducing the oxygen affinity of hemoglobin. This, in turn, augments oxygen unloading in the microvasculature and enhances the diffusion of oxygen from the blood to the tissues.

This study was designed to determine the dose of RSR13 that results in an increase in p50 of 10 mm Hg. This effect was reproducibly achieved with a dose of 100 mg/kg. Further increases in p50, while further increasing tissue oxygen tension, would also decrease the affinity of hemoglobin for oxygen such that hemoglobin saturation would be <90% while breathing room air. Three of the patients receiving RSR13 experienced such a decrease in arterial oxygen saturation. This expected side effect was easily managed by increasing the arterial oxygen tension with the administration of supplemental oxygen.

The ability to amplify physiologic tissue oxygenation indicates that RSR13 has potential application in the clinical conditions characterized by tissue hypoxia, including oncology, cardiovascular, and cerebrovascular events. Since its description, RSR13 reliably increases p50 and tissue oxygenation in a dose-dependent fashion in laboratory animals (5,6,10) and both increases brain oxygenation and reduces infarct size in a feline stroke model (8). Pharmacodynamic studies in awake volunteers demonstrated similar dose-related effects on p50 as this study (data on file at Allos Therapeutics, Denver, CO), and our results further indicate that the pharmacodynamic effects of RSR13 in humans are not altered in the presence of surgery or general anesthesia.

There were no significant hemodynamic effects seen with the RSR13 infusion. Awake volunteers report local irritation when the IV infusion exceeded 1.67 mg · kg−1 · min−1; none of our patients reported phlebitis or pain at the site of injection. Although fever was more common in RSR13 recipients, this finding was not seen in awake volunteers, and may be related to the more complex surgeries performed in the RSR13 recipients.

The finding of creatinine increases is concerning, despite the fact that all events were transient, none required more than conservative therapy or prolonged hospitalization, and that renal function returned to normal before or by the time of the follow-up visit. These renal findings may have been influenced by the patient’s disease. Of the six patients with increases in serum creatinine, two underwent nephrectomy, and all of the nephrectomy patients (four) received RSR13. As there were no placebo patients who underwent nephrectomy, we cannot determine whether these increases were the result of RSR13 or the surgical loss of a kidney. The patients with the most significant increases of creatinine were found, on post hoc analysis, to all have been volume depleted and oliguric before and during the period of the RSR13 infusion. Deliberate hypotension was a technique used in two of the RSR13 patients who experienced an increase in serum creatinine. Although this technique can decrease blood loss during surgery, it can also transiently affect renal function. Finally, many renally cleared drugs were also administered during the study, in particular vancomycin and some cephalosporins, such as cefotetan, which are occasionally associated with increases in creatinine and renal dysfunction. Additionally, many drugs affecting renal blood flow were administered in this study, such as nonsteroidal antiinflammatory drugs, aldosterone-converting enzyme inhibitors, and calcium channel inhibitors.

To evaluate the renal dysfunction observed in these patients, studies have been completed by using a rat model of acute renal dysfunction produced by uninephrectomy, salt restriction, volume depletion, prostaglandin inhibition (indomethacin), and radio contrast treatment (data on file at Allos Therapeutics, Denver, CO). This model of acute renal failure is similar to that described by Heyman et al. (11), and generally results in a doubling of serum creatinine within 24 hours of the interventions described. The most obvious lesion in this model is spotty necrosis of the thick ascending limb (TAL) of the loop of Henle, and the severity of the lesions correlates with the severity of renal dysfunction. The medullary TAL is particularly sensitive to this combination of insults, in part because of its naturally low Po2.

After performance of the noted insults, animals were either given RSR13 (150 mg/kg) or an equivalent volume of saline vehicle. Saline-treated rats had no changes in creatinine, CrCl, or blood urea nitrogen (BUN) over the subsequent three days, whereas RSR13 animals had transient elevations of serum creatinine and BUN, and a decrease in CrCl. All abnormal values returned to normal by Day 3. Renal histology postnecropsy demonstrated no necrosis, infiltrate, or inflammatory response in any animal. In follow-up studies, the same model was applied in rats without the administration of indomethacin. In the absence of prostaglandin inhibition by indomethacin, RSR13 administration did not affect serum creatinine, CrCl, or serum BUN. Similarly, rats who underwent all of the renal insults, including RSR13, but who also received a small dose of furosemide before or after the RSR13 administration, demonstrated no renal dysfunction.

