Association of etiology of pericardial fluid with mortality

Background

Etiologies of pericardial effusions vary and frequently reflect evidence of a systemic process. Studies to date have demonstrated poor outcomes for patients with malignant pericardial effusions. However, little is known of the outcome of patients with non-malignant pericardial effusions and whether pericardial cell counts are associated with outcome.

Purpose

To determine the association of pericardial fluid content with outcome in consecutive patients undergoing pericardiocentesis.

Methods

All consenting adult patients undergoing first pericardiocentesis between 3/2014-3/2024 were included, unless pericardiocentesis was performed in the acute setting of intraprocedural cardiac perforation.

Pericardial fluid was routinely sent for cell count and cytology.

Patients were grouped as follows: 1) malignant (malignant cells on cytology); 2) haemorrhagic (pericardial hemoglobin > 50% of circulating hemoglobin); 3) serous (total nucleated cell count < 500 cells/μL); 4) acute inflammation (neutrophil predominant); 5) chronic inflammation (lymphocyte predominant)

Results

Of the 1092 patients, 595 (54.4%) were male, with mean age 63±16 years. 73% had evidence of pericardial tamponade on echocardiography. Median pericardial fluid volume was 450 ml (IQR 300-650). Fluid analysis was available in 1030 (94%). 14.5% were malignant, 13.1% hemorrhagic, 13.8% serous, 33.6% with acute inflammation, and 25.0% chronic inflammation. Survival free from all-cause mortality was best in those patients with chronic inflammation or hemorrhagic effusions, intermediate in those with acute inflammation or serous fluid, and worst in those with malignant effusions (median survival 2.2 months) (Figure). Pericardial volumes were not different across pericardial fluid types nor associated with long-term outcome.

Conclusion

Malignant pericardial effusions are associated with extremely poor outcome with a median survival rate of 2.2 months. Patients with non-malignant pericardial effusions, particularly serous and but also acute inflammation, do not have a benign prognosis and close follow-up is warranted.figure