Pleural effusion hyaluronic acid as a prognostic marker in pleural malignant mesothelioma

Abstract

Background

Malignant mesothelioma (MM), a primarily asbestos-induced tumour, has a poor prognosis, with over-all 5-year survival less than 5%. Tumour biomarkers are being intensely investigated in MM as aids to diagnosis and prognosis. Hyaluronic acid (HA) is produced in MM but its role in prognostication remains uncertain.

Materials and methods

HA concentrations were determined in matching serum and pleural effusion of 96 MM patients, 26 lung cancer patients and 42 patients with benign effusions resulting from infectious, cardiac, renal, liver and rheumatoid diseases and compared to the current ‘best practice’ biomarker, mesothelin. Liver and kidney function were determined for each patient. Diagnostic accuracy was determined by area under the receiver operator characteristic curve (AUC) analysis following logistic regression modelling. Difference in survival between groups was determined by both log-rank test and Cox proportional hazards regression modelling.

Results

For effusion HA, the AUC (IQ range) was 0.89 (0.82–0.94) and for effusion mesothelin, it was 0.85 (0.78–0.90). Serum HA was not diagnostically useful. A combined measure of effusion HA, and serum and effusion mesothelin had an AUC of 0.92 (0.86–0.96), which was significantly higher than effusion mesothelin alone. Effusion HA had a biphasic distribution in MM patients, dichotomised at a concentration of 75 mg/L. The median survival of MM patients with high effusion HA was 18.0 (13.7–22.4) months, significantly longer than those with low HA effusion levels (12.6 months (8.4–16.8), p = 0.004). Serum HA, and effusion and serum mesothelin were not significant prognostic indicators.

Conclusion

This study demonstrates that a combined biomarker panel has greater diagnostic accuracy than effusion mesothelin alone, and that significant prognostic information is provided by effusion HA.

Introduction

Malignant mesothelioma (MM) is an asbestos-induced cancer which is difficult to diagnose and has a median post-diagnosis survival of less than 12 months. This poor survival may be related to the late clinical diagnosis of the disease [1], [2].

Hyaluronic acid (HA) is a large polysaccharide made predominantly by fibroblasts and is present in MM pleural effusions [3], [4], [5], [6], [7]. HA is rapidly removed from the circulation by the clearance receptor stabilin-2 [8], and has a plasma half-life of 2.5–5 min [9], [10]. Despite being a recognised feature of MM, the role of HA in diagnosis and prognostication remains uncertain.

Soluble mesothelin was identified as a specific serum biomarker for MM in 2003 [11] and is also elevated in MM effusions [12]. Mesothelin is the best characterised biomarker for MM however, recent meta-analysis found sensitivity was too low for early diagnosis and suggested further biomarker research be undertaken [13]. To determine if the sensitivity of the mesothelin biomarker for MM could be improved by the incorporation of effusion and serum HA we undertook this study in matching serum and pleural effusions. As there is increasing interest in biomarkers as prognostic indicators in MM and HA concentration has been reported to correlate with survival [14] and disease progression [15] the prognostic value of the two biomarkers in MM was also examined.

The effects of liver and renal function on HA and mesothelin concentrations were assessed. Serum HA is elevated in liver disease, with the damaged liver failing to remove HA from circulation; HA is a component of some liver disease diagnostic test panels for this reason [16]. Mesothelin is excreted via the kidney so renal function is also important in mesothelin biomarker assessments.