news

Oxygen therapy is one of the most commonly used methods in modern medicine, but there are still misconceptions about the indications for oxygen therapy, and improper use of oxygen can cause serious toxic reactions

Clinical evaluation of tissue hypoxia

The clinical manifestations of tissue hypoxia are varied and non-specific, with the most prominent symptoms including dyspnea, shortness of breath, tachycardia, respiratory distress, rapid changes in mental state, and arrhythmia. To determine the presence of tissue (visceral) hypoxia, serum lactate (elevated during ischemia and reduced cardiac output) and SvO2 (decreased during reduced cardiac output, anemia, arterial hypoxemia, and high metabolic rate) are helpful for clinical evaluation. However, lactate can be elevated in non hypoxic conditions, so a diagnosis cannot be made solely based on lactate elevation, as lactate can also be elevated in conditions of increased glycolysis, such as rapid growth of malignant tumors, early sepsis, metabolic disorders, and administration of catecholamines. Other laboratory values that indicate specific organ dysfunction are also important, such as elevated creatinine, troponin, or liver enzymes.

Clinical evaluation of arterial oxygenation status

Cyanosis. Cyanosis is usually a symptom that occurs in the late stage of hypoxia, and is often unreliable in diagnosing hypoxemia and hypoxia because it may not occur in anemia and poor blood flow perfusion, and it is difficult for people with darker skin to detect cyanosis.

Pulse oximetry monitoring. Non invasive pulse oximetry monitoring has been widely used for monitoring all diseases, and its estimated SaO2 is called SpO2. The principle of pulse oximetry monitoring is Bill’s law, which states that the concentration of an unknown substance in a solution can be determined by its absorption of light. When light passes through any tissue, most of it is absorbed by the tissue’s elements and blood. However, with each heartbeat, arterial blood undergoes pulsatile flow, allowing the pulse oximetry monitor to detect changes in light absorption at two wavelengths: 660 nanometers (red) and 940 nanometers (infrared). The absorption rates of reduced hemoglobin and oxygenated hemoglobin are different at these two wavelengths. After subtracting the absorption of non pulsatile tissues, the concentration of oxygenated hemoglobin relative to total hemoglobin can be calculated.

There are some limitations to monitoring pulse oximetry. Any substance in the blood that absorbs these wavelengths can interfere with measurement accuracy, including acquired hemoglobinopathies – carboxyhemoglobin and methemoglobinemia, methylene blue, and certain genetic hemoglobin variants. The absorption of carboxyhemoglobin at a wavelength of 660 nanometers is similar to that of oxygenated hemoglobin; Very little absorption at a wavelength of 940 nanometers. Therefore, regardless of the relative concentration of carbon monoxide saturated hemoglobin and oxygen saturated hemoglobin, SpO2 will remain constant (90%~95%). In methemoglobinemia, when heme iron is oxidized to the ferrous state, methemoglobin equalizes the absorption coefficients of two wavelengths. This results in SpO2 only varying within the range of 83% to 87% within a relatively wide concentration range of methemoglobin. In this case, four wavelengths of light are required for arterial blood oxygen measurement to distinguish between the four forms of hemoglobin.

Pulse oximetry monitoring relies on sufficient pulsatile blood flow; Therefore, pulse oximetry monitoring cannot be used in shock hypoperfusion or when using non pulsatile ventricular assist devices (where the cardiac output only accounts for a small portion of cardiac output). In severe tricuspid regurgitation, the concentration of deoxyhemoglobin in venous blood is high, and the pulsation of venous blood may lead to low blood oxygen saturation readings. In severe arterial hypoxemia (SaO2<75%), accuracy may also decrease as this technique has never been validated within this range. Finally, more and more people are realizing that pulse oximetry monitoring may overestimate arterial hemoglobin saturation by up to 5-10 percentage points, depending on the specific device used by darker skinned individuals.

PaO2/FIO2. The PaO2/FIO2 ratio (commonly referred to as the P/F ratio, ranging from 400 to 500 mm Hg) reflects the degree of abnormal oxygen exchange in the lungs, and is most useful in this context as mechanical ventilation can accurately set FIO2. A P/F ratio less than 300 mm Hg indicates clinically significant gas exchange abnormalities, while a P/F ratio less than 200 mm Hg indicates severe hypoxemia. The factors that affect the P/F ratio include ventilation settings, positive end expiratory pressure, and FIO2. The impact of changes in FIO2 on the P/F ratio varies depending on the nature of lung injury, shunt fraction, and the range of FIO2 changes. In the absence of PaO2, SpO2/FIO2 can serve as a reasonable alternative indicator.

