Myocardial and Systemic Vascular Responses to Low Concentrations of Carboxyhemoglobin

STEPHEN M. AYRES, M.D.,* STANLEY GIANNELLI, JR., M.D., HILTRUD MUELLER, M.D., AND ANTOINETTE CRISCITIELLO, R.N., B.S., M.A.

St. Vincent’s Hospital and Medical Center of New York, New York, NY 10011 and *The St. Vincent Hospital, Worcester, MA 01610

ABSTRACT Carboxyhemoglobin alters the oxyhemoglobin dissociation curve in such a manner that oxygen is released to the tissues with great difficulty and at a lower oxygen tension. The known effects on heart and brain of breathing low concentrations of carbon monoxide are primarily related to this leftward shift and perhaps also to its combination with myoglobin and certain iron-contain ing enzymes. Hemoglobin-oxygen equilibria in the presence of carboxyhemo globin resemble the equilibria of more primitive forms of hemoglobin and give rise to the suggestion that this decrease in the access to oxygen is a form of counter-evolution.

The human toxicology of carbon monox malities in patients with coronary artery ide was first studied in detail by Haldane disease. 1 in 1895.14 A series of courageous experi ments performed on himself led to the con Theoretical Considerations clusion that the symptoms of carbon mon The respiratory characteristics of hemo oxide intoxication closely resembled the globin may be defined by ( 1 ) the oxygen effects produced by exposure to decreased capacity and ( 2 ) the shape of the dissocia concentrations of oxygen. Although he did tion curve. Since one gram of hemoglobin not observe serious symptoms at rest until will reversibly bind 1.34 ml of oxygen, the his hemoglobin was at least one third sat oxygen capacity is closely related to hemo urated with carbon monoxide, exertion pro globin concentration. Invertebrates have duced mild dyspnea and palpitations when oxygen capacities ranging from one to two as little as 14 percent carboxyhemoglobin ml per 1 0 0 ml of blood; fish, amphibia and (COHb) was present. Recently presented reptiles carry between 5 and 15 ml per 100 data from our laboratory suggest that ml of blood while mammals carry between COHb concentrations as low as 5 percent 15 and 20 ml per 100 ml of blood. The may interfere with oxygen delivery to the capacity may be as high as 40 ml per 100 diseased myocardium. 5 More recent studies ml of blood in certain diving vertebrates. have confirmed this and demonstrated that Equal in importance to the oxygen ca low concentrations of carbon monoxide pacity of hemoglobin is the affinity of he may provoke electrocardiographic abnor moglobin for oxygen. Many attempts to

4 4 0 RESPONSES TO LOW CONCENTRATIONS OF COHb 4 4 1

describe mathematically the oxyhemoglo bin dissociation curve have been made, but the concept of loading and unloading ten sions, proposed by Krogh and Leitch in 1919,15 has survived as a simple and unique description of hemoglobin equilibria. The loading tension (ti) is the oxygen tension permitting hemoglobin to become 95 per cent saturated with blood; the unloading tension (tu) is the tension associated with half saturation. The former is related to ease of uptake within the lung, the latter to ease of unloading within the peripheral tissues. Generally, the two values move in the same direction so that a conflict exists F i g u r e 1. Oxyhemoglobin dissociation c u r v e s since a low tx is needed to insure adequate for normal adult hemoglobin myoglobin and vary ing mixtures of oxyhemoglobin and carboxyhemo saturation in the lung while a high tu is globin. The presence of significant amounts of needed to insure rapid diffusion of oxygen carboxyhemoglobin (curves C, D, and E ) reduces from blood to tissues. the total amount of oxygen carried and also changes the shape of the curve so that a lower Another characteristic of hemoglobin, oxygen tension results from the same oxygen con important in an understanding of the tent. The carboxyhemoglobin curves approach the pathophysiology of carboxyhemoglobin, is dissociation curves for human myoglobin. Clin. Pharmacol. Ther. 11:337, 1960. the Bohr shift. Decreasing pH shifts the curve to the right and raises the tu, enhanc ing tissue oxygenation; increasing pH shifts conventional oxygen saturation to demon the curve to the left and increases ti, which strate both of these effects. Increasing car benefits oxygen uptake in the lung but de boxyhemoglobin concentration decreases creases tissue oxygen tension by decreasing oxygen capacity and the tu. tu. In general, a pH decrease of 0.2 units in The tu oxyhemoglobin in the presence of human blood produces an increase in tu various concentrations of carboxyhemoglo from 27 to 34 mm Hg. One might predict bin may be calculated by equations first a rightward shift in conditions associated published in 1912 by Douglas, Haldane with decreased tissue oxygen availability and Haldane. 1 1 By way of footnote, John S. (low blood flow) and an increase with con Haldane and a young assistant, C. G. ditions associated with decreased atmo Douglas (of later altitude physiology fame), did the experimental work while spheric oxygen concentration (altitude). Haldane’s 20 year old son, John Burdon Sanderson Haldane, provided the math R e s p ir a t o r y C haracteristics ematical treatment. Unloading tensions for The respiratory characteristics of car a number of COHb mixtures together with boxyhemoglobin approach those of primi those from a number of more primitive tive hemoglobins. The oxygen capacity is animals are shown in figure 2. One can diminished by the volume of carbon mon envision hemoglobin evolution as an ascent oxide bound and the dissociation curve is up the tu ladder. Note that a human with a shifted to the left decreasing the tu. In COHb saturation of 30 percent has slipped figure 1 is shown the oxyhemoglobin dis well down the ladder and is somewhere sociation curve, plotting oxygen content between a newborn goat and an embryonic against oxygen tension instead of the more chick in the evolutionary ladder. 4 4 2 AYRES, GIANNELLI, MUELLER AND CRISCITIELLO

exchange. The coronary circulation differs markedly from the systemic circulation in its mode of adaptation to the increased oxygen requirements of stress. Peripheral tissues normally extract about 25 percent • 0 A Elephant • o pH 7.40 of the oxygen present in arterial blood and a Newborn Goal • pH 736 a Llama the remaining 75 percent serves as a re

15- serve supply. Total oxygen uptake may be mm Hg increased by increasing the amount of ox a Chick Embryo, 12 Days 10- ygen extracted from the perfusing blood and by increasing the rate of blood flow. In periods of stress, for example, mixed venous oxygen tension may drop from a 10 20 30 40 50 60 70 normal of 40 mm Hg to below 20 mm Hg. COHb, % Sat. In contrast, the exorbitant resting myocar

F i g u b e 2. Effect of increasing carboxyhemoglo- dial oxygen needs are met by extracting a bin saturation on unloading tension (tu) or P». much greater fraction of oxygen from the Note that a human with 15 percent carboxyhemo- perfusing coronary blood than is extracted globin has oxygen-hemoglobin equilibria interme diate between an elephant and a newborn goat. by the peripheral tissues from the systemic blood. Approximately 75 percent of oxygen Changes in gas exchange, following ex is extracted from the coronary circulation posure to carbon monoxide, are dictated by at rest and coronary venous blood is about changes in the oxyhemoglobin dissociation 25 percent saturated, corresponding to an curve in the presence of various concen oxygen tension of 20 mm Hg. If increased trations of COHb. The curve is shifted to myocardial oxygen needs were met by in the left, so that oxygen tension must de creasing oxygen extraction, as in the pe crease if the same arteriovenous oxygen ripheral circulation, coronary venous and difference and venous saturation are to be myocardial cellular oxygen tension would maintained. The decrease in mixed venous fall to dangerously low levels. The coronary oxygen tension is presumably the primary circulation avoids low oxygen tensions by event; other hemodynamic changes are increasing flow rate in response to in probably compensatory. The decrease in creased oxygen demands rather than in mixed venous and coronary sinus oxygen creasing oxygen extraction. Near maximal tension (and also in peripheral, tissue and resting extraction and the ability to in myocardial oxygen tension) appears to ini crease oxygen consumption by increasing tiate a series of stress responses which at flow allow performance of continuous con tempt to compensate for the decrease in tractile work at rest and also provide the oxygen tension. In general these responses ability to increase cardiac work with exer resemble those seen with increased adre cise. Coronary vascular disease may pre nergic activity. vent an increase in local coronary blood flow in response to need. In this situation, M y o c a r d ia l a n d S y s t e m ic O x y g e n the myocardium is forced to extract more E x c h a n g e oxygen but does this at the expense of a Changes in coronary blood flow and markedly reduced coronary venous and myocardial oxygen extraction can best be tissue oxygen tension. appreciated by considering the differences Myocardial oxygen consumption, the between myocardial and systemic oxygen product of coronary blood flow and the RESPONSES TO LOW CONCENTRATIONS OF COHb 4 4 3 difference between arterial and coronary diffusion gradient so that the arteriovenous sinus oxygen content, is a complex function difference (or extraction) is a function of of external ventricular work, myocardial arterial lactate or pyruvate concentration. wall tension and fiber length, rate of de The extraction ratio of either lactate or velopment of ventricular pressure and pyruvate (the arterio-coronary sinus differ heartbeat rate. 12 Increasing oxygen require ence divided by arterial concentration) is ments must be met by increasing oxygen a useful index for expressing extraction, consumption, otherwise myocardial ische which is relatively independent of arterial mia and necrosis will develop. If oxygen concentration. Although increasing arterial is plentiful, glucose and lactate are ex concentrations may increase the extraction tracted from the coronary perfusate and ratio slightly, the ratio does not decrease metabolized to pyruvate in the glycolytic with decreasing arterial concentration so cycle. Pyruvate from this source, together that a decrease in extraction ratio may be with directly extracted pyruvate and ac considered due to some intracellular altera etate fragments from fatty acid degrada tion. The extraction ratio for both metab tion, are converted to acetyl coenzyme A olites ranges from 15 to 30 percent when and oxidized in the citric acid cycle liber oxygen is plentiful. During periods of in ating energy, carbon dioxide and water. In adequate oxygenation, extraction decreases periods of insufficient oxygen supply, the and lactate and pyruvate may actually be citric acid oxidative cycle is inactivated and produced by the myocardium so that cor pyruvate accumulates in the cytoplasm. A onary sinus concentrations exceed arterial small amount of energy is produced by concentrations. glycolysis as the enzyme nicotinamide adenine dinucleotide ( NAD, or DPN in the C arboxyhemoglobinemia S y s t e m ic older classification) oxidizes one of the H emodynamics glycolytic intermediates and produces high Several investigators have studied the ef energy bonds. The cytoplasmic supply of fects of carboxyhemoglobinemia systemic NAD would soon be exhausted if a mech hemodynamics. Chiodi8 observed increases anism for its regeneration did not exist. in pulse rate and cardiac output with Accumulating pyruvate is reduced to lac COHb concentrations as low as 16 percent. tate by the reduced form of NAD ( NADH) Both Haldane14 and Haggard13 observed and, in the process, NADH is oxidized to hyperventilation during carbon monoxide NAD, permitting continued operation of inhalation in man and in experimental an the glycolytic cycle. In summary, inade imals although Chiodi et al8 were unable quate oxygen decreases the utilization of to measure significant increases in ventila pyruvate by the citric acid cycle, pyruvate tion in their studies. accumulates in the cytoplasm and lactate Chevalier et al7 evaluated the overall is produced from the accumulating pyru cardiorespiratory adaptation to exercise by vate. measuring oxygen consumption before, during and after exercise in normal subjects M e t a b o l ic E x t r a c t io n s prior to and after inhalation of sufficient Increasing cellular concentrations of carbon monoxide (0.5 percent) to raise lactate and pyruvate decrease the diffusion COHb concentration to an average of 3.95 gradient between coronary perfusate and percent. Total oxygen consumption aver cytoplasm and decrease the extraction of aged 5,990 ml before CO inhalation and these metabolites. Normally, arterial lactate 5,794 ml after inhalation, the difference and pyruvate concentrations determine the representing a slight training effect. The 4 4 4 AYRES, GIANNELLI, MUELLER AND CRISCITIELLO

■ CONTROL increased from 6.86 to 8.64 1 per min, while B FOLLOWING CO arterial carbon dioxide tension fell from 40 to 38 mm Hg. Systemic oxygen extraction ratios increased from 0.27 to 0.32, indicat ing more complete extraction of oxygen’ from perfusing arterial blood. Mixed ve nous oxygen tension fell from 39 to 31 mm Hg as a result of the leftward shift of the oxyhemoglobin dissociation curve. Arterial oxygen tension unexpectedly fell from an average value of 81 to 76 mm Hg.

F i g u r e 3 . Arterial oxygen tension (Pao2), mixed This finding was recently discussed in de venous oxygen tension (Pvo2), alveolar-arterial ox tail4 and agreement reached with Brody ygen difference (AaD), minute ventilation (Ve), cardiac output ( Qs ) and oxygen consumption and associates6 that it is related to enhance (Vo2) in a group of patients before and after ment of the venoarterial shunt effect and breathing five percent carbon monoxide for 45 is more prominent in patients who initially seconds. Brackets indicate standard error of dif ference between means. have low oxygen tensions. Increased ven tilation tends to minimize the decrease in arterial oxygen tension in conscious man, oxygen debt was 21 percent of total oxygen although the alveolar-arterial oxygen dif consumption prior to CO and rose to 24 ference remains elevated. Dogs ventilated percent (p < 0.05) afterwards, suggesting at constant tidal volumes show marked de that oxygen supply during exercise was in creases in arterial oxygen tension with car adequate for metabolic requirements. The bon monoxide breathing because of in overriding importance of this study was creased atelectasis and the inability to that COHb concentrations as low as 3.95 hyperventilate. Such observations suggest percent could call forth adaptive responses that carbon monoxide inhalation would on the part of the peripheral circulation. have a significant effect on arterial oxygen These investigators also observed small but tension in patients with pre-existing lung significant decreases in inspiratory capacity disease, individuals who are heavy smokers and total lung capacity together with sta and patients in coma from severe carbon tistically borderline decreases in vital ca monoxide poisoning. pacity following carbon monoxide inhala These studies were repeated using the tion. lower concentration of carbon monoxide. In 1965 and again in 1970, we reported Cardiac output did not change signifi the systemic hemodynamic and respiratory cantly, although arterial carbon dioxide response to acute elevation of COHb in tension decreased indicating hyperventila man by means of measurements performed tion. Changes in arterial and mixed venous during diagnostic cardiac catheterization.3’4 oxygen tensions were similar to those ob Saturations of COHb between 6 and 12 served with the higher concentrations. percent were achieved by the breathing of Myocardial studies were performed be either 5 percent CO for 30 to 45 seconds or fore and after the administration of either 0.1 percent CO for 8 to 15 minutes. 5 percent or 0.1 percent carbon monoxide In figure 3 are summarized our findings for the time periods listed. Patients were following administration of the higher con divided into two groups, those with cor centration. Cardiac output increased from onary artery disease and those with other 5.01 to 5.56 1 per min; minute ventilation cardiopulmonary disorders. r e s p o n s e s t o l o w concentrations o f COHb 4 4 5

The myocardial arteriovenous oxygen difference is more than twice the systemic difference because of the high oxygen re quirements of the contracting myocardium. Increased oxygen demands are met by in creasing coronary blood flow since further increases in oxygen extraction would so lower coronary venous oxygen tension that myocardial cells would face certain hy poxia. In figure 4 it is shown that the lowest concentration of carbon monoxide de creased the myocardial arteriovenous ox ygen difference an average of 6 . 6 and 7.9 percent (in the patients with other cardio pulmonary diseases and coronary disease respectively) while the higher concentra tion decreased the myocardial arteriove F i g u r e 5. Coronary blood flow ( C B F ) follow ing carbon monoxide breathing plotted against nous oxygen 25 and 30.5 percent. The same that obtained during control state. Line of identity figure also presents the effects of higher is shown and points to left of line indicate increase levels of carboxyhemoglobin in a series of in coronary blood flow with carbon monoxide. CAD refers to coronary artery disease; other indi canine experiments. cates all other heart disease. Ann. N. Y. Acad. Sci., Myocardial oxygen consumption could Oct. 5, 1970. only be maintained in the face of a de sponse to increasing COHb. Myocardial creasing arteriovenous oxygen difference tissue oxygen tension (coronary sinus mea by the adaptive maneuver of increasing surements) decreased in all but one patient blood flow. In figure 5 it is demonstrated (figure 6 ). that coronary blood flow did increase in all Several recent studies have confirmed but two of the studies, regardless of the our findings that relatively low concentra dose delivered. These changes were statis tions of carboxyhemoglobin can interfere tically significant for three of the four with myocardial and systemic oxygen de groups. Neither the presence of coronary livery in man. Dhinsda et al1 0 administered artery disease nor the concentration of carbon monoxide intermittently to three carbon monoxide appeared to alter the re- monkeys for 12 different 30 minute periods each day over an eight month period and Human Con;ne achieved carboxyhemoglobin levels of 1 1 to 12 percent. Cardiac output was un changed but the arteriovenous oxygen dif ference and oxygen consumption decreased while mixed venous oxygen difference was unchanged. DeBias et al9 administered 100 ppm carbon monoxide continuously to a 0 10 2 0 3 0 4 0 A COHb, % Soturotion series of monkeys achieving carboxyhemo globin saturations averaging 12.4 percent. F i g u r e 4. Ratio of coronary arteriovenous ox ygen difference during carbon monoxide breathing These animals exhibited an increase in (Ca-Ccsco) to control arteriovenous oxygen dif hematocrit from an average of 38 to 52 ference (Ca-Ccscontpoi) plotted on vertical axis percent and also developed significant elec against change in COHb saturation. Ann. N. Y. Acad. Sci., Oct. 5, 1970. trocardiographic abnormalities. T wave in- 4 4 6 AYRES, GIANNELLI, MUELLER AND CRISCITIELLO

glycerate and (5) increased oxygen ex traction. The response chosen appears to depend upon species, status of the vascular bed and intensity and duration of the hy poxic stress. Our studies in the resting human subject indicated that an increase in coronary blood flow was an almost universal re sponse to either high or low concentrations of carbon monoxide. In contrast, systemic blood flow increased only when the sub jects were exposed to the higher concentra tion. Vogel and Gleser1 6 used a lower con centration for longer periods of time and F i g u r e 6 . Coronary sinus oxygen tension ( Pcs02) also failed to observe increases in cardiac following carbon monoxide breathing plotted against that obtained during control state. Line of output. These investigators observed that identity is shown; points to right of line indicate exercise produced higher than control out decrease in coronary sinus oxygen tension with puts and lower maximal oxygen consump carbon monoxide. CAD indicates coronary artery disease; other refers to all other heart disease. Ann. tions, suggesting that both increased blood N. Y. Acad. Sci„ Oct. 5, 1970. flow and a reduction in oxygen require version was common, indicating myocardial ments were important responses to car-' ischemia. In addition, the animals uni boxyhemoglobin-induced hypoxia during^ formly developed increased P wave ampli exercise. tude suggesting the development of pul Dhinsda’s resting monkeys, intermittently monary hypertension. Vogel and Gleser16 exposed to low concentrations, minimized exposed healthy volunteers to somewhat hypoxia by decreasing oxygen require higher concentrations (225 ppm or 0.0225 ments10 while Debias’ monkeys, contin percent) for a sufficient period of time to uously exposed to somewhat lower con raise carboxyhemoglobin concentrations to centrations, developed polycythemia. The 18 to 20 percent. Resting cardiac output inability of this polycythemic response to and heart rate were not increased but were protect the monkeys against myocardial higher than control values at each of three hypoxia is revealed by the uniformly ob levels of exercise. Maximal oxygen con served electrocardiographic abnormalities. sumption was 23 percent less than control. None of the studies demonstrated marked increase in oxygen extraction. This response Adaptations to Carboxyhemoglobin would expose tissues to lowered oxygen Induced Hypoxia tensions and is probably the least desirable The data reviewed suggest that the of the physiologic responses suggested. mammalian cardiopulmonary system ex Most of these adaptive responses would hibits one or more of the following re probably serve admirably for normal man. sponses, which minimize cellular hypoxia The development of coronary atherosclero when exposed to carboxyhemoglobin—in sis, however, might be predicted to render duced hypoxia: ( 1 ) decrease in oxygen any of these adaptive responses ineffective requirements; ( 2 ) increase in blood flow; and expose the myocardium to potentially (3) polycythemia; (4) rightward shift of lethal hypoxia. This prediction has been the oxyhemoglobin dissociation curve me graphically confirmed by the recent obser diated through increases in 2,3 diphospho- vations of Aronow et al. 1 These investi RESPONSES TO LOW CONCENTRATIONS OF COHb 4 4 7 gators exposed 1 0 patients to heavy free 5. A y r e s , S. M., M ueller, H., Gregory, J., G iannelli, S., Jr., and Penny, J.: Systemic way automotive traffic by driving them in and myocardial hemodynamic responses to a station wagon for 90 minutes. Mean relatively small concentrations of carboxy COHb concentration rose from 1.12 to 5.08 hemoglobin. Arch. Environ. Health 18:699— percent saturation and then decreased to 709, 1969. 6. B r o d y , J. S. and Coburn, R. F.: Carbon 2.91 percent saturation two hours later. monoxide-induced arterial hypoxemia. Science Four of the subjects exhibited electrocar J 64:1297-1298, 1969. diographic changes during the exposure 7. Chevalier, R. B., Krumbolz, R. A., a n d R o s s , J. C . : Reaction of non-smokers to carbon period; they did not experience changes monoxide inhalation. J.A.M.A. 198:1061-1064, during a subsequent drive when provided 1966. with a source of uncontaminated air. Im 8. Chiodi, H., D ill, D. B., Consolazio, F., a n d H o r v a t h , S. M.: Respiratory and circulatory mediately following the exposure period, responses to acute carbon monoxide poisoning. the subjects were exercised on a bicycle Amer. J. Physiol. 134:683-693, 1941. until they developed angina pectoris. The 9. Debias, D. A., B a n e r j e e , C. M., a n d B i r k - time necessary to develop angina was 249.4 h e a d , N. C.: Effect of carbon monoxide in halation on myocardial infarction in monkeys. seconds prior to exposure. The time fell to Presented at A.M.A. Air Pollution Research 174.3 seconds after exposure and rose to Conference, October 2-3, 1972. 210.8 seconds two hours later. In addition, 10. Dhindsa, D. S., O c h s n e r , A. J., I l l , M e t c a l f e , J., and M alinow, M. R.: Hemody angina developed at lower heart rates and namic effects of carbon monoxide in monkeys. systemic blood pressures following expo Fed. Proc. 31:355, 1972. sure to freeway air. 11. D o u g l a s , C. G., H aldane, J. S., a n d H a l d a n e , J. B . S.: The laws of combination of iieferences hemoglobin with carbon monoxide and ox ygen. J. Physiol. 44:275-304, 1972. 1. A r o n o w , W. S., H a r r i s , C. N., I s b e l l , M. 12. G a u l t , J. H., Ross, J., and Braunwald, E.: W., R o k a w , S. N., and Im parato, B.: Effect of heavy freeway traffic on cardiopulmonary Contractile state of the left ventricle in man. function in angina. Chest 62:358, 1972. Circ. Res. 22:451-463, 1968. Haggard, H. 2. A y r e s , S. M. and Buehler, M. E.: The effects 13. W.: Studies in carbon monoxide of urban air pollution on health. Clin. Pharma asphyxia. 1. The behavior of the heart. Amer. col. Ther. 11:337-371, 1970. J. Physiol. 56:390-403, 1921. 3 . A y r e s , S. M., G iannelli, S., Jr., and Arm 14. H a l d a n e , J.: The action of carbonic oxide on s t r o n g , R. G .: Carboxyhemoglobin: hemody man. J. Physiol. 18:430-462, 1895. namic and respiratory responses to small con 15. K r o g h , A. and Leith, L.: The respiratory centrations. Science _Z 4.9:193-194, 1965. function of the blood in fishes. J. Physiol. 52: 4 . A y r e s , S. M., G iannelli, S., Jr., and M uel 288, 1919. l e r , H.: Myocardial and systemic responses to 16. V o g e l , J. A. and G leser, M. A.: Effect of carboxyhemoglobin. Ann. N. Y. Acad. Sci. 174: carbon monoxide on oxygen transport during 268-293, 1970. exercise. J. Appl. Physiol. 32:234-239, 1972.

Supported in part by grants from The Council for Tobacco Research USA, The United States Public Health Service National Air Pollution Control Administration Grant #AP-00838 “Chronic Lung Disease Response to Pollution-Free Air” and The United States Public Health Service National Heart Institute Training Grant #HL-05686 “Cardiopulmonary Physiology.”