| ||||||||||||||
|
|
|||||||||||||
Department of Anesthesiology, University of Virginia Health System, Charlottesville, Virginia
Address correspondence and reprint requests to Stuart M. Lowson, MBBs, Department of Anesthesiology, University of Virginia Health Sciences Center, Box 10010, Charlottesville, VA 22906-0010.
| Abstract |
|---|
|
|
|---|
Implications: Nitric oxide affects platelet function. We demonstrated that nitric oxide delivered into the gas inflow of the cardiopulmonary bypass circuit membrane oxygenator does not significantly alter platelet numbers or function.
| Introduction |
|---|
|
|
|---|
Nitric oxide (NO) and NO donors (i.e., nitroglycerin [NTG]) elevate 3',5'-cyclic guanosine monophosphate levels in platelets (8,9), and inhibit platelet aggregation in response to a variety of stimuli, such as adenosine diphosphate (ADP), arachidonic acid, collagen, and thrombin (10,11). NO inhibits platelet adhesion to healthy vascular endothelium (1215). Inhaled NO prolongs bleeding times in both animals (16,17) and humans (18,19). Furthermore, inhaled NO inhibits platelet aggregation in both animals (20) and patients with acute respiratory distress syndrome (21,22).
There is a theoretical basis for the hypothesis that NO may prevent platelet adhesion when blood comes into contact with a foreign surface. The large gas transfer surface of the membrane oxygenator (MO) of the CPB circuit is a suitable model in which to test this hypothesis. In vitro models have indicated that NO decreases platelet adhesion to the MO (2325). A study of CPB in pigs suggested that NO delivered in very high concentrations to the MO may increase platelet numbers by decreasing platelet aggregation (26). However, Mellgren et al. (27) recently demonstrated that 40 ppm NO delivered to the MO has minimal effects on platelets in humans. We hypothesized that delivering 100 ppm NO into the MO of the CPB circuit in cardiac surgical patients may increase platelet numbers, decrease platelet aggregation, and reduce postoperative bleeding.
| Methods |
|---|
|
|
|---|
Patients were divided into two groups on the basis of whether they were receiving preoperative aspirin. Each group, aspirin (ASA) versus nonaspirin (non-ASA), was randomized to either NO or a control group. In the control group, CPB was conducted routinely, and no investigational gas was added to the inlet gas flow. In the NO group, NO in nitrogen was added to the inlet gas flow of the MO of the CPB.
NO was delivered from a cylinder containing 1600 ppm NO in nitrogen. In pilot studies using cylinders with a lower NO concentration, we were unable to deliver an adequate and stable concentration of NO without diluting the oxygen delivery to the MO. This was a problem at the onset of CPB, when it is the practice to deliver 100% oxygen to the MO. NO was delivered into the side port of the inlet gas tubing just distal to the flowmeters. The concentration of NO in the fresh gas inflow was measured just proximal to the MO using an electrochemical NO sensor (Bedfont Scientific Ltd., Kent, UK). The delivery of NO was adjusted, using a flowmeter on the cylinder manifold, to provide a constant preset concentration of NO. NO delivery was started immediately prior to the onset of CPB (into the pump prime) and continued until after the end of CPB. The gas outflow from the MO was scavenged using the wall-mounted vacuum system.
In selecting the concentration of NO to deliver to the MO, we were guided by safety considerations. Our concern was the formation of met-hemoglobin, which is formed when NO combines with hemoglobin in red cells. We accepted an upper level for met-hemoglobin as 4% of total hemoglobin. In preliminary experiments (n = 30), NO concentrations of 20, 50, and 100 ppm were delivered for the duration of CPB, and met-hemoglobin levels were measured throughout CPB. A NO concentration of 100 ppm increased the met-hemoglobin levels over the first 90 min and then reached a plateau, but levels did not exceed 4%. Therefore, we selected a NO concentration of 100 ppm as the highest safe concentration.
The anesthesia and the use of vasoactive drugs was determined by the individual anesthesiologists. Patients were given a combination of fentanyl or sufentanil, midazolam, and isoflurane with pancuronium for neuromuscular blockade. Post-CPB patients were administered dopamine (5 µg · kg-1 · min-1). Epinephrine or NTG was used only if required by the patients cardiac function.
Apart from the addition of NO, CPB was conducted according to standard protocol for our institution. Heparin was given prior to CPB in a dose of 200300 units/kg. The activated clotting time was maintained greater than 400 s during CPB. The CPB circuit was primed with 1.7 L of Plasmalyte-A, 50 mL of 25% albumin, 88 mL of 25% mannitol, 7000 units of heparin, and 2 g cefazolin. The CPB circuit consisted of a MO (Optima, Cobe Cardiovascular, Inc., Arvada, CO), variable venous reservoir bag, arterial filter (heparin-coated), and cardiotomy reservoir. Cardipolegia consisted of modified Plegisol delivered by a Sarns MP4 system (Sarns, Ann Arbor, MI). The CPB flow was driven by a biomedicus biopump (Medtronic, Minneapolis, MN) and adjusted according to the mixed venous oxygen saturation. Patients were cooled to a temperature of 28°C30°C.
