Pulmonary Circulation

Anatomy

The pulmonary arterial blood undergoes oxygenation and filtration in the lung. The pulmonary arterial vascular circuit is comprised of the pulmonary arteries, the alveolar capillary network, and the pulmonary veins. Systemic venous blood exits the heart from the right ventricle through the main pulmonary artery, an anterior and intrapericardial structure. After a distance of approximately 3-5 cm, the main pulmonary artery bifurcates into right and left main pulmonary arteries. The right main pulmonary artery crosses the mediastinum to the right hilum posterior to the ascending aorta and superior vena cava (SVC) and anterior to the carina and the esophagus ( Fig. 8-1 ). Within the right hemithorax the artery branches first into upper and lower trunks, and then into segmental vessels that roughly follow the bronchial segments. Beyond this point pulmonary artery anatomy becomes extremely variable. Common variants of the right pulmonary artery include an accessory branch to the posterior segment of the upper lobe from the lower trunk, and two arteries to the middle lobe.

The left pulmonary artery courses superiorly and posteriorly toward the left hilum (see Fig. 8-1 ). In this location the artery is anterior to the descending thoracic aorta and closely approximated to the undersurface of the arch. At the hilum the left pulmonary artery gives off its first branches to the upper lobe. These are usually multiple individual or paired arteries. A single common upper lobe trunk is seen in fewer than 20% of individuals. The lingula is usually supplied next, from the descending portion of the left pulmonary artery before it divides into the lower lobe vessels. As with the right pulmonary artery, the lower lobe left pulmonary arterial branches approximate the segmental bronchial anatomy, but are highly variable.

The pulmonary arteries are elastic vessels that contain only small amounts of smooth muscle cells down to the level of fifth-order branches. Conversely, the pulmonary arterioles are very muscular. Because of the elastic nature of the pulmonary arteries and the extensive capillary network, the capacity of this vascular bed is enormous. Normal main pulmonary artery pressures are approximately 22/8 mm Hg, with a mean of 13 mm Hg.

The diameter of the individual pulmonary capillaries averages 8-9 μm. The capillary bed is vast (almost 90 m ² ), which permits efficient and rapid gas exchange at the alveolar level. This bed also performs another important function—filtration of solid waste from the venous blood before it reaches the left side of the heart and the systemic arteries. The small size but immense number of capillaries allows filtration of large quantities of particulate material without compromising gas exchange or blood flow. Active intrinsic thrombolytic and phagocytic systems in the lung rapidly dispose of normal physiologic debris. There are also anastomoses to the systemic (bronchial) arteries at the capillary level, although in normal individuals these have no clinical importance.

Oxygenated blood is returned to the left atrium by the pulmonary veins ( Fig. 8-2 ). Typically, one upper and two lower veins are formed from coalescence of the segmental veins in each lung. The right pulmonary veins lie inferior to the pulmonary artery and posterior to the SVC. The right middle lobe usually drains into the upper vein, but may empty directly into the left atrium. On the left, the pulmonary veins also lie inferior to the pulmonary artery, and anterior to the descending thoracic aorta. The lingular veins drain with the upper lobe segments. Anomalous venous drainage to the SVC, systemic thoracic veins, or coronary sinus may be found in association with congenital heart disease, pulmonary sequestration, or as an isolated occurrence ( Fig. 8-3 ).

At any point in time, 30% of pulmonary blood is in the arteries, 20% is in the capillaries, and 50% is in the veins. In the supine position the blood volume is relatively evenly distributed between the upper and lower lobes. When upright, the lower lobes are preferentially perfused.

The bronchial arteries, branches of the thoracic aorta, provide blood supply to the airways. Bronchial arteries are normally small vessels that are highly variable in number, but the most common pattern (45%) is two on the left and one on the right. The right bronchial artery arises from a common intercostal trunk in over 70% of individuals, but only 5% of left bronchial arteries have a common origin with an intercostal artery. One quarter of individuals will have single bronchial arteries bilaterally, whereas 30% will have four or more. Right and left bronchial arteries have a common origin in about 40% of individuals ( Figs. 8-4 and 8-5 ). Bronchial arteries usually are located on the anterolateral surface of the thoracic aorta just below the ligamentum arteriosum at the level of the T3-T4 vertebral bodies. Variant sites of origin include the inner surface of the aortic arch (15%), internal mammary, brachiocephalic, inferior thyroidal, and subclavian arteries. Bronchial arteries (especially those on the right) have communication with the anterior spinal artery in 10% of patients. Anastomoses may also be present with the coronary arteries. The venous drainage of the bronchi is through both the systemic veins of the thorax and the pulmonary veins.

