Hypoxemia without hypercapnia in COVID-19 patients; a proposed membrane-transfer mechanism
Hold your breath. After a little while, it will get uncomfortable. That may happen within a minute for many people; for those with exceptional fitness it could take up to three minutes. Regardless of those differences, everyone eventually reaches a point where the body reaches its limit and reflexively gasps for air.
You might assume that this discomfort is how your body tells you that you’re running out of oxygen but in fact, that’s not the case. The body detects respiratory distress based on the levels of absorbed Carbon Dioxide, which results in the acidification of blood, which is what your body interprets as a need to breathe, triggering discomfort and eventually, convulsive breathing.
Under normal circumstances, respiratory CO2 and O2 are inextricably linked. Your metabolism reacts sugars with O2 to produce energy, water and CO2 in a roughly 1 to 1 ratio with the O2 that was inhaled initially. Ordinarily, the air you breathe in is about 21% comprised of O2. What you breathe out is down to about 17% O2 and 4% CO2. The partial pressure of O2 decreases commensurately from 0.21 atmospheres to 0.17 atmospheres (if you are at sea-level), which is not that significant a decrease, particularly because hemoglobin, which transports the majority of that O2, becomes fully saturated at around 0.13 atmospheres.
With this mechanism in mind, it makes sense that the body uses CO2 as its alert system for respiratory distress. O2 can be absorbed in large amounts by hemoglobin, but only up to a certain point, after which it is saturated. Beyond that point, further oxygen can only be added by dissolving in the plasma. At normal atmospheric conditions, the amount of oxygen dissolved in plasma is roughly 1/20th of the amount that can bond to hemoglobin and so, in the range above the saturation point of hemoglobin, large changes in O2 partial pressure represent small changes in blood oxygen content. At low levels of oxygen, when hemoglobin sites are available, small changes in partial pressure represent significant swings in oxygen content.
CO2, on the other hand, is primarily transported by dissolution in plasma and therefore has a more linear relationship between its partial pressure and actual blood content, making it a better indicator of impaired respiration under normal circumstances.
The peculiar feature of this safety mechanism is that your body doesn’t know that it’s running out of oxygen unless it’s also accumulating carbon dioxide. Under normal circumstances this is perfectly fine; any impairment of breathing that threatens to deplete O2 will first elevate CO2. By the time the partial pressure of O2 in your lungs has dropped to the point that it cannot saturate your blood, the CO2 level will have risen past the point of severe discomfort.
This is why low O2 saturation is so closely associated with extreme respiratory discomfort. When a condition such as pneumonia or COPD inhibits a patient’s breathing to such an extent that their O2 saturation drops, their CO2 levels are sky-high and they show very clear signs of distress. There are some instances where this is not the case, however; it is well-established that over-exposure to high quantities of inert gases can cause suffocation before someone even knows that they are in trouble. So-called exit bags were a popular method of suicide for this reason. Freely breathing a gas mixture that contains very little oxygen can result in suffocation without any of the discomfort and trauma along the way.
COVID-19, troublingly, appears to have a similar effect. Patients present with severe hypoxemia but no distress whatsoever. Reports abound of patients sitting in ICUs, playing games on their phones and showing not the slightest sign of distress, while their O2 saturation reads catastrophically low — sometimes in the region of 80%, a level that is seldom observed except in critical circumstances. For reference, the most oxygen-intensive middle-distance running races seldom drop O2 saturation below 95%.
This phenomenon of hypoxemia in the absence of hypercapnia tells us that something very different to typical pneumonia is at play. The rate at which oxygen is transported into the volume of the lungs and the rate at which CO2 is transported out are both proportional to pulmonary flow-rate, so inhibited breathing cannot be the cause of one but not the other.
The cause must lie elsewhere. O2 consumption and CO2 production are inextricably linked by stoichiometry, and their transport through the body is also linked by virtue of the fact that they share the same transport media on both sides of the pulmonary interface, namely blood and air.
The one area where they are decoupled is in their respective rates of transfer across that interface. In general, the transfer of gas across a membrane is governed by the following equation derived from Fick’s Law:
The driving force for oxygen transfer over the lung/blood interface is the difference between the partial pressure of oxygen on either side of the blood-lung interface, which means that if mass transfer is obstructed then there must be a commensurate increase in the driving forces for transfer (partial pressure difference). This would suggest that an obstruction to gas transfer would result in higher CO2 partial pressure in the blood to drive its transfer into air, and a lower partial pressure of O2 in the blood to drive its transfer from air into blood. Because the O2 saturation of blood can change dramatically relative to changes in its partial pressure, O2 saturation would be affected more significantly than CO2 levels.
As evidenced in the above equation, there are several possible mechanisms by which an obstruction of gas transfer could occur.
The first possible mechanism is a decrease in available membrane area for gas transfer. This concept is in line with a theory espoused by numerous clinicians, which is that blood clots obstruct perfusion of blood into some percentage of lung tissue. However, the particular presentation of low O2 saturation without other clinical signs of hypercoagulopathy is very seldom observed in other clotting disorders, so this mechanism is either not complete or not a compelling explanation.
The second possible mechanism is an increase in the effective membrane thickness, possibly arising from widespread inflammation of the endothelium. This mechanism is also not entirely compelling; once again, low O2 saturation in isolation is not a common or likely presentation of systemic vasculitis.
The third possible mechanism is a decrease in the mass transfer coefficients for gas transfer. This could occur, for instance, if virus particles were to obstruct sites through which gas transfer takes place and is particularly compelling because this effect could potentially affect the respective mass transfer coefficients of O2 and CO2 differently, so it is the only mechanism that would allow for CO2 removal to be completely unaffected, thereby fully explaining the complete absence of hypercapnia.
This mechanism also offers a good explanation for the widespread effects of COVID-19 infection because it would also obstruct the delivery of oxygen from the blood into organs.
