Hypothyroid cardiac tamponade: clinical features, electrocardiography, pericardial fluid and management.

摘要

BACKGROUND: Cardiac tamponade associated with hypothyroidism has been reported, but few studies address the clinical features. This study aims to identify the patient characteristics and to propose a reasonable clinical approach for hypothyroid cardiac tamponade. METHOD: Patients admitted to Chang Gung Memorial Hospital between September 1998 and September 2008 with pericardial effusion secondary to hypothyroidism were enrolled. Cases involving cardiac tamponade were investigated. The clinical data, electrocardiography, echocardiography and aspirated fluid were examined. RESULTS: Thirty-six patients with moderate or large amount of pericardial effusion as a result of hypothyroidism were examined. Eight patients (22.2%) with both clinical and echocardiographic signs of tamponade were identified and were treated by pericardiocentesis or creation of pleural-pericardial window. These patients were characterized with normal sinus rhythm (80.75 +/- 13.45 beats/min), low voltage over limb leads (6 of 8, 75.0%), flat T wave (6 of 8, 75.0%) and clear yellowish pericardial fluid (7 of 8, 87.5%). Their heart rates were significantly lower (80.75 +/- 13.45 beats/min versus 112.75 +/- 12.87 beats/min, P < 0.01) than those of patients with cardiac tamponade from malignancy, autoimmune disease, tuberculosis and iatrogenic insult. Their interventricular septa were also significantly thicker (15.71 +/- 6.70 mm versus 11.70 +/- 2.11 mm, P = 0.02). Ten patients (27.8%) had echocardiographic signs of tamponade without paradoxical pulse and were successfully treated with thyroxine without pericardial drainage. CONCLUSION: For patients diagnosed with cardiac tamponade without sinus tachycardia, hypothyroidism should be highly suspected. Although emergent pericardiocentesis should be performed in clinical cardiac tamponade, patients with echocardiographic tamponade signs without a paradoxical pulse should be treated with thyroxine initially.