Clinical Review Lesson of the week

Restrictive-constrictive heart failure masquerading as liver disease

BMJ 1999; 318 doi: (Published 27 February 1999) Cite this as: BMJ 1999;318:585
  1. Martin D Lowe, specialist registrar,
  2. Alun A Harcombe, specialist registrar,
  3. Andrew A Grace, consultant,
  4. ichael C MPetch, consultant
  1. Department of Cardiology, Papworth Hospital, Cambridge CB3 8RE
  1. Correspondence to: Dr Petch
  • Accepted 26 June 1998

Restrictive cardiomyopathy or pericardial constriction should be considered as a cause of hepatomegaly before liver biopsy is undertaken

The clinical features of restrictive or constrictive heart disease can be non-specific, 1 2 and this may delay diagnosis. We describe three patients with heart disease whose initial presentation led to a wrong diagnosis of primary hepatic disease. Potentially hazardous investigations, including liver biopsies, were performed or planned in each case. The true diagnosis was usually prompted by further clinical examination, and all three patients responded well to appropriate treatment. The importance of making the correct diagnosis is emphasised by the potential for surgical cure in these patients.

Case reports

Case 1

An 18year old man presented with a five month history of fatigue, anorexia, abdominal distension, and dyspnoea on exertion. Physical examination showed that he was icteric, with an enlarged liver and ascites. Findings at cardiovascular examination were reported as normal, although his jugular venous pressure was not seen. Liver ultrasound and computed tomography showed diffuse hepatic enlargement, and dilated venous sinuses and septal fibrosis were seen in a liver biopsy specimen. The diagnosis remained unclear, but a raised jugular venous pressure and gallop rhythm were detected at further clinical examination. An electrocardiogram showed widened QRS complexes (>140 msec) and diffuse, non-specific repolarisation changes, and chest radiography showed cardiomegaly and pulmonary oedema. There was no evidence of pericardial thickening in thoracic computed tomograms, and at cardiac catheterisation the haemodynamic profile of a restrictive cardiomyopathy was found. The patient was treated with diuretic drugs and subsequent cardiac transplantation was successful.

Case 2

A 49year old man had an 18day history of epigastric discomfort, abdominal distension, and increasing shortness of breath. Physical examination showed that he had hepatomegaly, ascites, and ankle oedema. Apart from a resting tachycardia, findings at cardiovascular examination were normal. Liver function tests were abnormal, and abdominal paracentesis showed an exudate. A liver biopsy was planned, but before this could be performed pericardial calcification was found incidentally on an abdominal ultrasound scan and a second opinion was obtained. On further examination, the patient's jugular venous pressure was raised to the angle of the jaw and an early diastolic sound was noted. An electrocardiogram showed normal QRS voltages, but widespread T wave flattening, and a chest radiograph showed cardiomegaly. Noticeable pericardial thickening, confirming the clinical diagnosis of constrictive pericarditis, was seen on thoracic computed tomogram. Pericardectomy was performed and led to a complete resolution of symptoms.

Case 3

A 51year old woman had a nine month history of epigastric discomfort, abdominal swelling, and weight loss. At physical examination she seemed unwell, with hepatomegaly, ascites, and peripheral oedema, but cardiovascular examination showed no abnormalities. Results of liver function tests were abnormal, and an ultrasound scan confirmed diffuse hepatic enlargement. A chest radiogram showed cardiomegaly and bilateral pleural thickening. Expansion of the portal tracts with lymphocytic and neutrophilic infiltrates was seen in a liver biopsy specimen, while non-specific inflammatory changes were detected in a pleural biopsy specimen. A grossly raised jugular venous pressure, pulsus paradoxus, and gallop rhythm were detected at a later clinical examination. An electrocardiogram showed low voltage ventricular complexes and echocardiography revealed a large pericardial effusion. Pericardiocentesis and pericardectomy were performed; the pericardium was found to be grossly thickened and inflamed. No cause for the patient's disease was established, but she has made a full recovery.


These patients illustrate the difficulties encountered when constrictive or restrictive cardiac abnormalities are not considered in patients with hepatomegaly. All three patients had a high jugular venous pressure and an abnormal electrocardiogram and chest radiogram, although postero-anterior radiograms did not show pericardial calcification in those with constriction. The possibility of a constrictive or restrictive cardiac abnormality was not considered initially, and inappropriate investigations were performed before the true diagnosis became clear.

Hepatosplenomegaly and ascites are often present in patients with pericardial constriction or restrictive cardiomyopathy, and may suggest primary liver disease.3 Indeed, constrictive pericarditis has presented as hepatic coma.4 However, compared with patients with cirrhosis, prominent liver pulsations are reported in up to 70% of those with constriction,5 and the protein content of ascitic fluid is often high.6 Despite previous reports,710 and as illustrated here, potentially hazardous investigations such as liver biopsy11 may be undertaken in these patients. Histological findings are usually non-specific—diffuse centrilobular congestion, necrosis, and fibrosis are the most common, but there may be mild abnormalities such as patchy fibrosis without congestion.12 Patients with appreciable sinusoidal dilatation and haemorrhagic necrosis together with hepatic venular thrombus may be mistakenly diagnosed as having the Budd-Chiari syndrome. 8 9

The most important physical sign in pericardial constriction or restrictive cardiomyopathy is a raised jugular venous pressure, but this may be missed, particularly when it is raised above the angle of the jaw. The jugular venous pressure is usually measured with the patient lying at an angle of 45° to the horizontal, but when the venous pressure is grossly raised, a 90° sitting position may be preferable. Further pointers to making a correct diagnosis from the clinical examination include pulsus paradoxus and added heart sounds. Pulsus paradoxus is commonly found in effusive-constrictive pericarditis,13 where the presence of pericardial fluid in addition to visceral pericardial constriction results in a reduction of left ventricular stroke volume during inspiration. The “pericardial knock,”14 characteristic of constriction, occurs earlier in diastole and is of higher frequency than the third heart sound with restriction.

In constriction, the chest radiogram shows calcification of the pericardium, usually seen best on a lateral projection, in around half the cases.15 The heart size is usually normal or reduced, but is occasionally enlarged, and pulmonary congestion is sometimes present. An electrocardiogram may show atrial fibrillation, low QRS voltages, T wave flattening or inversion, and P mitrale16; similar changes are found in restrictive cardiomyopathy.1 Echocardiographic features of constrictive pericarditis include a thickened pericardium, abnormal ventricular septal motion, and appreciable respiratory variation in Doppler inflow velocities and isovolumic relaxation times17; increased wall thickness and a granular looking myocardium suggest restrictive heart muscle disease. Computed tomography and magnetic resonance imaging are particularly helpful in diagnosing constriction, 18 19 but left and right heart catheterisation may still be required to confirm primary cardiac disease and help distinguish restrictive cardiomyopathy from constrictive pericarditis.18 If haemodynamic studies are unable to establish the diagnosis, imaging and myocardial biopsy may be needed.

Tuberculosis remains an important cause of constrictive pericarditis, but other aetiologies including cardiac surgery, connective tissue disease, and malignancy have become increasingly prominent in recent years.20 In most cases, however, no specific cause is identified, and the condition is attributed to past subclinical viral pericarditis.21 Restrictive cardiomyopathy is also usually classified as idiopathic, although amyloidosis and other infiltrative diseases are recognised secondary causes.1 It is important to recognise restrictive and constrictive heart failure because these conditions can be treated successfully by heart transplantation and pericardectomy respectively. Systemic disease, requiring appropriate adjunctive treatment, may also be identified.

In conclusion, cardiac disease should be considered in the differential diagnosis of hepatomegaly, and the jugular venous pressure should be examined with care. Routine investigations such as chest radiography and electrocardiography should be performed before liver biopsy in all patients with unexplained liver disease. Echocardiography should be undertaken if there is any suspicion of heart disease. Liver histology may be abnormal if there is a cardiac abnormality; non-specific findings should be interpreted with particular care and should not be seen as confirmation of primary hepatic disease.


MDL is a British Heart Foundation junior research fellow and AAG is a Britsh Heart Foundation senior research fellow.

Contributors: MCP was the consultant responsible for the cardiological care of the patients described and is guarantor for the paper. MDL, AAH, and AAG collected data on each of the patients. All authors contributed to the writing of the manuscript.


View Abstract

Log in

Log in through your institution


* For online subscription