Incremental Diagnostic Value of Computed Tomography Attenuation in Differentiating Malignant Pericardial Effusion: A Retrospective Observational Study

OBJECTIVE
Malignant pericardial effusion (MPE) is associated with poor prognosis and often presents as cardiac tamponade. While cytology is the diagnostic gold standard, it has limited sensitivity. Computed tomography (CT) attenuation, measured in Hounsfield Units (HU), may reflect fluid composition and aid in the non-invasive differentiation of MPE.

METHODS
This retrospective, single-center study analyzed 102 patients who underwent percutaneous pericardiocentesis and thoracic CT. Patients were grouped as malignant or non-malignant based on pathological findings. CT attenuation was measured in three standardized axial levels, and its diagnostic value was evaluated using multivariate logistic regression and receiver operating characteristic (ROC) analysis. Two predictive models were compared: Model 1 (clinical and laboratory variables) and Model 2 (Model 1 + CT attenuation).

RESULTS
MPE was identified in 44 (43.1%) patients. CT attenuation values were significantly higher in the MPE group (median 24.4 HU vs. 9.3 HU, p < 0.001). In multivariate analysis, male sex, high pericardial fluid protein, low glucose, and high lactate dehydrogenase were independent predictors of MPE. When CT attenuation was added to the model (Model 2), it also emerged as an independent predictor (OR = 1.076, 95% CI: 1.026–1.128, p = 0.003). The inclusion of CT attenuation improved the model's diagnostic performance (AUC: 0.893 for Model 2 vs. 0.860 for Model 1). Model 2 demonstrated superior diagnostic performance (AUC = 0.893), with a CT attenuation cut-off of 16.45 HU achieving 88.2% sensitivity and 78.3% specificity.

CONCLUSIONS
CT attenuation adds incremental diagnostic value in identifying MPE when combined with conventional clinical and biochemical parameters. In clinical settings where rapid diagnosis is critical, its non-invasive and reproducible features may enhance the early detection of malignant conditions.