Diagnostic And Prognostic Value Of Serum Mesothelin/smrp In Patients With Epithelial Ovarian Cancer

Objective: Mesothelin (MSLN) is a 40-KDa cell surface antigen newly detected. It is a glycoprotein present on the surface of normal mesothelial cells and highly expressed in some cancers such as malignant mesothelioma, ovarian cancer, pancreatic cancer and other malignant tumors. It has three variants and one of them can shed from cell surface into blood forms soluble mesothelin related peptides/proteins (SMRP), with the relative molecular mass 42KDa～45KDa. SMRP can be detected in blood. In order to investigate the role that serum SMRP played in the early diagnosis and prognostic in epithelial ovarian cancer, serum SMRP concentrations was determined by commercially available enzyme-linked immunosorbent assay (ELISA) in samples from three groups of patients with ovarian cancer, benign ovarian masses and healthy female volunteers respectively. Cell bound mesothelin expression was evaluated by immunohistochemistry in ovarian cancer patients. The difference of serum SMRP levels has been compared between the three groups, various subtypes of ovarian cancer, across stages and grades. Serum SMRP levels were also measured before and after surgical cytoreduction in patients with EOC. Serum CA125 level was evaluated at equal pace in candidates mentioned above. Clinical retrospective status was also combined in order to generally evaluated SMRP as a new potential serum tumor maker for epithelial ovarian cancer.Methods:We recruited 56 primary epithelial ovarian cancer patients, 38 patients of ovarian benign masses and 34 healthy volunteers during March 2008 to April 2009 from the gynecology-oncology Department of Bethune International Peace Hospital and The Forth Hospital of Hebei Medical University. Corresponding archival surgical ablated paraffin-embedded EOC tissues were selected for use. 3-4mL peripheral blood was obtained from patients and healthy females by routine venepuncture (vacuum blood collection tube with coagulant). Sample was allowed to clot, centrifuged at 3000rpm for 15 minutes, then the supernatant liquid was aspirated and stored at under -20℃until use.1 Serum SMRP concentration was determined by a commercial ELISA assay, the MESOMARK kit, supplied by Fujirebio Diagnostics.2 Tissue sections obtained from archival paraffin-embedded tumor blocks from surgical resection were evaluated for mesothelin expression and artificial semi-quantitative analysis.3 Patients'serum CA125 concentration was routine assayed by Department of Nuclear Medicine of Bethune International Peace Hospital with Chemiluminescence.4 The statistical analysis of the data was done using SSPS version 13.0 for windows. Measurement data comparison, numeration data comparison, correlation analysis and diagnostic ability analysis were applied. A p value less than 0.05 resulting from a two-sided test was considered as statistically significant.Results:1 Serum SMRP levels in patients and healthy volunteers.1.1 Serum SMRP levels were different between the three groups: the median values measured in patients with ovarian cancer were significantly higher than in patients with either benign ovarian masses(Wilcoxon z=6.973, P<0.001) or healthy volunteers (z=7.153, P<0.001)after rank transformation. However, the SMRP levels between patients with benign ovarian masses and healthy volunteers were not statistically significantly different (Wilcoxon z=1.895, P=0.058>0.05).1.2 There was no significant difference between the histologic subtypes of ovarian cancer(Kruskal-Wallis H Test,χ2=5.736, P=0.057>0.05).1.3 The SMRP values measured in stageⅡ-ⅣEOC patients significantly higher than in stageⅠEOC patients, (Wilcoxon, z=3.184, P=0.001<0.05). 1.4 There was significant difference between the differentiation grades of ovarian cancer, (Kruskal-Wallis H Test,χ2=26.221, P=0.000<0.05).The higher differentiation grade correspond the lower level of serum SMRP.1.5 There was no significant difference between the histologic subtypes of benign ovarian masses. (Kruskal-Wallis H Test,χ2=6.948, P=0.225>0.05).1.6 Serum SMRP levels significantly decreased after surgical cytoreduction than preoperative in patients with EOC (Wilcoxon, z=5.129, P=0.000<0.05).2 Diagnostic ability of serum SMRP in distinguishing ovarian cancer from other samples. Receiver operating characteristic (ROC) curve analysis showed the diagnostic ability was promising. The area under the ROC curve (AUC) was 0.938, standard error was 0.026 and 95% confidence interval (CI) was 0.888–0.988.3 The assessment of the ELISA assay which detect serum SMRP as a biomarker for ovarian diseases. Optimal cutoff values which define the best diagnostic efficiency were calculated according to Youden-index (YI= sensitivity + specificity - 1), we made cut-off=1.6432nM. Consequently we get the criterion that means patients with SMRP serum concentrations more than 1.6432nM were judged positive/anomalism, less than or equal to 1.6432nM were judged negative/normal, with sensitivity at 0.821, specificity at 1.4 To assess the clinical potential of SMRP and CA125 in differentiation EOC from benign ovarian masses (using pathologic diagnosis as the Golden standard), the receiver operating characteristic (ROC) curve were made for the two tumor markers. The area under the receiver operating characteristic (ROC) curve (AUC) for serum SMRP was 0.925 with a cut-off value of 1.6432nM for differentiating EOC patients from benign ovarian masses patients, whereas the AUC for CA125 was 0.941 with a cut-off value of 35.0u/ml. Both SMRP and CA125 showed proximalis distinguish ability. The diagnostic test evaluating index was comparison of SMRP and CA125. The sensitivity (Se) was proximalis. SMRP has higher specificity(Sp), omission diagnose rate(β), positive predictive value(PV+), correct ratio (π), Youden-index (YI) , lower mistake diagnose rate(α) and negative predictive value(PV-) than CA125.5 Correlations between serum SMRP and CA125 values were calculated using Spearman's rank test, did not find significant Correlations between the two markers (p=0.334>0.05).6 Correlations between serum SMRP and clinical status such as ascites before surgical cytoreduction, primary tumor size, menopause state, were calculated using Spearman's rank test respective. No significant correlation has been found.7 Combine serum SMRP and CA125 may outperform SMRP or CA 125 alone, we obtained the sensitivity at 97.9%, and specificity at 81.6% in distinguishing EOC from benign ovarian masses.8 There was no significant correlation between serum SMRP and tissue mesothelin expression (p=0.178>0.05).Conclusion:1 Serum SMRP level evaluated by ELISA kit was significantly elevated in patients with EOC than in patients with either benign ovarian masses or in healthy volunteers.2 The diagnostic test evaluating showed that serum SMRP has a promising sensitivity and specificity. When compared with CA125, SMRP has higher specificity and the sensitivity was proximalis. These results have identified serum SMRP as a potential biomarker for EOC, which could be useful for early diagnosis.3 The rapid decrease in SMRP levels after surgery in patients with EOC suggests that serum SMRP may be a useful test to monitor treatment response in epithelial ovarian cancer.