Myocardial ischaemia

Ischaemia of the heart results from an imbalance between myocardial oxygen supply and demand, producing pain called angina (Boxes 11.2 and 11.3). Angina is usually a symptom of atherosclerotic coronary artery disease, which impedes myocardial oxygen supply. Other causes of coronary artery disease (Box 11.4) are rare. However, it is important to be vigilant for causes of angina due to increased myocardial oxygen demand, such as aortic stenosis. The history is diagnostic for angina if the location of the pain, its character, its relation to exertion and its duration are typical. The patient describes retrosternal pain which may radiate into the arms, the throat or the jaw. It has a constricting character, is provoked by exertion and relieved within minutes by rest. The patient’s threshold for angina is typically reduced after eating or in cold weather due to the diversion of blood to the gut and the increased myocardial work consequent upon peripheral vasoconstriction, respectively. Occasionally angina is provoked only by the first significant activity of the day, a phenomenon known as the ‘warm-up effect’ and due to myocardial preconditioning. Less commonly, myocardial ischaemia may manifest as breathlessness, fatigue or symptoms that the patient finds difficult to describe – ‘I just have to stop’ – in which case the clues to the diagnosis are the relation of symptoms to exertion, the presence of risk factors for coronary artery disease and the absence of an alternative explanation for the symptoms, such as heart failure.

Acute coronary syndromes

In these life-threatening cardiac emergencies, the pain is similar in location and character to angina but is usually more severe, more prolonged and unrelieved by rest (Box 11.5).

Pericarditis

This causes central chest pain, which is sharp in character and aggravated by deep inspiration, cough or postural changes. Characteristically, the pain is exacerbated by lying recumbent and reduced by sitting forward. Pericarditis is usually idiopathic or caused by Coxsackie B infection. It may also occur as a complication of myocardial infarction, but other causes are seen less commonly (Box 11.6).

Aortic dissection

This produces severe tearing pain in either the front or the back of the chest. The onset is abrupt, unlike the crescendo quality of ischaemic cardiac pain (Box 11.7).

Pulmonary embolism

Peripheral pulmonary embolism causes sudden-onset sharp, pleuritic chest pain, breathlessness and haemoptysis. Major, central pulmonary embolism presents with breathlessness, chest pain that can be indistinguishable from ischaemic chest pains and syncope. Risk factors for pulmonary embolism should be sought in the history (Box 11.8).

Rare cardiovascular causes of chest pain include mitral valve disease associated with massive left atrial dilatation. This causes discomfort in the back, sometimes associated with dysphagia due to oesophageal compression. Aortic aneurysms can also cause pain in the chest owing to local compression.

Anaemia

This may exacerbate angina and heart failure. Pallor of the mucous membranes is a useful but sometimes misleading physical sign, and diagnosis requires laboratory measurement of the haemoglobin concentration.

Cyanosis

This is a blue discoloration of the skin and mucous membranes caused by increased concentration of reduced haemoglobin in the superficial blood vessels. Peripheral cyanosis may result when cutaneous vasoconstriction slows the blood flow and increases oxygen extraction in the skin and the lips. It is physiological during cold exposure. It also occurs in heart failure, when reduced cardiac output produces reflex cutaneous vasoconstriction. In mitral stenosis, cyanosis over the malar area produces the characteristic mitral facies or malar flush.

Central cyanosis may result from the reduced arterial oxygen saturation caused by cardiac or pulmonary disease. It affects not only the skin and the lips but also the mucous membranes of the mouth. Cardiac causes include pulmonary oedema (which prevents adequate oxygenation of the blood) and congenital heart disease. Congenital defects associated with central cyanosis include those in which desaturated venous blood bypasses the lungs by (‘reversed’) shunting through septal defects or a patent ductus arteriosus (e.g. Eisenmenger’s syndrome, Fallot’s tetralogy).

Clubbing of the fingers and toes

In congenital cyanotic heart disease, clubbing is not present at birth but develops during infancy and may become very marked. Infective endocarditis is the only other cardiac cause of clubbing.

Other cutaneous and ocular signs of infective endocarditis

These are caused by immune complex deposition in the capillary circulation. A vasculitic rash is common, as are splinter haemorrhages in the nail bed, although these are a very non-specific finding. Other ‘classic’ manifestations of endocarditis including Osler’s nodes (tender erythematous nodules in the pulps of the fingers), Janeway lesions (painless erythematous lesions on the palms) and Roth’s spots (erythematous lesions in the optic fundi) are now rarely seen.

Coldness of the extremities

In patients hospitalized with severe heart failure, this is an important sign of reduced cardiac output. It is caused by reflex vasoconstriction of the cutaneous bed.

Pyrexia

Infective endocarditis is invariably associated with pyrexia, which may be low grade or ‘swinging’ in nature if paravalvular abscess develops. Pyrexia also occurs for the first 3 days after myocardial infarction.

Oedema

Subcutaneous oedema that pits on digital pressure is a cardinal feature of congestive heart failure. Pressure should be applied over a bony prominence (tibia, lateral malleoli, sacrum) to provide effective compression. Oedema is caused by salt and water retention by the kidney. Two mechanisms are responsible:

Salt and water retention expands plasma volume and increases the capillary hydrostatic pressure. Hydrostatic forces driving fluid out of the capillary exceed the osmotic forces reabsorbing it, so that oedema fluid accumulates in the interstitial space. The effect of gravity on capillary hydrostatic pressure ensures that oedema is most prominent around the ankles in the ambulant patient and over the sacrum in the bedridden patient. In advanced heart failure, oedema may involve the legs, genitalia and trunk. Transudation into the peritoneal cavity (ascites), the pleural and pericardial spaces may also occur.

Rate and rhythm

By convention, both are assessed by palpating the right radial pulse. Rate, expressed in beats per minute (bpm), is measured by counting the number of beats in a timed period of 15 seconds and multiplying by four. Normal sinus rhythm is regular, but in young patients may show phasic variation in rate during respiration (sinus arrhythmia). An irregular rhythm usually indicates atrial fibrillation, but may also be caused by frequent ectopic beats or self-limiting paroxysmal arrhythmias. In patients with atrial fibrillation, the rate should be measured by auscultation at the cardiac apex, because beats that follow very short diastolic intervals may create a ‘pulse deficit’ by not generating sufficient pressure to be palpable at the radial artery.

Character

This is defined by the volume and waveform of the pulse and should be evaluated at the right carotid artery (i.e. the pulse closest to the heart and least subject to damping and distortion in the arterial tree). Pulse volume provides a crude indication of stroke volume, being small in heart failure and large in aortic regurgitation. The waveform of the pulse is of greater diagnostic importance (Fig. 11.3). Severe aortic stenosis produces a slow-rising carotid pulse; the fixed obstruction restricts the rate at which blood can be ejected from the left ventricle. In aortic regurgitation, in diastole, the left ventricle receives not only its normal pulmonary venous return but also a proportion of the blood ejected into the aorta during the previous systole as it flows back through an incompetent valve. The resultant large stroke volume, vigorously ejected, produces a rapidly rising carotid pulse, which collapses in early diastole owing to backflow through the aortic valve. This collapsing pulse can be exaggerated at the radial artery by lifting the arm. In mixed aortic valve disease, a biphasic pulse with two systolic peaks is occasionally found. Alternating pulse – alternating high and low systolic peaks – occurs in severe left ventricular failure but the mechanism for this is unknown. Paradoxical pulse refers to an inspiratory decline in systolic pressure greater than 10 mmHg (Fig. 11.4). In normal circumstances, inspiration results in an increase in venous return as blood is ‘sucked into’ the thorax by the decline in intrathoracic pressure. This increases right ventricular stroke volume, but left ventricular stroke volume falls slightly (ventricular interdependence). When the heart is constrained in a ‘fixed box’ by a pericardial effusion (cardiac tamponade) or by thickened pericardium (pericardial constriction), the increased inspiratory right ventricular blood volume reduces left ventricular compliance resulting in a more pronounced reduction in left ventricular filling, stroke volume and systolic blood pressure during inspiration. ‘Pulsus paradoxus’ therefore represents an exaggeration of the normal inspiratory decline in systolic pressure and is not truly paradoxical. Pulsus paradoxus in acute severe asthma is thought to be due to negative pleural pressure increasing afterload and thereby impedence to left ventricular emptying.

Figure 11.3 The waveform of the pulse is characterized by the rate of rise of the carotid upstroke. Note that in aortic regurgitation, the upstroke is rapid and followed by abrupt diastolic ‘collapse’. In hypertrophic cardiomyopathy, the upstroke is also rapid and the pulse has a jerky character. In aortic stenosis, the upstroke is slow with a plateau.

Figure 11.4 Paradoxical pulse (radial artery pressure signal). The patient had severe tamponade. Note the exaggerated (>10 mmHg) decline in arterial pressure during inspiration.

Symmetry

Symmetry of the radial, brachial, carotid, femoral, popliteal and pedal pulses should be confirmed. A reduced or absent pulse indicates an obstruction more proximally in the arterial tree, caused usually by atherosclerosis or thromboembolism, less commonly by aortic dissection. Coarctation of the aorta causes symmetrical reduction and delay of the femoral pulses compared with the radial pulses (‘radiofemoral delay’), a sign that should be looked for in younger patients with hypertension.

