How Can We Help?
Pericardial disease.
Recognition of pericardial syndromes.
Consider the diagnosis if there is a large heart shadow and clear lung fields. There is often an associated pleural effusion that may be either exudative or transudative.
Other useful clues that this is acute pericardial disease rather than myopathic cardiac failure are:
- Central pleuritic chest pain
- A history of drenching night sweats
- Fever
- A history of perceived loss of weight. Although fluid accumulation may have resulted in net gain in weight, the catabolism associated with tuberculosis may cause facial thinning that is noticed by the patient and relatives.
- A pericardial friction rub. In order to say you have listened for this you need to try both with the patient supine and also sitting forward in inspiration.
The presence of a significant paradoxus is useful in distinguishing tamponade from a CMO, but is not of value for a large effusion without tamponade.
The ECG may demonstrate electrical alternans (varying QRS height with beat to beat variation in heart position within a bag of fluid):
The key clue on the CXR is the presence of a large cardiac shadow with clear lung fields – most individuals with cardiomyopathy at the late stage leading to an increased cardiac shadow will present in pulmonary oedema with clear congestion on the CXR.
Perspective – tamponade versus cardiomyopathy
A study from Malawi1 looked at differentiating tamponade from cardiomyopathy.
Signs felt by that group to be of value were a paradoxical rise in JVP on inspiration (sensitivity 24%,, specificity ‘100%’), a paradoxus of more than 10 mmHg (sens 52%) and a rub, which last was only detected in 4% but is also highly specific. All of these had excellent positive predictive values. An impalpable apex is also helpful (sens 84%, spec 0.96, LR+ 21.0, LR- 0.17.) Few signs had very good negative predictive values (i.e. excluded the diagnosis) although the absence of a murmur or a palpable apex were fair indicators:
(A paradoxus can be undetectable in spite of acute tamponade in patients with severe hypotension, those with right-sided pericardial adhesions, marked LVH, ASD or severe aortic regurgitation.)2
Perspective – diagnosis of acute tuberculous pericarditis
In a setting of high prevalence of both tuberculosis and HIV infection, a tuberculous aetiology in a patient with pericarditis should be considered likely, although purulent pericarditis (e.g. Salmonella non-typhi) and Kaposi’s, along with other rarer causes, are possibilities. The finding of tuberculosis at other sites is clearly helpful.
A combination of the clinical criteria of weight loss, fever and either a WCC < 10 x 109/l or a serum globulin (total protein minus albumin) > 40 g/l in a setting of high prevalence of both conditions had a sensitivity of 82% and specificity of 76% in a study using a selected group of patients with large effusions presenting to an academic centre.3
The same study found that in HIV positive individuals with a WCC < 10 x 109/l, the combination of a pericardial ADA > 40U/l with the finding of a lymphocytic predominance in the fluid (lymphs:polys>1) had a sensitivity of 96% and specificity of 97% (pericardial fluid criteria). Unlike earlier studies in predominantly HIV negative patients, bedside innoculation of pericardial fluid was only positive in 52%, and ZN stain, as in pleural effusions was positivity in a minority (2%).
|
Pre-test probability |
0.01 |
0.05 |
0.1 |
0.25 |
0.5 |
0.75 |
0.9 |
0.95 |
0.99 |
|
Revised probability | |||||||||
|
Pericardial fluid criteria positive |
0.24 |
0.63 |
0.78 |
0.91 |
0.97 |
0.99 |
1.00 |
1.00 |
1.00 |
|
Pericardial fluid criteria negative |
0.00 |
0.00 |
0.00 |
0.01 |
0.04 |
0.11 |
0.26 |
0.43 |
0.80 |
Management of acute tuberculous pericarditis.
The issue of using steroids in the management of tuberculous pericarditis has become clearer with the publication of the IMPI trial, where 1400 patients, two thirds of whom were HIV positive, were randomised to prednisolone or placebo. There was no statistically significant difference in the primary endpoint (a composite of death, tamponade requiring pericardiocentesis, or constriction.) When analysed individually, although there was no difference in mortality or tamponade, constriction was reduced from 7.8% to 4.4% (ARR 3.4%, NNT 30, 95% CI 17 to 123) and rate of hospitalisation from 25.2 to 20.7% (hazard ratio, 0.79; 95% CI, 0.63 to 0.99; P = 0.04.) This gain was at the cost of an increased HIV-related cancer rate, from 0.1% to 1.3% (NNH 85, and this reached statistical significance.)4
So it is currently difficult to justify the use of corticosteroids for this indication, particularly in HIV positive individuals, as for approximately every 3 constrictions prevented, one new HIV-related cancer will develop.
Constrictive pericarditis.
In resource constrained environments this is usually due to tuberculosis although occasional patients may have had radiotherapy as the cause. Patients commonly present like cor pulmonale without cyanosis or a right ventricular lift. Leg oedema may not be that striking. CXR classically shows a small cardiac shadow in the face of fluid overload. The key features on examination are a bit subtle and are often overlooked, so if in doubt, consider doing pulse oximetry – the well-saturated individual with gross peripheral oedema, a raised JVP and no evidence of nephrotic syndrome or chronic liver disease should certainly be considered to be a possible candidate. The CXR may show pericardial calcification if you request a well-penetrated lateral. Treatment is pericardiectomy, usually preceded by several months of TB treatment, as surgery is rumoured to cause flaring of quiescent disease
Round A, Hamilton W. Clinical diagnosis of tamponade in Malawi. Quarterly J Med. 1990;279:763-6 ↩
Spodick DH. Acute cardiac tamponade. N Engl J Med. 2003;349:684-90 ↩
Reuter H, Burgess L, van Vuuren W, et al. Diagnosing tuberculous pericarditis. QJM. 2006;99:827-39 ↩
Mayosi B, Ntsekhe J, Bosch S, et al.Prednisolone and Mycobacterium indicus pranii in tuberculous pericarditis. New Engl J Med. 2014;371:1121-1130 ↩