|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Internal Medicine, Divisions of Pneumology (Drs. Clerbaux, Detry, Veriter, Liistro, and Frans) and Gastroenterology (Drs. Geubel and Horsmans), Cliniques Universitaires Saint-Luc Brussels, Brussels, Belgium.
Correspondence to: Thierry Clerbaux, PhD, Division of Pneumology, Department of Internal Medicine, Cliniques Uuniversitaires Saint-Luc Brussels, Brussels, Belgium; e-mail: clerbaux{at}pneu.ucl.ac.be
Abstract
Study objectives: To trace the entire oxyhemoglobin dissociation curve (ODC) in a cohort of cirrhotic patients in stable condition who were candidates for orthotopic liver transplantation (OLT).
Design: Prospective cohort study.
Setting: A large academic hospital.
Patients and methods: We traced the entire ODC in whole blood in standard conditions (pH 7.4; PCO2, 40 mm Hg; temperature, 37°C) for 50 cirrhotic candidates for OLT (27 men and 23 women) and 50 age- and height-matched healthy subjects (27 men and 23 women). All subjects were nonsmokers or ex-smokers for at least 5 years. We also measured 2,3 diphosphoglycerate (2,3 DPG) in RBCs, plasma ions, and arterial blood gases in all subjects according to standard methods. Mixed venous blood was also obtained from the 50 cirrhotic patients.
Results: Mean ODC was the same in the two groups. However, for the cirrhotic patients, the dispersion of the PO2 values of oxygen saturation percentage (SO2%) from 20 to 80% was significantly larger (p < 0.01 to p < 0.0001). In the cirrhotic patients, the mean PO2 for half-saturation of hemoglobin (P50) was 7.11 + 0.14 mEq/L chloride (p < 0.001) plus 0.36 mEq/L inorganic phosphate (p < 0.05) plus 0.25 µmol/gram of hemoglobin (gHb) 2,3 DPG (p < 0.00002) in absolute numerical values. Sodium, potassium, and calcium, three plasma ions disturbed in cirrhotic patients, did not contribute to determine the mean P50.
Discussion: In patients with cirrhosis, increased dispersion of PO2 values for a given level of SO2% may be related to four factors: (1) an observed alteration of the enzymes controlling the phosphoglycerate shunt; (2) hypothyroidism, which may affect 7 to 20% of patients with primary biliary cirrhosis; (3) the type of ongoing treatment, eg, diuretics and/or propranolol; and (4) plasma ion disturbances.
Conclusions: We describe the ODC by three indexes: shape, position, and an index of dispersion of the PO2 values for a given level of SO2%. In addition, when the latter is increased, we suggest that other factors than pH, temperature, carbon dioxide, and inorganic phosphates are acting on the position of the ODC.
Key Words: liver cirrhosis oxyhemoglobin dissociation curve
Two indexes characterize the oxyhemoglobin dissociation curve (ODC): shape and position. The latter index is usually described by the PO2 at half-saturation of hemoglobin (P50). Historically, the first factors found to be able to influence hemoglobin affinity for oxygen were salts studied in hemoglobin solutions.1234567 Anions, such as chloride or inorganic phosphates, induce a right shift of the ODC that enhances tissue oxygenation, whereas cations such as sodium, potassium, or magnesium have an opposite effect.