Based on the findings from our clinical study and these studies in animals, we hypothesize that RSR13 treatment induces an imbalance between oxygen supply and demand in the medullary TAL of the nephron (mTAL). In experimental models, acute reversible tubular dysfunction in the mTAL occurs because of an imbalance between solute delivery and available oxygen. The normal response to such an imbalance is to decrease oxygen consumption by decreasing tubular function. As noted above, reduction of oxygen demand in the mTAL through the administration of small-dose furosemide prevented any RSR13-related renal dysfunction.

Prostaglandin inhibition by indomethacin reduces renal medulla oxygen content by up to 80%. Any event that further affects the imbalance between oxygen supply and demand results in the phenomenon of renal tubular shutdown to decrease renal oxygen consumption. Although these imbalances in oxygen supply and demand result in a decrease in renal function, no long-term injury occurs, as demonstrated by histological examination of the renal medulla in these rat studies.

With the limited number of patients in this study, and the limited information about each of them, we cannot absolutely quantify or qualify the degree of renal dysfunction that will be associated with the administration of RSR13. However, as all of the increases were transient and resolved without significant intervention, this question should not prohibit continued investigation of this drug in the surgical setting.

This pharmacodynamic and pharmacokinetic study of RSR13 in anesthetized surgical patients demonstrates a dose-related rightward shift in the oxygen dissociation curve with minimal adverse effects, and indicates that trials investigating the effect of this agent on ischemic conditions such as stroke or myocardial injury should be performed.


1. Perutz MF, Poyart C. Bezafibrate lowers oxygen affinity of haemoglobin. Lancet 1983; 2: 881–2.
2. Poyart C, Marden MC, Kister J. Bezafibrate derivatives as potent effectors of hemoglobin. Methods Enzymol 1994; 232: 496–513.
3. Randad RS, Mahran MA, Mehanna AS, Abraham DJ. Allosteric modifiers of hemoglobin. 1. Design, synthesis, testing, and structure: allosteric activity relationship of novel hemoglobin oxygen affinity decreasing agents. J Med Chem 1991; 34: 752–7.
4. Abraham DJ, Wireko FC, Randad RS, et al. Allosteric modifiers of hemoglobin: 2-[4-[[(3,5-Disubstituted anilino)carbonyl]-methyl]phenoxy]-2-methylpropionic acid derivatives that lower the oxygen affinity of hemoglobin in red cell suspensions, in whole blood, and in vivo in rats. Biochemistry 1992; 31: 9141–9.
5. Kunert MP, Laird JF, Abraham DJ, Lombard JH. Low-affinity hemoglobin increases tissue PO2 and decreases arteriolar diameter and flow in the rat cremaster muscle. Microvasc Res 1996; 52: 58–68.
6. Kunert MP, Liard JF, Abraham DJ. RSR-13, an allosteric effector of hemoglobin, increases systemic and iliac vascular resistance in rats. Am J Physiol 1996; 271: H602–13.
7. Wei EP, Randad RS, Levasseur JE, et al. Effect of local change in O2 saturation of hemoglobin on cerebral vasodilation from hy-poxia and hypotension. Am J Physiol 1993; 265: H1439–43.
8. Watson JC, Doppenberg EM, Bullock MR, et al. Effects of the allosteric modification of hemoglobin on brain oxygen and infarct size in a feline model of stroke. Stroke 1997; 28: 1624–30.
9. Gibaldi M, Perrier D. Pharmacokinetics. New York: Marcel Dekker Inc., 1982: 409–16.
10. Khandelwal SR, Randad RS, Lin PS, et al. Enhanced oxygenation in vivo by allosteric inhibitors of hemoglobin saturation. Am J Physiol 1993; 265: H1450–3.
11. Heyman SN. Acute renal failure with selective medullary injury in the rat. J Clin Invest 1988; 82: 401–12.
© 2001 International Anesthesia Research Society