Effects of Metabolic Alkalosis on Pulmonary Gas Exchange 1 , 2
SERGE BRIMIOULLE and ROBERT J. KAHN
Introduction
Metabolic alkalosis has been reported as a common acid-base disorder in hospitalized patients (1). It is generally associated with an increase in alveolar-arterial oxygen tension difference, which, in clinical conditions, is attributed to atelectasis resulting from the compensatory hypoventilation (1). Whether blood pH can affect pulmonary gas exchange when hypoventilation is prevented remains unclear (2). Haas and Bergofsky (3) reported that metabolic acidosis increased Pa02 because of a decrease in venous admixture (QS/QT) (3). On the contrary, QS/QT remained unchanged during metabolic acidosis and alkalosis in a study of Frans and coworkers (4), and modifications in Pa02 were attributed to shifts of the oxyhemoglobin dissociation curve. In a previous article (5), we reported the changes in blood gases during correction of metabolic alkalosis with infusion of hydrochloric acid (HCl). Patients treated during controlled mechanical ventilation showed a marked increase in Pa02 despite the absence of change in ventilation. Available data did not allow identification of the mechanism of this improvement in oxygenation. The Bohr effect probably contributed to increased Pa0 2, but this seemed insufficient to explain the extent of the observed changes. A suggested additional mechanism was a reduction of pulmonary ventilation/ perfusion (\r/Q) mismatching caused by a redistribution of blood flow away from hypoxic regions of the lung. Wetherefore studied the hemodynamic and gasometric changes during correction of severe metabolic alkalosis with HCI infusion in patients treated with continuous mechanical ventilation. The main purpose of the study was to determine whether the changes in oxygenation were due to extrapulmonary factors or to an improvement in pulmonary \r/Q relationships. Additional purposes were to assess whether changes did persist after the end of the HCI infusion, and whether HCI could be administered safely at the concentration of 1 mollL.
SUMMARY In order to investigate whether the changes in Pao, reported during acid-base disturbances are due to modifications of ventilation/perfusion relationships or only to extrapulmonary factors, we studied the hemodynamics and blood gases of eight critically III patients maintained in constant mechanical ventilation, before and after selective correction of metabolic alkalosis by infusion of 1 N hydrochloric acid (HCI). HCI infusion decreased arterial pH from 7.55 to 7.40 (p < 0.001) and increased Pao, from 76 to 98 mm Hg (p < 0.05) at the end of the study. Cardiac output and oxygen consumption did not change. In patients with initial venous admixture (OSIC:lT) < 20% (n = 4), as/aT did not change, and hemoglobin saturation decreased, whereas Pao, increased from 87 to 96 mm Hg (p < 0.10), Indicating a shift in the oxyhemoglobin dissociation curve caused by the Bohr effect. In patients with as/aT> 20% (n 4), as/aT decreased from 27 to 22% (p < 0.05), hemoglobin saturation increased from 93 to 96% (p < 0.05), and Pao, increased from 65 to 100 mm Hg (p < 0.05),which reflects an improvement in ventilation/perfusion relationships, probably because of enhanced hypoxic pulmonary vasoconstriction. These data indicate that metabolic alkalosis deteriorates pulmonary ventilation/perfusion relationships in patients with marked respiratory failure (as/aT> 20%), and that reversing this effect with HCI infusion can improve Pao, significantly. AM REV RESPIR DIS 1990; 141:1185-1189
=
Methods Eight critically ill patients were studied (six male and two female, 37 to 83 yr of age; mean, 64 yr) (table I) according to a protocol approved by the Committee for Medical Ethics of the Erasme University Hospital. Criteria for admission into the study were metabolic alkalosis (arterial pH> 7.50 and Paco, > 35 mm Hg) persisting after 24 h of adequate therapy, treatment with mechanical ventilation, and presence of a pulmonary artery SwanGanz catheter (Edwards Laboratories, Santa Ana, CA).