Alveolar arterial oxygen partial pressure (A-a PO2) difference. A-a PO2 differential measurement is the difference between the calculated alveolar oxygen partial pressure and the measured arterial oxygen partial pressure, used to measure the efficiency of gas exchange.

The “normal” A-a PO2 difference for breathing ambient air at sea level varies with age, ranging from 10 to 25 mm Hg (2.5+0.21 x age [years]). The second influencing factor is FIO2 or PAO2. If either of these two factors increases, the difference in A-a PO2 will increase. This is because gas exchange in alveolar capillaries occurs in the flatter part (slope) of the hemoglobin oxygen dissociation curve. Under the same degree of venous mixing, the difference in PO2 between mixed venous blood and arterial blood will increase. On the contrary, if the alveolar PO2 is low due to inadequate ventilation or high altitude, the A-a difference will be lower than normal, which may lead to underestimation or inaccurate diagnosis of pulmonary dysfunction.

Oxygenation index. Oxygenation index (OI) can be used in mechanically ventilated patients to assess the required ventilation support intensity for maintaining oxygenation. It includes mean airway pressure (MAP, in cm H2O), FIO2, and PaO2 (in mm Hg) or SpO2, and if it exceeds 40, it can be used as a standard for extracorporeal membrane oxygenation therapy. Normal value less than 4 cm H2O/mm Hg; Due to the uniform value of cm H2O/mm Hg (1.36), units are usually not included when reporting this ratio.

Indications for acute oxygen therapy
When patients experience difficulty breathing, oxygen supplementation is usually required before the diagnosis of hypoxemia. When the arterial partial pressure of oxygen (PaO2) is below 60 mm Hg, the most clear indication for oxygen uptake is arterial hypoxemia, which typically corresponds to arterial oxygen saturation (SaO2) or peripheral oxygen saturation (SpO2) of 89% to 90%. When PaO2 drops below 60 mm Hg, blood oxygen saturation may sharply decrease, leading to a significant decrease in arterial oxygen content and potentially causing tissue hypoxia.

In addition to arterial hypoxemia, oxygen supplementation may be necessary in rare cases. Severe anemia, trauma, and surgical critical patients can reduce tissue hypoxia by increasing arterial oxygen levels. For patients with carbon monoxide (CO) poisoning, supplementing oxygen can increase the dissolved oxygen content in the blood, replace CO bound to hemoglobin, and increase the proportion of oxygenated hemoglobin. After inhaling pure oxygen, the half-life of carboxyhemoglobin is 70-80 minutes, while the half-life when breathing ambient air is 320 minutes. Under hyperbaric oxygen conditions, the half-life of carboxyhemoglobin is shortened to less than 10 minutes after inhaling pure oxygen. Hyperbaric oxygen is generally used in situations with high levels of carboxyhemoglobin (>25%), cardiac ischemia, or sensory abnormalities.

Despite the lack of supporting data or inaccurate data, other diseases may also benefit from supplementing oxygen. Oxygen therapy is commonly used for cluster headache, sickle cell pain crisis, relief of respiratory distress without hypoxemia, pneumothorax, and mediastinal emphysema (promoting chest air absorption). There is evidence to suggest that intraoperative high oxygen can reduce the incidence of surgical site infections. However, supplementing oxygen does not seem to effectively reduce postoperative nausea/vomiting.

With the improvement of outpatient oxygen supply capacity, the use of long-term oxygen therapy (LTOT) is also increasing. The standards for implementing long-term oxygen therapy are already very clear. Long term oxygen therapy is commonly used for chronic obstructive pulmonary disease (COPD).
Two studies on patients with hypoxemic COPD provide supportive data for LTOT. The first study was the Nocturnal Oxygen Therapy Trial (NOTT) conducted in 1980, in which patients were randomly assigned to either nighttime (at least 12 hours) or continuous oxygen therapy. At 12 and 24 months, patients who only receive nighttime oxygen therapy have a higher mortality rate. The second experiment was the Medical Research Council Family Trial conducted in 1981, in which patients were randomly divided into two groups: those who did not receive oxygen or those who received oxygen for at least 15 hours per day. Similar to the NOTT test, the mortality rate in the anaerobic group was significantly higher. The subjects of both trials were non-smoking patients who received maximum treatment and had stable conditions, with a PaO2 below 55 mm Hg, or patients with polycythemia or pulmonary heart disease with a PaO2 below 60 mm Hg.

