Chapter 496. Cardiac Catherization

The role of cardiac catheterization for children with congenital heart disease has changed dramatically over the past 10 years. Improved anatomic imaging of complex congenital heart disease with echocardiography, computerized tomography (CT), and magnetic resonance imaging (MRI) has made catheter-based angiography nearly obsolete for many common conditions. Today, most surgical patients with intracardiac lesions are diagnosed by echocardiography and undergo surgery without cardiac catheterization. MRI or CT has supplanted cardiac catheterization as the preoperative imaging modality for aortic arch abnormalities and coronary anomalies in many institutions, and MRI is the primary modality for measuring ventricular volumes and semilunar valve insufficiency. Transesophageal echocardiography supplements direct visualization during surgery and identifies residual abnormalities that can be promptly addressed intraoperatively, reducing the need for postoperative cardiac catheterization. Cardiac catheterization remains the primary test for evaluating complex physiology in children with abnormal pulmonary vascular resistance and reactivity, complex single ventricle anatomy, multiple obstructions in the right or left heart, or lesions of peripheral pulmonary arteries not seen well with other imaging modalities. Patients may need cardiac catheterization to assess residual defects soon after surgery when surface echo-Doppler studies may be less accurate.

Cardiac catheterization can rarely lead to serious complications, including arrhythmias, arterial obstruction, reactions to contrast medium, hemorrhage, cardiac perforation, hypoxemic episodes, infections, and death. The mortality rate is 0.2%, with the highest risk occurring in premature infants, critically ill term neonates, and those patients undergoing complex interventional catheterization procedures. Older children at particular risk of death are those with a very high pulmonary vascular resistance and no means of shunting. Pulmonary hypertensive crises or vagal episodes may decrease systemic output and cause death during or soon after catheterization in such patients. About 3% of children may have significant but nonfatal complications. Radiation exposure is a concern, especially in children with complex disease requiring repeat catheterizations. However, the few longitudinal studies done in children after cardiac catheterization in childhood have not shown an increased rate of cancer in adult years.

The most commonly used vessels for cardiac catheterization are the femoral, internal jugular, and subclavian veins and the femoral artery. The catheters are inserted percutaneously. Local infections and arterial complications are extremely rare, and the same vessels may be used repeatedly. In patients with congenital heart disease, the femoral approach often permits passage of a venous catheter into the left side of the heart through a patent foramen ovale or atrial septal defect, avoiding the use of the artery. Patients who have many repeat catheterizations may develop obstruction of the femoral or internal jugular veins, prohibiting their future use. If needed, vascular access can be safely obtained through percutaneous cannulation of the hepatic veins, called a “transhepatic approach” (eFig. 496.1). Some centers use carotid artery cut down with direct sheath placement and surgical repair for access for balloon valvuloplasty for critical aortic valve stenosis or for stent placement in the patent ductus arteriosus in a neonate with pulmonary atresia.

Catheter manipulation is performed under fluoroscopy, with image intensification and pulsed radiation to reduce radiation exposure. Catheters, long thin plastic tubes with one or more holes at the end, come in many sizes, shapes, and materials. Angiographic catheters have multiple holes at their tip to allow rapid injection of contrast medium for angiography. Directional catheters come in many preformed shapes, have a single end hole, and are made out of stiff materials to provide excellent control (ie, “steerability”) to facilitate manipulation into specific chambers and vessels. Balloon-tipped catheters have soft shafts and, with the balloon inflated, tend to follow the course of blood flow (eFig. 496.2).

Measurements during the catheterization include oxygen saturations and pressures in the cardiac chambers and great vessels. This information is used to calculate blood flows, intracardiac shunts, and vascular resistances. Angiography in one or more chambers or vessels is then done to define the cardiac anatomy.

Oxygen Saturations

A small blood sample for oximetry is taken from each of the major vessels and cardiac chambers. Saturations are currently measured by spectrophotometry based on the difference between the maximal wavelength absorption of hemoglobin bound to oxygen (600 nm) versus deoxygenated hemoglobin (506.5 nm). Shunts are diagnosed by comparing oxygen saturations in great veins, chambers, and great vessels. This information must be interpreted with caution, as streaming may influence saturations, especially in the vena cavae and right atrium. For example, hepatic venous blood with medium saturation mixes poorly with highly saturated renal streams and less saturated femoral venous streams in the inferior vena cava. The lowest oxygen concentrations are found in coronary sinus blood, and because the coronary sinus enters just above the tricuspid valve, a sample in the right ventricle may be a little less saturated than one in the right atrium. Normally saturations should vary little through the right heart, less than 4%, so any increase in saturation of greater than 5% indicates a significant left to right shunt. An increase in oxygen saturation at a given level may be the result of a shunt at that level, a shunt at a more proximal level with streaming, or a shunt at a more distal level with regurgitation.

Although noninvasive methods, including echocardiography with color Doppler and nuclear medicine shunt studies, can be used to quantify cardiac shunts, calculation from oxygen saturation data remains the gold standard. Saturation data may be of limited use in quantifying multiple cardiac shunts. A patient with a large left-to-right atrial shunt has blood with very high oxygen saturation entering the right ventricle. A second shunt of equal magnitude at the ventricular level produces a relatively small additional increase in oxygen saturation. Consequently, multiple cardiac shunts are best delineated by using both oxygen saturation data and angiography.

Normal right-sided oxygen saturations range between 65% and 80%, depending on cardiac output and hemoglobin concentration. In patients breathing room air, left-sided saturations are usually 95% to 98%. This mild left-sided desaturation results from normal right-to-left shunting of desaturated pulmonary bronchial blood into the pulmonary veins.

Flows, Shunts, and Resistances

Flows and shunts may be calculated by the Fick (oxygen content) method or by indicator dilution techniques that today are done with room temperature saline and thermistor catheters. The thermodilution technique is accurate and reliable except in patients with significant pulmonary or tricuspid regurgitation, or a significant intracardiac shunt. This technique is therefore primarily used in heart transplant patients, those with isolated pulmonary hypertension or isolated obstructive lesions, such as aortic stenosis or coarctation. Because shunts and regurgitation are common in congenital heart disease, the Fick method, using oxygen saturation and blood hemoglobin concentration, is much more widely used. It makes use of the fact that flow through an organ can be estimated by measuring the concentration of a substance (indicator) in arterial blood flowing to that organ, the amount of that substance added to or removed from blood as it passes through that organ, and the concentration of the substance in venous blood leaving that organ. For example, if blood entering the lung has an oxygen content of 150 mL/L, and blood leaving the lung has an oxygen content of 200 mL/L, then each liter of blood passing through the lungs picks up 50 mL of oxygen. If the oxygen consumption is 200 mL/min, then each minute 200 mL of oxygen is taken up by the lungs and carried away by 200/50 = 4 L of blood. Hence the pulmonary flow (Q̇p) is 4 L/min, calculated as:

Where Q̇p is pulmonary flow (L/min), V̇O2 is oxygen consumption (mL/min), Cpv is pulmonary venous oxygen content (mL/L of blood), and Cpa is pulmonary arterial oxygen content (mL/L of blood).

Note that the calculation uses oxygen content, not oxygen saturation. Oxygen saturation refers to the percentage of hemoglobin in blood that is combined with oxygen, and is independent of hemoglobin concentration. In contrast, oxygen content refers to the total amount of oxygen in a volume of blood and includes physically dissolved oxygen in addition to that bound to hemoglobin. It is expressed as either milliliters of oxygen per deciliter (dL or 100 mL) or liter of blood or as volume percent. Clearly, the oxygen content of blood with 50% oxygen saturation and a hemoglobin concentration of 20 g/dL is nearly twice that of blood with 50% oxygen saturation and a hemoglobin concentration of 10 g/dL.

In patients breathing room air, the physically dissolved oxygen (0.3 mL/dL) is negligible in comparison with the oxygen bound to hemoglobin (13.6 mL/dL at 100% saturation, even in an anemic patient with a hemoglobin of 10g/dL). The oxygen capacity, or the total content of oxygen bound to hemoglobin when 100% saturated, is calculated by multiplying the patient’s hemoglobin concentration in gm/dL by the constant 1.36 mL O2/gm of hemoglobin times 10 to convert dL to liters. A close approximation of the arterial-venous (AV) difference in oxygen content may be obtained by multiplying the arterial-venous difference in oxygen saturation by the oxygen capacity. Hence,

where satPV is pulmonary venous oxygen saturation expressed as a fraction of 1, and satpa is pulmonary arterial oxygen saturation, and capacity = Hgb (in gm/dl) × 1.36 × 10.

With no right-to-left shunting, an aortic sample is preferable to a pulmonary venous sample because it is a mixture of blood from all four pulmonary veins. Similarly, with no left-to-right shunting, a pulmonary arterial sample is the best representative mixture of systemic venous blood.

In a steady state, the amount of oxygen utilized in the tissues is equal to the amount taken into the lungs, and systemic flow is calculated with the same equation and oxygen consumption. Without shunts the systemic arterial-venous (AV) oxygen difference is the same as the pulmonary AV oxygen difference, and the calculated pulmonary flow (Q̇p) equals the systemic flow (Q̇s) or cardiac output. If there are shunts, then pulmonary and systemic flows must be calculated separately, the former from oxygen contents in pulmonary veins and pulmonary artery, the latter from aortic and mixed systemic venous blood.

Flow and shunt calculations are shown in eFigures 496.2, 496.3, 496.4, and 496.5. In each, the oxygen consumption (V̇O2) is assumed to be 150 mL/min/m2, and the hemoglobin (Hb) concentration is assumed to be 14.7 g/dL blood or 147 g/L of blood. Because 1 g of Hb combines with 1.36 mL O2, then

In the equations, saturations are given as fractions of unity. Thus, 96% is given as 0.96.

Example 1: Left‐to-Right Shunt

The most distal saturation that is proximal to the left-to-right shunt is used in the calculation, as it is usually the best-mixed venous sample available. Left-to-right shunt = Q̇p – Q̇s = 3.75 L/min/m2.

Example 2: Right‐to-Left Shunt

Right-to-left shunt = Qs – Qp = 1.25 L/min/m2.

Example 3: Bidirectional Shunting

Calculation of bidirectional shunts requires the concept of effective pulmonary flow (Qep), defined as the volume of systemic venous blood per minute that flows through the lungs. This volume does not include blood that is shunted in either direction because shunted blood is not effective; it neither picks up oxygen nor delivers it. In a patient with only a left-to-right shunt, effective flow equals systemic flow (Q̇p – Q̇ep the left-to-right shunt), and in a patient with only a right-to-left shunt, effective flow equals pulmonary flow (Q̇s – Q̇ep is the right-to-left shunt). In patients with bidirectional shunting, the arteriovenous difference used to calculate effective pulmonary flow uses the mixed systemic venous and mixed pulmonary venous oxygen contents. So effective pulmonary flow is calculated by

Left-to-right shunt Q̇p– Q̇ep = 0.75 L/min/m2; right-to-left shunt = Q̇s – Q̇ep = 2.0 L/min/m2.

Pressures are recorded in each chamber and vessel entered and are monitored continuously during catheter manipulation to assist in determining catheter position. The volume of blood in a chamber or vessel, the size of the chamber or vessel, and the compliance of the walls will determine the pressure. Abnormalities in cardiac pressures can be used to diagnose congenital cardiac disease, so an understanding of normal physiology is key to recognizing abnormalities. Pressure signals from the transducers are sent to a recording device that displays them on a monitor and records them digitally for future reference. Normal pressures and common abnormalities are shown in Table 496-1.

Atrial Pressures

Atrial pressures are reported as an a wave, v wave, and mean pressure. Although there is a c wave and an X and Y descent, they are rarely used to detect clinical abnormalities. The a wave is an increase in pressure that occurs at the end of ventricular diastole generated by contraction of the atrium so it starts with the P wave of the EKG. The a wave peaks as the atrium completes contraction then the pressure decreases as the atrioventricular (AV) valves close with ventricular systole (eFig. 496.6). Atrial pressure continues to decrease during early ventricular systole, the X descent, then increases in late ventricular systole creating the v wave as forward flow fills the atrium against a closed AV valve. As the AV valves open at the start of ventricular diastole the atrial pressure falls, Y descent, resulting in early rapid filling of the ventricles in diastole.

Elevated atrial pressures signify increased intravascular volume, decreased compliance or hypoplasia of the atrium or decreased compliance of the downstream ventricle. Low atrial pressures typically indicate hypovolemia. Elevated a and v wave pressures indicate obstruction to emptying of the atria during diastole, either tricuspid or mitral stenosis or decreased ventricular compliance, a “stiff ventricle” takes more pressure to fill (eFig. 496.7). Mitral or tricuspid insufficiency will increase the v wave and shift the peak earlier in the cardiac cycle, sometimes termed a “regurgitant wave” (eFig. 496.8). All of these relationships are dependent on normal sinus rhythm. If there is AV block then this pattern will be disturbed and the ventricle will at times contract with an open AV valve resulting in a cannon a wave (eFig. 496.9).

Ventricular Pressures

Ventricular pressures are reported as end diastolic and systolic; no mean pressure is reported. In addition, the minimal ventricular pressure and the contour of the pressure tracing can be useful in detecting cardiac abnormalities. Normal ventricular pressure ranges are shown in Table 496-1. End diastolic pressure occurs just before systole begins and should be similar to the peak a wave atrial pressure. If the a wave is greater this indicates AV valve stenosis. Elevation of the end diastolic pressure indicates decreased compliance of the ventricle commonly due to muscular hypertrophy or diastolic dysfunction. It also occurs if there is restriction or constriction as with pericardial tamponade. The normal ventricular systolic pressure trace has a square shape. Elevation of ventricular systolic pressure with a late-peaked shape indicates dynamic ventricular outflow tract obstruction (sub-AS or sub-PS) (eFig. 496.10A,B). Although not typically reported, the minimal pressure the ventricle reaches during the end of isovolumetric relaxation should be below 4 mmHg. The ventricle actively relaxes against closed AV and semilunar (aortic or pulmonary) valves generating negative pressure until the AV valve opens and the ventricle begins to fill. Elevated minimal ventricular pressures and a rapid rise to end diastolic pressures (square root sign) indicate an abnormality with relaxation either due to diastolic dysfunction such as myocardial restriction or external limitation such as pericardial constriction (eFig. 496.11).

Arterial Pressures

Normal pulmonary artery and aortic systolic, diastolic, and mean pressures are reported in Table 496-1. At peak systole the pressure in the ventricle proximal to the artery should be identical to that in the artery. Elevation of the ventricular pressure suggests aortic or pulmonary valve stenosis (eFig. 496.12). Elevation of the arterial systolic pressure suggests downstream obstruction, either large vessel stenosis such as coarctation (eFig. 496.13A,B) or pulmonary artery stenosis or small vessel vascular resistance elevation (pulmonary or systemic hypertension). Elevated arterial diastolic pressure suggests limited distal runoff due to elevated vascular resistance. Low diastolic pressures on the other hand suggest brisk distal runoff from low vascular resistance or some abnormal run off such as aortic insufficiency, patent ductus arteriosus, aortic window, or systemic arteriovenous malformations in the systemic circulation or pulmonary insufficiency or pulmonary arteriovenous malformations in the pulmonary circulation (eFig. 496.14A,B).

Vascular Resistance

The resistance to blood flow in the pulmonary and systemic circulations is calculated from the equations:

where PA and PV are pulmonary artery and vein, respectively, LA and RA are left and right atria, respectively, Ao is aorta, RP is pulmonary vascular resistance (mm Hg/L per min/ m2commonly known as resistance units/m2 or Wood units), Q̇p is pulmonary flow (L/min/m2), RS is systemic vascular resistance (resistance units/m2), and Q̇s is systemic flow (L/min/m2). Most of the resistance to flow is at the arteriolar level, and the calculated resistance is inversely related to the cross-sectional area of the arteriolar lumen. In the systemic circulation, the calculated resistance is the combined value across several vascular beds in parallel. The normal systemic vascular resistance varies between 15 and 30 units/m2. Although high at birth, the pulmonary vascular resistance reaches values near low adult values after 2 to 4 weeks. So that these values can be compared at different ages, they are usually related to surface area. Normal values in children and adults are 1 to 3 units/m2. Elevation to 12 units/m2 suggests mild to moderate pulmonary vascular disease. Marked elevation to greater than 12 units/m2 often suggests significantly elevated pulmonary vascular resistance, which is unlikely to return to normal with vasodilators (“fixed” pulmonary vascular disease). When pulmonary vascular resistance is elevated patients are given inhaled pulmonary vasodilators (100 % oxygen, 40 ppm nitric oxide, or iloprost 2.5 to 5 mcg) in the catheterization laboratory and repeat pulmonary vascular resistance measured to determine vascular reactivity and the minimal pulmonary resistance. This information is often used to determine suitability for surgical repair or heart transplant, plan postoperative management, or to guide outpatient vasodilator therapy.

The cross-sectional area of a valve or vessel most accurately reflects the degree of obstruction. Although pressure gradient is commonly used as a measure of obstruction, it is determined by the cross-sectional area of the obstruction and the blood flow through the obstruction. In patients with valvar, aortic, or mitral stenosis, the effective valve area can be calculated from stroke volume and pressure measurements to more accurately assess the degree of obstruction. Gorlin and Gorlin in 19511 described a formula for calculating valve area that continues in use today. Flow through a valve only occurs during a portion of the cardiac cycle (for example, during systole for the aortic valve and during diastole for the mitral valve), so this must be considered when calculating valve areas. Systolic ejection time is measured from simultaneous left ventricular (LV) and aortic pressure tracings by determining the duration in seconds that the LV pressure exceeds that of the aortic trace in systole. Similarly, the diastolic filling time is measured from simultaneous left atrial (LA) (or pulmonary capillary wedge) and LV pressure tracings by determining the duration in seconds that the LV pressure exceeds that of the LA trace in diastole. The formulas for valve areas are listed below:

The mean gradient across a valve is most accurately determined by the area between the pressure tracings during ejection of blood through the valve. For the aortic valve this would be the area between the LV and aortic pressure tracings during systole, and for the mitral this would be the area between the LA and LV tracing during diastole. Current catheterization laboratory computerized hemodynamic systems easily, accurately, and automatically calculate these values (eFig. 496.15). Normal aortic and mitral valve areas are given in Table 496-1.

Positioning a catheter in a particular chamber and rapidly injecting iodinated contrast permits radiographic delineation of most cardiovascular structures. Multiple injection sites and angulation of the cameras are often necessary to obtain a complete anatomic diagnosis.

Angiograms are recorded digitally at 15, 30, or 60 frames per second. Biplane angiograms yield more information than single-plane angiograms while using the same volume of contrast material, although with greater radiation exposure. This is important for patients with complicated abnormalities who require multiple angiograms and in newborns who may have a low tolerance to contrast medium. A digital replay system permits immediate examination of angiograms, as well as a roadmap for further catheter manipulation. Figure 496-1A,B shows a right ventricular angiogram in the right anterior oblique (ROA) and lateral projections with Figure 496-2A,B shows a left ventricular injection in a RAO and LAO projection.

Additional Testing

Additional tests may be done during the procedure. Drug testing with isoproterenol or dobutamine infusion to simulate exercise can be useful in patients with mild valvar, subvalvar, or aortic obstruction who have symptoms or hypertrophy out of proportion to their resting gradients (eFig. 496.10A,B). Giving oxygen, nitric oxide, or iloprost to determine pulmonary vascular reactivity is common.

Endomyocardial biopsy is the gold standard for detecting rejection in cardiac transplant patients. Nearly all institutions screen their pediatric transplant patients with cardiac biopsies at regular intervals, although this is less common for neonates. Other, more controversial, indications for biopsy include new-onset cardiomyopathy and suspected myocarditis. Though in the past it was thought useful to biopsy both ventricles, right ventricular biopsy alone is now standard practice.1-10