Patients Metabolic alkalosis was attributed to previous hypercapnia (n = 5), acute renal failure (n = 5), hepatic failure (n = 2), massive blood transfusions (n = 2), hypokalemia (n = 2), administration of furosemide (n = 6), oral antacids (n = 3), and steroids (n = 3). Administration of furosemide had been discontinued at least 24 h before HCI infusion in three patients, but was maintained in three patients with oliguric renal failure. Management of metabolic alkalosis had consisted of (l) removal of causal factors, except when this was contraindicated by the underlying diseases, (2) correction of fluid deficits, (3) administration of potassium, as long as serum potassium concentration remained below 5.0 mmol/L, and (4) administration of chloride, as sodium and/or potassium solutions. Urine concentrations of sodium, potassium, and chloride were 18 to 74 (mean, 45), 41 to 88 (mean, 65), and 25 to 112 (mean, 75) mmol/L, respectively. All patients had respiratory failure, which was attributed to chronic obstructive pulmo-
nary disease (n = 5), bronchopneumonitis (n = 5), lung contusion (n = 2), noncardiogenic pulmonary edema (n = 2), and pulmonary embolism (n = 1). Patients were ventilated (Servo 9OOC; Elema, SoIna, Sweden) with an inspired oxygen fraction (F102) of 0.35 to 0.60 (mean, 0.42), a tidal volume of 10 to 15 (mean, 12) ml/kg, a respiratory rate of 8 to 13 (mean, 11) breaths/min, and, in four patients, a positive end-expiratory pressure of 5 to 10em H,o. Twohours before the study, the ventilator had been shifted from assist to control mode after respiratory rate had been, if needed, increased until the patient no longer attempted to trigger the ventilator. Respiratory conditions were then maintained unchanged until 12 h after the end of the HCI infusion. Patients were at rest during the measurements, and no specific sedation or paralysis was given for the study.
na Infusion The 1 N HCI solution (1,000 mmol of HCI in 1 L of sterile water) was prepared in a glass bottle and infused via an internal jugular vein into the superior vena cava at a constant rate of 1 ml/kg/h, In each patient, the adequate position of the central venous catheter had
(Received in original form May 26, 1989 and in revised form July 26, 1989) I From the Department of Intensi ve Care, Erasme University Hospital, Brussels, Belgium. , Correspondence and requests for reprints should be addressed to Serge Brimioulle, M.D., Department of Intensive Care, Erasme University Hospital, Route de Lennik, 808, B-1070Brussels, Belgium.
1185
1186
BRIMIOULLE AND KAHN
TABLE 1
decreased from 36 to 23 mmollL (p
Patient No.
1 2 3 4 5 6 7 8
Sex F M M F M M M M
Age (yr)
75 61 73 72 72 42 37 83
as/at (%)
Main Diagnosis Brainstem infarct, coma Septic shock, cardiac arrest Decompensated COPD Decompensated COPD Pulmonary embolism Hepatitis, liver transplantation Hepatitis, liver transplantation Decompensated COPD
been confirmed by a chest roentgenogram before the infusion. At l-h intervals, blood samples were withdrawn through an arterial catheter for immediate determination of arterial blood gases. HCl infusion was discontinued when base excess had decreased below - 4 mmollL.
Measurements Hemodynamic, gasometric, and biologic measurements were performed before the HCl infusion and at 2 and 12 h after the end of infusion. Systemic arterial pressure was obtained from a radial or femoral catheter, and pulmonary arterial pressure, right atrial pressure, and pulmonary capillary wedge pressure were obtained from the Swan-Ganz catheter (Edwards Laboratories), using Statham P50 transducers (Gould Inc., Oxnard, CA) connected to a bedside hemodynamic monitoring system (Sirecust 404; Siemens, Erlangen, FRG, or HP 78354A; Hewlett-Packard, Palo Alto, CA). The pressure tracings and an electrocardiograph lead were recorded on a four-channel recorder (HP 7404; Hewlett-Packard). Pressures were referenced to the midchest level, and were read at end expiration. Cardiac output was measured with the thermodilution method using five injections, starting at the beginning of expiration, of lO-mlice-cold boluses. Arterial and mixed venous blood gases (ABL 2 or ABL 3 analyzer; Radiometer, Copenhagen, Denmark) and hemoglobin oxygen saturations (lL 282 CO-Oximeter; Instrumentation Laboratories, Lexington, MA, or OSM 3 hemoximeter; Radiometer) were measured immediately after drawing the samples and corrected for patient temperature. Biologicdata included blood cells counts, serum electrolytes, lactate, blood urea nitrogen, creatinine, and total protein concentrations. Calculations The following formulas were used for gas exchange calculations: alveolar Po, (PAo" mm Hg) = {(barometric pressure - 47) x FlO,} - {Paco, x (I - (I - R) x Flo,)/R}, where R is the respiratory gas exchange ratio; arterial and mixed venous blood oxygen contents (Can, and Cvo" mlldl) = {hemoglobin concentration (g/dl) x hemoglobin saturation (070) x 0.0139} + {oxygen tension (mm Hg) x 0.0031}; QS/QT (% of blood flow) =
35 17
15 20 29 25 12 11
Outcome Death Death Survival Death Death Survival Survival Survival
(Ceo, - Cao,)/(Ceo, - Cvo,) x 100, where Ceo, is the end-capillary blood oxygen content, estimated from the PAo, and the calculated corresponding saturation (6).