These two experiments indicate that supplementing oxygen for more than 15 hours a day is better than completely not getting oxygen, and continuous oxygen therapy is better than only treating at night. The inclusion criteria for these trials are the basis for current medical insurance companies and ATS to develop LTOT guidelines. It is reasonable to infer that LTOT is also accepted for other hypoxic cardiovascular diseases, but there is currently a lack of relevant experimental evidence. A recent multicenter trial found no difference in the impact of oxygen therapy on mortality or quality of life for COPD patients with hypoxemia that did not meet the resting criteria or was only caused by exercise.

Doctors sometimes prescribe nighttime oxygen supplementation to patients who experience severe decrease in blood oxygen saturation during sleep. There is currently no clear evidence to support the use of this approach in patients with obstructive sleep apnea. For patients with obstructive sleep apnea or obesity hypopnea syndrome leading to poor nighttime breathing, non-invasive positive pressure ventilation rather than oxygen supplementation is the main treatment method

Another issue to consider is whether oxygen supplementation is needed during air travel. Most commercial aircraft typically increase the cabin pressure to an altitude equivalent to 8000 feet, with an inhaled oxygen tension of approximately 108 mm Hg. For patients with lung diseases, a decrease in inhaled oxygen tension (PiO2) can cause hypoxemia. Before traveling, patients should undergo a comprehensive medical evaluation, including arterial blood gas testing. If the patient’s PaO2 on the ground is ≥ 70 mm Hg (SpO2>95%), then their PaO2 during flight is likely to exceed 50 mm Hg, which is generally considered sufficient to cope with minimal physical activity. For patients with low SpO2 or PaO2, a 6-minute walk test or hypoxia simulation test can be considered, typically breathing 15% oxygen. If hypoxemia occurs during air travel, oxygen can be administered through a nasal cannula to increase oxygen intake.

Biochemical basis of oxygen poisoning

Oxygen toxicity is caused by the production of reactive oxygen species (ROS). ROS is an oxygen derived free radical with an unpaired orbital electron that can react with proteins, lipids, and nucleic acids, altering their structure and causing cellular damage. During normal mitochondrial metabolism, a small amount of ROS is produced as a signaling molecule. Immune cells also use ROS to kill pathogens. ROS includes superoxide, hydrogen peroxide (H2O2), and hydroxyl radicals. Excessive ROS will invariably exceed cellular defense functions, leading to death or inducing cell damage.

To limit the damage mediated by ROS generation, the antioxidant protection mechanism of cells can neutralize free radicals. Superoxide dismutase converts superoxide into H2O2, which is then converted into H2O and O2 by catalase and glutathione peroxidase. Glutathione is an important molecule that limits ROS damage. Other antioxidant molecules include alpha tocopherol (vitamin E), ascorbic acid (vitamin C), phospholipids, and cysteine. The human lung tissue contains high concentrations of extracellular antioxidants and superoxide dismutase isoenzymes, making it less toxic when exposed to higher concentrations of oxygen compared to other tissues.

Hyperoxia induced ROS mediated lung injury can be divided into two stages. Firstly, there is the exudative phase, characterized by the death of alveolar type 1 epithelial cells and endothelial cells, interstitial edema, and the filling of exudative neutrophils in the alveoli. Subsequently, there is a proliferation phase, during which endothelial cells and type 2 epithelial cells proliferate and cover the previously exposed basement membrane. The characteristics of the oxygen injury recovery period are fibroblast proliferation and interstitial fibrosis, but the capillary endothelium and alveolar epithelium still maintain a roughly normal appearance.
Clinical manifestations of pulmonary oxygen toxicity

The exposure level at which toxicity occurs is not yet clear. When FIO2 is less than 0.5, clinical toxicity generally does not occur. Early human studies have found that exposure to nearly 100% oxygen can cause sensory abnormalities, nausea, and bronchitis, as well as reduce lung capacity, lung diffusion capacity, lung compliance, PaO2, and pH. Other issues related to oxygen toxicity include absorptive atelectasis, oxygen induced hypercapnia, acute respiratory distress syndrome (ARDS), and neonatal bronchopulmonary dysplasia (BPD).
Absorbent atelectasis. Nitrogen is an inert gas that diffuses very slowly into the bloodstream compared to oxygen, thus playing a role in maintaining alveolar expansion. When using 100% oxygen, due to the oxygen absorption rate exceeding the delivery rate of fresh gas, nitrogen deficiency can lead to alveolar collapse in areas with lower alveolar ventilation perfusion ratio (V/Q). Especially during surgery, anesthesia and paralysis can lead to a decrease in residual lung function, promoting collapse of small airways and alveoli, resulting in rapid onset of atelectasis.