Once the patient was rewarmed to 37°C and satisfactorily weaned from CPB, protamine (2 mg/kg) was given to return the activated clotting time to the pre-CPB value. The post-CPB hematocrit was maintained at a value greater than 24% using cell saver and packed red blood cells (RBCs) if necessary. Postoperative platelet and fresh frozen plasma (FFP) transfusions were given based on chest tube drainage according to our institutional surgical protocol. The surgical staff was blinded as to the study group.
Measured Variables
Preoperative prothrombin and partial thromboplastin times were measured to ensure that these were normal before including the patient in the study. The duration of CPB and the total volume of blood products (including cell saver and packed RBCs) were recorded. Met-hemoglobin levels were measured at 30-min intervals throughout CPB. The requirement for inotropic drugs, vasoactive drugs, and/or NTG was also noted. The chest tube drainage at 18 h post-CPB and the patients need for blood products was recorded.
Platelet Aggregation and BTG
Arterial blood samples were taken after induction of anesthesia (pre-CPB), after 60 min on CPB, and 2 h after the end of CPB. The samples were used for measurement of total platelet numbers, platelet aggregation studies, and measurement of ß-thromboglobulin (BTG).
Blood samples were collected in 0.129 M buffered sodium citrate (1\X9 volume) vacuum tubes (Vacutainer, Beckton Dickerson, Rutherford, NJ) and immediately transported to the laboratory, centrifuged, and analyzed for the platelet aggregation responses. Platelet counts were measured on a Cell-Dyn 3000 (Abbott Diagnostics, Santa Clara, CA). Platelet-rich (PRP) and platelet-poor (PPP) plasma were prepared by centrifugation. Platelet aggregation studies were performed in siliconized cuvettes with stirring at 900 rpm at 37°C using a Chrono-Log model 490 optical aggregometer (Havertown, PA). The instrument was calibrated to measure turbidity as percent transmission (%T) (PPP = 100%T, PRP = 0%T). The platelet aggregation response to a range of concentrations of ADP (2.5, 5.0, 10.0, and 20 µM) was determined. The maximum slope of the change in turbidity and the amplitude of the maximum response were used to quantitate aggregation responses.
BTGwhich is a platelet release product normally stored in the platelet
2-granules and released when platelets are activatedwas measured in samples of PPP stored at -20°C. The BTG was analyzed by a commercially available enzyme immunoassay (Asserachrom ß-TG, Diagnostica Stago, Asnieres-Sur-Seine, France).
Statistical Analysis
Two-way multivariate analysis of variance was performed to examine for statistical differences in the continuous demographic variables, chest tube drainage, and duration of CPB between populations (ASA versus non-ASA) and between groups (NO versus control). A
2 test was performed to examine for equal gender representation for both groups and populations. Repeated-measures analysis of variance was performed to detect differences in the platelet numbers, platelet aggregation response, and the BTG levels between the two groups before, during, and after CPB. Data are presented as the means ± SD. Significance was considered for P < 0.05.
| Results |
|---|
|
|
|---|
Patients age, the duration of CPB, post-CPB hemoglobin, and the intraoperative administration of cell saver blood and allogenic packed RBCs were not different between the groups (Table 1). Postoperatively, the number of patients and quantity per patients requiring packed RBCs, platelets, and FFP were also not different between the groups. The chest tube drainage in the first 18 h post-CPB was not different between groups. Forty-four patients received dopamine (5 µg · kg-1 · min) and three patients received epinephrine (14 µg/min). Seven patients received NTG postoperatively, of which five were in the NO group and two in the control group (not statistically different). No complications resulted from delivery of NO to the MO. The met-hemoglobin level at the end of CPB was 3.4 ± 0.4% (range 2.93.8%) in the NO group versus 1.1 ± 0.3% (range 0.81.6%) in the control patients.
|
|
|
| Discussion |
|---|
|
|
|---|
Although there was a trend toward higher platelet numbers, post-CPB in those patients given NO (100 ppm) into the MO, the results did not reach statistical significance. This is very similar to the results of a clinical study by Mellgran et al. (27) that showed a nonsignificant trend toward increased platelet numbers using 40 ppm of NO. In our study, the slope of the platelet aggregation response to 2.5 µM ADP in blood samples taken on-CPB was significantly less in the NO group; otherwise, there were no other detectable intergroup differences in platelet aggregation in response to the other doses of ADP tested. Mellgren et al. (27) also did not detect an effect of NO on the platelet aggregation response to 0.54.0 µM ADP. Nevertheless, we cannot rule out that the dose range of ADP we used was too insensitive to detect a difference between the NO and control groups, or that other platelet aggregation tests may have detected a significant difference.
