| BACKGROUND AND AIMS:Venous thromboembolism (VTE), such as Deep-vein thrombosis (DVT) and Pulmonary thromboembolism (PE), often occurs after surgery for gynecological cancer and is one of the severe complications, which has extremely high lethal rate but usually be ignored for a long time. The high level of D-dimer was associated with occurrence of VTE. Herein we investigated the association of D-dimer level before operation with VTE. Other risk factors for VTE in gynecological cancer, particularly in the ovarian cancer, were also evaluated.METHODS:We performed a retrospective study of448patients with gynecologic cancers who underwent the detection of plasma D-dimer between July2007to December2010from Fudan University Shanghai Cancer Center, including141cases of ovarian cancer,252cases of cervical cancer and55cases of endometrial cancer. Clinical characteristics were collected regarding D-dimer level, age, body mass index (BMI), pathological type, FIGO stage, VTE prevention, diagnosis and treatment. For ovarian cancer patients, hematologic indexes (prothrombin time, PT; white blood cell number, WBC; platelet number, PLT) and operation information (operation time, intra-operative ascites and bleeding, intra-and post-operative transfusion) were also identified. In addition, medical records for40cases of ovarian benign tumor and10cases of ovarian borderline tumor were collected to compare with ovarian cancer.Follow-up study to analyze post-operative VTE for above patients (353cases available) was further conducted. Statistics analysis was performed by Graphpad Prism5.0and SPSS15.0software. ANOVA analysis and Newman-kueuls test were used to compare differences of D-dimer level before operation among disease groups. Meanwhile, Bivariate correlation analysis was selected to analyze the relation between D-dimer and other indexes. Finally, Logistic regression analysis was performed to evaluate possible risk factors for VTE.RESULTS:Among448patients with gynecological malignant tumor, VTE occurred in13patients (2.9%), which were3cases (0.8%) of PE in ovarian cancer and10cases (2.8%) of DVT, including8cases (5.7%,8/141) in ovarian cancer,3cases (1.2%,3/252) in cervical cancer and2cases (3.6%,2/55) in endometrial cancer, with2events (15.4%) occurred before operation,7events (53.8%) diagnosed within one week postoperatively and11events (84.6%) found within one month after surgery. In addition, the PE incidience in patients with ovarian cancer is2.1%(3/141).Significant differences of D-dimer level before operation among gynecological cancer were observed by ANOVA analysis (F=67.10, P＜0.0001). Further study by Newman-kueuls test showed that there were obvious differences between ovarian cancer and cervical cancer (q=15.67, P＜0.05), ovarian cancer and endometrial cancer (q=10.99, P<0.05). Significant differences of D-dimer level before operation among three types of ovarian tumor were obtained by ANOVA analysis (F=10.01, P<0.0001). Newman-kueuls test indicated that significant differences were observed between ovarian cancer and ovarian benign tumor (q=5.08, P＜0.05), ovarian cancer and ovarian borderline tumor (q=4.4, P＜0.05). Furthermore, Mann-whitney T test displayed that D-dimer level before operation was different between stage Ⅰ/Ⅱ and stage III/IV patient in ovarian cancer (P=0.003).Bivariate correlation analysis suggested that D-dimer level before operation was largely correlated with platelet (Pearson r=0.4969, P＜0.001) but not WBC. D-dimer level before operation were still associated with PT (Pearson r=0.1738, P=0.0462). However, there was no correlation of D-dimer level before operation with any of these three hematological parameters in ovarian benign tumor or ovarian borderline tumor. The association of D-dimer level before operation with BMI and age were not significant in ovarian cancer, cervical cancer or endometrial cancer.In the Logistic regression analysis of448person diagnosed with gynecological cancer, the level of D-dimer higher than310μg/L in184patients (41.4%) were grouped as high risk factor (OR=3.3,95%confidence interval [CI],1.05-10.17, P=0.047),with5.3%(8/151) incidence of VTE. Furthermore, the positive predictive value was5.3%as well as the negative predictive value was98.3%when the level of D-dimer was310μg/L. Compared with ovarian cancer, cervical cancer was lower risk type of disease (OR=0.2,95%CI,0.05-0.77,P=0.019).In the Logistic regression analysis of141person diagnosised with ovarian cancer, the level of D-dimer higher than780μg/L in38(27%) patients was considered as significant risk factor (OR=5.05,95%CI,1.15-22.29, P=0.032) as the incidence of VTE was13.2%(5/38), meanwhile ascites more than1500ml (OR=5.95,95%CI, 1.34-26.45,P=0.019) in33(24.3%) patients, which had15.2%(5/33) incidence of VTE, was also considered as high risk factor。In addition, as the level of D-dimer was780μg/L, the positive predictive value was14.8%while the negative predictive value was97.8%.CONCLUSIONS:There was a close relationship between D-dimer and platelet in patients with ovarian cancer. Logistic regression analysis suggested that the incidence of VTE in ovarian cancer was highest and the level of D-dimer before operation higher than310μg/L could be considered as a predictive risk factor for patients with gynecological cancer. Moreover, Ascites more than1500ml and the level of D-dimer before operation higher than780μg/L were two significant risk factors of VTE in patients with ovarian cancer. |
PDF Full Text Request