The diagnostic accuracy of HE4 in the differential diagnosis of pleural effusions

Background

Pleural effusions are challenging to diagnose, with approximately 20–50% of malignant effusions not diagnosed by cytology. Human epididymal protein 4 (HE4) may be useful in the differential diagnosis of pleural effusions. In serum, this biomarker shows false-positive results in some benign diseases. The aim of this study was to evaluate the diagnostic utility of HE4 in this setting and to identify false positives.

Methods

Concentrations of HE4, adenosine deaminase, % polynuclear cells, and C-reactive protein, were determined in 238 pleural fluid samples and the estimated glomerular filtration rate (eGFr) in serum.

Results

HE4 values ​​differed significantly (p < 0.01) between malignant [median (IQR)] [1065 (2085)] pmol/L and benign effusions [699 (589)] pmol/L. HE4 concentrations in gynecological and pulmonary tumors were significantly higher than in other tumors.

For a cut-off point of 3050 pmol/L, 22 % sensitivity and 100 % specificity were obtained.

In patients with benign disease, significant increases in HE4 were identified only in those with eGFr < 30 mL/min/1.73 m² [1050(596)] pmol/L, and not in those with eGFr > 30 mL/min/1.73 m² [597(532)] pmol/L).

Two cut-offs were established for maximum specificity, depending on the eGFr: 3050 pmol/L for eGFr < 30 mL/min/1.73 m² and 1992 pmol/L for eGFr > 30 mL/min/1.73 m². A sensitivity of 28.5 % was obtained for patients with eGFr > 30 mL/min/1.73 m² and 36.3 % for patients with eGFr < 30 mL/min/1.73 m². The sensitivity using a specific cut-off point was 29.7 %.

Conclusions

The determination of HE4 in pleural fluids demonstrates high specificity and low sensitivity. The use of specific cutoff points that are clinically adjusted improves sensitivity while maintaining maximum specificity.