Intrapericardial Treatment of Neoplastic Pericardial Effusions
STUDY OBJECTIVES To evaluate the acute effect of minocycline on the pericardium in the experimental animal and in the human with malignant pericardial disease. DESIGN A prospective study in open-chest dogs and in humans. SETTING Experimental surgery laboratory, medical school; coronary care unit, university hospital. METHODS Twenty-three open-chest dogs were divided into four groups according to the solution injected intrapericardially: (1) minocycline, 5 mg/kg; (2) minocycline, 10 mg/kg; (3) normal saline solution, 100 mL, followed by minocycline, 10 mg/kg; (4) a mixture of 50 mL of the dog's own blood mixed ex vivo with minocycline, 10 mg/kg to evaluate the effect of rising pH of minocycline solution. The extent of myocardial injury is evaluated by measuring ST-T segment deviation in six standard bipolar leads and in three unipolar electrograms recorded over the left ventricular pericardial surface. The pH of the various minocycline solutions is measured. Nine consecutive patients with malignant cardiac tamponade receiving minocycline intrapericardially are evaluated for the appearance of chest pain and ECG changes. RESULTS Minocycline (5 and 10 mg/kg) caused marked, transient ST-T segment deviation in all dogs, whether or not saline solution was previously injected into the pericardial sac. Prior mixing of minocycline with blood markedly increased the acidic pH of the minocycline solution and significantly reduced the extent of ST-T segment deviation. Four of nine patients had chest pain during minocycline injection. None had ST-T segment changes. CONCLUSION Minocycline causes a marked, transient injury to the epicardial-pericardial surface. Our animal and in vitro studies indicate that this acute injury is probably partly related to the acidic pH of the minocycline solution. Our experimental findings suggest that this minocycline-induced injury may be reduced by raising the pH of the solution either ex vivo (eg, by mixing minocycline with previously withdrawn pericardial fluid) or in vivo (eg, by leaving 200 to 300 mL of pericardial fluid prior to minocycline injection). Limited experience in the human with malignant cardiac tamponade indicates that intrapericardial minocycline is usually well tolerated, although severe chest pain may appear.