No abstract

The clinical features with particular reference to tamponade and mediastinal adenopathy were studied in tuberculous pericardial effusion. Tamponade is a frequent complication and the recognition of tuberculous etiology can be difficult. Involvement of the pericardium is mostly from mediastinal lymph nodes that have not been studied. This was a prospective cohort study. All patients had large effusions, and underwent pericardiocentesis and chest computed tomography. Patients with tuberculosis had specific therapy. Others with viral/idiopathic effusion served as controls for the computed tomography studies. There were 26 patients with tuberculosis: 18 had tamponade on echocardiography. All had symptoms. Fever (n = 23) and dyspnea (n = 20) were the most frequent presenting symptoms. Pericardial rub was heard in 14, and 3 had enlarged cervical or axillary nodes. Pulmonary tuberculosis was present in 6. Tuberculin skin test measured 17 +/- 3.3 mm. The biopsy specimen showed a granuloma in 22 of 24. All 26 had mediastinal lymph nodes > 10 mm with a mean size of 19.5 +/- 8.6 mm that disappeared (81%) or regressed (19%) on treatment (p < 0.001). Aortopulmonary nodes were most frequently enlarged (65.4%) and hilar the least. Three required pericardiectomy. At follow-up all were doing well. None with viral/idiopathic effusion had lymph node enlargement. Fever, dyspnea, and tamponade were frequent with tuberculosis. The prognosis was good with specific therapy. Mediastinal nodes were enlarged in all and only with tuberculosis and not with viral/idiopathic effusion. Nodes disappeared or regressed with treatment. In the appropriate clinical context, mediastinal lymph node enlargement on chest computed tomography along with a strongly positive skin test results could help in the diagnosis of a tuberculous etiology of pericardial effusion.

There are varying reports on the electrocardiogram in pericardial effusions. Some correlate low QRS voltage with tamponade and the size of the effusion while others do not. Low voltage also appears to vary with the etiology. There are no reports on the influence of pericardial thickness or changes in the P voltage. The authors studied 43 patients with large effusions of whom 26 had tuberculosis and the remaining had viral/idiopathic etiology. Pericardial thickness was measured at chest computed tomography. They found no correlation between the low QRS voltage and tamponade, size of the effusion, etiology, or pericardial thickness. Low voltage of the P wave and T-wave changes were more frequent than low QRS voltage.