Cardiomyopathies are diseases of cardiac muscle of unknown cause, and do not include rare specific heart muscle diseases. Myocarditis is not included in this group of diseases since its aetiology is known. There are three specific types; hypertrophic, dilated, and restrictive. Their management is essentially medical, with surgical referral as an extreme course.

Originally called hypertrophic obstructive cardiomyopathy, this disease is characterized by massive left ventricular hypertrophy, classically causing asymmetric septal hypertrophy and gross papillary muscle hypertrophy. Systolic anterior motion of the anterior mitral leaflet may exacerbate the outflow obstruction often caused by septal hypertrophy: this is not seen in all cases, particularly with progression of the disease, and for this reason the term ‘obstructive’ has now been omitted. The hypertrophy produces a ventricle with cavity obliteration on systole, and non-compliant walls causing abnormal diastolic relaxation. Echocardiography is the main diagnostic tool, angiography being used to measure gradients on withdrawal from the left ventricular cavity. The aetiology is unknown, but there is a familial element. Studies have identified links with abnormally high catecholamine receptor density in the outflow tract and with catecholamine overproduction generally. There are also links with hyperthyroidism and hyperinsulinaemia, and the disease may be a reflection of a genetically determined disorder of catecholamine handling by the fetal myocardium.

Sudden death is common, with a 3.5 per cent annual mortality rate. Factors associated with a high risk of sudden death are young age at the time of diagnosis (5.9 per cent annual mortality if less than 30 years), positive family history, and acute deterioration in symptoms. There are no haemodynamic predictors of a worse outcome. Death is usually due to ventricular dysrrhythmias, but myocardial infarction also occurs due to the immense muscle bulk and to retrograde accessory pathway conduction causing supraventricular tachycardias. Treatment of dysrhythmia with &bgr;-blockers may decrease the incidence of sudden death. Amiodarone has been shown to abolish bursts of ventricular tachycardia in about 77 per cent of patients.

Treatment is primarily medical; surgery is reserved for unresponsive cases. In older series the surgical mortality was quoted as being 27 per cent. More recent reports show a zero operative mortality rate and a postoperative mortality for the surgical group of 1.8 per cent per year. Given the medical mortality of 3.5 per cent the indications for surgery may well expand.

Medical treatment for those without ventricular dysrhythmias includes &bgr;-blockers and verapamil. Both of these agents improve clinical manifestation and haemodynamics; each may be efficacious where the other has failed. Supraventricular dysrhythmias may be responsive to amiodarone. Congestive cardiac failure developing in patients with advanced stage disease has a very poor outlook; such patients are managed medically.

Surgery has been reserved for the 10 to 15 per cent of patients who do not benefit from medical treatment. The gradient across the left ventricular outflow tract is relieved by myotomy or myomectomy (first described by Morrow) with enlargement of the outflow tract. However, Fighali et al. claimed that improvement was greater if mitral valve replacement was performed at the same time, as the pathological anterior leaflet of this valve causes much of the obstruction. Though clinical improvement was noted in all the patients, systolic anterior motion of the mitral valve was not abolished. Outflow gradients and left ventricular end-diastolic pressures were much better in those treated with valve replacement. In the not uncommon patient with concomitant hypertrophic cardiomyopathy and coronary artery disease, combined bypass grafting, myotomy/myomectomy, and mitral valve replacement is safe and improves symptoms and prognosis compared to isolated coronary artery grafting.

Dilated cardiomyopathy is a ‘final common pathway’ for many cardiac muscle pathologies which result in ventricular dilatation; this is commonly biventricular but occasionally only the right is affected. Patients present with congestive cardiac failure; grossly impaired systolic function and an increase in the cardiothoracic ratio can be seen. Systemic embolization is common, occurring at the rate of four episodes/100 patient–years, though this risk can be minimized by oral anticoagulation. On electrocardiography there are non-specific changes with some delay of left ventricular conduction.

The diagnosis is one of exclusion, though many aetiological factors have been implicated. These include chronic severe hypertension, persistent alcohol abuse, pregnancy, and autoimmune disease triggered by viral infection, especially coxsackie virus, or previous rheumatic fever. At autopsy the heart is overweight and dilated, with no evidence of ventricular hypertrophy or coronary artery disease. There is extensive fibrosis of the ventricular wall with cellular infiltration, suggestive on occasions of myocarditis. In these cases levels of antibodies against coxsackie virus are often raised. This subgroup of patients has had a short clinical history. In most cases, however, there is no biopsy evidence of myocardial inflammation and the history is longer.