Oxygen induced hypercapnia. Severe COPD patients are prone to severe hypercapnia when exposed to high concentrations of oxygen during the worsening of their condition. The mechanism of this hypercapnia is that the ability of hypoxemia to drive respiration is inhibited. However, in any patient, there are two other mechanisms at play to varying degrees.
The hypoxemia in COPD patients is the result of low alveolar partial pressure of oxygen (PAO2) in the low V/Q region. In order to minimize the impact of these low V/Q regions on hypoxemia, two reactions of the pulmonary circulation – hypoxic pulmonary vasoconstriction (HPV) and hypercapnic pulmonary vasoconstriction – will transfer blood flow to well ventilated areas. When oxygen supplementation increases PAO2, HPV significantly decreases, increasing perfusion in these areas, resulting in areas with lower V/Q ratios. These lung tissues are now rich in oxygen but have weaker ability to eliminate CO2. The increased perfusion of these lung tissues comes at the cost of sacrificing areas with better ventilation, which cannot release large amounts of CO2 as before, leading to hypercapnia.

Another reason is the weakened Haldane effect, which means that compared to oxygenated blood, deoxygenated blood can carry more CO2. When hemoglobin is deoxygenated, it binds more protons (H+) and CO2 in the form of amino esters. As the concentration of deoxyhemoglobin decreases during oxygen therapy, the buffering capacity of CO2 and H+also decreases, thereby weakening the ability of venous blood to transport CO2 and leading to an increase in PaCO2.

When supplying oxygen to patients with chronic CO2 retention or high-risk patients, especially in the case of extreme hypoxemia, it is extremely important to fine adjust FIO2 to maintain SpO2 in the range of 88%~90%. Multiple case reports indicate that failure to regulate O2 can lead to adverse consequences; A randomized study conducted on patients with acute exacerbation of CODP on their way to the hospital has unquestionably proven this. Compared with patients without oxygen restriction, patients randomly assigned to supplement oxygen to maintain SpO2 within the range of 88% to 92% had significantly lower mortality rates (7% vs. 2%).

ARDS and BPD. People have long discovered that oxygen toxicity is associated with the pathophysiology of ARDS. In non-human mammals, exposure to 100% oxygen can lead to diffuse alveolar damage and ultimately death. However, the exact evidence of oxygen toxicity in patients with severe lung diseases is difficult to distinguish from the damage caused by underlying diseases. In addition, many inflammatory diseases can induce upregulation of antioxidant defense function. Therefore, most studies have failed to demonstrate a correlation between excessive oxygen exposure and acute lung injury or ARDS.

Pulmonary hyaline membrane disease is a disease caused by a lack of surface active substances, characterized by alveolar collapse and inflammation. Premature newborns with hyaline membrane disease typically require inhalation of high concentrations of oxygen. Oxygen toxicity is considered a major factor in the pathogenesis of BPD, even occurring in newborns who do not require mechanical ventilation. Newborns are particularly susceptible to high oxygen damage because their cellular antioxidant defense functions have not yet fully developed and matured; Retinopathy of prematurity is a disease associated with repeated hypoxia/hyperoxia stress, and this effect has been confirmed in retinopathy of prematurity.
The synergistic effect of pulmonary oxygen toxicity

There are several drugs that can enhance oxygen toxicity. Oxygen increases the ROS produced by bleomycin and inactivates bleomycin hydrolase. In hamsters, high oxygen partial pressure can exacerbate bleomycin induced lung injury, and case reports have also described ARDS in patients who have received bleomycin treatment and were exposed to high FIO2 during the perioperative period. However, a prospective trial failed to demonstrate an association between high concentration oxygen exposure, previous exposure to bleomycin, and severe postoperative pulmonary dysfunction. Paraquat is a commercial herbicide that is another enhancer of oxygen toxicity. Therefore, when dealing with patients with paraquat poisoning and exposure to bleomycin, FIO2 should be minimized as much as possible. Other drugs that may exacerbate oxygen toxicity include disulfiram and nitrofurantoin. Protein and nutrient deficiencies can lead to high oxygen damage, which may be due to a lack of thiol containing amino acids that are crucial for glutathione synthesis, as well as a lack of antioxidant vitamins A and E.
Oxygen toxicity in other organ systems

Hyperoxia can cause toxic reactions to organs outside the lungs. A large multicenter retrospective cohort study showed an association between increased mortality and high oxygen levels after successful cardiopulmonary resuscitation (CPR). The study found that patients with PaO2 greater than 300 mm Hg after CPR had an in-hospital mortality risk ratio of 1.8 (95% CI, 1.8-2.2) compared to patients with normal blood oxygen or hypoxemia. The reason for the increased mortality rate is the deterioration of central nervous system function after cardiac arrest caused by ROS mediated high oxygen reperfusion injury. A recent study also described an increased mortality rate in patients with hypoxemia after intubation in the emergency department, which is closely related to the degree of elevated PaO2.

For patients with brain injury and stroke, providing oxygen to those without hypoxemia seems to have no benefit. A study conducted by a trauma center found that compared to patients with normal blood oxygen levels, patients with traumatic brain injury who received high oxygen (PaO2>200 mm Hg) treatment had a higher mortality rate and lower Glasgow Coma Score upon discharge. Another study on patients receiving hyperbaric oxygen therapy showed poor neurological prognosis. In a large multicenter trial, supplementing oxygen to acute stroke patients without hypoxemia (saturation greater than 96%) had no benefit in mortality or functional prognosis.

In acute myocardial infarction (AMI), oxygen supplementation is a commonly used therapy, but the value of oxygen therapy for such patients is still controversial. Oxygen is necessary in the treatment of acute myocardial infarction patients with concomitant hypoxemia, as it can save lives. However, the benefits of traditional oxygen supplementation in the absence of hypoxemia are not yet clear. In the late 1970s, a double-blind randomized trial enrolled 157 patients with uncomplicated acute myocardial infarction and compared oxygen therapy (6 L/min) with no oxygen therapy. It was found that patients receiving oxygen therapy had a higher incidence of sinus tachycardia and a greater increase in myocardial enzymes, but there was no difference in mortality rate.

In ST segment elevation acute myocardial infarction patients without hypoxemia, nasal cannula oxygen therapy at 8 L/min is not beneficial compared to inhaling ambient air. In another study on oxygen inhalation at 6 L/min and inhalation of ambient air, there was no difference in 1-year mortality and readmission rates among patients with acute myocardial infarction. Controlling blood oxygen saturation between 98% to 100% and 90% to 94% has no benefit in patients with cardiac arrest outside the hospital. The potential harmful effects of high oxygen on acute myocardial infarction include coronary artery constriction, disrupted microcirculation blood flow distribution, increased functional oxygen shunt, decreased oxygen consumption, and increased ROS damage in the successfully reperfusion area.

Finally, clinical trials and meta-analyses investigated the appropriate SpO2 target values for critically ill hospitalized patients. A single center, open label randomized trial comparing conservative oxygen therapy (SpO2 target 94%~98%) with traditional therapy (SpO2 value 97%~100%) was conducted on 434 patients in the intensive care unit. The mortality rate in the intensive care unit of patients randomly assigned to receive conservative oxygen therapy has improved, with lower rates of shock, liver failure, and bacteremia. A subsequent meta-analysis included 25 clinical trials that recruited over 16000 hospitalized patients with varying diagnoses, including stroke, trauma, sepsis, myocardial infarction, and emergency surgery. The results of this meta-analysis showed that patients receiving conservative oxygen therapy strategies had an increased in-hospital mortality rate (relative risk, 1.21; 95% CI, 1.03-1.43).

However, two subsequent large-scale trials failed to demonstrate any impact of conservative oxygen therapy strategies on the number of days without ventilators in patients with lung disease or the 28 day survival rate in ARDS patients. Recently, a study of 2541 patients receiving mechanical ventilation found that targeted oxygen supplementation within three different SpO2 ranges (88%~92%, 92%~96%, 96%~100%) did not affect outcomes such as survival days, mortality, cardiac arrest, arrhythmia, myocardial infarction, stroke, or pneumothorax without mechanical ventilation within 28 days. Based on these data, the British Thoracic Society guidelines recommend a target SpO2 range of 94% to 98% for most adult hospitalized patients. This is reasonable because SpO2 within this range (considering the ± 2%~3% error of pulse oximeters) corresponds to a PaO2 range of 65-100 mm Hg, which is safe and sufficient for blood oxygen levels. For patients at risk of hypercapnic respiratory failure, 88% to 92% is a safer target to avoid hypercapnia caused by O2.