Hyaluronan and N-ERC/mesothelin as key biomarkers in a specific two-step model to predict pleural malignant mesothelioma.

Abstract

PURPOSE: Diagnosis of malignant mesothelioma is challenging. The first available diagnostic material is often an effusion and biochemical analysis of soluble markers may provide additional diagnostic information. This study aimed to establish a predictive model using biomarkers from pleural effusions, to allow early and accurate diagnosis. PATIENTS AND METHODS: Effusions were collected prospectively from 190 consecutive patients at a regional referral centre. Hyaluronan, N-ERC/mesothelin, C-ERC/mesothelin, osteopontin, syndecan-1, syndecan-2, and thioredoxin were measured using ELISA and HPLC. A predictive model was generated and validated using a second prospective set of 375 effusions collected consecutively at a different referral centre. RESULTS: Biochemical markers significantly associated with mesothelioma were hyaluronan (odds ratio, 95% CI: 8.82, 4.82-20.39), N-ERC/mesothelin (4.81, 3.19-7.93), CERC/mesothelin (3.58, 2.43-5.59) and syndecan-1 (1.34, 1.03-1.77). A two-step model using hyaluronan and N-ERC/mesothelin, and combining a threshold decision rule with logistic regression, yielded good discrimination with an area under the ROC curve of 0.99 (95% CI: 0.97-1.00) in the model generation dataset and 0.83 (0.74-0.91) in the validation dataset, respectively. CONCLUSIONS: A two-step model using hyaluronan and N-ERC/mesothelin predicts mesothelioma with high specificity. This method can be performed on the first available effusion and could be a useful adjunct to the morphological diagnosis of mesothelioma.