Before 1967, four groups of authors studied some parts of the ODC of cirrhotic patients by measuring a few PO2/oxygen saturation percentage (SO2%) points in each curve: two articles89 reported normal values, whereas the others1011 showed a rightward shift of the ODC. In 1967, two groups of authors1213 showed that in RBCs, the organic phosphate 2,3-diphosphoglycerate (2,3 DPG) plays a major role in regulating the position of the ODC, although inorganic phosphates play a minor, but not negligible, role in this regard.1214
A 1974 review article15 did not discuss the effects of ions on hemoglobin but considered that only four factors were able to influence the position of the ODC: pH, body temperature, carbon dioxide content, and organic phosphates. Subsequently, other authors1617181920212223 observed an increase in 2,3 DPG and P50 of cirrhotic patients. Unfortunately, the number of subjects was small in one group,23 while some series1822 included not only cirrhotics but patients with other liver diseases. Plasma ions were not measured. Finally, the position of the ODC was characterized by P50 instead of the whole ODC, which would have given much more information.24 In short, all articles except two89 reported a rightward shift of the ODC in cirrhotic patients. Nevertheless, in vivo252627 and in vitro142829 data confirmed that ions play a role in regulating the position of the ODC. The aim of the present work was to measure the entire ODC in whole blood in a homogeneous group of patients with cirrhosis, all of whom were candidates for an orthotopic liver transplantation (OLT).
Materials and Methods
We examined two groups of 50 subjects: one group of patients with cirrhosis (27 men and 23 women), all candidates for OLT, and one group of 50 gender-, age-, and height-matched control subjects. Mean age ± SD was 54 ± 7 years in the cirrhotic patients and 53 ± 8 years in the control subjects (not significant [NS]). Mean height was 1.72 ± 0.05 m in cirrhotic patients and 1.74 ± 0.06 m in normal subjects (NS). All subjects were lifelong nonsmokers or ex-smokers since at least 5 years of age. Criteria for diagnosis and liver serobiochemistry followed routine clinical methods. The origin of cirrhosis was as follows: cryptogenetic (n = 12), alcoholic (n = 7), postviral type B or C (n = 20), and primary biliary cirrhosis (n = 11). Four patients had a Child-Pugh score A, 20 patients had a score B, and 26 patients had score C. Ascites was present in 26 cirrhotic patients. Twenty-two cirrhotic patients were treated with spironolactone, and the other 4 patients received furosemide. Ten patients received propranolol for prophylaxis of digestive bleeding due to portal hypertension.30
Due to the occurrence of severe hemodynamic complications in the early experience of our OLT surgical group, we performed a complete hemodynamic investigation with agreement of the local ethics committee and informed consent of the patients. We therefore had the opportunity to sample and analyze mixed venous blood from the pulmonary artery.
The entire ODC on whole blood was traced under standard conditions (pH 7.40; PCO2, 40 mm Hg; temperature, 37°C) and corrected to the ODC in vivo by using actual values of pH, PCO2, and temperature. The method has been described previously.3132 The advantage of this method is to describe the ODC at all levels of oxygen saturation and thereby to give the resulting function of oxygen loading and unloading.24 Briefly, 10 mL of blood was sampled and tonometered33 with a reducing gas mixture (5.6% carbon dioxide in pure nitrogen); the entire ODC was then measured by a dynamic process by introducing a second gas mixture (5.6% carbon dioxide in pure oxygen). During this process, two variables were measured continuously: SO2% by photometry, and PO2 using a PO2 electrode (Eschweiler; Kiel, Germany). A plot was continuously made of hemoglobin oxygen saturation as a function of PO2. In addition, 10 mL of blood was sampled for measurements of 2,3 DPG, plasma ions, and arterial and mixed venous blood gases. For each curve, one hundred pairs of PO2 and SO2% values were processed by a homemade computer program in order to obtain the entire curve. The accuracy of the method, as expressed by the SD of the P50, was 0.1 mm Hg for six curves obtained from the same blood sample. Biological variation, estimated from 16 curves traced from four blood samples taken every Monday during 4 weeks from the same healthy nonsmoking volunteer, was 0.3 mm Hg.32
The plasma concentration of inorganic phosphates was determined using a commercial kit (Phosphorus Inorganic Kit No. 670; Sigma Chemical; St Louis, MO). The plasma chloride concentration was determined by a titrimetric method (Merckotest; Merck; Darmstad, Germany). Plasma sodium, potassium, and calcium concentrations were assayed in a blood gas analyzer (model 288; Ciba-Corning Diagnostics; Medfield, MA). Hemoglobin and carboxyhemoglobin were measured by an oximeter (OSM III; Radiometer; Copenhagen, Denmark). 2,3 DPG was measured by an enzymatic method (kit No. 35-UV; Sigma). The hemoglobin oxygen capacity was measured by a Lex-O2-Con galvanic cell (Lexington Instrument Corporation; Waltham, MA).34
Statistical Analysis
A Student t test for unpaired data was applied to compare PO2 values at 11 levels of SO2 in the two groups. To test the dispersion of the PO2 values for a single level of SO2, we used the F Fisher one-tailed test, F = S2x/S2y, where S2x and S2y are the variances of the data.35 The effect of 2,3 DPG and ions on P50 was assessed using univariate linear regression. To select the plasma ions that independently affect the P50, all of them were submitted to a forward stepwise linear regression using T-to-enter statistics to check the significance of change in explained P50 variance (r2); p < 0.05 was considered significant (analysis of variance).