Statistics Results are expressed as mean values ± standard error. Statistical analysis consisted of an analysis of variance for repeated measures in the same patients (7). Postinfusion values were compared with preinfusion values using the method of contrasts (7); p values less than 0.05 were accepted as indicating statistical significance.
Results
Metabolic alkalosis was corrected after infusion of 4 to 8 (mean, 6) mmollkg of HCI, and acid-base status remained normal 12 h after the end of the infusion (table 2). No clinical complication was observed during or after the infusion. Arterial pH decreased from 7.55 to 7.38 (p < 0.001), and bicarbonate concentration
Hemodynamic data 2 h after the end of the HCI infusion showed an increase in pulmonary arterial pressure from 21 to 26 mm Hg (p < 0.05), without increase in cardiac output or in pulmonary capillary wedge pressure. Pan, increased from 76 to 114 mm Hg (p < 0.001), and Sao 2 increased from 96 to 97070 (p < 0.10). Pvo2 increased from 33 to 37 mm Hg (p < 0.01), whereas Svo 2 remained unchanged. QsI QT decreased from 21 to 16% (p < 0.01). Significant changes in Pao, and QS/QT persisted 12 h after the end of the infusion, whereas pulmonary arterial pressure returned almost to the initialleveI. Oxygen consumption did not change during the study. In order to investigate whether the effects of pH were dependent on the severity of the respiratory failure, patients were separated into two groups according to whether their initial QS/QT was below or above 20% (table 3 and figure 1). Patients with less venous admixture (n = 4) showed a moderate increase in Pao2 without improvement in Sao, or QS/QT. Patients with more severerespiratory failure (n = 4) had a greater and sustained increase in Pao., together with a significant and persistent improvement in Sao 2 and QS/QT. Other variables showed similar changes in the two groups of patients. HCI infusion was associated with some increase in white blood cells (p < 0.001) and a decrease in platelets (p < 0.01) (ta-
TABLE 2 ACID·BASE STATUS, HEMODYNAMICS, AND GAS EXCHANGE BEFORE AND AFTER HCI INFUSION
Arterial pH Paco" mm Hg HCO;, mmol/L
Pre-HCI
HCI at 2 h
HCI at 12 h
7.55 ± 0.Q1 40 ± 2 36 ± 1
7.38 ± 0.02:1: 39 ± 2 23 ± 1:1:
7.40 ± 0.02:1: 40 ± 2 25 ± 1:1:
HR, beats/min CO, Llmin/m' Psa, mm Hg Pra, mm fig Ppa, mm Hg Ppcw, mm Hg
97 3.5 81 8 21 11
± ± ± ± ± ±
8 0.4 6 1 2 2
90 ± 7 3.2 ± 0.4 90 ± 7t 8±2 26 ± 3' 11 ± 2
89 3.6 89 8 22 12
± ± ± ± ± ±
7 0.4 7t 1 2 2
Pao" mm Hg Pvo" mm Hg Sao" % svo, % as/aT, %
76 33 96 64 21
± ± ± ± ±
6 1 1 3 3
114 37 97 64 16
13:1: 1t 1 3 2t
98 37 97 66 18
± ± ± ± ±
7' 1t 1 3 2'
± ± ± ± ±
Definition of abbreviations: HR = heart rate; CO = cardiac output; Psa = systemic arterial pressure; Pra = right atrial pressure; Ppa = pulmonary arterial pressure; Ppcw = pulmonary capillary wedge pressure; PiIo, = mixed venous 0, pressure; Silo, = mixed venous 0, saturation; QS/QT = venous admixture. • p < 0.05 versus pre-HC!. t p < 0.01 versus pre-HCI. p < 0.001 versus pre-HCI.
