Carbon monoxide (CO)

The usual sources of carbon monoxide causing poisoning are car exhaust (unleaded petrol contains about one tenth the amount of CO of leaded petrol), charcoal burners, faulty heaters, fires, and in industry.

An unusual source of carbon monoxide poisoning is the paint-stripping agent, methylene chloride. It can be absorbed by ingestion, inhalation or across the skin and then undergoes hepatic metabolism to form CO.

Carbon monoxide (CO) is the most common cause of fatal and disabling poisoning in most developed countries. Exposure is most commonly from suicide attempts using car exhaust, and accidental exposures from incomplete combustion in charcoal burners, faulty heaters, fires, and industrial accidents. Other toxic exposures should be considered; suicide attempts commonly involve drug overdose and fire victims often have coexisting cyanide poisoning.

CO is carried in the blood carried to haemoglobin as COHb. COHb cannot carry oxygen and the functional hypoxia could explain most symptoms and signs and long term consequences. However, additional mechanisms may include the inhibition of mitochondrial cytochrome oxidase producing cellular hypoxia and free radical production and reperfusion injury.

The target organs of CO poisoning are principally the heart and brain. Of particular concern is the potential for delayed neurotoxicity including Parkinsonism, memory and concentration impairment.

The management of CO poisoning is straightforward in terms of enhancing elimination. The administration of 100% oxygen will displace CO from haemoglobin and decrease the effective half-life from 4 hours to 40 minutes. The half-life is initially even shorter as hyperventilation and increased cardiac output also increase the rate of elimination. Thus by the time hyperbaric oxygen (HBO) might be applied to further decrease the effective half-life to around 20 minutes, most CO is already eliminated. Clinical trials do not provide good evidence that HBO reduces delayed neuropsychiatric sequelae, but all studies to date have had serious flaws.

The 'hypoxic' mechanisms are relatively straightforward. The initial clinical effects (at COHb concentrations less than 40–60%) occur despite compensated hypoxaemia. Oxygen delivery to the brain is maintained by a compensatory increase in cardiac output (as with anaemia).

As COHb concentrations rise, the heart rapidly becomes unable to compensate via higher cardiac output. Note there is also impaired oxygen delivery to the heart. Hypotension and reduced cardiac output exacerbate severe tissue hypoxia and death will rapidly occur unless there is intervention. Hypoxic brain damage is readily explainable in this group of patients.

Other mechanisms are less clear. CO also inhibits cytochrome oxidases in mitochondria. There are post-exposure inflammatory responses, but it is unclear if these differ substantially from those seen with other hypoxic/ischemic injuries.

Other mechanisms are less clear. CO also binds to cytochromes in the mitochondria. Falling concentrations of CO stimulate neutrophilic leucocytes to bind to endothelial cells, migrate through the vessel wall and to subsequently cause peroxidative tissue damage.

COHb concentration in blood is a function of the CO concentration in inspired air and the time of exposure. Perceptible clinical effects occur within 2 hours of exposure at concentrations as low as 0.01% (100 ppm). Decreased barometric pressure, increased activity, rate of ventilation, high metabolic rate, and anaemia (factors that increase breathing rate and pulmonary blood flow) increase uptake (& elimination) of CO.

Tobacco smokers have elevated concentrations of COHb (3 to 10%) and therefore will reach toxic concentrations slightly earlier in any exposure.

At equilibration, atmospheric carbon monoxide concentrations of 50, 100, and 200 ppm produce average carboxyhaemoglobin concentrations of 8, 16, and 30% respectively. This reflects the very high affinity of haemoglobin for CO (about 200 fold higher than for oxygen).

Less than 1% of the absorbed CO is metabolised endogenously to carbon dioxide.

Carbon monoxide is eliminated unchanged from the lungs in an exponential manner. The biological half-life of CO in a sedentary healthy adult is 4–5 hours. This half-life decreases to approximately 40–80 minutes with administration of 100% oxygen and to 23 minutes when hyperbaric (2 atmospheres) oxygen is used.

Hyperventilation also greatly shortens the half-life of CO, and indeed may be more effective than HBO in that regard. It is likely that most severely poisoned patients have a much shorter half-life than seen at low concentrations due to the compensatory hyperventilation and increased cardiac output. Thus by the time most people present to hospital a large proportion of CO has already been eliminated.

Typical clinical symptoms and signs relative to COHb (Normal = 0.5%):

• <10% nil (commonly found in smokers)
• 10 - 20% nil or vague nondescript symptoms
• 30 - 40% headache, tachycardia, confusion, weakness, nausea, vomiting, collapse
• 50 - 60% coma, convulsions, Cheyne-Stokes breathing, arrhythmias, ECG changes
• 70 - 80% circulatory and ventilatory failure, cardiac arrest, death

The skin is classically cherry pink although severely ill patients are often pale or cyanosed.
The %COHb changes rapidly, variably, but generally to a greater extent in the severely poisoned. Thus on arrival at emergency departments there is very little relationship between prior symptoms and signs and the measured COHb%.

Pre-existing cerebral or cardiovascular disease, anaemia, and volume depletion or cardiac failure increases toxicity (for a given COHb). These people all have a reduced ability to compensate by increasing cardiac output or redistributing blood supply to vital organs.

COHb concentrations (plus or minus back calculation based on estimated half-life since removal) are a very poor indicator of peak COHb or exposure. The correlation with acute and long-term clinical effects is also generally poor. This is partly due to these inaccuracies in estimating half-life and also the substantial variability in ability to compensate. Thus a high %COHb can confirm (or possibly exclude) the diagnosis but should not be used as a guide for treatment or long term prognosis.

Patients should have a baseline ECG, repeated 6 hours later. ECG monitoring may be indicated for those with evidence of myocardial ischemic injury.

A high lactate may occur with hypoxia but an unexpectedly high lactate might indicate cyanide poisoning. Troponin should be measured in anyone with loss of consciousness, chest pain, ECG abnormalities or any other indicators of hypoxic injury.

In suicide attempts, the diagnosis of CO poisoning is generally apparent from the circumstances when the person is found. Inquiries should be made as to the possibility of other agents taken in deliberate self-poisonings or, in unconscious patients, the ECG, paracetamol concentration, and electrolytes should be reviewed with this possibility in mind.

A large proportion of victims of smoke inhalation also have cyanide poisoning. This rarely leads to a change in management (due to problems with administering the cyanide antidotes in this setting) but should be suspected when CNS effects are out of proportion with COHb concentrations and if there is a marked lactic acidosis.

The prognosis is worse if there is evidence of tissue hypoxia, risk factors for hypoxia, or reduced ability to compensate. Common factors linked to long term morbidity and mortality are:

• coma at the time of discovery
• neurological symptoms/signs upon wakening
• raised troponin and/or cardiac stunning
• older age
• underlying cerebrovascular or cardiovascular disease

These factors are linked to both long term neuropsychological sequelae and medium term mortality.
The %COHb at the time of presentation is NOT a good prognostic indicator.

If there is impaired consciousness, ensure the airway is maintained with intubation if necessary.
Ensure the patient is quiet and resting. Muscle activity will increase oxygen demand and/or lactic acidosis.
Monitor for complications.

Note: Metabolic acidosis should not be treated directly unless the acidosis itself contributes to toxicity (pH < 7.0). It should rapidly respond to improved oxygenation and ventilation. Acidosis shifts the haemoglobin-oxygen dissociation curve to the right thereby increasing the tissue availability of the oxygen carried as oxyhaemoglobin. Further, sodium bicarbonate is likely to lead to rebound alkalaemia as the lactic acidosis resolves.

100% oxygen should be administered with mechanically assisted ventilation if necessary. In patients able to tolerate it, CPAP (continuous positive airway pressure) by mask may allow 100% oxygen delivery without intubation. Four to six hours of 100% normobaric oxygen will remove over 90% of the carbon monoxide. Oxygen toxicity is unlikely with less than 24 hours treatment but the risk increases with increasing exposure.

When immediately available, HBO should be considered for patients with serious carbon monoxide poisoning. Oxygen at 2–3 atmospheres will reduce the half-life of COHb to about 20 minutes and causes almost instantaneous reversal of tissue hypoxia due to oxygenation of tissue from oxygen dissolved in the plasma. However, in practice delays mean most CO has already been eliminated before patients could feasibly enter the chamber.

There is controversy on the benefits, risks, and indications for HBO. Most trials have shown no benefit and the major one that claims to relies heavily on post hoc data manipulations to make this case. Transferring patients between hospitals, particularly over long distances, is clearly not justified based on current evidence.

Complications of HBO therapy include

• decompression sickness
• rupture of tympanic membranes
• damaged sinuses
• oxygen toxicity
• problems due to lack of monitoring

Contraindications include

• chest trauma
• other major comorbidity (e.g. serious drug overdose, severe burns)
• uncooperative patient

Animal models suggest possible benefits from agents that might protect against oxidative damage during hypoxic/reperfusion injury. Numerous experimental (i.e. never used in humans) agents have also been suggested. Their use cannot be recommended outside of clinical trials.

The most promising agent to date is erythropoetin. Seven days of treatment reduced long term sequelae and neurological biomarkers in a small RCT. However, the high cost of this treatment is likely to limit the use of this treatment without further supporting studies.

Pregnancy

Due to low oxygen pressures, the high affinity of foetal haemoglobin for CO, and the much longer half-life of CO in the foetal circulation, the foetus is particularly susceptible to CO poisoning. There is also likely to be added benefit from HBO in this setting (to shorten the half-life of CO and to deliver oxygen independent of haemoglobin). HBO appears to be safe in pregnancy although the outcome of significant CO poisoning in the mother is often foetal death or neurological damage.

Delayed neurological sequelae occur with relapses, usually within a week of the exposure and after initial recovery. Long-term follow up is also essential as more subtle defects can develop or become apparent over a few weeks to months. The most common problems encountered are depressed mood (even in those accidentally exposed) and difficulty with higher intellectual functions (especially short-term memory). More severe problems include Parkinsonism and speech problems. Severe damage is uncommon but milder damage is very common in those with an initial coma. Neuropsychiatric testing may detect subtle defects not apparent on crude mini-mental state testing. Anyone severely poisoned with CO should be advised about risks and carefully assessed.

Long term cardiac sequelae may also occur. Early cardiac injury has been associated with a substantially lower life expectancy although it is unclear if this reflects the baseline risk factors or is caused by the hypoxic injury.

Annane D, Chadda K, Gajdos P, et al. Hyperbaric oxygen therapy for acute domestic carbon monoxide poisoning: two randomized controlled trials. Intensive Care Med 2011; 37(3): 486-92 PMID21125215
Buckley NA, Juurlink DN, Isbister G, Bennett MH, Lavonas EJ. Hyperbaric oxygen for carbon monoxide poisoning. Cochrane Database Syst Rev 2011; (4): CD002041 PMID21491385
Chiew A, Buckley NA. Carbon monoxide poisoning in the 21st century. Crit Care 2014; 18(2): 221 Fulltext
Elkharrat D, Raphael JC, Korach JM, et al. Acute carbon monoxide intoxication and hyperbaric oxygen in pregnancy. Intensive Care Med 1991; 17(5): 289-92 PMID1939875
Fisher J, Rucker J, Sommer LZ, et al. Isocapnic hyperpnea accelerates carbon monoxide elimination. Am J Resp Crit Care Med 1999; 159(4 Pt 1): 1289-92 PMID10194179
Hampson NB, Dunn SL; UHMCS/CDC CO Poisoning Surveillance Group. Symptoms of carbon monoxide poisoning do not correlate with the initial carboxyhemoglobin level. Undersea Hyperb Med 2012; 39(2): 657-65 PMID22530449
Henry CR, Satran D, Lindgren B, et al. Myocardial injury and long-term mortality following moderate to severe carbon monoxide poisoning. JAMA 2006; 295(4): 398-402 PMID16434630
Koren G, Sharav T, Pastuszak A, et al. A multicenter, prospective study of fetal outcome following accidental carbon monoxide poisoning in pregnancy. Reprod Toxicol 1991; 5(5): 397-403 PMID1806148
Pang L, Bian M, Zang X, et al. Neuroprotective Effects of Erythropoietin in Patients with Carbon Monoxide Poisoning. J Biochem Mol Toxicol 2013; 27(5): 266-71 PMID23554277
Park E, Ahn J, Min Y-G, et al. The usefulness of the serum s100b protein for predicting delayed neurological sequelae in acute carbon monoxide poisoning. Clin Toxicol 2012; 50(3): 183-8 PMID22372785
Pepe, G, Castelli M, Nazerian P, et al. Delayed neuropsychological sequelae after carbon monoxide poisoning: predictive risk factors in the Emergency Department. A retrospective study. Scand J Trauma Resusc Emerg Med 2011; 19: 16 PMID21414211 Fulltext
Thom SR, Fisher D, Xu YA, Notarfrancesco K, Ischiropoulos H. Adaptive responses and apoptosis in endothelial cells exposed to carbon monoxide. Proc Natl Acad Sci USA 2000; 97(3): 1305-10 PMID10655526