Feasibility Study On Clinical Assay Of Ovarian Cancer Platinum-resistance Related Annexin A3Protein In Human Peripheral Blood

Backgroud and Purpose:Epithelial ovarian cancer (EOC) is the leading cause of death among the gynecological cancers with5-years survival around20%to30%in the patients with advanced disease. Resistance of tumor cell to chemotherapeutic drugs is one of the main causes of treatment failure. Platinum is the first-line drug against epithelial ovarian cancer, and there is a clear explanation and definition of clinical platinum-resistance. But understanding of its mechanism is relatively in the initial stage and predicting the platinum-resistance is still in its infancy actually.Since2003, our research team has studied on platinum-resistance related proteins in EOC. The platinum-resistance related protein Annexin A3was selected by comparative proteomics. Validation of Annexin A3by immunohistochemistry identified its significantly high expression in tumor tissues of platinum-resistant patients, whose disease-free survival was much shorter. In animal experiments, when cisplatin treatment was administrated to nude mice xenografted with Annexin A3overpression tumors, the inhibition of tumor growth was significantly reduced. After transfecting Annexin A3-expressing plasmid into drug sensitive cells, the intracellular concentration of platinum and platinum-DNA binding capacity lowered. The Annexin A3overexpression cells showed increase in resistance to platinum compound, while there was no cross-resistance to paclitaxel, epirubicin or other drugs. It suggested that Annexin A3might be the specific platinum-resistant protein. It was found that Annexin A3could be detected in culture medium supernatant of ovarian cancer cell lines, and its concentration was positively correlated with its intra-cellular levels. In vivo, Annexin A3was also detected in serum of ovarian cancer patients, which demonstrated that Annexin A3was a secreted protein.The aims of this research were exploring the feasibility of quantitative detection of Annexin A3protein in the serum by ELISA method on the basis of previous work, analyzing the distribution and expression of serum Annexin A3protein in healthy women, benign ovarian tumor patients and ovarian cancer patients and studying the correlation between the platinum-resistance and serum Annexin A3level of ovarian cancer patients.Methods:1. Objectives:the healthy women, the benign ovarian tumor patients and the ovarian cancer patients were continuously enrolled in Peking Union Medical College Hospital from Dec2009to Feb2010.Inclusion criteria of healthy women included normal gynecological physical examination, cervical cytology and gynecological ultrasonography. Additional entry criteria included serum CA125assay. Women with gynecological or systemic malignancies were excluded. Age distribution was matched with ovarian cancer patients.Inclusion criteria of benign ovarian tumor patients included inpatients who were treated surgically with definite pathological diagnosis. Additional entry criteria included serum CA125assay before the operation.Inclusion criteria of ovarian cancer patients included epithelial ovarian cancer patients who were treated and followed up in Peking Union Medical College Hospital. Patients with other gynecological or systemic malignancies were excluded. Platinum-resistant disease was defined as occurrence of disease progression within6months after completing therapy or disease progression during treatment with platinum-based chemotherapy.2. Collecting the specimens and clinical records:all serum specimens were from the remainder of clinical serum CA125assay, which were kept at4℃refrigerator for no more than7days. Then serum samples were packed in several0.5mL Eppendorfs and transferred to-80℃refrigerator. All the samples were unfrozen only once just before the ELISA assay. The CA125levels of the same serum samples and the corresponding clinical data were recorded.3. Detecting the serum concentration of Annexin A3by ELISA kit. Each sample was set with3duplicates. Mean of the three measured samples was the actual value. The quality control process include:formulating the Standard Operative Procedures (SOP), training operators, carrying out measurement at one time, setting the standard products and standard curve under the guidance of clinical laboratory experts, utilizing quality control products whose concentration were already known, and measuring by dual-wave length absorbance.4. All the clinical data were analyzed retrospectively. The correlations between Annexin A3level and the other clinical variables in platinum sensitive and resistant cases were stratified analyzed. The correlation between the Annexin A3level and platinum-free interval of ovarian cancer patients were determined.5. SPSS17.0software was utilized for statistical analysis. The Crubbs’ method was employed for the quality control test of ELISA. The Kolmogorov-Smirnov test was adopted to test the normal distribution of the quality control samples and others. The t test was employed to compare the means of two independent-samples with normal distribution. The Mann-Whitney nonparametric test was employed to compare the distrbution of two independent-samples with unknown distribution. The Kruskal-Wallis nonparametric test was employed to compare the distribution of multiple independent-samples with unknown distribution.Results:1.271cases were continuously enrolled, including102healthy women (age range:31-78yo, median age:52yo),47benign ovarian tumor patients (age range:17-61yo, median age:36yo) and122ovarian cancer patients (age range:31-81yo, median age:53yo), and the last consisted of63platinum-sensitive patients and59platinum-resistant patients. Age distributions of ovarian cancer patients and healthy women was not significantly different, whereas ones of ovarian cancer patients and benign ovarian tumor patients was significantly different (p<0.05).2.10plates (96-well) were used for assay of total271specimens. Quality control was tested by Crubbs’ method. Quality control was satisfactory in plates1to9, whereas the quality control of plate10did not meet requirements. Total26values from8benign ovarian tumor patients and18healthy women in plate10samples were excluded. The Kolmogorov-Smirnov test of quality control samples showed normal distribution of Annexin A3in three groups (p=0.919,0.751and0.931respectively).6cases of extremely high value and8cases of extremely low value, which exceeded standard curve limits, were excluded.3. Annexin A3protein could be detected in all271samples. After removing extreme values,231samples were analyzed totally. By Kolmogorov-Smirnov test, Annexin A3level distribution in benign ovarian tumor patients showed normal distribution (p=0.387), in healthy women and ovarian cancer patients showed non-normal distribution (p=0.028and0.007, respectively). Serum Annexin A3level in healthy women ranged0.261-4.74ng/mL (mean:0.867ng/mL, median:0.709ng/mL), in benign ovarian tumor patients ranged0.263-2.46ng/mL (mean:0.815ng/mL, median:0.740ng/mL), in ovarian cancer patients ranged0.201-2.84ng/mL (mean:0.689ng/mL, median:0.567ng/mL). By Kruskal-Wallis test, there was no significant difference among these three groups. The correlation between Annexin A3expression and CA125level was analyzed. Healthy women were divided into two groups according to serum CA125level (<10U/mL and≥10U/mL), Annexin A3expression was not significantly different between these two groups. Also, Annexin A3expression was not significantly different neither between two groups from benign ovarian tumor patients (serum CA125<35U/mL and CA125> 35U/mL). Whereas among ovarian cancer patients, Annexin A3concentration was significantly higher in the patients whose serum CA125≥100U/mL than the patients whose serum CA125<100U/mL (p=0.043). The means of these two groups were0.840ng/mL and0.645ng/mL respectively.4. Epithelial ovarian cancer patients were divided into platinum-sensitive and platinum-resistant groups based on disease-free interval. There was no significant difference of Annexin A3levels between the two groups in the stratified analysis which were stratified by the age, FIGO staging, pathological types, tumor differentiation and the total number of the platinum-based chemotherapy cycles. Furthermore, ovarian cancer patients were divided into two groups according to platinum-free interval (PFI<6months and PFI>6months). Stratified analysis was employed based on Annexin A3level. The most upper stratum was≥0.8ng/mL which was recommended as the lowest boundary of the standard curve by the kit instructions, and the lower limit of stratum was≥0.2ng/mL which was the lowest concentration the kit could distinguish. The interval of the strata was0.2ng/mL. Analysis showed that the mean and median of Annexin A3of each group increased simultaneously in accord with the increasing Annexin A3level stratum. When Annexin A3level≥0.6ng/mL, the expressions of Annexin A3in two groups were both normal distribution, so that t test was employed to compare the means between two groups. In Annexin A3≥0.6ng/mL stratum, the mean of Annexin A3level in PFI≤6months group was1.17±0.615ng/mL, while in PFI>6months group was0.924±0.250ng/mL(p=0.048). In Annexin A3≥0.8ng/mL stratum, the mean of Annexin A3level in PFI≤6months group was1.44±0.647ng/mL, while that in PFI>6months group was1.07±0.214ng/mL(p=0.028). Since homogeneity of variance in both groups were not accepted (p<0.05in both groups), so t’test was adopted instead to compare the means. In Annexin A3>0.6ng/mL stratum, p=0.095; in the Annexin A3≥0.8ng/mL stratum, p=0.070. It nearly reached the statistical significance.Conclusions:1.For the first time, this study continuously collected large number of clinical specimens of human peripheral serum and measured platinum-resistant related protein Annexin A3level.271cases were studied totally, including102healthy women,47benign ovarian tumor patients and122ovarian cancer patients.2. Annexin A3expressions in human peripheral blood could be detected by quantitative high-throughput ELISA method. Annexin A3protein had been measured in all271specimens. The distributions of Annexin A3level in healthy women and ovarian cancer patients were non-normal distribution. More than150measured values were below 0.8ng/mL (which was recommended by the Elisa kit as lowest limit of standard curve), suggesting that the sensitivity of ELISA kit available was not satisfactory. Development of more sensitive assay can improve on measurement of Annexin A3protein in human peripheral blood.3. In this study, there was no significantly different of Annexin A3levels in peripheral blood among healthy women, benign ovarian tumor patients and ovarian cancer patients. Annexin A3and serum CA125levels in healthy women and benign ovarian tumor patients were not significantly correlated. In contrast, Annexin A3expression significantly correlated with high level CA125in ovarian cancer patients, suggesting that Annexin A3protein might be ovarian cancer specific protein. In the result, it seemed that Annexin A3level in ovarian cancer patients was lower than that in healthy women and benign ovarian tumor patients. That may be the result of lower tumor volumn after the surgery and chemotherapy. Measurement of serum Annexin A3in ovarian cancer patients before surgery and chemotherapy might identify whether Annexin A3is a platinum-resistant biomarker.4. Epithelial ovarian cancer patients were divided into platinum-sensitive and platinum-resistant groups based on disease-free interval. There was no significant difference of Annexin A3level between the two groups in the stratified analysis which was stratified by the age, FIGO staging, CA125level, pathological types, tumor differentiation and the total number of the platinum-based chemotherapy cycles. Furthermore, ovarian cancer patients were divided into two groups based on platinum-free interval (PFI<6months and PFI>6months). Stratified analysis was employed based on Annexin A3level. When Annexin A3≥0.6ng/mL, Annexin A3level in former group was higher than that in the latter group. It nearly reached the statistical significance. It highly suggested that Annexin A3is an ovarian cancer-specific platinum-resistant protein, and the conclusions could be improved by increasing the number of samples enrolled in the study.