Establishment Of Primary Ovarian Cancer Xenograft Models In Nude Mice For Efficacy And Gene Mutation Test

Objectives1. To establish a variety of primary ovarian cancer xenograft models in nude mice using the fresh tissues of epithelial ovarian carcinoma under sterile condition from the operationroom. To observe the growth characteristics of the xenogrfts and discuss the relationship between the clinical-pathological factors and tumor take rate. Compare the morphological character of the xenografts derived from ovarian cancers and cell lines with the original human tumors. It provides a research platform for systematic study of pathogenesis, biological characteristics, preclinical pharmacodynamics experiment, and gene phenotype of ovarian cancer.2. To test the efficacy of taisu(taxol), carboplation and cisplation on the human ovarian carcinoma xenografts. To evaluate the application value of the nude mice xenograft models in pharmacodynamics by comparing the efficacy with clinical outcome.3. To detect the genes alterations in fifth-generation xenografts by Next Generation Sequencing and select the related genes with ovarian cancer by searching the literature. It provides the basis for new anticancer drug.Methods1. The collected epithelial ovarian cancer tissues which were confirmed by frozen sections were inoculated subcutaneously in nude mice within6hours (4spots per mouse,3-5mice per case). The growth of the xenografts was observed. The definition of tumor take rate is that the xenograft can grow progressively after implantation and could be serially passaged within6month. Xenografts which were larger than800mm3within60days during the following passage were regarded as success. The weight of the mice and the volume of xenografts were measured from passage3and the growth curves were drawn. The pathological features of P1xenografts, P5xenografts, the corresponding human ovanrian cancer and xenografts originated from ovarian cancer cell lines were compared.2. The P5xenografts of ID17, ID34, ID49and ID50were randomly devided into4 groups (blank controller, taisu, carboplatin, and cisplatin), respectively. The efficacy of the above drugs were evaluated by T/C value. T and C were the mean volume of the treated and control groups respectively on a given day. The experimental result and clinical efficacy were compared.3. The genes alterations were detected using Angilent SureSelect target enrichment capture system and Illumina Solexa HiSeq2000sequencing paltform. A total of12P5xenografts were analyzed. The results were compared with NCBI reference genome build37.3. The results were analysed by retrievaling HMGD in version of profession and BIC database.Results1. Overall tumor take rate was53.13%(34/64). The take rate of xenografts in subsequent passages was94.12%(32/34). The delay from initial implantation to the first passage was2-3months for most of the xenografts, and the time between subsequent passages was3-8weeks. No abdominal fluid was observed in the tumor bearing mice. Metastatic spread was observed in only3cases. The xenografts originated from human were more similar with human ovarian cancer than xenografts originated from ovarian cancer cell lines.2. There were no significant difference in tumor take rate among different FIGO stages (χ2=3.022;P=0.388) and among different histological types except for mucious adenocarcinoma (χ2=0.997;P=0.802). The tumor take rate of mucious adeno-carcinoma was0. The tumor take rate of Stage Ⅲ./Ⅳ serous adenocarcinoma was higher than Stage Ⅰ/Ⅱ (P=0.018). The tumor take rate of non-serous adenocarcinoms including endometrioid carcinoma, clear cell carcinoma, and low differential cancer was higher than serous adenocarcinoma for Stage Ⅰ/Ⅱ (P=0.044).3. T/C value of ID17was44.93%,27.1%,11.77%in taisu, carboplatin, and cisplatin group, respectively. T/C value of ID34was65.15%,57.81%,30.67%in taisu, carboplatin, and cisplatin group, respectively. T/C value of ID49was49.50%,28.97%,6.79%in taisu, carboplatin, and cisplatin group, respectively. T/C value of ID50was68.97%,16.64%,34.02%in taisu, carboplatin, and cisplatin group, respectively. All the drugs had anticancer activity (P<0.01of all), which were similar with clinical initial treatments.4. There were many mutation genes in12xenografts. Partial of them were the key molecules in EGFR signal transduction pathway such as EGFR, KRAS, PIK3CA, AKT1, mTOR. EGFR with KRAS, PIK3CA, AKT1, mTOR mutation simultaneously were presented in5,3,2,2cases respectively. BRCA1and BRCA2mutation were observed in7and12xenogratfs respectively. There were6cases with BRCA1and TP53mutation and10tumors with BRCA2and TP53mutation. Similarly, the number of cases with BRCAl and PARP1mutation or with BRCA2and PARP1mutation were5and10respectively. Multisite mutation was detected in genes of BRCA1, BRCA2, and TP53. Single point mutation was identified in genes of EGFR, KRAS, AKT1, PIK3CA, mTOR, PTEN and PARP1. CTNNB1and BRAF were wild-type in all of12xenegrafts. Harmful mutations in the study include p.R521K of EGFR; p.V762A and p.K940R of PARP1; p.W353Stop, p.P871L, p.E1038G, p.K1183R, p.S1613G and p.R1751Stop of BRCA1; p.N372H, p.H523R, p.I3412V, p.A2466V of BRCA2; p.RZ2P, p.R273C、p.R175H and p.C141Y of TP53. All the above mutations were proved to be SNPs except p.H523R of BRCA2and p.C141Y of TP53.Conclusions1. A total of34primary ovarian cancer xenograft models in nude mice were established. It provided a platform for further study. But the tumor biological behaviors such as growth pattern and metastasis ability were different from their originated tumors.2. Cisplatin, carboplatin and taisu could inhibit tumor growth within experimental period. It was similar with clinical initial treatment. The xenograft models in nude mice could be used to explore new anticancer drugs and to test drug susceptibility.3. Xenografts maintain the characteristics of human ovarian cancer gene mutation and morphological features.