# Gas Exchange

- DOI:
- 10.1093/med/9780190670085.003.0002

I use the term *gas exchange* to encompass several processes that allow the respiratory system to maintain a normal arterial partial pressure of O_{2} and CO_{2} (PaO_{2}, PaCO_{2}). Even though I separated ventilation and gas exchange in the introduction to Chapter 1, ventilation is actually an essential part of gas exchange because it delivers O_{2}, eliminates CO_{2}, and determines ventilation–perfusion ratios.

The components of normal gas exchange are:

• Delivery of oxygen

• Excretion of carbon dioxide

• Matching of ventilation and perfusion

• Gas diffusion

Before discussing each of these components, it’s important to review the concept of gas partial pressure.

Partial Pressure

The *total pressure* (P_{T}) produced by a *gas mixture* is equal to the sum of the pressures generated by each of its components.

The pressure contributed by each gas is referred to as its *partial pressure*, which is equal to total pressure multiplied by the *fractional concentration* (F) of each gas in the mixture. For example, at sea level, the total pressure of the atmosphere (barometric pressure; P_{B}) is 760 mmHg. Since the fractional concentration of O_{2} (FO_{2}) in dry air is 0.21 (i.e., 21% of the gas molecules are O_{2}), its partial pressure (PO_{2}) is calculated as:

Similarly, the partial pressure of nitrogen (PN_{2}) is 760 mmHg x 0.79, or 600 mmHg.

Partial pressure is important because O_{2} and CO_{2} molecules diffuse between alveolar gas and pulmonary capillary blood and between systemic capillary blood and the tissues along their partial pressure gradients, and diffusion continues until the partial pressures are equal. Gas diffusion is discussed in much more detail later in this chapter.

Delivery of Oxygen

Ventilation is responsible for delivering O_{2} molecules to the alveoli, and this is the first step in transferring O_{2} from outside the body to the arterial blood. As just mentioned, the PO_{2} of dry air at sea level is 160 mmHg. Once gas enters the upper and lower airways, it is heated and humidified. This introduces another gas into the mixture—water. Since the partial pressure of water (P_{H2O}) at body temperature is 47 mmHg, the *inspired* partial pressure of oxygen (P_{I}O_{2}) becomes the product of FO_{2} and the difference between barometric and water pressure.

Here, F_{I}O_{2} is the fractional concentration of *inspired* oxygen.

When gas reaches the alveoli, the PO_{2} falls even further as O_{2} and CO_{2} molecules are exchanged across the alveolar–capillary interface. Although there’s no way to measure the mean PO_{2} of the gas in all the alveoli of the lungs $({\text{P}\overline{\text{A}}\text{O}}_{\text{2}})$, we can estimate it by using the *alveolar gas equation*.

This equation has two components. The first part, within the brackets, is identical to the right side of Equation 2.3 and equals the PO_{2} of the gas within the conducting airways (i.e., the airways from the mouth to the terminal bronchioles). The second part equals the drop in P_{I}O_{2} caused by the diffusion of O_{2} into the capillary blood. In this portion of the equation, ${\text{P}\overline{\text{A}}\text{CO}}_{\text{2}}$ is the mean alveolar PCO_{2}, which is assumed to equal arterial PCO_{2}, and R is the ratio of CO_{2} molecules entering to O_{2} molecules leaving the alveolar gas. Let’s assume for a moment that R is 1.0. In that case, ${\text{P}\overline{\text{A}}\text{CO}}_{\text{2}}$ would increase and ${\text{P}\overline{\text{A}}\text{O}}_{\text{2}}$ would fall by the same amount, and we could simply subtract PaCO_{2} from P_{I}O_{2} to get the mean alveolar PO_{2}. Normally, however, R is less than 1.0, and for the alveolar gas equation, it is assumed to equal 0.8. So, the decrease in PO_{2} between the conducting airways and the alveoli will be PaCO_{2} divided by 0.8 (or multiplied by 1.25). If PaCO_{2} is 40 mmHg, we get:

As you can see, ${\text{P}\overline{\text{A}}\text{O}}_{\text{2}}$ varies directly with barometric pressure and F_{I}O_{2} and inversely with PaCO_{2}. The progressive fall in PO_{2} between the outside of the body and the alveoli is part of the *oxygen cascade*, which continues as O_{2} enters the arterial blood and the tissues (Figure 2.1).

You can think of the calculated ${\text{P}\overline{\text{A}}\text{O}}_{\text{2}}$ as being the highest possible PaO_{2} for a given P_{B}, F_{I}O_{2}, and PaCO_{2}. That’s because the alveolar gas equation tells us what the ${\text{P}\overline{\text{A}}\text{O}}_{\text{2}}$ *and* the PaO_{2} *would be* if the lungs were “perfect”—that is, if every alveolus had the same ratio of ventilation to perfusion. In fact, as you’ll soon learn, there’s no such thing as perfect lungs, and that’s why there’s always a difference between the calculated ${\text{P}\overline{\text{A}}\text{O}}_{\text{2}}$ and the measured PaO_{2}.

Carbon Dioxide Excretion

Carbon dioxide is a normal byproduct of cellular metabolism and continuously diffuses from the tissues into the systemic capillary blood, and from the pulmonary capillary blood into the alveolar gas. The mean alveolar and arterial PCO_{2} are determined by the balance between the rates at which CO_{2} is produced by the tissues (${\dot{\text{V}}}_{\text{P}}{\text{CO}}_{\text{2}}$) and excreted by ventilation (${\dot{\text{V}}}_{\text{E}}{\text{CO}}_{\text{2}}$).

If, for example, CO_{2} is produced faster than it’s eliminated, PaCO_{2} will rise. If CO_{2} is removed faster than it’s produced, PaCO_{2} will fall.

Although ventilation delivers O_{2} and removes CO_{2}, not all of the gas entering and leaving the lungs takes part in this process. In fact, the *tidal volume* (V_{T}) can be divided into two components. The first is the *alveolar volume* (V_{A}), which is the portion that reaches optimally perfused alveoli and is responsible for the exchange of O_{2} and CO_{2}. The second is referred to as the *dead space volume* (V_{D}) because it does not participate in gas exchange.

The total or *physiologic dead space* volume is divided into two components—airway and alveolar. The *airway dead space* is the volume of gas that remains in the conducting airways at the end of inspiration. Since this gas never reaches the alveoli, it cannot remove CO_{2}. *Alveolar dead space* is the volume of gas that goes to non-perfused or under-perfused alveoli, and it will be discussed later in this chapter.

If each component of Equation 2.6 is multiplied by the respiratory rate, volume is converted to volume per minute, and the relationship between the gas leaving the lungs (minute ventilation; ${\dot{\text{V}}}_{\text{E}}$), optimally perfused alveoli (alveolar ventilation; ${\dot{\text{V}}}_{\text{A}}$), and the physiologic dead space (dead space ventilation; ${\dot{\text{V}}}_{\text{D}}$) can be expressed as:

Carbon dioxide excretion is directly proportional to alveolar ventilation, and this allows us to rewrite Equation 2.5 as:

which, according to Equation 2.7, can also be written as:

Equation 2.9 tells us two important things. First, an increase in PaCO_{2} (hypercapnia) can result from a drop in ${\dot{\text{V}}}_{\text{E}}$, an increase in CO_{2} production, or a rise in ${\dot{\text{V}}}_{\text{D}}$. Second, any change in CO_{2} production or ${\dot{\text{V}}}_{\text{D}}$ must be matched by a change in ${\dot{\text{V}}}_{\text{E}}$ if PaCO_{2} is to remain constant.

The physiologic dead space is usually expressed as a fraction of the tidal volume (V_{D}/V_{T}). When V_{D}/V_{T} is high, V_{A} is low, and each breath is relatively ineffective at eliminating CO_{2}. When V_{D}/V_{T} is low, V_{A} is high, and much more CO_{2} is excreted. The importance of V_{D}/V_{T} can be emphasized by rewriting Equation 2.9 as:

This shows that ${\dot{\text{V}}}_{\text{E}}$ must increase and decrease with V_{D}/V_{T} if PaCO_{2} is to remain constant. Although V_{D}/V_{T} varies directly with the volume of physiologic dead space, clinically significant changes are more often due to variations in tidal volume. That is, a low V_{T} increases the ${\dot{\text{V}}}_{\text{E}}$ needed to maintain a given PaCO_{2}, whereas a high V_{T} decreases the required ${\dot{\text{V}}}_{\text{E}}$.

Matching of Ventilation and Perfusion

As shown in Figure 2.2, O_{2} is delivered by ventilation and removed by perfusion, and CO_{2} is delivered by perfusion and removed by ventilation. It follows that the partial pressure of oxygen and carbon dioxide in the gas of *each* alveolus (P_{A}O_{2}; P_{A}CO_{2}) must be determined by its *ratio* of ventilation to perfusion ($\dot{\text{V}}/\dot{\text{Q}}$). It’s important to recognize that here P_{A}O_{2} and P_{A}CO_{2} refer to the partial pressure of gas in *individual alveoli*, whereas ${\text{P}\overline{\text{A}}\text{O}}_{\text{2}}$ and ${\text{P}\overline{\text{A}}\text{CO}}_{\text{2}}$ refer to mean values of *all* alveolar gas.

Figure 2.3 illustrates the effect of three different ventilation–perfusion ratios. An “ideal” alveolus has a $\dot{\text{V}}/\dot{\text{Q}}$ that allows the ratio of CO_{2} to O_{2} exchange (R) to equal the ratio of total body CO_{2} production to O_{2} consumption (the respiratory quotient; RQ). If R and RQ are assumed to be 0.8, this ideal $\dot{\text{V}}/\dot{\text{Q}}$ is very close to 1.0, and P_{A}O_{2} and P_{A}CO_{2} are approximately 100 mmHg and 40 mmHg, respectively. If an alveolus receives blood flow but no ventilation, $\dot{\text{V}}/\dot{\text{Q}}$ is zero, O_{2} cannot enter and CO_{2} cannot leave, and the P_{A}O_{2} and P_{A}CO_{2} will be the same as in the mixed venous blood. If ventilation is intact but perfusion is absent, $\dot{\text{V}}/\dot{\text{Q}}$ is infinity, O_{2} is not removed and CO_{2} cannot enter, and the P_{A}O_{2} and P_{A}CO_{2} will be the same as in the conducting airways.

In fact, as shown in Figure 2.4, there is an infinite number of ventilation–perfusion ratios that may exist within individual alveoli, and every $\dot{\text{V}}/\dot{\text{Q}}$ between zero and infinity produces a unique P_{A}O_{2} and P_{A}CO_{2}. As $\dot{\text{V}}/\dot{\text{Q}}$ increases, P_{A}O_{2} rises and P_{A}CO_{2} falls as ventilation delivers more O_{2} and removes more CO_{2}. As $\dot{\text{V}}/\dot{\text{Q}}$ decreases, the opposite occurs; P_{A}O_{2} falls and P_{A}CO_{2} rises.

In a theoretical *perfect lung*, every alveolus has the same $\dot{\text{V}}/\dot{\text{Q}}$ and the same P_{A}O_{2} and P_{A}CO_{2}. Normal lungs are not perfect, because they have a *distribution* of ratios representing both high and low $\dot{\text{V}}/\dot{\text{Q}}$ alveoli. This modest degree of $\dot{\text{V}}\text{/}\dot{\text{Q}}$ *mismatching* occurs because ventilation and perfusion increase at different rates from the less to the more dependent regions of the lungs. All lung diseases, regardless of whether the airways, parenchyma, or vasculature are primarily affected, generate both abnormally high and low $\dot{\text{V}}/\dot{\text{Q}}$ regions. This is because reduced ventilation or perfusion to some alveoli must be accompanied by increased ventilation or perfusion to others if total ventilation and perfusion are unchanged.

Mismatching of ventilation and perfusion is important because it interferes with the ability of the respiratory system to maintain a normal PaO_{2} and PaCO_{2}. You can see how and why this occurs by studying Figures 2.5 and 2.6, which show lung models that consist of two “compartments.” Each compartment could represent an individual alveolus or a region of a lung that contains any number of alveoli, but for our purposes, let’s assume that each is an entire lung (pretend that you’re looking at a chest X-ray). The blood leaving each lung, which we’ll call the pulmonary venous blood, combines to form the arterial blood.

In Figure 2.5, both lungs receive the same amount of ventilation and perfusion, so their $\dot{\text{V}}/\dot{\text{Q}}$ ratios are identical (in this case, $\dot{\text{V}}/\dot{\text{Q}}\text{=1}\text{.0}$). Notice that under these conditions both lungs have the same P_{A}O_{2} and pulmonary venous PO_{2} (P_{V}O_{2}), and that the PO_{2} and hemoglobin saturation (SO_{2}) of the mixed (arterial) blood are identical to those of the blood leaving each lung. Because these lungs are “perfect,” the averaged alveolar and arterial PO_{2} are equal, and the difference between them (the A–a gradient) is zero.

In Figure 2.6, total ventilation and perfusion are unchanged, but narrowing of the airway to the left lung has altered the $\dot{\text{V}}/\dot{\text{Q}}$ ratios. Now, the right lung receives more ventilation than perfusion $(\dot{\text{V}}/\dot{\text{Q}}\text{= 1}\text{.5})$, and the left lung has more perfusion than ventilation $(\dot{\text{V}}/\dot{\text{Q}}\text{= 0}\text{.5})$. As you would expect, when compared to the values in Figure 2.5, the high $\dot{\text{V}}/\dot{\text{Q}}$ of the right lung has increased P_{A}O_{2} and P_{V}O_{2} and lowered P_{A}CO_{2} and pulmonary venous PCO_{2} (P_{V}CO_{2}). The low $\dot{\text{V}}/\dot{\text{Q}}$ of the left lung has had the opposite effect.

Now things really get interesting. When we combine the blood from the two lungs, we find that the PaO_{2} has fallen from 100 mmHg in Figure 2.5 to 60 mmHg, and the A–a gradient has gone from 0 to 42.5. Why did this happen? To answer this question, you first need to understand that the PO_{2} of a blood mixture is determined by the average *oxygen content*, not the average PO_{2}.

The oxygen content of blood (CxO_{2}) is determined by its hemoglobin concentration (Hb) and saturation (SO_{2}) and is expressed as milliliters of O_{2} per deciliter of blood (ml/dl).

In this equation, 1.34 is the milliliters of O_{2} carried by one gram of completely saturated hemoglobin (ml/g), and Hb is expressed in grams per deciliter of blood (g/dl). When we average the O_{2} contents of the blood leaving the two lungs (C_{V}O_{2}) in Figure 2.6 and solve Equation 2.11 for SO_{2}, we find that the saturation of the arterial blood is 90%, which, based on the normal hemoglobin dissociation curve, corresponds to a PaO_{2} of 60 mmHg. Since the hemoglobin concentration of the blood coming from both lungs is the same, we can actually take a shortcut and simply average the saturations rather than the O_{2} contents of the pulmonary venous blood. Try it and see.

Okay, but how did $\dot{\text{V}}/\dot{\text{Q}}$ mismatching produce such a dramatic drop in the PaO_{2}? The explanation lies in the nonlinear shape of the hemoglobin dissociation curve. Look at Figure 2.7. If we combine two blood samples, each with a PO_{2} of 80 mmHg and an SO_{2} of 95%, the PO_{2} and SO_{2} will be unchanged in the mixture. But what if we were to decrease the PO_{2} in one sample by 30 mmHg and increase it by the same amount in the other? As you can see, hemoglobin saturation (and O_{2} content) is affected much more by a fall in PO_{2} than by an equivalent increase. So, when we combine these blood samples, the average SO_{2} is 84%, which corresponds to a PO_{2} of 55 mmHg. Figure 2.7 demonstrates an essential concept: Blood from high $\dot{\text{V}}/\dot{\text{Q}}$ alveoli can never compensate for the drop in SO_{2} and O_{2} content caused by low $\dot{\text{V}}/\dot{\text{Q}}$ regions.