The prognosis is poor. Some 77 per cent of patients die within 2 years, but the surviving 23 per cent improve over this time; this probably reflects the heterogeneous nature of this diagnosis. Risk factors include older age (97 per cent 6-year mortality if over 55 years, compared with 66 per cent if younger), larger cardiothoracic ratio (86 per cent 6-year mortality if over 0.55 compared to 40 per cent if less), and lower cardiac index (89 per cent 6-year mortality if less than 3 l/min compared to 35 per cent if greater). Systemic embolization is a common complication, particularly in patients in atrial fibrillation (33 per cent compared to 14 per cent).

Treatment of this condition is unsatisfactory. The mainstay of management is medical treatment of the congestive failure. The use of angiotensin-converting enzyme inhibitors in this group is expanding. Oral anticoagulation should be considered, in view of the risk of embolism. Those with a short history and evidence of non-specific inflammation on biopsy may benefit from immunosuppression with steroids and azathioprine. Cardiac transplantation is the surgical option for suitable individuals.

Endomyocardial fibrosis is the most common form of restrictive cardiomyopathy. It primarily affects the endocardium, extending into the myocardium and the base of the papillary muscles. The condition mainly affects young and middle-aged men, and is most common in Central Africa, less common in South America and Asia, and rare elsewhere. The aetiology is unknown, though it has been described in many conditions associated with an eosinophilia, such as Loeffler's endomyocardial fibrosis. This eosinophilic type is more common in temperate climates, whilst in tropical climates there is less of an eosinophilic response. Pathologically the active phase shows inflammation, oedema, necrosis, and cellular infiltration, which is later replaced by organized fibrous tissue. There is restriction of ventricular filling with obliteration of the ventricular inflow by the fibrotic process and associated atrioventricular valve dysfunction. The outflow tract of the ventricles is spared.

Endomyocardial fibrosis is diagnosed angiographically and on haemodynamic criteria, though echocardiography is proving to be of increasing benefit. Angiography shows obliteration of the apical inflow portion of the affected ventricle with calcification of the affected portion and dilatation and increased contractions in the outflow portion of the ventricle. There are often associated outpouchings of the ventricular wall and filling defects due to the presence of clot. The fibrosis may involve either ventricle or both. Medical treatment aims to control congestive cardiac failure, and produces an average survival of 2 years. Surgical treatment may be palliative or corrective. In those with recurrent pericardial effusions pericardectomy, pericardial windows, and pericardioperitoneal shunts have been used with some success. If the fibrosis only involves the right ventricle, a Glenn shunt (superior vena cava to pulmonary artery) may be used to reduce the volume of blood returning to the right ventricle and to augment pulmonary flow. Dubost described stripping the endocardium from the ventricle(s) and replacing the antrioventricular valve(s). A combination of the above techniques was associated with an early operative mortality rate of 19 per cent (higher for those with biventricular disease). After periods ranging from 8 months to 8 years clinical, radiological, and haemodynamic criteria showed continued improvement.

Apart from this specific type of endocardial restriction, true restrictive cardiomyopathy with severe impairment of diastolic function occurs in patients with amyloidosis and sarcoidosis. Though these are usually systemic diseases, the former has been described limited to the heart in a familial type. In such cases cardiac transplantation remains the only option.

Pulmonary embolism is a major risk after any form of prolonged immobility, including convalescence after surgery. It is estimated that 15 to 20 per cent of deaths in acute district hospitals are due to pulmonary embolism: 20 000 patients die each year in British hospitals, while a further 40 000 suffer non-fatal events. The mechanism of development and propagation of such embolism is well recognized. Venous thrombosis occurs at the time of immobility and may affect the lower limb and abdominal venous systems.

Acute minor emboli cause tachypnoea, pleuritic chest pain, and, occasionally, mild haemoptysis. Examination is often unremarkable, and as the lungs have a very active fibrinolytic system most emboli are lysed spontaneously. Acute major embolism causes severe dyspnoea, dull central chest pain, circulatory collapse with elevation of central venous pressure, and hypoxia, hypocapnia, and acidosis. The electrocardiogram shows changes associated with right ventricular strain while the chest radiography confirms poor lung performance with oligaemic lung fields. Death is usually rapid: 60 per cent of patients die within the first 2 h. Many survivors suffer further pulmonary emboli over the next 72 h, and without treatment these are usually fatal. Chronic pulmonary embolism is a rare progression of the disease which presents with gradually increasing dyspnoea and is characterized by pulmonary hypertension secondary to vascular occlusion. The fibrinolytic system of the lungs is capable of lysing most of the clot. Subacute pulmonary embolism can be defined as an acute embolic episode occurring against a background of chronic embolic disease.

Treatment of pulmonary embolism is aimed at dispersing the existing pulmonary occlusion and preventing further emboli. The vast majority of patients are anticoagulated, initially with intravenous heparin (0.5–1 mg/kg 6-hourly). This is converted to oral anticoagulation with warfarin to maintain an INR of 2.5 to 3. The optimal duration of treatment is still not clear, but providing all risk factors have been removed it is generally agreed that 3 to 6 months of anticoagulation should be adequate.

Massive pulmonary embolism requires aggressive treatment. The mainstay of treatment, while maintaining heparinization, has recently included the increasingly more widespread use of thrombolytics. Heparin reduces the risk of death from further emboli, but does not significantly alter the obstruction to pulmonary artery bloodflow since heparin has no thrombolytic action.

Thrombolytics, such as streptokinase, urokinase, and tissue plasminogen activator, can fulfil both requirements for adequate treatment when administered systemically (250 000 units loading dose followed by 100 000 units/h infusion over the next 24 h), but their use may be associated with a high rate of bleeding complications. Emboli clear much faster on treatment with thrombolysis rather than with heparin, but though a reduction in mortality has been shown this does not reach statistical significance. There is a synergistic effect between heparin and streptokinase: a more recent approach has been to use low-dose (25 000 units loading dose and 10 000 units/hour infusion) streptokinase infused directly into the clot or into the origin of the occluded vessel in association with heparin (1000 units/h infusion initially, adjusted to achieve a partial thromboplastin time of 1.5–2). This is as effective in producing clot lysis as is high-dose systemic streptokinase, judged by clinical, angiographic, and blood gas changes with significantly fewer bleeding problems, though this complication was not prevented. It is therefore not useful in the immediate postoperative situation.

Surgery for pulmonary embolism is performed in few patients, as most die before operation can be undertaken. A subgroup of patients, however, suffer further major emboli and decompensate late. Attempts to predict this occurrence from haemodynamic data have been unsuccessful. However, patients presenting with a collapsing circulation needing cardiac or respiratory support should be referred for surgical management. Patients in whom thrombolytics are contraindicated may also be referred for surgery.

The surgical approach without cardiopulmonary bypass was first described by Trendelenburg in 1908 and was first carried out successfully in 1924. The inflow occlusion technique was reported later by Lewis. Today most patients are transferred to a cardiothoracic centre, where embolectomy is performed with cardiopulmonary bypass. This treatment is associated with a mortality rate of 29 per cent against an overall mortality of 51 per cent without bypass, though the former patients are a selected subgroup.

Further embolization occurs in 4 per cent of patients after a successful embolectomy. The indications for operating on the inferior vena cava are still debated, some arguing for intervention after all pulmonary emboli while others operate only after recurrent embolization. Procedures used involve interruption or plication of the inferior vena cava, or the insertion of filters or umbrellas. They often fail acutely due to right ventricular failure as the right ventricle is critically dependent on a high filling pressure. The overall mortality rate varies from 6 to 27 per cent: morbidity is due to severe oedema, and a 2 to 4 per cent incidence of recurrent embolization. There is little evidence that these procedures are better than heparinization. Reservation of these procedures for high-risk patients or those in whom anticoagulation is contraindicated would seem the best approach.

Since 1984 the Greenfield filter, which can be inserted percutaneously under sedation and local anaesthesia, either from the femoral or the internal jugular vein, has been developed. Insertion is associated with an operative mortality rate of 4.6 per cent. There is a 4 per cent incidence of further emboli and morbidity rates have been variably quoted as between 4.3 and 38 per cent, depending on the criteria used for assessment. Leg oedema still occurs and proceeds to ulceration in 5 per cent; 4 per cent suffer caval occlusion. Despite this morbidity, the use of this device is becoming increasingly common in high-risk patients and after recurrent embolization.

Chronic pulmonary embolic disease with pulmonary hypertension carries a poor prognosis. Prevention of further embolism is often attempted as described above, though a significant proportion still suffer repeated events. Anticoagulation may help prevent further emboli but obviously will not treat the primary disease, and thrombolysis has nothing to offer as the clot is well organized. These patients are increasingly frequently being treated by pulmonary thromboendarterectomy, and lung or heart-lung transplantation. Experience with all such procedures is limited.

The most common cardiac tumours are secondaries, usually from the lung in men and from the breast in women. Primary cardiac tumours are rare in all age groups, with the incidence being variably quoted as 1.7 to 10 per 100 000. They are mainly benign (65–80 per cent), but early detection is essential as successful resection is possible only if cardiac function is preserved. Small cardiac tumours, apart from myxomas, usually lie intramurally and cause dysrhythmias by interfering with the conduction tissue. Bradyrhythmias and tachyrhythmias may occur. Larger tumours project into the cardiac cavities and cause obstruction of the valves, the cavities, and the ventricular outflow tracts. Presentation is thus often non-specific and a high index of suspicion must be maintained: unexplained murmurs, congestive cardiac failure, and dysrhythmias may be the first signs.

Electrocardiography is usually normal or non-specifically abnormal. Chest radiography is also usually normal, though chamber enlargement and tumour calcification may be seen. Echocardiography is the best screening investigation, allowing assessment of the size, shape, location, attachment, and mobility of the tumour, as well as displacement of the coronary arteries by the tumour mass. However, echocardiography may miss large tumours in various chambers, possibly due to the relative homogeneous state of certain tumours and to the lack of normal tissue interfaces due to the large tumour size. Thus the gold standard still remains cardiac catheterization.

Treatment is by surgical resection as far as possible: when complete excision cannot be achieved, relief of the obstruction alone can give good long-term results. Enucleation or conservative resection of ventricular tumours should be undertaken. Atrial tumours can be radically excised and the defect in the atrial wall repaired with pericardium.

This is the most common tumour of the heart, comprising 50 per cent of benign tumours. It is more common in females and rare before adulthood, though it has been described in a newborn. They are usually solitary though biatrial myxomas with a common attachment at the fossa ovalis have been described, as have complex and familial forms. These last syndromes are associated with earlier development of tumours at atypical and multiple locations. Recurrences are also more common. In the complex myxoma syndrome, cutaneous lentigos, myxomas at various sites, and endocrine tumours accompany the cardiac tumour.

Cardiac myxomas usually originate in the left atrium (75 per cent); 20 per cent are seen in the right atrium and 4 per cent in the ventricles. Other sites have been reported (both atrioventricular valves, inferior vena cava, and both ventriculoarterial valves), but the stalk usually originates from the inferior rim of the fossa ovalis. Pathologically, they are soft semitranslucent tumours; histologically they contain large amounts of amorphous matrix with syncytial groups of elongated cells, closely related to thin-walled capillaries. In the past there has been some question over the origin of these tumours, some claiming that they are areas of organized thrombus rather than true neoplasms, and hence better called Lambi's excrescences. This however does not now appear appropriate.

Myxomas may present with cardiac failure due to valve obstruction or regurgitation and impingement into the cardiac chambers. This may be further manifested as pulmonary hypertension. They may also present with embolization, either systemic or pulmonary, due to thrombotic embolization or fragmentation of the tumour. Constitutional symptoms (fever, malaise, and weight loss) are also common, probably due to an autoimmune phenomenon. Physical examination is often unrewarding, though the murmur of mitral stenosis and the tumour ‘plop’ are the classic findings. Diagnosis is best achieved by echocardiography; catheterization is best avoided because of the risk of tumour fragmentation. When the diagnosis has been made, surgery should be undertaken urgently in view of the risk of embolization.

Since the first report of successful resection by Crafoord in 1954, the surgical mortality rate has fallen to 5 per cent. Care must be taken in handling the tumour to ensure that no fragments are left behind, and all four cardiac chambers must be examined, since multiple tumours may occur. Excision should be wide around the base of the stalk, as local recurrences occur and the tumour may become locally invasive. If this precaution is taken single tumours rarely recur, although this is more common in the complex syndrome, which is associated with an overall recurrence rate of 6.3 per cent. The 20-year survival is 91 per cent, but occasionally the tumours may metastasize to brain and bone.

Eighty-five per cent of these tumours occur in infants and children, usually in the free wall of the left ventricle or the interventricular septum. They comprise 5 per cent of primary cardiac tumours and are usually single, round, firm, well-encapsulated masses, with a false capsule that allows enucleation. They are entirely benign and show no malignant transformation. On cutting, the surface bulges convexly and the fibrous tissue intermingled with muscle has a whorled appearance. Occasionally they calcify. They often interfere with the conduction system so that sudden death is common, occurring in 30 per cent of patients.

This tumour is primarily a tumour of childhood. Fifty per cent of patients die within the first month of life and 80 per cent by 1 year. In 50 per cent of patients it is associated with tuberosclerosis. Rhabdomyomas are often multiple in the interventricular septum and adjacent free wall; diffuse myocardial disease leads to heart failure and death. The tumours are usually well-circumscribed, non-encapsulated, whitish nodules, and they may grow to a large size, projecting into the ventricular cavities. The individual cells are large and vacuolated, with delicate radiating strands of cytoplasm. They are filled with glycogen. Whether they are true neoplasms or simply overgrowths of the Purkinje cells and thus hamartomas remains uncertain, but spontaneous regression has been described. They are certainly congenital, as intrauterine diagnosis has been confirmed at neonatal operation.

This very uncommon tumour usually occurs in adults. Most commonly it is located subendocardially and appears as a yellow mass projecting into the cardiac chamber. Rarely it is intramural and unencapsulated. It is slow growing and may calcify following fat necrosis. Microscopically it comprises fatty tissue with interlacing muscle fibres.

Haemangiomas and benign teratomas occur infrequently. Primary malignant tumours are extremely rare: they are usually haemangiosarcomas and small cell sarcomas. Early death is usually due to rapid growth and dissemination.

Historically, effusive pericarditis and constrictive pericarditis were attributed to mycobacterial infection, but other factors, such as infection, metabolic, trauma, infiltration, radiation, autoimmune disease and idiopathic causes, are now more common. The diagnosis can be confirmed by echocardiography; nuclear magnetic resonance imaging allows the aetiology to be ascertained and the differential diagnosis of restrictive cardiomyopathy to be excluded. Cardiac catheterization shows an elevated diastolic plateau between the pressures in all chambers and prominent ‘x’ and ‘y’ descents in the right atrial pressure tracing.

Infection with bacteria (usually Mycobacterium tuberculosis, Staphylococcus aureus, and Escherichia coli) fungi, parasites, and viruses may all cause acute pericarditis which can develop into constrictive pericarditis. This is most common in nutritionally, immunologically, or traumatically compromised patients. There is a high mortality rate associated with pyogenic pericarditis (77 per cent); after diagnosis by pericardiocentesis treatment with appropriate antibiotics and immediate pericardectomy should be undertaken.

Uraemic pericarditis has become less of a problem with the establishment of long-term haemodialysis. The incidence in patients on haemodialysis is 20 per cent; in 61 per cent the condition develops within 6 months of starting dialysis. Medical management with aggressive haemodialysis, pericardiocentesis, and steroids is successful in 50 per cent of patients. Surgery should be reserved for those with acute tamponade, unresponsive or recurrent effusions, or those in whom constriction occurs. Except when the effusion is posterior or constriction has occurred, an anterior pericardectomy or formation of a large pericardial window gives excellent results with minimal morbidity and mortality.

Constrictive pericarditis may occur following cardiac surgery. The incidence is between 0.15 and 0.3 per cent and it occurs from 2 to 40 months after surgery. The mechanism of injury is unknown but both visceral and parietal layers are affected by inflammation and infiltration. It may be caused by air-drying of the serosal layer or chemical exposure; outbreaks have been seen following the use of povidone-iodine solution to wash out the pericardium at the end of surgery. Cold injury or the persistence of organized clot may also be a cause. It is in all respects typical constrictive pericarditis with the usual angiographic findings, and although medical therapy with diuretics, angiotensin converting enzyme inhibitors, steroids, and non-steroidal anti-inflammatory drugs may be attempted, further surgery may be required.

Neoplasms infiltrating the heart are most commonly breast and lung carcinoma, lymphoma, and leukaemia. Except for lung carcinoma constrictive pericarditis it is not usually a terminal event and prolonged survival may be expected. Treatment of effusive pericarditis should initially be by pericardiocentesis plus chemotherapy (both systemic and by instillation directly into the pericardium) and radiotherapy; pericardectomy is reserved for those with good survival prospects or in whom medical treatment fails or constriction develops. Constrictive pericarditis may also occur 1 to 4 years after radiation treatment, usually when more than 45 Gy have been used on the mediastinum.

Numerous surgical approaches have been described for pericardectomy, including median sternotomy, anterolateral thoracotomy, and bilateral thoracotomy. The extent of the decortication is however crucial, and radical pericardectomy is essential: pericardium overlying all intrapericardial structures including the atria and venae cavae is removed, and all ventricular surfaces including the anterolateral and diaphragmatic surfaces are decorticated. The depth of decortication must be adequate: failure of surgery may be due to failure to remove a constrictive epicardial peel. In some of these cases there is a definite plane between the two layers. Surgery to remove the epicardial peel carries an increased risk of disrupting the coronary vessels, and it may prove more prudent to cross-hatch the pericardium, allowing the ventricle to bulge through the gaps. Cardiopulmonary bypass is generally only required if other cardiac lesions are being corrected at the same operation but should be available on standby, and used if the posterior surface is affected.

Pericardectomy may produce an immediate or a delayed improvement. Incomplete pericardectomy due to retention of an epicardial peel is occasionally blamed for a postoperative low output state, but it may also be caused by a progressive deterioration in cardiac function due to myocardial atrophy. In support of this a low output state is associated with preoperative disability, and there is often excessive dilatation of the ventricle at the end of surgery. In such cases the improvement following surgery may be protracted.

Kawasaki's disease, first described in 1967, is an acute, multisystem vasculitic disorder affecting children and infants. It has a predilection for Japanese and other Asian populations, though it has been reported in all racial groups and particularly in young children, 50 per cent of cases being diagnosed by the age of 2 years and 80 per cent by the age of 4. The male to female ratio is 1.6 : 1.0. The aetiology is unknown, though an infectious agent is suspected due to its occurrence of epidemics in different locations. Direct spread, however, has not been proven. Clinically and pathologically there is a spectrum of severity from incomplete forms to aggressive forms which resemble infantile polyarteritis nodosa. In all cases there are prominent vasculitic features. The diagnostic criteria are given in Table 1 526. Incomplete forms are, however, common, especially in children under 6 months. These may be associated with complications as severe as in those patients who fulfil all diagnostic criteria for the disease. A high index of suspicion is therefore essential in children with a fever of unknown aetiology; as the disease is progressive, frequently repeated examinations are required.

All layers of the heart are affected. Pericarditis is common in the acute phase of the disease and may be associated with effusions. It is usually transient and settles without long-term complications. Myocarditis and myocardial fibrosis are detectable clinically in 30 per cent of patients, but are present in all patients on routine biopsy. There are slight, non-specific electrocardiographic changes but no rise in cardiac enzyme levels, suggesting that the dysfunction is due more to infiltration rather than myolysis. Diagnosis is therefore difficult, but clinical evidence of congestive cardiac failure and echocardiographic demonstration of impaired left ventricular function are probably the most reliable indicators. The endocardium is also affected, and valvular dysfunction occurs in both acute and subacute phases of the disease. Ten per cent of patients show leaflet inflammation or dilatation of the annulus due to myocarditis or myocardial ischaemia. Aortic regurgitation is less common (5 per cent), and the tricuspid valve is rarely involved. Echocardiographic monitoring of these individuals has shown that all valvular abnormalities resolve spontaneously.

The most dangerous of all complications is arterial disease, initially aneurysmal and later occlusive. This occurs in the coronary arteries as well as in the peripheral circulation, and starts most commonly in the subacute stage of the disease in those who suffered a more acute phase. Arterial disease is also more common in younger patients and in those treated with steroids. The administration of gammaglobulin reduces the incidence of coronary abnormalities in these patients, and attempts to identify those at risk have been made. Echocardiographic evidence suggests that pericardial effusions or mitral regurgitation in the acute phase are reproducible predictors of later coronary disease. Aortic regurgitation likewise is seen in all patients with coronary lesions. There is, however, no evidence that treating these patients with gammaglobulin in the acute phase reduces the risk of aneurysms. Aneurysms may be fusiform or saccular, and a high incidence of thrombosis leads to myocardial infarction in the subacute or convalescent phases. Two-dimensional echocardiography demonstrates 92 per cent of aneurysms shown on angiography and is the method of choice for screening. The investigation should be repeated every 4 weeks in patients shown to have aneurysms until they have all disappeared, usually within 6 months to 2 years of the onset of Kawasaki's disease. The use of angiography can be restricted to those with a history of ischaemia, persistent mitral regurgitation, coronary calcification on plain chest radiography, or persistent aneurysms on echocardiography.

Treatment of this disease involves frequent, thorough physical examination of the child, regular echocardiography to detect the development of cardiac complications, and drug therapy. Those with aneurysms of the coronary arteries are at greatest risk of developing occlusive disease: echocardiographic assessment of these individuals should continue until all aneurysms have disappeared. Aspirin and dipyridamole are claimed to decrease the risk of sudden death due to the thrombosis of these aneurysms chronically, while thrombolytics are used in the acute phase. However, stenoses of the coronary arteries develop at the sites of previous aneurysms, and long-term treatment with aspirin and regular exercise testing is therefore advocated. Steroids produce no benefit in this disease and increase the risk of aneurysm formation. Administration of gammaglobulin reduced the incidence of coronary artery abnormalities, and early identification of patients at risk may therefore allow a reduction in this severe complication. Though not applicable in the acute phase of the disease, conventional coronary artery surgery may be required in those with persistent aneurysms or those who develop stenoses at the point of previous aneurysms in the convalescent or chronic stage of the disease.

Anderson DM, Raff GL, Ports TA, Brundage BH, Parmley WM, Chatterjee K. Hypertrophic obstructive cardiomyopathy. Effects of acute and chronic verapamil treatment on left ventricular systolic and diastolic function. Br Heart J, 1984; 51: 523–9.

Cherian G, et al. Endomyocardial fibrosis: report on the haemodynamic data in 29 patients and review of the results of surgery. Am Heart J, 1983; 105: 659–65.

Fighali S, Krajcer Z, Leachman RD. Septal myomectomy and mitral valve replacement for idopathic hypertrophic subaortic stenosis: short and long term follow-up. J Am Coll Cardiol, 1984; 3: 1127–34.

Fuster V, Gersh BJ, Guilaini ER, Tajik AJ, Brandenburg RO, Frye RL. The natural history of idopathic dilated cardiomyopathy. Am J Cardiol, 1981; 47: 525–31.

Gill CC, Duda AM, Kirazume H, Kramer JR, Loop FD. Idiopathic hypertrophic subaortic stenosis and coronary atherosclerosis. Results of coronary artery bypass alone and myomectomy combined with coronary artery bypass. J Thoracic Cardiovasc Surg, 1982; 84: 856–60.

McKenna W, Deanfield J, Faruqui A, England D, Oakley C, Goodwin J. Prognosis in hypertrophic cardiomyopathy: role of age and clinical, electrocardiographic and haemodynamic features. Am J Cardiol, 1981; 47: 532–8.

Maron BJ, Epstein SE. Hypertrophic cardiomyopathy; recent observations regarding the specificity of three hallmarks of the disease: asymmetric septal hypertrophy, septal disorganisation, and anterior systolic motion of the anterior mitral leaflet. Am J Cardiol, 1980; 45: 141–53.

Spirito P, Maron BJ, Rosing DR. Morphologic determinants of hemodynamic state after ventricular septal myotomy-myomectomy in patients with obstructive hypertrophic cardiomyopathy: M mode and two-dimensional echocardiographic assessment. Circulation, 1984; 70: 984–95.

Pulmonary embolism

Golueke PJ, Garrett WV, Thompson JE, Smith BL, Talkington CM. Interruption of the vena cava by means of the Greenfield filter: expanding the indications. Surgery, 1988; 103: 111–7.

Gray HH, Miller GAH, Paneth M. Pulmonary embolectomy: its place in the management of pulmonary embolism. Lancet, 1988; i: 1441–5.

Leeper KV, et al. Treatment of massive acute pulmonary embolism: the use of low doses of intrapulmonary arterial streptokinase combined with full doses of systemic heparin. Chest, 1988; 93: 234–9.

Cardiac tumours

Arciniegas E, Hakimi M, Farooki ZQ, Truccone NJ, Green EW. Primary cardiac tumours in children. J Thoracic Cardiovasc Surg, 1960; 79: 582–91.

Attar S, Lee Y, Singleton R, Scherlis L, David R, McLaughlin JS. Cardiac myxoma. Ann Thoracic Surg, 1980; 29: 397–405.

Bortolotti U, et al. Surgical excision of intracardiac myxomas: a 20 year follow up. Ann Thoracic Surg, 1990; 49: 449–53.

Bulkley BH, Hutchins GM. Atrial myxomas: a fifty year review. Am Heart J, 1979; 97: 639–43.

Come PC, Riley MF, Markis JE, Malagold M. Limitations of echocardiographic techniques in evaluation of left atrial masses. Am J Cardiol, 1981; 48: 947–53.

Corno A, de Simone G, Catena G, Marcelletti C. Cardiac rhabdomyoma: surgical treatment in the neonate. J Thoracic Cardiovasc Surg, 1984; 87: 725–31.

Lazzara RR, Park SB, Magovern GJ. Cardiac myxomas: results of surgical treatment. J Cardiovasc Surg, 1991; 32: 824–7.

Silverman NA. Primary cardiac tumors. Ann Surg, 1980; 191: 127–38.

Wold LE, Lie JT. Cardiac myxomas: a clinicopathological profile. Am J Pathol, 1980; 101: 219–40.

Pericarditis

Culliford AT, Lipton M, Spencer FC. Operation for chronic constrictive pericarditis: do the surgical approach and degree of pericardial resection influence the outcome significantly. Ann Thoracic Surg, 1980; 29: 146–52.

Frame JR, Lucas SK, Pederson JA, Elkins RC. Surgical treatment of pericarditis in the dialysis patient. Am J Surg, 1983; 146: 800–3.

Kutcher MA, King SB, Alimurung BN, Craver JM, Logue RB. Constrictive pericarditis as a complication of cardiac surgery: recognition of an entity. Am J Cardiol, 1982; 50: 742–9.

McCaughan BC, et al. Early and late results of pericardectomy for constrictive pericarditis. J Thoracic Cardiovasc Surg, 1985; 89: 340–50.

Miller JI, Mansour KA, Hatcher CR. Pericardectomy: current indications, concepts and results in a university center. Ann Thoracic Surg, 1982; 34: 40–5.

Soulen RL, Stark DD, Higgins CB. Magnetic resonance imaging of constrictive pericardial disease. Am J Cardiol, 1985; 55: 480–4.

Sutton FJ, Whitley NO, Applefield MM. The role of echocardiography and computed tomography in the evaluation of constrictive pericarditis. Am Heart J, 1985; 109: 350–5.

Walsh TJ, Baughman KL, Gardner TJ, Bulkley BH. Constrictive epicarditis as a cause of delayed or absent response to pericardectomy; a clinicopathological study. J Thoracic Cardiovasc Surg, 1982; 83: 126–32.

Kawasaki's disease

Burns JC, et al. Clinical spectrum of Kawasaki disease in infants younger than 6 months of age. J Pediatr, 1986; 109: 759–63.

Gidding SS, Duffy E, Pajcic S, Berdusis K, Shulman ST. Usefulness of echocardiographic evidence of pericardial effusion and mitral regurgitation during the acute stage in predicting development of coronary arterial aneurysms in the late stage of Kawasaki disease. Am J Cardiol, 1987; 60: 76–9.

Hicks RV, Melish ME. Kawasaki syndrome. Pediatr Rheumatol, 1986; 33: 1151–73.