*
20%). Acknowledgment The writers thank P. Lejeune, M.D. for his constructive review of the manuscript. References 1. Hodgkin JE, Soeprono FF, Chan DM. Incidence of metabolic alkalemia in hospitalized patients. Crit Care Med 1980; 12:725-8. 2. Nattie EE. Gas exchange in acid-base disturbances. In: Fahri LE, Tenney SM, eds. Handbook of physiology. Vol. 4. The respiratory system, gas exchange. Bethesda: American Physiological Society, 1987; 421-38. 3. Haas F, Bergofsky EH. Effect of pulmonary vasoconstriction on balance between alveolar ventilation and perfusion. J Appl Physio11968; 24:491-7. 4. Frans A, Turek Z, Yokota H, Kreuzer F. Effects of variations in blood hydrogen ion concentration on pulmonary gas exchange of artificially ventilated dogs. Pfluegers Arch 1979; 380:35-9. 5. Brimioulle S, Vincent JL, Dufaye P, Berre J, Degaute JP, Kahn RJ. Hydrochloric acid infusion for treatment of metabolic alkalosis: effects on acidbase balance and oxygenation. Crit Care Med 1985; 13:738-42. 6. Severinghaus JW. Simple, accurate equations for human blood O. dissociation computations. J Appl Physiol 1979; 46:599-602. 7. Winer BJ. Statistical principles in experimental design. New York: McGraw-Hill, 1971. 8. Grover RF, Wagner WW, McMurtry IF, Reeves JT. Pulmonary circulation. In: Shepherd JT, Ab-
METABOLIC ALKALOSIS AND PULMONARY GAS EXCHANGE
boud FM, eds, Handbook of physiology. Vol. 3. The cardiovascular system, peripheral circulation and organ blood flow. Bethesda: American Physiological Society, 1983; 103-36. 9. Fishman AP. Pulmonary circulation. In: Fishman AP, Fisher AB, eds. Handbook of physiology. Vol. I. The respiratory system, circulation and nonrespiratory functions. Bethesda: American Physiological Society, 1985; 93-165. 10. Sylvester JT, Gottlieb JE, Rock P, Wetzel RC. Acute hypoxic responses. In: Bergofsky EH, ed, Abnormal pulmonary circulation. New York: Churchill Livingstone, 1986; 127-65. 11. Benumof JL, PirioAF, Johanson I, Trousdale FR. Interaction of Pvo2 with PAo2 on hypoxicpulmonary vasoconstriction. J Appl Physiol1981; 51: 871-4. 12. Pease RD, Benumof JL, Trousdale FR. PAo2 and PV02interaction on hypoxic pulmonary vaso-
constriction. J Appl Physiol 1982; 53:134-9. 13. Hughes JD, Rubin LJ. Relation between mixed venous oxygen tension and pulmonary vascular tone during normoxic, hyperoxic and hypoxic ventilation in dogs. Am J Cardiol 1984; 54:1118-23. 14. West JB, Wagner PD. Pulmonary gas exchange. In: West JB, ed. Bioengineering aspects of the lung. New York: Marcel Dekker, 1977; 361-457. 15. Orr JA, Shams H, Fedde MR, Scheid P. Cardiorespiratory changes during HCI infusion unrelated to decreases in circulating blood pH. J Appl Physiol 1987; 62:2362-70. 16. Shams H, Peskar BA, Scheid P. Acid infusion elicits thromboxane A,-mediated effects on respiration and pulmonary hemodynamics in the cat. Respir Physiol 1988; 71:169-83. 17. Shirer HW, Erichsen DF, Orr JA. Cardiorespiratory responses to HCI vs lactic acid infusion. J Appl Physiol 1988; 65:534-40.
1189 18. Lloyd TC Jr. Influence of blood pH on hypoxic pulmonary vasoconstriction. J Appl Physiol 1966; 21:358-64. 19. Marshall C, Lindgren L, Marshall BE. Metabolic and respiratory hydrogen ion effects on hypoxic pulmonary vasoconstriction. J Appl Physiol 1984; 57:545-50. 20. Kwun KB, Boucherit T, Wong J, Richards Y, Bryan-Brown C. Treatment of metabolic alkalosis with intravenous infusion of concentrated hydrochloric acid. Am J Surg 1983; 146:328-30. 21. Adrogue HJ, Madias NE. Changes in plasma potassium concentrations during acute acid-base disturbances. Am J Med 1981; 71:456-67. 22. Tobin RB. In vivo influences of hydrogen ions on lactate and pyruvate in blood. Am J Physiol 1964; 207:601-5.