Previous studies have shown that CPB is associated with
-granule release from platelets and that BTG, a constituent of
2-granules, increases during CPB (1,3,4,28,29). We chose to measure BTG levels as a marker of platelet activation because most of the study patients were taking aspirin, which does not affect BTG levels (30). NO (100 ppm) had no demonstrable effect on BTG levels. Although this suggests that NO administration may not influence platelet activation and
2-granule secretion on CPB, it should be noted that measurements of BTG levels produced highly variable results.
In vitro studies of CPB have demonstrated that NO, when delivered into the inlet gas flow to the MO, can protect platelet numbers and function. Mellgren et al. (23) noted significantly higher circulating platelet counts when NO (15, 40, or 75 ppm) was added to the MO of a CPB circuit, primed with heparin-treated human blood, compared with control CPB circuits (23). Plasma BTG increased progressively throughout the experiment in both NO and control circuits; however, the levels were significantly lower with NO versus controls, suggesting decreased platelet
2-granule release. There was no significant difference in mean platelet serotonin content or platelet membrane GPIb expression between controls and NO. Decreased release of BTG, but not serotonin, suggests that NO was efficient in protecting
2, but not dense granules, from releasing their contents. Keh et al. (24) also demonstrated that platelet counts were higher in an in vitro CPB circuit receiving 20 ppm NO. In contrast, Konishi et al. (25) were unable to show higher circulating platelet counts with either 100 or 200 ppm NO; however, platelet deposition on the extracorporeal circuit was almost completely prevented by the administration of 200 ppm NO. The divergent results between our clinical study and the in vitro studies may be related to slight differences in NO concentration or to the slightly longer CPB times in our study; however, it is more likely that the differences are related to surgical stress and more complex biological interactions between the vasculature and platelets associated with in vivo clinical studies.
A platelet-sparing effect of NO has been demonstrated in a porcine model of CPB using very large concentrations of NO (500 and 1000 ppm) (26). Although there was no significant difference in the circulating platelet count between NO and control groups, NO caused a significant decrease in the percentage of platelets adherent to the circuit, compared with controls (4% vs 25% of total pre-CPB platelet count). It is possible that higher concentrations of NO may be required in vivo to produce significant platelet-sparing effects, or that the effects are measurable only at the foreign surface and do not significantly affect circulating platelet numbers. We did not test concentrations higher than 100 ppm because of the risk of high methemoglobin levels.
In the only other similar study, Mellgren et al. (27) administered NO (40 ppm) into the MO of a CPB circuit in 20 patients scheduled for cardiac surgery. They found that NO administration did not significantly affect platelet aggregation or platelet counts. Similar to our results, they demonstrated that there was a trend for platelet numbers to be greater in the NO group throughout CPB and at three h post-CPB, but this did not attain statistical significance. Administration of NO resulted in a significant decrease in the expression of platelet membrane expression of GPIb, suggesting reduced potential for platelet adhesiveness. BTG levels increased during CPB, but were not influenced by NO.
We attempted to standardize the postoperative use of catecholamines and nitrates because of the reported effects of these agents on platelet function. Epinephrine is an established agonist of platelet aggregation, which is thought to be mediated via the
-adrenergic receptor (31,32). In our study, similar numbers of patients in both the NO and control groups were given epinephrine post-CPB. Nitrates have been shown to prolong bleeding time (33) and inhibit platelet aggregation (34,35). However, Muikko et al. (35) found that pre- or post-CPB nitrates administered in clinically relevant doses had no effect on bleeding time or platelet function in patients undergoing coronary artery bypass graft surgery. In our study, there was a trend, but not a significant excess of, post-CPB use of NTG in the control groups. Furthermore, there was only a small number of patients that received NTG. We, therefore, do not believe that either epinephrine or NTG administration influenced our results.
Because of safety considerations, we selected a low-risk population of patients scheduled for cardiac surgery. We excluded any patient scheduled for repeat cardiac surgery, CPB duration greater than 180 min, or patients requiring an IABP to assist with weaning from CPB. Thus, this was not a population at high risk of postoperative platelet dysfunction. It is possible that the trend toward platelet-sparing effects that we observed and in the study by Mellgran et al. may yield statistically positive results in a large trial of higher risk patients. Likewise, we were unable to demonstrate a difference in postoperative blood loss or the need for blood products; however, we only enrolled patients with a relatively low risk of major bleeding complications.
In conclusion, we demonstrated that 100 ppm NO may be added to the MO of the CPB circuit without increasing met-hemoglobin levels above 4%. We were unable to demonstrate a platelet-sparing effect based on platelet number, platelet aggregation to ADP, BTG levels, and postoperative bleeding in low-risk patients. Further studies will be required to determine if different NO concentrations added to the MO produce a platelet-sparing effect in high-risk cardiac surgical patients.
| Acknowledgments |
|---|
We acknowledge the assistance of the Departments of Cardiothoracic Surgery and Cardiac Perfusion, Dr. J. Humphries of Hematology, and Dr. W. Chang of the Department of Biostatistics at the University of Virginia.
| References |
|---|
|
|
|---|
-granule release. Blood 1980;65:32335.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|