Key Collateral Pathways

Pulmonary arteries are considered end arteries, in that few normal intrapulmonary anastomoses exist. Acute proximal occlusion of a normal pulmonary artery segment usually results in distal infarction of the subtended lung parenchyma. Congenital proximal pulmonary artery obstruction is relieved by flow through a patent ductus arteriosum, as well as the bronchial arteries and other mediastinal arteries. In adults with longstanding acquired pulmonary artery occlusions, reconstitution of peripheral pulmonary arteries by small distal intrapulmonary collaterals can occur. The bronchial arteries can provide collateral supply to both the lung parenchyma and the pulmonary arteries ( Fig. 8-6 ). Almost every artery that supplies the thorax (including the diaphragm) can potentially provide collateral supply to the pulmonary arteries. When systemic arteries provide substantial collateral flow to the pulmonary arteries, a measurable left-to-right shunt may develop.

Bronchial arteries and the lung parenchyma have multiple potential sources of collateral supply ( Box 8-1 ). These usually develop in response to increased arterial flow demands in the lung tissues from chronic pulmonary infections, granulomatous diseases, and tumors, as well as in congenital or acquired pulmonary artery obstruction ( Fig. 8-7 ). Knowledge of these collateral pathways becomes important during interventions for bronchial artery bleeding. Successful occlusion of the bronchial arteries may fail to control bleeding in patients with well-developed collaterals.

Imaging

Pulmonary Circulation

The optimum imaging modality for the pulmonary vasculature depends on the clinical question, the vessels of interest, available technology, and available technique. Multidetector computed tomography angiography (CTA) is the most widely used imaging technique; it is readily available, fast, and accurate, and it provides comprehensive imaging. Scans performed to evaluate for pulmonary artery embolism follow a different protocol than those to evaluated bronchial arteries. The scan direction (diaphragm to apex for pulmonary embolism, apex to diaphragm for systemic arteries) varies depending on the clinical question and the number of detectors ( Fig. 8-8 ). When possible a peripheral intravenous catheter is used. Injection of contrast through a vein in the right upper extremity minimizes artifact from dense contrast in the left brachiocephalic vein. Power injection of 80-120 mL of contrast at 3-5 mL/sec is crucial to obtain satisfactory images. A scan delay of approximately 20-25 seconds is often used. Prospective cardiac triggering improves image quality but is not routinely used. Streak artifact from contrast in the SVC can interfere with evaluation of the main right pulmonary artery. Delayed scans are useful when evaluating central veins and vascular masses.

Careful postprocessing on an independent workstation facilitates inspection of the pulmonary vasculature. One of the great advantages of pulmonary CTA is the vast amount of additional information about the lung parenchyma, mediastinal structures, and thoracic arteries that can be acquired by simply viewing the same data at different window levels. Patients with suspected pulmonary arterial pathology often have alternate thoracic disease processes that account for or contribute to their symptoms.

Pulmonary arterial magnetic resonance angiography (MRA) usually requires contrast enhancement with gadolinium to obtain satisfactory images ( Fig. 8-9 ). The surrounding aerated lung and cardiac motion limits conventional spin-echo images to evaluation of the central pulmonary arteries. The intrinsic black-blood nature of these images is useful for depiction of central vascular tumors or thrombi. Acquisition times for conventional time-of-flight (TOF) angiographic techniques are too long, and signal loss from in-plane flow is problematic. The most promising techniques are breath-hold fast three-dimensional (3-D) gradient echo time-resolved sequences with bolus injection of gadolinium. Ultra-fast scanners can image the contrast bolus at each step in the pulmonary circuit. Very powerful and fast gradients are required to obtain the extremely short echo times used for time-resolved pulmonary MRA. Limitations of pulmonary MRA are the lack of discrimination of small peripheral vessels with current technology, limited availability, and long duration of the scans compared to CTA.

Catheter angiography remains an extremely important though seldom used diagnostic tool for imaging the peripheral pulmonary arterial circulation. More often, pulmonary angiography is performed as part of an intervention. Selective pulmonary arteriography with a pigtail catheter and low-osmolar contrast agents is more invasive than either CTA or MRA, but safe and definitive in experienced hands.

The patient’s electrocardiogram should be reviewed before pulmonary angiography to evaluate for left bundle branch block. If present, temporary pacing (either external or internal) should be in place before a catheter is introduced into the right heart because transient right bundle branch block can be caused by catheter manipulation. In patients with preexisting left bundle branch, this will result in complete heart block.

Contraindications to pulmonary angiography are severe pulmonary hypertension with an end-diastolic right ventricular pressure of 20 mm Hg or more, unstable ventricular arrhythmias, and untreatable severe contrast allergy.

Pulmonary angiography can be performed from either the femoral (most commonly) or jugular venous approach. Nonselective injection of contrast into the vena cava, right atrium, or main pulmonary artery usually does not provide satisfactory visualization of the peripheral pulmonary vessels and requires very large volumes of contrast. Selective pulmonary angiography is safer and provides the best images. Large pigtail catheters (6- or 7-French) that can tolerate high flow rates without whipping in the artery or unwinding are used. The catheter is advanced to the right atrium and then through the heart selectively into a right or left pulmonary artery. Several techniques can be used to catheterize the pulmonary arteries, including the use of a deflecting wire to direct a standard pigtail catheter through the valves or the manipulation of a preshaped catheter ( Figs. 8-10 and 8-11 ). Ventricular arrhythmias are common because the catheter is manipulated through the right ventricle, but they are almost always self-limited. When elevated right ventricular pressures are suspected, a pressure measurement should be obtained in this location. A pressure measurement should always be obtained in the pulmonary artery before contrast injection. Careful and frequent flushing of the catheter prevents thrombus formation in the lumen, which, if injected into the lung during the study, mimics small emboli originating from peripheral veins. In patients with pulmonary arteriovenous fistulas or malformations, small thrombi or air bubbles could result in a stroke. The main concern related to contrast injection in the pulmonary artery is causing acute right heart failure due to increase afterload. Normal contrast injection rates are 25-30 mL/sec for 2 seconds (see Fig. 8-1 ). Patients with chronic pulmonary artery hypertension are often well compensated and may not require alteration in contrast injection parameters. In patients with uncompensated pulmonary hypertension, the rate and volume of contrast should be reduced to avoid acute right heart failure, sometimes to less than half of the volumes described. There are no generally agreed upon or validated guidelines for contrast reduction; observation of a small test dose (8-10 mL) of contrast injected by hand can be very helpful. Brisk flow, even in the setting of extremely elevated pressures, indicates adequate right ventricular function and the ability to tolerate normal injection rates. Slow flow and delayed washout of test contrast in the pulmonary arteries in the setting of a normal or elevated heart rate suggests a severely impaired right ventricle, and contrast volumes should be reduced appropriately.

The choice of imaging projection depends on the clinical question, but usually an anteroposterior (AP) view of the entire lung and a posterior oblique magnification view of the lung base with the catheter advanced beyond the upper lobe branches is sufficient for pulmonary embolism studies (see Fig. 8-1 ). When basilar lung segments are atelectatic, the anterior oblique view may provide the best view of the lower lobe pulmonary arteries. Rapid filming (at least four frames per second) is necessary. Imaging should be continued into the venous phase to evaluate the pulmonary veins (see Figs. 8-2 and 8-3 ). Diluting contrast containing more than 30% iodine by one third with saline reduces burnout against aerated lung in some digital subtraction angiograph systems. In rare instances, a balloon occlusion catheter may be necessary to isolate a pulmonary artery segment to image a peripheral embolus or high-flow abnormality ( Fig. 8-12 ).

There is a lot of lore about the proper way to remove a pigtail catheter from the pulmonary artery. The theoretical concern is that the catheter will become entangled in valve elements and will either tear them or become trapped. The most conservative approach is to open the pigtail with a soft-tipped guidewire in the pulmonary artery before removing it. Gentle withdrawal of the closed pigtail while observing under fluoroscopy works well. The pigtail frequently opens while exiting through the tricuspid valve, with no sequelae. Do not spin the pigtail, especially if it seems to get stuck, because it may indeed become entangled valve chordae. Either push it back into the heart or insert a soft straight guidewire to open the pig.

Overall complication rates of pulmonary angiography are less than 3%, with the majority related to access site issues or contrast-induced nephropathy. The mortality rate is less than 1% when the procedure is performed for suspected thromboembolic disease, with death occurring more commonly in critically ill patients with acute right heart strain, pulmonary hypertension, or elevated cardiac troponins due to the thromboembolic event ( Box 8-2 ).

Box 8-2

Risk Factors for Pulmonary Angiography

• Complete left bundle branch block

• Severe uncompensated right heart failure

• Severe pulmonary hypertension

• Acute myocardial infarction

• Pulmonary edema

• History of anaphylaxis to iodinated contrast

Right ventricular overload is avoided by reduction of contrast rate and volume in patients with acutely elevated pulmonary artery pressures. Once induced, right ventricular decompensation is difficult to reverse.

Acute Pulmonary Embolism

Acute thrombotic pulmonary embolism (PE) occurs when thrombus that has formed in the systemic veins breaks free and is carried by the venous return to the heart and, ultimately, the lungs ( Box 8-3 ). In situ formation of thrombus in the pulmonary arteries is rare and usually related to a surgical procedure, proximal obstruction, or pulmonary artery tumor. The actual incidence of PE is not known but is suspected to be more than 600,000 cases per year in the United States. Untreated, it is thought that the mortality rate of acute PE is 30%.

The clinical effects and prognosis of PE are dependent upon the following factors: degree and level of obstruction; baseline conditions of the pulmonary vasculature and lung parenchyma; and the status of the heart ( Table 8-1 ). Healthy patients with normal lungs and hearts can tolerate massive pulmonary emboli, whereas older patients with end-stage pulmonary diseases may succumb to relatively small emboli. In general, large emboli lodge in the central pulmonary arteries and present with cardiopulmonary collapse, elevated right heart pressures, and hypoxia with evidence of poor oxygen exchange (the so-called “death embolus”) ( Fig. 8-13 ). Small emboli lodge in peripheral pulmonary arteries and cause infarcts, which result in pleuritic chest pain and tachypnea but stable hemodynamic parameters and normal oxygenation ( Fig. 8-14 ). Subsegmental emboli tend to be multiple, bilateral, and in the lower lobes.

Lower extremity deep vein thrombosis (DVT) is present in up to 75% of patients with documented PE, but is often asymptomatic. Conversely, 50% of patients with symptomatic DVT have abnormal ventilation/perfusion ( $\dot{\text{V}}/\dot{\text{Q}}$ ) scans that are consistent with PE, despite the lack of pulmonary symptoms.

The clinical diagnosis of PE remains elusive, with several models proposed that incorporate symptoms, clinical parameters, and history to arrive at a probability of PE. The D-dimer test provides additional information when negative in patients with a low to moderate probability of PE. However, no scoring system has proven sufficient to obviate imaging evaluation.

Suspicion of acute PE is the most common indication for imaging the pulmonary vasculature. Plain radiographs are useful only to identify alternate explanations for the patient’s pulmonary symptoms and as a correlate for other studies. Peripheral wedge-shaped infiltrates can be seen in pulmonary infarcts, but emboli by themselves are not visible.

The $\dot{\text{V}}/\dot{\text{Q}}$ scan was once commonly used in the evaluation of patients with suspected PE and still retains some utility. Unlike other imaging tests, the $\dot{\text{V}}/\dot{\text{Q}}$ scan compares pulmonary perfusion with pulmonary ventilation. Ventilation of nonperfused areas of lung that are normal on chest radiograph allows a presumptive diagnosis of pulmonary artery obstruction. Conversely, PE does not explain perfusion of areas of nonventilated lung. This modality was validated in a landmark multicenter trial (Prospective Investigation of Pulmonary Embolism Diagnosis, or PIOPED) that comprised 933 patients and compared $\dot{\text{V}}/\dot{\text{Q}}$ to pulmonary angiography. The PIOPED II trial, which compared pulmonary CTA, CT venography of the pelvis and lower extremities, $\dot{\text{V}}/\dot{\text{Q}}$ scan, and compression venous ultrasound of the lower extremities suggested that perfusion scans can be interpreted without the ventilation component in patients with normal chest radiographs with an accuracy equivalent to pulmonary CTA. Uncooperative patients with obvious chest radiograph abnormalities are poor candidates for $\dot{\text{V}}/\dot{\text{Q}}$ scans because perfusion abnormalities may result from pneumonia, atelectasis, and other nonembolic causes.

Refer to Nuclear Medicine: The Requisites for a complete description of $\dot{\text{V}}/\dot{\text{Q}}$ scan interpretation ( Box 8-4 ). In summary, the results of $\dot{\text{V}}/\dot{\text{Q}}$ scans are expressed in terms of probability of pulmonary embolism. High probability scans have sensitivity of 77% for detection of PE, and low probability or normal scans have a specificity of 98% (no PE). However, a definitive diagnosis or exclusion of PE based on $\dot{\text{V}}/\dot{\text{Q}}$ scanning alone is possible in about 50% of patients. For this reason, a number of strategies have been proposed to enhance the diagnosis of PE using $\dot{\text{V}}/\dot{\text{Q}}$ scans, including combining the results with lower extremity venous ultrasound and clinical probability scores.