Carbon Monoxide

Exposure to carbon monoxide (CO) in large amounts or even small amounts for a long duration often results in death. CO is an odorless, colorless and tasteless gas in the exhaust produced in gasoline engines. Boats release CO through the exhaust ports of vessels that are either idling or underway or running generators. When inhaled, CO rapidly replaces oxygen in tissues. People are surprised they can get CO poisoning when their activity is outdoors but it is a silent killer.


RtSUM.E
On estime que 50% de toutes les intoxications mortelles aux Etats-Unis sont attribuables a l'oxyde de carbone. ll est difficile d'estimer le nombre d'intoxications non mortelles parce que les signes et sympt6mes de l'intoxication a l'oxyde de carbone sont nombreux et miment d'autres conditions pathologiques. ll est grave de poser un diagnostic erronne puisque le patient retournera souvent dans son environnement contamine. Ceux qui ne recoivent pas le traitement approprie ont un risque significatif de 10% a 12% de developper des sequelles neurologiques tardives. Le diagnostic d'intoxication a l'oxyde de carbone repose sur une histoire detaillee, un examen physique minutieux et un indice de suspicion eleve. D that carbon monoxide poisoning is the leading cause of death due to poisoning in the United States, accounting for an estimated 3500 fatalities each year.' The exact incidence of non-lethal poisoning is more difficult to gauge. Some have estimated a 2:1 ratio of survival; however, many others believe this ratio is considerably underestimated.
Underestimation occurs because presenting symptoms often lead to misdiagnoses ranging from flulike viral illness to acu te myocardial infarction. Grace and Platt2 suggest that carbon monoxide poisoning could well rank along with syphilis, tuberculosis, and subdural hematoma in its ability to simulate other systemic illnesses.
Carbon monoxide poisoning is rarely mentioned in medical school core lectures.
Most textbooks of general medicine do not cover it well, and only the newer editions of emergency texts deal with the subject in detail.
Carbon monoxide, as a pure gas, is colorless, odorless, and tasteless. Knowledge of its effects goes back centuries. During Cicero's time (106 to 43 Bc), charcoal fumes were documented as an effective means of suicide or execution. During the Dr Aniol is a Resident in the ,jrtheastern Ontario Fimil MlIedicine Progam in Sudburv.
15th and 16th centuries, coal became a popular source of fuel for domestic heating. In 1616, the purpose of the chimney was described, "that thereby, the smoke ... might be dispersed abroad in the aire for fear of hurting the bodies of men."' Workplace problems were also noted. One physician, Ramazzini (1633 to 1714), wrote a manual in which he described diseases afflicting workers. During his studies he noted that a large number of materials handled by workers produced "noxious vapors." Describing instances in coal miners and in confectioners preparing roasted nuts, he commented that burning coals had a certain "potent pernicious quality, which unless you give it a free outlet to escape by can kill a man almost on the instant."' J.S. Haldane, the respiratory physiologist, investigated many mining accidents in England at the end of the 19th century. He discovered that up to 95% of the miners died not from mechanical injuries or burns but from carbon monoxide poisoning. It was Haldane who suggested that rescue workers carry either a mouse or a canary with them into the mines. These creatures, having a large relative respired gas volume to body mass and a high metabolic rate, would show symptoms of distress and warn of high carbon monoxide levels before the rescuers were adversely affected. Haldane's studies led him to develop breathing equipment for these environments. During the late 1700s, gas converted from coal and coke was used as an illuminant. Gaswork companies were created and gas lighting became popular. In 1798 a factory in Birmingham, England, adopted gas lighting for the first time. In 1805 the first public gasworks was constructed at Salford, and in 1808 gas street lighting was installed in London.:$ From then on, manufactured gas became widely used and, wvith coal gas having a carbon monoxide concentration of 7.4%o, the problem of carbon mounoxide poisoninig also lecame xvidespread.