Jugular venous pressure

The jugular venous pressure (JVP) should be assessed from the waveform of the internal jugular vein which lies adjacent to the medial border of the sternocleidomastoid muscle. Distention of the external jugular vein is a useful clue to an elevated JVP but, strictly speaking, it should not be used because it can be compressed as it passes under the clavicle. The JVP is measured in centimetres vertically from the sternal angle to the top of the venous waveform. The normal upper limit is 4 cm. This is about 9 cm above the right atrium and corresponds to a pressure of 6 mmHg. Elevation of the JVP indicates a raised right atrial pressure unless the superior vena cava is obstructed, producing engorgement of the neck veins (Box 11.12). During inspiration, the pressure within the chest decreases and there is a fall in the JVP. In constrictive pericarditis, and less commonly in tamponade, inspiration produces a paradoxical rise in the JVP (Kussmaul’s sign) because the increased venous return that occurs during inspiration cannot be accommodated within the constrained right side of the heart (Fig. 11.5).

Figure 11.5 Kussmaul’s sign. Jugular venous pressure recording in a patient with tamponade. The venous pressure is raised and there is a particularly prominent systolic ‘x’ descent, giving the waveform of the JVP an unusually dynamic appearance. Note the inspiratory rise in atrial pressure (Kussmaul’s sign) reflecting the inability of the tamponaded right heart to accommodate the inspiratory increase in venous return.

Waveform of jugular venous pulses

In sinus rhythm, the jugular venous pulse has a double waveform attributable to the ‘a’ and ‘v’ waves separated by the ‘x’ and ‘y’ descents. The ‘a’ wave is produced by atrial systole. It is followed by the ‘x’ descent (marking descent of the tricuspid valve ring), which is interrupted by the diminutive ‘c’ wave caused by tricuspid valve closure. Atrial pressure then rises again, producing the ‘v’ wave as the atrium fills passively during ventricular systole. The decline in atrial pressure as the tricuspid valve opens to allow ventricular filling produces the ‘y’ descent. Important abnormalities of the pattern of deflections are shown in Figure 11.6.

Figure 11.6 Waveform of the jugular venous pulse. (A) The ECG is portrayed at the top of the illustration. Note how electrical events precede mechanical events in the cardiac cycle. Thus the P wave (atrial depolarization) and the QRS complex (ventricular depolarization) precede the ‘a’ and ‘v’ waves, respectively, of the JVP. (B) Normal JVP. The ‘a’ wave produced by atrial systole is the most prominent deflection. It is followed by the ‘x’ descent, interrupted by the small ‘c’ wave marking tricuspid valve closure. Atrial pressure then rises again (‘v’ wave) as the atrium fills passively during ventricular systole. The decline in atrial pressure as the tricuspid valve opens produces the ‘y’ descent. (C) Giant ‘a’ wave. Forceful atrial contraction against a stenosed tricuspid valve or a non-compliant hypertrophied right ventricle produces an unusually prominent ‘a’ wave. (D) Cannon ‘a’ wave. This is caused by atrial systole against a closed tricuspid valve. It occurs when atrial and ventricular rhythms are dissociated (complete heart block, ventricular tachycardia) and marks coincident atrial and ventricular systole. (E) Giant ‘v’ wave. This is an important sign of tricuspid regurgitation. The regurgitant jet produces pulsatile systolic waves in the JVP. (F) Prominent ‘x’ and ‘y’ descents. These occur in constrictive pericarditis and give the JVP an unusually dynamic appearance. In tamponade, only the ‘x’ descent is usually exaggerated.

In atrial fibrillation, there is no atrial contraction. Consequently, there is no ‘a’ wave and the jugular venous pulse loses its double waveform. It is not always easy to differentiate venous from arterial pulsations in the neck, but several features help to distinguish the jugular venous pulse from the carotid arterial pulse (Box 11.13).

First sound (S1)

This corresponds to mitral and tricuspid valve closure at the onset of systole. It is accentuated in mitral stenosis because prolonged diastolic filling through the narrowed valve ensures that the thickened leaflets are widely separated at the onset of systole. Thus valve closure generates unusually vigorous vibrations. In advanced mitral stenosis, the valve is rigid and immobile and S1 becomes soft again.

Second sound (S2)

This corresponds to aortic and pulmonary valve closure following ventricular ejection. S2 is single during expiration. Inspiration, however, causes physiological splitting into aortic followed by pulmonary components because increased venous return to the right side of the heart delays pulmonary valve closure. Important abnormalities of S2 are illustrated in Figure 11.7.

Figure 11.7 Splitting of the second heart sound. The first sound, representing mitral and tricuspid closure, is usually single, but the aortic and pulmonary components of the second sound normally split during inspiration as increased venous return delays right ventricular emptying. Abnormal splitting of the second heart sound is an important sign of heart disease.

Third and fourth sounds (S3, S4)

These low-frequency sounds occur early and late in diastole, respectively. When present, they give a characteristic ‘gallop’ to the cardiac rhythm. Both sounds are best heard with the bell of the stethoscope at the cardiac apex. They are caused by abrupt tensing of the ventricular walls following rapid diastolic filling. Rapid filling occurs early in diastole (S3) following atrioventricular valve opening, and again late in diastole (S4) due to atrial contraction. S3 is physiological in children and young adults but usually disappears after the age of 40. It also occurs in high-output states caused by anaemia, fever, pregnancy and thyrotoxicosis. After the age of 40, S3 is nearly always pathological, usually indicating left ventricular failure or, less commonly, mitral regurgitation or constrictive pericarditis. In the elderly, S4 is sometimes physiological. More commonly, however, it is pathological, and occurs when vigorous atrial contraction late in diastole is required to augment filling of a hypertrophied, non-compliant ventricle (e.g. hypertension, aortic stenosis, hypertrophic cardiomyopathy).

Systolic clicks and opening snaps

Valve opening, unlike valve closure, is normally silent. In aortic stenosis, however, valve opening produces a click in early systole which precedes the ejection murmur. The click is only audible if the valve cusps are pliant and non-calcified, and is particularly prominent in the congenitally bicuspid valve. A click later in systole suggests mitral valve prolapse, particularly when followed by a murmur. In mitral stenosis, elevated left atrial pressure causes forceful opening of the thickened valve leaflets. This generates a snap early in diastole that precedes the mid-diastolic murmur.

Heart murmurs

These are caused by turbulent flow within the heart and great vessels (Fig. 11.8). Occasionally the turbulence is caused by increased flow through a normal valve – usually aortic or pulmonary – producing an ‘innocent’ murmur. However, murmurs may also indicate valve disease or abnormal communications between the left and right sides of the heart (e.g. septal defects).

Figure 11.8 Heart murmurs. These are caused by turbulent flow within the heart and great vessels, and may indicate valve disease. Heart murmurs may be depicted graphically as shown in this illustration. CM, continuous murmur; EC, ejection click; EDM, early diastolic murmur; LSM, late systolic murmur; MDM, mid-diastolic murmur; MSC, mid-systolic click; MSM, mid-systolic murmur; OS, opening snap; PSA, presystolic accentuation of murmur; PSM, pansystolic murmur; S3, third heart sound; S4, fourth heart sound. Parentheses indicate those auscultatory findings that are not constant.

Rheumatic heart disease has become much less common in developed countries, although it remains common elsewhere, and is the cause of many of the classic heart murmurs (Box 11.15). Degenerative valve disease (calcific aortic stenosis, mitral regurgitation due to chordal rupture) is increasingly common. Heart murmurs are defined by four characteristics: loudness, quality, location and timing.

The loudness of a murmur reflects the degree of turbulence. This relates to the volume and velocity of flow and not the severity of the cardiac lesion. Loudness is graded on a scale of 1 (barely audible) to 6 (audible even without application of the stethoscope to the chest wall). The quality of a murmur relates to its frequency and is best described as low, medium or high-pitched. The location of a murmur on the chest wall depends on its site of origin and has led to the description of four valve areas (see above). Some murmurs radiate, depending on the velocity and direction of blood flow. The sound of the high-velocity systolic flow in aortic stenosis and mitral regurgitation, for example, is directed towards the neck and the axilla, respectively; that of the high-velocity diastolic flow in aortic regurgitation is directed towards the left sternal edge. Murmurs are timed according to the phase of systole or diastole during which they are audible. It is inadequate to describe the timing of a murmur as systolic or diastolic without more specific reference to the length of the murmur and the phase of systole or diastole during which it is heard: systolic murmurs are either mid-systolic, pansystolic or late systolic; diastolic murmurs are either early diastolic, mid-diastolic or presystolic in timing. Continuous murmurs are audible in both phases of the cardiac cycle.

A mid-systolic (‘ejection’) murmur is caused by turbulence in the left or right ventricular outflow tract during ejection. It starts following opening of the aortic or pulmonary valve, reaches a crescendo in mid-systole and disappears before the second heart sound. The murmur is loudest in the aortic area (with radiation to the neck) when it arises from the left ventricular outflow tract, and in the pulmonary area when it arises from the right ventricular outflow tract. It is best heard with the diaphragm of the stethoscope while the patient sits forward. Important causes of aortic ejection murmurs are aortic stenosis and hypertrophic cardiomyopathy.

Aortic regurgitation also produces an ejection murmur due to increased stroke volume and velocity of ejection. Pulmonary ejection murmurs may be caused by pulmonary stenosis or infundibular stenosis (as in Fallot’s tetralogy).

In atrial septal defect, the pulmonary ejection murmur results from right ventricular volume loading and consequent increased blood flow through the pulmonary valve and does not indicate organic valvular disease. ‘Innocent’ murmurs unrelated to heart disease are always mid-systolic in timing and are caused by turbulent flow in the left (sometimes right) ventricular outflow tract. In most cases, there is no clear cause, but they may reflect a hyperkinetic circulation in conditions such as anaemia, pregnancy, thyrotoxicosis or fever. They are rarely louder than grade 3, often vary with posture, may disappear on exertion and are not associated with other signs of heart disease.

Pansystolic murmurs are audible throughout systole from the first to the second heart sounds. They are caused by regurgitation through incompetent atrioventricular valves and by ventricular septal defects. The pansystolic murmur of mitral regurgitation is loudest at the cardiac apex and radiates into the left axilla. It is best heard using the diaphragm of the stethoscope with the patient lying on the left side. The murmurs of tricuspid regurgitation and ventricular septal defect are loudest at the lower left sternal edge. Inspiration accentuates the murmur of tricuspid regurgitation because the increased venous return to the right side of the heart increases the regurgitant volume. Mitral valve prolapse may also produce a pansystolic murmur but, more commonly, prolapse occurs in mid-systole, producing a click followed by a late-systolic murmur.

Early diastolic murmurs are high-pitched and start immediately after the second heart sound, fading away in mid-diastole. They are caused by regurgitation through incompetent aortic and pulmonary valves and are best heard using the diaphragm of the stethoscope while the patient leans forward. The early diastolic murmur of aortic regurgitation radiates from the aortic area to the left sternal edge, where it is usually easier to hear, in maintained expiration with the patient leaning forward. Pulmonary regurgitation is loudest at the pulmonary area. Mid-diastolic murmurs are caused by turbulent flow through the atrioventricular valves. They start after valve opening, relatively late after the second sound, and continue for a variable period during mid-diastole. Mitral stenosis is the principal cause of a mid-diastolic murmur which is best heard at the cardiac apex using the bell of the stethoscope while the patient lies on the left side. Increased flow across a non-stenotic mitral valve occurs in ventricular septal defect and mitral regurgitation and may produce a mid-diastolic murmur. In severe aortic regurgitation, preclosure of the anterior leaflet of the mitral valve by the regurgitant jet may produce mitral turbulence associated with a mid-diastolic murmur (Austin Flint murmur). A mid-diastolic murmur at the lower left sternal edge, accentuated by inspiration, is caused by tricuspid stenosis and also by conditions that increase tricuspid flow (e.g. atrial septal defect, tricuspid regurgitation).

In mitral or tricuspid stenosis, atrial systole produces a presystolic murmur immediately before the first heart sound. The murmur is perceived as an accentuation of the mid-diastolic murmur associated with these conditions. Because presystolic murmurs are generated by atrial systole, they do not occur in patients with atrial fibrillation.

Continuous murmurs are heard during systole and diastole, and are uninterrupted by valve closure. The commonest cardiac cause is patent ductus arteriosus, in which flow from the high-pressure aorta to the low-pressure pulmonary artery continues throughout the cardiac cycle, producing a murmur over the base of the heart which, though continuously audible, is loudest at end systole and diminishes during diastole. Ruptured sinus of Valsalva aneurysm also produces a continuous murmur.

Friction rubs and venous hums

A friction rub occurs in pericarditis. It is best heard in maintained expiration with the patient leaning forward as a high-pitched scratching noise audible during any part of the cardiac cycle and over any part of the left precordium. A continuous venous hum at the base of the heart reflects hyperkinetic jugular venous flow. It is particularly common in infants and usually disappears on lying flat.

Generation of electrical activity

The stimulus for every normal ventricular contraction (sinus beat) begins with depolarization of an area of specialized conducting tissue high in the right atrium called the sinoatrial (SA) node. The depolarization spreads through the walls of the atria causing contraction of the atrial muscle before reaching another area of specialized conducting tissue in the lower part of the right atrium called the atrioventricular (AV) node. Conduction through the AV node is relatively slow which allows atrial contraction to be completed and the ventricles to fill before depolarization travels down the bundle of His and then into the left and right bundle branches. The left bundle branch divides further into the left anterior fascicle and the left posterior fascicle. From here, the depolarization spreads through the Purkinje fibres in the ventricular muscle which stimulates ventricular contraction. Once ventricular contraction has occurred, the muscle cells repolarize and the ventricles relax to allow ventricular filling to occur.

The wave of depolarization that spreads through the heart during each cardiac cycle has vector properties defined by its direction and magnitude. The net direction of the wave changes continuously during each cardiac cycle and the ECG deflections change accordingly, being positive as the wave approaches the recording electrode and negative as it moves away. Electrodes orientated along the axis of the wave record larger deflections than those orientated at right angles. Nevertheless, the size of the deflections is determined principally by the magnitude of the wave, which is a function of muscle mass. Thus the ECG deflection produced by depolarization of the atria (P wave) is smaller than that produced by the depolarization of the more muscular ventricles (QRS complex). Ventricular repolarization produces the T wave.

Inscription of the QRS complex

The ventricular depolarization vector can be resolved into two components:

Left ventricular depolarization dominates the second vector component, the resultant direction of which is from right to left. Thus electrodes orientated to the left ventricle record a small negative deflection (Q wave) as the septal depolarization vector moves away, followed by a large positive deflection (R wave) as the ventricular depolarization vector approaches. The sequence of deflections for electrodes orientated towards the right ventricle is in the opposite direction (Fig. 11.9).

Figure 11.9 Inscription of the QRS complex. The septal depolarization vector (1) produces the initial deflection of the QRS complex. The ventricular free-wall depolarization vector (2) produces the second deflection, which is usually more pronounced. Lead aVR is orientated towards the cavity of the left ventricle and records an entirely negative deflection.

Any positive deflection is termed an R wave. A negative deflection before the R wave is termed a Q wave (this must be the first deflection of the complex), whereas a negative deflection following the R wave is termed an S wave.

Electrical axis

Because the mean direction of the ventricular depolarization vector (the electrical axis) shows a wide range of normality, there is corresponding variation in QRS patterns consistent with a normal ECG. Thus correct interpretation of the ECG must take account of the electrical axis. The frontal plane axis is determined by identifying the limb lead in which the net QRS deflection (positive and negative) is least pronounced. This lead must be at right angles to the frontal plane electrical axis, which is defined using an arbitrary hexaxial reference system (Fig. 11.10).

Figure 11.10 Mean frontal QRS axis. This is the mean direction of the left ventricular depolarization vector in those leads (I–aVF) that lie in the frontal plane of the heart. It lies at right angles to the lead in which the net QRS deflexion is least pronounced. It is quantified using a hexaxial reference system. The QRS axis shows a wide range of normality from −30° to 90°. Thus despite the different ECG patterns in this illustration, only recordings (A) and (G) are abnormal, owing to left and right axis deviation, respectively.

Heart rate

The ECG is usually recorded at a paper speed of 25 mm/s. Thus each large square (5 mm) represents 0.20 s. The heart rate (bpm) is conveniently calculated by counting the number of large squares between consecutive R waves and dividing this into 300.

Rhythm

In normal sinus rhythm, P waves precede each QRS complex and the rhythm is regular. Absence of P waves and an irregular rhythm indicate atrial fibrillation.

Electrical axis

Evaluation of the frontal place QRS axis is described above.

P-wave morphology

The duration should not exceed 0.10 s, prolongation indicating left atrial enlargement, often the result of mitral valve disease or left ventricular failure. Tall-peaked ‘pulmonary’ P waves indicate right atrial enlargement, caused usually by pulmonary hypertension and right ventricular failure.

PR interval

The normal duration is 0.12-0.20 s measured from the onset of the P wave to the first deflection of the QRS complex. Prolongation indicates delayed atrioventricular conduction (first-degree heart block). Shortening indicates rapid conduction through an accessory pathway bypassing the atrioventricular node (Wolff-Parkinson-White (WPW) syndrome).

QRS morphology

The QRS duration should not exceed 0.12 s. Prolongation indicates slow ventricular depolarization due to bundle branch block (Fig. 11.12), pre-excitation (WPW syndrome), ventricular tachycardia or hypokalaemia.

Figure 11.12 Bundle branch block. (A) Left bundle branch block (LBBB); the entire sequence of ventricular depolarization is abnormal, resulting in a broad QRS complex with large slurred or notched R waves in I and V₆. (B) Right bundle branch block (RBBB); right ventricular depolarization is delayed, resulting in a broad QRS complex with an ‘rSR’ pattern in V₁ and prominent S waves in I and V₆.

Exaggerated QRS deflections indicate ventricular hypertrophy (Fig. 11.13). The voltage criteria for left ventricular hypertrophy are fulfilled when the sum of the S and R wave deflections in leads V₁ and V₆ exceeds 35 mm (3.5 mV). Right ventricular hypertrophy causes tall R waves in the right ventricular leads (V₁ and V₂). A dominant R wave in lead V₁ can also be caused by right bundle branch block, posterior myocardial infarction, WPW syndrome and dextrocardia. Diminished QRS deflections occur in myxoedema and also when pericardial effusion or obesity electrically insulates the heart. The presence of pathological Q waves (duration >0.04 s) usually indicates previous myocardial infarction.

Figure 11.13 Ventricular hypertrophy. (A) Left ventricular hypertrophy. The QRS voltage deflections are exaggerated such that the sum of S and R waves in V₁ and V₆, respectively, exceeds 35 mm. T-wave inversion in V₅ and V₆ indicates left ventricular ‘strain’. (B) Right ventricular hypertrophy. Prominent R waves in V₁ and V₂ associated with T-wave inversion are shown.

QT interval

This is measured from the onset of the QRS complex to the end of the T wave and represents the duration of electrical systole (mechanical systole starts between the QRS complex and the T wave). The QT interval (0.35-0.45 s) is very rate sensitive, shortening as heart rate increases. Abnormal prolongation of the QT interval predisposes to ventricular arrhythmias and may be congenital or occur in response to hypokalaemia, rheumatic fever or drugs (e.g. quinidine, amiodarone, tricyclic antidepressants). Shortening of the QT interval is caused by hyperkalaemia and digoxin therapy.

ST segment morphology

Minor ST elevation reflecting early repolarization may occur as a normal variant (Fig. 11.14), particularly in subjects of African or West Indian origin. Pathological elevation (>2.0 mm above the isoelectric line) occurs in acute myocardial infarction, variant angina and pericarditis. Horizontal ST depression indicates myocardial ischaemia. Other important causes of ST depression are digoxin therapy and hypokalaemia.

Figure 11.14 ST segment morphology: common causes of ST segment elevation and depression. Note that depression of the J point (junction between the QRS complex and ST segment) is physiological during exertion and does not signify myocardial ischaemia. Planar depression of the ST segment, on the other hand, is strongly suggestive of myocardial ischaemia.

T-wave morphology

The orientation of the T wave should be directionally similar to the QRS complex. Thus T-wave inversion is normal in leads with dominantly negative QRS complexes (aVR, V₁ and sometimes lead III). Pathological T-wave inversion occurs as a non-specific response to various stimuli (e.g. viral infection, hypothermia). More important causes of T-wave inversion are ventricular hypertrophy, myocardial ischaemia and myocardial infarction. Exaggerated peaking of the T wave is the earliest ECG change in ST elevation myocardial infarction. It also occurs in hyperkalaemia.

Stable angina

The ECG is often normal in patients with stable angina unless there is a history of myocardial infarction, when pathological Q waves or T-wave inversion may be present.

Exercise stress testing

The patient is usually exercised on a treadmill, the speed and slope of which can be adjusted to increase the workload gradually. In patients with coronary artery disease, exercise-induced increases in myocardial oxygen demand may outstrip oxygen delivery through the atheromatous arteries, resulting in regional ischaemia. This causes planar or downsloping ST segment depression, with reversal during recovery (Fig. 11.15). The ready availability of the exercise test means that it is one of the most widely used tests for evaluating the patient with chest pain but its diagnostic accuracy is limited to about 70%, false positive or false negative results being common when the pretest probability of coronary disease is very low (as in young women) or very high (as in elderly patients with typical symptoms), respectively. Non-invasive imaging modalities (see later: magnetic resonance perfusion imaging, stress echocardiography, CT coronary angiography) have higher rates of diagnostic accuracy and are now preferred in the assessment of patients with suspected coronary artery disease, where available. The exercise ECG also provides prognostic information in patients who are known to have coronary artery disease: an increased risk of myocardial infarction or sudden death is indicated by ST depression very early during exercise, by an exertional fall in blood pressure or by exercise-induced ventricular arrhythmias. In these cases, urgent coronary angiography is required.

Figure 11.15 Exercise ECG: ischaemic changes in inferior standard leads (II, III and aVF). At rest, the ST segments are isoelectric. Exercise causes tachycardia and provokes 3 mm of downsloping ST depression in leads II, III and aVF. The changes reverse during recovery. The findings suggest exertional ischaemia affecting the inferior wall of the heart. The probability of coronary artery disease is high.

Acute coronary syndromes

Acute myocardial infarction and unstable angina present similarly with unprovoked – often severe – ischaemic cardiac pain. Reliable differentiation between the acute coronary syndromes on clinical grounds cannot be made and requires measurement of cardiac biomarkers, in particular troponin I or T (see below), 12 hours after the onset of symptoms, and observation of the 12-lead ECG. The combination of typical symptoms plus raised troponins is diagnostic of myocardial infarction, which is categorized as ST elevation myocardial infarction (STEMI) or non-ST elevation myocardial infarction (non-STEMI) by the ECG findings. Typical symptoms unassociated with either troponin release or ST elevation are diagnosed as troponin-negative acute coronary syndrome or unstable angina. It therefore follows that acute myocardial infarction and unstable angina may be associated with a completely normal ECG or with ST depression (Fig. 11.16) or T-wave changes, the diagnosis depending on the presence or absence of raised troponins.

Figure 11.16 Unstable angina or non-ST elevation myocardial infarction (depending on troponin release): 12-lead ECG showing planar/downsloping ST depression in the inferolateral territory.

In STEMI, the evolution of ECG changes is characteristic, although it may be aborted by timely reperfusion therapy (thrombolysis or primary stenting). Peaking of the T wave followed by ST segment elevation occurs during the first hour of pain (Fig. 11.17). The changes are regional, and reciprocal ST depression may be seen in the opposite ECG leads. Usually a pathological Q wave develops during the following 24 hours and persists indefinitely. The ST segment returns to the isoelectric line within 2-3 days, and T-wave inversion may occur. The ECG is a useful indicator of infarct location. Changes in leads II, III and aVF indicate inferior infarction (Fig. 11.18), whereas changes in leads V₁–V₆ indicate anteroseptal (V₁–V₃) or anterolateral (V₁–V₆) infarction (Fig. 11.19). When the infarct is located posteriorly, ECG changes may be difficult to detect, but dominant R waves in leads V₁ and V₂ often develop (see Fig. 11.18).

Figure 11.17 Acute myocardial infarction: evolution of ECG changes. Elevation of the ST segment occurs during the first hour of chest pain. The Q wave develops during the subsequent 24 hours and usually persists indefinitely. Within a day of the attack, the ST segment usually returns to the isoelectric line and T-wave inversion may occur.

Figure 11.18 Acute inferolateral infarction. ECG 2 hours after the onset of chest pain. Typical ST elevation in leads II, III and aVF is diagnostic of inferior myocardial infarction. ST elevation in leads V₄-V₆ indicates lateral extension. There is reciprocal ST depression in lead aVL. Prominent R waves associated with ST depression in leads V₁ and V₂ indicate posterior wall infarction. This pattern may reflect occlusion of the right coronary artery or a dominant circumflex coronary artery.

Figure 11.19 Acute anterior infarction. ECG 1 hour after the onset of chest pain. Typical ST elevation in leads V₂–V₅ is diagnostic of anterior myocardial infarction. Additional ST elevation in standard leads I and aVL indicates lateral extension of the infarct. This pattern usually reflects proximal occlusion of the left anterior descending coronary artery.

In-hospital ECG monitoring

Patients who have had out-of-hospital cardiac arrest or severe, arrhythmia-induced heart failure should undergo continuous ECG monitoring in hospital under the surveillance of staff trained in the recognition and treatment of arrhythmia. Patients with acute myocardial infarction should undergo ECG monitoring for 24 hours, after which time the risk of ventricular arrhythmia falls dramatically.

Ambulatory (Holter) ECG monitoring

Patients with frequent palpitation or dizzy attacks are commonly investigated by means of an ambulatory 24-hour ECG. The availability of portable cassette recorders allows this to be performed as an outpatient. Analysis of the tape identifies any cardiac arrhythmias that occurred during the monitoring period (Fig. 11.20). A patient diary allows correlation between symptoms and heart rhythm.

Figure 11.20 Ventricular tachycardia: Holter recording. When tachycardias are paroxysmal in nature, continuous ECG monitoring is often necessary to document the arrhythmia. Here a Holter recording illustrates a long burst of rapid VT lasting a total of 6 minutes. Preceding the VT there is second-degree heart block (arrows).

Patient-activated ECG recording

For patients with infrequent symptoms, the detection rate with 24-hour ambulatory monitoring is low and patient-activated recorders are more useful. The patient can keep the recorder for several weeks and activate it when symptoms occur. Rhythm strips are stored for later analysis.

Implantable loop recording

Patients in whom there is clinical suspicion of serious arrhythmia but whose symptoms occur less than once a month pose particular diagnostic difficulty. In this group, a miniaturized recording device can be implanted subcutaneously using local anaesthetic and interrogated electronically through the skin in the event of symptoms.

Exercise testing

The ECG recorded during exercise may be helpful when there is a history of exertional palpitation. Arrhythmias provoked by ischaemia or increased sympathetic activity are more likely to be detected during exercise.

Tilt testing

When malignant vasovagal syndrome is suspected, ECG and blood pressure recordings during tilting from supine to erect posture can be helpful. Abnormal bradycardia or hypotension sufficient to produce presyncope or syncope indicate a ‘vasodepressor response’ or ‘cardioinhibitory response’, respectively, and are strongly suggestive of the diagnosis.

Electrophysiological study

This technique requires cardiac catheterization with catheter-mounted electrodes. Premature stimuli are introduced into the atria or ventricles with a view to stimulating re-entry arrhythmias. In the normal heart, sustained arrhythmias are rarely provoked by premature stimuli. Thus arrhythmia provocation is usually diagnostic, particularly when the arrhythmia reproduces symptoms. Electrophysiological study can identify accessory pathways, and areas of focal atrial or ventricular ectopy as the prelude to radiofrequency ablation of the arrhythmia substrate.

Atrial ectopic beats

These rarely indicate heart disease. They often occur spontaneously, but may be provoked by toxic stimuli such as caffeine, alcohol and cigarette smoking. They are caused by the premature discharge of an atrial ectopic focus; an early and often bizarre P wave is essential for the diagnosis. The premature impulse enters and depolarizes the sinus node such that a partially compensatory pause occurs before the next sinus beat during resetting of the sinus node.

Atrial fibrillation

Prevalence increases with age and it is common in hypertensive heart disease, mitral valve disease, thyrotoxicosis and left ventricular failure. It can be precipitated by pneumonia, major surgery and by various toxic stimuli, particularly alcohol. Atrial activity is chaotic and mechanically ineffective. P waves are therefore absent and replaced by irregular fibrillatory waves (rate 400-600/min). The long refractory period of the atrioventricular node ensures that only some of the atrial impulses are conducted, to produce an irregular ventricular rate of 130-200 bpm. If the atrioventricular node is diseased, the ventricular rate is slower, but in the presence of a rapidly conducting accessory pathway in WPW syndrome, dangerous ventricular rates above 300 bpm may occur.

Atrial flutter

This is less common than atrial fibrillation but occurs under identical circumstances. Re-entry mechanisms produce an atrial rate close to 300 bpm. The normal atrioventricular node conducts with 2 : 1 block, giving a ventricular rate of 150 bpm. Higher degrees of block may reflect intrinsic disease of the atrioventricular node or the effects of nodal blocking drugs. The ECG characteristically shows sawtooth flutter waves, which are most clearly seen when the block is increased by carotid sinus pressure.

Atrioventricular nodal re-entry tachycardia (AVNRT)

The abnormal atrionodal pathway provides the basis for a small re-entry circuit. In sinus rhythm, the electrocardiogram is usually normal, although occasionally the PR interval is short (Lown-Ganong-Levine syndrome). During tachycardia, the rate is 150-250 bpm (see Fig. 11.21). The arrhythmia is usually self-limiting. Sustained AVNRT will sometimes respond to carotid sinus pressure. If this fails, intravenous adenosine or verapamil are usually effective by blocking the re-entry circuit within the atrioventricular node. Antitachycardia pacing or direct current cardioversion may also be used. Many patients are now being treated by catheter ablation to destroy the abnormal atrionodal pathway and avoid the need for long-term drug therapy (see below).

Wolff-Parkinson-White syndrome

This congenital disorder, which affects 0.12% of the population, is caused by an accessory pathway (bundle of Kent) between the atria and the ventricles. During sinus rhythm, atrial impulses conduct more rapidly through the accessory pathway than the atrioventricular node, such that the initial phase of ventricular depolarization occurs early (pre-excitation) and spreads slowly through the ventricles by abnormal pathways. This produces a short PR interval and slurring of the initial QRS deflection (δ wave). The remainder of ventricular depolarization, however, is rapid because the delayed arrival of the impulse conducted through the atrioventricular node rapidly completes ventricular depolarization by normal His-Purkinje pathways (Fig. 11.22). Cardiac arrhythmias affect about 60% of patients with WPW syndrome and are usually re-entrant (rate 150-250 bpm) triggered by an atrial premature beat. In most patients, the re-entry arrhythmia is ‘orthodromic’, with anterograde conduction through the atrioventricular node and retrograde conduction through the accessory pathway (atrioventricular re-entry tachycardia (AVRT), Fig 11.23). This results in a narrow complex tachycardia (without pre-excitation) that is indistinguishable from AVNRT. Occasionally, the re-entry circuit is in the opposite direction (‘antidromic’), producing a very broad, pre-excited tachycardia.

Figure 11.22 WPW syndrome: 12-lead ECG. Ventricular pre-excitation is reflected on the surface ECG by a short PR interval and a slurred upstroke to the QRS complex (δ wave). The remainder of the QRS complex is normal because delayed arrival of the impulse conducted through the AV node rapidly completes ventricular depolarization through normal His-Purkinje pathways.

Figure 11.23 WPW syndrome: re-entry tachycardia recorded at fast paper speed. Just after the third pre-excited (broad) complex, a premature atrial pacing stimulus (S) initiates an impulse that is blocked in the bundle of Kent but conducted normally through the AV node, producing ventricular depolarization without pre-excitation. Thus the QRS complex is narrow and lacks a δ wave. The impulse is conducted retrogradely through the bundle of Kent, re-enters the proximal conducting system and completes the re-entry circuit, initiating a self-sustaining orthodromic re-entry tachycardia (last three complexes).

Patients with WPW syndrome are more prone than the general population to atrial fibrillation. If the accessory pathway is able to conduct the fibrillatory impulses rapidly to the ventricles, it may result in ventricular fibrillation and sudden death. Digoxin (and to a lesser extent verapamil) should be avoided because it shortens the refractory period of the accessory pathway and can heighten the risk. Patients with dangerous accessory pathways of this type require catheter ablation of the pathway. Ablation therapy also cures AVNRT and is the treatment of choice in patients with frequent attacks.

Ventricular premature beats

These may occur in normal individuals, either spontaneously or in response to toxic stimuli such as caffeine or sympathomimetic drugs. They are caused by the premature discharge of a ventricular ectopic focus that produces an early and broad QRS complex (Fig. 11.24). The premature impulse may be conducted backwards into the atria, producing a retrograde P wave, but penetration of the sinus node is rare. Thus, resetting of the sinus node does not usually occur and there is a fully compensatory pause before the next sinus beat.

Figure 11.24 Multifocal, ventricular ectopic beats. Frequent broad complex ectopic beats are seen early after the sinus beats. Note, however, that the ectopic beats have two different morphologies, indicating that they arise from different foci. Note also that the coupling interval (interval between QRS complex and ectopic beat) is identical for beats arising from any particular focus.

Ventricular tachycardia

This is always pathological. It is defined as three or more consecutive ventricular beats at a rate above 120 per minute. Ventricular depolarization inevitably occurs slowly by abnormal pathways, producing a broad QRS complex. This distinguishes it from most atrial and junctional tachycardias which have a narrow QRS complex, although differential diagnosis may be more difficult for atrial or junctional tachycardias with a broad QRS complex caused by rate-related or pre-existing bundle branch block (Fig. 11.25). Nevertheless, ventricular tachycardia can usually be identified by careful scrutiny of the 12-lead ECG (Fig. 11.26). Support for the diagnosis is provided by a very broad QRS complex (>140 ms), extreme left or right axis deviation, concordance of the QRS deflections in V₁–V₆ (either all positive or all negative) and configurational features of the QRS complex, including an ‘rSR’ complex in V₁ and a QS complex in V₆. Confirmation of the diagnosis is provided by any evidence of AV dissociation: either P waves, at a slower rate than the QRS complexes, ‘marching through’ the tachycardia (Fig. 11.27A), or ventricular capture and/or fusion beats, in which the dissociated atrial rhythm penetrates the ventricle by conduction through the AV node and interrupts the tachycardia, producing a normal ventricular complex (capture, Fig. 11.27A) or, more commonly, a broad hybrid complex (fusion) that is part sinus and part ventricular in origin (Fig. 11.27B). Torsades de pointes, a broad complex tachycardia with changing wavefronts, also provides unequivocal evidence of ventricular tachycardia and is particularly characteristic of the arrhythmia that complicates long QT syndrome, often resulting in sudden death. The syndrome may be inherited as an autosomal dominant (Romano-Ward syndrome) or as an autosomal recessive (Lange-Nielsen syndrome) trait, when it is associated with congenital deafness.

Figure 11.25 Paroxysmal AV nodal re-entrant tachycardia. This Holter recording shows sinus rhythm giving way to a broad complex tachycardia. However, this is clearly the result of temporary bundle branch block because it converts spontaneously to a narrow complex tachycardia, confirming that the arrhythmia is junctional, not ventricular, in origin.

Figure 11.26 Ventricular tachycardia: 12-lead ECG. The recording shows a broad complex tachycardia. The following features suggest or confirm the ventricular origin of the tachycardia: very broad QRS complex (>140 ms); extreme right axis deviation; atrioventricular dissociation – note the dissociated P waves in lead V₁; the ‘rSR’ complex in V₁.

Figure 11.27 (A) Ventricular tachycardia: AV dissociation. P waves (arrowed) can be seen ‘marching through’ the tachycardia, confirming its ventricular origin. The tachycardia is interrupted by a narrow capture beat. (B) Ventricular tachycardia (VT): fusion. In this example, VT is initiated by a very early ventricular ectopic beat (morphologically similar to the previous isolated ectopic beat) and is interrupted by a fusion beat (arrowed), confirming the ventricular origin of the tachycardia.

Ventricular fibrillation

This occurs most commonly in severe myocardial ischaemia, either with or without frank infarction. It is a completely disorganized arrhythmia characterized by irregular fibrillatory waves with no discernible QRS complexes. There is no effective cardiac output and death is inevitable unless resuscitation with direct current cardioversion is instituted rapidly.

Sinus bradycardia (<50 bpm)

This is physiological during sleep and in trained athletes, but in other circumstances often reflects sinoatrial disease, particularly when the heart rate fails to increase normally with exercise.

Sinoatrial block

If the sinus impulse is blocked and fails to trigger atrial depolarization, a pause occurs in the ECG. No P wave is seen during the pause owing to the absence of atrial depolarization. The electrically ‘silent’ sinus discharge, however, continues uninterrupted. Thus, the pause is always a precise multiple of preceding PP intervals. Sinoatrial block that cannot be abolished by atropine-induced vagal inhibition usually indicates sinoatrial disease, particularly with pauses longer than 2 seconds.

Sinus arrest

Failure of sinus node discharge produces a pause on the ECG that bears no relation to the preceding PP interval. Pauses longer than 2 seconds are usually pathological. Prolonged pauses are often terminated by an escape beat from a ‘junctional’ focus in the bundle of His.

Bradycardia-tachycardia syndrome

In this syndrome, atrial bradycardias are interspersed by paroxysmal tachyarrhythmias, usually atrial fibrillation. Nevertheless, it is the bradycardia that usually causes symptoms, particularly dizzy attacks and blackouts.

First-degree atrioventricular block

Delayed atrioventricular conduction causes prolongation of the PR interval (>0.20 s). Ventricular depolarization occurs rapidly by normal His-Purkinje pathways and the QRS complex is usually narrow.

Second-degree atrioventricular block: Mobitz type I (Wenckebach)

This commonly occurs in inferior myocardial infarction. Successive sinus beats find the atrioventricular node increasingly refractory until failure of conduction occurs. The delay permits recovery of nodal function, and the process may then repeat itself. The ECG shows progressive prolongation of the PR interval, culminating in a dropped beat. Block is within the atrioventricular node itself and ventricular depolarization occurs rapidly by normal pathways. Thus the QRS complex is usually narrow.

Second-degree atrioventricular block: Mobitz type II

This indicates advanced conducting tissue disease affecting the bundle branches. The ECG typically shows a normal PR interval with bundle branch block in conducted beats, and intermittent block in the other bundle branch resulting in complete failure of atrioventricular conduction and dropped beats.

Third-degree (complete) atrioventricular block

The atrial and ventricular rhythms are ‘dissociated’ because none of the atrial impulses are conducted. Thus the ECG shows regular P waves (unless the atrium is fibrillating) and regular but slower QRS complexes occurring independently of each other. When block is within the atrioventricular node (e.g. inferior myocardial infarction, congenital atrioventricular block), a junctional escape rhythm with a reliable rate (40-60 bpm) takes over (see Fig. 11.29). Ventricular depolarization occurs rapidly by normal pathways, producing a narrow QRS complex. However, when block is within the bundle branches (e.g. idiopathic fibrosis), there is always extensive conducting tissue disease. The ventricular escape rhythm is slow and unreliable, with a broad QRS complex (see Fig. 11.29).

Right bundle branch block

This may be a congenital defect but is more commonly the result of organic conducting tissue disease. Right ventricular depolarization is delayed, resulting in a broad QRS complex with an ‘rSR’ pattern in lead V₁ and prominent S waves in leads I and V₆.

Left bundle branch block

This always indicates organic conducting tissue disease. The entire sequence of ventricular depolarization is abnormal, resulting in a broad QRS complex with large slurred or notched R waves in leads I and V₆.

Ventricular dilatation

The PA chest X-ray does not reliably distinguish left from right ventricular dilatation. For this, the lateral chest X-ray is more helpful. Dilatation of the posteriorly located left ventricle encroaches on the retrocardiac space, whereas dilatation of the anteriorly located right ventricle encroaches on the retrosternal space.

Atrial dilatation

Right atrial dilatation is usually due to right ventricular failure, but occurs as an isolated finding in tricuspid stenosis and Ebstein’s anomaly. It produces cardiac enlargement without specific radiographic signs.

Left atrial dilatation occurs in left ventricular failure and mitral valve disease (Fig. 11.32). Radiographic signs are:

Figure 11.32 Left atrial dilatation. This is a penetrated PA chest X-ray in a patient with mitral stenosis. The dilated left atrium causes a bulge on the left heart border below the pulmonary artery which is also dilated, widening of the carina and the double-density sign at the right heart border.

Vascular dilatation

Aortic dilatation caused by aneurysm or dissection may produce widening of the entire upper mediastinum. Localized dilatation of the proximal aorta occurs in aortic valve disease and produces a prominence in the right upper mediastinum (Fig. 11.33). Dilatation of the main pulmonary artery occurs in pulmonary hypertension and pulmonary stenosis and produces a prominence below the aortic knuckle (Fig. 11.34).

Figure 11.33 Chest X-ray in a patient with Marfan’s syndrome. Note the dilatation of the ascending aorta.

Figure 11.34 Atrial septal defect: chest X-ray. Note the prominent proximal pulmonary arteries and the pulmonary plethora reflecting increased pulmonary flow.

Intracardiac calcification

Because the radiodensity of cardiac tissue is similar to that of blood, intracardiac structures can rarely be identified unless they are calcified. Valvular, pericardial or myocardial calcification may occur, and usually indicate important disease of these structures. Calcification is best appreciated on the deeply penetrated lateral chest X-ray.

Altered pulmonary flow

Increments in pulmonary flow sufficient to cause radiographic abnormalities are caused by left-to-right intracardiac shunts (e.g. atrial septal defect; see Fig 11.34, ventricular septal defect, patent ductus arteriosus). Prominence of the vascular markings gives the lung fields a plethoric appearance. Reductions in pulmonary flow, on the other hand, cause reduced vascular markings. This may be regional (e.g. pulmonary embolism) or global (e.g. severe pulmonary hypertension).

Increased left atrial pressure

This occurs in mitral stenosis and left ventricular failure and produces corresponding rises in pulmonary venous and pulmonary capillary pressures. Prominence of the upper lobe veins is an early radiographic finding. As the left atrial and pulmonary capillary pressures rise above 18 mmHg, transudation into the lung produces interstitial pulmonary oedema, characterized by prominence of the interlobular septa, particularly at the lung bases (Kerley B lines). Further elevation of pressure leads to alveolar pulmonary oedema, characterized by perihilar ‘bat’s-wing’ shadowing (Fig. 11.35).

Figure 11.35 Chest X-ray in acute left ventricular failure: the patient had severe pulmonary oedema caused by acute myocardial infarction. The heart is not yet enlarged, but there is prominent alveolar pulmonary oedema in a perihilar (‘bat’s-wing’) distribution. Note the bilateral pleural effusions.

Other lung field abnormalities

Physics

A transducer containing a piezoelectric element converts electrical energy into an ultrasound beam that is directed towards the heart. The beam is reflected when it strikes an interface between tissues of different densities. The reflected ultrasound, or echo, is converted back to electrical energy by the piezoelectric element, which permits the construction of an image using two basic units of information:

Density differences within the heart are greatest between the blood-filled chambers and the myocardial and valvular tissues, all of which are clearly visible on the echocardiogram. Because the depth of the myocardial and valvular tissues with respect to the transducer changes constantly throughout the cardiac cycle, the time taken for echo reflection changes accordingly. Thus, real-time imaging throughout the cardiac cycle provides a dynamic record of cardiac function.

M-mode echocardiogram

This provides a unidimensional ‘ice-pick’ view through the heart. Continuous recording on photographic paper provides an additional time dimension, thereby permitting appreciation of the dynamic component of the cardiac image. By convention, cardiac structures closest to the transducer are displayed at the top of the record and more distant structures are displayed below. Thus, on the transthoracic M-mode echocardiogram, anteriorly located (‘right-sided’) structures lie above the posteriorly located (‘left-sided’) structures, but on the transoesophageal echocardiogram, the display is reversed. M-mode is particularly useful for measuring chamber dimensions and left ventricular wall thickness, and for timing events within the cardiac cycle (Fig. 11.36A).

Figure 11.36 (A) M-mode echocardiography. The figure shows a sweep as the transducer is angulated from the left ventricle to the aortic root. (B) Transthoracic 2D echocardiography. Parasternal long-axis and apical four-chamber views are shown. The dots are a 1-cm scale. AV, aortic valve; CW, chest wall; IVS, interventricular septum; LA, left atrium; LV, left ventricle; MV, mitral valve; PW, posterior LV wall; RV, right ventricle.

Two-dimensional echocardiogram

This provides more detailed information about morphology than the M-mode recording. By projecting a fan of echoes in an arc of up to 80°, a 2D ‘slice’ through the heart can be obtained, the precise view depending on the location and angulation of the transducer (Fig. 11.36B).

Congenital heart disease

Echocardiography, particularly the 2D technique, has revolutionized the diagnosis of congenital heart disease, in the majority of cases obviating the need for invasive investigation by cardiac catheterization (Fig. 11.37). The relationships of the cardiac chambers and their connections with the great vessels are readily determined. Valvular abnormalities and septal defects can also be recognized. Recent technology has permitted in utero fetal imaging for the antenatal diagnosis of cardiac defects.

Figure 11.37 Atrial septal defect: 2D echocardiogram (subcostal view). In this view, good views of the interatrial septum can usually be obtained without resorting to transoesophageal echocardiography. Note the atrial septal defect (arrowed) and the dilatation of the right atrium (RA).

Myocardial disease

Echocardiography permits accurate assessment of cardiac dilatation, hypertrophy and contractile function. Dilated cardiomyopathy produces ventricular dilatation with global contractile impairment (Fig. 11.38). This must be distinguished from the regional contractile impairment that follows myocardial infarction in patients with coronary artery disease (Fig. 11.39). Hypertrophic cardiomyopathy is characterized by thickening (hypertrophy) of the left ventricular myocardium, usually with disproportionate involvement of the interventricular septum (asymmetric septal hypertrophy). In aortic and hypertensive heart disease, on the other hand, left ventricular hypertrophy is usually symmetrical (Fig. 11.40).

Figure 11.38 Dilated cardiomyopathy: echocardiogram. This M-mode study shows severe dilatation of the left ventricular cavity and severe global contractile impairment. The patient later underwent successful heart transplantation.

Figure 11.39 Heart failure: echocardiogram. This M-mode study shows considerable dilatation of the left ventricle. Note that the interventricular septum (IVS) is almost akinetic, but the posterior wall (PW) is contracting normally. Regional contractile impairment of this type indicates coronary heart disease. The phonocardiogram recorded simultaneously shows normal first and second heart sounds and also a third heart sound (arrowed).

Figure 11.40 Aortic stenosis and left ventricular hypertrophy: 2D echocardiogram (long-axis view). The aortic valve (AV) is grossly thickened and calcified. Concentric left ventricular hypertrophy is present (arrowed).

Valvular disease

Echocardiography is of particular value for identifying both structural and dynamic valvular abnormalities and any associated chamber dilatation or hypertrophy (Boxes 11.18–11.21). The severity of valvular involvement in congenital, rheumatic, degenerative and infective disease may thus be defined; the technique is diagnostic for bicuspid aortic valve and mitral valve prolapse, and readily identifies valve thickening and calcification in rheumatic and calcific disease (Figs 11.40 and 11.41). Vegetations in infective endocarditis can usually be visualized if they are large enough (>3 mm, Fig. 11.42). The transoesophageal approach is usually necessary for endocarditis involving prosthetic heart valves.

Figure 11.41 Mitral stenosis: 2D echocardiogram (parasternal long-axis view). The mitral valve leaflets are densely thickened (arrowed) and the left atrium is severely dilated.

Figure 11.42 Infective endocarditis: transoesophageal echocardiogram. A vegetation (arrowed) is adherent to the aortic valve leaflet.

Pericardial disease

Although the echocardiogram is of little value in constrictive pericarditis, it is the most sensitive technique available for the diagnosis of pericardial effusion (Fig. 11.43). The effusion appears as an echo-free space distributed around the ventricles but usually avoiding the potential space behind the left atrium.

Figure 11.43 Pericardial effusion: 2D echocardiogram (parasternal long-axis view). Note the echo-free space (arrowed) around the heart, but not behind the left atrium.

Other clinical applications

Intracardiac tumours, particularly myxomas (see Fig. 11.1) and thrombi (Fig. 11.44), are readily visualized by echocardiography. The transoesophageal technique is more sensitive for identifying thrombus in the left atrial appendage and is also helpful for diagnosing aortic disease, such as aneurysm and dissection, as it provides better views of the thoracic aorta than is possible with conventional 2D echocardiography (Fig. 11.45).

Figure 11.44 Mitral stenosis: 2D echocardiogram (long-axis view). The left atrium is severely dilated and a large thrombus (arrowed) is visible, emphasizing the importance of anticoagulation in patients with mitral valve disease and atrial fibrillation.

Figure 11.45 Aortic dissection: transoesophageal echocardiogram. Right panel: this long-axis transoesophageal echocardiogram reveals the dilated aortic root and an S-shaped flap (arrowed) traversing the lumen. Left panel: same patient with colour Doppler superimposed to show flow in the true lumen.

Physics

According to the Doppler principle, when an ultrasound beam is directed towards the bloodstream, the frequency of the sound waves reflected from the blood cells is altered. The frequency shift or Doppler effect is related to the direction and velocity of flow. If continuous-wave Doppler is used, blood flow at any point along the path of the ultrasound beam is detected, such that a ‘clean’ Doppler signal from the area of interest may be difficult to obtain. Pulsed Doppler, however, has a range-gating facility that permits frequency sampling from any specific point within the heart, preselected on the echocardiogram. This lends greater precision to the technique. Nevertheless, pulsed Doppler is less able than continuous-wave Doppler to quantify very high-velocity jets, such as those that occur in aortic stenosis.

Colour-flow mapping

This has been a major technological advance. Instead of the unidirectional ultrasound beam used in continuous-wave and pulsed Doppler imaging, the beam is rotated through an arc. Frequency sampling throughout the arc permits the construction of a colour-coded map, red indicating flow towards and blue away from the transducer. Colour-flow data can be superimposed on the standard 2D echocardiogram to identify precisely the patterns of flow within the four chambers of the heart. This simplifies the interpretation of Doppler imaging and provides more useful qualitative data, although it is less useful for quantitative assessment of valve gradients, which requires the precision of conventional Doppler technique.

Radionuclide ventriculography

This method is used for assessment of ventricular function. Red cells labelled with technetium-99m (^99mTc) are allowed to equilibrate in the blood pool and the heart is then imaged under the gamma camera. The waxing and waning of radioactivity within the ventricular chambers during diastole and systole, respectively, permits the construction of a dynamic ventriculogram. Left ventricular contractile function can be evaluated quantitatively by calculating the ejection fraction, or qualitatively by observing wall movement. In clinical practice, this has now been replaced by other techniques for the assessment of left ventricular function such as echocardiography and magnetic resonance imaging.

Myocardial perfusion scintigraphy

This method is used for the diagnosis of coronary artery disease (Fig. 11.47). The patient is ‘stressed’ in an attempt to provoke myocardial ischaemia either by a standardized exercise test, an intravenous dobutamine infusion or an intravenous adenosine infusion. Isotope is injected intravenously at peak stress and the heart is imaged under a gamma camera. Thallium-201 (²⁰¹Tl) has now given way to ^99mTc-labelled methoxy-isobutyl-isonitrile (MIBI) which provides better image quality. Isotope is distributed homogeneously in normally perfused myocardium, ischaemic or infarcted areas appearing as scintigraphic defects. If ²⁰¹Tl is used, repeat imaging after 2-4 hours’ rest permits the reassessment of scintigraphic defects; those that disappear (reversible defects) indicate areas of stress-induced ischaemia, while those that persist (fixed defects) indicate infarcted myocardium. If ^99mTc-labelled MIBI is used, resting images for assessment of reversibility require a separate injection of isotope 24 hours after (or before) the stress images.

Figure 11.47 Isotope perfusion scan. These are tomographic slices across the short axis of the left ventricle. A posterior wall defect is seen during stress, but it largely disappears during rest as isotope ‘redistributes’ into the ischaemic area. A smaller fixed defect is seen in the anterior wall, indicating infarction in that territory.

Positron emission tomography

Positron emission tomography (PET) scanning is used to determine myocardial ‘viability’ in patients with heart failure (see stress echocardiography, above). Simultaneous assessment of myocardial perfusion using ¹³N ammonia and glucose uptake using a glucose analogue permits the identification of viable but dysfunctional myocardium, in which perfusion is impaired but metabolic activity in terms of glucose uptake remains normal. This perfusion-metabolic ‘mismatch’ indicates viable muscle likely to respond favourably to revascularization by angioplasty or bypass surgery.

Pulmonary scintigraphy (radioisotope imaging)

This method is used for the diagnosis of pulmonary embolism (Fig. 11.48). ^99mTc-labelled microspheres injected intravenously become trapped within the pulmonary capillaries. The normal pulmonary perfusion scintigram shows a homogeneous distribution of radioactivity throughout both lung fields. Pulmonary embolism causes regional impairment of pulmonary flow, which results in a perfusion defect on the scintigram; however, the appearance is non-specific and occurs in many other pulmonary disorders, particularly chronic obstructive pulmonary disease. Specificity is enhanced by simultaneous ventilation scintigraphy (a ventilation/perfusion or V/Q scan). Inhaled xenon-133 (¹³³Xe) is distributed homogeneously throughout the normal lung, and in pulmonary embolism (unlike other pulmonary disorders), distribution remains homogeneous. Thus, a scintigraphic perfusion defect not ‘matched’ by a ventilation defect is highly specific for pulmonary embolism. CT pulmonary angiography is now the diagnostic test of choice for pulmonary embolism, but V/Q scanning is still valuable in patients with suspected pulmonary embolism who have severe renal impairment in whom X-ray contrast is contraindicated.

Figure 11.48 Ventilation (right) and perfusion (left) lung scans in pulmonary embolism. Contrast the homogeneous distribution of isotope in the ventilation scan with the regional defects in the perfusion scan.

Aortic root angiography

Injection of contrast into the aortic root demonstrates the vascular anatomy in suspected aneurysm or dissection, and also permits evaluation of aortic valve function (Fig. 11.55). The normal aortic valve prevents diastolic backflow of contrast, but in aortic regurgitation, variable opacification of the left ventricle occurs, depending on the severity of the valve lesion.

Figure 11.55 Aortic root angiogram: normal study. Contrast injection provides an X-ray image of the ascending aorta. The coronary arteries arising from the sinuses of Valsalva are clearly seen.

Left ventricular angiography

Contrast injection into the left ventricle defines ventricular anatomy and wall motion, and also permits evaluation of mitral valve function. Dilatation of the ventricle and contractile dysfunction occurs in left ventricular failure. An akinetic segment denotes previous myocardial infarction. Exaggerated contractile function with systolic obliteration of the cavity occurs in hypertrophic cardiomyopathy. Filling defects within the ventricular lumen may indicate thrombus or neoplasm. The normal mitral valve prevents systolic backflow of contrast into the left atrium, but in mitral regurgitation, variable atrial opacification occurs, depending on the severity of the valve lesion (Fig. 11.56).

Figure 11.56 Left ventricular angiogram showing mitral regurgitation. Contrast injection into the left ventricle has resulted in prompt opacification of the left atrium owing to backflow across the diseased mitral valve.

Coronary angiography

This is the most reliable technique for diagnostic imaging of the coronary arteries and assessment of lesion severity. Indications are summarized in Box 11.22. The technique requires selective injection of contrast into the left and right coronary arteries to opacify the lumen (Fig. 11.57), and multiple views in different projections are necessary for a complete study. Intraluminal filling defects or occlusions indicate coronary artery disease, which is nearly always caused by atherosclerosis. In stable patients, coronary angiography often reveals a stenosis or stenoses of intermediate severity. Additional information regarding the functional significance and anatomic severity of the disease is required before treatment decisions can be made. This can be obtained either from non-invasive myocardial perfusion imaging that looks for ischaemia in the territory of the affected artery (see MRI perfusion imaging) or from catheter laboratory-based techniques.

Figure 11.57 Coronary angiograms. (A) Left anterior descending disease (arrowed). This tight stenosis threatens the coronary supply to the anterior wall of the left ventricle. (B) Right coronary artery disease (arrowed). Serial stenoses in this dominant right coronary artery threaten the supply to the inferior wall of the heart.

Haemodynamic evaluation of valvular stenosis

In the normal heart, there is no pressure gradient across an open valve. Such a gradient usually indicates valvular stenosis (Fig. 11.59) and, as stenosis worsens, the pressure gradient increases. This therefore provides a useful index of the severity of stenosis. However, it must be recognized that the pressure gradient is influenced by the flow through the valve. For example, if cardiac output is very low, the gradient may be small despite the presence of severe stenosis. This applies particularly to the aortic valve because flow velocity is normally high. In patients with poor left ventricular function, the 2D appearances of the valve on echocardiography and estimates of valve area should be considered before concluding that significant aortic stenosis is not present.

Figure 11.59 Valvular stenosis: pressure signals. (A) Aortic stenosis: simultaneous left ventricular (LV) and aortic (AO) pressure signals. In the normal heart, the pressure signals should be essentially similar throughout systole. Here there is a peak systolic gradient of about 35 mmHg across the aortic valve. Note the prominent ‘a’ wave (arrowed), reflecting the major contribution that atrial systole makes to filling of the hypertrophied, non-compliant ventricle. Note also the pulsus alternans indicating left ventricular failure. (B) Mitral stenosis: simultaneous recordings of the pulmonary artery wedge (PAW) and left ventricular (LV) pressure signals. In the normal heart, the pressure signals should be superimposed throughout diastole. Here there is a pressure gradient >10 mmHg, indicating severe mitral stenosis. Note that the patient is in atrial fibrillation and the pressure gradient varies inversely with the RR interval, tending to increase as the RR interval shortens.

Haemodynamic evaluation of intracardiac shunts

Left-to-right intracardiac shunts through atrial or ventricular septal defects introduce ‘arterialized’ blood into the right side of the heart. This results in an abrupt increase or step-up in the oxygen saturation of venous blood at the level of the shunt, which can be detected by right heart catheterization. Thus, by drawing serial blood samples for oxygen saturation from the pulmonary artery, right ventricle, right atrium and vena cava, the shunt may be localized to the site at which the step-up in oxygen saturation occurs. The magnitude of the step-up is related to the size of the shunt, but precise quantification of the shunt requires measurement of pulmonary and systemic blood flow. The extent to which the pulmonary-systemic flow ratio exceeds 1 is a measure of the size of the shunt.

Haemodynamic evaluation of constriction and tamponade

In constrictive pericarditis and tamponade (Boxes 11.23 and 11.24), diastolic relaxation of the ventricles is impeded, preventing adequate filling. Compensatory increments in atrial pressures occur to help maintain ventricular filling, and because these disorders usually affect both ventricles equally, the filling pressures also equilibrate. Thus, simultaneous left- and right-sided recordings in constrictive pericarditis and tamponade show characteristic elevation and equalization of the filling pressures (atrial or ventricular end-diastolic), with loss of the normal differential (Fig. 11.60). Similar physiology characterizes restrictive cardiomyopathy, in which infiltrative disease (usually amyloid in the UK) impedes relaxation of the ventricles.

Figure 11.60 Pericardial constriction: left (LV) and right (RV) ventricular pressure signals. Like tamponade and restrictive cardiomyopathy, constriction usually affects both ventricles equally and the diastolic pressures must rise and equilibrate to maintain ventricular filling. Thus, in diastole, the pressure signals are superimposed with a typical ‘dip and plateau’ configuration (‘square root’ sign).

Measurement of cardiac output

This is usually calculated by thermodilution using a Swan-Ganz catheter with a right atrial portal and a terminal thermistor positioned in the pulmonary artery. A known volume of cold saline (usually 10 ml) is injected into the right atrium and the temperature reduction in the pulmonary artery is recorded at the thermistor. The contour of the cooling curve is dependent on cardiac output, which is calculated by measuring the area under the curve using a bedside computer.

Cardiac output can also be measured by a variety of non-invasive methods including oesophageal Doppler probe, pulse waveform analysis, bioimpedance and echocardiography. Stroke volume can be calculated using 2D echocardiography from the estimated difference in left ventricular volumes at the end of systole and diastole. Alternatively, Doppler echocardiography can be used to measure stroke volume.

Cardiac output measurement is performed most commonly in intensive care units to characterize cardiovascular status and monitor responses to treatment. It is particularly valuable in the shocked patient whose intravascular volume status cannot be determined by clinical assessment.

Cardiac enzymes and other markers of myocardial injury

Following myocardial infarction, enzymes and structural proteins released into the circulation by the necrosing myocytes provide biochemical markers of injury. In the UK, until recently, creatine kinase (CK) was the most widely used marker, the MB isoenzyme being the most specific for myocardial injury. However, newer serum markers of myocardial injury, particularly the highly specific troponins T and I, have replaced enzymatic markers. Indeed, these form the basis of the updated definition of myocardial infarction, which is diagnosed in any patient with raised circulating troponins who presents with cardiac chest pain, or who develops regional ST elevation on the 12-lead ECG. Implicit in this definition is the fact that myocardial infarction may be diagnosed in patients without ST elevation. In non-ST elevation myocardial infarction (non-STEMI), the increase in circulating troponin concentration is a useful measure of risk, troponin concentration correlating directly with rates of recurrent myocardial infarction and death, and it is widely used to guide decisions about cardiac catheterization and coronary revascularization.

Renal function

Renal function should always be measured in the cardiac patient. Renal dysfunction is a major cause (and consequence) of hypertension whereas, in heart failure, progressive deterioration of renal function is almost inevitable as perfusion of the kidneys becomes threatened. Renal function may also deteriorate in response to treatment with diuretics and angiotensin converting enzyme (ACE) inhibitors, and careful monitoring is essential as these drugs are introduced. Cardiac drugs that are excreted through the kidneys (e.g. digoxin) should be used cautiously if renal function is impaired. In the patient with established renal failure, accelerated coronary artery disease and heart failure commonly occur, reflecting the combined effects of dyslipidaemia, hypertension, arterial endothelial dysfunction, anaemia and volume loading on the cardiovascular system. Cardiovascular disease is the major cause of death in patients with renal failure.

Electrolytes

Patients with hypokalaemia are at risk of lethal cardiac arrhythmias, while hyperkalaemia may cause bradyarrhythmias and heart block. Patients presenting with acute coronary syndromes commonly have hypokalaemia, due to the effects of sympathoadrenal activation on membrane-bound sodium–potassium ATPase, and this may require correction. In patients on thiazide or loop diuretics, potassium supplements (or potassium-sparing diuretics) are usually necessary to protect against hypokalaemia unless ACE inhibitors are also given. The combination of an ACE inhibitor and an aldosterone antagonist is used in patients with severe heart failure and can be complicated by hyperkalaemia. Serum potassium should always be measured in patients with hypertension, not only to provide a baseline before the introduction of diuretic therapy, but also as a simple screening test for primary aldosteronism. Hyponatraemia is common with chronic diuretic use.

Glucose and lipids

Blood glucose and lipid profiles (triglycerides, total cholesterol and high-density lipoprotein (HDL) and low-density lipoprotein (LDL) cholesterol) should be measured in all patients with suspected vascular disease. In high-risk individuals, particularly those with diabetes, and in all patients with established coronary artery disease, statin therapy is essential regardless of the baseline lipid profile, and reduces the cardiovascular event rate by about 25%. Statins lower LDL (‘bad’) cholesterol, but if HDL (‘good’) cholesterol is low, treatment with nicotinic acid or fibrates should be added. Treatment of diabetes protects against microvascular complications, and in hyperglycaemic patients with acute myocardial infarction, infusion of insulin and glucose is usually recommended.

Blood culture

In suspected infective endocarditis (Box 11.25), treatment must not be delayed beyond the time necessary to obtain three to four blood samples for culture (Table 11.2). Aerobic, anaerobic and fungal cultures should be performed. Occasionally, bone marrow cultures are helpful for detection of Candida and Brucella endocarditis. Coxiella and Chlamydia can never be cultured from the blood and must be diagnosed by serological tests. Failure to detect bacteraemia may be due to pretreatment with antibiotics, inadequate sampling (up to six blood samples should be taken over 24 hours) or infection with unusual microorganisms.

Table 11.2 Organisms implicated in endocarditis

* These drugs are not bactericidal, and valve replacement is nearly always necessary to eradicate infection.

Serology

If a recent streptococcal throat infection can be confirmed by demonstrating an elevated serum antistreptolysin O titre, Jones’s criteria may be used for the diagnosis of rheumatic fever (Box 11.26). The presence of two major criteria, or one major and two minor criteria, indicates a high probability of rheumatic fever. In suspected viral pericarditis or myocarditis, the aetiological diagnosis depends on the demonstration of elevated viral antibody titres in acute serum samples, which decline during convalescence. Virus may sometimes be cultured from throat swabs and stools.