| Backgroud and PurposeOvarian cancer is the most lethal gynecological malignancy. Resistance of tumor cells to chemotherapeutic drugs is one of the main causes of treatment failure. Using a biomarker to predict clinical platinum resistance is important to treat patients individually and eventually improve prognosis.In previous studies of our research team, the platinum-resistance related protein Annexin A3was selected by comparative proteomics and further validated both in vitro and vivo. Further studies also demonstrated that Annexin A3was a secreted protein. It could be secreted outside the cells by exosomes and be detected both in culture medium supernatant of ovarian cancer cell lines and serum of ovarian cancer patients. This characteristic enabled Annexin A3as a potential serological platinum resistant biomarker used in clinic. In order to explore the feasibility of quantitative detection of Annexin A3protein in the serum, our research team used the only commercially available detection kit—Annexin A3enzyme-linked immunosorbent assay (ELISA) kit. However, this kit was found the disadvantages such as low sensitivity, poor reproducibility and so on in subsequent clinical detection of large sample size. These outstanding problems resulted in poor accuracy of experimental data. Therefore, developing a new immunoassay kit for high-throughput and stably detecting the expression level of Annexin A3protein in human peripheral blood has become a key issue to be solved in our further study.Chemiluminescence immunoassay (CLIA) is a new immunological detection method, which develops following radioimmunoassay, enzyme immunoassay and fluorescence immunoassay. It combines highly specific antigen-antibody reaction with high sensitivity chemiluminescent reaction, and has many advantages such as high sensitivity and specificity, good reproducibility, simple and convenient operation, without radioactive contamination and so on. CLIA has become the mainstream method in clinical detection of protein molecule.Based on above background, the present study aimed to prepare the Annexin A3CLIA kit which can reach clinical detection requirements, and then applied it to quantitatively detect serum Annexin A3protein in different women population in order to understand its expression distribution in human peripheral blood, determine its normal range initially, and explore its predictive value for platinum chemoresistance in ovarian cancer patients as a serological marker. Methods1,Six kinds of commercially available anti-Annexin A3polyclonal or monoclonal antibodies were evaluated for their affinity with the antigen—human Annexin A3recombinant protein using double-antibody sandwich chemical luminescence enzyme immunoassay. The best pair was chosen as the capture antibody and detection antibody to prepare chemiluminescence immunoassay kit. Capture antibody was coated on microplate strips in different concentrations and detection antibody was conjugated with horseradish peroxidase (HRP) with modified sodium iodide method which was also diluted in different concentrations. They could form a double-antibody sandwich complex with Annexin A3antigen standard of different concentrations. Luminescence analyzer could detect the light intensity after adding light-emitting substrate which was proportional to the concentration of standard Annexin A3protein. The optimal working concentration of capture antibody and HRP labeled detection antibody was determined with cross-titration method for the best accuracy in measuring known concentrations of antigen. After optimizing the response mode, the final optimal detection system was established and the CLIA kit was prepared successfully.2,Assess the Annexin A3CLIA kit in its sensitivity, precision (also called reproducibility) and dilution linearity, and determine the detection range by plotting the precision curve. The CLIA kit was also compared with the purchased ELISA kit by simultaneously detecting Annexin A3protein in sixty serum samples from healthy women and patients with benign ovarian tumor or epithelial ovarian cancer.3,The serum specimens of healthy women at health checkup, benign or borderline ovarian tumor patients before surgery and epithelial ovarian cancer patients before treatment were collected continuously in Peking Union Medical College Hospital from September2009to April2012, and were detected by CLIA kits. The normal reference range of Annexin A3in healthy women was determined by95%confidence interval. The serum specimens of epithelial ovarian cancer patients before surgery or chemotherapy, after surgery but before chemotherapy, during chemotherapy and after chemotherapy were collected prospectively and continuously, all of which were detected by CLIA kits too. The relationship between the concentration of Annexin A3and the various clinicopathologic factors, surgery or chemotherapy was analyzed. All of the ovarian cancer patients were followed up in order to determine the chemotherapy outcome and disease status. According to the definition of clinical platinum chemoresistance, they were divided into the sensitive group and resistant group. The dynamic changes and expression differences of Annexin A3were analyzed in the two groups. By correlation analysis, the relationship between Annexin A3and CA125at different treatment points was evaluated. Diagnostic test was used to assess the diagnostic or predictive value of CA125and Annexin A3.Receiver operating characteristic (ROC) curve was drawn and the best diagnostic cutoff value was determined by the maximum Youden index which is equal to the sensitivity and specificity minus1, and then the diagnostic sensitivity, specificity, accuracy (total coincidence rate), false positive rate, the false negative rate, positive predictive value and negative predictive value were calculated. The ovarian cancer patients were also divided into the high-expression group and the low-expression group using the upper limit of normal for Annexin A3as cutoff value. The correlation between the Annexin A3level before treatment and progression free survival (PFS) was assessed by COX regression analysis and survival analysis.Results1,Using a rabbit anti-Annexin A3polyclonal antibody as capture antibody and a mouse anti-Annexin A3polyclonal antibody as detecting antibody to detect Annexin A3standard of different concentration showed the highest antigen-antibody binding specificity based on the largest correlation coefficient of the standard curve and the highest signal to noise ratio (SNR, the light intensity ratio of maximum value and zero value standard proteins),0.9982and19.56respectively. So this antibody pair was the best for preparing CLIA kit. The optimal working concentration of capture antibody and HRP labeled detection antibody was5μg/mL and1:250respectively which was also determined by the largest correlation coefficient0.9987and the highest SNR59.33. The reaction mode of one-step was better than two-step because of the similar correlation coefficient (0.9983vs0.9986) but higher SNR (49.11vs45.57) and easier to operate.2,The sensitivity of Annexin A3CLIA kit was0.08ng/ml. The intra-assay precision was5.92～8.63%and the inter-assay precision was5.15～6.42%. It had good dilution linearity and the correlation coefficient of the fitting curve between sample dilution and light intensity was0.9997.The work scope of the kit was determined as0.1～1.6ng/ml for the coefficient of variation all less than10%. Detecting sixty blood samples showed the results as follows. The median measured value was0.36ng/ml (0.1～2.03ng/ml) by ELISA kit and0.95ng/ml (0.04～2.77ng/ml) by CLIA kit.80%(48/60) measured value were lower than the minimum recommended concentration (0.8ng/ml) of the standard curve, and30%(18/60) were lower than the minimum detectable concentration (0.2ng/mL) indicate by ELISA kit instructions. While using CLIA kit,90%(54/60) were in the range of the standard curve, only two lower than the minimum detectable concentration (0.08ng/ml). And the other four cases higher than the upper limit of the standard curve (1.6ng/ml) were re-tested by diluted two times, four times and eight times. It showed a good agreement among the final calculated values of different dilutions, and the coefficients of variation were all less than3%. The standard or samples were added at the same time in a one-step detection mode and the reaction time was only two hours. While the ELISA kit applied three-step mode by adding the standard or samples first, then detection antibody and last enzyme which made the reaction time up to4.5hours.3,113healthy women,70patients with benign ovarian tumor,14patients with borderline ovarian tumor and96patients with epithelial ovarian cancer were enrolled into the study. The median concentration of serum Annexin A3in the four groups was0.75ng/ml (0.15～1.54ng/ml),0.66ng/ml (0.04～1.79ng/ml),0.65ng/ml (0.14～3.29ng/ml) and1.12ng/ml (0.20～6.18ng/ml) respectively. The concentration of Annexin A3before treatment in ovarian cancer patients was significantly higher than that in healthy women and patients with benign ovarian tumor (P=0.000). There was no significant difference between healthy women and patients with benign ovarian tumors (P=0.322). The level of Annexin A3had no significant difference in patients with borderline ovarian tumor compared with the other three groups respectively (P values were0.144,0.132and0.236). Kolmogorov-Smirnov test showed normal distribution of Annexin A3in healthy women (Z=0.723, P=0.673). The upper limit of normal for Annexin A3was defined as1.45ng/ml based on the mean value plus1.96standard deviations.There was no correlation between the Annexin A3level before treatment and the age, histological type, histological grade, clinical stage, or lymph node metastasis in ovarian cancer patients (P values were0.985,0.631,0.347,0.441and0.480). The median concentrations of Annexin A3before and after surgery were1.11ng/ml (0.43～4.55ng/ml) and1.23ng/ml (0.4～4.32ng/ml) respectively, and there was no significant difference by comparison (P=0.226). The median concentration of Annexin A3before chemotherapy was1.15ng/ml (0.42～6.18ng/ml) and significantly reduced to0.23ng/ml (0.02～3.OOng/ml) during chemotherapy (P=0.000). After chemotherapy, the median concentration of Annexin A3was0.24ng/ml(0.03～1.82ng/ml), significantly lower than that before chemotherapy (P=0.000) but without significant difference compared with that during chemotherapy (P=0.487).Up to April30,2012, the median follow-up time was10.73months (2.23to32.17months) in96patients with ovarian cancer, of whom56cases could be clearly judged the outcome of platinum chemoresistance, including41(73.21%) platinum-sensitive patients and15(26.79%) drug-resistant patients. During the follow-up period,3patients were lost,21patients had progressive disease, and3patients were still undergoing the neoadjuvant chemotherapy. The level of Annexin A3during chemotherapy or after chemotherapy was significantly lower than that before chemotherapy either in the sensitive group or the resistant group (P=0.000). But there was no significant difference between the treatment points during chemotherapy and after chemotherapy in both groups (P values were0.605and0.296). The level of Annexin A3before treatment in the resistant group was significantly higher than that in the sensitive group (P=0.012), and the median concentrations were2.21ng/ml (0.5-6.18ng/ml) and0.80ng/ml (0.20-2.43ng/ml) respectively. But the levels of Annexin A3during or after chemotherapy showed no significant difference between the two groups (P values were0.937and0.624). The level of CA125before treatment showed no significant difference between the two groups (P=0.271) and had no diagnostic value to platinum chemoresistance because the area under ROC curve was only0.597(P=0.271). While the CA125level during chemotherapy in the resistant group was significantly higher than that in the sensitive group (P=0.003) and the AUC was0.755(P=0.014).The best diagnostic cutoff value was24.60U/ml and at this point, the diagnostic sensitivity was60.00%, the specificity was90.24%, the accuracy was82.14%, the false positive rate was9.76%, the false negative rate was40.00%, the positive predictive value was69.23%and the negative predictive value was86.01%. By correlation analysis, the expression level of Annexin A3was unrelated to CA125at any treatment points (P values were0.384,0.594and0.883). Using the upper limit of normal1.45ng/ml as cutoff value, the ovarian cancer patients before treatment was divided into the high-expression group and the low-expression group and the platinum resistance ratio in the high-expression group was significantly higher than that in the low-expression group (55.56%vs13.16%, P=0.001). Logistic regression analysis showed that the level of Annexin A3before treatment was one of the platinum-resistant risk factors (P=0.004, OR=9.241), which meant the platinum resistance risk of high-expression group was9.241times that of the low-expression group. Diagnostic test showed that the AUC of Annexin A3before treatment was0.744(P=0.006) and the best cutoff point was1.645ng/ml. At this point, the sensitivity, specificity, accuracy, false positive rate, false negative rate, positive predictive value and negative predictive value of the prediction to platinum resistance were66.70%,82.93%,78.57%,17.07%,33.30%,58.82%and87.18%respectively. Cox hazard regression model analysis showed that the level of Annexin A3before treatment was an independent risk factor for disease progression (P=0.001, RR=9.863), which meant the risk of disease progression in the high-expression group was9.863times that in the low-expression group. By the Kaplan-Meier survival analysis, the median PFS of the two groups were27.60months (2.23to32.17months) and12.77months (4.90to25.80months) respectively and the PFS of high-expression group was significantly shorter than that of low-expression group (P=0.001).Conclusions1,In this study we chose a pair of antibody as the capture antibody and detection antibody from six kinds of commercially available anti-Annexin A3polyclonal or monoclonal antibodies which had the best affinity to human Annexin A3recombinant protein. And the Annexin A3protein chemiluminescence immunoassay kit was successfully prepared for the first time.2,The CLIA kit had high sensitivity, good reproducibility and dilution linearity, and convenient operation.Compared with currently commercially available ELISA kit, the CLIA kit had higher sensitivity (0.08ng/ml vs0.2ng/ml), better discrimination to serum with low concentration of Annexin A3protein, and fewer steps. The CLIA kit was better than the ELISA kit and could meet the requirements of clinical detection. But the CLIA kit also had a shortcoming of the low upper limit of standard curve, which resulted in the re-test of high value serum samples after dilution. Further improvement was needed in order to be to increase the linear detection range.3,Annexin A3protein could be detected differential expression in the peripheral blood of different women population by the CLIA kit and its upper limit of normal was initially defined as1.45ng/ml. The level of Annexin A3before treatment in ovarian cancer patients was significantly higher than that in the healthy women or in benign ovarian tumor patients. The level of Annexin A3didn't change after the removal of tumors by surgery but could gradually decreased to normal with chemotherapy and there was no significant difference between the sensitive group and the resistant group. The level of CA125before treatment had no significant difference between the sensitive group and the resistant group, but during chemotherapy the level of CA125in the resistant group was significantly higher than that in sensitive group. The expression level of Annexin A3was unrelated to CA125at any treatment points.The differential expression of Annexin A3before treatment and CA125during chemotherapy between the sensitive and resistant group suggested they had predictive and diagnostic value in judging the platinum chemoresistance respectively. Besides the level of Annexin A3before treatment was the independent risk factor for disease progression.
|
PDF Full Text Request