SERGE BRIMIOULLE and ROBERT J. KAHN
Introduction
Metabolic alkalosis has been reported as a common acid-base disorder in hospitalized patients (1). It is generally associated with an increase in alveolar-arterial oxygen tension difference, which, in clinical conditions, is attributed to atelectasis resulting from the compensatory hypoventilation (1). Whether blood pH can affect pulmonary gas exchange when hypoventilation is prevented remains unclear (2). Haas and Bergofsky (3) reported that metabolic acidosis increased Pa02 because of a decrease in venous admixture (QS/QT) (3). On the contrary, QS/QT remained unchanged during metabolic acidosis and alkalosis in a study of Frans and coworkers (4), and modifications in Pa02 were attributed to shifts of the oxyhemoglobin dissociation curve. In a previous article (5), we reported the changes in blood gases during correction of metabolic alkalosis with infusion of hydrochloric acid (HCl). Patients treated during controlled mechanical ventilation showed a marked increase in Pa02 despite the absence of change in ventilation. Available data did not allow identification of the mechanism of this improvement in oxygenation. The Bohr effect probably contributed to increased Pa0 2, but this seemed insufficient to explain the extent of the observed changes. A suggested additional mechanism was a reduction of pulmonary ventilation/ perfusion (\r/Q) mismatching caused by a redistribution of blood flow away from hypoxic regions of the lung. Wetherefore studied the hemodynamic and gasometric changes during correction of severe metabolic alkalosis with HCI infusion in patients treated with continuous mechanical ventilation. The main purpose of the study was to determine whether the changes in oxygenation were due to extrapulmonary factors or to an improvement in pulmonary \r/Q relationships. Additional purposes were to assess whether changes did persist after the end of the HCI infusion, and whether HCI could be administered safely at the concentration of 1 mollL.
SUMMARY In order to investigate whether the changes in Pao, reported during acid-base disturbances are due to modifications of ventilation/perfusion relationships or only to extrapulmonary factors, we studied the hemodynamics and blood gases of eight critically III patients maintained in constant mechanical ventilation, before and after selective correction of metabolic alkalosis by infusion of 1 N hydrochloric acid (HCI). HCI infusion decreased arterial pH from 7.55 to 7.40 (p < 0.001) and increased Pao, from 76 to 98 mm Hg (p < 0.05) at the end of the study. Cardiac output and oxygen consumption did not change. In patients with initial venous admixture (OSIC:lT) < 20% (n = 4), as/aT did not change, and hemoglobin saturation decreased, whereas Pao, increased from 87 to 96 mm Hg (p < 0.10), Indicating a shift in the oxyhemoglobin dissociation curve caused by the Bohr effect. In patients with as/aT> 20% (n 4), as/aT decreased from 27 to 22% (p < 0.05), hemoglobin saturation increased from 93 to 96% (p < 0.05), and Pao, increased from 65 to 100 mm Hg (p < 0.05),which reflects an improvement in ventilation/perfusion relationships, probably because of enhanced hypoxic pulmonary vasoconstriction. These data indicate that metabolic alkalosis deteriorates pulmonary ventilation/perfusion relationships in patients with marked respiratory failure (as/aT> 20%), and that reversing this effect with HCI infusion can improve Pao, significantly. AM REV RESPIR DIS 1990; 141:1185-1189
=
Methods Eight critically ill patients were studied (six male and two female, 37 to 83 yr of age; mean, 64 yr) (table I) according to a protocol approved by the Committee for Medical Ethics of the Erasme University Hospital. Criteria for admission into the study were metabolic alkalosis (arterial pH> 7.50 and Paco, > 35 mm Hg) persisting after 24 h of adequate therapy, treatment with mechanical ventilation, and presence of a pulmonary artery SwanGanz catheter (Edwards Laboratories, Santa Ana, CA).
Patients Metabolic alkalosis was attributed to previous hypercapnia (n = 5), acute renal failure (n = 5), hepatic failure (n = 2), massive blood transfusions (n = 2), hypokalemia (n = 2), administration of furosemide (n = 6), oral antacids (n = 3), and steroids (n = 3). Administration of furosemide had been discontinued at least 24 h before HCI infusion in three patients, but was maintained in three patients with oliguric renal failure. Management of metabolic alkalosis had consisted of (l) removal of causal factors, except when this was contraindicated by the underlying diseases, (2) correction of fluid deficits, (3) administration of potassium, as long as serum potassium concentration remained below 5.0 mmol/L, and (4) administration of chloride, as sodium and/or potassium solutions. Urine concentrations of sodium, potassium, and chloride were 18 to 74 (mean, 45), 41 to 88 (mean, 65), and 25 to 112 (mean, 75) mmol/L, respectively. All patients had respiratory failure, which was attributed to chronic obstructive pulmo-
nary disease (n = 5), bronchopneumonitis (n = 5), lung contusion (n = 2), noncardiogenic pulmonary edema (n = 2), and pulmonary embolism (n = 1). Patients were ventilated (Servo 9OOC; Elema, SoIna, Sweden) with an inspired oxygen fraction (F102) of 0.35 to 0.60 (mean, 0.42), a tidal volume of 10 to 15 (mean, 12) ml/kg, a respiratory rate of 8 to 13 (mean, 11) breaths/min, and, in four patients, a positive end-expiratory pressure of 5 to 10em H,o. Twohours before the study, the ventilator had been shifted from assist to control mode after respiratory rate had been, if needed, increased until the patient no longer attempted to trigger the ventilator. Respiratory conditions were then maintained unchanged until 12 h after the end of the HCI infusion. Patients were at rest during the measurements, and no specific sedation or paralysis was given for the study.
na Infusion The 1 N HCI solution (1,000 mmol of HCI in 1 L of sterile water) was prepared in a glass bottle and infused via an internal jugular vein into the superior vena cava at a constant rate of 1 ml/kg/h, In each patient, the adequate position of the central venous catheter had
(Received in original form May 26, 1989 and in revised form July 26, 1989) I From the Department of Intensi ve Care, Erasme University Hospital, Brussels, Belgium. , Correspondence and requests for reprints should be addressed to Serge Brimioulle, M.D., Department of Intensive Care, Erasme University Hospital, Route de Lennik, 808, B-1070Brussels, Belgium.
1185
1186
BRIMIOULLE AND KAHN
TABLE 1
decreased from 36 to 23 mmollL (p
Patient No.
1 2 3 4 5 6 7 8
Sex F M M F M M M M
Age (yr)
75 61 73 72 72 42 37 83
as/at (%)
Main Diagnosis Brainstem infarct, coma Septic shock, cardiac arrest Decompensated COPD Decompensated COPD Pulmonary embolism Hepatitis, liver transplantation Hepatitis, liver transplantation Decompensated COPD
been confirmed by a chest roentgenogram before the infusion. At l-h intervals, blood samples were withdrawn through an arterial catheter for immediate determination of arterial blood gases. HCl infusion was discontinued when base excess had decreased below - 4 mmollL.
Measurements Hemodynamic, gasometric, and biologic measurements were performed before the HCl infusion and at 2 and 12 h after the end of infusion. Systemic arterial pressure was obtained from a radial or femoral catheter, and pulmonary arterial pressure, right atrial pressure, and pulmonary capillary wedge pressure were obtained from the Swan-Ganz catheter (Edwards Laboratories), using Statham P50 transducers (Gould Inc., Oxnard, CA) connected to a bedside hemodynamic monitoring system (Sirecust 404; Siemens, Erlangen, FRG, or HP 78354A; Hewlett-Packard, Palo Alto, CA). The pressure tracings and an electrocardiograph lead were recorded on a four-channel recorder (HP 7404; Hewlett-Packard). Pressures were referenced to the midchest level, and were read at end expiration. Cardiac output was measured with the thermodilution method using five injections, starting at the beginning of expiration, of lO-mlice-cold boluses. Arterial and mixed venous blood gases (ABL 2 or ABL 3 analyzer; Radiometer, Copenhagen, Denmark) and hemoglobin oxygen saturations (lL 282 CO-Oximeter; Instrumentation Laboratories, Lexington, MA, or OSM 3 hemoximeter; Radiometer) were measured immediately after drawing the samples and corrected for patient temperature. Biologicdata included blood cells counts, serum electrolytes, lactate, blood urea nitrogen, creatinine, and total protein concentrations. Calculations The following formulas were used for gas exchange calculations: alveolar Po, (PAo" mm Hg) = {(barometric pressure - 47) x FlO,} - {Paco, x (I - (I - R) x Flo,)/R}, where R is the respiratory gas exchange ratio; arterial and mixed venous blood oxygen contents (Can, and Cvo" mlldl) = {hemoglobin concentration (g/dl) x hemoglobin saturation (070) x 0.0139} + {oxygen tension (mm Hg) x 0.0031}; QS/QT (% of blood flow) =
35 17
15 20 29 25 12 11
Outcome Death Death Survival Death Death Survival Survival Survival
(Ceo, - Cao,)/(Ceo, - Cvo,) x 100, where Ceo, is the end-capillary blood oxygen content, estimated from the PAo, and the calculated corresponding saturation (6).
Statistics Results are expressed as mean values ± standard error. Statistical analysis consisted of an analysis of variance for repeated measures in the same patients (7). Postinfusion values were compared with preinfusion values using the method of contrasts (7); p values less than 0.05 were accepted as indicating statistical significance.
Results
Metabolic alkalosis was corrected after infusion of 4 to 8 (mean, 6) mmollkg of HCI, and acid-base status remained normal 12 h after the end of the infusion (table 2). No clinical complication was observed during or after the infusion. Arterial pH decreased from 7.55 to 7.38 (p < 0.001), and bicarbonate concentration
Hemodynamic data 2 h after the end of the HCI infusion showed an increase in pulmonary arterial pressure from 21 to 26 mm Hg (p < 0.05), without increase in cardiac output or in pulmonary capillary wedge pressure. Pan, increased from 76 to 114 mm Hg (p < 0.001), and Sao 2 increased from 96 to 97070 (p < 0.10). Pvo2 increased from 33 to 37 mm Hg (p < 0.01), whereas Svo 2 remained unchanged. QsI QT decreased from 21 to 16% (p < 0.01). Significant changes in Pao, and QS/QT persisted 12 h after the end of the infusion, whereas pulmonary arterial pressure returned almost to the initialleveI. Oxygen consumption did not change during the study. In order to investigate whether the effects of pH were dependent on the severity of the respiratory failure, patients were separated into two groups according to whether their initial QS/QT was below or above 20% (table 3 and figure 1). Patients with less venous admixture (n = 4) showed a moderate increase in Pao2 without improvement in Sao, or QS/QT. Patients with more severerespiratory failure (n = 4) had a greater and sustained increase in Pao., together with a significant and persistent improvement in Sao 2 and QS/QT. Other variables showed similar changes in the two groups of patients. HCI infusion was associated with some increase in white blood cells (p < 0.001) and a decrease in platelets (p < 0.01) (ta-
TABLE 2 ACID·BASE STATUS, HEMODYNAMICS, AND GAS EXCHANGE BEFORE AND AFTER HCI INFUSION
Arterial pH Paco" mm Hg HCO;, mmol/L
Pre-HCI
HCI at 2 h
HCI at 12 h
7.55 ± 0.Q1 40 ± 2 36 ± 1
7.38 ± 0.02:1: 39 ± 2 23 ± 1:1:
7.40 ± 0.02:1: 40 ± 2 25 ± 1:1:
HR, beats/min CO, Llmin/m' Psa, mm Hg Pra, mm fig Ppa, mm Hg Ppcw, mm Hg
97 3.5 81 8 21 11
± ± ± ± ± ±
8 0.4 6 1 2 2
90 ± 7 3.2 ± 0.4 90 ± 7t 8±2 26 ± 3' 11 ± 2
89 3.6 89 8 22 12
± ± ± ± ± ±
7 0.4 7t 1 2 2
Pao" mm Hg Pvo" mm Hg Sao" % svo, % as/aT, %
76 33 96 64 21
± ± ± ± ±
6 1 1 3 3
114 37 97 64 16
13:1: 1t 1 3 2t
98 37 97 66 18
± ± ± ± ±
7' 1t 1 3 2'
± ± ± ± ±
Definition of abbreviations: HR = heart rate; CO = cardiac output; Psa = systemic arterial pressure; Pra = right atrial pressure; Ppa = pulmonary arterial pressure; Ppcw = pulmonary capillary wedge pressure; PiIo, = mixed venous 0, pressure; Silo, = mixed venous 0, saturation; QS/QT = venous admixture. • p < 0.05 versus pre-HC!. t p < 0.01 versus pre-HCI. p < 0.001 versus pre-HCI.
*
20%). Acknowledgment The writers thank P. Lejeune, M.D. for his constructive review of the manuscript. References 1. Hodgkin JE, Soeprono FF, Chan DM. Incidence of metabolic alkalemia in hospitalized patients. Crit Care Med 1980; 12:725-8. 2. Nattie EE. Gas exchange in acid-base disturbances. In: Fahri LE, Tenney SM, eds. Handbook of physiology. Vol. 4. The respiratory system, gas exchange. Bethesda: American Physiological Society, 1987; 421-38. 3. Haas F, Bergofsky EH. Effect of pulmonary vasoconstriction on balance between alveolar ventilation and perfusion. J Appl Physio11968; 24:491-7. 4. Frans A, Turek Z, Yokota H, Kreuzer F. Effects of variations in blood hydrogen ion concentration on pulmonary gas exchange of artificially ventilated dogs. Pfluegers Arch 1979; 380:35-9. 5. Brimioulle S, Vincent JL, Dufaye P, Berre J, Degaute JP, Kahn RJ. Hydrochloric acid infusion for treatment of metabolic alkalosis: effects on acidbase balance and oxygenation. Crit Care Med 1985; 13:738-42. 6. Severinghaus JW. Simple, accurate equations for human blood O. dissociation computations. J Appl Physiol 1979; 46:599-602. 7. Winer BJ. Statistical principles in experimental design. New York: McGraw-Hill, 1971. 8. Grover RF, Wagner WW, McMurtry IF, Reeves JT. Pulmonary circulation. In: Shepherd JT, Ab-
METABOLIC ALKALOSIS AND PULMONARY GAS EXCHANGE
boud FM, eds, Handbook of physiology. Vol. 3. The cardiovascular system, peripheral circulation and organ blood flow. Bethesda: American Physiological Society, 1983; 103-36. 9. Fishman AP. Pulmonary circulation. In: Fishman AP, Fisher AB, eds. Handbook of physiology. Vol. I. The respiratory system, circulation and nonrespiratory functions. Bethesda: American Physiological Society, 1985; 93-165. 10. Sylvester JT, Gottlieb JE, Rock P, Wetzel RC. Acute hypoxic responses. In: Bergofsky EH, ed, Abnormal pulmonary circulation. New York: Churchill Livingstone, 1986; 127-65. 11. Benumof JL, PirioAF, Johanson I, Trousdale FR. Interaction of Pvo2 with PAo2 on hypoxicpulmonary vasoconstriction. J Appl Physiol1981; 51: 871-4. 12. Pease RD, Benumof JL, Trousdale FR. PAo2 and PV02interaction on hypoxic pulmonary vaso-
constriction. J Appl Physiol 1982; 53:134-9. 13. Hughes JD, Rubin LJ. Relation between mixed venous oxygen tension and pulmonary vascular tone during normoxic, hyperoxic and hypoxic ventilation in dogs. Am J Cardiol 1984; 54:1118-23. 14. West JB, Wagner PD. Pulmonary gas exchange. In: West JB, ed. Bioengineering aspects of the lung. New York: Marcel Dekker, 1977; 361-457. 15. Orr JA, Shams H, Fedde MR, Scheid P. Cardiorespiratory changes during HCI infusion unrelated to decreases in circulating blood pH. J Appl Physiol 1987; 62:2362-70. 16. Shams H, Peskar BA, Scheid P. Acid infusion elicits thromboxane A,-mediated effects on respiration and pulmonary hemodynamics in the cat. Respir Physiol 1988; 71:169-83. 17. Shirer HW, Erichsen DF, Orr JA. Cardiorespiratory responses to HCI vs lactic acid infusion. J Appl Physiol 1988; 65:534-40.
1189 18. Lloyd TC Jr. Influence of blood pH on hypoxic pulmonary vasoconstriction. J Appl Physiol 1966; 21:358-64. 19. Marshall C, Lindgren L, Marshall BE. Metabolic and respiratory hydrogen ion effects on hypoxic pulmonary vasoconstriction. J Appl Physiol 1984; 57:545-50. 20. Kwun KB, Boucherit T, Wong J, Richards Y, Bryan-Brown C. Treatment of metabolic alkalosis with intravenous infusion of concentrated hydrochloric acid. Am J Surg 1983; 146:328-30. 21. Adrogue HJ, Madias NE. Changes in plasma potassium concentrations during acute acid-base disturbances. Am J Med 1981; 71:456-67. 22. Tobin RB. In vivo influences of hydrogen ions on lactate and pyruvate in blood. Am J Physiol 1964; 207:601-5.
