*3.4. Shed SDC-1 and VEGF Levels Correlate to Patient Survival in Malignant Mesothelioma*

In order to determine the prognostic value of pleural effusion derived Angiopoietin-1, HGF, MMP-7, Osteopontin, TIMP-1, Galectin, Mesothelin, NRG1-b1, SDC-1 and VEGF, we divided patients in two groups depending on the established cut-off value for all analytes. Strikingly, malignant mesothelioma patients with high levels of SDC-1 (>19.28 ng/mL) and VEGF (>0.7 ng/mL) have significantly worse prognosis in comparison to the patients with low levels of SDC-1 (<19.28 ng/mL) and VEGF (<0.7 ng/mL). Median survival time of malignant mesothelioma patients with high VEGF level was significantly shorter (2.6 months) compared to low VEGF level (18 months) (*p* = 0.0003) (Figure 4A). Median survival time of malignant mesothelioma patients with high level of SDC-1 was significantly shorter (2.4 months) compared to patients with low shed SDC-1 level (9.6 months) (*p* = 0.03) (Figure 4B).

**Figure 4.** VEGF and shed SDC-1 have prognostic value in malignant mesothelioma patients. Patients were separated into "high" and "low" VEGF and shed SDC-1 level by the online web application Cutoff Finder (VEGF cutoff = 0.7 ng/mL and shed SDC-1 = 19.28 ng/mL). Malignant mesothelioma patients with high level of VEGF (**A**) and shed SDC-1 (**B**) have shorter survival time compared to malignant mesothelioma patients with low VEGF and shed SDC-1 levels. *P-*values are ≤0.05.

### ≤ *3.5. Correlation between Shed SDC-1 and Other Biomarkers*

The correlation between shed SDC-1 and other biomarkers was explored in malignant mesothelioma patients and in all malignant cases (malignant mesothelioma and adenocarcinoma). In all malignant cases, we showed a significant weak-positive correlation between shed SDC-1 and Angiopoietin-1 (*p* = 0.004 and r = 0.3) and a significant weak-negative correlation between shed SDC-1 and Mesothelin (*p* = 0.004 and r = −0.3) (Figure 5). We did not find any significant correlation between

expression of shed SDC-1 and other seven angiogenic-related biomarkers (HGF, MMP-7, Osteopontin, TIMP-1, Galectin, VEGF and NRG1-b1).

−

−

**Figure 5.** Correlation between shed SDC-1 and other biomarkers in all malignant cases. (**A**) Expression of shed SDC-1 was significantly and positively correlated with Angiopoietin-1 (*p* = 0.004 and r = 0.3), whereas, shed SDC-1 expression was negatively correlated with Mesothelin (*p* = 0.004 and r = −0.3) (**B**). −

− In malignant mesothelioma cases, shed SDC-1 expression was significantly and positively correlated with HGF (*p* = 0.02; r = 0.3) and NRG1-b1 (*p* = 0.001; r = 0.4) (Figure 6). No significant correlations were found between shed SDC-1 and other seven biomarkers (Angiopoietin-1, Mesothelin, MMP-7, Galectin-1, Osteopontin, TIMP-1 and VEGF).

**Figure 6.** Correlation between shed SDC-1 and other biomarkers in malignant mesothelioma cases. (**A**) Expression of shed SDC-1 was significantly and positively correlated with HGF (*p* = 0.02 and r = 0.3) and NRG1-β1 (*p* = 0.001 and r = 0.4) (**B**) in malignant mesothelioma cases.

### β *3.6. Diagnostic Value of Individual Biomarkers for Malignant Mesothelioma*

β β β Based on ROC curve analyses, Galectin-1, Mesothelin, Osteopontin, NRG1-β1 and shed SDC-1 performed the best diagnostic capacity to distinguish malignant mesothelioma from benign effusions. MMP-7 and angiopoietin-1 have no discriminative capacity to distinguish between malignant mesothelioma and benign effusions (Figure 7). The resulting areas under the curves (AUC) are shown in (Figure 7).

β

β **Figure 7.** Diagnostic efficiency of individual biomarkers. Receiver operating characteristic (ROC) analysis showing the specificity and sensitivity for each individual angiogenesis related biomarkers. Area under the curve (AUC) values show that Galectin-1 (AUC = 0.99), NRG1-β1 (AUC = 0.97), Osteopontin (AUC = 0.95), Mesothelin (AUC = 0.94) and shed SDC-1(AUC = 0.93) have the most diagnostic value for malignant pleural mesothelioma.