Sources
Carboni monoxide found in the human body results from endogenous production and exogenous uptake. Endogenous production is due to the catabolism of hemeconitaininig compounds to bilirubin and carbon monoxide. This produces baseline carboxvhemoglobin (HbCO) the union betwveein carbon monoxide and hemoglobin levels of less than 1% in healthy adults. In adults wvith hemolytic anemia or induction of hepatic microsomal heme enzymes, including cytochrome P-450 and possibly B5, HbC(O levels may exceed 5o.' Carbon monoxide is produced naturallv in the environment by the sponitaneous oxidatioin of organic matter. T'odax, carbon monoxide is produced mainly from the incomplete combustion of orgainic fuels. Obvious sources are vehicle exhaust, fires, and inadequately vented furniaces, water heaters, and space heaters. Automobiles are by far the largest producer of carbon monoxide. Researchers estimate that automobiles are responsible for more than half the total global output of carbon monoxide, 3 about 57.1 million tonnes per year. By comparison, all other forms of transportation combined (ie, shipping, flight, rail, diesel) release only 2.6 million tonnes. Vehicle emission standards have slightlv recluced the output from newer vehicles. TI'he catalytic convertor, an insulated chamber containiing a porous catalytic material, is one device used in an attempt to reduce carboni monioxide emissioIns Vehicle exhaust is particularly hazardous in enclosed areas like garages, repair shops, Nwarchouses, .ship's holds, andl indoor ice rinks. Researchers studving the effects of carbon moinoxide levels in indoor ice rinks have compiled 15 cases of children who were initially believed to have had food poisoning.' Other less obvious exogenous sources include incomplete combustion of canned fuels for heating food (ie, Sterno), cigarette smoke, and methylene chloride. Methylene chloride, a clear, colorless liquid with a sweet and pleasant odor, once used as a general anesthetic, is now commonly used in painit strippers, degreasing solvents, and solvents used to make photographic film. MIethylelne chloride is absorbed into the body at a rate dependent upon the minute respiratory volume and the ventilation of the Nworking space. It is then metabolized by the body to produce carbon monoxide andl HbCO. Exposure to methylenie chloride for 3 11ours can prodtice HbCO levels of 5%( to I0°/(,. Absorbed methylene chloride is released slowly from body stores and converted slowly to HbCO. Because the HbCO resulting from the conversion of methylene chloride has a half-life twice that of HbCO produced by carbon monoxide gas exposure, the levels of cardiovascular stress, as well as other symptoms, may be prolonged.' See Table I for a more complete list of sotirces.

Pathophysiology
The effects of carbon monoxid( are related to its interactioin wvith the body's various pigmenits. Thle major target is the pigment fouInd in red blood cells, hemoglobiin. In 1857, Bernard demonistrated that carboin monoxide could result in hypoxia by reversibly combining with hemoglobin.' Mlyoglobin (a muscle pigmenit) aind cytochrome P-450 (located within the mitochondria) also appear to be targeted. T hese pigmenits all cointain iron, which tends to react with carbon moinoxide.
Heme, occurring free in solutioin, has ain affinity for carbon monoxide that is approximately 25 000 times greater thain its affinity for oxygen. However, when heme molecules become bound in the form of heemoglobin, the arrangemeint of proteins reduces this affinity 220 to 250 times, depeindinig OIn the type of hemoglobiin.8" ExogenIous uptake of carbon monoxide generallx results froim inhaliig a gas mix-tuIre contaminaiiiltcd with carboni moioxide. (Oine exceptioIn is carbon mnoiloxide prodtuced by, the bodx 's metabolism of meth 1ene chloride.) OInce iinhialed, carbon monoxide diffuses quite readily across the alveolar-capillary membrane. As a result, the proportion of carbon monoxide gas in the alveolus is similar to that in pulmonary capillary blood and at the red blood cell membrane. Once at this site, the carbon monoxide binds to the hemoglobin molecule in only a few hundred milliseconds at body temperature.
The binding of carbon monoxide to hemoglobin appears to affect the transport of oxygen two ways. Under normal conditions, cellular function requires approximately 5 mL of oxygen per 100 mL of blood. The presence of HbCO will decrease arterial oxygen content, as HbCO carries no oxygen, resulting in an effect much like anemia. Differing from anemia, however, the binding of carbon monoxide to hemoglobin also results in a tighter binding of the remaining oxygen, causing a leftward displacement of the oxyhemoglobin dissociation curve. This increased affinity of hemoglobiin for oxygen is related to the elimination of the first of the four sites of heme oxygenation by carbon monoxide.]7" Under normal circumstances, when oxygen binds with the first heme unit, it will alter the hemoglobin structure such that further heme-oxygen interactions are facilitated. If carbon monoxide binds this first heme unit, then any further heme-oxygen interactions are not easily achieved. The remaining oxygen will be bound more tightly to the hemoglobin molecule.
Explanations of the health effects of carbon monoxide poisoning have concentrated on this mechanism. Clinical observation and animal data, however, have suggested that carbon monoxide toxicity involves other mechanisms, resulting in tissue hypoxia separate from the impairment of oxygen transport. We know that other heme proteins (myoglobin, cytochrome P-450, peroxidases, catalases) also bind carbon monoxide and that this extravascular uptake approaches IO% to 15%)o of the total body burden of carbon monoxide.' Research has demonstrated that these proteins bind a sufficient quantity of carbon monoxide to inhibit their function in vitro. " Raybourn et al"' studied the effects of hypoxia and carbon monoxide on the activity of cerebellar Purkinje's cells and concluded that carbon monoxide induced a greater cellular toxicity than hypoxia alone. Photodissociation of carbon monoxide from the presumed binding site led to the belief that this site was cytochrome a,3 oxidase. Raybourn and other researchers ' hypothesize that carbon monoxide binds competitively to cytochrome a, the terminal oxidase of the mitochondrial electron transport chain, and thereby blocks cellular respiration. Others have evidence suggesting that this is not likely to be involved clinically in carbon monoxide poisoning. H The effects upon the binding of myoglobin as a mechanism depends upon the function of myoglobin and how important it is to cellular oxygenation. MIyoglobin may be involved in supplying oxygeni to the mitochondria of skeletal as well as heart muscle by helping facilitate the diffusion of oxygen into the cytoplasm. :l Coburn' has suggested that the ratio of cardiac carboxymyoglobin to circulatiing HbCO is approximately 3:1. As a result, individuals with HbCO levels of 10%/( have approximately 30%/o of their myoglobin saturated, which could decrease the oxygen reserve available to the myocardium.
Myocardial depressioni is a potential result of this, often confirmed clinically on electrocardiographic examination.
Carbon monoxide toxicity can thus be explained by the reduction of hemoglobin oxygen-carrying capacity, a leftward shift of the hemoglobin dissociation curve, and the possibility of effects from myoglobin binding.
NMost carbon monoxide is eliminated from the body across the alveolar-capillary membrane. A small amount, however, is oxidized within the body to carbon dioxide. Elimination, like uptake, is determined by alveolar ventilation, partial pressure of oxygen and carbon monoxide, pulmonary diffusing capacity, and rate of endogenous carbon monoxide production. NMany researchers believe the half-life of carbon monoxide (the time required to eliminate half the amount carried within the body) follows a single exponential curve. The half-lives for carbon monoxide have been determined for airas well as for oxygenbreathing mediums and will be discussed later ( The signs and symptoms of carbon monoxide poisoning cover a wide spectrum, affecting most of the body's systems. Those systems having the greatest oxygen demand and greatest amount of blood flow, particularly the heart and central nervous system, will be affected to the greatest degree. Carbon monoxide poisoning may mimic many conditions, including alcohol or drug intoxication, psychiatric disorders, flulike illnesses, and otlhers that can lead to misdiagnosis. The concomitant effects of alcohol or drugs furtlher compound the diagnostic problem.
Haldane:3 wvas among the first researchers to document the effects of carbon monoxide on the human body. In his laboratory he exposed himself, as wvell as some of his colleagues, to high levels of carbon monoxide. 'These high-level exposures were of short durationi and Haldane had oxygen available for breathing afterward.
Such exposures had no long-term effects. Haldane noted that, at blood HbCO levels greater than 50(% saturation, vision, hearing, and intelligenice were impaired. He also noted that in miners, who suffered long, mid-level carboni monoxide exposures, the nerVous system was likely to remain affected for a period stretching from days to monithls, during which coginition and memory were impaired.:' T he earliest complaints of carboon monoxide poisoninig appear to be fronital healdache and nausea. Other complaints include vomiting, dizziness, ataxia, weaknless, visual disturbances, ColnfUsioIl, Coginitive impairmenit, clyspnea, palpitations, anid loss of consciousnless. This short list readily inidicates howv maniy otlher diagnoses can be considered. Individuals with any degree of coronary artery narrowing due to atherosclerosis may be especially affected by any carbon monooxide conitaminiated enlVironment. lThe heart will be required to increase its workload to proxvide ain iIcreased supply of oxygen for the hypoxic peripheral tissues. l)ecreased oxygeni supply to the myocardiumn anid increased demancd by the myocardium will lead to angiina, myocardial damage, anid arrhythmia.
Anidersoni et al'' studied 10 meni with stable aInginla. Thllese patienits wvere exposed to car-boin moinoxide levels of 50 ppm aind 1 00 p)p)m1l, conlccietIatioIns similar to those found in many urban areas, for 4 hours. This exposure produced HbCO levels ranging from 2.9% to 4.50%. The patients underwent treadmill-exercise electrocardiography after each exposure. The authors demonstrated that exercise time before the onset of angina was shortened _______________________________________________________________________________________________ L absorbance spectra that are similar in the red wavelength range. As a result, the oximeter is unable to differentiate the two and could overestimate the oxygen saturation of arterial blood.211 A co-oximeter is, therefore, needed to differentiate. Metabolic acidosis may be present because of lactic acid production. Electrocardiographic results are frequently  (Table 3).
Carbon monoxide poisoning is often difficult to diagnose. It is obvious in the victim of a house fire or in a comatose patienit found in a closed garage with the car running. It can, however, be much less obvious when an individual arrives at the emergency room with an acute onset of nausea and vomiting, diarrhea, headache, and weakness. Patients with vague symptoms or flulike symptoms should be questioned about the possibility of exposure to carbon monoxide, especially during the colder months.
Of those affected, researchers estimate that approximately 10%/o to 12% will go on to develop late neurological sequelae. These sequelae also cover a wide spectrum. Grinker first described them in 1926. 21 Since his reports, many others have described various features (Table 4l. These signs and symptoms do not correlate well with pathological findings. The lesions themselves are often associated with patchy white matter damage and have been referred to as myelinopathy or leukoencephalopathy. Damage to gray matter has been seen with striking frequency in the basal ganglia, especially the globus pallidus. These findings have been demonstrated by autopsy computed tomography, and magnetic resonance imaging.- '2 In the past, several theories have postulated the mechanism producing this brain damage; however, none have been able to explain the effects adequately. More recently, a theory by Thom2' has gained much support. He proposed that the initial mechlanism leading to demyelination is the production of hydroxyl-free radicals and lipid peroxidation in the brain folloving global ischemia (Figure 1). Carboin monoxide causes a hypoxic-ischemic injury of brain tissue, which results in transferrin releasing iron and electrons. T he free iron in the presence of ascorbic acid fuels hydroxyl radical production. Lipid peroxides result anid lead to demyelination. This theory is supported by both in vitro and in vivo evidence.

Treatment
Carbon monoxide poisoning treatment shoulld begin with assessment of the patient's airwNay; breathing, and circulation. Fire victims may very wvell develop an airway obstruction resulting from thermal or chemical injury. If the physician suspects the preseince of laryngeal edema, laryngoscopy should be done and the patient intubated if required. If the patient is not breathing, assistance is necessary. Cardiac imnoInitoring is essential, as some cardiac ischemia may be presenlt.
TFhe most widely used anid accepted treatment of carboni monoxide poisoninig is oxygen. Haldane" demonstrated that the bindiing of carbon monoxide to hemoglobiin wvas anitagonized if 'oxyge under a high 2130 (Caadian Familv PI/sicianl \V0.38: Sel/aembo 1992 L -. I partial pressure was given, and that the recovery process was speeded up. The experience of other researchers also demonstrated that hyperbaric oxygen was more useful in reducing morbidity and mortality in carbon monoxide poisoning than oxygen at ambient pressure.'),( Dr Kindwall of Milwaukee'2 compiled results from literature comparing carbon monoxide elimination half-time with the breathing medium (Table 2).
Hyperbaric oxygen dramatically shortens HbCO half-life, provides oxygen to tissues independent of that delivered by hemoglobin, and, as theorized by Thom,2-1 may have a protective mechanism on the brain, reducing the degree of neurological sequelae. Thom's experiments revealed that lipid peroxidation could be seen in rat brains following carbon monoxide exposure, but that the lipid peroxidation could be prevented by hyperbaric oxygen treatment at 3 ata. This effect was not seen with oxygen treatment at normal pressure.
Which cases of carbon monoxide poisoning should receive hyperbaric oxygen treatment continues to be controversial. I)uring a 26-month period from 1983 to 1985, Mathieu et a128 undertook a followup assessment of all patients admitted for carbon monoxide poisoning to the researchers' facility. Any patients with neurological abnormality or impairment in level of consciousness were treated with hyperbaric oxygenation, for 90 minutes at 2.5 ata. The remainder received normobaric oxygen for 12 hours.
Suggested factors affecting mortality were patient comatose when admitted and patient older than 60 years. 2" Factors related to immediate complications were impairment or loss of consciousness, age above 60 years, and HbCO level higher than 40%. Factors related to long-term sequelae were patient comatose on admission and patient older than 60 years. Therefore, patients should be referred to a hyperbaric uInit for treatment when they are unconscious, show neurological signs, and are older than 60 years or younger than 2 years. There is no contraindication to treating infants, children, or pregnant women.  Peripheral neuropathy patients or two critical care patients. The multiplace chamber is used for treating carbon monoxide poisoning so that an attendant can be present with the patient at all times. Approximately 25% of these patients require critical care treatment and life support. They are intubated and ventilated, either manually or by a Monoghan ventilator, chosen because it is pressure driven. Electrical devices are prohibited inside the chamber due to the risk of fire in the high oxygen environment. Those patients who do not require intubation or ventilatory assistance receive oxygen either by a tightly fitting mask or head tent. The head tent is a closed, flexible, clear plastic cylinder sealed around the neck by a latex seal. The tent is connected to a closed circuit rebreathing system. Oxygen is delivered at a rate of 10 L/min or as high as is required to keep the head tent inflated.
At times, ingenuity plays a role in treatment. A 4-week-old infant with carbon monoxide poisoning required treatment at TIoronto General Hospital. The infant did not require intubation, and a mask would not have fitted. The staff sealed the infant inside two head tents connected together. With this system connected to the closed circuit rebreathing system, treatment was carried out successfully ( Figure 2). 1

Conclusion
Carbon monoxide is a very dangerous gas; exposure can be fatal. The toxicity resulting from exposure can be explained by the reduction of the hemoglobin's oxygen-carrying capacity, a leftward shift of the oxyhemoglobin dissociation curve, the possibility of effects from myoglobin binding, and the effects of lipid peroxidation following global cellular ischemia. Those surviving carbon monoxide poisoning may go on to develop late neurological sequelae.
Diagnosing carbon monoxide poisoning is not easy, and misdiagnosis occurs because the symptoms with which it may present are non-specific. Physicians must obtain a detailed history and physical examination, and possess an index of suspicion. After diagnosis, proper treatment can be initiated with normobaric or hyperbaric oxygen. When diagnosis is not made, the patient may be returned to the contaminated environment with potentially long-term or fatal consequences. Even if diagnosis is made late and the patient appears recov-ered or to be recovering, he or she should be referred for consideration of hyperbaric oxygen treatment to prevent late neurological sequelae.