Results
The mean ODC was the same in the two groups (Fig 1 ). However, the dispersion of the PO2 values for the cirrhotic patients was significantly (p < 0.01 to p < 0.0001) increased: SO2% of 20 to 80% (Table 1 ). Table 1 shows the SD of the PO2 values for 11 levels of oxygen saturation and related indexes (mean ± SD) for control subjects (n = 50) and cirrhotic patients (n = 50). Table 2 shows the effect of dispersion (1.96 x SD) of the PO2 values around the SO2% in terms of oxygen content (volume percentage).
|
|
|
|
|
Although the mean ODC was identical in the two groups (Fig 1), in cirrhotic patients the dispersions of the PO2 values for different levels of SO2% were significantly increased, from 20 to 80% saturation. In terms of oxygen transport, the maximal effect was not observed at P50 but at PO2 at 20% of saturation decreasing progressively to PO2 at 45% of saturation (Tables 1 , 2; Fig 2 ). This was revealed by tracing the entire ODC, instead of just measuring the P50, a method that gives more information about the loading and unloading of hemoglobin.2432
|
We have already described the same pattern, ie, a normal mean ODC and an increased dispersion of the PO2 values for different levels of SO2% in patients with chronic obstructive lung disease27 and in patients with severe comorbid illnesses.25 When the patients of this last group were treated, their ODC normalized, suggesting that the increased dispersion is reversible.2529
Why does the dispersion of PO2 values for different levels of SO2% increase in patients with cirrhosis? Synthesis and breakdown of 2,3 DPG are controlled by several enzymes in the phosphoglycerate cycle of Rapoport and Luebering, a side-shuttle of the main Embden-Meyerhoff pathway36 (Fig 3 ). Factors controlling synthesis and breakdown of 2,3 DPG include the following: (1) The concentration of 2,3 DPG itself: a negative feedback mechanism inhibits the activity of the phosphoglycerate mutase.37 (2) Hydrogen ion concentration: in alkalosis, the rate of glycolysis is increased and the activity of 2,3 DPG phosphatase decreases; therefore, 2,3 DPG is increased.38 As a matter of fact, the 2,3 DPG level is negatively correlated with hydrogen ion concentration, pH being significantly higher in cirrhotics than in control subjects. This fact had likely induced a 2,3 DPG synthesis, thus a rightward shift of the ODC (Fig 4 ). (3) Inorganic phosphate concentration: 2,3 DPG is low in hypophosphatemia39 and high in hyperphosphatemia.40 In cirrhotic patients, it is likely that the increase in inorganic phosphate induced a rightward shift of the ODC by two concomitant mechanisms: a direct action of plasma ions on hemoglobin,14262829 and a stimulation of 2,3 DPG synthesis.26
|
|
We attribute the increased dispersion of the PO2 values in our cirrhotic patients to four factors: (1) Alterations in the activities of enzymes controlling the phosphoglycerate shunt are observed in liver cirrhosis50: glucokinase activity is low or absent, hexokinase increases, and the glucokinase/hexokinase ratio is reduced. Glucose-6-phosphatase is also lower in cirrhotics.51 In alcoholic liver cirrhosis and primary biliary cirrhosis, the activity of glucokinase is < 10 of the activity in control subjects.52 (2) Thyroxine stimulates 2,3 DPG synthesis53; hypothyroidism is present in 7%54 to 20%55 of patients with primary biliary cirrhosis. 2,3 DPG was not measured in these patients, but it is likely that its concentration was decreased and therefore that the ODC was shifted to the left. (3) Plasma ions are disturbed in cirrhosis56: for instance, sodium and water excretion are impaired,57 especially in the presence of ascites and when diuretics are administered.58 Hyponatremia or hypernatremia and hypokalemia are frequently present.59 In our results, disturbed plasma ions in cirrhotic patients included sodium, calcium, and potassium (Table 4). It is likely that these alterations contributed to increase the dispersion of the PO2 values for different levels of SO2%. However, they did not contribute to determine the P50 (Table 3, last equation). (4) Various drugs are administered to cirrhotic patients: diuretics influence the position of the ODC,60 and propranolol alters hemoglobin affinity in patients with coronary artery disease by a mechanism not mediated by 2,3 DPG.60
What Is the Clinical Relevance of Our Study?
Above 80% saturation, the ODC dispersions of the two groups are identical, and therefore the loading of oxygen from pulmonary alveoli to hemoglobin was normal in cirrhotic patients. However, as all our cirrhotic patients were anemic, a leftward shift of the ODC had a detrimental effect on tissue oxygenation. We do not know the pattern of the ODC during exercise, which is accompanied by a marked increase in oxygen consumption. Finally, some tissues are highly sensitive to hypoxia.
How Can an Increased Dispersion of the PO2 Values for Different Levels of SO2% Be Revealed, and What Is the Usefulness of This Finding?
The two conditions for revealing an abnormal increase of the PO2 dispersion are an appropriate methodology32 and a homogeneous group of patients. This implies that factors other than pH, temperature, carbon dioxide content, and organic phosphates15 are able to influence the position of the ODC. This must stimulate a search for more factors than the four factors mentioned above,15 for instance plasma ions,2527 diuretics, and propranolol.60
Leaving aside carbon monoxide intoxication and hemoglobinopathies (such as thalassemia, human hemoglobin with abnormal affinity, and impaired heme-heme interaction such as hemoglobin Warsaw or South Milwaukee),6162 three patterns for the ODC have been described: (1) a leftward shift induced by hypothermia, alkalosis, or RBC hexokinase deficiency63; (2) a rightward shift mediated by hyperthermia, acidosis, or an increased synthesis of 2,3 DPG induced by chronic hypoxia41424344454647484963 or thyroxine intake53; and (3) normal ODC and 2,3 DPG despite severe hypoxemia of chronic lung disease.64 Based on our observations, we propose a fourth pattern: a normal mean ODC with an increased dispersion of PO2 values at different levels of SO2%.2527
Acknowledgements
The authors thank Professor F. Kreuzer, Department of Physiology, University of Nijmegen, the Netherlands; Professor J. Lebacq, Department of Physiology of our faculty; Dr. M. De Kock, who performed cardiac catheterization; Professor A. Robert, Department of Statistics; and Mrs. My Linh Cao for technical assistance.
Footnotes
Abbreviations: 2,3 DPG = 2,3-diphosphoglycerate; gHb = gram of hemoglobin; NS = not significant; ODC = oxyhemoglobin dissociation curve; OLT = orthotopic liver transplantation; SO2% = oxygen saturation percentage; P50 = PO2 for half-saturation of hemoglobin
Dr. Clerbaux and Dr. Frans benefited from a "Crédit personnel aux chercheurs" from the Belgian Fonds National de la Recherche Scientifique.
Received for publication August 6, 2004. Accepted for publication July 8, 2005.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |