Dorsal Respiratory Groups

As shown in Figure 11-1, the DRG consist of mainly inspiratory neurons located bilaterally in an area called the nucleus of the tractus solitarius. These neurons send impulses to the phrenic and external intercostal motor nerves in the spinal cord, providing the main stimulus for inspiration.¹ Many DRG nerves extend into the VRG, but few VRG fibers extend into the DRG. Thus, reciprocal inhibition is an unlikely explanation for rhythmic, spontaneous breathing.¹ The DRG consists of two neuron populations, one of which is inhibited by deep lung inflation (causing cessation of inspiratory effort), whereas the other is excited by lung inflation (causing continued inspiratory effort).¹ These neurons are involved in the Hering-Breuer and Head’s reflexes described later in this chapter.

The vagus and glossopharyngeal nerves transmit many sensory impulses to the DRG from the lungs, airways, peripheral chemoreceptors, and joint proprioceptors. These impulses modify the basic breathing pattern generated in the medulla.

Ventral Respiratory Groups

VRG neurons are located bilaterally in the medulla in two separate nuclei that contain both inspiratory and expiratory neurons (see Figure 11-1). Some inspiratory VRG neurons send motor impulses through the vagus nerve to the laryngeal and pharyngeal muscles, abducting the vocal cords and increasing the diameter of the glottis. Other VRG inspiratory neurons transmit impulses to the diaphragm and external intercostal muscles. Still other VRG neurons have mostly expiratory discharge patterns and send impulses to the internal intercostal and abdominal expiratory muscles. The most rostral (toward the head) part of the VRG, Bötzinger’s complex (see Figure 11-1), contains the only expiratory neurons known to inhibit the inspiratory VRG and DRG impulses.²

Respiratory Rhythm Generation

The exact origin of the basic rhythmic pattern of breathing is unknown; no single group of pacemaker cells have been identified. The Bötzinger complex and a structure located between it and the VRG, the pre-Bötzinger complex, are thought to be responsible for rhythmic breathing.² Two predominant theories of rhythm generation are the pacemaker hypothesis and the network hypothesis. The pacemaker hypothesis holds that certain medullary cells have intrinsic pacemaker properties—that is, rhythmic self-exciting characteristics—and that these cells drive other medullary neurons. The network hypothesis suggests that rhythmic breathing is the result of a particular pattern of interconnections between neurons dispersed throughout the rostral VRG, the pre-Bötzinger complex, and the Bötzinger complex. This hypothesis assumes that certain populations of inspiratory and expiratory neurons inhibit one another and that one of the neuron types fires in a self-limiting way, such that it becomes less responsive the longer it fires. There is no definitive proof of either hypothesis; the precise origin of respiratory rhythm generation remains elusive.

Inspiratory Ramp Signal

The dorsal and ventral inspiratory neurons do not send an abrupt burst of impulses to the inspiratory muscles; instead, their firing rate increases gradually after expiration ceases, creating a smoothly increasing ramp signal (Figure 11-2). This ramp signal leads to a progressively stronger contraction of inspiratory muscles, smoothly and gradually inflating the lungs instead of filling them in an abrupt inspiratory gasp. During exercise, various peripheral reflexes and receptors influence the medullary neurons, steepening the inspiratory ramp signal and filling the lungs more rapidly.

As the inspiratory ramp signal strengthens, inhibitory neurons begin to fire with increasing frequency.² After approximately 2 seconds, these restraining impulses become strong enough to switch off the inspiratory signal abruptly. Expiration proceeds for about 3 seconds.³ As expiration begins, inspiratory neurons fire briefly, “braking” the early phase of expiration by maintaining some inspiratory muscle tone (see Figure 11-2). Inspiratory neuronal activity completely stops in the last phase of expiration. The inhibitory neurons that switch off the inspiratory ramp arise from the pneumotaxic center and pulmonary stretch receptors, which are discussed in the next section.

Apneustic and Pneumotaxic Centers

The two respiratory centers in the pons (apneustic and pneumotaxic) have been identified through animal brain transection experiments. If the brainstem is severed at the midpons level (see Figure 11-1, level II), and the vagus nerves connecting this area with the lower parts of the brain and spinal cord are also cut, a breathing pattern called apneusis results (Figure 11-3). Apneusis consists of prolonged inspiratory gasps interrupted by occasional expirations. The apneustic center is thought to be in a region in the lower pons, but it has never been anatomically identified with certainty.² Its existence and function can be demonstrated only if its connections to the higher pneumotaxic center are severed and the vagus nerves are cut. Under these circumstances, the area in the lower pons (i.e., the apneustic center) sends signals to the DRG neurons, preventing the inspiratory ramp signal from being switched off.³ Apparently, vagal and pneumotaxic center impulses hold the apneustic center’s stimulatory effect on DRG neurons in check because apneusis does not occur if the vagus is left intact (see Figure 11-3).

The pneumotaxic centers are bilateral groups of neurons in the upper pons (see Figure 11-1). These centers control the off-switch point of the DRG’s inspiratory ramp signal, sending inhibitory impulses to medullary inspiratory neurons; thus, pneumotaxic signals control the length of inspiration. Strong pneumotaxic signals shorten the inspiratory time and increase the respiratory rate. Weak signals produce prolonged inspiration and large tidal volumes. The primary function of the pneumotaxic center is to limit inspiration and hold apneustic center impulses in check. The exact nature of the interaction between the apneustic and pneumotaxic centers remains poorly understood; they apparently work together to control the depth of inspiration.³ The effect of the pontine centers on the lower medullary centers seems to be fine-tuning the respiratory pattern because the transection of the brainstem between the pons and medulla (see Figures 11-1, level III, and 11-3) results in an irregular breathing pattern.¹

Hering-Breuer Inflation Reflex

The Hering-Breuer inflation reflex, described by Hering and Breuer in 1868, is generated by stretch receptors located in smooth muscle of large and small airways. These receptors are called slowly adapting receptors because their activity continues as long as the stimulus persists (see Chapter 2). When stretched, these receptors send inhibitory impulses through the vagus nerve to the DRG neurons, stopping further inspiration. Thus, the Hering-Breuer reflex has an effect similar to that of the pneumotaxic center.³ The Hering-Breuer reflex in adults is activated only at large tidal volumes (≥800 to 1000 mL) and is apparently not an important control mechanism in quiet breathing. However, this reflex is important in regulating the respiratory rate and depth during moderate to strenuous exercise.²

Diseases associated with low lung compliance (i.e., strong elastic retractive forces) may stimulate the inflation stretch receptors by increasing the mechanical stress acting on the airways during inspiratory efforts.² This stimulation tends to increase the respiratory rate because it sends an early inhibitory signal, shortening the inspiratory phase.

Various combinations of respiratory rate and tidal volume can produce the level of ventilation needed to meet metabolic gas exchange demands, but there is always a specific combination that results in the least work. The information stretch receptors send to the medullary centers, which can be consciously sensed, probably helps to establish this optimal combination. The sensation of dyspnea is thought to be produced when the medullary centers demand a level of ventilation greater than the actual ventilation achieved.²

Hering-Breuer Deflation Reflex

Hering and Breuer also observed that sudden collapse of the lung stimulates strong inspiratory efforts and increases the respiratory rate. This deflation reflex may be the result of decreased stretch receptor activity, or it may be caused by the stimulation of other receptors, such as the irritant receptors and J-receptors (discussed later). Although it is unclear which receptors are involved, it is clear that the vagus nerve is the pathway (as it is for the Hering-Breuer reflex) and that the effect is hyperpnea.¹ The deflation reflex is probably responsible for the hyperpnea observed with pneumothorax.

Head’s Paradoxical Reflex

In 1889, Head observed that if the Hering-Breuer reflex is partially blocked by cooling the vagus nerve, lung hyperinflation causes a further increase in inspiratory effort, opposite of the Hering-Breuer reflex. The receptors for this paradoxical reflex are called rapidly adapting receptors because they stop firing promptly after a volume change occurs. Head’s reflex may help maintain large tidal volumes during exercise and may be involved in periodic deep sighs during quiet breathing. Periodic sighs help prevent alveolar collapse or atelectasis. This reflex also may be involved in stimulating the first breaths of a newborn infant.¹

Irritant Receptors

Rapidly adapting irritant receptors in the epithelium of the larger conducting airways have vagal sensory nerve fibers. When stimulated by inhaled irritants or by mechanical factors, they cause reflex bronchoconstriction, coughing, sneezing, tachypnea, and narrowing of the glottis. Some of these reflexes, called vagovagal reflexes, have both sensory and motor vagal components; sensory nerve stimulation results in motor reflexes that cause laryngospasm, bronchospasm, coughing, and slowing of the heartbeat. Endotracheal intubation, airway suctioning, and bronchoscopy readily elicit vagovagal reflexes. Physical stimulation of the conducting airways, as with suctioning or bronchoscopy, may cause a severe case of bronchospasm, coughing, and laryngospasm.

J-Receptors

C-fibers in the lung parenchyma near pulmonary capillaries are called juxtacapillary receptors, or J-receptors (see Chapter 1). They are stimulated by alveolar inflammatory processes (pneumonia), pulmonary vascular congestion (congestive heart failure), and edema. J-receptor stimulation causes rapid, shallow breathing; a sensation of dyspnea; and expiratory narrowing of the glottis. Expiratory narrowing of the glottis causes grunting on expiration, especially in infants.

Peripheral Proprioceptors

Proprioceptors (positional sensors) in muscles, tendons, and joints, and pain receptors in muscles and skin send stimulatory impulses to the medullary centers, increasing inspiratory activity and hyperpnea.² For this reason, moving the limbs, slapping the skin, and other painful stimuli stimulate ventilation in patients who have respiratory depression. Splashing cold water on the skin has a similar effect.

Proprioceptors in joints and tendons may be important in initiating and maintaining increased ventilation at the beginning of exercise. Passive limb movement around a joint increases the breathing rate in anesthetized animals and unanesthetized humans.³

Muscle Spindles

Muscle spindles in the diaphragm and intercostal muscles are part of a reflex arc that helps the muscles adjust to an increased load. Muscle spindles are stretch-sensing elements located on intrafusal muscle fibers, which are arranged in parallel with the main extrafusal muscle fibers (Figure 11-4). The extrafusal skeletal muscle fibers that elevate the ribs (see Figure 11-4) are innervated by different motor fibers (alpha motor fibers) than the intrafusal spindle fibers (gamma motor fibers). As the main extrafusal muscle fiber and the intrafusal fibers contract together in parallel, the spindle-sensing element stretches and sends impulses over spindle afferent nerves directly to the spinal cord (see Figure 11-4). The number of impulses sent is proportional to the spindle’s stretch. The spindle’s afferent (sensory) nerve synapses directly with the same alpha motor neuron in the spinal cord that sends motor impulses back to the main extrafusal muscle fiber; this creates a single synapse reflex arc. When stimulated by the spindle’s afferent nerve signal, the alpha motor neuron sends signals to only the main extrafusal muscle fibers, causing them to contract with greater force, which unloads and shortens the adjacent intrafusal muscle fibers. As a result, the stretch-sensitive spindle is unloaded, and its impulses cease. In this way, inspiratory muscle force automatically adjusts to the load imposed by decreased lung compliance or increased airway resistance.

Central (Medullary) Chemoreceptors

Hydrogen ions, not CO₂ molecules, stimulate highly responsive chemosensitive nerve cells, located bilaterally in the medulla. Nevertheless, these central chemoreceptors are extremely responsive to CO₂ because the [H⁺] surrounding them is dependent on the reaction between CO₂ and H₂O in their local environment. The chemoreceptors are not in direct contact with arterial blood. Instead, they are bathed in the cerebral spinal fluid, separated from the blood by a semipermeable membrane called the blood-brain barrier. This membrane is almost impermeable to H⁺ and HCO₃⁻, but it is freely permeable to CO₂. When the PaCO₂ rises, CO₂ diffuses rapidly through the blood-brain barrier into the CSF. In the CSF, CO₂ reacts with H₂O to form hydrogen ions (Figure 11-5); this stimulates the central chemoreceptors, which in turn stimulate the medullary inspiratory neurons. Because CSF contains no protein buffers, CO₂ diffusion from the blood into the CSF increases [H⁺] almost instantly, exciting the central chemoreceptors within seconds. Through this mechanism, alveolar ventilation increases by approximately 2 to 3 L/min for each millimeter of mercury increase in PaCO₂.⁴ In this indirect fashion, PaCO₂ is the principal minute-to-minute stimulus for ventilation mediated through the central chemoreceptors.

Figure 11-6 compares the ventilatory response to increased PaCO₂ with the ventilatory response to increased arterial [H⁺] arising from fixed acid accumulation (i.e., acidemia from a nonrespiratory cause). A great change in ventilation occurs between PaCO₂ values of 40 mm Hg and 80 mm Hg compared with a relatively small change in ventilation caused by a fixed acid–induced fall in arterial pH from 7.40 to 7.00 (see Figure 11-6).

Hydrogen ions in the arterial blood do not readily diffuse across the blood-brain barrier and thus cannot stimulate the medullary chemoreceptors to any great extent; the effect of hydrogen ions is mediated mainly through the less responsive peripheral chemoreceptors, which are discussed in the next section.

The stimulatory effect of CO₂ on the central chemoreceptors gradually declines over 1 to 2 days because of renal compensatory responses. The kidneys increase blood bicarbonate ion [HCO₃⁻] concentration in response to respiratory acidosis (see Chapter 10), which returns blood pH to the normal range. The increased number of bicarbonate ions in the blood eventually diffuses across the blood-brain barrier into the CSF where they buffer H⁺ and bring the CSF pH back to normal; this removes the stimulus to the chemoreceptors, and ventilation decreases. Although an increase in PaCO₂ has an extremely powerful effect on ventilation, its effect is greatly weakened after 1 or 2 days of adaptation.

Peripheral Chemoreceptors

The peripheral chemoreceptors are small, highly vascular tissues known as the carotid and aortic bodies. The carotid bodies are small organs located bilaterally in bifurcations of the common carotid arteries (Figure 11-7). The aortic bodies are found in the arch of the aorta. These neural structures increase their firing rates in response to increased arterial [H⁺] regardless of its origin—that is, whether from fixed acid accumulation or increased CO₂. The carotid bodies contain far more chemoreceptors than the aortic bodies; they send their impulses over the glossopharyngeal nerve to medullary DRG and VRG, whereas the aortic bodies send their impulses over the vagus nerve (see Figure 11-7). The carotid bodies exert much more influence over the respiratory centers than do the aortic bodies, especially with respect to arterial hypoxemia and acidemia.¹

The carotid bodies receive an extremely high blood flow, about 20 times their weight each minute.³ This extremely high flow rate means the carotid bodies have little time to remove oxygen from the blood, so the venous blood leaving the carotid bodies has almost the same oxygen content as the arterial blood entering them. The carotid bodies are exposed at all times to arterial (not venous) blood, and they sense arterial (not venous) [H⁺].

Response to Decreased Arterial Oxygen

Traditionally, it was believed that the carotid bodies directly sensed decreased PaO₂ levels, implying that arterial hypoxemia represents an independent drive to breathe—the so-called hypoxic drive. Although peripheral chemoreceptors fire more frequently in the presence of arterial hypoxemia, they do so only because hypoxemia makes them more sensitive to hydrogen ions.² That is, in the presence of hypoxemia, carotid body sensitivity to a given [H⁺] increases; in this way, hypoxia increases ventilation for any given pH. Conversely, a very high PO₂ (hyperoxia) decreases carotid body sensitivity to [H⁺]. Thus, the carotid bodies respond to arterial hypoxemia but only because hypoxia makes them more sensitive to [H⁺]; this implies that if the arterial [H⁺] is extremely low, as in severe alkalemia, hypoxemia has little effect on the carotid bodies.² Simply stated, the ultimate effect of hypoxemia is to increase the neural firing rate of the peripheral chemoreceptors, which increases minute ventilation.

Because carotid body tissues have extremely high blood flow rates, they respond to decreased PaO₂ (in the indirect way just described) rather than to an actual decrease in arterial oxygen content. That is, the carotid bodies’ extraction of oxygen from each unit of rapidly flowing blood is so small that their oxygen needs are met entirely by dissolved oxygen in the plasma. This is why conditions associated with low arterial oxygen content but normal PaO₂ (e.g., anemia and carbon monoxide poisoning) do not stimulate ventilation.⁴

When pH and PaCO₂ are normal (i.e., pH = 7.40 and PaCO₂ = 40 mm Hg), the carotid bodies’ rate of nerve impulse transmission does not increase significantly until PaO₂ decreases to about 60 mm Hg.⁴ As PaO₂ decreases from 60 mm Hg to 30 mm Hg, the rate of impulse transmission increases sharply and linearly because hypoxemia makes the carotid bodies much more sensitive to a pH of 7.40 (Figure 11-8). A decrease in PaO₂ from 60 mm Hg to 30 mm Hg corresponds to the sharpest decrease in oxygen content on the oxygen-hemoglobin equilibrium curve—that is, the steepest part of the curve. Arterial hypoxemia does not stimulate ventilation greatly until the PaO₂ decreases to values less than 60 mm Hg; this is why ventilation increases as one ascends to high altitudes. Decreased barometric pressure at high altitudes decreases inspired and arterial PO₂ values, which increases the sensitivity of peripheral chemoreceptors to hydrogen ions. However, the resulting increase in ventilation is less than expected because hyperventilation increases arterial pH by decreasing PaCO₂ and subsequently, blood [H⁺]. The increase in arterial pH depresses the medullary respiratory center, counteracting the excitatory effect of decreased PaO₂ on peripheral chemoreceptors. However, lung mechanics in certain conditions, such as advanced chronic obstructive pulmonary disease, are so deranged that the stimulatory effect of hypoxemia on ventilation fails to decrease PaCO₂ regardless of the patient’s effort. In such instances, there is no alkalosis to counteract stimulatory effects of hypoxemia on ventilation.

Control of Breathing during Chronic Hypercapnia

A sudden increase in PaCO₂ causes an immediate increase in ventilation because CO₂ rapidly diffuses from the blood into the CSF, increasing [H⁺] surrounding the central chemoreceptors. On the other hand, if PaCO₂ increases gradually over many years, as might occur in the development of severe COPD and steadily worsening lung mechanics, the kidneys compensate by increasing the plasma bicarbonate concentration, keeping the arterial pH within normal limits (see Chapter 10). As plasma bicarbonate levels increase, these ions slowly diffuse across the blood-brain barrier, keeping CSF pH in its normal range. Because the central chemoreceptors respond to [H⁺], not the CO₂ molecule, they sense a normal pH environment, even though the PaCO₂ is abnormally high.

This adaptation explains why chronically elevated PaCO₂ in patients with severe COPD does not overly stimulate ventilation. Instead, the hypoxemia that accompanies chronic hypercapnia becomes the minute-to-minute breathing stimulus in the roundabout way discussed previously: Hypoxemia increases the peripheral chemoreceptors’ sensitivity to [H⁺], increasing the nerve impulses they transmit to the medulla, stimulating ventilation. Patients with severe COPD are invariably hypoxemic when breathing room air because their lungs have severe mismatches in ventilation and blood flow. It stands to reason that breathing supplemental oxygen would increase the PaO₂ and make the carotid bodies less sensitive to [H⁺], which would decrease ventilation further and thus increase PaCO₂

Oxygen-Associated Hypercapnia

An acute increase in PaCO₂ sometimes occurs after oxygen is given to patients with severe COPD who are chronically hypoxemic and hypercapnic; the reason for this phenomenon continues to be a subject of much debate and misunderstanding. The traditional explanation has been that oxygen breathing removes hypoxemia’s stimulatory effect on the peripheral chemoreceptors. As explained in the previous section, the ventilatory stimulus of the peripheral chemoreceptors increases significantly when the PaO₂ decreases to less than 60 mm Hg; this is true in hypoxemic patients with or without a history of chronic hypercapnia. If breathing supplemental oxygen increases the PaO₂ to levels greater than 60 mm Hg, it stands to reason that the stimulatory effect of the peripheral chemoreceptors on ventilation would diminish, which would decrease ventilation and increase PaCO₂. This explanation for the increase in PaCO₂ after oxygen administration is widely accepted in reference to hypoxemic patients without COPD—for example, a patient with pneumonia whose hypoxemia is corrected by O₂ administration; it seems to be challenged only in reference to patients with COPD who are chronically hypercapnic. The major reason

CLINICAL FOCUS 11-3   Chronic Physiological Effect of Low Inspired Oxygen at High Altitude

You and some friends decide to vacation in the mountains of Colorado and climb a 14,000-foot peak (barometric pressure = 460 mm Hg). You are currently at sea level. You drive to your destination in 1 day, and the next morning you hike to the peak of the 14,000-foot mountain and set up camp for the night.

Questions and Discussion

1. At the top of the mountain, what is the inspired PO₂ (P_IO₂) in your trachea?

< ?xml:namespace prefix = "mml" />PIO2=0.21(460−47)=87mm Hg

2. What is your alveolar PO₂ (P_AO₂) if your ventilation does not change and P_ACO₂ remains at 40 mm Hg?

PAO2=0.21(460−47)−(40×1.2)=39mm Hg

PaO₂ would be a few millimeters of mercury lower than this because of a normal P(A-a)O₂ difference (secondary to normal anatomical shunting). Actually, your ventilation does not remain the same; hypoxia makes your peripheral chemoreceptors more sensitive, which stimulates ventilation quite strongly at this PaO₂.

3. If your ventilation increases enough to reduce your PaCO₂ to 20 mm Hg, what is your P_AO₂?

PAO2=0.21(460−47)−(20×1.2)=63mm Hg

(Because of the normal P(A-a)O₂, PaO₂ would be in the upper 50s.)

4. What is the long-term physiological effect of high altitude on your ventilation?

A PaCO₂ of 20 mm Hg immediately decreases CSF PCO₂ to 20 mm Hg and creates an even greater CSF alkalosis than it creates in the blood because the blood-brain barrier is nearly impermeable to HCO₃⁻ ions. That is, as CO₂ rapidly diffuses out of the CSF in response to the low PaCO₂, HCO₃⁻ cannot follow. Your central chemoreceptors are exposed to an alkaline environment, which works to suppress their ventilatory stimulus, counteracting the peripheral chemoreceptor stimulus arising from hypoxia. These events limit the extent to which hypoxia increases your ventilation. However, as time progresses, HCO₃⁻ gradually diffuses out of the CSF across the blood-brain barrier and into the blood, which brings the CSF pH back toward normal. At the same time, your kidneys compensate for the hyperventilation-induced alkalemia by excreting HCO₃⁻ ions; over the next 24 hours, this brings your blood pH back toward normal—even though the PaCO₂ is still low. With the inhibiting effects of an alkalotic CSF removed, the hypoxic stimulus gradually increases ventilation again. On arriving at the mountaintop, about 24 hours pass before your minute ventilation increases to its maximum level. This considerable increase in ventilation greatly improves your ability to tolerate the low P_IO₂.

why alternative explanations have been sought is that the reduction in minute ventilation after oxygen breathing in patients with advanced COPD is not always severe enough to account for the increased PaCO₂.^5,6

The question is not whether hypoxemia heightens peripheral chemoreceptor sensitivity to [H⁺] in hypercapnic patients with COPD (it does) but whether oxygen administration decreases this sensitivity sufficiently to account for the observed increase in PaCO₂. An additional pertinent question is whether an acute increase in PaCO₂ on top of already elevated PaCO₂ stimulates the medullary chemoreceptors in these patients as it does in healthy patients.

Several investigators have suggested that oxygen breathing worsens the lungs’ ventilation-perfusion (V˙/Q˙) relationships, creating more dead space.5,6 Other investigators suggested that oxygen-induced hypercapnia is caused by the combined effects of hypoxic stimulus removal and redistribution of V˙/Q˙ relationships in the lungs.^7–9

V˙/Q˙ relationships may worsen in the lungs after oxygen administration because improved oxygenation abolishes hypoxic pulmonary vasoconstriction in poorly ventilated lung regions. The resulting reduction in vascular resistance shifts more blood flow to these underventilated areas, drawing blood flow away from well-ventilated regions (Figure 11-9). As poorly ventilated regions receive more blood flow, they become even less ventilated as oxygen-rich inspired gas washes out resident nitrogen gas, making alveoli more subject to absorption atelectasis; that is, oxygen may be absorbed by the pulmonary circulation more rapidly than the slowed ventilation can replenish it (notice the further decreased V˙ in Figure 11-9, B). As a result, inspired gas flows preferentially to the already well-ventilated alveoli (see Figure 11-9, B), increasing their V˙/Q˙ ratios. The increased V˙/Q˙ in these alveoli is further exaggerated because they lose some blood flow to the poorly ventilated regions, which now have lower vascular resistance owing to oxygen breathing. In the end, more blood flow is directed to poorly ventilated alveoli, which takes blood flow away from well-ventilated alveoli. The key point is that when already underventilated alveoli receive additional blood flow, arterial blood PCO₂ increases further. These events occur without a decline in overall minute ventilation.

Although some investigators believe the mechanisms just described explain why oxygen administration induces hypercapnia in patients with COPD, other studies support an equally important role for oxygen suppression of the hypoxic ventilatory stimulus.7–9 These studies demonstrated that in exacerbated, chronically hypercapnic patients with COPD, oxygen administration significantly reduced their ventilation^8,9 and elevated the PaCO₂ level required to stimulate ventilation via the medullary chemoreceptors.⁷

The diagnosis of COPD on a patient’s medical record does not automatically mean the patient has a chronically increased PaCO₂ or that hypercapnia will follow oxygen administration. These characteristics are present only in patients with severe end-stage disease, which includes only a small percentage of patients with the diagnosis of COPD; therefore, concern about oxygen-induced hypercapnia and acidemia is not warranted in most patients with a diagnosis of COPD. In any event, the fear of inducing hypoventilation and hypercapnia is never a justifiable rationale for withholding oxygen from acutely hypoxemic patients with COPD. Tissue oxygenation is an overriding priority; oxygen must never be withheld for any reason from hypoxemic COPD patients experiencing exacerbations of their disease. The clinician must be prepared to mechanically support the patient’s ventilation if severe hypoventilation occurs.

Carbon Dioxide and Cerebral Blood Flow

CO₂ plays an important role in regulating cerebral blood flow (CBF). Its effect is mediated through the formation of hydrogen ions by CO₂.¹² Increased PCO₂ dilates cerebral vessels, increasing CBF, whereas decreased PCO₂ constricts cerebral vessels and reduces CBF. In patients with traumatic brain injury (TBI), the brain swells acutely; this increases the intracranial pressure (ICP) in the rigid skull to such high levels that blood supply to the brain might be cut off, causing cerebral hypoxia (ischemia)—increased ICP may exceed cerebral arterial pressure and stop blood flow. Normal ICP is less than 10 mm Hg; a pressure of 20 mm Hg places the brain at risk for ischemia and requires active treatment to reduce the ICP.¹² The brain has a high oxygen requirement; it accounts for about 20% of the entire body’s resting oxygen consumption. To meet this oxygen demand, the body maintains a constant CBF of about 50 to 60 mL per 100 g of brain tissue per minute; CBF less than 18 mL/100 g/min is the threshold for cerebral ischemia.¹⁸ To sustain the CBF, cerebral perfusion pressure (CPP) of about 60 mm Hg must be maintained. CPP is the difference between mean arterial pressure and ICP; cerebral ischemia is likely to occur if CPP decreases to less than 55 mm Hg.¹²

Mechanical hyperventilation has been used for many years in TBI to reduce PaCO₂ and reduce CBF and ICP. In patients with TBI, a cerebral blood volume reduction of only 0.5 to 0.7 mL decreases ICP by 1 mm Hg; on the other hand, for every 1-mm Hg acute reduction in PaCO₂ (between 20 mm Hg and 60 mm Hg), there is a 3% reduction in CBF.¹² Thus, although an acute reduction in PaCO₂ reduces ICP, it also reduces CBF and potentially causes cerebral ischemia. In any event, hypoventilation in a patient with head trauma and a preexisting elevated ICP is especially dangerous because the resulting hypercapnia dilates cerebral vessels and elevates ICP even more.