Comparison of open subxiphoid pericardial drainage with percutaneous catheter drainage for symptomatic pericardial effusion

BACKGROUND: The optimal therapy for symptomatic pericardial effusions remains controversial. This paper compares outcomes after the two most commonly used techniques, percutaneous catheter drainage and operative subxiphoid pericardial drainage. METHODS: We performed a 5-year retrospective, single-institution study to analyze outcomes after either percutaneous catheter drainage or subxiphoid open pericardial drainage for symptomatic pericardial effusions. RESULTS: Symptomatic pericardial effusions in 246 patients were treated by open pericardiotomy and tube drainage (n = 150) or percutaneous catheter drainage (n = 96). Drainage duration, total drainage volume, and duration of follow-up (2.6 years) were similar in both groups. Effusions were classified malignant in 79 (32%) patients and benign in 167 (68%) patients. No direct procedural mortality occurred, but the hospital mortality was 16 patients (10.7%) in the open group and 22 (22.9%) in the percutaneous group (p = 0.01) The 5-year survival rate was 51% in the open group versus 45% in the percutaneous group, despite a greater percentage of the open group having a preoperative malignant diagnosis (35% versus 28%). Symptomatic effusions recurred in 16.5% of the percutaneous group compared with 4.6% in the open group (p = 0.002), and sclerosis did not appear to reduce recurrence rates (10.7% with sclerosis versus 15.6% without; p > 0.05). The diagnosis of malignancy was confirmed in 16 of 27 (59%) percutaneous procedures performed on patients with known malignancy. In the open group, cytologic and pathologic evaluation of the pericardial specimen revealed malignancy in 32 of 52 (62%) patients with known malignancy. CONCLUSIONS: Subxiphoid and percutaneous pericardial drainage of symptomatic pericardial effusions can be performed safely; however, death occurs from underlying disease. Open subxiphoid pericardial drainage with pericardial biopsy appears to decrease recurrence but does not improve diagnostic accuracy of malignancy over cytology alone.