Establishment And Characteristics Of Patient-derived Xenograft Models In Human Ovarian Cancer

Background Ovarian carcinoma is characterized as hardly discoverable, high metastasis, poor prognosis.It is one of the most common malignancy among female population with the highest mortality rate comparing to other gynecological malignancies.Because there are no pathological signs at the early stage of ovarian carcinoma,patients only went to hospital when the pain from waist and abdominal occurs or the digestive tract symptom was detectable.With those reasons,2/3 of the patients are diagnosed at an the advanced stage of the illness,and more than of 70% of them has tumor transferred or metastasis of pelvic lymphnode. Presently,according to the NCCN guidelines,the established standard therapy of advanced ovarian carcinoma is debulking surgery with platinum-containing chemotherapy followed.The Paclitaxel and platinum-containing chemotherapy is still the main scheme for dealing with the advanced ovarian carcinoma.Although platinum-based chemotherapy has made new progress and skill of the surgeon is progressing too,the 5-year survival rate did not significantly increase.The main cause is drug resistance to lead to tumor recurrence.In general,we use cell lines and cell-lines xenografts to develop OC,but these cell lines have been cultured in vitro for several years,their genomic profiles and biological characteristics have been irreversibly altered and rarely recapitulateheterogeneity of the parental tumor and the host microenvironment.To improve the survival rate of ovarian carcinoma, the seeking for an accurate preclinical models seems essential,which preserved histopathological and genotypic characteristics of original patient samples, and therapeutic response is consistent with original human tumor patients.In our experiment,we will establish patient-derived xenograft(PDX) models in ovarian cancer,then,correlations between histopathological and genotypic characteristics of the original patient samples and PDX models will be analyzed.Our ultimate goal is develop an accurate preclinical model for precision medicine of ovarian cancer. Objective 1. Establishing patient-derived xenograft(PDX) models in ovarian cancer; 2. Correlations between histopathological and molecular characteristics of the original patient samples and PDX models will be analyzed. Methods Our experiment is approved by Ethics Committee. 1. Forty two OC patient specimens were obtained at initial surgery from Xijing hospital,confirmed by intraoperative frozen and postoperative Pathological diagnosis, surgically removed tissues are implanted subcutaneously into SCID mice,named P1. 2. Xenografted OC tumors(~400mm3) in SCID mice were further implanted subcutaneously into nude mice for further generations of these models,named P2,then consecutive mouse-to-mouse passages,named P3,P4,P5,respectively. 3. Recording clinical data,including age,clinical stages,histologic subtype, operation timing,ascites and the time of tumor relapse,then statistically analyzing the relationships between model success rate and clinical data. 4. The frozen tumors were resuscitated into SCID mice. 5. Histopathological and phenotypic characteristics of the original patient samples and PDX models are compared. 6. The original patient tissues and xenograft tissues,genomic DNA was extracted from each tissue sample,P53 exons 5-9 were amplified by polymerase chain reaction(PCR),then sequenced.Results 1. The model success rate was 35.7%(15/42) in the first generation of SCID mice,and 100% in subsequent generations.Statistical analysis show that tumor success rate is correlative with malignant ascites(P=0.047)and the time of tumor relapse(P=0.001),but no correlation with patients' age(P=0.137),histological subgroup(P=0.461),grade(P=0.222),or operation timing(P=0.694). 2. The frozen tumors were resuscitated into SCID mice,tumor success rate was 100%. 3. Histopathological and phenotypic characteristics of the original patient samples and PDX models are similar. 4. DNA sequence between original patient samples and PDX models is also consist ent. Conclusions 1. The model success rate was 35.7%(15/42) in the first generation of SCID mice,and 100% in subsequent generations,and there is a correlation between model success rate and malignant ascites,the result also show that model success rate may be as a prognostic factor. 2. Histopathological and molecular characteristics of the original patient samples and PDX models are similar,our PDX model can be as an accurate preclinical model to precision medicine of ovarian cancer.