Figure 2.8 shows what happens when one lung receives no ventilation $(\dot{\text{V}}/\dot{\text{Q}}\text{= 0})$. Since no O_{2} or CO_{2} exchange can occur, the left lung becomes a right-to-left shunt, and mixed venous blood enters the arterial circulation. This causes an even greater fall in PaO_{2} despite the higher PO_{2} of the blood leaving the right lung.

What about the PaCO_{2} in Figures 2.6 and 2.8? Just like the PaO_{2}, the PaCO_{2} is determined by the average *CO*_{2} *content* of the blood coming from the two lungs. But notice that the PaCO_{2} is actually very close to the mean PCO_{2} of the pulmonary venous blood. That’s because the relationship between CO_{2} content and PCO_{2} is nearly linear in the physiological range. This means that high $\dot{\text{V}}/\dot{\text{Q}}$ alveoli can (almost) compensate for the high CO_{2} content of blood coming from low $\dot{\text{V}}/\dot{\text{Q}}$ alveoli.

So, alveoli with low $\dot{\text{V}}/\dot{\text{Q}}$ reduce the PaO_{2}, increase the A–a gradient, and increase the PaCO_{2}. But what about high $\dot{\text{V}}/\dot{\text{Q}}$ alveoli? They, too, interfere with gas exchange because they create *alveolar dead space*. Figure 2.9 shows an extreme example in which the left lung receives ventilation but no perfusion $(\dot{\text{V}}/\dot{\text{Q}}\text{=\hspace{0.17em}}\infty )$. Clearly, the gas entering and leaving the alveoli of this lung cannot assist with CO_{2} excretion and is functionally no different from the gas that fills the conducting airways.

Although it may not be nearly as obvious, alveolar dead space is also produced whenever alveoli receive too little blood flow for their amount of ventilation (i.e., whenever $\dot{\text{V}}/\dot{\text{Q}}\text{>}1$). It may help you to think of these alveoli as having *wasted ventilation* rather than dead space. For example, when an alveolus receives 10 times as much ventilation as perfusion $(\dot{\text{V}}/\dot{\text{Q}}\text{=}10)$, only 10% of the ventilation is needed to excrete CO_{2}, and the other 90% is wasted. Similarly, the right lung in Figures 2.6 and 2.8 also generates excess alveolar dead space.

So what is the clinical effect of high $\dot{\text{V}}/\dot{\text{Q}}$ alveoli? Remember from Equations 2.8 and 2.9 that PaCO_{2} is inversely related to alveolar ventilation, which is the difference between minute and dead space ventilation. By creating alveolar dead space, high $\dot{\text{V}}/\dot{\text{Q}}$ alveoli increase dead space ventilation and reduce alveolar ventilation, which, according to our equations, leads to an increase in PaCO_{2}.

At this point, it would seem that both high and low $\dot{\text{V}}/\dot{\text{Q}}$ alveoli can increase PaCO_{2}; but let’s take a closer look. Although it may be conceptually useful to think that high $\dot{\text{V}}/\dot{\text{Q}}$ alveoli increase PaCO_{2} through their effect on dead space ventilation, that’s not really what happens. As shown in our two-compartment models, it is the alveoli with *low*$\dot{\text{V}}/\dot{\text{Q}}$ ratios that reduce CO_{2} excretion and are actually responsible for any rise in the PaCO_{2}. In fact, as shown in Figures 2.6 and 2.8, blood from high $\dot{\text{V}}/\dot{\text{Q}}$ alveoli has a *low* PCO_{2}, and alveoli with no perfusion (Figure 2.9) contribute no blood at all to the systemic circulation.

In “real life,” any rise in PaCO_{2} resulting from $\dot{\text{V}}/\dot{\text{Q}}$ mismatching is quickly detected by central and peripheral chemoreceptors, which increase respiratory drive and minute ventilation. This is the increase in ventilation that, according to Equation 2.9, compensates for the rise in dead space ventilation and normalizes the PaCO_{2}. But, as you can see, it is triggered by the impaired CO_{2} excretion of low $\dot{\text{V}}/\dot{\text{Q}}$ alveoli and actually has nothing to do with increased dead space. Furthermore, it is primarily the increase in ventilation to low $\dot{\text{V}}/\dot{\text{Q}}$ alveoli that augments CO_{2} excretion (by increasing $\dot{\text{V}}/\dot{\text{Q}}$ ratios) and returns the PaCO_{2} to normal.

Of course, our two-compartment models are a gross oversimplification. In reality, there are hundreds of millions of compartments (alveoli), each of which contributes blood with a specific content of O_{2} and CO_{2} that is determined by its ratio of ventilation to perfusion. The final result is the same, though. High $\dot{\text{V}}/\dot{\text{Q}}$ alveoli increase alveolar dead space and dead space ventilation, whereas low $\dot{\text{V}}/\dot{\text{Q}}$ alveoli cause a drop in the PaO_{2}, increase the A–a gradient, impair CO_{2} excretion, and stimulate a compensatory rise in minute ventilation.

Gas Diffusion

The final component of gas exchange is the *diffusion* of O_{2} and CO_{2} molecules between alveolar gas and pulmonary capillary blood. During this journey, O_{2} moves through the alveolar epithelium, the capillary endothelium, both basement membranes, and the plasma (where some dissolves) before entering the erythrocyte and combining with hemoglobin. Carbon dioxide molecules, of course, move in the opposite direction.

The factors that determine the rate at which gas molecules cross the alveolar–capillary interface ($\dot{\text{V}}$) are shown by this modification of Fick’s law for diffusion:

In this equation, A is the total area of contact between gas and blood in the lungs, (P_{1} – P_{2}) is the difference between the partial pressure of the gas in the alveolus and the blood, and D is the distance that the molecules must travel.

Oxygen moves from a mean P_{A}O_{2} of around 100 mmHg to a mixed venous PO_{2} of about 40 mmHg. Carbon dioxide diffuses from a mixed venous PCO_{2} of about 46 mmHg to a P_{A}CO_{2} of approximately 40 mmHg. Remember, though, that the actual PO_{2} and PCO_{2} of the gas in each alveolus are determined by its $\dot{\text{V}}/\dot{\text{Q}}$ ratio.

The lungs have an enormous gas–blood interface of approximately 100 square meters, and the alveolar-capillary membrane is incredibly thin at 0.2–0.5 μm. So it’s not surprising that O_{2} and CO_{2} normally equilibrate very rapidly between alveolar gas and capillary blood. It’s estimated that blood normally spends about 0.75 second in each alveolar capillary. But, as shown in Figure 2.10A, it normally takes only about one-third of this time for the alveolar and capillary PO_{2} and PCO_{2} to equilibrate. Even during exercise, when transit time is reduced by the increase in cardiac output, equilibration occurs before blood leaves the alveoli.

Diseases that decrease the area of the gas–blood interface or increase diffusion distance, however, slow the rate of diffusion. If diffusion impairment is severe (Figure 2.10B), there may be insufficient time for equilibration to occur, and this will produce a gradient between alveolar and end-capillary PO_{2} and PCO_{2}. As you can see from Figure 2.10, because of the larger partial pressure gradient, the PaO_{2} will be affected much more than the PaCO_{2}. If diffusion impairment does cause the PaCO_{2} to rise, a compensatory increase in minute ventilation will be needed to return it to normal.

## Additional Reading

Petersson J, Glenny RW. Gas exchange and ventilation–perfusion relationships in the lung. *Eur Respir J.* 2014;44:1023–1041.Find this resource:

Wagner PD. The physiological basis of pulmonary gas exchange: Implications for clinical interpretation of arterial blood gases. *Eur Respir J.* 2015;45:227–243.Find this resource: