第一部分：

目的：耐药复发卵巢癌患者的预后差，单药化疗的有效率低，抗血管生成药物联合化疗能够改善铂耐药复发卵巢癌的预后。本研究试图探索口服的选择性VEGFR2抑制剂阿帕替尼在耐药复发卵巢癌中的作用，并通过动态监测患者治疗中血浆VEGF、ctDNA的变化，探索VEGF、ctDNA与疗效的相关性，判断其对预测疗效有无指导意义。

方法：自2018年3月22日至2021年1月28日，该研究共招募了152例有可评价病灶的、非黏液性的铂难治性或铂耐药复发上皮性卵巢癌、输卵管或腹膜癌患者。入组患者1：1随机接受多柔比星脂质体（PLD）单药或PLD联合阿帕替尼治疗，阿帕替尼给药剂量为250mg qd，PLD 给药剂量为40mg/m²，每四周一次， 共计6-8个周期，联合用药组在化疗结束后阿帕替尼维持，直至病情进展、发生不可耐受的毒性、或患者自行退出。治疗前完善CT检查，同时留存患者的石蜡标本及血标本。治疗中每2个周期按照RECIST标准进行评价疗效，并再次留取患者血液样本。针对中国医学科学院院肿瘤医院单中心入组的73例患者，采用ELISA方法检测其血浆中VEGF的浓度，NGS+靶向扩增深度测序检测其血浆中ctDNA fraction，并动态监测其变化。

结果：PLD+阿帕替尼联合用药组78例患者，PLD单药组74例患者。全组分析，中位随访时间为10.4个月。联合用药组和单药组的中位PFS分别为5.8、3.3个月（P＜0.001；HR=0.44，95%CI：0.28-0.69）；中位OS分别为21.8、14.5个月（P=0.038；HR=0.56，95%CI：0.33-0.97）。联合用药组的ORR和DCR均较单药组明显提高（ORR：43.1% vs 10.9%，P＜0.001；DCR：84.6% vs 57.8%，P＜0.001）。中国医学科学院肿瘤医院单中心结果显示：联合用药组和单药组的中位PFS为5.6、 2.8个月（P=0.005；HR=0.42，95%CI：0.23-0.78），ORR分别为40.0%、7.4%（P=0.004），DCR分别为65.7%、51.8%（P=0.270）。联合组较单药组的中位OS稍有延长，但无统计学差异（21.7 vs 14.5个月，P=0.631）。治疗前患者血浆VEGF的中位值为48.0pg/ml，在PLD联合阿帕替尼组，当血浆VEGF浓度≥93.7pg/ml时，治疗后疾病得到控制的敏感性为41.2%，特异性为76.9%（AUC=0.593，95%CI：0.385-0.801）。在PLD单药组，当治疗前血浆VEGF水平≥36.6pg/ml时，患者治疗中疾病进展的敏感性为76.9%，特异性为50.0%（AUC=0.519，95%CI：0.244-0.794）。对于接受联合治疗的患者，当化疗2程后VEGF浓度升高≥基线水平的30.8%时，对判断患者病情进展的敏感性为66.7%，特异性为56.2%（AUC=0.569， 95%CI:0.331-0.807）。但在单药化疗组，化疗2程后VEGF变化对预测是否会发生疾病进展无意义。81.8%（30/37）患者的血浆中检测到了ctDNA，中位ctDNA fraction为1.3%（IQR：0.07%-8.03%）。在联合用药组，治疗前ctDNA fraction高于中位值时，PFS明显延长（9.2 vs 3.7个月，P=0.028，HR：0.224，95% CI：0.059-0.848）。当治疗前ctDNA fraction≥1.1%时，患者治疗后疾病得到控制的敏感性为70.6%，特异性为83.3%，(AUC=0.716, 95%CI:0.455-0.976)。化疗2程后ctDNA fraction下降≤基线值的8.8%或较基线升高，患者可能发生疾病进展，敏感性为80%，特异性为64.3%(AUC=0.700, 95%CI:0.465-0.935)。

结论：与PLD单药相比，PLD联合阿帕替尼能提高耐药复发卵巢癌患者的ORR及DCR，也能够延长患者的PFS及OS。对于铂耐药复发卵巢癌患者，治疗前血浆VEGF浓度高的患者可能更能从抗血管联合化疗中获益。对于PLD联合阿帕替尼治疗的患者，治疗前ctDNA fraction、治疗中ctDNA fraction变化对预测联合用药的疗效有一定的价值。

第二部分：

目的：卵巢透明细胞癌是上皮性卵巢癌中第三常见组织类型，早期OCCC患者的预后存在争议，术后辅助化疗的最佳疗程目前仍不清楚。本研究试图探讨I期卵巢透明细胞癌的亚分期结局，以及增加辅助化疗疗程是否能改善生存。

方法：回顾性分析自1999年2月至2018年12月中国医学科学院肿瘤医院手术治疗的I期卵巢透明细胞癌患者102例。影响无瘤生存的预后因素采用COX回归模型进行评估，采用Kaplan-Meier绘制生存曲线，并用Log rank检验比较各组间的差异。P＜0.05具有统计学意义。

结果：中位随访时间为40.5个月。IA、IC1、IC2、IC3期患者例数分别为31（30.4%）、17（16.7%）、25（24.5%）、 17（16.7%）。5年无瘤生存率为82.8%，10年无瘤生存率为78.8%。5年总体生存率（OS）为97.9%。术中肿瘤破裂（IC1期）患者与IA期患者的无瘤生存时间（DFS）无明显差异（P=0.538,OR=0.024），但术前肿瘤破裂（IC2期）或腹水细胞学阳性(IC3期)较IA和IC1期患者的5年无瘤生存率明显降低（72.6% vs 95.1%，P=0.039，OR=5.051）。术后化疗4程与更多疗程相比，5年无瘤生存率无明显差异（83.9%vs81.7%）。单因素分析提示年龄、肿瘤大小及CA199水平与患者的无瘤生存时间相关，且有统计学意义，但多因素分析未发现影响患者无瘤生存时间的独立预后因素。

结论：I期卵巢透明细胞癌患者总的预后较好，但术前肿瘤破裂或腹水细胞学阳性的IC2和IC3期患者5年DFS率明显降低。术后辅助化疗＞4程也许并不改善患者的预后。CA199明显升高可能是提示I期OCCC预后差的一个潜在指标。

Part one:

Objective: The prognosis of patients with platinum-resistant recurrent ovarian cancer is poor, and the response rate of single drug chemotherapy is low. Anti-angiogenic therapy combined with chemotherapy could improve the prognosis of patients with platinum-resistant recurrent ovarian cancer. This study was conducted to evaluate the efficacy of apatinib an oral tyrosine kinase inhibitor that selectively inhibits VEGFR-2, in combination with PLD in patients with platinum-resistant recurrent OC, and to explore whether the levels of VEGF/ctDNA before treatment and changes of VEGF/ctDNA after treatment could be used as biomarkers of treatment response for patients with platinum-resistant recurrent ovarian cancer.

Method: From Mar 22, 2018 to January 28, 2021, a total of 152 patients with at least 1 assessable lesion and histologically confirmed non-mucinous ovarian, primary peritoneal cancer, or fallopian-tube cancer were recruited. Patients were randomly assigned (1:1) to receive PLD alone (group B, 40mg/m² IV every 4 weeks for up to 6 or 8 cycles) or with apatinib 250mg orally once daily (group A) until disease progression, unacceptable toxicity, or consent withdrawal. Before treatment, paraffin and blood samples should be kept, and baseline disease evaluation should be done by abdominopelvic and chest CT. During the treatment, tumor response was evaluated according to RECIST standard every two cycles, and blood samples were collected again. A total of 73 patients were enrolled in Cancer Hospital, Chinese Academy of Medical Sciences. The concentration of VEGF in plasma was detected by ELISA, and ctDNA fraction in plasma was detected by NGS combined with targeted amplification and deep sequencing.

Results: There were 78 patients in PLD combined with apatinib group and 74 patients in PLD monotherapy group. The median follow-up was 13.4 months. The median PFS was 5.8 months with aptinib-containing therapy versus 3.3 months with PLD alone (HR:0.44, 95% CI 0.28-0.69, P<0.001). And the median OS was 21.8 months, 14.5 months respectively(P=0.038). The ORR and DCR were significantly improved in the apatinib plus PLD group (ORR：43.1% vs 10.9%，P＜0.001；DCR：84.6% vs 57.8%) compared with in the PLD group. The results of single center suggested that the median PFS was also significantly improved in the aptinib-containing therapy group compared with in the PLD group(5.6 vs 2.8 months, P=0.005；HR=0.42，95%CI：0.23-0.78). The ORR was 40.0% (14/35) versus 7.4% (2/27) for combination and PLD monotherapy group, respectively (P=0.004), and DCR was 65.7% (23/35) and 51.8% (14/27), respectively (P=0.270). No significant difference was seen in OS between treatment groups, although the median OS of combination group was slightly longer than that of monotherapy group(21.7 vs 14.5 months, P=0.631). The median VEGF prior to treatment was 48.0pg/ml. In PLD combined with apatinib group, patients with more than 93.7pg/ml pre-treatment VEGF could have disease control with 41.2% sensitivity and 76.9% specificity(AUC 0.593, 95% CI 0.385-0.801). And patients in PLD monotherapy group with more than 36.6pg/ml pre-treatment VEGF could have disease progression with 76.9% sensitivity and 50.0% specificity(AUC=0.519，95%CI：0.244-0.794). A VEGF increase of 30.8% was associated with progression disease in the combination group with 66.7% sensitivity and 56.2% specificity(AUC=0.569, 95%CI:0.331-0.807). There was no correlation in PLD monotherapy group. The ctDNA was detected in 81.8%(30/37) patients, and the median ctDNA fraction was 1.3% (inter-quartile range IQR: 0.07% - 8.03%). In the combination group, PFS was significantly prolonged when ctDNA fraction was higher than the median value before treatment (P = 0.028, HR: 0.224, 95% CI: 0.059-0.848). patients with more than 1.1% pre-treatment ctDNA fraction could have disease control with 70.6% sensitivity and 83.3% specificity (AUC=0.716, 95%CI:0.455-0.976). And patients with ctDNA fraction decreased by less than 8.8% or increased compared with the baseline value, could have disease progression, with sensitivity of 80% and specificity of 64.3% (AUC=0.700, 95%CI:0.465-0.935) .

Conclusion: Compared with PLD alone, PLD combined with apatinib could prolong PFS，OS, and improve ORR, DCR in patients with platinum-resistant recurrent ovarian cancer. In platinum resistant patients with recurrent ovarian cancer, patients with high plasma VEGF concentration before treatment might benefit more from anti-angiogenic combination chemotherapy. Pre-treatment ctDNA fraction, ctDNA fraction changes during treatment might have any value in predicting the tumor response in patients treated with PLD and apatinib.

Part two:

Objective: Ovarian clear cell cancer (OCCC) is the third most common epithelial ovarian cancer. The prognostic relevance of the different substages in early stage OCCC remains controversial, and the optimal duration of adjuvant chemotherapy for early stage ovarian cancer is unclear. This research sought to explore the prognostic impact of adjuvant chemotherapy and tumor substage on stage I ovarian clear cell carcinoma (OCCC). 

Method: Data of 102 patients with stage I OCCC who underwent surgery at the National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College from February 1999 to December 2018 was retrospectively analyzed. Prognostic factors were evaluated using the Cox Regression Model. The disease-free survival (DFS) and overall survival (OS) were assessed by the Kaplan-Meier method and compared between different groups with the log-rank test. P < 0.05 was considered statistically significant.

Results: The median follow-up duration was 40.5 months. Thirty-one (30.4%) patients were at stage IA, and 17 (16.7%), 5 (24.5%) and 17 (16.7%) patients were at stage IC1, IC2 and IC3 respectively. The 5-year and 10-year DFS rates of the entire cohort were 82.8% and 78.8% respectively, and the 5-year OS was 97.9%. Patients at stages ICI (intraoperatively ruptured tumor) and IA had similar DFS (P=0.538, OR=0.024), and that of patients at stages IC2 (tumor ruptured preoperatively or tumor on ovarian surface) or IC3 (ascites or peritoneal washings with positive cytology) was significantly lower (72.6% vs 95.1%, P=0.039, OR=5.051). The 5-year DFS of patients receiving four (83.9%) and more than four (81.7%) cycles adjuvant chemotherapy were similar. Furthermore, univariate analysis showed that age, tumor size and CA199 levels were significantly correlated with DFS, although none of these variables were identified as independent prognostic factors in the multivariate analysis. 

Conclusion: Patients with stage I OCCC have overall good prognosis. However, tumor surface involvement or positive cytology can worsen prognosis, and intraoperative tumor rupture may not affect the prognosis of OCCC. The prognosis may not be improved by more than four cycles chemotherapy following surgery. The remarkable increased CA199 may be an potential indicator of poor prognosis in stage I OCCC.